THE JOURNAL OF THE ALZHEIMER’S ASSOCIATION
Volume 3, Issue 4
CONTENTS
Editorial
261 Zenith Award Program of the Alzheimer’s
Association
Zaven S. Khachaturian
262 Introduction: The Third Canadian Consensus
Conference on the Diagnosis and Treatment
of Dementia, 2006
Howard Chertkow
Review Articles
266 Mild cognitive impairment and cognitive
impairment, no dementia: Part A, concept
and diagnosis
Howard Chertkow, Ziad Nasreddine,
Yves Joanette, Valérie Drolet, John Kirk,
Fadi Massoud, Sylvie Belleville, Howard Bergman
283 Mild cognitive impairment and cognitive
impairment, no dementia: Part B, therapy
Fadi Massoud, Sylvie Belleville, John Kirk,
Howard Chertkow, Ziad Nasreddine,
Yves Joanette, Howard Bergman,
Morris Freedman
292 Clinical diagnosis of dementia
Alain Robillard
299 Neuropsychological testing and assessment
for dementia
Claudia Jacova, Andrew Kertesz, Mervin Blair,
John Fisk, Howard H. Feldman
318 Cobalamin and homocysteine in older adults:
Do we need to test for serum levels in the
work-up of dementia?
Angeles Garcia
325 The role of biologic markers in the diagnosis
of Alzheimer’s disease
Hyman M. Schipper
October 2007
333 Structural neuroimaging in the diagnosis of
dementia
Tiffany Chow
336 Functional neuroimaging in the diagnosis of
dementia
Michael Borrie
341 General risk factors for dementia:
A systematic evidence review
Christopher Patterson, John Feightner,
Angeles Garcia, Chris MacKnight
348 Primary prevention of dementia
Christopher Patterson, John Feightner,
Angeles Garcia, Chris MacKnight
355 Management of mild to moderate
Alzheimer’s disease and dementia
David B. Hogan, Peter Bailey, Anne Carswell,
Barry Clarke, Carole Cohen, Dorothy Forbes,
Malcolm Man-Son-Hing, Krista Lanctôt,
Debra Morgan, Lilian Thorpe
385 Clinical practice guidelines for severe
Alzheimer’s disease
Nathan Herrmann, Serge Gauthier, Paul G. Lysy
398 Management of dementia with a
cerebrovascular component
Christian Bocti, Sandra Black, Chris Frank
404 Disclosure of the diagnosis of dementia
John D. Fisk, B. Lynn Beattie, Martha Donnelly,
Anna Byszewski, Frank J. Molnar
411 Ethical considerations for decision making
for treatment and research participation
John D. Fisk, B. Lynn Beattie, Martha Donnelly
418 Genetics and dementia: Risk factors,
diagnosis, and management
Ging-Yuek Robin Hsiung, A. Dessa Sadovnick
Contents continued on next page
Contents continued
428 Toward a revision of criteria for the
dementias
Kenneth Rockwood, Rémi W Bouchard,
Richard Camicioli, Gabriel Léger
Perspectives
441 Commentary on “The Third Canadian
Consensus Conference on the Diagnosis and
Treatment of Dementia:” Clinical guidelines
are not enough—System-wide, population-
based programs are needed to improve the
care of patients with Alzheimer’s disease and
related dementias
Howard Fillit
444 Commentary on “The Third Canadian
Consensus Conference on the Diagnosis and
Treatment of Dementia”
Bruno Vellas, Emma Reynish, Philippe Robert
446 Commentary on “The Third Canadian
Consensus Conference on the Diagnosis
and Treatment of Dementia”—An appraisal
Paul S. Aisen
Association Pages
449 Alzheimer’s Association Update

Cover image courtesy of Drs. William Jagust and Ansgar Furst, University of California, Berkeley and Lawrence
Berkeley National Laboratory.
The top pair of images is a patient with frontotemporal lobar degeneration (FTLD), the center pair is a patient with
Alzheimer’s disease (AD), and the bottom pair is a normal older adult. Images on the left column are FDG-PET
superimposed on the MRI. Images on the right column are PET with Pittsburgh compound B (PIB) superimposed on the
MRI. Left column images show frontal hypometabolism in FTLD patient, parietal hypometabolism in AD patient, and
normal metabolism in the normal older adult. Right column images show PIB binding only in the AD patient and not
the FTLD patient or normal adult.
 THE JOURNAL OF THE ALZHEIMER’S ASSOCIATION
Alzheimer’s & Dementia (ISSN 1552-5260) is published quarterly by
Elsevier, 360 Park Avenue South, New York, NY 10010-1710. Months
of issue are January, April, July, and October. Business Offices: 1600 John
F. Kennedy Boulevard, Suite 1800, Philadelphia, PA 19103-2899. Edi-
torial Office: 11830 Westline Industrial Dr, St Louis, MO 63146-3318.
Accounting and Circulation Offices: 6277 Sea Harbor Drive, Orlando,
FL 32887-4800.
CUSTOMER SERVICE (orders, claims, online, change of address):
Elsevier Periodicals Customer Service, 6277 Sea Harbor Drive, Orlando,
FL 32887-4800. Tel: (800) 654-2452 (U.S. and Canada); (407) 345-4000
(outside U.S. and Canada). Fax: (800) 225-4030 (U.S. and Canada); (407)
363-9661 (outside U.S. and Canada). E-mail: elspcs@elsevier.com Ad-
dress changes must be submitted four weeks in advance.
YEARLY SUBSCRIPTION RATES: United States and possessions:
Individual $105.00; Institution $229.00; Student/Resident: $70.00. All
other countries (prices include airspeed delivery): Individual $105.00;
Institution $229.00; Student/Resident $70.00. Single Issues $58.00. To
receive student/resident rate, orders must be accompanied by name of
affiliated institution, date of term and the signature of program/residency
coordinator on institution letterhead. Orders will be billed at the indi-
vidual rate until proof of status is received. Current prices are in effect for
back volumes and back issues.
Further information on this journal is available from the Publisher or
from this journal’s website (http://www.alzheimersanddementia.org).
Information on other Elsevier products is available through Elsevier’s
website (http://www.elsevier.com).
Advertising information. Advertising orders and inquiries can be sent
to: USA, Canada, and South America, Hank Blaney, Elsevier, 360
Park Ave South, New York, NY 10010: phone (212) 633-3648; fax:
(212) 633-3820; e-mail, h.blaney@elsevier.com. Classified advertising
orders and inquiries can be sent to Simone Imbert, Elsevier, 360 Park Ave
South, New York, NY 10010: phone (212) 462-1908, fax (212) 633-
3820, e-mail, s.imbert@elsevier.com. Europe and the rest of the
world, Julie Toop; phone ⫹44 (0) 1865 843016; fax ⫹44 (0) 1865
843976; e-mail media@elsevier.com.
Author inquiries. For inquiries relating to the submission of articles
(including electronic submission where available), please visit www.
elsevier.com/authors. This site also provides the facility to track ac-
cepted articles and set up e-mail alerts to inform you of when an article’s
status has changed, as well as detailed artwork guidelines, copyright
information, frequently asked questions, and more. Please see Informa-
tion for Authors for individual journal requirements. Contact details for
questions arising after acceptance of an article, especially those relating to
proofs, are provided after registration of an article for publication. Re-
prints. For queries about author offprints, e-mail authorsupport@
elsevier.com. To order 100 or more reprints for educational, commercial,
or promotional use, contact the Commercial Reprints Department,
Elsevier Inc., 360 Park Avenue South, New York, NY 10010-1710. Fax:
(212) 462-1935; e-mail reprints@elsevier.com. Reprints of single articles
available online may be obtained by purchasing Pay-Per-View access for
$32 per article on the journal website, www.alzheimersanddementia.org.
© 2007 The Alzheimer’s Association. All rights reserved. This
journal and the individual contributions contained in it are protected
under copyright by The Alzheimer’s Association and the following terms
and conditions apply to their use:
Photocopying. Single photocopies of single articles may be made for
personal use as allowed by national copyright laws. Permission of the
Publisher and payment of a fee is required for all other photocopying,
including multiple or systematic copying, copying for advertising or
promotional purposes, resale, and all forms of document delivery. Special
rates are available for educational institutions that wish to make photo-
copies for non-profit educational classroom use.
Permission may be sought directly from Elsevier’s Global Rights
Department in Oxford, UK: phone 215-239-3804 or ⫹44 (0)1865
843830, fax ⫹44 (0)1865 853333, e-mail healthpermissions@elsevier.
com. Requests may also be completed online via the Elsevier homepage
(http://www.elsevier.com/permissions).
In the USA, users may clear permissions and make payments through
the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, USA; phone: (978) 750-8400, fax: (978) 750-4744, and in
the UK through the Copyright Licensing Agency Rapid Clearance
Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP,
UK; phone: (⫹44) 20 7631 5555; fax: (⫹44) 20 7631 5500. Other
countries may have a local reprographic rights agency for payments.
Derivative Works. Subscribers may reproduce tables of contents or
prepare lists of articles including abstracts for internal circulation within
their institutions. Permission of the Publisher is required for resale or
distribution outside the institution.
Permission of the Publisher is required for all other derivative works,
including compilations and translations.
Electronic Storage or Usage. Permission of the Publisher is required
to store or use electronically any material contained in this journal,
including any article or part of an article.
Except as outlined above, no part of this publication may be repro-
duced, stored in a retrieval system or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission of the Publisher.
Address permissions requests to: Elsevier Rights Department, at the
fax and e-mail addresses noted above.
Notice. No responsibility is assumed by the Publisher for any injury
and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods,
products, instructions or ideas contained in the material herein. Because
of rapid advances in the medical sciences, in particular, independent
verification of diagnoses and drug dosages should be made.
Although all advertising material is expected to conform to ethical
(medical) standards, inclusion in this publication does not constitute a
guarantee or endorsement of the quality or value of such product or of
the claims made of it by its manufacturer.
 THE JOURNAL OF THE ALZHEIMER’S ASSOCIATION

Volume 3, Issue 4 October 2007

EDITORIAL BOARD
Editor-in-Chief Editorial Advisory Board
Zaven S. Khachaturian Philippe Amouyel, Lille, France
451 Hungerford Drive Carol Barnes, Tucson, AZ
Suite 119-355 Anders Björklund, Lund, Sweden
Rockville, MD 20850 Francois Boller, Paris, France
Tel. 301-294-7201; Fax 301-294-7203 Heiko Braak, Frankfurt, Germany
E-Mail: ADJ@kra.net Jason Brandt, Baltimore, MD
John C. S. Breitner, Seattle, WA
Editors-at-Large Henry Brodaty, Sydney, Australia
Floyd Bloom, San Diego, CA Louis Burgio, Chapel Hill, NC
Paul Greengard, New York, NY Carl Cotman, Irvine, CA
Peggye Dilworth-Anderson, Chapel Hill, NC
Associate Editors Fred Gage, La Jolla, CA
Paul Coleman, Rochester, NY Alison Goate, St Louis, MO
Francoise Forette, Paris, France Michel Goedert, Cambridge, United Kingdom
Virginia Lee, Philadelphia, PA Jeff Kaye, Portland, OR
Marsel Mesulam, Chicago, IL David Knopman, Rochester, MN
John Trojanowski, Philadelphia, PA Jae-Young Koh, Seoul, South Korea
Michael Weiner, San Francisco, CA Lewis Kuller, Pittsburgh, PA
Eckhard Mandelkow, Hamburg, Germany
Eva Mandelkow, Hamburg, Germany
Richard Mayeux, New York, NY
Steven M. Paul, Indianapolis, IN
Ron Petersen, Rochester, MN
Steve Post, Cleveland, OH
Teresa Radebaugh, El Dorado, KS
Joseph Rogers, Phoenix, AZ
Allen Roses, Durham, NC
Barbara Sahakian, Cambridge, United Kingdom
Mary Sano, New York, NY
Lon S. Schneider, Los Angeles, CA
Dennis Selkoe, Boston, MA
Sangram Sisodia, Chicago, IL
Philip Sloane, Chapel Hill, NC
Publication Staff Christine Van Broeckhoven, Antwerp, Belgium
Managing Editor Sandra Weintraub, Chicago, IL
Ara S. Khachaturian Kathleen Welsh-Bohmer, Durham, NC
Tel. 301-309-6730; Fax: 301-309-6724 Bengt Winblad, Stockholm, Sweden
E-mail: ADJ_MgrEd@kra.net

Publisher
Elsevier Inc.
11830 Westline Industrial Dr, St Louis, MO 63146-3318
 THE JOURNAL OF THE ALZHEIMER’S ASSOCIATION

Volume 3, Issue 4 October 2007

Alzheimer’s Association
MEDICAL AND SCIENTIFIC ADVISORY COUNCIL
Chair William E. Klunk, MD, PhD Chairs Emeriti
Samuel E. Gandy, MD, PhD University of Pittsburgh School of Marilyn S. Albert, PhD
Thomas Jefferson University Medicine Leonard Berg, MD
Farber Institute for Neurosciences Department of Psychiatry Steven DeKosky, MD
Philadelphia, PA Pittsburgh, PA David Drachman, MD
Robert Katzman, MD
Vice Chair John C. Morris, MD William Markesbery, MD
Ronald C. Petersen, MD, PhD Washington University School of Allen Roses, MD
Mayo Clinic Medicine
Department of Neurology Department of Neurology Alzheimer’s Association
Rochester, MN St. Louis, MO William H. Thies, PhD
Vice President, Medical and
Members Ralph A. Nixon, MD, PhD Scientific Relations
Joseph D. Buxbaum, PhD New York University School of Chicago, IL
Mount Sinai School of Medicine Medicine
Departments of Psychiatry, Departments of Psychiatry and Cell Maria Carrillo, PhD
Neuroscience, Geriatrics and Adult Biology Director, Medical and Scientific
Development New York, NY Relations
New York, NY Chicago, IL
James W. Simpkins, PhD
Peggye Dilworth-Anderson, PhD University of North Texas Health Sam Fazio, PhD
University of North Carolina, Science Center Director, Medical and Scientific
Chapel Hill Department of Pharmacology and Relations
School of Public Health Neuroscience Chicago, IL
Center for Aging and Diversity Fort Worth, TX
UNC Institute on Aging Paula Moore
Chapel Hill, NC Yaakov Stern, PhD Director, Medical and Scientific
Columbia University Relations
Steven H. Ferris, PhD New York, NY Chicago, IL
New York University School of
Medicine Linda Teri, PhD Consultants
Silberstein Institute for Aging and University of Washington
Zaven Khachaturian, PhD
Dementia Seattle, WA
Senior Science Adviser
New York, NY
Potomac, MD
Linda J. Van Eldik, PhD
Claudia H. Kawas, MD Northwestern University Feinberg
University of California, Irvine School of Medicine
Department of Neurology Department of Cell and Molecular
Department of Neurobiology and Biology
Behavior Chicago, IL
Irvine, CA
 Alzheimer’s & Dementia 3 (2007) 261
Editorial
Zenith Award Program of the Alzheimer’s Association
The difficulty of finding funds to sustain the expansion of
Alzheimer’s research at a reasonable rate above inflation is
not a new problem for the field. Thirty years ago while
developing the nascent fields of brain aging and dementia
research at the National Institute on Aging (NIA) at the
National Institutes of Health (NIH), the two perennial chal-
lenges I faced were (1) how to obtain set-aside funds from
U. S. Congress (which are specifically targeted appropria-
tions for Alzheimer’s research) and (2) how to attract new
talent with fresh ideas to the field of Alzheimer’s research.
Today, dementia research is confronted with the same
crises as three decades ago but for different reasons. The
current dilemma in funding stems from the extraordinary
discrepancy between the available supply of research dol-
lars and the ever-increasing demand for these funds.
During the early years of developing a national program
of Alzheimer’s research, the only successful strategy in
achieving the dual objectives of increasing the appropria-
tions for Alzheimer’s research and recruiting new investi-
gators to the field was to create new funding mechanisms.
The development of special programs (or funding mecha-
nisms) became the most effective tool for increasing NIA’s
budget by providing Congress specific targets for appropri-
ations, as well as the means for advancing the Institute’s
programmatic goals. For example, programs such as the
Alzheimer’s Disease Research Centers (ADRCs), the Con-
sortium to Establish Registries for Alzheimer’s Disease
(CERAD), Alzheimer’s Disease Clinical Studies (ADCS),
Leadership and Excellence in Alzheimer’s Disease (LEAD),
Alzheimer’s Disease Education and Referral (ADEAR)
center, and the Cross-Cultural and Epidemiological Studies
Initiative were primarily responsible for the dramatic
growth of federal funds allocated to Alzheimer’s research.
As an illustration, the LEAD program, at $1 million per
award, increased NIA’s budget by $10 million annually and
increased the visibility and prestige of Alzheimer’s research
nationally as a result of the high-caliber recipients of these
awards. At the same time, this program became a potent
way of recruiting highly talented young investigators to the
field under the tutelage of senior scientists.
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.08.003
As the major private funding entity in the U. S., the
Alzheimer’s Association has historically faced the same
challenges as the NIA. Like the NIA, the Alzheimer’s As-
sociation seeks to develop programs that are attractive to po-
tential donors while addressing the needs of the medical and
scientific programs of the Association. The Zenith Award was
created by the Alzheimer’s Association in 1991 as a unique
program intended not only to attract major gifts and donors but
also to provide special recognition and support to established
investigators with a track record of creativity. A notice of this
award appears in this issue of Alzheimer’s & Dementia, and the
details may be found at http://www.alz.org/professionals_
and_researchers_11284.asp.
During the 16-year history of the Zenith Award, the list
of recipients of this special grant represents the “who’s
who” of dementia research. During the same period, the
Zenith Award has played a crucial role in assembling an
impressive cadre of highly dedicated donors that pledge to
become Zenith Fellows by donating $1 million each to
support the Association’s scientific programs. The unusual
aspect of the Zenith Award is that the Fellows as a group
make the final selection of recipients—selected from the entire
array of applicants being evaluated for scientific merit— by
the normal peer-review process of the Association.
The prospects for improving funding for Alzheimer’s
research through the NIH in the foreseeable future are dim.
Therefore, more than ever before, the burden of funding
high risk– high payoff projects falls on the Alzheimer’s
Association. By expanding the Zenith program signifi-
cantly, the Association can create novel funding mecha-
nisms that will attract major gifts to the Association, attract
much needed talent to the field, and cultivate creative new
ideas for research. In this way the Association will lend
invaluable speed to our quest for a cure.
Zaven S. Khachaturian
Editor-in-Chief
E-mail address: adj@kra.net
 Alzheimer’s & Dementia 3 (2007) 262–265

Introduction: The Third Canadian Consensus Conference on the

Diagnosis and Treatment of Dementia, 2006*
Howard Chertkowa,b,c,d,**
a
Bloomfield Centre for Research in Aging, Lady Davis Institute for Medical Research, Sir Mortimer B. Davis -Jewish General Hospital (Department of
Clinical Neuroscience), McGill University, Montreal, Quebec, Canada
b
Division of Geriatric Medicine, Department of Medicine, Sir Mortimer B. Davis -Jewish General Hospital, McGill University, Montreal,
Quebec, Canada
c
Department of Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada
d
Centre de recherche, Institut Universitaire de Gériatrie de Montréal, Montreal, Quebec, Canada

In this issue of Alzheimers & Dementia, something
unique will be presented—a set of wide-ranging recommen-
dations developed by the entire academic community of
Canada involved in clinical research and care of Alzhei-
mer’s disease and dementia patients. Despite constant re-
search progress on all fronts and despite ongoing contro-
versies, these articles prove that there is a large body of
clinical practice regarding dementia that is commonly ac-
cepted and agreed on by the authorities in the field, whether
they are from Neurology, Geriatric Medicine, Geriatric Psy-
chiatry, Neuropsychology, or Family Practice. Because of
the manageable size and cohesion of our clinical research
and academic community, it was possible in Canada to have
all of our experts exchange viewpoints online and meet
together in one room. During the course of 1 year, we were
able to hammer out an evidence-based set of recommenda-
*EDITOR’S NOTE: This series of papers from the Third Canadian
Consensus Conference on the Diagnosis and Treatment of Dementia will
serve not only to educate family physicians, but also to further standardize
guidelines and broaden the research infrastructure. The utility of the exer-
cise initiated by the Canadian consortium is necessitated by prospective
plans in several parts of the globe to conduct comprehensive population-
based initiatives to identify individuals at risk for dementia and other
memory disorders. These new super-surveys envision enrolling cohorts
with many thousands of volunteers at a time when funding resources are
diminishing. An important task facing these new initiatives will be an
increased reliance on general medical practitioners to augment and econ-
omize case detection and risk factor identification in large populations.
Although this is a timely and important idea for the field, the practical
applications of the concept will clearly require further deliberations and
validation by other groups. We encourage our readers and other interested
individuals to submit commentaries or share their perspective on any of the
recommendations proposed by the Canadian consortium.
**Corresponding author. Tel.: 514-340-8260; Fax: 514-340-8295.
E-mail address: howard.chertkow@mcgill.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.012
tions, gaining approval by at least 80% of the group, which
we intend to become the core of dementia care and practice
in our country (and perhaps beyond our borders as well!).
The Alzheimer’s disease academic community in Can-
ada has a tradition of developing and disseminating national
standards for physicians. In 1989 Dr Mark Clarfield, a
geriatrician at McGill University, chaired a Canadian Con-
sensus Conference on Dementia (CCCD) that led to a series
of recommendations that were eventually published in the
Canadian Medical Association Journal (CMAJ) [1]. In
1998 a second Canadian Consensus Conference on Demen-
tia (Second CCCD) was held under the joint direction of Drs
Christopher Patterson and Serge Gauthier to deal with re-
cent developments such as the introduction of cholinester-
ase inhibitors. The consensus statements agreed to at the
conference were published in the CMAJ [2,3], with the
background articles supporting the statements appearing
shortly afterwards in the Canadian Journal of Neurological
Sciences (Can J Neurol Sci 2001;28[Suppl 1]:S1–121). They
have had a significant impact on the care of patients with
dementia in Canada, particularly in terms of giving guid-
ance to practicing physicians and direction to continuing
medical education (CME)/continuing professional develop-
ment (CPD) activities. These publications have been widely
disseminated to Canadian physicians and other stakehold-
ers. After the Second CCCD, the Consortium of Canadian
Centres for Clinical Cognitive Research (C5R) was charged
with the responsibility of evaluating the impact of the Sec-
ond CCCD and organizing the next consensus conference at
an opportune time. The timing of the next meeting would be
based on the pace and significance of advances in the field.
Have there been sufficient advances to warrant a rethink-
ing of diagnostic and therapeutic recommendations directed
 H. Chertkow / Alzheimer’s & Dementia 3 (2007) 262–265
at practicing physicians? With my colleagues, I answered in
the affirmative. On the diagnostic side, for example, there
have been substantial developments in neuroimaging. The
potential role of homocysteine as a risk factor for Alz-
heimer’s disease (AD) has become recognized. Should
the recommended modest battery of laboratory investi-
gations for patients with seemingly typical AD be altered
because of these and other findings? The recognition that
vascular risk factors like hypertension and hyperlipid-
emia increase the likelihood of AD as well as vascular
dementia offers enhanced opportunities for prevention of
dementia. New diagnoses such as mild cognitive impair-
ment (MCI) have become widely used. Mixed dementias
seem to be growing in importance with technical ad-
vances in imaging.
On the therapeutic side, a better understanding of the
needs of caregivers and how to support them has arisen.
Novel nonpharmacologic management modalities have
been developed and studies, notwithstanding the inherent
difficulties of research in this area. Guidelines on dementia
and driving have been introduced. The cholinesterase inhib-
itors have become widely prescribed, although their utility
continues to be debated. Is there still consensus that they
should be offered routinely to patients with mild to moder-
ate AD? Do they have a role for those at a severe stage?
Memantine has been introduced, and physicians are uncer-
tain about its place in therapy. Although routine use of
vitamin E was not recommended, this micronutrient did
receive dissenting support at the Second CCCD. It has come
under attack as a therapeutic intervention.
In 1998 it was believed that “the reduced risk [for AD]
associated with long-term estrogen use in epidemiological
studies may provide an additional potential benefit to con-
sider when weighing the pros and cons of estrogen therapy.”
The results of the Women’s Health Initiative Memory Study
[4,5] seemingly dashed this hope. There are new data
concerning a number of other therapies designed to pre-
vent, delay the onset, and/or slow the progression of a
dementia.
There emerged a general agreement last year that another
meeting had clearly become necessary to review the field
and determine the current consensus among physicians and
other health care providers who deal with patients with a
dementia. As then president of C5R, it fell to me to provide
leadership to the process that culminated in the Third
Canadian Consensus Conference on the Diagnosis and Treat-
ment of Dementia (CCCDTD3), which was held in March
2006 in Montreal. The organization, planning, and structure of
the conference adhered as far as possible to the guidelines
established by the AGREE collaboration (The AGREE Col-
laboration: Appraisal of guidelines for research and evaluation
[AGREE] instrument; www.agreecollaboration.org).
A Steering Committee of nationally eminent experts in
different aspects of dementia was formed. Members (listed
alphabetically) were the following:
263
Dr Howard Bergman (Joseph Kaufmann Professor of
Geriatric Medicine at McGill University; President
of Canadian Geriatric Society; and director of the
Quebec Research Network on Aging)
Dr Howard Chertkow (Professor of Neurology and
Geriatric Medicine at McGill University and
past president of C5R; Chair of the Steering
Committee)
Dr John W. Feightner (Family physician; Professor of
Family Practice, University of Western Ontario)
Dr Howard Feldman (Professor and Head, Division of
Neurology, University of British Columbia; acted
as liaison to the Canadian Neurological Society)
Dr Serge G. Gauthier (Professor of Neurology, Med-
icine, and Psychiatry at McGill University; chief of
the Alzheimer Disease Research Unit at the McGill
Centre for Studies in Aging)
Dr Nathan Herrmann (Geriatric psychiatrist; Profes-
sor and Director of Geriatric Psychiatry, Sunny-
brook Hospital, University of Toronto; acted as
liaison with the Canadian Academy of Geriatric
Psychiatry)
Dr David Hogan (Brenda Stafford Foundation Chair
and Professor in Geriatric Medicine, University of
Calgary)
Dr Yves Joanette (Neuropsychology and Speech lan-
guage Pathologist; Directeur, Centre de recherche
de l’Institut universitaire de gériatrie de Montréal,
Université de Montréal)
Dr Christopher J. Patterson (Professor of Geriatric
Medicine; McMaster University).
Financial and/or in-kind support was obtained from the
Canadian Institutes of Health Research, Fonds de la Recher-
che en Santé du Québec, Alzheimer Society of Canada,
C5R, Canadian Neurological Society, Canadian College of
Family Physicians, Canadian Academy of Geriatric Psychi-
atrists, Quebec Research Network on Aging, and Canadian
Geriatric Society. Pharmaceutical companies could make
unrestricted contributions of $10,000 each and attend the
consensus meeting as observers only. The total contribution
of pharmaceutical sources to the conference was less than
50% of the cost of putting on the consensus conference. All
grants were received in the form of unrestricted grants, and
no organization or company had any input into the form or
content of the recommendations.
The Steering Committee met frequently by teleconfer-
ence to establish the objectives and methods and plan for the
consensus meetings and to set up working groups. Partici-
pants on the working groups were chosen on the basis
of known expertise in dementia or a related area, reputation
for delivering high quality work quickly, recognition as a
national opinion leader, and/or being acknowledged as both
open-minded and collaborative. Participants agreed to abide
by the rules established by the Steering Committee for
264 H. Chertkow / Alzheimer’s & Dementia 3 (2007) 262–265
carrying out their task. Conflicts of interest were declared
beforehand, with written declarations provided by all
participants.
The topic areas were as follows:
Topic 1: Concept, utility, and management of MCI
and cognitive impairment no dementia (CIND)
(pharmacologic and nonpharmacologic manage-
ment) (H. Chertkow, coordinator)
Topic 2: Criteria for “dementia”, AD, “mixed dementia
states” and others (Kenneth Rockwood, coordinator)
Topic 3: Dementia diagnosis and differential diagnosis
of dementia for the primary care practitioner and
consultant: Clinical, laboratory, imaging, markers
(Howard Feldman, coordinator)
Topic 4: Assessment and management of risk factors
(genetic and environmental) and primary preven-
tion strategies (Christopher Patterson, coordinator)
Topic 5: Management of mild to moderate AD (David
Hogan, coordinator)
Topic 6: Management of severe AD (Serge Gauthier
and Nathan Hermann, coordinators)
Topic 7: Management of dementia with a cerebrovascu-
lar component (Christian Bocti and Sandra Black,
coordinators)
Topic 8: Ethical issues (eg, the use of placebos in re-
search, end of life decisions, terminal care, giving the
diagnosis) (John Fisk, coordinator)
Topic 9: Genetic aspects of dementia (Robin Hsiung
and Dessa Sadovnick, coordinators)
After carrying out appropriate literature search(es),
each working group prepared background article(s) and
discussed/agreed to a series of draft recommendations in
their assigned topic area. The article(s) and the recommen-
dations were then posted on an on-line website. Feedback
on them was solicited from all participants. This process
extended for several months.
Literature searches were carried out in both PubMed and
Embase going back to 1996, the end date of the literature
review for the Second CCCD. These searches were supple-
mented by articles selected from the authors’ files. A pre-
liminary list of potential articles was obtained through
these searches. The list was reduced by two individuals
on each subcommittee to relevant articles on the basis of
methodologic quality, presentation of original data, and
non-overlap with other reports. For therapy, a standard-
ized analysis of the quality of the evidence was under-
taken, integrating a rating of the quality of the report by
using the Jadad scale [6] and such quantitative aspects as
statistical significance and magnitude of effects (www.
naturalstandards.com).
The quality of the available evidence was graded (I, II, or
III) by using the following system adapted from the Cana-
dian Task Force on Preventive Health Care [7].
(I) Evidence obtained from at least one properly ran-
domized controlled trial
(II-1) Evidence obtained from well-designed con-
trolled trials without randomization, or
(II-2) Evidence obtained from well-designed cohort or
case-control analytic studies preferably from more
than one center or research group, or
(II-3) Evidence obtained from comparisons between
times or places with or without the intervention.
Dramatic results in uncontrolled experiments are
included in this category.
(III) Opinions of respected authorities based on clin-
ical experience, descriptive studies, or reports of
expert committees.
The strength of recommendations was graded by using
the following system used in the previous CCCD [2,8]:
A: There is good evidence to support this maneuver.
B: There is fair evidence to support this maneuver.
C: There is insufficient evidence to recommend for or
against this maneuver, but recommendations might
be made on other grounds.
D: There is a fair evidence to recommend against this
procedure.
E: There is good evidence to recommend against this
procedure.
Ideally, as in the previous CCCD [8], grade A and E
recommendations were supported by Level 1 evidence, but
this could not always be the case; sometimes Level 2 evi-
dence was deemed sufficient. Note that a C recommendation
does not imply that the maneuver was useless or harmful;
there was simply insufficient evidence to make a stronger
recommendation. In the background articles, the grading
and strength of supporting evidence are given.
Participants were required to read the background arti-
cles prepared by each working group and vote on recom-
mendations. All recommendations receiving at least 80%
voting support were considered to have achieved consensus.
Those not achieving this degree of approval were either
modified (with the amended recommendation then brought
forward for further discussion and approval at the March
2006 meeting) or (rarely) were noted in the results but stated
to have achieved only weak consensus.
A 2-day meeting of all participants was held in Montreal.
At that meeting a brief overview was provided by each
working group. Then there was a presentation followed by
discussion and a vote on each modified recommendation.
Voting was open and transparent, and the process for for-
mulating recommendations was outlined in detail before the
conference. Several recommendations had to be reworked
to finally meet approval by the group.
A total of 214 separate recommendations were posted on
the website. One hundred forty-seven recommendations
were ultimately approved with strong consensus on the Web
 H. Chertkow / Alzheimer’s & Dementia 3 (2007) 262–265
or after subsequent modification and voting at the meeting.
Four other recommendations achieved weak consensus. The
others were rejected (or modified and incorporated into
other recommendations).
The Steering Committee for the CCCDTD3 decided
early in its deliberations that the work done should (1)
address the needs of family physicians and (2) address the
needs of specialists. To satisfy the first goal, a liaison with
the College of Family Physicians of Canada was estab-
lished. A representative (Dr John W. Feightner) of the
College sat on the organizing committee, and several family
practitioners (Drs Martha Donnelly and Chrisopher Frank)
were designated by the College to attend the meeting. Fur-
thermore, a family practitioner sat as an adviser on each
working group. That physician was charged with reviewing
the recommendations from a family practice perspective to
ensure their relevance and practicality. It is hoped that this
will help ensure the relevance and practicality of the rec-
ommendations agreed to.
Translating, disseminating, and implementing our rec-
ommendations received particular emphasis. During the last
half-day of the meeting, a session chaired by Dr Morris
Freedman focused on how to further our goals in the area of
knowledge translation. At this session Dr Freedman was
elected to chair an implementation committee that will fo-
cus on disseminating the final set of recommendations.
Work is being done at a number of levels:
1. Publication of the 18 background articles in this issue
of Alzheimer’s & Dementia.
2. Development of a series of five case-based articles for
the CMAJ that would highlight the most important
recommendations. These will focus on (1) risk factors
and prevention of AD, (2) MCI, (3) investigating
dementia, (4) mild and moderate dementia, and (5)
severe dementia.
3. Development of a tool kit/practice aids for distribu-
tion to Canadian family practitioners.
4. Production of course material for distribution to peer
and specialist teachers across Canada. The hope is
that this material will be used for small group, inter-
active educational activities developed for Canadian
practitioners.
5. Assessment of the impact of our recommendations on
clinical practice and patient outcomes.
These activities of the implementation committee are
currently underway.
In conclusion, the CCCDTD3 was a time-consuming but
a rewarding enterprise for all those who participated. Only
265
time will tell, however, whether we will improve both the
care provided by Canadian physicians and the outcomes of
patients with a dementia.
Author Disclosures
Howard Chertkow has received support from Pfizer Can-
ada (Advisory Board, Speaker, Grant Recipient), Neuro-
chem Inc. (Advisory Board), and Lundbeck Canada (Advi-
sory Board, Speaker).
Acknowledgments
My lab support staff of Victor Whitehead, Shelley So-
loman, Randi Pilon, and Kathy D’Souza provided invalu-
able support at all stages of this process. Dr. Anca Raus
carried out many of the literature services. MedPlan Com-
munications Inc. provided outstanding meeting and web
support for this process, including a sophisticated online
voting platform.
References
[1] Clarfield M. Assessing dementia: the Canadian Consensus. CMAJ
1991;144:851–3.
[2] Patterson CJ, Gauthier S, Bergman H, Cohen CA, Feightner JW, Feldman
H, et al. The recognition, assessment and management of dementing
disorders: conclusions from the Canadian Consensus Conference on De-
mentia. CMAJ 1999;160(Suppl):S1–15.
[3] Patterson CJ, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. Canadian Consensus Conference on Dementia: a
physician’s guide to using the recommendations. CMAJ 1999;160:
1738 – 42.
[4] Shumaker SA, Egault C, Rapp SR, Thal L, Wallace RB, Ockene JK, et
al. Estrogen plus progestin and the incidence of dementia and mild
cognitive impairment in postmenopausal women: the Women’s Health
Initiative Memory Study: a randomized controlled trial. JAMA 2003;
289:2651– 62.
[5] Shumaker SA, Legault C, Kuller L, Rapp SR, Thal L, Lane DS, et al.
Conjugated equine estrogens and incidence of probable dementia and
mild cognitive impairment in postmenopausal women: Women’s
Health Initiative Memory Study. JAMA 2004;291:2947–58.
[6] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan
DJ, et al. Assessing the quality of reports of randomized clinical trials:
is blinding necessary? Control Clin Trials 1996;17:1–12.
[7] Woolf SH, Battista RN, Anderson GM, Logan AG, Wang E. Assessing
the clinical effectiveness of preventive maneuvers: analytic principles
and systematic methods in reviewing evidence and developing clinical
practice recommendations—a report by the Canadian Task Force on
the Periodic Health Examination. J Clin Epidemiol 1990;43:891–905.
[8] Patterson C, Gauthier S, Bergman H, Cohen C, Feightner JW, Feldman
H, et al. The recognition, assessment and management of dementing
disorders: conclusions from the Canadian Consensus Conference on
Dementia. Can J Neurol Sci 2001;28(Suppl 1):S3–16.
 Alzheimer’s & Dementia 3 (2007) 266 –282

Review Articles

Mild cognitive impairment and cognitive impairment, no dementia:

Part A, concept and diagnosis
Howard Chertkowa,b,c,d,*, Ziad Nasreddinee, Yves Joanetted,f, Valérie Droletd, John Kirkb,
Fadi Massoudg, Sylvie Bellevilled,h, Howard Bergmana,b
a
Bloomfield Centre for Research in Aging, Lady Davis Institute for Medical Research, Sir Mortimer B. Davis-Jewish General Hospital,
McGill University, Montreal, Quebec, Canada
b
Division of Geriatric Medicine, Department of Medicine, Sir Mortimer B. Davis-Jewish General Hospital, McGill University, Montreal,
Quebec, Canada
c
Department of Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada
d
Centre de recherche, Institut Universitaire de Gériatrie de Montréal, Montreal, Quebec, Canada
e
Département de médecine, service de neurologie, Hôpital Charles LeMoyne and Université de Sherbrooke, Montreal, Quebec, Canada
f
Faculté de Médicine, Université de Montréal, Montréal, Quebec, Canada
g
Service de Gériatrie, Centre Hospitalier de l’Université de Montréal, and Département de Médecine, Université de Montréal, Montréal,
Quebec, Canada
h
Department de Psychologie, Université de Montréal, Montréal, Quebec, Canada

Abstract Mild cognitive impairment (MCI) and cognitive impairment, no dementia (CIND) are contro-
versial emerging terms that encompass the clinical state between elderly normal cognition and
dementia. This article reviews recent work on the classification of MCI and CIND, their prognosis,
and diagnostic approaches and presents evidence-based recommendations approved at the meeting
of the Third Consensus Conference on the Diagnosis and Treatment of Dementia held in Montreal
in March 2006. New short tools such as the Montreal Cognitive Assessment are making it easier for
family physicians to confidently attach the label of MCI.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Mild cognitive impairment; Alzheimer’s disease; Dementia; Diagnosis; Cognitive impairment, no dementia

1. Introduction In this article we will review the concept of MCI as well as

Dementia, mild cognitive impairment (MCI), and normal
aging represent a continuum of cognitive states encountered
in the elderly. As our ability to treat dementia such as
Alzheimer’s disease (AD) improves, there is corresponding
increase in our interest in mild cognitive decline that might
presage AD and dementia. This interest is motivated by
patient requests for prognosis and treatment. Treating phy-
sicians are currently at sea in a wealth of research informa-
tion and a dearth of practical suggestions for patient man-
agement. This report constitutes a background article for the
Third Canadian Consensus Conference on the Diagnosis
and Treatment of Dementia (CCCDTD3), held March 2006.
*Corresponding author. Tel.: 514-340-8260; Fax: 514-340-8295.
E-mail address: howard.chertkow@mcgill.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.013
cognitive impairment, no dementia (CIND), their prognosis,
the search for useful prognostic markers that predict pro-
gression to dementia, and current developments in treatment
of MCI. The recommendations presented here were ap-
proved by at least 80% of the votes cast by the 45 experts
who attended the CCCDTD3.
2. Methods
An evidence-based review was carried out to search for
references for CIND and MCI by using the terms “Ques-
tionable dementia” OR “Age-associated cognitive decline”
OR “Cognitive impairment” OR “Cognitive impairment no
Dementia” OR “Mild Cognitive Impairment” as primary
search terms. These were combined with the secondary
search terms “concept,” “criteria,” “diagnosis,” “therapy,”
 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
or “management.” Searches were carried out in both
PubMed and Embase going back to 1996 (the last date of
the Second Canadian Consensus Conference on Treatment
of Dementia searches). This search was supplemented by
articles in the authors’ files. The articles chosen were broken
down further into articles dealing with (1) the concept,
prognosis, and pathology of the entities, (2) diagnosis, (3)
nonpharmacologic therapies, and (4) pharmacologic thera-
pies. A preliminary list of 1,729 potential articles was ob-
tained through this search. The list was reduced to 314
articles on the basis of methodologic quality, presentation of
original data, and non-overlap with other reports. These
included 80 articles regarding diagnosis, 28 regarding phar-
macologic therapy, 16 related to nonpharmacologic therapy,
and the rest related to concept, natural history, and progno-
sis of the terms being considered.
For therapy, a standardized analysis of the quality of the
evidence was undertaken, integrating a rating of the quality
of the report by using the Jadad scale [1] and such quanti-
tative aspects as statistical significance and magnitude of
effects (see also www.naturalstandard.com).
Recommendations were rated on the basis of rules of
evidence developed by the Canadian Task Force on the
periodic health examination [2]. Ratings for recommenda-
tions were ranked at the following levels:
(I) Evidence obtained from at least one properly random-
ized controlled trial
(II)
(1) Evidence obtained from well-designed controlled tri-
als without randomization, or
(2) Evidence obtained from well-designed cohort or
case control analytic studies, preferably from more
than one center or research group, or
(3) Evidence obtained from comparisons between times
or places with or without the intervention. Dramatic
results in uncontrolled experiments are included in
this category.
(III) Opinions of respected authorities based on clinical
experience, descriptive studies, or reports of expert com-
mittees.
Recommendations are graded as follows:
A: There is good evidence to support this maneuver.
B: There is fair evidence to support this maneuver.
C: There is insufficient evidence to recommend for or
against this maneuver, but recommendations might
be made on other grounds.
D: There is a fair evidence to recommend against this
procedure.
E: There is good evidence to recommend against this
procedure.
267
3. CIND, MCI, and transition cognitive states of the
elderly
The Canadian Study of Health and Aging (CSHA) has
estimated the prevalence of dementia to be 8% in the pop-
ulation older than age 65 years [3]. The best estimates of
lifetime risk of AD range from 14.5% to 26.2% of the
population [4]. With the accepted criteria of McKann et al
[5], most physicians feel fairly confident now in diagnosing
AD and dementia. Significant impairment in two or more
domains of cognition, significant impairment in day-to-day
functional activities, and lack of other explanations for the
decline are the mainstays of dementia diagnosis. The main
limitation in diagnosing dementia is now finding the requi-
site time to carry out the assessment!
What is more challenging are the patients who do not
have dementia, but who are not “normal”. First, the concept
of “normality” in an elderly person is itself controversial.
Some researchers stress that maintenance of the same level
of performance shown by a young person is the goal, and
one definition, that of age-associated memory impairment
(AAMI) [6], compares the performance of elderly subjects
with that of younger persons. Up to 80% of individuals in
their 80s will fall into the AAMI category by demonstrating
memory performance at least 1 standard deviation (SD)
below mean test values for younger subjects [7]. Longitu-
dinal follow-up shows this group to be quite heterogeneous,
containing individuals preserving their cognition and a sub-
group deteriorating toward dementia [8]. The argument that
such a label captures a legitimate target group for drug
therapies that prevent cognitive decline [9] has proved con-
troversial. In essence, many of the individuals falling within
the AAMI group are normal. This underlines the majority
view that normality must be determined with respect to a
particular age group. Cognitive aging is thus considered a
normal phenomenon and not a pathologic condition. An
older individual whose memory performance and day-to-
day function is “about the same as others his or her age”
[10] is arguably normal.
Note that normality does not require total absence of
disease (three fourths of normal elderly subjects take a
medication, for instance) but performance at the same level
as others their age. This assessment is largely subjective and
is influenced by the nuance of questions. When asked “do
you have memory problems?”, one study found that 67% of
elderly subjects replied in the affirmative. When the ques-
tion was rephrased “Do you have memory problems relative
to most people your age?”, only 6% replied positively [10].
Retesting elderly on memory tests during a period of 2 years
will reveal that about one third of them are exhibiting some
subtle but measurable memory decline over time [11]. On
the revised Wechsler Adult Intelligence Scale IQ scale, an
85-year-old can achieve an IQ score of 100 by answering
only half as many questions correctly as a 21-year-old!
Hence, to claim that an elderly person shows “cognitive
268 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
impairment,” it is important to be comparing that person
with a group of individuals similar in age.
While supporting this concept of normal cognitive aging
in terms of clinical assessment, it must be noted that there
are recent results that call this notion into serious question.
Sliwinski et al [12] derived measures of normal cognitive
aging from a study group, but they then followed those
individuals for many years, during which a significant por-
tion of their normal elderly group progressed to MCI or
dementia. They examined the effects of preclinical demen-
tia on estimates of normal cognitive function in the aged.
After excluding preclinical dementia cases, they found that
many changes previously deemed normal cognitive aging
simply disappeared. The results indicate that by failing to
exclude preclinical dementia, conventional normative stud-
ies underestimate the mean, overestimate the variance, and
overestimate the effect of age on cognitive measures. With
a different approach, Bennett et al [13] assessed 134 elderly
individuals in a longitudinal study who had been classified
as normal and who soon thereafter died and came to au-
topsy. They found that more than one third met National
Institute on Aging (NIA)–Reagan pathologic criteria for
intermediate likelihood of AD, and these individuals actu-
ally scored about one fourth standard unit lower on tests of
episodic memory. Again, cognitive aging was being driven
by preclinical AD.
In the CSHA, subjects in a population study were cate-
gorized as no cognitive loss, CIND, or dementia [3,14].
CSHA defined CIND on the basis of a consensus conference
of physician, nurse, and neuropsychologist, integrating all
available information from clinical and neuropsychological
assessment. Mild objective memory problems falling be-
tween these two poles of normality and dementia are a
common phenomenon in older people; the CSHA docu-
mented a 16.8% prevalence of cognitive impairment with-
out dementia in the elderly [15,16]. Individuals falling into
this category exhibited an increased risk of death and even-
tual dementia (about 50% after a 5-year follow-up period)
[17]. CIND as a label has the benefit that inclusion depends
on a conclusion simply that there is cognitive impairment,
without exclusions related to the degree of decline, presence
of functional impairment, or suppositions regarding under-
lying etiology. The CSHA label does not appear to have
been based solely on psychometric cutoff scores, although
psychometric tests were administered.
A number of limitations, however, are inherent in using
this term. One drawback is that an individual might be
considered as having CIND as a result of lifelong static
brain damage, mental retardation, schizophrenia, and a
range of pathologies that would be obvious to the treating
physicians. The prognosis and diagnostic approach to CIND
would not be as relevant to the treating family physician or
specialist, who is focused on patients complaining of a new
decline in memory. When CIND as a label was applied to a
cohort derived from referral Memory Clinics, the causes
appeared quite heterogeneous [18]. A subclassification of
CIND included amnestic 25.1%, vascular cognitive impair-
ment 18.1%, psychiatric 17.2%, neurologic 7.3%, medical/
toxic metabolic 3.5%, mixed 7.6%, and not specified 19.0%.
Another drawback is that there is no agreement on how the
cutoff scores should be set or which tests should be used.
The Kungsholmen Project [19] used the label CIND, but
they based the diagnosis on scoring at least 1 SD below the
age and education specific mean on the Mini-Mental State
Examination (MMSE) score derived from a normative
group [20].
A different approach has been used in the Washington
University longitudinal cohort study of memory loss. These
individuals were referred to an Alzheimer Disease Research
Center and classified according to severity of decline as-
sessed by a semistructured interview of the client/patient
and the caregiver, scored in a scale, the Clinical Dementia
Rating scale, developed in 1982 [21]. Degree of ability in
areas such as memory, orientation, social and functional
activities, and hobbies were assessed and combined to give
a single numeric score. In this scale, 0 indicated normal
function with no loss, 1.0 indicated significant loss almost
always diagnostic of dementia, and a transition level, 0.5
(termed questionable dementia), indicated some very mild
loss that might or might not be classified as dementia. It has
been pointed out, however, that you can be CDR 0.5 and
nevertheless meet diagnostic criteria for AD and dementia,
thus making this grouping significantly different from MCI.
What is important to note is that primacy is given to a
detailed and expert history taken from patient and family,
focusing on sometimes subtle features of decline and loss
(rather than a particular cutoff score on any test) as the
critical feature. Morris et al [22] have argued that this
approach is valid and accurate and takes into account an
individual-oriented concept of degenerative disease that is
more sensitive than neuropsychological testing. Many (but
not all) CDR 0.5 subjects would be classified as MCI. The
expertise to apply this approach might be limited to only a
few experts in the field.
There are other global measures, such as Reisberg’s
Global Deterioration Scale [23], that also focus on “wors-
ening” from a previous level. Most MCI individuals score as
Stage 3, earliest subtle deficits, on the Global Deterioration
Scale, but it is not clear that the two concepts are inter-
changeable [24]. Both share the view that they are defining
MCI as a predementia stage.
A somewhat different approach focuses on patients’ and
families’ subjective complaints of memory and cognitive
loss as the starting point. A person’s degree of complaints
might not be equivalent to their actual objective degree of
decline, of course. Clearly, elderly subjects might complain
of memory loss as a result of anxiety, dementia in family
members or friends, or mild depression. At the same time,
studies show that memory complaints in elderly people
should no longer be considered merely as an innocent age-
 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
related phenomenon or a symptom of depression. Instead,
these complaints deserve to be taken seriously, at least as a
possible early sign of actual decline that could ultimately
lead to dementia [25]. At the same time unless there is
objective supporting evidence of real memory loss, the
prognosis is relatively benign [26,27].
Historically, this conception of subjective complaint of
cognitive loss, accompanied by objective supporting evi-
dence of loss, was originally suggested by Levy and an
International Psychogeriatric Association working group
[28]. They termed it age-associated cognitive decline
(AACD). Currently, the most widespread term in use to
characterize this group is MCI, derived from the World
Health Organization and adapted by a number of centers
[29 –31]. MCI is a clinical label that includes elderly sub-
jects with short-term or long-term memory impairment and
with no significant daily functional disability. The original
diagnosis of MCI required subjective report of cognitive
decline from a former level, gradual in onset, and present
for at least 6 months. This subjective complaint is supple-
mented by objective evidence of memory and learning de-
cline, with other cognitive domains remaining “generally
intact” [32]. There was no clear delineation as to how the
presence of memory loss was to be established. In all cases,
this term excluded individuals with significant depression,
delirium, mental retardation, or other psychiatric disorders
likely responsible for the impairment. If the memory loss
was severe and accompanied by significant functional im-
pairment, then the individual met clinical criteria for de-
mentia and was no longer MCI. The concept was intended
to capture and classify patients who seem to have a cogni-
tive problem that one would be loathe to label as normal and
yet clearly is not severe enough to qualify as dementia.
Modifications of the MCI concept have developed during
the past decade. Although Petersen et al [31,33] stressed
that MCI should be based on clinical judgment, they men-
tioned that most MCI subjects fell 1.5 SDs below norms on
memory tests. Subsequently it has become usual to assign
an objective level of impairment as performance either 1.0
SD or usually 1.5 SDs below published norms for a memory
test. Furthermore, there has been a push for greater speci-
ficity in terms of using three or four patterns of subgroups.
One subgroup would be elderly individuals whose deficits
are restricted to memory, a so-called amnestic or amnestic
single MCI [33,34]. Petersen et al [31] describe most MCI
subjects as showing only memory impairment, and these
amnestic MCI are more likely to go on to AD. Other
subgroups (amnesic multiple) are characterized by impair-
ment extending beyond memory to include subtle but mea-
surable changes in executive function, attention, naming
[35]. A third subgroup would involve individuals with im-
pairment in a single non-memory cognitive domain, while
memory remains normal. These could sometimes also in-
volve multiple non-memory domains as a fourth subgroup
[31]. Presumably such subjects are more likely to go on to
269
another neurodegenerative disease, such as primary pro-
gressive aphasia. Alternatively, they might be exhibiting
evidence of vascular damage (now termed vascular MCI or
vascular cognitive impairment). We and others have found
that most subjects initially labeled amnestic single MCI by
the clinician in fact exhibit subtle deficits in other neuro-
psychological and reaction time tests [36,37]. The clinician
should not be surprised when an MCI individual performs
slightly abnormally on placement of hands on clock draw-
ing. The validity of these subtypes, in terms of presumed
different prognoses for developing AD, is quite unproven at
this point. The details of how a subgroup should be applied
are somewhat vague, as are which tests to use and how to
apply them. It is not at all clear that practicing clinicians
need to delineate these subgroups in any formal manner, or
that these subgroups impact significantly on prognosis [38].
In fact, there have been numerous slightly different opera-
tionalized criteria that have appeared in the literature.
In our experience, use of different norms and different
tests, however, can significantly alter inclusion across sub-
groups [39,40]. A recent European Consortium on AD rec-
ommended that general medical practitioners should focus
on establishing the presence of an MCI syndrome, whereas
specialists could focus on the second stage of establishing
the MCI subgroup and the third stage of analyzing the
etiologic subtype (ie, degenerative, vascular, traumatic) in
each case [41].
MCI is a controversial concept and is now the focus of
natural history studies, biomarker studies, along with AD
prevention studies. The MCI stage might be the optimum
stage at which to intervene with preventative therapies. It
will increasingly become a label used by neurologists, ger-
iatricians, and family physicians who treat elderly cogni-
tively impaired individuals. Current interest is largely pred-
icated on the fact that patients classified as MCI have a high
rate of conversion to dementia, particularly to AD. It is now
becoming clear that many MCI subjects have AD pathology
in its earliest stages and will progress to AD if given
sufficient follow-up. In most clinic-based studies of MCI,
about 44% of individuals meeting the criteria for this diag-
nosis convert to AD during a 3-year follow-up, giving an
annual conversion rate of about 15% per year [42]. What is
controversial is whether all MCI individuals progress to AD
with time. In a 10-year follow-up in a Maastricht Memory
Clinic, Visser et al [43] found the risk of progression to
dementia to be 48% in 64 MCI subjects. In one Canadian
group of 89 MCI subjects accumulated in a Memory Clinic
setting, it appeared that about 25% would not progress even
10 years after onset of memory problems [44]. In population-
based studies, the prognosis of mild cognitive deficits seems
much less ominous. Ritchie et al [45] found that only 22%
of such a group went on to a degenerative dementia during
8 years of follow-up. In contrast, Morris et al [46], with
CDR 0.5 as their target group, have argued that with suffi-
270 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
Table 1
General criteria for MCI
Subjective complaint of memory loss
Objective impairment of memory
Generally preserved other cognitive ability
Preserved basic day-to-day functioning
No other obvious medical neurologic or psychiatric explanation for the
memory problems
Individual does not meet criteria for dementia
NOTE. Modified from Petersen and Smith, 1999.
cient time, most or all subjects in this category would
deteriorate to dementia, specifically AD.
Much of the disagreement might center on different
diagnostic criteria being used to assemble cohorts of sub-
jects for study and different means of collecting subjects.
There are multiple different operational definitions of
groups presented or discussed as MCI (Table 2). Depending
on the diagnostic criteria used, the prevalence and incidence
of MCI vary by a factor of 8 times. The CSHA term CIND
[16] would have encompassed virtually all MCI individuals,
along with numerous subjects who would not have met the
MCI criteria (some had alcoholism, large strokes, or mental
retardation, for instance). It is estimated that one half to two
thirds of subjects with CIND would meet criteria for MCI.
But is that an improvement? In fact, is one definition supe-
rior to the others?
We would argue that these concepts are designed for
different goals and have different strengths. A definition like
CIND that can be satisfied by objective test cutoff scores is
better designed for community and population studies.
What it lacks, however, is ecologic validity in the Memory
Clinic or clinical setting. In the doctor’s office, patients are
presenting with a memory complaint. Thus, definitions such
as MCI that begin with the individual’s complaint are more
likely to correspond with clinical reality for a physician
[47]. There are several weak points regarding the MCI
definition, however. One is the lack of agreed on criteria for
determining objective memory loss, the key feature of MCI.
In the clinic, one might accept impaired three-word recall as
sufficient evidence of memory impairment in a hitherto high
functioning individual now complaining of memory loss.
An alternative approach favored by psychological research-
ers, however, is to insist that MCI must be operationalized
to be meaningful [48]. This is generally done by including
only individuals exhibiting abnormal performance at least 1
SD or even 1.5 SDs below the norm for age-appropriate
populations on standardized verbal memory tests [32,45].
This attempt to establish statistical cutoff criteria for MCI,
while well-motivated, is fraught with difficulty; who de-
cides which memory tests to be used, which norms, which
cutoff point? Can memory tests be adequate for culturally
and educationally heterogeneous populations? Are the
norms reliable for a particular cultural setting? Rigorous use
of psychometric cutoff points provides more reliability to
the diagnosis and is useful for randomized clinical trials, but
it restricts the diagnosis of MCI to centers having easy
access to neuropsychological evaluation. Furthermore, as
Morris et al [22] have pointed out, these cutoff scores lose
the strength of individual assessment that emphasizes de-
cline from a previous level of performance rather than a
single test assessment.
Similarly, Petersen et al [31,33] had initially suggested
that non-memory domains be “generally intact” in MCI, and
in terms of bedside mental status testing, this is a reasonable
criterion to apply. In response to this, Ritchie and Touchon
[49,50] applied this notion that to satisfy the criteria for
MCI, all other domains besides verbal delayed memory
must be within 1.0 SD of the age-adjusted norm. Unfortu-
nately, when such constraints were firmly applied, there
were few putative MCI subjects who still existed. In fact,
detailed cognitive and neuropsychological testing shows
that a range of subtle “low normal” test results is the
majority finding in groups of MCI individuals [51]. These
neuropsychological test abnormalities, however, are not so
evident with the much cruder bedside testing carried out by
medical clinicians. Because one major goal of using the
terms MCI and dementia is to allow general physicians to
better assess, diagnose, and treat patients they see, the di-
agnostic criteria might be more meaningful when viewed
with respect to clinical mental status assessment and simple
bedside testing, rather than being dependent on more accu-
rate (and often unavailable) neuropsychological criteria.
There are other problems with overreliance on quantita-
tive criteria for diagnosing MCI. It is a fact that many
elderly individuals who would be rated as cognitively intact
are in fact substantially impaired on tasks that tap into a
range of cognitive systems [52]. Statistically speaking, there
are normal subjects who will sit for their whole lives at the
bottom of any gaussian curve of neuropsychological results,
and these individuals must not be falsely labeled as MCI.
Cultural, educational, and attentional factors (ie, impaired
attention as a result of stress, anxiety, or depression) can
also impact on neuropsychological testing. Thus, clinical
judgment from a physician or psychologist is essential in
assessing whether an MCI label is warranted. Neuropsycho-
logical tests can be used, however, to support such clinical
decision making.
One critical element in the diagnosis of dementia is
significant impairment in functional ability. Unfortunately,
there are few agreed on criteria for evaluation of function or
what significant functional impairment entails. About 31%
of MCI subjects likely exhibit subtle functional changes;
having to maintain written lists of things might well be
considered a functional alteration [53]. The key thing is that
such changes should only affect higher functions and not
represent significant impairment, although this is crudely
defined. In fact, one group defining MCI suggests that
impairment in complex functions should be an inclusion
criterion for MCI.
Table 2
Different criteria for MCI
Study Reference Complaint Non-memory Functional Indication of Normal Tests Used to Assess Objective Cutoffs That Were
Required? Cognitive Deficits Deficits Allowed? Intellectual Functioning Used in the Definition
Allowed?

Petersen et al [30], 1999
Morris et al [46], 2001
Ritchie et al [49], 2001
Y Not indicated N Y WAIS-R, WMS-R, Auditory Verbal Learning n/a
Test, Wide-Range Achievement Test-III,
MMSE, DRS, Free and Cued Selective
Reminding Test, Boston Naming Test,
Controlled Oral Word Association Test,
Category Fluency Procedures
Not indicated Y Y Not indicated CDR CDR ⫽ 0.5
Y N N Y Examen cognitive par ordinateur (ECO) n/a

H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282

Bennett et al [94], 2002
Darby et al [122], 2002
Hänninen et al [123], 2002
Kabani et al [124], 2002
Larrieu et al [99], 2002
Busse et al [125], 2003
Fisk and Rockwood [63], 2003
Lambon et al [126], 2003
N Y
Y N
N Y
Y Y
Y Y
Examined as a Y N
clinical variable
Examined as a N Examined as a
clinical variable
Y N
Not indicated Not indicated Extensive neuropsychological battery Population-specific cutoff
scores used
N Not indicated CERAD word list recall ⬍6
N MMSE ⱖ20 CDR CDR ⫽ 0.5
Buschke Selective Reminding Test (total ⱖ1.5 SD below healthy
immediate recall during 6 trials ⫹ delayed subsample cutoff
recall)
Visual reproduction Test (WMS) (immediate
and delayed recall)
Logical Memory Test (WMS) (immediate
and delayed recall)
Boston Naming Test (abbreviated)
Verbal Fluency Test
Trail making Test (A⫹B)
MMSE
N Not indicated CDR CDR ⫽ 0.5
Tests of explicit memory ⱖ1 SD below age-adjusted
norms
Y MMSE ⬍1 SD from Benton Visual Retention test ⱖ1.5 SD below age &
age & education education adjusted means
adjusted cohort
means
Not indicated Structured Mental Status (SIDAM) ⬎1 SD below age & education
associated norms on subtests
of SIDAM
Y Benton Visual Retention test; Buschke Cued No cutoffs set; diagnosis based
clinical variable recall test, Digit Span, and auditory verbal on clinical consensus
learning test; WAIS-R Similarities,
Comprehension, Digit Symbol, Block
design; Token test; COWA; Animal
naming.
Y MMSE ⬎24 Logical Memory (immediate) ⱖ1.5 SD below norms
DRS (memory)

271
Table 2

272
Continued
Study Reference Complaint Non-memory Functional Deficits Indication of Normal Tests Used to Assess Objective Cutoffs That Were
Required? Cognitive Deficits Allowed? Intellectual Functioning Used in the Definition
Allowed?

Lopez et al [127], Necessary for Y Y Not indicated n/a ⱖ1.5 SD below a sample of
2003 “probable” MCI, 250 unimpaired subjects
but not
“possible” MCI
Farlow et al [128], — Y Y Not indicated CDR CDR ⫽ 0.5
2004 (InDDEx NYU Paragraph recall ⬍9
study) HAM-D ⬍13
HAM-D item 1 ⱖ1
Feldman et al [129], N Not indicated Not indicated Not indicated CDR CDR ⬍1

H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282

2004 NYU paragraph recall ⬍9
Ganguli et al [34], Y Not indicated N MMSE ⬎24 CERAD 10 word list delayed recall ⬎1 SD below mean of cohort
2004
Grundman et al [130], Y Not indicated Y MMSE ⱖ24 Logical Memory II ⱖ8 (⫹16 y education)
2004 (ADCS-MIS Hamilton Dep. Rating Scale ⬍4 (8 to 15 y education)
study) ⱖ2 (0 –7 y education)
⬎12
Royall et al [37], 2004 Not indicated Examined as a Y Not indicated MMSE, CLOX 1,2 (clock drawing test), COWA ⱖ1.5 SD below average score
clinical variable (verbal fluency), EXIT 25 (executive of the sample on each
interview), Trail A ⫹ B, CVLT (verbal measure
memory), DRS: MEM
Solfrizzi et al [131], N Y Y MMSE ⬍1.5 SD from Babcock Story Recall Test Score in the lowest 10th
2004 age & education percentile of the cohort
adjusted means
DeJager et al [132], N N Y Not indicated CERAD 10-word list (free and delayed recall), ⱖ1.5 SD below norms
2005 Verbal Paired Associates (cued recall test),
Pattern and spatial recognition from the
CANTAB, CANTAB Paired Associates
Learning (6-items), The Placing Test (TPT)
Devanand et al [133], Y Y Y MMSE ⱖ22 MMSE Delayed recall ⱖ2 of 3 items recalled
2005 or or ⱖ1 SD below norms
if Spanish speaking Selective Reminding Task ⱖ10 points below WAIS-R
with education ⬍5 y or verbal IQ score
MMSE ⱖ18 WAIS-R performance IQ score If no deficits on objective tests,
memory complaint ⫹
informant’s confirmation of
decline and functional decline
Geslani et al [134], Y Not indicated N WAIS-R similarities or CVLT ⱖ1.5 SD below age & sex
2005 digit symbol ⬍0.5 adjusted norms
SD from mean
Kumar et al [135], Y Not indicated N MMSE ⬎25 CVLT ⬍2
2005
Nasreddine et al [112], Y Y Y Y Rey Auditory Verbal Learning Test At least 1 SD below norms
2005 2 subtests of WMS
(delayed visual reproduction and logical
memory)
Table 2
Continued
Study Reference Complaint Non-memory Functional Deficits Indication of Normal Tests Used to Assess Objective Cutoffs That Were Used in the
Required? Cognitive Deficits Allowed? Intellectual Functioning Definition
Allowed?

H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282

Purser et al [136], 2005 N Y Examined as SPMSQ ⬍4 20-item immediate word recall test ⬍4
clinical variable
Visser et al [137], 2005 Not indicated Y Y Not indicated Global Deterioration Scale 3
Blessed Dementia Rating scale ⱖ0.5 on first 8 items if GDS ⫽ 2
20-item story recall Cutoff based on centile score estimations
for LM, RAVLT, NYU paragraph
recall
Gal Int-11 Not indicated Y Y Not indicated CDR CDR ⫽ 0.5
(as described in Visser et al NYU Paragraph recall ⬍11
[137], 2005)
Ampakine study (as Not indicated Y Y MMSE ⬎24 CDR CDR ⫽ 0.5
described in Visser Logical Memory II ⱕ12 (⫹16 y education)
et al [137], 2005) Geriatric Dep. Scale ⱕ4 (8 to 15 y education)
ⱕ2 (0 –7 y education)
⬎6
Piracetam study (as Not indicated Y Y Not indicated CDR CDR ⫽ 0.5
described in Visser et al Logical Memory I ⬍10
[137], 2005) Logical Memory II ⱖ5 pts below LMI
HDRS ⬎17
Rofecoxib Not indicated Y Y MMSE ⬎23 CDR CDR ⫽ 0.5
study (as described in Visser RAVLT total learning ⬍38
et al [137], 2005) HDRS ⬍12
Verghese et al [138], 2006 Y Y N Verbal IQ ⬎84 Blessed test 5-item memory phrase ⱖ3 errors (ⱖ1.5 SD below norms)

Abbreviations: CANTAB, Cambridge Neuropsychological Test Automated Battery; COWA, controlled oral word memory test; CVLT, California Verbal Learning Test; DRS: MEM, Dementia Rating
Scale: Memory Subscale; GDS, Global Deterioration Scale; HAMD, Hamilton depression scale; HDRS, Hamilton Depression Rating Scale; LM, logical memory; RAVLT, Rey Auditory Verbal Learning
Test; SPMSQ, Short Portable Mental Status Questionnaire; WAIS-R, revised Wechsler Adult Intelligence Scale; WMS-R, Wechsler memory scale–revised.

273
274 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
There are many diagnostic issues that arise in character-
izing MCI. Should patients with mild depression be ex-
cluded from consideration? Some studies find depression in
up to 60% of individuals with MCI who go on to AD, and
presence of depression might in fact be a useful prognostic
sign in MCI individuals [54]. For the above reasons, it is
usual to encounter subtle but significant heterogeneity in
MCI patients [55]. MCI is perhaps best viewed as a some-
what heterogeneous clinical syndrome that includes many
subjects in a dementia “prodrome state” who might some
day be distinguishable as such.
4. The Neuropathology of MCI
Part of the confusion concerning MCI is based on our
evolving understanding of the brain processes occurring in
the condition. It is clear that pathologically at least, some of
the MCI subjects are showing the very earliest stages of AD
[56]. This is characterized by neurofibrillary tangles, usually
in the hippocampus as well as the entorhinal cortex. When
AD ensues, the pathology has spread to the lateral temporal
cortex as well. Senile plaques occur only later on. Markes-
bery et al [57] reported the findings of 10 individuals dying
with amnestic MCI, compared with AD and normal con-
trols. The group showed increased neocortical neuritic
plaque density compared with normal controls, virtually as
severe as the AD brains. A study of individuals with CDR
0.5 impairment showed similar findings [46]. These findings
fit with the notion that many and indeed most MCI individ-
uals will go on to develop clinical dementia caused by AD.
The problems with this conclusion are three-fold. First, it is
known that some individuals who are cognitively normal
will also show AD pathology at autopsy [58]. In the study
by Knopman et al [58] of 39 cognitively normal elderly
individuals, four met NIA-RI criteria for intermediate like-
lihood of AD, seven cases met Consortium to Establish a
Registry for Alzheimer Disease (CERAD) criteria for pos-
sible AD, and 19 met Khachaturian criteria for AD. Simi-
larly, Bennett et al [59] found that between one third and
one half of normal elderly brains had a significant amount of
AD pathology. Second, it is also likely that MCI individuals
with a more benign outlook will not be presenting for
autopsies. There is no way to ascertain that those presenting
for autopsy are representative of the entire group. Finally, a
clinical-pathologic study of MCI from the Religious Orders
Study [59] showed that individuals dying with MCI were
intermediate between normal and AD not only in terms of
plaques and tangles but also in terms of cerebral infarcts and
Lewy body pathology. This stresses the nonspecificity of
MCI in terms of dementia etiology.
Of great interest now that MCI brains are being sec-
tioned, there appears to be no real increase in choline acetyl-
transferase (ChAT) activity in the cortex at the MCI or early
AD stage [60]. In one recent study, levels of hippocampal
ChAT activity were increased in MCI subjects, although not
related to memory loss. This up-regulation was speculated
to be compensatory in nature [61]. This is important infor-
mation in gauging the likelihood that MCI subjects will
respond to treatment with cholinesterase inhibitors.
5. Prognosis in MCI
Is it possible to delineate at initial presentation which
MCI subjects are most likely to progress to AD during a
limited period of time such as 5 years? Numerous attempts
have been made to delineate prognostic markers in MCI,
most involving rather small samples of subjects followed
for limited periods of time. Daly et al [62] studied 123
subjects recruited through the Boston media who all met
criteria of being CDR 0.5 on the CDR Scale. After 3 years
of follow-up, only 19% had progressed to dementia (AD),
2% progressed as a result of strokes, and 5% had become
normal. It was found that most of their group was milder in
terms of cognitive deficits than other studies, and that this
explained the lower progression/conversion to AD rate.
Daly et al found that whereas 4% of subjects scoring a CDR
sum of boxes of 1.0 (almost normal cognition) went on to
progress, 67% of those with a total box score of 3.0 (almost
demented already) progressed to AD. This demonstrated a
not surprising effect of severity; within the broad range of
individuals labeled with mild memory loss, those most
impaired to begin with were more likely to progress during
follow-up.
A study by Fisk et al [63] has called into question the
need to define MCI with subjective as well as objective
measures. With a variety of MCI definitions and subjects
derived from the CSHA, they found that changes of defini-
tion changed the prevalence of MCI but did not change the
proportion who progressed to dementia during a period of 5
years, which remained about 8% to 10% per year. In addi-
tion, all groups included a subgroup who had improved after
5 years.
A number of studies have looked at the impact of APOE
genotype on risk of developing AD and progression of
cognitive decline. The results are rather complex. APOE e4
genotype has been found to be a significant risk factor for
CIND in the Canadian ACCORD study [64]. In several
studies the presence of APOE e4 in cognitively impaired
elderly subjects was also associated with a greater chance of
memory decline during a period of 3 years [65,66]. Petersen
et al [32] suggested that the MCI individual’s APOE status
might be useful in prognostication. In their study, Petersen
et al found a major effect of APOE e4 load on progression
to AD from MCI. In other studies, APOE genotype ap-
peared to have little impact on the degree of decline [67].
A series of studies have looked at other predictors of
cognitive decline (ie, changing from normal to MCI) or
progression (from normal or MCI) to AD/dementia. One
recent study targeted high homocysteine levels as another
biologic risk factor, but this remains to be confirmed [68].
 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
High levels of low-density lipoprotein or a high ratio of
low-density lipoprotein/high-density lipoprotein were asso-
ciated with increased risk of later diagnosis of AD in the
Baltimore Longitudinal Study [69]. Serum markers for in-
flammation such as interleukin 6 have also been found to be
associated with cognitive decline in healthy elderly men and
women [70]. To what extent these biochemical markers,
alone or together, can be used in individual prognostication
remains to be elucidated.
Imaging markers might be useful in prognostication in
MCI, but these have mainly been used in academic research
settings. On magnetic resonance imaging scans, the formal
measurement of hippocampal volumes [71] and the rate of
change of global or hippocampal volumes [72,73] are pow-
erful tools to predict progression to AD. Even earlier alter-
ations in entorhinal cortex can be detected with positron
emission tomography [74,75], and these can be potentially
useful in predicting progression to AD even when cognition
is normal [76]. Although visual inspection of standard sin-
gle photo emission computed tomography (SPECT) brain
images is not useful in prognostication [77], the sophisti-
cated quantification of SPECT images might allow fairly ac-
curate prognostication [78]. Promising reports have emerged
with use of magnetic resonance spectroscopy [79] in predicting
onset of AD in patients with MCI. Altered patterns of positron
emission tomography activation can be detected during atten-
tional tasks [80] and with functional magnetic resonance im-
aging during memory tasks in MCI individuals [81]. These
could potentially be used for prognostication. Imaging remains
a very active field of diagnostic investigation that should pro-
vide helpful indicators in the coming years.
Neuropsychological and cognitive measures have been
assessed for their predictive capacity. It is unclear whether
cognitive tests are robust enough to allow prediction at the
individual patient level, as opposed to large subject groups.
At the same time, MCI individuals who complain, exhibit
global cognitive impairment, and exhibit episodic memory
deficits are at highest risk for progression to AD [82]. We
have found that MCI subjects with any disorientation to
time, or even subtle problems with clock drawing, are more
likely to progress to AD [44]. In fact, the presence of any
quantifiable abnormalities beyond memory constitutes a risk
for progression from MCI to dementia [83]. Batteries of
different formal neuropsychological tests have been used
together with the objective of stratifying risk of progression
to AD [84 – 87]. Tests of delayed verbal recall and executive
function appear to have the best discriminating power for
prediction. Although promising, these tests are not suffi-
ciently robust to prognosticate in all MCI cases.
Currently a long series of other biomarkers are under
investigation: smell testing [88], blood heme oxygenase 1
[89], CSF-tau, and A beta 42 [90] to note only a few.
What the majority of such studies conclude is largely
consistent and consists of three findings. (1) There appear to
be notable and measurable deficits in individuals who de-
275
velop AD a good number of years before the diagnosis is
officially made [91]. (2) Individuals with MCI or CIND are
at increased risk of dementia [92–95]. (3) At the same time,
a nontrivial percentage of MCI individuals do not progress
to dementia and even appear to become normal over time
[62,92,94,95].
6. Conclusions regarding the concept of MCI
MCI as a clinical entity continues to be debated. Morris
et al [46,96] argue that there is no such independent entity
as MCI, simply AD at various levels of severity, with some
individuals who have AD not even meeting criteria for
dementia. What will change in the future, they argue, is the
threshold at which we are willing to attach a diagnosis of
AD based on clinical information and biomarkers.
Another approach is to view MCI as a high risk group for
both death and development of dementia in the future [97].
This latter underlines the fact that pathologically, many
MCI individuals already have common conditions respon-
sible for age-related dementia, including AD, Lewy body
disease, and cerebral infarction [59].
A rather different perspective emerges from several com-
munity studies of MCI. In a Pennsylvania study of the
elderly, 4% were found to meet operational criteria for MCI.
About one fourth of these reverted to normal during 10
years of follow-up [34]. Similarly, a French population
study found the MCI diagnosis to be remarkably unstable,
with a significant proportion reverting to normal during
follow-up [98]. In a cohort of MCI individuals derived from
the CSHA, fully one third were considered to have no
cognitive impairment after 5 years. In the PAQUID study in
France, a similar instability of diagnosis and high reversion
rate to normal were noted [99]. Ritchie and Touchon [50]
concluded that it would be erroneous to consider MCI as a
clear early stage of AD or dementia.
This view is not limited to population studies, however.
Heterogeneity in the prognosis of MCI during the long term
emerges in cohorts collected from Memory Clinics in both
Europe [100] and Canada [101]. The MCI diagnostic cate-
gory appears to contain many individuals who will ulti-
mately progress to AD (or other dementias), along with a
subgroup, usually about one fourth of patients seen in Mem-
ory Clinic settings, who will not progress.
Should MCI be considered as a distinct entity? The
construct of MCI has been criticized as being difficult to
operationalize [102,103]. Some authors insist that better
approaches must be found, perhaps on the basis of change
scores and measures of cognitive decline over time
[102,104]. Some have urged criteria to be developed for
“prodromal/predementia Alzheimer Disease” [95,105]. The
conclusions of different groups stand in considerable oppo-
sition; work by Morris et al [46] suggests that individuals
with MCI already have mild dementia, whereas population
276 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282

studies and clinic studies of others stress the heterogeneity
and unpredictability of the outcome of MCI subjects.
There is a danger, however, of “throwing out the baby
with the bath water.” Attaching a diagnostic label to indi-
viduals who are neither (in the opinion of their physician)
normal nor demented accurately captures the clinical di-
lemma and uncertainty [106]. We know enough to advise
such patients on increased risk, arrange for closer follow-up,
and even investigate for underlying causes of memory loss.
The group has already been targeted for possible secondary
prevention therapy [107–109]. This, it would seem, is pref-
erable to the alternative, mislabeling them as normal and
misinforming the families as to a benign prognosis.
7. Recommendations approved at the CCCDTD3
meeting regarding concept of MCI:
1. Physicians should be aware that most dementias
might be preceded by a recognizable phase of mild
cognitive decline. Physicians should be familiar with
the concept of MCI (or CIND) as a high risk state for
decline and dementia (Grade B, Level 3).
2. There is currently inadequate evidence to recommend
one term or label (MCI, CIND) over another (Grade
B, Level 3).
3. There is inadequate evidence to advise MCI patients
and their families that the patient is already exhibiting
signs of dementia or to treat MCI as equivalent to
dementia (Grade C, Level 2).
4. There is fair evidence that physicians should
closely monitor individuals who have MCI or
CIND because of the known increased risk of both
dementia and death that has been documented
(Grade B, Level 2).
8. Diagnosis of MCI and screening for MCI by
physicians
The question as to how physicians should approach the
diagnosis of MCI or CIND is an important one. Is there an
agreed or accepted recommended way to diagnose MCI, or
does the approach vary depending on the definition used, eg,
MCI vs CIND? In 2001, the Quality Standards Subcommit-
tee of the American Academy of Neurology recommended
that MCI should be detected. There was sufficient evidence
suggesting that subjects with MCI were at increased risk for
developing dementia or AD when compared with similarly
aged individuals in the general population. Screening in-
struments such as the MMSE were found useful, as were
neuropsychological batteries [31].
Until 2001, there were no specific cognitive screening
instruments to detect MCI. Although the MMSE was con-
sidered useful, it had low sensitivity to detect MCI because
most subjects scored in the normal range on the test [33].
Since that time, several screening instruments have been
developed to help screen for MCI (Table 3). We will look at
the instruments currently available, validated in MCI,
CIND, and CDR 0.5.
The following screening instruments have shown prom-
ising results with well-designed studies based on the MCI
criteria of Petersen et al [33]:
The DemTect [110] evaluates immediate and delayed
recall of a word list, number transcoding, verbal fluency,
and reverse digit span. It was validated with 363 English-

Table 3
Brief cognitive screening tests for detection of MCI
Test First Author Community vs Subjects, Normal/MCI Cognitive Time to Administer Sensitivity/Specificity
Memory Clinic Domain(s) (min) (%)

CWL Shankle [118], 2005 Memory clinic and 119/95 Memory 15–20 (estimate) 82/91, 94/89 with CA
community statistical technique
MoCA Nasreddine [112], 2005 Memory clinic 90/94 Multiple 10 90/87
DemTect Kalbe [110], 2004 Memory clinic 145/97 Multiple 8–10 80/92
CMC Xu [115], 2002 Memory clinic 351 subjects. Multiple ⱕ15 (estimate) 83/80
retrospective design
RI-48-Adams Ivanoiu [111], 2005 Memory clinic 38/25 Memory 20–25 77/100
CDT Yamamoto [113], 2004 Memory clinic 41/48 Clock only ⱕ5 (estimate) 75/76
ECO Artero [116], 2003 Memory clinic 60/308 Multiple 5–10 73/99
Extended delayed Loewenstein [48], 2000 Memory clinic 52/24 Memory ⱕ20 (estimate) 70/84
recall MMSE
Mini-Cog Borson [119], 2005 Memory clinic 140/71 Multiple 3 55/83
STMS Tang-Wai [117], 2003 Memory clinic 788/29 Multiple ⱕ10 (estimate) AUC, 0.82; ⬍80/⬍80
Letter F and Canning [121], 2004 Memory clinic 46/25 (CIND) Language 3 (estimate) F: 48/57, Animal: 44/96
Animal fluency
6-item screener Callahan [120], 2002 Memory clinic and 708/44 Memory and 1.5 97/49
community orientation
ABCS Molloy [114], 2005 Memory clinic 111/124 Multiple 3 AUC, 0.7

Abbreviations: ABCS, AB Cognitive Screen; RI-48: Rappel Indicé.

 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
speaking subjects (97 MCI, 121 AD, 145 normal controls
[NC]). The test achieved a sensitivity of 80% and specificity
of 92% to detect MCI and separate it from NC. Positive
predictive value (PPV) and negative predictive value (NPV)
were not reported. MCI subjects were recruited in a memory
clinic. The DemTect can be administered in 8 to 10 minutes.
Rappel-indice-48-Adams (RI-48) [111] evaluates de-
layed cued-recall of 48 items. It was tested in 107 French-
speaking subjects (25 MCI, 22 AD, 22 subjective memory
complaints), 38 NC). The test achieved a sensitivity of 77%
and specificity of 100% to detect MCI and separate it from
NC. MCI subjects were recruited in a memory clinic. The
RI-48 can be administered in 20 to 25 minutes.
The Montreal Cognitive Assessment (MoCA) [112]
evaluates delayed recall, verbal fluency, visuospatial skills,
clock drawing, executive functions, calculation, abstraction,
language, orientation, attention, and concentration. It was
validated in French and English with 277 subjects (94 MCI,
93 AD, 90 NC). The test achieved a sensitivity of 90% and
specificity of 87% to detect MCI and separate it from NC.
The PPV was 89%, and the NPV was 91%. MCI subjects
were recruited in a memory clinic (in an academic center)
and in the community. The MoCA can be administered in
10 minutes.
Clock Drawing Test (CDT) was tested in 89 Japanese-
speaking subjects (41 normal elderly controls and 48 MCI
individuals) with 75% sensitivity and 76% specificity to
detect MCI and separate it from NC [113].
The AB Cognitive Screen (ABCS) evaluates delayed
recall, orientation, registration, clock drawing, and word
fluency. It was assessed in 235 English-speaking subjects
(124 MCI, 111 NC). The MCI subjects’ diagnosis was not
confirmed with neuropsychological testing, and demonstra-
tion of objective memory deficits was not considered as part
of the MCI criteria. The area under the curve (AUC) was
0.7 for distinguishing normal subjects from MCI. No
sensitivity or specificity was reported. MCI subjects were
recruited in a memory clinic. The test can be adminis-
tered in 3 minutes [114].
The Modified Mini-Mental State Examination with Ex-
tended Delayed Recall evaluates delayed recall of the three
MMSE words at three 5-minute intervals. It was assessed in
102 subjects (52 NC, 24 MCI, 26 dementia). Sensitivity and
specificity for MCI were 70.8% and 84.6%, respectively.
Subjects were recruited in a memory clinic [47].
The following screening instrument was studied by using
a retrospective predictive accuracy design:
The Combined Mini-mental-cognitive Capacity (CMC)
combines the MMSE with the Cognitive Capacity Screen-
ing Examination evaluating multiple cognitive domains and
are both scored out of 30 points. The total score of the CMC
is 47 points, scoring only once overlapping items in both
tests. The CMC was tested in English in 351 subjects eval-
uated for memory impairment at a memory clinic. Eighty-
four (23.9%) developed dementia (47 AD) in follow-up up
277
to 6 years (average, 3.89 years). Accuracy was determined on
the retrospective capability of the test to detect MCI on the
basis of final diagnosis of AD with follow-up. With specificity
set at 80%, sensitivity was 83%. PPV was 46%, and NPV was
96%. Administration time was not reported [115].
The following screening instruments were studied by
using a prospective predictive accuracy design.
The Examen Cognitif par Ordinateur (ECO) evaluates
multiple cognitive domains, including working memory,
verbal and visuospatial memory, implicit memory, language
skills, verbal fluency, visuospatial skills, and attention. It
was determined to be a good predictive test for determining
which subjects (308 with cognitive complaints, 60 without
cognitive complaints) will convert to AD in a 2-year pro-
spective study. Three subtests (delayed story recall, verbal
fluency, visuospatial construction) when combined achieved a
sensitivity of 73% and specificity of 99% to predict conversion
to AD. The test subtests could be administered in 5 to 10
minutes [116].
The Short Test of Mental Status (STMS) evaluates de-
layed recall, orientation, attention, learning, calculation,
construction, abstraction, and information. It was shown to
be as insensitive (less then 80% for both sensitivity and
specificity) as the MMSE for prevalent MCI and modest in
its ability to predict conversion to AD, with follow-up
average of 5.6 years [117].
The following screening instruments have shown prom-
ising results with well-designed studies based on CDR 0.5
criteria (questionable dementia) [21,22]:
The CERAD 10-word list (CWL) [118] evaluates imme-
diate and delayed verbal memory of a list of 10-words. It
has shown promising results with well-designed studies
based on CDR 0.5 criteria. It was tested in English, with 471
subjects classified according to CDR score of 0, 0.5, or 1
(NC 119, MCI 95, mild dementia 257). The test achieved a
sensitivity of 82% and specificity of 91% to detect MCI and
separate it from NC. With correspondence analysis, a tech-
nique that creates weighted scores from the individual per-
formance profile, the sensitivity and specificity reached
94% and 89%, respectively. PPV was 100%, and NPV was
97%. MCI subjects were recruited in a memory clinic and in
the community. Administration time is not reported.
The Mini-Cog [119] evaluates word learning, delayed
recall, and clock drawing. It was tested in English with 371
multiethnic subjects (71 MCI, 112 AD, 48 other dementia,
140 NC). The test achieved a sensitivity of 55% and spec-
ificity of 83% to detect MCI and separate it from NC. MCI
subjects were recruited in a memory clinic. The Mini-Cog
can be administered in 3 minutes.
The following screening instruments have shown either
low sensitivity or specificity (less than 80%) results with
well-designed studies, on the basis of CIND criteria rather
than MCI:
The 6-Item Screener [120] evaluates delayed recall and
orientation. It was tested in English with 995 subjects (244
278 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
CIND, 43 dementia, 708 NC). The test achieved a sensitiv-
ity of 97% and specificity of 49% to detect MCI and sepa-
rate it from NC in the community. PPV was 41%, and NPV
was 98.3. MCI subjects were recruited in a memory clinic
and in the community. The 6-Item Screener can be admin-
istered in 1.5 minutes in person or over the phone.
Letter F Fluency and Animal Fluency [121] achieved
sensitivities of 48% and 44% and specificities of 57% and
96%, respectively, to detect MCI and separate it from NC.
In conclusion, there is no accepted, agreed on, or recom-
mended way to diagnose or screen for MCI at the present
time according to the literature. No studies exist that com-
pare MCI screening tests validated according to different
definitions for mild memory loss, namely CDR 0.5, CIND,
and MCI as defined by Petersen [30 –32]. It appears that no
clear consensus exists in the literature for a specific ap-
proach for diagnosing MCI.
One can formulate the slightly different question as to
what approach to diagnosis makes sense for general prac-
titioners and specialists? The CWL was validated in CDR
0.5, and its administration time is not clear. Only three other
cognitive screening tests have shown high (greater than
80%) sensitivity and specificity for diagnosis of MCI and
therefore are the most promising tests currently available.
Furthermore, they are all easy to use in a clinical setting and
can be administered in less than 15 minutes. The three tests
are the MoCA, the DemTec, and the CMC.
9. Recommendations approved at the CCCDTD3
meeting regarding diagnostic maneuvers
1. In cases in which there is suspicion of cognitive
impairment or concern about the patient’s cognitive
status and the MMSE score is in the normal range (24
to 30), tests such as the MoCA, DemTect, or CMC
could be administered. These would help to demon-
strate objective cognitive loss (Grade B, Level 2).
2. There is good evidence that the addition of in-depth
neuropsychological testing can be recommended to
aid in the confirmation of the diagnosis (Grade A,
Level 1).
10. Areas for future research
(1) There is a need for long-term studies of the natural
history of MCI in the general physician or memory
clinic setting, as opposed to population studies.
(2) There is a need for formal comparison of the
different diagnostic approaches and labels to com-
pare their natural history and prognosis in a clin-
ical setting.
(3) There is a strong need for development of better AD
biomarkers. Ultimately, establishment of such biomar-
kers will make it possible to diagnose AD at a much
earlier point in time and make the current confusion
regarding MCI a thing of the past.
(4) Different clinical assessment algorithms for MCI
screening by first-line physicians need to be tested.
There is a need to assess and adapt simple memory
clinic tools to be used by first-line physicians in the
community setting.
(5) We should also assess the predictive value of screen-
ing measures in assessing risk of conversion to AD.
(Table 1).
Author Disclosures
Howard Chertkow has received support from Pfizer Can-
ada (Advisory Board, Speaker, Grant Recipient), Neuro-
chem Inc. (Advisory Board), and Lundbeck Canada (Advi-
sory Board, Speaker).
Ziad Nasreddine has received support from Pfizer (Con-
sultant, Adviser, Program Organizer, Lecturer), Novartis
(Lecturer, Drug Trials, Advisor), Janssen Ortho Inc. (Lec-
turer, drug trials, advisor), Neurochem (Drug Trial), Myriad
(Drug Trial), and Sanofi-Aventis (Drug Trial).
Fadi Massoud has received support from Janssen Ortho
Inc. (Advisory Board, CME Speaker, Research Funds),
Lundbeck Canada (Advisory Board, CME Speaker, Re-
search Funds), Novartis (Advisory Board, CME Speaker,
Research Funds), and Pfizer (Advisory Board, CME
Speaker, Research Funds).
Howard Bergman has received support from Pfizer
(Drug Trial) and Novartis (Drug Trial).
Acknowledgments
This work was supported by grants from a number of
agencies. Howard Chertkow is supported by operating
grants from the CIHR (Canadian Institutes for Health Re-
search), the Alzheimer Society of Canada, and the Fonds de
la recherche en santé du Québec (FRSQ).
Howard Chertkow and Sylvie Belleville receive “cher-
cheur national” awards from the FRSQ (Fonds de la Re-
cherche en Santé du Québec.
Howard Bergman is supported through the Dr Joseph
Kaufmann Chair in Geriatric Medicine at McGill Univer-
sity.
Yves Joanette was supported by a grant from the Cana-
dian Institutes of Health Research (MOP-15006).
Sylvie Belleville is supported by research grants from the
Canadian Institutes of Health Research (CIHR), Heart and
Stroke Foundation of Canada, Alzheimer’s Society of Can-
ada, CAREC, and NSERC.
References
[1] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Ga-
vaghan DJ, et al. Assessing the quality of reports of randomized
 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
clinical trials: is blinding necessary? Control Clin Trials 1996;
17:1–12.
[2] Woolf SH, Battista RN, Anerson G, Logan A, Wang, E. Asssessing
the clinical effectiveness of preventive manoeuvers: analytical prin-
ciples and systematic methods in reviewing evidence and develop-
ing clinical practice recommendations. J Clin Epidemiol 1990;43:
891–905.
[3] CSHA. The incidence of dementia in Canada: the Canadian Study of
Health and Aging Working Group. Neurology 2000;55:66 –73.
[4] Drachman D. If we live long enough, will we all be demented?
Neurology 1994;44:1563–5.
[5] McKann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM. Clinical diagnosis of Alzheimer’s disease: report of the
NINCDS-ADRDA work group under the auspices of the Depart-
ment of Health Services Task Force on Alzheimer’s disease. Neu-
rology 1984;34:939 – 44.
[6] Crook T, Bartus RT, Ferris SH, Whitehouse P, Cohen GD, Gershon
SG. Age-associated memory impairment: proposed diagnostic cri-
teria and measures of clinical change: report of a National Institute
of Mental Helath Work Group. Dev Neuropsychol 1986;2:261–76.
[7] Blackford RC, LaRue A. Criteria for diagnosing age-associated
memory impairment: proposed revisions from the field. Dev Neu-
ropsychol 1989;5:295–306.
[8] Hanninen T, Hallikainen M, Koivisto K, Helkala EL, Reinikainen
KJ, Soininen H, et al. A follow-up study of age-associated memory
impairment: neuropsychological predictors of dementia. J Am Geri-
atr Soc 1995;43:1007–15.
[9] Larrabee GJ. Age-associated memory impairment: definition and
psychometric characteristics. Aging Neuropsychology & Cognition
1996;3:118 –31.
[10] Salthouse TA. Theoretical perspectives on cognitive aging. Hills-
dale, NJ. Lawrence Erlbaum Associates, Inc; 1991. p. 438.
[11] Collie A, Maruff P, Shafiq-Antonacci R, Smith M, Hallup M,
Schofield PR, et al. Memory decline in healthy older people: impli-
cations for identifying mild cognitive impairment. Neurology 2001;
56:1533– 8.
[12] Sliwinski M, Lipton RB, Buschke H, Stewart W. The effects of
preclinical dementia on estimates of normal cognitive functioning in
aging. J Gerontol B Psychol Sci Soc Sci 1996;51:217–25.
[13] Bennett DA, Schneider JA, Arvanitakis Z, Kelly JF, Aggarwal NF,
Shah RC, et al. Neuropathology of older persons without cognitive
impairment from two community-based studies. Neurology 2006;
66:1837– 44.
[14] Tuokko H, Frerichs RJ. Cognitive impairment with no dementia
(CIND): longitudinal studies, the findings and the issues. Clin Neu-
ropsychol 2000;14:504 –25.
[15] Graham JE, Rockwood K, Beattie BL, Eastwood R, Gauthier S,
Tuokko H, et al. Prevalence and severity of cognitive impairment
with and without dementia in an elderly population. Lancet 1997;
349:1793– 6.
[16] Ebly EM, Hogan DB, Parhad IM. Cognitive impairment in the
nondemented elderly: results from the Canadian Study of Health and
Aging. Arch Neurol 1995;52:612–9.
[17] Tuokko H, Frerichs R, Graham J, Rockwood K, Kristjansson B,
Fisk J, et al. Five-year follow-up of cognitive impairment with no
dementia. Arch Neurol 2003;60:577– 82.
[18] Feldman H, Levy AR, Hsiung GY, Peters KR, Donald A, Black SE,
et al. A Canadian cohort study of cognitive impairment and related
dementias (ACCORD): study methods and baseline results. Neuro-
epidemiology 2003;22:265–74.
[19] Palmer K, Wang H-X, Backman L, Winblad B, Fratiglioni L. Dif-
ferential evolution in non-demented older persons: results from the
Kungsholmen Project. Am J.Psychiatry 2002;159:436 – 42.
279
[20] Folstein M, Anthony JC, Parhad I, Duffy B, Gruenberg EM. The
meaning of cognitive impairment in the elderly. J Am Geriatr Soc
1985;33:228 –35.
[21] Hughes C, Berg L, Danziger W, Coben L, Martin R. A new clinical
scale for the staging of dementia. Br J Psychiatry 1982;140:566 –72.
[22] Morris JC, McKeel DW Jr, Storandt M, Rubin EH, Price JL, Grant
EA, et al. Very mild Alzheimer’s disease: informant-based clinical,
psychometric, and pathologic distinction from normal aging. Neu-
rology 1991;41:469 –78.
[23] Reisberg B, Ferris SH, de Leon MJ, Crook T. Global Deterioration
Scale (GDS). Psychopharmacol Bull 1988;24:661–3.
[24] Kluger A, Ferris SH, Golomb J, Mittelman MS, Reisberg B. Neu-
ropsychological prediction of decline to dementia in nondemented
elderly. J Geriatr Psychiatry Neurol 1999;12:168 –79.
[25] Jonker C, Geerlings MI, Schmand B. Are memory complaints pre-
dictive for dementia? a review of clinical and population-based
studies. Int J Geriatr Psychiatry 2000;15:983–91.
[26] O’Brien J, Beats B, Hill K, Howard R, Sahakian B, Levy R. Do
subjective memory complaints precede dementia? a three-year
follow-up of patients with supposed “benign senescent forgetful-
ness”. Int J Geriatr Psychiatry 1992;7:481– 6.
[27] Flicker C, Ferris SH, Reisberg B. A longitudinal study of cognitive
function in elderly persons with subjective memory complaints.
J Am Geriatr Soc 1993;41:1029 –32.
[28] Levy R. Report of the working party of the International Psychoge-
riatric Association in collaboratino with the World Health Organi-
zation: aging-associated cognitive decline [published erratum ap-
pears in Int Psychogeriatr 1994;6:133]. Int Psychogeriatr 1994;
6:63– 8.
[29] Smith AD, Jobst K. Use of structural imaging to study the progres-
sion of Alzheimer’s disease. Br Med Bull 1996;52:3.
[30] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG,
Kokmen E. Mild cognitive impairment: clinical characterization
and outcome. Arch Neurol 1999;56:303– 8.
[31] Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL,
DeKosky ST. Practice parameter: early detection of dementia: mild
cognitive impairment (an evidence-based review)—report of the
Quality Standards Subcommittee of the American Academy of Neu-
rology. Neurology 2001;56:1133– 42.
[32] Petersen RC, Smith GE, Ivnik RJ, Tangalos EG, Schaid DJ, Thi-
bodeau SN, et al. Apolipoprotein E status as a predictor of the
development of Alzheimer’s disease in memory-impaired individu-
als [published erratum appears in JAMA 1995;274:538]. JAMA
1995;273:1274 – 8.
[33] Petersen RC. Mild cognitive impairment as a diagnostic entity.
J Intern Med 2004;256:183–94.
[34] Ganguli M, Dodge HH, Shen C, DeKosky ST. Mild cognitive
impairment, amnestic type: an epidemiologic study. Neurology
2004;63:115–21.
[35] Lopez O, Jagust W, DeKosky S, Becker J. Risk factors for mild
cognitive impairment in a population study. Neurology 2002;58:
A238.
[36] Chertkow H. Mild memory loss in the elderly: can we predict
development in dementia? Can J Neurol Sci 1998;25:1–S4.
[37] Royall DR, Chiodo LK, Polk MJ. Misclassification is likely in the
assessment of mild cognitive impairment. Neuroepidemiology 2004;
23:185–91.
[38] Winblad B, Palmer K, Kivipelto M, Jelic V, Fratiglioni L, Wahlund
LO, et al. Mild cognitive impairment: beyond controversies, towards
a consensus—report of the International Working Group on Mild
Cognitive Impairment. J Intern Med 2004;256:240 – 6.
[39] Chertkow H. What are the characteristics of patients with mild
cognitive impairment who do not progress to dementia? The Amer-
ican Neurological Association, San Francisco, CA, Oct 21, 2003.
280 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
[40] Chertkow HM, Duong A, Philips N, Bergman H, Babins L, Kelner N.
Mild cognitive impairment. the danger of basing the diagnosis on
published normative groups (P3-072). Paper presented at ICAD-
Alzheimer’s Association 10th International Conference on Alzheimer’s
Disease and Related Disorders, 2006, Madrid, Spain.
[41] Portet F, Ousset PJ, Visser PJ, Frisoni GB, Nobili F, Scheltens P, et
al. Mild cognitive impairment (MCI) in medical practice: a critical
review of the concept and new diagnostic procedure—report of the
MCI Working Group of the European Consortium on Alzheimer’s
Disease. J Neurol Neurosurg Psychiatry 2006;77:714 – 8.
[42] Grundman M, Petersen RC, Morris JC. ADSC Cooperative Study:
rate of dementia of the Alzheimer type (DAT) in subjects with mild
cognitive impairment. Neurology 1996;46:A403.
[43] Visser PJ, Kester A, Jolles J, Verhey F. Ten-year risk of dementia
in subjects with mild cognitive impairment. Neurology 2006;67:
1201–7.
[44] Chertkow H, Verret L, Bergman H, Wolfson C, McKelvey R.
Predicting progression to dementia in elderly subjects with mild
cognitive impairment. a multidisciplinary approach. Paper presented
at 2001 Contemporary Clinical Issues Plenary Session, 53rd Annual
Meeting American Academy of Neurology, 2001, Philadelphia, PA.
[45] Ritchie K, Leibovici D, Ledesert B, Touchon J. A typology of
sub-clinical senescent cognitive disorder. B J Psychiatry 1996;168:
470 – 6.
[46] Morris JC, Storandt M, Miller JP, McKeel DW, Price JL, Rubin EH,
et al. Mild cognitive impairment represents early-stage Alzheimer
disease. Arch Neurol 2001;58:397– 405.
[47] Luis CA, Loewenstein DA, Acevedo A, Barker WW, Duara R. Mild
cognitive impairment: directions for future research. Neurology
2003;61:438 – 44.
[48] Loewenstein DA, Barker WW, Harwood DG, Luis C, Acevedo A,
Rodriguez I, et al. Utility of a modified Mini-Mental State Exami-
nation with extended delayed recall in screening for mild cognitive
impairment and dementia among community dwelling elders. Int J
Geriatr Psychiatry 2000;15:434 – 40.
[49] Ritchie K, Artero S, Touchon J. Classification criteria for mild cog-
nitive impairment: a population-based validation study. Neurology
2001;56:37– 42.
[50] Ritchie K, Touchon J. Mild cognitive impairment: conceptual basis
and current nosological status. Lancet 2000;355:225– 8.
[51] Chertkow HM, Bergman H, Wolfson C, Babins L, Kelner N. Stan-
dard neuropsychological tests do not predict development of Alz-
heimer’s Disease in indivduals with “Age-associated Cognitive De-
cline”. Can J Neurol Sci 1998;25(Suppl 1):S27.
[52] Craik F. Memory changes in normal aging. Curr Dir Psychol Sci
1994;3:1558.
[53] Artero S, Touchon J, Ritchie K. Disability and mild cognitive
impairment: a longitudinal population-based study. Int J Geriatr
Psychiatry 2001;16:1092–7.
[54] Visser P, Verhey F. Distinction between preclinical Alzheimer’s
disease and depression. J Am Geriatr Soc 2000;48:479 – 84.
[55] Lautenschlager NT, Riemenschneider M, Drzezga A, Kurz AF.
Primary degenerative mild cognitive impairment: study population,
clinical, brain imaging and biochemical findings. Dement Geriatr
Cogn Disord 2001;12:379 – 86.
[56] Bouras C. Evolution of AD pathology in the aging brain. Paper
presented at American Academy of Neurology, 2002, Geneva,
Switzerland.
[57] Markesbery W, Schmitt FA, Kryscio RH, Davis D, Smith CD,
Wekstein D. Neuropatholic substrate of mild cognitive impairment.
Arch Neurol 2006;63:38 – 46.
[58] Knopman DS, Parisi JE, Salviati A, Floriach-Robert M, Boeve
BF, Ivnik RJ, et al. Neuropathology of cognitively normal el-
derly. J Neuropathol Exp Neurol 2003;62:1087–95.
[59] Bennett DA, Schneider JA, Bienias JL, Evans DA, Wilson RS. Mild
cognitive impairment is related to Alzheimer disease pathology and
cerebral infarctions. Neurology 2005;64:83441.
[60] Davis KL, Mohs RC, Marin D, Purohit DP, Perl DP, Lantz M, et al.
Cholinergic markers in elderly patients with early signs of Alzhei-
mer disease. JAMA 1999;281:1401– 6.
[61] DeKosky ST, Ikonomovic M, Cochran EJ, Kordower JH, Mufson
EJ. Cholinergic upregulation in hippocampus in mild cognitive
impairment. relation to Alzheimer Neuropathology. Paper presented
at 54th Annual Meeting of the American Academy of Neurology,
Denver, Colorado, April 13-20, 2002.
[62] Daly E, Zaitchik D, Copeland M, Schmahmann J, Gunther J, Albert
M. Predicting conversion to Alzheimer disease using standardized
clinical information. Arch Neurol 2000;57:675– 80.
[63] Fisk JD, Merry HR, Rockwood K. Variations in case definition
affect prevalence but not outcomes of mild cognitive impairment.
Neurology 2003;61:1179 – 84.
[64] Hsiung G-Y, Johnston D, Donald A, Feldman H. An analysis of
vascular risk factors and ApoE genotype in subtypes of dementia in
the ACCORD study. Neurology 2002;58(Suppl 3):A107.
[65] Jonker C, Schmand B, Lindeboom J, Havekes LM, Launer LJ.
Association between apolipoprotein E epsilon4 and the rate of cog-
nitive decline in community-dwelling elderly individuals with and
without dementia. Arch Neurol 1998;55:1065–9.
[66] Dik MG, Jonker C, Bouter LM, Geerlings MI, van Kamp GJ, Deeg
DJ. APOE-epsilon4 is associated with memory decline in cogni-
tively impaired elderly. Neurology 2000;54:1492–7.
[67] Small BJ, Basun H, Backman L. Three-year changes in cognitive
performance as a function of apolipoprotein E genotype: evidence
from very old adults without dementia. Psychology & Aging 1998;
13:80 –7.
[68] Lehmann M, Gottfries CG, Regland B. Identification of cognitive
impairment in the elderly: homocysteine is an early marker. Dement
Geriatr Cogn Disord 1999;10:12–20.
[69] Corrada M, Kawas C. Plasma total cholesterol, LDL, HDL levels
and the risk of Alzheimer disease: the Baltimore Longitudinal Study
of Aging. Neurology 2002;58:A106.
[70] Yaffe K, Simonsick E, Launer L, Lindquist K, Penninx BW, Pahor
M, et al. Inflammatory markers and cognitive decline in well-
functioning elders: the Health ABC Study. Neurology 2002;58:
A106.
[71] Jack CR, Petersen RC, Xu Y, O’Brien PC, Smith GE, Ivnik RJ, et
al. Rate of medial temporal lobe atrophy in typical aging and
Alzheimers disease. Neurology 1998;51:993–9.
[72] Fox NC. Increased rates of atrophy in early and preclinical AD:
studies with registration of serial MRI. Neurobiol Aging 2000;21:
S74.
[73] Jack CR Jr, Petersen RC, Xu Y, O’Brien PC, Smith GE, Ivnik RJ,
et al. Rates of hippocampal atrophy correlate with change in clinical
status in aging and AD. Neurology 2000;55:484–9.
[74] de Leon MJ, Convit A. Longitudinal measures of hippocampal
formation pathology and CSF measures of phospho-tau and amyloid
beta: early markers for Alzheimer disease. Paper presented at Sev-
enth International Geneva/Springfield symposium on advances in
Alzheimer Therapy, 2002; Geneva, Switzerland.
[75] de Leon MJ, Convit A, Wolf OT, Tarshish CY, DeSanti S, Rusinek
H, et al. Prediction of cognitive decline in normal elderly subjects
with 2-[(18)F]fluoro-2-deoxy-D-glucose/poitron-emission tomogra-
phy (FDG/PET). Proc Natl Acad Sci U S A 2001;98:10966 –71.
[76] Reiman EM, Caselli RJ, Yun LS, Chen K, Bandy D, Minoshima S,
et al. Preclinical evidence of Alzheimer’s disease in persons ho-
mozygous for the epsilon 4 allele for apolipoprotein E. N Engl
J Med 1996;334:752– 8.
[77] McKelvey R, Bergman H, Stern J, Rush C, Zahirney G, Chertkow
H. Lack of prognostic significance of SPECT abnormalities in non-
 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
demented elderly subjects with memory loss. Can J Neurol Sci
1999;26:23– 8.
[78] Johnson KA, Jones K, Holman BL, Becker JA, Spiers PA, Satlin A,
et al. Preclinical prediction of Alzheimer’s disease using SPECT.
Neurology 1998;50:1563–71.
[79] Kantarci K, Jack CR Jr, Xu YC, Campeau NJ, O’Brien PC, Smith
GE, et al. Regional metabolic patterns in mild cognitive impairment
and Alzheimer’s disease: a 1H MRS study. Neurology 2000;55:
210 –7.
[80] Nordberg A. PET activation in MCI and AD patients. Paper pre-
sented at Seventh Annual Springfield/Geneva symposium on ad-
vances in Alzheimer Therapy, 2002, Geneva, Switzerland.
[81] Bookheimer SY, Strojwas M, Cohen MS, Saunders AM, Pericak-
Vance MA, Mazziotta JC, et al. Patterns of brain activation in people
at risk for Alzheimer’s disease. N Engl J Med 2000;343:450 – 6.
[82] Backman L. The cognitive transition from normal aging to AD.
Paper presented at the Seventh Springfield/Geneva symposium on
advances in Alzheimer Therapy, 2002, Geneva, Switzerland.
[83] Bozoki A, Giordani B, Heidebrink J, Berent S, Foster N. Mild
cognitive impairments predict dementia in non-demented elderly
patients with memory loss. Arch Neurol 2001;58:411– 6.
[84] Tierney MC, Szalai JP, Snow WG, Fisher RH. The prediction of
Alzheimer disease: the role of patient and informant perceptions of
cognitive deficits. Arch Neurol 1996;53:423–7.
[85] Devanand DP, Folz M, Gorlyn M, Moeller JR, Stern Y. Question-
able dementia: clinical course and predictors of outcome. J Am
Geriatr Soc 1997;45:321– 8.
[86] Devanand DP, Michaels-Marston KS, Liu X, Pelton GH, Padilla M,
Marder K, et al. Olfactory deficits in patients with mild cognitive
impairment predict Alzheimer’s disease at follow-up. Am J Psychi-
atry 2000;157:1399 – 405.
[87] Albert MS, Moss MB, Tanzi R, Jones K. Preclinical prediction of
AD using neuropsychological tests. J Int Neuropsychol Soc 2001;
7:631–9.
[88] Graves AB, Bowen JD, Rajaram L, McCormick WC, McCurry SM,
Schellenberg GD, et al. Impaired olfaction as a marker for cognitive
decline: interaction with apolipoprotein E epsilon4 status. Neurol-
ogy 1999;53:1480 –7.
[89] Schipper HM, Chertkow H, Mehindate K, Frankel D, Melmed C,
Bergman H. Evaluation of heme oxygenase-1 as a systemic biolog-
ical marker of sporadic AD. Neurology 2000;54:1297–304.
[90] Andreasen N, Minthon L, Vanmechelen E, Vanderstichele H, Dav-
idsson P, Winblad P, et al. Cerebrospinal fluid tau and Abeta42 as
predictors of development of Alzheimer’s disease in patients with
mild cognitive impairment. Neurosci Lett 1999;273:5– 8.
[91] Saxton J, Lopex O, Ratcliff G, Dulberg C, Fried L, Carlson MC,
Kuller L. Preclinical Alzheimer disease: neuropsychological test
performance 1.5 to 8 years prior to onset. Neurology 2004;63:
2341–7.
[92] Frisoni GB, Padovani A, Wahlund LO. The predementia diagnosis
of Alzheimer disease. Alzheimer Dis Assoc Disord 2004;18:51–3.
[93] Bennett DA. Mild cognitive impairment. Clin Geriatr Med 2004;20:
15–25.
[94] Bennett DA, Wilson RS, Schneider JA, Evans DA, Beckett LA,
Aggarwal NT, et al. Natural history of mild cognitive impairment in
older persons. Neurology 2002;59:198 –205.
[95] Dubois B, Albert ML. Amnestic MCI or prodromal Alzheimer’s
disease? Lancet Neurol 2004;3:246 – 8.
[96] Morris J. Mild cognitive impairment is early-stage Alzheimer dis-
ease. Arch Neurol 2006;63:15– 6.
[97] Bennett DA. Update on mild cognitive impairment. Curr Neurol
Neurosci Rep 2003;3:379 – 84.
[98] Ritchie K. Mild cognitive impairment: an epidemiological perspec-
tive. Dialogues in Clinical Neuroscience 2004;6:401– 8.
281
[99] Larrieu S, Letenneur L, Orgogozo JM, Fabrigoule C, Amieva H, Le
Carret N, et al. Incidence and outcome of mild cognitive impairment
in a population-based prospective cohort. Neurology 2002;59:
1594 –9.
[100] Wolf H, Grunwald M, Ecke GM, Zedlick D, Bettin S, Dannenberg
C, et al. The prognosis of mild cognitive impairment in the elderly.
J Neural Transm Suppl 1998;54:31–50.
[101] Chertkow H. Five year follow-up of mild cognitive impairment.
predictors of progression—the 8th International Conference on Alz-
heimer’s Disease and Related Disorders. Stockholm, Sweden, July
2002.
[102] Davis HS, Rockwood K. Conceptualization of mild cognitive im-
pairment: a review. Int J Geriatr Psychiatry 2004;19:313–9.
[103] Petersen RC, Ivnik RJ, Tangalos EG, Kokmen E. Mild cognitive
impairment: clinical characterization and outcome— correction.
Arch Neurol 1999;56:6.
[104] Collie A, Maruff P. An analysis of systems of classifying mild
cognitive impairment in older people. Aust N Z J Psychiatry 2002;
36:133– 40.
[105] Feldman HH, Jacova C. Mild cognitive impairment. Am J Geriatr
Psychiatry 2005;13:645–55.
[106] DeCarli C. Mild cognitive impairment: prevalence, prognosis, aeti-
ology, and treatment. Lancet Neurol 2003;2:15–21.
[107] Salloway S, Ferris S, Kluger A, Goldman R, Griesing T, Kumar D,
et al. Efficacy of donepezil in mild cognitive impairment: a random-
ized placebo-controlled trial. Neurology 2004;63:651–7.
[108] Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins RV,
et al. Current concepts in mild cognitive impairment. Arch Neurol
2001;58:1985–92.
[109] Jelic V, Winblad B. Treatment of mild cognitive impairment: ratio-
nale, present and future strategies. Acta Neurol Scand Suppl 2003;
179:83–93.
[110] Kalbe E, Kessler J, Calabrese P, Smith R, Passmore AP, Brand M,
et al. DemTect: a new, sensitive cognitive screening test to support
the diagnosis of mild cognitive impairment and early dementia. Int
J Geriatr Psychiatry 2004;19:136 – 43.
[111] Ivanoiu A, Adam S, Van der Linden M, Salmon E, Juillerat AC,
Mulligan R, et al. Memory evaluation with a new cued recall test in
patients with mild cognitive impairment and Alzheimer’s disease.
J Neurol 2005;252:47–55.
[112] Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, White-
head V, Collin V, et al. The Montreal Cognitive Assessment,
MoCA: a brief screening tool for mild cognitive impairment. J Am
Geriatr Soc 2005;53:695–9.
[113] Yamamoto S, Mogi N, Umegaki H, Suzuki H, Ando F, Shimokata
H, et al. The clock drawing test as a valid screening method for mild
cognitive impairment. Dement Geriatr Cogn Disord 2004;18:172–9.
[114] Molloy DW. Screening for mild cognitive impairment: comparing
the SMMSE and the ABCS. Can J Psychiatry 2005;50:52– 8.
[115] Xu G, Meyer JS, Huang Y, Chen G, Chowdhury M, Quach M.
Cross-cultural comparison of mild cognitive impairment between
China and USA. Curr Alzheimer Res 2004;1:55– 61.
[116] Artero S, Ritchie K. The detection of mild cognitive impairment in
the general practice setting. Aging Ment Health 2003;7:251– 8.
[117] Tang-Wai DF, Knopman DS, Geda YE, Edland SD, Smith GE,
Ivnik RJ, et al. Comparison of the short test of mental status and the
mini-mental state examination in mild cognitive impairment. Arch
Neurol 2003;60:1777– 81.
[118] Shankle WR, Romney AK, Hara J, Fortier D, Dick MB, Chen JM,
et al. Methods to improve the detection of mild cognitive impair-
ment. Proc Natl Acad Sci U S A 2005;102:4919 –24.
[119] Borson S, Scanlan JM, Watanabe J, Tu SP, Lessig M. Simplifying
detection of cognitive impairment: comparison of the Mini-Cog and
Mini-Mental State Examination in a multiethnic sample. J Am
Geriatr Soc 2005;53:871– 4.
282 H. Chertkow et al. / Alzheimer’s & Dementia 3 (2007) 266 –282
[120] Callahan CM, Unverzagt FW, Hui SL, Perkins AJ, Hendrie HC.
Six-item screener to identify cognitive impairment among
potential subjects for clinical research. Med Care 2002;40:771–
81.
[121] Canning SJ, Leach L, Stuss D, Ngo L, Black SE. Diagnostic utility
of abbreviated fluency measures in Alzheimer disease and vascular
dementia. Neurology 2004;62:556 – 62.
[122] Darby D, Maruff P, Collie A, McStephen M. Mild cognitive im-
pairment can be detected by multiple assessments in a single day.
Neurology. Oct 8 2002;59:1042– 6.
[123] Hanninen T, Hallikainen M, Tuomainen S, Vanhanen M, Soininen
H. Prevalence of mild cognitive impairment: a population-based
study in elderly subjects. Acta Neurol Scand 2002;106:148 –54.
[124] Kabani NJ, Sled JG, Chertkow H. Magnetization transfer ratio in
mild cognitive impairment and dementia of Alzheimer’s type. Neu-
roimage 2002;15:604 –10.
[125] Busse A, Bischkopf J, Riedel-Heller SG, Angermeyer MC. Subclas-
sifications for mild cognitive impairment: prevalence and predictive
validity. Psychol Med 2003;33:1029 –38.
[126] Lambon Ralph MA, Patterson K, Graham N, Dawson K, Hodges JR.
Homogeneity and heterogeneity in mild cognitive impairment and
Alzheimer’s disease: a cross-sectional and longitudinal study of 55
cases. Brain 2003;126:2350 – 62.
[127] Lopez OL, Jagust WJ, DeKosky ST, Becker JT, Fitzpatrick A,
Dulberg C, et al. Prevalence and classification of mild cognitive
impairment in the Cardiovascular Health Study Cognition Study:
part 1. Arch Neurol 2003;60:1385–9.
[128] Farlow MR, He Y, Tekin S, Xu J, Lane R, Charles HC. Impact of
APOE in mild cognitive impairment. Neurology. Nov 23 2004;
63(10):1898 –1901.
[129] Feldman H, Scheltens P, Scarpini E, Hermann N, Mesenbrink P,
Mancione L, et al. Behavioral symptoms in mild cognitive impair-
ment. Neurology 2004;62:1199 –201.
[130] Grundman M, Petersen RC, Ferris SH, Thomas RG, Aisen PS,
Bennett DA, et al. Mild cognitive impairment can be distinguished
from Alzheimer disease and normal aging for clinical trials. Arch
Neurol 2004;61:59 – 66.
[131] Solfrizzi V, Panza F, Colacicco AM, D’Introno A, Capurso C,
Torres F, et al. Vascular risk factors, incidence of MCI, and rates of
progression to dementia. Neurology 2004;63:1882–91.
[132] de Jager CA, Budge MM. Stability and predictability of the classi-
fication of mild cognitive impairment as assessed by episodic mem-
ory test performance over time. Neurocase 2005;11:72–9.
[133] Devanand DP, Pelton GH, Zamora D, Liu X, Tabert MH, Goodkind
M, et al. Predictive utility of apolipoprotein E genotype for Alzhei-
mer disease in outpatients with mild cognitive impairment. Arch
Neurol 2005;62:975– 80.
[134] Geslani DM, Tierney MC, Herrmann N, Szalai JP. Mild cognitive
impairment: an operational definition and its conversion rate to
Alzheimer’s disease. Dement Geriatr Cogn Disord 2005;19(5-6):
383–389.
[135] Kumar R, Dear KB, Christensen H, Ilschner S, Jorm AF, Meslin
C, et al. Prevalence of mild cognitive impairment in 60- to
64-year-old community-dwelling individuals: The Personality
and Total Health through Life 60⫹ Study. Dement Geriatr Cogn
Disord 2005;19:67–74.
[136] Purser JL, Fillenbaum GG, Pieper CF, Wallace RB. Mild cognitive
impairment and 10-year trajectories of disability in the Iowa Estab-
lished Populations for Epidemiologic Studies of the Elderly cohort.
J Am Geriatr So 2005;53:1966 –72.
[137] Visser PJ, Scheltens P, Verhey FR. Do MCI criteria in drug trials
accurately identify subjects with predementia Alzheimer’s disease?
J Neurol Neurosurg Psychiatry 2005;76:1348 –54.
[138] Verghese J, Levalley A, Derby C, Kuslansky G, Katz M, Hall C,
et al. Leisure activities and the risk of amnestic mild cognitive
impairment in the elderly. Neurology 2006;66:821.
 Alzheimer’s & Dementia 3 (2007) 283–291

Mild cognitive impairment and cognitive impairment, no dementia:

Part B, therapy
Fadi Massouda,**, Sylvie Bellevilleb,c, John Kirkd, Howard Chertkowb,d,e,f,*,
Ziad Nasreddineg, Yves Joanetteb,h, Howard Bergmand,e, Morris Freedmani
a
Service de Gériatrie, Centre Hospitalier de l’Université de Montréal, and Département de Médecine, Université de Montréal, Montreal, Quebec, Canada
b
Centre de recherche, Institut Universitaire de Gériatrie de Montréal, Montreal, Quebec, Canada
c
Department de Psychologie, Université de Montréal, Montreal, Quebec, Canada
d
Division of Geriatric Medicine, Department of Medicine, Sir Mortimer B. Davis-Jewish General Hospital, McGill University, Montreal, Quebec, Canada
e
Bloomfield Centre for Research in Aging, Lady Davis Institute for Medical Research, Sir Mortimer B. Davis-Jewish General Hospital, McGill
University, Montreal, Quebec, Canada
f
Department of Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada
g
Département de médecine, service de neurologie, Hôpital Charles LeMoyne and Université de Sherbrooke, Montreal, Quebec, Canada
h
Faculté de Médicine, Université de Montréal, Montreal, Quebec, Canada
i
Department of Medicine (Neurology), University of Toronto, and Behavioural Neurology Program, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada

Abstract Mild cognitive impairment (MCI) and cognitive impairment, no dementia (CIND) might be the
optimum stage at which to intervene with preventative therapies. This article reviews recent work
on the possible treatment and presents evidence-based recommendations approved at the meeting of
the Third Consensus Conference on the Diagnosis and Treatment of Dementia held in Montreal in
March 2006. A number of promising nonpharmacologic interventions have been examined. Asso-
ciations exist with both cognitive and physical activity that suggest that both of these, together or
separately, can delay progression to dementia. Similarly, case control studies as well as prospective
long-term studies suggest a number of low toxicity interventions and supplements that might
significantly impact on MCI progression; folate, B6, and B12 to lower homocysteine levels,
omega-fatty acids, and anti-oxidants (fruit juices or red wine) are good examples. In selected
genotypes such as individuals with APOE e4, therapy with donepezil might slow progression. The
concern, however, is that none of these therapies (including cholinesterase inhibitors) have dem-
onstrated a clinically meaningful effect with randomized, placebo-controlled studies. Just as ran-
domized controlled studies have failed to support primary prevention of dementia by using estrogen
or nonsteroidal anti-inflammatory drugs (NSAIDs), there exists the possibility that well-designed
randomized controlled trials might fail to definitively demonstrate putative or promising mild
cognitive impairment interventions. Pharmacologic interventions and nonpharmacologic therapies,
while tantalizing, are currently for the most part insufficiently proven to allow serious consideration
by physicians. Recommendation were supported for a general “healthy lifestyle” including physical
exercise, healthy nutrition, smoking cessation, and mental stimulation. Close monitoring and
treatment of vascular risk factors are justified and were also supported.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: MCI; CIND; Cognitive impairment; Psychosocial intervention; Cognitive intervention; Exercise

*Corresponding author. Tel.: (514) 340-8260; Fax: (514) 340-8925.

E-mail address: howard.chertkow@mcgill.ca
**Additional Corresponding author. Tel.: (514) 890-8000 26769; Fax: (514) 412-7506.
E-mail address: fadimassoud@videotron.ca

1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.002
284 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291
1. Treatment for mild cognitive impairment
The preceding article reviewed the concepts and diagno-
sis of mild cognitive impairment (MCI) and cognitive im-
pairment, no dementia (CIND), reviewed our approach to an
evidence-based review of the literature, and presented the
recommendations that received consensus at the Third Con-
sensus Conference on the Diagnosis and Treatment of De-
mentia (CCCDTD3) held in Montreal in March 2006.This
second part will use the same approach to review therapy
for MCI and CIND. The treatment of MCI has been the
subject of a number of recent chapters and reviews [1,2].
There have been relatively few randomized controlled trials
of any therapy sufficient to rank as Level 1 evidence. Nev-
ertheless, there are a number of potential interventions both
pharmacologic and nonpharmacologic that deserve to be
addressed.
2. Nonpharmacologic treatment for MCI
2.1. Cognitive intervention in MCI
There have been several relevant studies in this area. We
will first review those carried out in normal elderly and then
in MCI. A full listing of these studies is seen in Table 1.
Longitudinal cohort studies of healthy elderly persons show
that engagement in stimulating cognitive activities (engaged
lifestyle; novel and intellectually challenging activities) is
associated with better memory and verbal abilities [3]. In a
case-control study, participation in intellectually stimulating
and social activities in midlife has been associated with
reduced risk of developing Alzheimer’s disease (AD) [4]. In
a longitudinal cohort study of healthy elderly persons (av-
erage follow-up, 4.5 years), a participant’s frequency of
participation in common cognitive activities at baseline was
associated with reduced risk of clinical diagnosis of AD and
reduced cognitive decline (annualized change on global
cognition, working memory, and perceptual speed) during
the follow-up period [5]. However, the methodology in
these studies is based on association; therefore, the direction
of causality remains to be clarified. For example, it is
unclear whether cognitive activities have a protective effect
on the development of cognitive deficits in aging, or
whether reduced engagement in cognitive activities is an
early sign of AD.
Verhaeghen et al [6] conducted a quantitative review of
studies measuring the efficacy of memory intervention stud-
ies in healthy aging. They reported that memory training
improved performance on targeted memory tasks and that
the effect sizes for the training effect were in the moderate
range. One large scale randomized controlled trial on cog-
nitive interventions (memory, speed, or reasoning vs no
training) was completed in a sample of 2,832 healthy older
adults [7]. The results indicated improved performance after
training on the cognitive domains that were targeted by the
interventions. The positive effects were sustained during a
2-year follow-up, and the effect sizes were moderate to
large. Thus there is good evidence that cognitive training
increases cognitive efficacy on target measures in healthy
older adults.
Two nonrandomized studies and two randomized con-
trolled trials (RCTs) have been reported on the effect of
cognitive training in MCI. With an RCT design, Olazaran
et al [8] reported decreased depression and improved cog-
nition (cognitive subscale of the Alzheimer’s Disease As-
sessment Scale, cognitive portion) in a mixed group com-
prising 72 AD and 12 MCI patients after a 1-year program
of cognitive-motor stimulation plus psychosocial compared
with psychosocial support. In a small-scale (n ⫽ 18) RCT
trial, Rapp et al [9] compared cognitive intervention with no
treatment. They reported improved subjective memory and
long-term maintenance during a 6-month period but no
effect on objective tests of memory. Gunther et al [10]
reported long-term improvement in cognitive performance
(working memory and verbal episodic memory) in a pre-
post comparison study of computer-assisted cognitive train-
ing in persons with MCI. Finally, Belleville et al [11]
compared the effect of a multifactorial memory training
program with a no-training condition (28 MCI participants)
and reported larger memory improvement on post-test in the
trained MCI participants compared with the untrained ones.
Moderate to large effect sizes were obtained for the training
effect on target episodic memory measures. Thus, studies
investigating the effect of cognitive intervention in MCI
provide encouraging findings. However, the effort required
to implement such therapeutic measures is not trivial, and
large scale cognitive intervention for MCI would require
considerable resources. Before widespread recommendation
of this therapy can occur, more replication studies are re-
quired with properly controlled RCT designs, larger sample
sizes, and analyses that control for type 1 error.
In conclusion, longitudinal cohort studies of healthy
older adults indicate that engagement in intellectually stim-
ulating activities is associated with decreased risk of AD
and decreased cognitive decline. However, the evidence at
the present time is insufficient to conclude that organized
cognitive intervention is beneficial to preventing progres-
sion in MCI or warrants prescription. On the other hand,
given that there is little or no “downside” to cognitive
activity, it is not unreasonable for physicians and therapists
to promote engagement in cognitive activity as part of an
overall healthy lifestyle formulation for elderly individuals
with and without memory loss.
2.2. Physical training in MCI
Several longitudinal cohort studies carried out in nor-
mal elderly individuals indicated that physical exercise is
associated with reduced cognitive decline and reduced
risk of dementia. These studies looked at outcomes such
as a change score on the Mini-Mental State Examination
Table 1
Studies of non-pharmacological treatment of MCI
Authors Year Type Population Outcome Measures Results Notes

Physical training
Colcombe and 2003 Meta-analysis (18 RCTs) Healthy and clinical aged Cognitive tests ES, .478; largest on executive Not MCI; controlled
Kramer [15] intervention: ??
Heyn et al [16] 2004 Meta-analysis (30 RCTs) Demented Physical, functional, behavioral, ES, .62; large effect on most Heterogeneous group; 9 with
and cognitive measures components mild impairment; controlled
intervention: ??
Kramer et al [13] 2004 Literature review of Healthy aged Cognitive tests Improves executive functions
RCT and prospective
studies
Lytle et al [12] 2004 Prospective study Community sample Decline on MMSE over a Protects against severe
2-year time decline

F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291

Laurin et al [13] 2001 Prospective Community sample Cognitive impairment Physical activity reduces risk
of cognitive impairment,
AD, and dementia 5 y later
Psychosocial
intervention
Ishizaki et al [45] 2002 Prospective study Mild AD (CDR .05) and Improvement on cognition Small N; a specific intervention
controls and affective measures (eg, reminiscence); no placebo
Cognitive
intervention
Kramer et al [3] 2004 Literature review of Healthy aging Cognitive tests Improves targeted cognitive
RCTs and prospective functions (cognition)
studies; cognitive
intervention and effect
of leisure and
environment
Wilson et al [5] 2002 Prospective study; Longitudinal cohort Diagnosis of AD after a mean Reduces the risk of AD and
cognitively stimulating follow-up of 4.5 yrs the decline on global
activities cognition, WM and speed.
Lindstrom et al [4] 2005 Retrospective study of AD and controls Diagnosis of AD TV viewing increases risk of Limitations: association direction
leisure activity (TV AD; intellectual and social unclear; retrospective; based
viewing in midlife) activities reduce it on questionnaire in surrogate
respondents
Verhagen et al [6] 1992 Literature review of Healthy aging Memory Improved performance (ES, Negative effect of lower mental
intervention studies 0.7; 0.2 in controls) status
Ball et al [7] 2002 RCT Healthy aging Cognition Improves targeted cognitive High quality study; follow-up to
functions; long-standing appear soon
effect (2 years); no
generalisation.
Poon et al [46] 2005 RCT of videoconference MCI and early AD Cognition Equivalent effect of VC and Small N; no placebo
of cognitive and Face to face training
psychosocial
intervention
Olazaran et al [8] 2005 RCT of cognitive MCI and AD Cognition, depression, Positive effect on cognitive No distinction between MCI and
stimulation (1yr functional impact and depression AD; no control of type 1 error
program)

285
286 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291

(MMSE) 2 years later [12] and risk of dementia 5 years later

Not randomized; no treatment as

No-treatment group as control;
[13]. However, there are also studies that failed to find a

control; small N (26 MCI)

protective effect of physical exercise on cognitive decline

No group of untrained
and on incident dementia [14]. Two recent meta-analyses

small N (9 MCI)
have been published regarding the impact of physical exer-
cise programs on the cognitive function of older adults

participants
[15,16]. Colcombe and Kramer [15] included 18 RCT stud-
ies in their meta-analysis, and Heyn et al [16] included 30
Notes

studies. Both meta-analyses reported moderate effect sizes

for the exercise training effect on global cognitive scores,
with a larger effect on tasks measuring executive control
maintenance but no effect
Positive effect on subjective

Positive effect on cognitive

reported by Colcombe and Kramer. There are immense

measure and long-term

on objective measures

implications of such research in terms of potential public

Pre-post training effect

health measures to prevent dementia and cognitive decline.

For fitness: more studies are needed to assess modality, safety, intensity. Problems: blinding, small N; long-term; appropriate control; to target MCI.
and well-being

However, more studies, particularly RCTs, are needed to

assess the optimal exercise training modalities in older
adults, particularly in terms of intensity and duration. Safety
Results

issues will also need to be better addressed. Furthermore,

there were methodologic limitations in past studies regard-
ing appropriate blinding procedures and sample size. Fi-
nally, no studies have been carried out specifically with
MCI persons to assess the effect of physical training on their
Cognition and well-being

cognitive capacities and cognitive decline.

In conclusion, longitudinal cohort studies of healthy
Outcome Measures

For psychosocial intervention: more studies are needed in MCI to assess potential efficacy, feasibility, safety.

older adults indicate that physical training might be protec-

tive against cognitive decline. RCTs have provided compel-
For cognitive intervention: encouraging results but more RCT studies are needed in MCI with larger N.
Cognition

Cognition

ling evidence for a beneficial effect of physical training on

the cognitive performance of healthy older adults. However,
the evidence is insufficient to conclude that physical train-
ing is beneficial to preventing progression in MCI or war-
rants prescription. Furthermore, safety issues will need to be
better addressed. Keeping this in mind, physicians and ther-
MCI and controls

apists might promote physical activity at an intensity level

that is adapted to the person’s overall physical capacities as
Population

part of a healthy lifestyle for older individuals with and

without memory loss.
MCI

MCI

2.3. Psychosocial intervention

Clinical trial of cognitive

cognitive intervention

There are few data in this area, and no suggestions will

be reviewed at this time.
Pre-post study of
RCT of memory
intervention

intervention

2.4. Cognitive intervention in MCI

Type

2.4.1. Recommendation 7
The evidence at the present time is insufficient to con-
clude that organized cognitive intervention is beneficial to
2002

2006

2003
Year

preventing progression in MCI or warrants prescription

(Grade C, Level 1).
Gunther et al [10]

2.4.2. Recommendation 8
Rapp et al [9]

et al [11]

There is fair evidence that physicians and therapists

Belleville
Continued

should promote engagement in cognitive activity as part of

Authors
Table 1

an overall healthy lifestyle formulation for elderly individ-

uals with and without memory loss (Grade B, Level 2).
 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291 287

Table 2 suggest early cholinergic “functional” modifications such as

Drugs under investigation to prevent progression to AD choline transport and acetylcholine release at this stage [18].
CHEIs All three cholinesterase inhibitors (ChEIs) currently avail-
Anti-inflammatories able in Canada have been used in clinical trials of MCI.
Estrogen
Statins
Salloway et al [19] conducted a 24-week, multicenter,
Cholesterol-lowering drugs randomized, double-blind placebo-controlled study in 270
Anti-amyloid drugs (beta-secretase inhibitors, gamma-secretase inhibitors, individuals with MCI, comparing donepezil 10 mg daily
GAG mimetics, amyloid immunotherapy) with placebo. There were no significant beneficial effects on
Various anti-oxidants including vitamin E the two primary outcome measures used, the New York
AMPAkines
University Paragraph Delayed Recall test and the Alzheimer
Disease Cooperative Study Clinician’s Global Impression
of Change for MCI (ADCS-CGIC-MCI). Subjects on done-
2.5. Physical training as therapy in MCI pezil exhibited statistically significant benefit on two a pri-
2.5.1. Recommendation 9 ori secondary outcome measures, the Alzheimer’s Disease
There is fair evidence that physicians and therapists Assessment Scale– cognitive subscale (ADAS-cog) and the
should promote physical activity at an intensity level that is Patient Global Assessment (PGA). More individuals treated
adapted to the person’s overall physical capacities as part of with donepezil also had adverse events.
a healthy lifestyle for older individuals with and without The Alzheimer’s Disease Cooperative Study Group
memory loss (Grade B, Level 2). (ADCS) led by Petersen conducted a 3-year double-blind
study randomizing individuals with amnestic MCI to vita-
2.5.2. Recommendation 10 min E 2000 IU daily, donepezil 10 mg daily, and placebo
Current evidence is insufficient to conclude that a spe- [20]. In this study, the primary outcome was conversion to
cific program of physical training warrants prescription in possible or probable AD. At the end of the study, there were
MCI patients to prevent progression to dementia (Grade C, no differences in the progression to AD between the three
Level 3). groups. However, on preplanned interim analyses, donepezil-
treated individuals had a slightly reduced likelihood of pro-
3. Pharmacologic approaches to MCI gression to AD during the first year of the study. On a priori
pharmacogenetic analyses, the study confirmed that posses-
Several pharmacologic treatment approaches have been sion of the APOE e4 allele is a major predictor of progres-
attempted in MCI (Table 2). The formal review of the sion to AD and indicated that most of the treatment effect
evidence is presented in Table 3. occurred among the APOE e4 carriers. On secondary anal-
ysis confined to APOE e4 carriers, the effect of donepezil
3.1. Cholinesterase inhibitors
was significant at 12, 24, and 36 months. “In fact, a closer
Although evidence suggests the absence of significant look at their data reveals no convincing evidence of a
cholinergic deficits in MCI [17], several lines of evidence difference in treatment effect according to APOE e4 carrier

Table 3
Standard evidence table for pharmacologic therapy in MCI
Agent Design Author, Year N Statistical Significance Quality Magnitude Absolute Numbers Comments
of of Benefit Risk Needed
Study* Reduction to Treat

Donepezil RCT Salloway et al [19], 2004 270 On secondary outcome 4/5 NA NA NA

measures
Donepezil RCT Petersen et al [20], 2005 769 None in primary outcome; 4/5 NA NA NA
some positive
secondary analyses
Rivastigmine RCT Feldman et al [23], 2007 1018 No 5/5 NA NA NA
(InDDEX)
Galantamine RCT Scheltens et al [24], 2004
Rofecoxib RCT Thal et al [26], 2005 1457 No 5/5 NA NA NA
Estrogen RCT Shumaker et al [28], 2003 4532 Increased risk of dementia 5/5 NA NA NA
(hazard ratio, 2.1)in
treatment group
Ginkgo biloba RCT van Dongen et al [30], 2003 214 No 5/5 NA NA NA

* Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. The quality of reports of randomized clinical trials: is blinding
necessary? Control Clin Trials 1996;1:1–12.
288 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291
status; the effect of donepezil was similar among APOE e4
carriers and noncarriers (hazard ratio for progression of 0.66
as compared with 0.80 for the entire cohort, with overlap-
ping confidence intervals). These numbers suggest that the
observed difference in significance might have been due to
analogous differences in statistical power, because AD de-
veloped in about twice as many APOE e4 carriers (who are
more likely to be on a course toward AD) as noncarriers
within the 3 years after enrolment” [21]. Also, the authors
indicate that “ . . . there are insufficient data to warrant rec-
ommending APOE e4 genotyping in persons with mild
cognitive impairment, and our results cannot be used to
make this recommendation, since the study was not statis-
tically powered to determine the effects of treatment in
separate groups of APOE e4 carriers and noncarriers.”
Results from a 4-year double-blind, placebo-controlled,
clinical trial of rivastigmine in MCI individuals (InDDEX)
showed no difference in the likelihood of progression to AD
between the treatment group and those on placebo [22]. At
the time of the CCCDTD3, preliminary data had been pub-
lished in abstract format only. Although the final published
form of the study came out after the CCCDTD3 voting [23],
there was no meaningful change in the recommendations
based on this study. The overall rate of progression from
MCI to AD in this randomized clinical trial was much lower
than predicted.
Two clinical trials with galantamine in MCI similarly
showed no difference in the probability of conversion to
AD. However, there was a reduced rate of whole-brain
atrophy in patients treated with galantamine [24]. Therapy
with galantamine was associated with an increased all-cause
mortality risk that might be explained, at least in part, by the
very low mortality rate in the placebo groups. These results
are published as abstracts only; therefore, there is little
information concerning the methodology.
3.1.1. Recommendation 11
There is currently insufficient evidence to recommend
for the use of ChEIs in MCI (Grade C, Level 1).
3.2. Anti-inflammatory agents
Several large well-conducted epidemiologic studies have
shown a negative association between the use of NSAIDs
and the development of AD [25]. In the only clinical trial of
NSAIDs in MCI, Thal et al [26] randomized patients to rofe-
coxib 25 mg and placebo daily for up to 4 years. Individuals on
rofecoxib did not exhibit significant benefit in terms of clinical
conversion to clinical AD, which was the primary outcome
measure, or on secondary outcome measures.
3.2.1. Recommendation 12
There is currently fair evidence to recommend against
the use of NSAIDs in MCI (Grade D, Level 1).
3.3. Hormone replacement therapy
Estrogen has been shown to have neuroprotective effects
in several studies. It is considered a neurotrophic factor for
cholinergic neurons; it stimulates cerebral cellular growth,
contributes to synaptic reorganization, and decreases beta-
amyloid production. It is also associated with antioxidant
properties and beneficial vascular effects [27]. In a meta-
analysis of 12 observational studies, estrogen is associated
with a decreased risk of developing AD, with an odds ratio
of 0.69 [27]. In the Women’s Health Initiative Memory
Study, 4,532 women were randomized to estrogen 0.625 mg
daily with medroxyprogesterone 2.5 mg daily and placebo
and were followed up between 3 and 4 years [28]. Not only
did this study fail to demonstrate any benefit on the risk of
developing MCI and dementia, it actually showed a dou-
bling (hazard ratio, 2.05) of the risk of dementia in the
hormone-treated group.
In the Multiple Outcomes of Raloxifene Evaluation
(MORE) study, postmenopausal women were randomized
to a selective estrogen receptor modulator, raloxifene, at
doses of 60 mg and 120 mg daily and placebo [29]. At
3-year follow-up, women taking 120 mg of raloxifene had a
33% decreased risk of developing MCI. This benefit was not
observed in the group taking raloxifene 60 mg daily.
3.3.1. Recommendation 13
There is currently fair evidence to recommend against
the use of estrogen replacement therapy in MCI (Grade D,
Level 1).
3.4. Ginkgo biloba
Ginkgo biloba is commonly used for memory impair-
ment because of its presumably beneficial antioxidant, anti-
platelet, and neuroprotective properties. The only relevant
randomized, placebo-controlled clinical trial of ginkgo bi-
loba was conducted by van Dongen et al [30] in patients
with dementia and Age-Associated Memory Impairment
(AAMI). This category includes individuals with memory
impairment when compared with young controls. At the end
of the 36-week-long trial, there were no statistically signif-
icant differences between the ginkgo biloba and placebo
groups. A Cochrane Database systematic review of ginkgo
biloba for cognitive impairment concludes that the three
most recent clinical trials, which are methodologically more
sound, show inconsistent results [31]. They add that “there
is a need for a large trial using modern methodology . . .”.
3.4.1. Recommendation 14
There is currently fair evidence to recommend against
the use of ginkgo biloba in MCI (Grade D, Level 1).
3.5. Vitamin E
Epidemiologic data suggest a negative association be-
tween ingestion of vitamin E from natural [32] and artificial
sources [33] and developing AD. The potential neurocog-
 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291
nitive benefit of vitamin E was further supported by a
randomized clinical trial showing it delayed reaching sig-
nificant clinical milestones in patients with AD [34]. The
only trial of vitamin E in MCI, discussed in the ChEI
section, was negative [20]. Concerns have been raised about
the safety of vitamin E since the publication of a meta-
analysis of 19 clinical trials that suggested an increased
all-cause mortality in individuals ingesting 400 or more IU
daily [35].
3.5.1. Recommendation 15
There is currently fair evidence to recommend against
the use of vitamin E in MCI (Grade D, Level 1).
3.6. Treatment of vascular risk factors
Several epidemiologic studies have shown the associa-
tion between vascular risk factors and cognitive impairment
[36 –39]. Clinical trials have evaluated the role of primary
prevention of vascular risk factors and the risk of cognitive
deterioration. In the Systolic Hypertension in EURope Trial
(SYST-EUR), 2,418 subjects were randomized to active
treatment with nitrendipine with the possible addition of
enalapril and hydrochlorothiazide or placebo [40]. In the
active treatment group, the incidence of dementia was re-
duced by 50% at 2 years. In the Perindopril Protection
Against Recurrent Stroke Study (PROGRESS) 6,015 pa-
tients with previous stroke or transient ischemic attack were
randomized to receive perindopril with or without indapam-
ide or placebo [41]. At the end of the 3.9-year follow-up, the
risk of dementia and cognitive decline was significantly
reduced in the active treatment group. The risk was reduced
even further in subjects with recurring stroke. The two large
primary prevention trials with statins have reached some-
what different conclusions; statins do reduce cardiovascular
and cerebrovascular events in primary prevention, but they
do not provide any benefit on cognitive function [42,43].
These seemingly counterintuitive results can be explained
by some of the methodologic limitations of these trials,
namely the relatively short duration of follow-up and the
limited, and probably incomplete, evaluation of cognitive
function. One clinical trial evaluated the role of hyperten-
sion treatment in individuals with mild cognitive deficits
broadly defined as a MMSE score between 20 and 28 [44].
In this study, captopril was compared with bendrofluazide
in 81 subjects with systolodiastolic hypertension for 24
weeks. There were no statistically significant differences be-
tween the two drugs, but patients with the best response to
treatment in terms of reduction of their diastolic blood pressure
significantly improved on two cognitive tests. Interestingly,
this study suggests that reduction of blood pressure in elderly
hypertensive patients with MCI is not hazardous.
3.6.1. Recommendation 16
Because vascular risk factors and comorbidities impact
on the development and expression of dementia, they
289
should be screened for and treated optimally in MCI (Grade
B, Level 2).
Other pharmacologic agents that have been considered in
MCI have shown either marginal or inconsistent benefit.
These include neurotrophic medications (phosphatidylser-
ine, acetyl-l-carnitine, piracetam), memory stimulants (am-
pakines, N-methyl-D-aspartate modulators, CREB modula-
tors), antioxidants, vitamin supplements, and omega fatty
acids.
4. Pharmacologic treatment recommendations for
MCI (summary)
4.1. Recommendation 11
There is currently insufficient evidence to recommend
for the use of ChEIs in MCI (Grade C, Level 1).
4.2. Recommendation 12
There is currently fair evidence to recommend against
the use of NSAIDs in MCI (Grade D, Level 1).
4.3. Recommendation 13
There is currently fair evidence to recommend against
the use of estrogen replacement therapy in MCI (Grade D,
Level 1).
4.4. Recommendation 14
There is currently fair evidence to recommend against
the use of ginkgo biloba in MCI (Grade D, Level 1).
4.5. Recommendation 15
There is currently fair evidence to recommend against
the use of vitamin E in MCI (Grade D, Level 1).
4.6. Recommendation 16
Because vascular risk factors and comorbidities impact
on the development and expression of dementia, they
should be screened for and treated optimally in MCI (Grade
B, Level 2).
In conclusion, an evidence-based review of the literature
provides evidence that leisure activities, cognitive stimula-
tion, and physical activity should be promoted as part of a
healthy lifestyle in elderly individuals and those with MCI.
Vascular risk factors should be treated optimally in these
individuals as well. No other specific therapies can yet be
recommended or supported as having adequate demonstra-
tion of efficacy. There is certainly a need for larger scale
trials to demonstrate efficacy of some of the pharmacologic
interventions, and our recommendations might have to be
altered in the future as such trials are completed.
290 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291
Author Disclosures
Fadi Massoud has had associations with Janssen Ortho
Inc. (Advisory Board, CME Speaker, Research Funds),
Lundbeck Canada (Advisory Board, CME Speaker, Re-
search Funds), Novartis (Advisory Board, CME Speaker,
Research Funds), and Pfizer (Advisory Board, CME
Speaker, Research Funds).
Howard Chertkow has had associations with Pfizer Can-
ada (Advisory Board, Speaker, Grant Recipient), Neuro-
chem Inc. (Advisory Board), and Lundbeck Canada (Advi-
sory Board, Speaker).
Ziad Nasreddine has had associations with Pfizer (Con-
sultant, Adviser, Program Organizer, Lecturer), Novartis
(Lecturer, Drug Trials, Advisor), Janssen Ortho Inc. (Lec-
turer, Drug Trials, Advisor), Neurochem (Drug Trial), Myr-
iad (Drug Trial), and Sanofi-Aventis (Drug Trial).
Howard Bergman has had associations with Pfizer (Drug
Trial) and Novartis (Drug Trial).
Morris Freedman has had associations with Pfizer (Ad-
visory Board, CME Speaker, Research Funds), Janssen-
Ortho Inc. (Advisory Board, CME Speaker), Novartis (Ad-
visory Board, CME Speaker), Lundbeck (CME Speaker),
Eli Lilly (Research Funds), and Orphan Medical (Research
Funds).
Acknowledgments
This work was supported by grants from a number of
agencies. Howard Chertkow is supported by operating
grants from the CIHR (Canadian Institutes for Health Re-
search), the Alzheimer Society of Canada, and the Fonds de
la recherche en santé du Québec (FRSQ).
Howard Chertkow and Sylvie Belleville receive “cher-
cheur national” awards from the FRSQ (Fonds de la Re-
cherche en Santé du Québec).
Howard Bergman is supported through the Dr Joseph Kauf-
mann Chair in Geriatric Medicine at McGill University.
Yves Joanette was supported by a grant from the Cana-
dian Institutes of Health Research (MOP-15006).
Sylvie Belleville is supported by research grants from the
Canadian Institutes of Health Research (CIHR), Heart and
Stroke Foundation of Canada, Alzheimer’s Society of Can-
ada, CAREC, and NSERC.
References
[1] Chertkow H. Mild cognitive Impairment and very early stage Alz-
heimer’s disease. In: Gauthier S, ed. Clinical diagnosis and manage-
ment of Alzheimer’s disease, third edition. London: Informa UK;
2006. p. 205–21.
[2] Chertkow H. Emerging pharmacological therapies for mild cognitive
impairment. In: Hultsch IHTaD, ed. Mild cognitive impairment: inter-
national perspectives. New York: Taylor & Francis; 2006. p. 217– 43.
[3] Kramer AF, Bherer L, Colcombe SJ, Dong W, Greenough WT.
Environmental influences on cognitive and brain plasticity during
aging. J Gerontol A Biol Sci Med Sci 2004;59:M940 –57.
[4] Lindstrom HA, Fritsch T, Petot G, Smyth KA, Chen CH, Debanne
SM, et al. The relationships between television viewing in midlife and
the development of Alzheimer’s disease in a case-control study. Brain
Cogn 2005;58:157– 65.
[5] Wilson RS, Barnes LL, Mendes de Leon CF, Aggarwal NT, Schnei-
der JS, Bach J, et al. Participation in cognitively stimulating activities
and risk of incident Alzheimer disease. JAMA 2002;287:742– 8.
[6] Verhaeghen P, Marcoen A, Goossens L. Improving memory perfor-
mance in the aged through mnemonic training: a meta-analytic study.
Psychol Aging 1992;7:242–51.
[7] Ball K, Berch DB, Helmers KF, Jobe JB, Leveck MD, Marsike M, et
al. Effects of cognitive training interventions with older adults: a
randomized controlled trial. JAMA 2002;288:2271– 81.
[8] Olazarán J, Muniz R, Resiberg B, Pena-Casanova J, del Ser T,
Cruz-Jentoft AJ, et al. Benefits of cognitive-motor intervention in
MCI and mild to moderate Alzheimer disease. Neurology 2004;63:
2348 –53.
[9] Rapp S, Brenes G, Marsh AP. Memory enhancement training for
older adults with mild cognitive impairment: a preliminary study.
Aging Ment Health 2002;6:5–11.
[10] Gunther VK, Schafer P, Holzner BJ, Kemmler GW. Long-term im-
provements in cognitive performance through computer-assisted cog-
nitive training: a pilot study in a residential home for older people.
Aging Ment Health 2003;7:200 – 6.
[11] Belleville S, Gilbert B, Fontaine F, Gagnon L, Menard E, Gauthier S.
Improvement of episodic memory in persons with mild cognitive
impairment and healthy older adults: evidence from a cognitive in-
tervention program. Dement Geriatr Cogn Disord 2006;22:486 –99.
[12] Lytle ME, Vander Bilt J, Pandav RS, Dodge HH, Ganguli M. Exer-
cise level and cognitive decline: the MoVIES project. Alzheimer Dis
Assoc Disord 2004;18:57– 64.
[13] Laurin D, Verreault R, Lindsay J, MacPherson K, Rockwood K.
Physical activity and risk of cognitive impairment and dementia in
elderly persons. Arch Neurol 2001;58:498 –504.
[14] Wilson RS, Beckett LA, Barnes LL, Schneider JA, Bach J, Evans
DA, et al. Individual differences in rates of change in cognitive
abilities of older persons. Psychol Aging 2002;17:179 –93.
[15] Colcombe S, Kramer AF. Fitness effects on the cognitive function of
older adults: a meta-analytic study. Psychol Sci 2003;14:125–30.
[16] Heyn P, Abreu BC, Ottenbacher KJ. The effects of exercise training
on elderly persons with cognitive impairment and dementia: a meta-
analysis. Arch Phys Med Rehabil 2004;85:1694 –704.
[17] DeKosky ST, Ikonomovic MD, Styren SD, Beckett L, Wisniewski S,
Bennett DA, et al. Upregulation of choline acetyltransferase activity
in hippocampus and frontal cortex of elderly subjects with mild
cognitive impairment. Ann Neurol 2002;51:145–55.
[18] Terry AV Jr, Buccafusco JJ. The cholinergic hypothesis of age and
Alzheimer’s disease-related cognitive deficits: recent challenges and
their implications for novel drug development. J Pharmacol Exp Ther
2003;306:821–7.
[19] Salloway S, Ferris S, Kluger A, Goldman R, Griesing T, Kumar D, et
al. Efficacy of donepezil in mild cognitive impairment: a randomized
placebo-controlled trial. Neurology 2004;63:651–7.
[20] Petersen RC, Thomas RG, Grundman M, Bennet D, Doody R, Ferris
S, et al. Vitamin E and donepezil for the treatment of mild cognitive
impairment. N Engl J Med 2005;352:2379 – 88.
[21] Blacker D. Mild cognitive impairment: no benefit from vitamin E,
little from donepezil. N Engl J Med 2005;352:2439 – 41.
[22] Ferris S, Feldman H, Sfikas N, Mancione L, Lane R. Time to clinical
diagnosis of dementia in MCI subjects receiving rivastigmine: a
randomized, double-blind, placebo-controlled study (abstract). Int
Psychogeriatr 2005;17:242.
[23] Feldman HH, Ferris S, Winblad B, Sfikas N, Mancione L, He Y, et al.
Effect of rivastigmine on delay to diagnosis of Alzheimer’s disease
 F. Massoud et al. / Alzheimer’s & Dementia 3 (2007) 283–291
from mild cognitive impairment: the InDDEx study. Lancet Neurol
2007;6:501–12.
[24] Scheltens P, Fox NC, Barkhof F, Gold M. Effect of galantamine
treatment on brain atrophy as assessed by MRI in patients with mild
cognitive impairment. Neurobiol Aging 2004;25(Suppl 2):S270 –1.
[25] Etminan M, Gill S, Samii A. Effect of non-steroidal anti-inflammatory
drugs on risk of Alzheimer’s disease: systematic review and meta-
analysis of observational studies. BMJ 2003;327:128 –31.
[26] Thal LJ, Ferris SH, Kirby L, Block GA, Lines CR, Yuen E, et al. A
randomized, double-blind, study of rofecoxib in patients with mild
cognitive impairment. Neuropsychopharmacology 2005;30:1204 –15.
[27] Massoud F, Yaffe K, Sano M. Estrogen. In: Qizillbash N, ed. Evidence-
based dementia practice. Oxford, UK: Blackwell Science; 2002. p.
523– 6.
[28] Shumaker SA, Legault C, Rapp SR, Thal L, Wallace RB, Ockene
JK, et al. Estrogen plus progestin and the incidence of dementia
and mild cognitive impairment in postmenopausal women: the
Women’s Health Initiative Memory Study—a randomized con-
trolled trial. JAMA 2003;289:2651– 62.
[29] Yaffe K, Krueger K, Cummings SR, Blackwell T, Henderson VW,
Sarkar S, et al. Effect of raloxifene on prevention of dementia and
cognitive impairment in older women: the Multiple Outcomes of
Raloxifene Evaluation (MORE) randomized trial. Am J Psychiatry
2005;162:683–90.
[30] van Dongen M, van Rossum E, Kessels A, Sielhorst H, Knipschild P.
Ginkgo for elderly people with dementia and age-associated memory
impairment: a randomized clinical trial. J Clin Epidemiol 2003;56:
367–76.
[31] Birks J, Grimley EV, Van Dongen M. Ginkgo biloba for cogni-
tive impairment and dementia. Cochrane Database Syst Rev 2002.
CD003120.
[32] Morris MC, Evans DA, Bienias JL, Tangney CC, Wilson RS.
Vitamin E and cognitive decline in older persons. Arch Neurol
2002;59:1125–32.
[33] Engelhart MJ, Gerlings MI, Ruitenberg A, van Swieten JC, Hofman
A, Witteman JC, et al. Dietary intake of antioxidants and risk of
Alzheimer disease. JAMA 2002;287:3223–9.
[34] Sano M, Ernesto C, Thomas RG, Klauber MR, Schafer K, Grund-
man M, et al. A controlled trial of selegiline, alpha-tocopherol, or
both as treatment for Alzheimer’s disease: the Alzheimer’s Dis-
ease Cooperative Study. New Engl J Med 1997;336:1216 –22.
291
[35] Miller ER 3rd, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ,
Guallar E. Meta-analysis: high-dosage vitamin E supplementation
may increase all-cause mortality. Ann Intern Med 2005;142:37– 46.
[36] Ott A, Stolk RP, van Harskamp F, Pols HA, Hofman A, Breteler MM.
Diabetes mellitus and the risk of dementia: the Rotterdam Study.
Neurology 1999;53:1937– 42.
[37] Yaffe K, Blackwell T, Kanaya AM, Davidowitz N, Barrett-Connor E,
Krueger K. Diabetes, impaired fasting glucose, and development of
cognitive impairment in older women. Neurology 2004;63:658 – 63.
[38] Yaffe K, Kanaya A, Lindquist K, Simonsick EM, Harris T, Shorr
RI, et al. The metabolic syndrome, inflammation, and risk of
cognitive decline. JAMA 2004;292:2237– 42.
[39] Elias MF, Elias PK, Sullivan LM, Wolf PA, D’Agostino RB.
Lower cognitive function in the presence of obesity and hyperten-
sion: the Framingham heart study. Int J Obes Relat Metab Disord
2003;27:260 – 8.
[40] Forette F, Seux ML, Staessen JA, Thijs L, Birkenhäger WH, Babar-
skiene MR, et al. Prevention of dementia in randomised double-blind
placebo-controlled Systolic Hypertension in Europe (Syst-Eur) trial.
Lancet 1998;352:1347–51.
[41] Tzourio C, Anderson C, Chapman N, Woodward N, Neal B, Mac-
Mahon S, et al. Effects of blood pressure lowering with perindopril
and indapamide therapy on dementia and cognitive decline in patients
with cerebrovascular disease. Arch Intern Med 2003;163:1069 –75.
[42] Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe
SM, et al. Pravastatin in elderly individuals at risk of vascular
disease (PROSPER): a randomised controlled trial. Lancet 2002;360:
1623–30.
[43] Collins R, Armitage J, Parish S, Sleight P, Peto R. Effects of cholesterol-
lowering with simvastatin on stroke and other major vascular events in
20536 people with cerebrovascular disease or other high-risk conditions.
Lancet 2004;363:757– 67.
[44] Starr JM, Whalley LJ, Deary IJ. The effects of antihypertensive
treatment on cognitive function: results from the HOPE study. J Am
Geriatr Soc 1996;44:411–5.
[45] Ishizaki J, Meguro K, Ohe K, Kimura E, Tsuchiya E, Ishii H, et al.
Therapeutic psychosocial intervention for elderly subjects with very
mild Alzheimer disease in a community: the tajiri project. Alzheimer
Dis Assoc Disord 2002;16:261–9.
[46] Poon P, Hui E, Dai D, Kwok T, Woo J. Cognitive intervention for
community-dwelling older persons with memory problems: telemedi-
cine versus face-to-face treatment. Int J Geriatr Psychiatry 2005;20:
285– 6.
 Alzheimer’s & Dementia 3 (2007) 292–298

Clinical diagnosis of dementia

Alain Robillard
Division of Neurology, Department of Medicine, Hôpital Maisonneuve-Rosemont & Université de Montréal, Montréal, Quebec, Canada

Abstract This paper presents recommendations deriving from the Third Canadian Consensus Conference on
the Diagnosis and Treatment of Dementia, concerning the clinical diagnosis of dementia. There are
currently no universally accepted biological or radiological markers of dementia. In the absence of
these, the diagnosis of dementia remains a clinical exercise aiming to integrate all available clinical
and laboratory information. It is proposed that the currently used National Institute of Neurological
and Communicative Disorders and Stroke/Alzheimer’s Disease and Related Disorders Association
(NINCDS/ADRA) criteria for diagnosis of Alzheimer’s disease (AD) be retained. The currently
available vascular dementia (VaD) diagnostic criteria have variable accuracy. An integrative
approach to VaD diagnosis based on all the available evidence (history, vascular risk factors,
physical exam, clinical course, neuroimaging, cognitive impairment pattern) is recommended. The
separation of Lewy body dementia (DLB) from Parkinson’s disease dementia (PDD) is based on the
dominant clinical presenting feature of each syndrome, and relies on the duration of this feature:
long duration of parkinsonian “motor” syndrome preceding dementia for PDD versus early/initial
dementia accompanied by extrapyramidal symptoms for DLB. It is recognized that it impossible
clinically to characterize DLB with (pathologically) coexisting AD changes. The Frontotemporal
group of dementia syndromes are discussed in regards to their typical clinical pictures, recognizing
that their neuropathological substrate are not predictable from their mode of presentation. Finally,
the particular rapid time sequence of evolution of the dementias due to prior disease is recognized
as the clinically most useful distinguishing feature of these syndromes.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Diagnosis; Alzheimer; Lewy body; Frontotemporal; Vascular; Behavior

1. Introduction paper in this volume) by Rockwood et al, which addresses

This paper presents recommendations deriving from the
Third Canadian Consensus Conference on the Diagnosis
and Treatment of Dementia, concerning the clinical diag-
nosis of dementia. These recommendations are the result of
a standardized literature review (see series Introduction for
methods) and all of the recommendations were supported by
at least 80% of the participants at the national meeting held
in Montreal in 2006. Our focus was on whether there were
clearly agreed upon clinical diagnostic criteria for different
dementia entities, and if so, what those recommended cri-
teria were. Note that there is a later paper (in fact the final
Corresponding author: Alain Robillard Tel.: (514) 252-3528; Fax:
(514) 252-3529.
E-mail address: robillal@videotron.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.08.002
the question of clinical criteria from a different perspective.
That paper critically reviews the literature and proposes
changes in the current criteria. The recommendations of this
paper pertain to how physicians should apply the current
criteria as presently understood. The purpose of the Rock-
wood et al paper is to propose changes which might lead in
the future to more cogent or coherent conceptualization of
the dementias. These are complementary goals.
The diagnosis of dementia and of its various etiologies
remain at its core a clinical exercise based on the presenting
and dominant symptoms, their order of appearance and the
evolution of symptoms and signs over time. A number of
classification criteria based on this procedure have been
proposed through the years for Alzheimer’s disease (AD)
and vascular dementia (VaD), with various precision when
compared with diagnosis based on pathological outcome.
 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
The same applies to the Frontotemporal group of dementias,
where pathological diagnosis is more difficult to predict on
clinical grounds alone. The problem is similar in the Lewy
Body dementias (Parkinson’s disease dementia and Lewy
Body dementia) where there exist a considerable overlap in
the presenting symptoms, and where again neuropathology
cannot always be relied on to “confirm” a clinical diagnosis.
Here the sequence of appearance of the cardinal symptoms
in both syndromes is central to a clinical diagnosis. In all of
these dementias, a descriptive and “integrative” approach to
diagnosis is proposed in the following discussions.
The essential symptoms of dementia are “. . . an acquired
impairment in short and long-term memory, associated with
impairment in abstract thinking, impaired judgment, other
disturbances of higher cortical function, or personality
changes. The disturbance is severe enough to interfere with
work or usual social activities or relationships with others.
The diagnosis of dementia is not made if these symptoms
occur in the presence of Delirium.” This definition is the
most widely used in practice and is based on the Diagnostic
and Statistical Manual [of Mental Disorders], Third Edition,
Revised (DSM-IIIR), DSM-IV, and the National Institute of
Neurological and Communicative Disorders and Stroke/
Alzheimer’s Disease and Related Disorders Association
(NINCDS/ADRDA) [1–3].
This DSM III-R definition of dementia has good reliabil-
ity particularly with respect to Alzheimer’s disease. The
DSM-IV version has not been validated, although it is
identical to the former definition [2,4,5].
After identifying the presence of a dementia syndrome,
the specific causes can be recognized by using clinical
criteria for the various clinical profiles.
2. AD
AD is characterized by gradual onset and continued
decline of memory and at least one additional cognitive
domain that is not explained by another systemic or neuro-
logic disorder [3].
Both the reliability of the clinical diagnosis and inter-
rater reliability have been subjected to review, with more
than 14 studies addressing these concerns [6 –23]. The gen-
eral consensus is that on the basis of Class I and Class II
studies and with neuropathologic data as confirming evi-
dence of clinical diagnosis, there is very good sensitivity of
clinical criteria (average, 81%; range, 49% to 100%), at the
expense of specificity (average, 70%; range, 47% to 100%),
with the NINCDS/ADRDA definition of “probable” AD.
“Possible” AD, as a diagnostic category, achieves higher
sensitivity (average, 93%) but much lower specificity (av-
erage, 48%); this reflects the fact that there are many com-
mon features between different types of dementia.
“Mild” AD (as defined by a Mini Mental State Exami-
nation [MMSE] score of ⱖ24) can also be diagnosed ac-
cording to clinical criteria, with a high degree of certainty
293
(89%, compared with pathology) (Class II, one trial). The
diagnostic accuracy is enhanced in this group of patients by
the use of a more comprehensive neuropsychological test
battery [24].
Patients with mild cognitive impairment (MCI), as de-
fined by Petersen et al [25], have a memory impairment
beyond that expected for age and education, yet they are not
demented. These subjects have become the focus of many
prediction studies and are diagnosed according to clinical
criteria. Debate is still ongoing regarding the specific sub-
types of MCI (amnestic and non-amnestic, single or multi-
ple domains), with definitions used on the basis of a pre-
sumed etiology. With the criteria of Petersen et al for
amnestic MCI, it has been shown that a high proportion of
these patients go on to AD [25–28]. The definition of MCI
is difficult to apply in a clinical setting, and the available
diagnostic instruments used for diagnosing dementia lack
sensitivity to MCI. A new tool, the Montreal Cognitive
Assessment (MoCA), has been proposed and validated in
one trial [29], demonstrating a higher sensitivity for MCI
(100% when judged against the MMSE, with a cutoff score
of 26), with good specificity (87%). A separate paper by
Chertkow et al in this issue further pursues the concept of
MCI and derives evidence-based recommendations for MCI
diagnosis.
3. VaD
VaD is a dementia syndrome resulting from cerebrovas-
cular damage. This syndrome can occur as the result of a
single strategically placed brain infarct or as the result of
numerous discrete smaller (and most often subcortical) le-
sions. The progression is classically in a stepwise fashion,
but increasingly a slow insidious progress is recognized.
The clinical hallmark of VaD is dysexecutive syndrome.
Four consensus criteria for VaD are in use: the State of
California AD Diagnostic and Treatment Centers criteria
(the California criteria), the National Institute of Neurologic
Disorders and Stroke and the Association Internationale
pour la Recherche et l’Enseignement en Neurosciences
(NINDS-AIREN) criteria, the Hachinski Ischemic Score
(HIS) modified by Rosen, and those found in the DSM-IV.
They all have poor sensitivity at the cost of high specificity,
when judged against the best clinical judgment (including
neuroimaging) of trained clinicians [30] (Class II evidence).
In studies that compared clinical diagnoses and neuropatho-
logic findings, the NINDS-AIREN and the California crite-
ria (as well as DSM-IIIR) had very low sensitivity but
higher specificity. One Class I study [31] reported the sen-
sitivity (43%) and specificity (95%) of the NINDS-AIREN
criteria for VaD. Five Class II studies [32–36] reported
sensitivity and specificity of the diagnosis of VaD with any
criteria. The results showed low sensitivity (average of 5
studies, 50%; range, 20% to 89%) but high specificity (av-
erage across 5 studies, 87%; range, 64% to 98%) for the
294 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
HIS, DSM-IIIR, NINDS-AIREN, or California clinical cri-
teria. Only one retrospective (Class II) study [35] showed
better sensitivity and specificity (both 89%) with the HIS.
This lack of sensitivity in the available diagnostic criteria
is confounded by the presence on neuropathologic exami-
nation of considerable overlap between vascular and degen-
erative (AD) pathology (some vascular pathology exists in
29% to 41% of dementia cases coming to autopsy), whereas
pure vascular pathology accounts for dementia in only 9%
to 10% [12,14], leading some to propose that pure VaD or
AD is rare [12–17].
A descriptive approach to the diagnosis of VaD is pro-
posed, which would take into account the neuropsycholog-
ical profile of the dementia syndrome, neuroimaging, and
vascular risk factors [37,38].
4. Lewy body disease
The third report of the Dementia with Lewy Bodies
(DLB) consortium was published in December 2005 [39].
The central feature of DLB remains the presence of a
dementia syndrome of sufficient intensity to interfere with
social or occupational function, although memory impair-
ment might not be prominent at diagnosis, whereas deficits
on tests of attention, visuospatial ability, and executive
function might be more prominent. The core features in-
clude spontaneous parkinsonian symptoms, fluctuating cog-
nition with pronounced variations in attention and alertness,
and recurrent well-formed and detailed visual hallucina-
tions. The consensus report also details suggestive as well as
supportive features that add a degree of probability to the
diagnosis. Clinically the most important supportive symp-
toms remain rapid eye movement sleep behavior disorder
and severe neuroleptic sensitivity.
The time sequence in which symptoms of DLB appear is
key to differentiation from Parkinson’s disease dementia
(PDD); typically dementia appears after a number of years
in PD (10 years usually), whereas parkinsonian symptoms
are central to the initial diagnosis of DLB. There are no
distinguishing clinical features that can discriminate be-
tween PDD and DLB at any given point in time [40].
The sensitivity of the DLB diagnosis based on pathology
has been subject to controversy [41] due in large part to the
original pathologic reports stating that the presence of any
LBs in whatever distribution was sufficient for diagnosis
[42,43]. One Class I study [12] investigated the diagnostic
accuracy of DLB criteria against neuropathologic findings;
sensitivity was low at 22%, but specificity was high (100%).
Six Class II studies also showed low sensitivities (average
across 5 studies, 58%; range, 30% to 75%) but higher
specificities (average across 5 studies, 87%; range, 71% to
100%) for the Consortium diagnostic criteria of DLB
[10,44 – 48]. New pathologic criteria have been proposed by
the DLB Consortium that will take into account the con-
comitant presence of AD pathology in the future [39].
The recognition that AD and DLB neuropathology can
coexist to various degrees in patients presenting initially
with either AD or DLB symptomatology has lead to at-
tempts at clinical differentiation. In a Class II study, Lopez
et al [49] reported that in patients with pathologically diag-
nosed AD and DLB, sensitivity of the clinical diagnostic
criteria improved as dementia worsened, at the expense of
specificity. There were no distinguishing features between
AD and AD with DLB in mild dementia. In moderate
dementia, the presence of extrapyramidal signs (EPS) and
major depression were associated with AD ⫹ DLB, and in
severe dementia, EPS and diurnal hypersomnia were indi-
cators of AD ⫹ DLB.
Del Ser et al [50] noted in a Class II study that the clinical
features of AD ⫹ DLB patients were similar to those of AD
patients except for more frequent acute-subacute onset and a
fluctuating evolution. Differentiation between three groups
of patients (pure DLB, AD ⫹ LB, and AD) or between both
groups with LB (DLB) from AD could only be attained in
70% of cases. Accuracy was excellent for the diagnosis of
pure DLB but only mediocre for separating AD ⫹ LB as
well as the pure DLB and DLB ⫹ AD groups from AD.
In a Class II retrospective study of 90 patients with
autopsy-proven DLB, Merdes et al [51] noted that only 27%
of these demonstrated both visual hallucinations and spon-
taneous EPS. Subjects with lower tangle burden on Braak
staging had a higher frequency of visual hallucinations
(65%) than did subjects with DLB with higher Braak stages
(33%) and showed a slightly greater but not significant
degree of EPS. Although clinical diagnostic accuracy for
DLB was relatively low (49%), it was higher for subjects
with low (75%) compared with high (39%) Braak stages.
The degree of concomitant AD tangle pathology had an
important influence on the clinical characteristics and the
accuracy of clinical diagnosis of DLB.
5. Frontotemporal dementia
The clinical presentation in frontotemporal dementia
(FTD) reflects the distribution of the pathologic changes
rather than the exact histologic subtype of the disease. Three
clinical syndromes are recognized, all of insidious onset.
5.1. Frontal variant FTD
Frontal variant FTD (dementia of frontal type), in which
changes in social behavior and personality predominate,
reflects the orbitobasal frontal lobe focus of the pathology.
5.2. Semantic dementia
In semantic dementia (progressive fluent aphasia) there
is a breakdown in the conceptual database that underlies
language production and comprehension. Patients with se-
mantic dementia have asymmetric anterolateral temporal
atrophy with relative sparing of the hippocampal formation,
typically worse on the left side.
 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
5.3. Progressive nonfluent aphasia
In progressive nonfluent aphasia the phonologic and syn-
tactic components of language are affected in association
with left perisylvian atrophy [52].
The original Lund-Manchester criteria [53] and the Con-
sensus revision [54] have been tested against pathology,
with low specificity; an autopsy-based Class II study (based
on retrospective clinical diagnoses) showed that most pa-
tients with FTD fulfill diagnostic criteria for AD [55]. In
contrast, a Class II study without autopsy confirmation [56]
found that the Lund-Manchester criteria differentiated
100% of FTD and AD patients. A recent Class II retrospec-
tive study concluded that FTD, corticobasal degeneration,
and progressive supranuclear palsy have overlapping clini-
cal features, with FTD being an unpredictable disease in
terms of its tau-positive biochemistry [57].
6. Normal pressure hydrocephalus
The classic clinical presentation of normal pressure hy-
drocephalus (NPH) is one of progressive gait apraxia, uri-
nary incontinence, and ultimately dementia [58 – 60]. There
have been a number of reports of amelioration of the clinical
picture through ventriculoperitoneal shunting but no posi-
tive predictive test to differentiate those patients who might
benefit from the procedure from those who will not. One
trial of temporary external lumbar cerebrospinal fluid (CSF)
drainage proved negative because of a high rate of false-
negative results [61] (Class II); although the predictive
value of a positive external lumbar drainage was high, the
costs and invasiveness of the test and the possibility of
serious test-related complications limit its usefulness in
managing patients with presumed NPH.
Lately there has been interest for low-flow CSF drainage
as a treatment for AD. Advancing age is the key risk factor
for developing AD. With aging, abnormal metabolism and
clearance of amyloid beta proteins (A␤) can lead to accu-
mulation of these proteins in brain parenchyma, resulting in
plaque formation. It has been postulated that age-associated
impairment in the CSF circulation and cerebral vasculature
could play a role, through diminished clearance of A␤, in
the high incidence of AD in the elderly [62,63]. The result
of one pilot trial of low-flow CSF drainage has been re-
ported, indicating a small stabilizing effect on dementia
severity rating (Class III) [64]. There is one ongoing trial for
which no results have been published.
7. Prion diseases
The clinical diagnosis of Creutzfeldt-Jakob disease (CJD)
rests on the occurrence of a rapidly progressing dementia
syndrome, associated with myoclonus and a characteristic
electroencephalogram (EEG) pattern of periodic sharp waves.
Probable CJD is diagnosed if patients have rapidly progres-
295
sive dementia, periodic sharp waves in the EEG, and at least
two of the following four findings: myoclonus, visual and/or
cerebellar symptoms, pyramidal and/or extrapyramidal
signs, and akinetic mutism. Cases that fulfill these criteria
but do not have a typical EEG are classified as possible
CJD, according to the criteria of Masters et al [65]. Other
criteria have been proposed (French [66] and European
criteria [67]). In a prospective Class I study Poser et al [68]
tested the diagnostic criteria for CJD. Diagnosis was con-
firmed in 188 autopsy-confirmed cases (97%) of 193 cases
diagnosed with probable CJD. In this cohort, among 54
cases diagnosed with possible CJD, the diagnosis was con-
firmed in 44 (81%). Only two pathologically diagnosed CJD
cases were found among 111 patients who had been given
other clinical diagnoses. Thus the criteria achieved high
sensitivity and specificity.
In a Class II retrospective trial Brandel et al [69] com-
pared three sets of clinical diagnostic criteria in a cohort of
428 patients, showing that the three sets of criteria had a
high specificity and low sensitivity as judged against neu-
ropathology; all three criteria had poor negative predictive
value (average, 44%; range, 29% to 57%).
An immunoassay for the detection of the 14-3-3 protein
in CSF was described in 1996 [70] that had a reported
specificity of 99% and a sensitivity of 96% for the diagnosis
of CJD among patients with dementia who had not had a
stroke within 1 month of testing. During the ensuing years
it has been shown that the 14-3-3 assay is not valid to
discriminate between unselected rapidly progressing de-
mentia syndromes. Burkhard et al [71] did a Class I study
but validated against clinical diagnosis, showing 14 of 100
positive patients, of whom only two had a final diagnosis of
CJD. There is also variation in the sensitivity of the 14-3-3
assay among CJD subtypes. In a Class II study Castellani
et al [72] reported sensitivity between 77% and 94%, ac-
cording to subtypes of CJD; the lower sensitivity was re-
ported in the less frequent atypical subtypes.
Magnetic resonance imaging (MRI), especially with
diffusion-weighted imaging, in a Class II study but with a
small number of subjects [73], showed evidence of a com-
parable sensitivity to 14-3-3 assay. In a prospective study of
219 patients suspected of having CJD, the presence of basal
ganglia high signal on T2-weighted MRI showed a sensi-
tivity of 63% and a specificity of 88% in definite/probable
CJD [74].
8. Recommendations approved at the Third Canadian
Consensus Conference on Diagnosis and Treatment
of Dementia
(1) The diagnosis of dementia remains clinical. There
is good evidence to retain the diagnostic criteria
currently in use (Grade A, Level 2).
296 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
(2) The sensitivity of clinical diagnosis for possible or
probable AD based on the NINCDS-ADRDA cri-
teria remains high. The specificity is lower. The
continued use of the NINCDS-ADRDA criteria is
recommended (Grade A, Level 1).
(3) Mild AD can be diagnosed with a high degree of
specificity, when the presenting clinical picture is
one of memory impairment (Grade B, Level 1).
(4) The currently available VaD diagnostic criteria
have variable accuracy. An integrative approach to
VaD diagnosis based on all the available evidence
(history, vascular risk factors, physical exam, clin-
ical course, neuroimaging, cognitive impairment
pattern) is recommended (Grade B, Level 2).
(5) The clinical features of DLB and PDD overlap con-
siderably. At present, DLB should be diagnosed when
this pattern of dementia occurs before or concurrently
with parkinsonism. PDD can be diagnosed when de-
mentia occurs in the context of well-established PD,
generally after many years (Grade B, Level 3).
(6) There is frequent coexistence of AD and LB
neuropathology in subjects presenting with the
initial clinical picture of either pathology. At
present, it is impossible to propose clinical
guidelines that would permit separation of the
two diagnoses and AD ⫹ DLB with a high spec-
ificity (Grade A, Level 2).
(7) In patients presenting primarily with progressive
decline in language or praxis or with prominent
changes in behavior or personality, Pick complex dis-
ease (also known as FTD or frontotemporal lobar
degeneration) should be considered. These disorders
of Pick complex include semantic dementia, FTD
behavioral variant, primary progressive aphasia,
corticobasal degeneration, progressive supranuclear
palsy, and FTD with motor neuron disease. These
disorders have clinical features that are distinctive
and are best referred for specialist care (Grade A,
Level 2).
(8) When gait apraxia or urinary incontinence occurs
early in the course of dementia, NPH should be
considered and should be supported by computed
tomography or MRI. Specialist referral is advised
for further assessment. The diagnostic work-up
might include the removal of a large CSF volume
with documentation of clinical response, if surgical
intervention is considered an appropriate option
(Grade B, Level 2).
(9) Rapidly progressing dementia associated with my-
oclonus and an EEG with the presence of periodic
sharp waves is typical of CJD. There are three
clinical criteria sets in use to diagnose CJD, all with
some specificity. The recognition of rapid progres-
sion in a dementia syndrome should by itself sug-
gest the possibility of CJD (Grade A, Level 2).
(10) CJD is a distinctively rapidly progressive dementia
that in Canada requires reporting to the CJD Sur-
veillance Network and requires special infection
control procedures. Its diagnosis is supported by a
positive 14-3-3 test on CSF and by an abnormal
MRI scan, with either basal ganglia high signal on
T2 or abnormalities on diffusion-weighted imaging,
and by a progressively worsening EEG with peri-
odic complexes. Specialist referral is recommended
(Grade B, Level 2).
Author Disclosures
Alain Robillard has received support from Pfizer Canada
(Speaker, Administrative Board Member, Investigator), No-
vartis (Speaker, Administrative Board Member, Investiga-
tor), Janssen Ortho Inc. (Speaker, Administrative Board
Member, Investigator), Lundbeck Canada (Speaker, Ad-
ministrative Board Member, Investigator), Sanofi-Aventis
(Clinical Investigator).
References
[1] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders. 4th ed. Washington, DC: American Psychiatric
Association; 1994.
[2] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders. 3rd ed–revised. Washington, DC: American
Psychiatric Association; 1987.
[3] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM. Clinical diagnosis of Alzheimer’s disease: report of the
NINCDS-ADRDA Work Group under the auspices of Department of
Health and Human Services Task Force on Alzheimer’s Disease.
Neurology 1984;34:939 – 44.
[4] O’Connor DW, Blessed G, Cooper B, Jonker C, Morris JC, Presnell
IB, et al. Cross-national interrater reliability of dementia diagnosis in
the elderly and factors associated with disagreement. Neurology
1996;47:1194 –9.
[5] Graham JE, Rockwood K, Beattie BL, McDowell I, Eastwood R,
Gauthier S. Standardization of the diagnosis of dementia in the
Canadian Study of Health and Aging. Neuroepidemiology 1996;15:
246 –56.
[6] Larson EB, McCurry SM, Graves AB, Bowen JD, Rice MM, Mc-
Cormick WC, et al. Standardization of the clinical diagnosis of the
dementia syndrome and its subtypes in a cross-sectional study: the
Ni-Hon-Sea experience. J Gerontol A Biol Sci Med Sci 1998;53A:
M313–9.
[7] Fratiglioni L, Grut M, Forsell Y, Viitanen M, Winblad B. Clinical
diagnosis of Alzheimer’s disease and other dementias in a population
survey: agreement and causes of disagreement in applying Diagnostic
and Statistical Manual of Mental Disorders, Revised Third Edition,
criteria. Arch Neurol 1992;49:927–2.
[8] Farrer LA, Cupples LA, Blackburn S, Kiely DK, Auerbach S, Grow-
don JH, et al. Interrater agreement for diagnosis of Alzheimer’s
disease: the MIRAGE study. Neurology 1994;44:652– 6.
[9] Blacker D, Albert MS, Bassett SS, Go RC, Harrell LE, Folstein MF.
Reliability and validity of NINCDS-ADRDA criteria for Alzheimer’s
disease: the National Institute of Mental Health Genetics Initiative.
Arch Neurol 1994;51:1198 –204.
 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
[10] Lopez OL, Litvan I, Catt KE, Stowe R, Klunk W, Kaufer DI, et al.
Accuracy of four clinical diagnostic criteria for the diagnosis of
neurodegenerative dementias. Neurology 1999;53:1292–9.
[11] Lopez OL, Becker JT, Klunk W, Saxton J, Hamilton RL, Kaufer DI,
et al. Research evaluation and diagnosis of probable Alzheimer’s
disease over the last two decades: I. Neurology 2000;55:1854 – 62.
[12] Forette F, Henry JF, Orgogozo JM, et al. Reliability of clinical criteria
for the diagnosis of dementia: a longitudinal multicenter study. Arch
Neurol 1989;46:646 – 8.
[13] Holmes C, Cairns N, Lantos P, Mann A. Validity of current clinical
criteria for Alzheimer’s disease, vascular dementia and dementia with
Lewy bodies. Br J Psychiatry 1999;174:45–50.
[14] Jobst KA, Barnetson LP, Shepstone BJ. Accurate prediction of
histologically confirmed Alzheimer’s disease and the differential
diagnosis of dementia: the use of NINCDS-ADRDA and DSM-
III-R criteria, spect, X-ray CT, and apo E4 in medial temporal lobe
dementias—Oxford Project to Investigate Memory and Aging. Int
Psychogeriatr 1998;10:271–302.
[15] Lim A, Tsuang D, Kukull W, Nochlin D, Leverenz J, McCormick W,
et al. Clinico-neuropathological correlation of Alzheimer’s disease in
a community-based case series. J Am Geriatr Soc 1999;47:564 –9.
[16] Wade JP, Mirsen TR, Hachinski VC, Fisman M, Lau C, Merskey H.
The clinical diagnosis of Alzheimer’s disease. Arch Neurol 1987;44:
24 –9.
[17] Victoroff J, Mack WJ, Lyness SA, Chui HC. Multicenter clinicopath-
ological correlation in dementia. Am J Psychiatry 1995;152:1476 –
84.
[18] Galasko D, Hansen LA, Katzman R, Wiederholt W, Masliah E, Terry
R, et al. Clinical-neuropathological correlations in Alzheimer’s dis-
ease and related dementias. Arch Neurol 1994;51:888 –95.
[19] Mayeux R, Saunders AM, Shea S, Mirra S, Evans D, Roses AD, et al.
Utility of the apolipoprotein E genotype in the diagnosis of Alzhei-
mer’s disease: Alzheimer’s Disease Centers Consortium on Apoli-
poprotein E and Alzheimer’s Disease. N Engl J Med 1998;338:506 –
11.
[20] Klatka LA, Schiffer RB, Powers JM, Kazee AM. Incorrect diagnosis
of Alzheimer’s disease: a clinicopathologic study. Arch Neurol 1997;
53:35– 42.
[21] Gearing M, Mirra SS, Hedreen JC, Sumi SM, Hansen LA, Heyman
A. The Consortium to Establish a Registry for Alzheimer’s Disease
(CERAD): part X—neuropathology confirmation of the clinical diag-
nosis of Alzheimer’s disease. Neurology 1995;45:461– 6.
[22] Berg L, McKeel DW Jr., Miller JP, Storandt M, Rubin EH, Morris JC,
et al. Clinicopathologic studies in cognitively healthy aging and
Alzheimer’s disease: relation of histologic markers to dementia se-
verity, age, sex, and apolipoprotein E genotype. Arch Neurol 1998;
55:326 –35.
[23] Massoud F, Devi G, Stern Y, Lawton A, Goldman JE, Liu Y, et al. A
clinicopathological comparison of community-based and clinic-based
cohorts of patients with dementia. Arch Neurol 1999;56:1368 –73.
[24] Salmon DP, Thomas RG, Pay MM, Booth A, Hofstetter CR, Thal LJ,
et al. Alzheimer’s disease can be accurately diagnosed in very mildly
impaired individuals. Neurology 2002;59:1022– 8.
[25] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG, Kokmen
E. Mild cognitive impairment: clinical characterization and outcome.
Arch Neurol 1999;56:303– 8.
[26] Petersen RC, Morris JC. Mild cognitive impairment as a clinical
entity and treatment target. Arch Neurol 2005;62:1160 –3.
[27] Petersen RC. Mild cognitive impairment clinical trials. Nat Rev Drug
Discov 2003;2:646 –53.
[28] Grundman M, Petersen RC, Ferris S, Thomas RG, Aisen PS, Bennett
DA, et al. Mild cognitive impairment can be distinguished from
Alzheimer disease and normal aging for clinical trials. Arch Neurol
2004;61:59 – 66.
297
[29] Petersen RC, Thomas RG, Grundman M, Bennett D, Doody R, Ferris
S, et al. Vitamin E and donepezil in the treatment of mild cognitive
impairment. N Engl J Med. 2005;352:2379 – 88.
[30] Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead
V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief
screening tool for mild cognitive impairment. J Am Geriatr Soc
2005;53:695–9.
[31] Rockwood K, Davis H, MacKnight C, Vandorpe R, Gauthier S,
Guzman A, et al. The Consortium to Investigate Vascular Impairment
of Cognition: methods and first findings. Can J Neurol Sci 2003;30:
237– 43.
[32] Gold G, Giannakopoulos P, Montes-Paixao Junior C, et al. Sensitivity
and specificity of newly proposed clinical criteria for possible vas-
cular dementia. Neurology 1997;49:690 – 4.
[33] Gold G, Bouras C, Canuto A, Bergallo Herrmenn FR, Hof P, Mayor
PA, et al. Clinicopathological validation study of four sets of clinical
criteria for vascular dementia. Am J Psychiatry 2002;159:82–7.
[34] Pohjasvaara T, Mantyla R, Ylikoski R, Kaste M, Erkinjuntti T.
Comparison of different clinical criteria (DSM-III, ADDTC, ICD-
10,NINDS-AIREN, DSM-IV) for the diagnosis of vascular dementia.
Stroke 2000;31:2952–7.
[35] Cosentino SA, Jefferson AL, Carey M, Price CC, Davis-Garrett K,
Swenson R, et al. The clinical diagnosis of vascular dementia: a
comparison among four classification systems and a proposal for a
new paradigm. Clin Neuropsychol 2004;18:6 –21.
[36] Chui HC, Mack W, Jackson JE, Mungas D, Reed BR, Tinklenberg J,
et al. Clinical criteria for the diagnosis of vascular dementia: a
multicenter study of comparability and interrater reliability. Arch
Neurol 2000;57:191– 6.
[37] Graham NL, Emery T, Hodges JR. Distinctive cognitive profiles in
Alzheimer’s disease and subcortical vascular dementia. J Neurol
Neurosurg Psychiatry 2004;75:61–71.
[38] Tierney MC, Black SE, Szalai JP, Snow WG, Fisher RH, Nadon G,
et al. Recognition memory and verbal fluency differentiate probable
Alzheimer disease from subcortical ischemic vascular dementia. Arch
Neurol 2001;58:1654 –9.
[39] McKeith G, Dickson DW, Lowe J, Emre M, O’Brien JT, Feldman H,
et al. Diagnosis and management of dementia with Lewy bodies: third
report of the DLB consortium Neurology 2005;65:1863–72.
[40] Noe E, Marder K, Bell KL, Jacobs DM, Manly JJ, Stern Y. Com-
parison of dementia with Lewy bodies to Alzheimer’s disease and
Parkinson’s disease with dementia. Mov Disord 2004;19:60 –7.
[41] Litvan I, Bhatia KP, Burn DJ, Goetz CG, Lang AE, McKeith I, et al.
SIC Task Force Appraisal of clinical diagnostic criteria for parkin-
sonian disorders. Mov Disord 2003;18:467– 86.
[42] Kosaka K. Lewy bodies in cerebral cortex: report of three cases. Acta
Neuropathol 1978;42:127–34.
[43] Kosaka K, Yoshimura M, Ikeda K, Budka H. Diffuse type of Lewy
body disease: progressive dementia with abundant cortical Lewy
bodies and senile changes of varying degree: a new disease? Clin
Neuropathol 1984;3:185–92.
[44] Litvan I, MacIntyre A, Goetz CG, Wenning GK, Jellinger K, Verny
M, et al. Accuracy of the clinical diagnoses of Lewy body disease,
Parkinson disease, and dementia with Lewy bodies: a clinicopatho-
logic study. Arch Neurol 1998;55:969 –78.
[45] Luis CA, Barker WW, Gajaraj K, Harwood D, Petersen R, Kashuba
A, et al. Sensitivity and specificity of three clinical criteria for de-
mentia with Lewy bodies in an autopsy-verified sample. Int J Geriatr
Psychiatry 1999;14:526 –33.
[46] Mega MS, Masterman DL, Benson DF, Vinters HV. Dementia with
Lewy bodies: reliability and validity of clinical and pathologic crite-
ria. Neurology 1996;47:1403–9.
[47] Verghese J, Crystal HA, Dickson DW, Lipton RB. Validity of clinical
criteria for the diagnosis of dementia with Lewy bodies. Neurology
1999;53:1974 – 82.
298 A. Robillard / Alzheimer’s & Dementia 3 (2007) 292–298
[48] Lopez OL, Becker JT, Kaufer D, Hamilton R, Sweet R, Klunk W,
et al. Research evaluation and prospective diagnosis of dementia with
Lewy bodies. Arch Neurol 2002;59:43– 6.
[49] Lopez OL, Hamilton RL, Becker JT, Wisniewski S, Kaufer DI,
DeKosky ST. Severity of cognitive impairment and the clinical di-
agnosis of AD with Lewy bodies. Neurology 2000;54:1780 –7.
[50] Del Ser T, Hachinski V, Merskey H, Munoz DG. Clinical and patho-
logic features of two groups of patients with dementia with Lewy
bodies: effect of coexisting Alzheimer-type lesion load. Alzheimer
Dis Assoc Disord 2001;15:31– 44.
[51] Merdes AR, Hansen LA, Jeste DV, Galasko D, Hofstetter CR, Ho GJ,
et al. Influence of Alzheimer pathology on clinical diagnostic accu-
racy in dementia with Lewy bodies. Neurology 2003;60:1586 –90.
[52] Hodges JR. Frontotemporal dementia (Pick’s disease): clinical fea-
tures and assessment. Neurology 2001;56(Suppl 4):S6 –10.
[53] Clinical and neuropathological criteria for frontotemporal dementia.
J Neurol Neurosurg Psychiatry 1994;57:416 – 8.
[54] Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S, et
al. Frontotemporal lobar degeneration: a consensus on clinical diag-
nostic criteria. Neurology 1998;51:1546 –54.
[55] Varma AR, Snowden JS, Lloyd JJ, Talbot PR, Mann DM, Neary D.
Evaluation of the NINCDS-ADRDA criteria in the differentiation of
Alzheimer’s disease and frontotemporal dementia. J Neurol Neuro-
surg Psychiatry 1999;66:184 – 8.
[56] Miller BL, Ikonte C, Ponton M, Levy M, Boone K, Darby A, et al. A
study of the Lund-Manchester research criteria for frontotemporal
dementia: clinical and single-photon emission CT correlations. Neu-
rology 1997;48:937– 42.
[57] Josephs KA, Petersen RC, Knopman DS, Boeve BF, Whitwell JL,
Duffy JR, et al. Clinicopathologic analysis of frontotemporal and
corticobasal degenerations and PSP. Neurology 2006;66:41– 8.
[58] Hakim S, Adams RD. The special clinical problem of symptomatic
hydrocephalus with normal cerebrospinal fluid pressure: observations
on cerebrospinal fluid hydrodynamics. J Neurol Sci. 1965;2:307–27.
[59] Vanneste JA. Three decades of normal pressure hydrocephalus: are
we wiser now? J Neurol Neurosurg Psychiatry 1994;57:1021–5.
[60] Vanneste J, Augustijn P, Dirven C, Tan WF, Goedhart ZD. Shunting
normal-pressure hydrocephalus: do the benefits outweigh the risks?
Neurology 1992;42:54 – 8.
[61] Josephs KA, Petersen RC, Knopman DS, Boeve BF, Whitwell JL,
Duffy JR, et al. The value of temporary external lumbar CSF drainage
in predicting the outcome of shunting on normal pressure hydroceph-
alus. J Neurol Neurosurg Psychiatry 2002;72:503– 6.
[62] Silverberg GD, Mayo M, Saul T, Rubenstein E, McGuire D. Alzhei-
mer’s disease, normal□pressure hydrocephalus, and senescent
changes in CSF circulatory physiology: a hypothesis. Lancet Neurol-
2003;2:506 –11.
[63] Silverberg GD, Heit G, Huhn S, Jaffe RA, Chang SD, Bronte-Stewart
H, et al. CSF production is reduced in dementia of the Alzheimer’s
type. Neurology 2001;57:1763– 6.
[64] Silverberg GD, Levinthal E, Sullivan EV, Bloch DA, Chang SD,
Leverenz J, et al. Assessment of low-flow CSF drainage as a treat-
ment for AD: results of a randomized pilot study. Neurology 2002;
59:1139 – 45.
[65] Masters CL, Harris JO, Gajdusek DC, Gibbs CJ Jr, Bernoulli C,
Asher DM. Creutzfeldt-Jakob disease: patterns of worldwide occur-
rence and the significance of familial and sporadic clustering. Ann
Neurol 1979;5:177– 88.
[66] Cathala F, Brown P, Castaigne P, Gajdusek DC. La maladie de
Creutzfeldt-Jakob en France continentale: etude rétrospective de
1968 –1977. Rev Neurol 1979;5:439 –54.
[67] Concerted action of the EU. Surveillance of Creutzfeldt-Jakob disease
in the European community: minutes of the second meeting in Rome,
July 3, 1993, and the third meeting in Paris, April 29 –30, 1994.
[68] Poser S, Mollenhauer B, Kraubeta A, Zerr I, Steinhoff BJ, Schroeter
A, et al. How to improve the clinical diagnosis of Creutzfeldt–Jakob
disease. Brain 1999;122:2345–51.
[69] Brandel JP, Delasnerie-Laupretre N, Laplanche JL, Hauw JJ, Alpero-
vitch A. Diagnosis of Creutzfeldt-Jakob disease: effect of clinical
criteria on incidence estimates. Neurology 2000;54:1095–9.
[70] Hsich G, Kenney K, Gibbs CJ, Lee KH, Harrington MG. The 14-3-3
brain protein in cerebrospinal fluid as a marker for transmissible
spongiform encephalopathies. N Engl J Med 1996;335:924 –30.
[71] Burkhard PR, Sanchez JC, Landis T, Hochstrasser D. CSF detection
of the 14-3-3 protein in unselected patients with dementia. Neurology
1997;56:1528 –33.
[72] Castellani RJ, Colucci M, Xie Z, Zou W, Li C, Parchi P, et al.
Sensitivity of 14-3-3 protein test varies in subtypes of sporadic
Creutzfeldt-Jakob disease. Neurology 2004;63:436 – 42.
[73] Shiga Y, Miyazawa K, Sato S, Fukushima R, Shibuya S, Sato Y, et
al. Diffusion-weighted MRI abnormalities as an early diagnostic
marker for Creutzfeldt-Jakob disease. Neurology 2004;63:443–9.
[74] Meissner B, Köhler K, Körtner K, Bartl M, Jastrow U, Mollenhauer
B, et al. Sporadic Creutzfeldt-Jakob disease: magnetic resonance
imaging and clinical findings. Neurology 2004;63:450 – 6.
 Alzheimer’s & Dementia 3 (2007) 299 –317

Neuropsychological testing and assessment for dementia

Claudia Jacovaa, Andrew Kerteszb, Mervin Blairb, John Fiskc, Howard H. Feldmana,*
a
Division of Neurology, Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
b
Department of Clinical Neurological Sciences, University of Western Ontario, London, Ontario, Canada
c
QE II Health Sciences Centre, Departments of Psychiatry, Medicine and Psychology, Dalhousie University, Halifax, Nova Scotia, Canada

Abstract This evidence-based review examines the utility of brief cognitive tests and neuropsychological
testing (NPT) in the detection and diagnosis of Mild Cognitive Impairment (MCI) and dementia. All
patients presenting with cognitive complaints are recommended to a brief screening test adminis-
tered to document the presence and severity of memory/cognitive deficits. There is fair evidence to
support the use of a range of new screening tests that can detect MCI and mild dementia with higher
sensitivity (ⱖ80%) than the Mini-Mental State Exam (MMSE). NPT should be part of a clinically
integrative approach to the diagnosis and differential diagnosis of dementia. It should be applied
selectively to address specific clinical and diagnostic issues including: 1) The distinction between
normal cognitive functioning in the aged, MCI and early dementia: there is fair evidence that NPT
can document the presence of specific diagnostic criteria and provide additional useful information
on the pattern of memory/cognitive impairment. 2) The evaluation of risk for Alzheimer Disease
(AD) or other types of dementia in persons with MCI: there is fair evidence that NPT measures or
profiles can predict progression to dementia (predictive accuracy ranges from ⬃80 to 100%,
sensitivities from 53 to 80%, and specificities from 67 to 99%). 3) Differential diagnosis: There is
fair evidence that NPT can complement clinical history and neuroimaging in determining the
dementia etiology. Different dementia types have distinguishable NPT profiles though these may be
stage-dependent, and increased sensitivity may be at the expense of specificity. 4) When NPT is part
of a comprehensive assessment, which also entails clinical interviews and consideration of other
clinical data, there is good evidence that it can contribute to management decisions in MCI and
dementia, including the determination of retained and impaired cognitive abilities, their functional
and vocational impact, and opportunities for cognitive rehabilitation.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: ●●●

1. Introduction
The neuropsychological evaluation of suspected cogni-
tive impairment and dementia is conducted at different
levels of comprehensiveness and professional specializa-
tion. At a basic level, brief cognitive tests are administered
to screen for the presence of cognitive impairment and to
obtain a global index of cognitive functioning. At an inter-
mediate level, neuropsychological testing is performed with
batteries of selected standardized tests that probe a range of
*Corresponding author. Tel.: 604-822-7979; Fax: (604) 822-7703.
E-mail address: hfeldman@interchange.ubc.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.011
cognitive domains and that yield scores indexing the func-
tioning of these domains. In the professional practice of
neuropsychology, such testing is used to answer a variety of
clinical questions including, but not limited to, differential
diagnosis. At an advanced level, this testing is an integrative
part of a clinical neuropsychological assessment, which also
entails clinical interviews and the interpretation of data
obtained from other sources to render a diagnostic opinion
and recommendations for management [1].
The primary aim of this article is to determine the utility
of brief cognitive screening tests and neuropsychological
testing in the detection and diagnosis of dementia, whereas
a secondary aim is to define the role of neuropsychological
300 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317

Table 1
Studies of brief cognitive tests for the detection of MCI
Study Reference Test Time to Sample Setting Findings Evidence Methodologic
Administer Description Level Observations

Nasreddine et al [15], MoCA ⱕ10 min 94 aMCI, 93 AD, Two memory clinics SE/SP, 90%/87% II-2 Samples from different
2005 90 normal (MMSE, 18%/ clinics; reliability
100%) estimate; education
correction
Kalbe et al [16], 2004 DemTect ⱕ10 min 97 aMCI, 121 Memory assessment SE/SP, 80%/92% II-2 Age & education
AD, 145 center (MMSE, 69%/ correction;
normal 77%) transformation
algorithm into
MMSE
Loewenstein et al [17], 3-trial delayed ⱕ5 min with 24 MCI, 26 Community memory SE/SP, 83%/90% II-2 Inclusion limited to
2000 recall of 3 5-min dementia, 52 screening program (MMSE, 71%/ subjects for whom
MMSE intervals normal 85%) diagnostic consensus
words existed between
neuropsychologist
and neurologist
Tang-Wai et al [18], STMS ⱕ10 min 129 aMCI, 788 Community sample & AUC, 0.74 III Retrospective series;
2003 normal, 235 referrals to Mayo (MMSE, 0.70) test scores available
AD Clinic ADRC to clinicians
Eibenstein et al [19], SSST ⱕ10 min 29 aMCI, 29 Alzheimer Evaluation AUC, 0.84; SE/SP, II-2 Test awkward to
2005 normal Unit 84%/72% administer (patient
needs to fast); small
series

Abbreviations: AUC, area under the curve; SE, sensitivity; SP, specificity.

assessment in the management of mild cognitive impair-
ment (MCI) and dementia. Brief cognitive tests and neuro-
psychological testing can be appropriately applied across a
range of clinical/diagnostic issues. They might be called on
to discriminate MCI from normal aging, to estimate the risk
of progression from MCI to dementia, and to identify the
etiology, severity, and prognosis of dementia. The current
article aims to provide evidence-based recommendations on
the optimal use of brief cognitive screening tests and neu-
ropsychological testing.
2. Methods
This evidence-based review included studies that as-
sessed the discriminative validity of brief cognitive tests and
neuropsychological tests in the normal aging to dementia
spectrum. Studies were identified by searching the PubMed
and Embase databases with the following search strings:
(1) “Questionable dementia” OR “Cognitive Impairment”
OR “Cognitive-Impairment-Not-Dementia” OR “Mild Cog-
nitive Impairment” AND “Diagnosis” AND “Cognitive
tests” OR “Screening tests” OR “Neuropsychological tests”;
(2) “Dementia” OR “Senile dementia” OR “Alzheimer’s
Disease” AND “Diagnosis” AND “Cognitive tests” OR
“Screening tests” OR “Neuropsychological tests”. Prefer-
ence was given to empirical studies that (1) were published
between 1996 and 2005; (2) were cross-sectional cohort
studies (evidence level II-2), prospective cohort studies
(evidence level II-2), or retrospective case-control studies
(evidence level III); and (3) reported on sensitivity and
specificity, odds ratios, or relative risks. Where needed,
older studies, studies reporting group differences, and clin-
ical qualitative studies were also consulted. For studies on
neuropsychological testing a further selection was applied
by including only those that evaluated a battery of tests
covering different domains, as is done in the standard clin-
ical application of neuropsychological testing, not studies of
single neuropsychological tests or domains. Detailed infor-
mation on all included empirical studies is presented in
summary Tables 1 to 6. Meta-analyses, review articles, and
position papers were also considered.
3. Brief cognitive tests
Brief cognitive tests are designed to address memory/
cognitive complaints and to objectively identify cognitive
impairment that warrants diagnostic work-up and treatment.
The sensitivity and specificity of these tests are not fixed
properties but vary with the setting. It is often the case that
the sensitivity and specificity of a test are estimated in the
specialty clinic setting, with samples of convenience.
These estimates might be inflated, particularly with re-
gard to sensitivity, because cases are preselected and
potentially more severe [2]. The test might perform less
well in other settings including general practice. It has
been proposed that population-based studies are better
able to estimate a test’s sensitivity and specificity in
general practice than clinic-based studies [3].
Table 2
Studies of brief cognitive tests for the detection of dementia and its subtypes
Study Reference Test Time to Sample Description Setting Findings Evidence Methodologic Observations
Administer Level

Tombaugh et al [6], 3MS ⱕ20 min 119 AD, 406 not Community sample 3MS overall comparable to II-2 Large population-based sample, good
1996 cognitively impaired MMSE; AUC, 0.93 (MMSE, reliability; age- and education-specific
0.91); 3MS possibly better accuracy estimates and norms
for low education
Powlishta et al [28], CDT ⱕ5 min 60 AD, 15 normal Study of healthy aging Very mild dementia (CDR 0.5) III Retrospective; examines CDR ⫽ 0.5, 1, ⱖ2
2002 and AD vs no dementia; SE/SP, separately; estimates for six scoring
20%–60%/60%–93% methods
De Lepeleire 7 tests (brief to ⱕ10 min 50 Dementia, 101 no GP offices AUC, 0.92 for functional index II-2 Assesses diagnostic gain of instruments
et al [29], 2005 NPT computer dementia followed by CDT ranked from least to most burdensome
battery)

C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317

Nasreddine et al [15], MoCA ⱕ10 min 93 AD, 94 aMCI, 90 Two memory clinics AD vs normal: SE/SP, 100%/ II-2 Samples from different clinics; good
2005 normal 87% (MMSE 78%/100%) reliability; education correction
Tang-Wai et al [18], STMS ⱕ10 min 235 AD, 129 aMCI, Community sample & AD vs normal: AUC, 0.96 III Retrospective series; test scores available to
2003 788 normal referrals to Mayo (MMSE, 0.94) AD vs MCI: clinicians; discrimination of AD from
Clinic ADRC AUC, 0.86 (MMSE, 0.85) MCI
Kalbe et al [16], 2004 DemTect ⱕ10 min 121 AD, 97 aMCI, 145 Memory assessment AD vs normal: SE/SP, 100%/ II-2 Age/education correction; transformation
normal center 92% (MMSE, 92%/86%) algorithm into MMSE; discrimination of
AD vs MCI: SE/SP, AD from MCI
85%/81%
Kuslansky et al [33], MIS ⱕ5 min 28 AD, 212 no Community participants AD vs no dementia (includes II-2 Replication study in community sample;
2002 dementia in Einstein Aging CDR 0 & 0.5): SE/SP, strong theoretical foundation for testing
Study 86%/97% paradigm
Solomon et al [31], 7-Minute Screen ⱕ10 min 60 AD, 60 normal Memory clinic SE/SP, 92%/96% II-2 1000⫻ randomly resampled runs; reliability
1998 community estimates; but polarized groups (clinic-
referred AD vs community normal)
Loewenstein et al [17], 3-trial delayed ⱕ5 min with 26 Dementia, 24 MCI, Community memory SE/SP, 92%/94% (MMSE, II-2 Inclusion limited to subjects for whom
2000 recall of MMSE 3 5-min 52 normal screening program 100%/85%) diagnostic consensus existed between
words intervals neuropsychologist & neurologist
Brodaty et al [34], GPCOG ⱕ5 min 82 Dementia, 50 no GP offices Dementia vs no dementia: SE/ II-2 Test designed/validated in GP setting in
2002 dementia but meeting SP, 82%–85%/83%–86% groups with graded severity; requires
some DSM criteria, (MMSE, 81%/76%) informant in ⬃75% of cases; good
20 normal reliability
Kilada et al [35], 2005 Blessed Memory ⱕ5 min with 378 Dementia, 378 BLSA, Washington Dementia vs normal/ III Retrospective study of samples from two
Test & animal brief delay normal University ADRC nondemented: SE/SP, 71%/ large cohort studies; dementia with CDR
fluency 98% (MMSE, 44%/98%) ⬎1 excluded; follow-up of false
negatives
Storey et al [36], 2004 RUDAS ⱕ10 min 45 Dementia, 45 normal Community SE/SP, 89%/98% II-2 Sound test development for cross-cultural
(culturally purpose, good reliability; insensitive to
heterogeneous sample demographic factors
w/ESL subjects)
Darvesh et al [37], BNA Short Form ⱕ20 min 29 Dementia, 29 Behavioural Neurology SE/SP, 93%/93% (MMSE, II-2 Polarized groups; validation of short form
2005 matched normal Clinic (patients), 79%/97%) within long form; good reliability
community (normal)

301
Abbreviations: AUC, area under the curve; BLSA, Baltimore Longitudinal Study of Aging; ESL, English as a second language; SE, sensitivity; SP, specificity.
 302
Table 3
Neuropsychological studies of discrimination between normal aging, MCI, and early dementia
Study Reference Sample Description Setting Design Findings Evidence Methodologic Observations
Level

Peters et al [43], 205 CIND, 68 NCI Dementia clinic Cross-sectional Measures of learning, memory, visuoconstruction, II-2 NCI subjects were clinic-referred, hence
2004 (ACCORD) & cognitive flexibility discriminate between not normal controls: discriminative
CIND and NCI with 75%–78% accuracy; inter- ability of NPT might be
test scatter and impairment count indicate underestimated and not generalizable
substantial heterogeneity to population at large
Peters et al [44], 461 CIND (CSHA), Community (CSHA), Cross-sectional Five subgroups characterized by distinct II-2 Use of base (CSHA) and replication
2005 166 CIND dementia clinic neuropsychological impairment patterns (ACCORD) sample for identification
(ACCORD) (ACCORD) of subgroups
Lopez et al [45], 38 MCI, 374 Community (CHS) Cross-sectional MCI multiple cognitive deficits type (74%) more II-2 Neuropsychological criteria for
2006 normal frequent than MCI amnestic type (26%) classification into MCI subtypes not

C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317

independently validated (eg,
clinically, by neuroimaging)
Petersen et al [10], 76 aMCI, 48 mild Mayo AD Center, Cross-sectional aMCI & mild AD impaired on memory measures II-2 aMCI classified clinically,
1999 AD, 234 normal community (⬃1.5 SD below normal subjects); aMCI subtly independently of NPT; comparison of
impaired on non-memory measures (⬃0.5 SD), aMCI is with very mild AD (CDR ⫽
mild AD more substantially impaired (ⱖ1 SD) 0.5 for both groups)
Grundman et al [46], 769 aMCI, 107 ADCS centers Cross-sectional aMCI had prominent memory impairment (⬎1 SD); II-2 Very large and carefully characterized
2004 normal (aMCI from more subtle impairments (⬍1 SD) on language, cohort of aMCI
ADCS-MIS trial) attention, and executive functioning tasks
Kramer et al [47], 22 aMCI, 33 mild UCSF Memory and Cross-sectional ⬎50% of aMCI scored ⱖ1 SD below controls on II-2 aMCI classified according to Petersen
2006 AD, 35 normal Aging Center ⱖ4 non-memory measures including executive criteria plus CDR impairment only on
functioning and fluency tasks memory and orientation domains
Steenhuis et al [53], 450 CIND, 219 Community (CSHA) Cross-sectional Memory, reasoning, and fluency tasks differentiated II-2 Final diagnoses not independent of
1995 dementia, 591 CIND from normal (77% accuracy); memory, NPT, although same pattern found for
normal fluency, and visuospatial tasks differentiated preliminary clinical diagnoses
dementia from CIND (82% accuracy)
Lambon Ralph et al 18 MCI, 38 AD Memory Clinic Longitudinal (2 y) Stage pattern of NP deficits in MCI⬍mild II-2 Cross-sequential analysis of natural
[54], 2003 AD⬍moderate AD: episodic memory history combining cross-sectional and
deficits⬎semantic memory deficits⬎language longitudinal data
deficits; longitudinal change in MCI and AD
identical, albeit starting at different points
Rubin et al [57], 82 cognitively Community Longitudinal (12 y) Change in general cognitive factor regression-fitted; II-2 Serial assessments; only those before
1998 healthy CDR ⫽ intercept but not slope significantly different CDR change were included in
0, 27 w/CDR ⬎0 between stable group and group that developed regression; visual representation of
at f/u CDR ⬎0 cognitive/CDR change
Amieva et al [58], 1,265 nondemented Population-based Longitudinal (9 y) Lower initial scores and accelerated decline ⬃3 y II-2 Serial assessments entered into linear
2005 subjects, 215 AD study (PAQUID) before diagnosis in episodic memory, abstract mixed model with modification to
at f/u reasoning & semantic fluency; decline was allow for non-linear effects; separate
steeper for highly educated subjects in analyses for low and high education
acceleration phase
Chen et al [59], 551 nondemented Community Longitudinal (3.5 y) Rate ratios (cases/controls who declined ⱖ1 SD on II-2 No comparison between predictive
2001 subjects, 68 AD test): executive function, 3.7; verbal learning & utility of baseline and change scores
at f/u recall, 2.6–3.6; praxis, 2.5; naming, 2.1

Abbreviations: f/u, follow-up; NCI, not cognitively impaired.

Table 4
Neuropsychological studies of prediction of dementia or AD
Study Reference Sample Description
Artero & Ritchie [66], 308 Subjects w/cognitive
2003 complaints, 19 AD at f/u
Chen P et al [59], 603 nondemented subjects,
2001 120 AD at f/u
Blackwell et al [69], 43 Subjects with questionable Memory clinic
2003 dementia, 11 AD at f/u
Tierney et al [63], 551 nondemented subjects, 77 Population-based study 5 or 10
2005 AD at f/u [5-y study]; 263
nondemented subjects, 47
AD at f/u [10-y study]
Tierney et al [64], 123 subjects with memory GP referred to study
1996 impairment, 29 AD at f/u
Elias et al [67], 2000 1043 nondemented subjects
with 106 AD at f/u
Howieson et al [68], 47 normal subjects, 16 AD at Population-based study ⱖ2
1997 f/u
Palmer et al [70], 121 subjects with memory
2003 complaints, 38 dementia at
f/u
Bowen et al [62], 21 subjects with isolated
1997 memory loss, 10 dementia
at f/u
Devanand et al [71], 62 subjects w/questionable
1997 dementia, 31 dementia at
f/u
Fabrigoule et al [76], 1159 nondemented subjects,
1998 16 incipient AD, 9 other
dementia
Setting Follow-up Follow-up Findings Evidence Methodologic Observations
Duration (y) Diagnosis Level
GP referred to study 2 AD Delayed story recall, semantic II-2 Heterogeneous (MCI, normal) and
(Eugeria Study) fluency, copying abstract figure: insufficiently characterized
SE/SP, 73%/99% cohort at baseline
Population-based study 1.5 AD Delayed recall ⫹ Trails B: AUC, 0.83 II-2 Heterogeneous (MCI, normal) and
(MoVIES) insufficiently characterized
cohort at baseline; test scores
available for diagnosis
2.5 AD Visuospatial associative learning ⫹ II-2 Battery covers only episodic and
confrontation naming: 100% semantic memory
accuracy
Delayed verbal recall ⫹ semantic
C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
AD II-2 Heterogeneous cohort composed
(CHSA) fluency ⫹ information (5-y f/u): of NCI and CIND; predictive
SE/SP, 74%/83% Delayed verbal parameters validated by
recall (10-y f/u): SE/SP, 73%/70% bootstrapping
2 AD Delayed verbal recall ⫹ mental II-2 Model validated on subsample
control: SE/SP, 76%/94% w/matched MMSE; research
NPT battery different from
clinical NPT battery
Population-based study ⱖ5 or ⱖ10 AD (y of AD after ⱖ5 y: logical memory II-2 Very long surveillance period
(Framingham Study) diagnosis) retention. OR 1.53; similarities, OR (22 y); Cox regression for
1.32; paired associate learning, OR prediction of time to diagnosis;
1.29 AD after ⱖ10 y: logical NPT battery used for later
memory retention, OR 1.17; diagnosis
imilarities, OR 1.35
AD/questionable Logical memory ⫹ naming: accuracy, II-2 Cohort of very old subjects (ⱖ80
(Oregon Brain dementia 83% y of age); diagnosis CDR-based
Aging Study) and inclusive of questionable
dementia (CDR 0.5)
Population-based study 3 Any dementia Episodic memory impairment: SE/SP, II-2 Investigates the added value of
(Kungsholmen) 50%/89%; verbal fluency NPT tests over MMSE score
impairment: SE/SP, 48%/92%; but very limited NPT battery
visuospatial impairment: SE/SP,
26%/46%
HMO registry 4 Any dementia No NPT test predictive of diagnosis II-2 Very small sample; NPT results
available for clinical diagnosis
Memory Disorders ⱕ7 Dementia (87% SRT long-term retrieval, digit symbol, II-2 NPT results available for clinical
Clinic AD) picture arrangement, block design, diagnosis; high rate of
semantic fluency: SE/SP, 67/67% reversion (41%)
for demented; SE/SP, 61%/78%
for AD
Population-based study 2 y AD and General cognitive factor reflecting II-2 Principal component analysis of
(PAQUID) dementia executive control processes: OR ⫽ NPT scores to address
2.3 for dementia, OR ⫽ 3.3 for AD colinearity problem; very low
number of incident cases
303

Table 4
Continued
Study Reference Sample Description
Arnaiz et al [65], 303 subjects with MCI
2004 (number of incident AD
not stated)
Tian et al [75], 2003 129 subjects w/questionable
dementia, 37 dementia
at f/u
Bozoki et al [72], 48 subjects with isolated
2001 memory impairment, 16
AD at f/u
Peters et al [44], 2005 CIND subjects
Ritchie et al [77], 283 subjects with recent
1996 cognitive decline, 19
dementia at f/u
Herlitz et al [74], 285 nondemented subjects, 70 Population-based study 3
1997 dementia at f/u
Abbreviations: AUC, area under the curve; f/u, follow-up; NCI, not cognitively impaired; OR, odds ratio; SE, sensitivity; SP, specificity; SRT, Selective Reminding Test.
Setting Follow-up Follow-up Findings
Duration (y) Diagnosis
Mayo Clinic Max 5–10 y AD Number of impaired domains (⫺1.5
SD): SE/SP, 80%/75%; delayed
recall best predictor in both
samples
Memory Disorders 2 Any dementia Predictive models unstable; better w/
Clinic categorical than continuous NP
variables; total immediate recall
most consistent predictor SE/SP,
⬃53%–64%/67%–83%
Michigan ADRC 2 AD Subjects classified into M⫺ (memory
database impairment only) and M⫹
(additional non-memory
impairment), relative risk 4.6 for
M⫹; no single test predictive
Community (CSHA), 2–5 y Any dementia Membership in episodic memory
dementia clinic dysfunction subgroup: OR 2.27 to
(ACCORD) 3.48
GP research network 3 Any dementia Membership in general cognitive
(Eugeria) impairment with language
dysfunction (OR 4.4) or memory
impairment (OR 3.9) cluster
Any dementia & Episodic memory measures & Trails-
(Kungsholmen) AD based dementia diagnoses w/out
DSM confirmation had 65% higher
DSM dementia at f/u than test-
based normal diagnoses
304
Evidence Methodologic Observations
Level
II-2 Cross-national study; outcome is
time to diagnosis with Cox
C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
regression; model validated in
each sample separately
II-2 Clinically well characterized
cohort; Cox regression for
prediction with different
variants to establish stability
III Retrospective database review;
repeated NP test results
available to treating clinician
II-2 Use of two independent samples
for cluster solution and
prediction
II-2 Preliminary report with few cases
and non-completion of
surveillance period; no
independent sample for cluster
validation
II-2 Analysis of discordant diagnostic
classifications based on NPT
and DSM
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317 305

Table 5
Neuropsychological differentiation of AD from depressive dementia, DLB, and PDD
Comparisons Study Reference & Sample Description Findings Conclusion (Evidence Level in Parentheses)
Design

AD vs depressive Christensen et al [153], Reviewed 154 articles Depressives ⬎ AD: paired associate Depression has widespread effects on
dementia 1997, meta-analysis with 225 learning task, anomalous sentence cognitive functioning including
comparisons repetition test, Boston naming & intelligence, problem solving, & memory;
WAIS block design; tests depressives do better on untimed tasks
requiring little cognitive capacity (II); clinical examination & depression
or effort (eg, nonspeeded tests inventories are basic for diagnosis (III)
superior to speeded tests); no
convincing difference between
the groups on recall vs
recognition task
AD vs DLB, Collerton et al [93], Reviewed 21 articles: AD ⬎ DLB on visuoperceptual/ Composite measures (eg, MMSE) might
PDD 2003, meta-analysis 312 DLB vs 380 semantic tasks regardless of underestimate severity of DLB because
on DLB AD, 48 DLB vs disease stage or diagnostic of inadequate testing of attentional,
65 PD criteria, attentional/executive/ executive, or visuoperceptual
timed motor responses performance; memory deficits of DLB
Simard et al [154], Reviewed 68 studies DLB ⬍ AD on spatial working are less severe than their visuoperceptual/
2000, review of on cognitive & memory (eg, Trails B), attentional/executive deficits (II); similar
DLB behavioral changes visuospatial function (eg, more cognitive profiles in DLB & PDD
in DLB impaired on clock copy vs free probably reflect superimposition of
drawing) subcortical deficits on those typically
Aarsland et al [83], Mild to severe Mild-moderate DLB ⬍ PDD associated with AD
2003, cross-sectional dementia patients: conceptualization subscore;
60 DLB, 35 PDD, severe DLB ⫽ PDD on all
49 PSP, 29 AD; subscores; PDD, DLB, & PSP ⬎
comparison on AD on memory subscores; AD ⬎
Mattis dementia PDD & DLB on
rating scale initiation/perseveration &
construction subscores; AD ⬎
DLB on conceptualization
subscore; PSP ⬎ PDD & DLB
on memory subscores

Abbreviations: PSP, progressive supranuclear palsy; WAIS, Weschler Adult Intelligence Scale.

By design, brief tests have a limited measurement range,
and both ceiling and floor effects can threaten their validity.
Ceiling effects are a particularly important threat when mild
symptoms are being evaluated. For example, the widely
used 30-item Mini-Mental State Examination (MMSE) [4]
has lacked sensitivity to mild degrees of impairment [5].
With the conventional cutoff level at ⬍24 for dementia, the
test has only a 6-point scale for discrimination between MCI
or very mild dementia and normal functioning, with score
overlap and skew as a result of the test’s ceiling. The
MMSE has become a benchmark against which newer tests
are compared to establish their improved sensitivity to mild
degrees of impairment.
When using brief cognitive tests, it is important to realize
that no single test has utility along the entire spectrum of
severity. Different tests will need to be chosen depending on
whether MCI or dementia is suspected. Some brief cogni-
tive tests have been specifically developed to detect Alzhei-
mer’s disease (AD) and might not be able to detect other
dementia subtypes because of their emphasis on memory
abilities. It is also critical to understand that performance on
most brief cognitive tests is affected by demographic vari-
ables, primarily age and education [5,6]. Typically, perfor-
mance is lower in older age and lower education groups.
When single cutoff scores are used across different age
and education levels, the evident risk is an increased rate
of false-positive classifications among older subjects and
those with lower education. However, test users should
be mindful that a “correction” for age and education
effects, through application of different cutoff scores or
adjustment, also carries risks. Both old age and low
education have consistently been associated with an in-
creased risk for cognitive impairment and dementia [7].
The application of adjusted cutoff scores can reduce the
sensitivity of tests, with reduced identification of true
positives in these at-risk groups [8].
3.1. Brief cognitive tests for the detection of MCI
MCI is a diagnostic label that has been variously defined
but that generally refers to individuals with memory and/or
cognitive impairment not sufficiently severe to fulfill de-
mentia criteria. In the current article, we apply the term MCI
in this general sense, thus covering many proposed specific
306 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317

Table 6
Neuropsychological differentiation of AD from FTD and VaD
Comparisons Study Reference & Sample Description Findings Recommendation
Design

AD vs FTD Kramer et al [155], 2003, 21 FTD, 14 SD, 30 AD AD and SD ⬍ FTD on Tests of verbal memory, language,
AD vs PPA case control 52 FTD-bv, 52 AD verbal memory and and executive function
AD vs SD Kertesz et al [106], 2003, 26 FTD naming; FTD worse on discriminate AD from FTD, but a
case control executive function behavioral inventory improves the
Mathunarath et al [156], FTD ⬎ AD on memory, but distinction - Level II evidence;
2000, case control lower language scores; A language/memory ratio (ACE) and
FBI best discriminates the the FBI can be adapted to a
groups clinical community service
[verbal fluency ⫹ language]/ - Level III and II evidence
[orientation ⫹ memory] Formal language testing is helpful,
ratio best discriminates but clinical experience is
the groups necessary for interpretation.
Kertesz et al [106], 2003, 67 PPA, 99 AD Anomia at start; change in Detailed language and semantic
case control 11 NFPA, 10 SD, 11 fluency is stage testing may be necessary for
Gorno-Tempini, 2004 logopenic progressive dependent; logopenia, accurate distinctions
[157], case control aphasia aphemia, Brocha type
Hodges et al [158], 1999, 9 FTD, 9 AD, 9 SD speech in PPA; Wernicke
case control type speech in AD
Apraxia and agrammatism in
NFPA; fluent semantic
memory deficit in SD;
anomia in LP
Naming and semantic
knowledge in SD
AD vs VaD Looi et al [87], 1999, Meta-analysis of 27 Better verbal memory and AD is different from VaD on
VAD vs FTD meta-analysis articles worse executive function neuropsychological testing, but the
AD vs Huntington, Sjogren et al [159], 1996, 11 VAD, 21 FTD in VaD differences are stage-related and
PSP, CBD case control Behavior worse in FTD subtle. Mixed dementia resembles
whereas memory is more AD. Neuropsychological testing is
impaired in VAD, but supportive, to be used in specialist
there is no overlap setting only
VaD and FTD overlap, but memory
and language tests and behavioral
quantification helps - Level II
Butters et al [160], 1985, Many case control studies Subcortical dementia: frontal
evidence
case control and reviews on deficit; recognition
The concept of subcortical dementia
subcortical dementia memory better than recall;
is useful and can be supported by
CBD has apraxia and
neuropsychology - Level II
aphasia
evidence. Most of these conditions
are diagnosed by the movement
disorder

Abbreviations: ACE, Addenbrooke’s Cognitive Examination; CBD, corticobasal degeneration; FBI, frontal behavior inventory; NFPA, progressive
nonfluent aphasia; PPA, primary progressive aphasia; SD, semantic dementia.

definitions [9]. We further distinguish the entities of amnes-
tic MCI (aMCI) and Cognitive Impairment Not Dementia
(CIND). The term aMCI was developed to capture the
preclinical stage of AD, with emphasis on memory loss
within the context of otherwise normal cognitive function-
ing [10]. The term CIND was developed for population
epidemiologic studies to denote any state of cognitive func-
tioning that falls outside of normal range yet does not meet
dementia criteria [11]. Depending on the definition that is
used, the prevalence of MCI ranges from as low as 1% [12]
to ⬃40% in the population and 30% in clinic-referred sam-
ples [13]. The rate of progression from MCI to dementia is
10% to 15% per year in clinic-referred samples [9] and
appears similar in population samples [12]. This MCI pro-
gression rate compares with 1% to 2% among matched
healthy elderly [9]. An American Academy of Neurology
Practice Parameter has recommended the identification and
further clinical evaluation of MCI [14].
Recently a number of new instruments have been
developed for the detection of MCI: the Montreal Cog-
nitive Assessment (MoCA) [15], the DemTect [16], a
three-trial delayed recall adaptation of the MMSE [17],
the Short Test of Mental Status (STMS) [18], and the
Sniffin’ Sticks Screening Test (SSST) [19]. Table 1 sum-
marizes studies of these five tests, with clinically diag-
nosed MCI as the standard.
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
The MoCA is a 30-point test that covers multiple cog-
nitive domains. The MoCA has been found to have good
internal consistency and test-retest reliability and was able
to correctly identify 90% of a large sample of aMCI subjects
from two different clinics [15]. The DemTect, scored on an
18-point scale, has similar coverage of domains. Although
its reliability remains to be determined, its discriminative
validity was superior to the MMSE in a large clinic sample
of aMCI patients, compared with normal controls. DemTect
scores can be transformed into MMSE by means of a simple
algorithm [16]. Three delayed recall trials of the three
MMSE items at 5-minute intervals adding to total of 9
points compared favorably with the total MMSE score in a
community sample of subjects classified as MCI and those
deemed cognitively normal [17]. There was a low diagnos-
tic agreement rate (57%), and cases in which no consensus
diagnosis could be reached (29%) were excluded, suggest-
ing problems with the reliability of the gold standard in this
study. Evaluation of this repeated recall paradigm is there-
fore not possible until further studies are carried out. The
38-point STMS is an expansion of the MMSE in the do-
mains of learning/recall and abstract reasoning. This mea-
sure did not perform better than the MMSE in separating
aMCI from normal, both achieving an area under the curve
of ⬃0.70 [18].
Because AD neuropathology affects the limbic system
that is involved in olfactory function, deficits in olfaction
have been investigated as a diagnostic marker of preclinical
AD [20,21] and aMCI. The SSST was developed as a
bedside odor identification test for early detection of AD.
The SSST separated aMCI patients from normal controls
with good accuracy (area under the curve, 0.84), but
cognitive tests tend to be more accurate, with both higher
sensitivity and specificity. In addition, the test is awk-
ward to administer, with the subject having to fast before
the test [19].
3.2. Brief cognitive tests for the detection of dementia and
its subtypes
Table 2 summarizes 13 studies that have examined brief
cognitive tests for the detection of AD or dementia in
general. In the development of these instruments the aim has
been to increase sensitivity levels to early or mild dementia
above those achieved with the MMSE. The Modified Mini-
Mental State Examination (3MS) [22] was developed as a
test with increased sensitivity to dementia by adding tasks
known to be impaired early in the disease, including delayed
recall and a verbal fluency test for the semantic category of
animals. In the 3MS both ceiling and floor of the MMSE are
extended, and the range of possible scores is increased from
30 to 100. The 3MS is quite widely used in clinic and
research settings. It served as a screening test for dementia
and referral for clinical assessment in the Canadian Study of
Health and Aging (CSHA) [23], allowing it particular fa-
307
miliarity in Canada. Tombaugh et al [6] have compared the
3MS and MMSE in CSHA participants who were diagnosed
with AD or cognitively intact. The 3MS did not have sig-
nificantly improved sensitivity over the MMSE, but its
accuracy tended to be about 5% higher in subjects with low
education (0 to 8 years).
The Clock Drawing Test (CDT) has achieved wide-
spread clinical utilization and has been studied extensively
[24]. A Canadian survey found that primary practitioners
considered the CDT an acceptable cognitive screening test
[25]. A review of validation studies identified consistently
good levels of sensitivity and specificity for the CDT in the
early detection of dementia, irrespective of the scoring sys-
tems that were used [26]. Recent evidence, however, invites
caution because sensitivity was found to be very low for
most scoring systems when dementia was very mild, at a
Clinical Dementia Rating (CDR) global stage [27] of 0.5
[28]. CDT scores added significantly to an overall 92%
accuracy in a hierarchical classification algorithm that was
developed for use in general practice and that used age, sex,
a functional status measure, and the CDT in a stepwise
fashion [29]. Earlier studies have also suggested that the
CDT might perform better in combination with other instru-
ments, for example, the MMSE [30].
New instruments for the detection of MCI have also been
investigated either with AD or with dementia in general as
the standard. Whereas the MoCA and DemTect had perfect
sensitivity for AD, performing better than the MMSE
[15,16], the STMS was comparable to the MMSE [18]. The
7-Minute Screen (7MS) is a multiple domain test with very
high sensitivity and specificity for AD, which was calculated in
1000 classification iterations with random subsamples [31].
These measurement properties were subsequently con-
firmed in a primary care setting [32]. The Memory Impair-
ment Screen (MIS) is a theoretically grounded attempt at
disentangling deficits caused by AD from those caused by
normal aging. This 8-point scale had high sensitivity and
specificity (⬎85%) for AD in a community sample includ-
ing nondemented controls, some with a CDR of 0.5 [33].
The three-trial delayed recall adaptation of the MMSE per-
formed similarly well in the discrimination of dementia of
any cause (sensitivity and specificity, ⬎90%), but caution is
required regarding the gold standard in this study as a result
of the aforementioned problems with diagnostic agreement
and subject exclusion [17].
The General Practitioner Assessment of Cognition
(GPCOG) is a 15-point test specifically developed for
primary care. A patient section and potentially an infor-
mant section are administered in a two-step procedure.
This instrument discriminated those who met Diagnostic
and Statistical Manual of Mental Disorders, Fourth Edi-
tion (DSM-IV) criteria for dementia from those who did
not with good levels of sensitivity and specificity
(⬎85%). Nearly 40% of false positives in this study were
subjects who met some but not all DSM-IV criteria for
308 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
dementia [34]. A brief paradigm consisting of the Blessed
Memory test plus animal fluency has also been found to
have better sensitivity than the MMSE at equal specificity in
retrospective studies of two large cohorts, with exclusion of
cases with CDR ⬎1 [35].
The Rowland Universal Dementia Assessment Scale
(RUDAS) was designed to detect dementia in cross-cultural
settings. The RUDAS items were developed with endorse-
ment of cross-cultural and cross-discipline advisory panels.
The test performed well (sensitivity and specificity, 89%
and 98%, respectively) in a culturally heterogeneous
community sample in Australia, with more than 50%
having English as a second language and/or with educa-
tion ⬍6 years [36]. The potential of detection bias caused
by Western culture idiosyncrasies has been recognized
for some MMSE items, and the RUDAS might offer a
valid alternative in the assessment of subjects with non-
Western background.
The Behavioural Neurology Assessment (BNA) Short
Form is a 114-point scale that assesses multiple domains but
requires greater administration time than most other brief
cognitive tests. The BNA has been found more sensitive to
dementia than the MMSE [37] and raises the question of
whether longer assessments generally have added utility that
justifies their extra time expenditure. This specific question
was addressed in a comparative study of three brief tests
(MMSE, Orientation-Memory-Concentration [OMC] test,
Mental Status Questionnaire [MSQ]) and two abbreviated
test batteries (Dementia Rating Scale [DRS], Ottawa Mental
Status Examination [OMSE]). The longer tests did not offer
advantages over the brief tests in terms of their reliability
and sensitivity to dementia of any type. Of note, no test or
mini-battery was able to reliably differentiate AD from
vascular dementia (VaD) [38].
The discrimination between MCI and dementia is critical
for monitoring and treatment decisions, yet this issue is
rarely addressed in the evaluation of brief cognitive tests.
Only the DemTect and the STMS have been examined for
this purpose. Although both were fairly accurate in separat-
ing AD from aMCI, with ⬃85% accuracy and sensitivity/
specificity [16,18], the ability of all other instruments to
make similar distinctions remains to be determined. On the
basis of the available evidence it is not possible to determine
whether this discrimination can in fact be reliably achieved
by brief cognitive tests, or whether it is an issue that might
have to be deferred to a neuropsychological evaluation.
In summary, most brief cognitive tests can detect mild
dementia accurately, with the possible exception of the
CDT. Where newer tests have been directly compared with
the MMSE, the MoCA, the DemTect, the BNA, and the
GPCOG but not the 3MS appear to provide better accuracy
as a result of their higher sensitivity. Among these the
GPCOG is the only instrument that has been validated
directly in primary care settings. Tests developed and vali-
dated solely with AD samples might not perform equally
well with heterogeneous dementia samples that include
non-AD subtypes. In this latter context, tests sampling mul-
tiple domains are likely to perform better than those
weighted toward or sampling only episodic memory (eg,
MIS, three-trial delayed recall). At present, no brief cogni-
tive test has been developed to differentiate between sub-
types of dementia, and because of the complexities of these
distinctions, it seems unlikely that one will emerge [38].
4. Neuropsychological testing
There is no unitary view on the role of neuropsycholog-
ical testing (NPT) in the diagnostic work-up of dementia.
NPT is typically performed after administration of a brief
cognitive test and other clinical work-up to probe the func-
tioning of memory and other cognitive domains with stan-
dardized tasks. A systematic study of the usefulness of the
1994 American Academy of Neurology practice parameters
for the evaluation of dementia found evidence that NPT has
utility in the assessment of MCI and in cases in which
memory impairment is not a predominant feature [39]. A
background article to the Second Canadian Consensus Con-
ference on Dementia concluded that although NPT might
have utility in borderline cases, it is not required in the
routine diagnostic work-up of suspected dementia [40]. The
National Institute of Neurological and Communicative Dis-
orders and Stroke—Alzheimer Disease and Related Dis-
orders (NINCDS-ADRDA) criteria require confirmatory
evidence of impairment on neuropsychological tests for the
diagnosis of probable AD [41]. At a practical level, NPT
involves often long administration times and is frequently a
limited resource, not widely accessible to the clinician.
In this article we propose on the basis of current evidence
that the diagnosis and differential diagnosis of dementia
must be considered a clinically integrative process. NPT
alone should not be used for these purposes. Instead, NPT
should be applied selectively to aid in (1) addressing the
distinction between normal cognitive functioning in the
aged, MCI, and early dementia; (2) addressing the risk of
progression to dementia in persons with MCI diagnoses;
and (3) the differential diagnosis of dementia and other
syndromes of cognitive impairment. We evaluate the
evidence in support of each of these applications in the
following paragraphs. In our evaluation of studies that
support the utility of NPT in the differential diagnosis of
dementia we have chosen to take a clinically based qual-
itative approach aimed at delineating differential NPT
profiles for the dementias, rather than examining general
accuracy parameters.
4.1. The distinction between normal aging, MCI, and
early dementia
In the cognitive spectrum from normal aging to dementia,
NPT might be useful in the differentiation of the bordering
states of MCI/normal functioning and early dementia/MCI.
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
In each case, the dividing line is in part definable by criteria
based on neuropsychological test performance. MCI defini-
tions might require objective impairment in memory [33] or
in any of the cognitive domains of learning, memory, atten-
tion, thinking, language, or visuospatial function [42] for
diagnosis. The NINCDS-ADRDA clinical criteria for prob-
able AD require abnormal performance in two or more areas
of cognition including orientation, memory, language,
praxis, attention, visual perception, and problem solving
[41]. There is in consequence a degree of interdependence
between the clinical differentiation of MCI or dementia and
NPT. Beyond its role in the fulfillment of specific diagnostic
criteria, however, NPT can provide additional information
that might be useful in the discrimination of MCI and early
dementia. The reviewed studies are summarized in Table 3.
4.1.1. Distinguishing MCI
In broadly defined MCI groups there is considerable
heterogeneity with respect to neuropsychological test per-
formance. In the Canadian Cohort Study of Cognitive Im-
pairment and Related Dementias (ACCORD), subjects clas-
sified as CIND differed from those deemed not cognitively
impaired on a range of memory and non-memory measures,
with measures of learning and recall, visuoconstruction, and
cognitive flexibility providing the best discrimination be-
tween the groups [43]. However, performance patterns in
CIND were not uniform within and between individuals,
and cluster analysis identified five distinct subgroups, with
only two having memory impairment [44]. In the Cardio-
vascular Health Study (CHS) subjects were generally clas-
sified as MCI if they had impairment in any domain (mem-
ory and/or non-memory). Within this broad grouping, 74%
of subjects met the criteria for MCI-multiple cognitive def-
icits type and had neuropsychological impairments in at
least two domains, whereas the remainder met criteria for
aMCI, with memory impairment only and no other domains
impaired [45].
Studies comparing groups of aMCI patients with normal
controls on comprehensive NPT batteries have consistently
found that in addition to the expected episodic memory im-
pairment on learning and delayed recall measures, even in this
narrowly defined MCI group subtle difficulties are manifest on
tests of language (naming and fluency), attention, and execu-
tive functioning [10,46,47]. Whereas differences between
aMCI and controls were generally ⬍1 standard deviation (SD)
on non-memory measures [10,46], one study reported that
more than half of their aMCI subjects scored at least 1 SD
below control means on ⱖ50% of tests [47].
There is no generally accepted NPT battery for MCI.
However, the evidence underscores that regardless of the
specific MCI criteria, NPT should extensively cover a range
of cognitive domains. In addition, serial testing has been
recommended to establish the consistency over time of
memory and/or other cognitive impairments [48].
309
4.1.2. Distinguishing mild dementia from MCI
NPT can discriminate quite accurately (sensitivity,
ⱖ80%, specificity, ⱖ90%) between patients with AD or
dementia, even at a very mild stage, and cognitively normal
elderly [49 –51]. The utility of NPT in the discrimination
between MCI and mild dementia has been less systemati-
cally investigated, in part perhaps because the focus in this
discrimination is on the requirement that cognitive deficits
have an adverse impact on social and occupational func-
tioning, which defines the threshold for dementia according
to DSM criteria [52]. MCI groups generally score at levels
intermediate between those of normal controls and those of
patients with mild AD or dementia [10,47,53]. Episodic
memory impairment in aMCI has been reported to be sim-
ilar in severity to that found in mild AD, whereas perfor-
mance in other domains, particularly language, is closer to
that of controls. This gradient of severity, with non-memory
measures best differentiating between aMCI and AD, is
evident also in cross-sequential analyses of neuropsycho-
logical test performance in groups of aMCI and mild and
moderate AD patients. These analyses identified a staged
pattern of neuropsychological deficits, starting in the epi-
sodic memory domain, spreading to the semantic memory
and language domains with AD onset, and subsequently, as
patients progress to moderate AD, to the perception and
attention domain [54].
Evidence of progressive memory and cognitive loss is a
requirement for AD diagnosis in the NINCDS-ADRDA
criteria [41], and there is agreement among a number of
studies that rates of decline on serial NPT accelerate in the
3- to 5-year time window before AD diagnosis, with abrupt
decline shortly before a diagnosis can be rendered [55]. The
change point in preclinical AD has been defined as the time
before diagnosis, where on average the rate of decline
among those who will develop AD and those who will not,
will begin to diverge. On the basis of longitudinal data on
episodic memory test performance, this point was estimated
to occur 5 years before diagnosis [56]. A 12-year study
found that subjects whose CDR increased above 0 during
the surveillance period did not have steeper cognitive de-
cline before the CDR increase when compared with those
whose CDR remained 0. Abrupt and general deterioration of
performance was observed only at the time of CDR change
[57]. A 9-year study reported parallel cognitive change for
those who later developed AD at follow-up and those who
did not develop dementia until 3 years before diagnosis
when cognitive decline accelerated in the incident AD sub-
jects, particularly in those subjects with high education [58].
The ratio of incipient AD cases to controls who declined by
⬎1 SD during a 2-year period was 4 AD subjects for every
control subject on executive function tests, 2.5 to 3.5 on
episodic memory tests, and 2 on language tests [59]. Al-
though none of these longitudinal studies have specifically
addressed the prognostic/diagnostic utility of rates of de-
cline in MCI, evidence of significant loss on serial NPT in
310 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
these subjects, particularly in the memory, language, and
executive functioning domains, should be considered sup-
portive of development of AD.
There is some evidence to suggest that the detection of
preclinical change during short time periods (1 to 2 years) is
better when computer-administered tasks are used rather
than standard examiner-administered NPTs. For example,
1-year decline in aMCI subjects was detected by a comput-
erized continuous learning task but not by the standard
neuropsychological Consortium to Establish a Registry for
Alzheimer’s Disease battery [60]. Similarly the Cambridge
Neuropsychological Test Automated Battery paired associ-
ate learning task but not standard neuropsychological mea-
sures including the revised Wechsler Adult Intelligence
Scale and Wechsler Memory Scale–revised subtests, verbal
and category fluency, Rey Complex Figure, Austin Maze,
and Rey Auditory Verbal Learning test detected a linear
2-year decline in subjects with questionable dementia who
were retrospectively classified as deteriorating [61]. How-
ever, because neither study provided diagnostic validation
that the declining subjects had dementia or AD at the end of
the observation period, these findings must be considered
preliminary.
4.2. The prediction of progression to dementia
The neuropsychological features that predict progression
to dementia or AD in cohorts of nondemented elderly or
patients with MCI have been investigated extensively. The
typical research design has been to follow population-based
or clinic-referred cohorts for a number of years and to
compare the baseline neuropsychological performance of
those who progress to dementia or AD at some point in the
surveillance period with the performance of those who re-
main nondemented. The present review has identified 17
studies, summarized in Table 3. The majority are popula-
tion-based studies, and the length of follow-up ranges from
2 to 22 years.
The majority of studies have found that initial neuropsy-
chological scores predict progression to dementia and AD,
with one exception, likely as a result of the very small
sample size of the study [62]. Predictive accuracy has
ranged from ⬃80% to 100%, sensitivities from 53% to
80%, and specificities from 67% to 99%. Episodic memory
impairment is the most consistently identified predictor,
whether measured by delayed word recall [59,63– 65), de-
layed story recall [66 – 68], or associative learning [67,69].
This is true for samples of clinic-referred subjects with
MCI, questionable dementia, or memory complaints but
also for population-based samples of elderly who were
cognitively normal at study entry. Many studies have also
identified predictive measures in other domains, again re-
gardless of setting and sample: semantic memory/language
[63,66,68 –70], attention and executive functioning [59,64],
abstract reasoning [67], and visuospatial abilities [71].
Studies involving longer times to diagnosis typically
identify fewer neuropsychological predictors than those in-
volving shorter times [63,67]. In the CSHA, the only pre-
dictor of AD emerging from the 10-year study was delayed
verbal recall, whereas the 5-year study revealed semantic
memory measures as additional predictors [63]. These find-
ings suggest that the earliest warning sign of dementia or
AD might be an episodic memory deficit, and that as the
preclinical stage evolves toward diagnosable dementia, def-
icits in other domains also begin to have prognostic signif-
icance. A study with 2-year follow-up of memory-impaired
subjects in fact identified an almost five-fold risk of pro-
gression to AD for those who had impairments in non-
memory domains [72].
A meta-analysis of 47 studies calculated pooled effect
sizes as well as percent score overlap (O/L) between the
incident AD and control groups. These were Cohen’s d ⫽
1.0, O/L 42% for episodic memory; d ⫽ 0.8, O/L 57% for
verbal abilities; d ⫽ 0.6, O/L 62% for visuospatial abilities;
d ⫽ 0.6, O/L 62% for attention; and d ⫽ 1.1, O/L 42% for
executive functioning [73]. The overlap statistic suggests
that prediction, although useful, is not likely to be perfect or
even highly accurate in most cases. Because effect sizes
were larger for shorter (⬍3 years) than longer follow-up, it
can be assumed that increasing predictive accuracy is
achieved as time to diagnosis approaches.
This meta-analytic study identifies episodic memory and
executive functioning deficits as the best predictors of de-
mentia. Corroboration of this point can also be found from
a follow-up study of discordant neuropsychological diag-
noses of dementia in the population-based Kungsholmen
Project. Subjects classified as demented on the basis of NPT
but not diagnosed as demented clinically had an increased
likelihood of receiving a dementia diagnosis at the next
evaluation, particularly if the initial cognitive classification
was based on episodic memory and executive functioning
deficits [74].
It would currently be premature to extrapolate from the
existing evidence that specific neuropsychological tests can
be used for prognostic purposes. A problem that is often
overlooked in predictive studies that analyze multiple neu-
ropsychological scores is colinearity, that is, the tendency
for test scores to vary systematically with each other as a
function of disease severity. This can give rise to overin-
clusion of interrelated scores and to predictor sets that are
too large to be practically useful [71]. There might also be
significant instability of the predictive models. Tian et al
[75] used the same set of neuropsychological predictor
variables in three variants of Cox regressions (Enter, For-
wards Stepwise, and Backwards Stepwise), with time to
diagnosis of dementia as the dependent variable. Each
model was found to retain different variables, although each
consistently included a measure of episodic memory. A
strategy that is capable of explicitly addressing colinearity,
although not very user friendly, is principal component
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
analysis, which extracts the latent cognitive factor(s) that
might be measured by multiple tests. Most neuropsycholog-
ical test scores in the Personnes Agées Quid (PAQUID)
study loaded highly on the first factor, interpreted as reflect-
ing general executive control processes. When other factors
were generated at eigenvalues ⬎0.75, the first factor re-
mained the only reliable predictor of progression to AD
among a very large cohort of elderly subjects. Low scores
on this factor were associated with a two- to three-fold risk
of dementia or AD [76].
An entirely different approach to prediction of dementia
has focused on the distinction of profiles or subgroups. This
subgroup/profile approach has the advantage of using all the
available neuropsychological information, capitalizing on
the relationship between test scores, and bypassing the issue
of colinearity. In the ACCORD and CSHA cohorts, a sub-
group of CIND subjects defined by episodic memory dys-
function alone had a two- to three-fold risk of dementia at
follow-up compared with subgroups defined by other dys-
functions [44]. Two subgroups of subjects with recent cog-
nitive decline had a four-fold risk of dementia; one was
defined by multiple cognitive impairments and the other by
episodic memory impairment [77].
In summary, NPT has been shown to provide useful
prognostic information in both nonselected cohorts of el-
derly and selected groups with evidence of MCI. Episodic
memory tests appear to have the greatest predictive accu-
racy, possibly in combination with executive functioning
tests. In part, this might reflect the fact that episodic memory
impairment is the hallmark of AD, which, in turn, is the most
prevalent form of dementia. The utility of NPT might be
further enhanced when performance on multiple tests is scru-
tinized for patterns of impairment. The diagnostic/prognostic
value of NPT would be further strengthened if, in addition
to group norms, statistics such as likelihood ratios were
developed for tests that are routinely used [78].
4.3. The differential diagnosis of dementia
NPT of memory, other cognitive domains, and behavior
complements clinical history and neuroimaging in differen-
tial diagnosis and in documenting the natural cognitive
progression of various dementing illnesses. Distinguishable
neuropsychological profiles are seen in patients diagnosed
with various dementias including AD, VaD, dementia with
Lewy bodies (DLB), dementia associated with Parkinson’s
disease (PDD), and frontotemporal dementia (FTD). FTD
encompasses distinct clinical subtypes including a behavioral
variant (FTD-bv), progressive nonfluent aphasia (PNFA), se-
mantic dementia (SD), as well as corticobasal degeneration
(CBD) and progressive supranuclear palsy (PSP) [79,80].
Episodic memory deficit is a primary feature of AD in
contrast to FTD, where it is spared initially [81,82], and to
a lesser extent in PDD, DLB, and VaD [83– 87]. Memory
for recent events or names is affected early in AD, but with
311
the progression of the illness, remote memory becomes
involved, along with semantic memory. Decreased sensitiv-
ity to the semantic properties of information in AD inter-
feres with encoding of new material [88 –90]. Orientation
and episodic memory are relatively preserved in FTD. The
complaint of forgetfulness on history and variable perfor-
mance on memory testing are usually related to inattention.
FTD-bv patients might have impaired test results on imme-
diate and delayed recall of a story, yet they are able to recall
personally relevant events, which is out of keeping with
their cognitive test results [91,92]. Poor recall in PNFA and
semantic dementia is likely a result of the aphasic distur-
bance in these conditions. Recognition memory is typically
better than recall in VaD and dementia arising from sub-
cortical neurodegenerative processes.
Core executive functions assessed in NPT are planning,
concept formation, sequential organization, and attention.
When matched in severity, AD patients have an advantage
over DLB patients on tests of working memory, attention,
initiation, and perseveration, despite having poorer episodic
memory [83,85,93]. Qualitative analysis of neuropsycho-
logical performance shows more inattention, distractibility,
impaired set shifting, incoherence, confabulation, perse-
veration, and intrusion in DLB compared with AD groups
[94]. As with other disorders of the basal ganglia, PDD
patients exhibit impaired problem solving, set maintenance,
and shifting attention, especially to novel stimuli [86,95].
PDD patients also have similar fluctuations in attention as
seen in DLB [96], which partially accounts for mixed per-
formances in these groups on cognitive testing. On memory
tasks PDD and DLB patients, similarly to those with FTD
and VaD, benefit from recognition or semantic cueing par-
adigms at retrieval [92,97–101], whereas these paradigms
are largely ineffective in AD [102,103]. The improvement
in recognition and recall as a result of retrieval cues sug-
gests underlying impairment of working memory in FTD,
PDD, and DLB. Traditional tests of frontal lobe function,
such as the Wisconsin Card Sorting Test, Trail Making Test,
and Stroop Tests, have high sensitivity to executive dys-
function, but they also have low specificity to dementia
[104 –106].
Language deficits are usually considered to occur in 8%
to 10% of early AD patients [107,108], although aphasia is
invariably present in the advanced stages [109]. Word find-
ing difficulties with circumlocutions and fluent, articulated,
and syntactically preserved speech are often seen in early
cases [108,109], whereas the later dissolution of language in
AD resembles anomic, transcortical sensory, Wernicke’s,
and, finally, global aphasia [108]. Although language abil-
ities in PDD are largely intact compared with AD [110,111],
PD patients might have decreased information content,
difficulty comprehending complex sentences, anomia, and
reduced fluency that likely relate to the dysexecutive syn-
drome seen in this condition [112,113]. Language distur-
bance is also an early feature in some variants of FTD.
312 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
Phonemic paraphasias, word finding and picture naming
difficulty, and decreased word list generation (letter more
than semantic fluency) are usually evident during language
testing in patients with PFNA. FTD-bv patients might also
manifest anomia and reduction in spontaneous conversa-
tion. Although fluency might remain intact in the early
stages, semantic dementia patients manifest decreased
speech content and impaired performance on tests that re-
quire access to semantic knowledge. They have deficits in
category fluency, noun definitions and comprehension, pic-
ture naming, reading irregular words, and impoverished
results on the Pyramids and Palm Trees Tests, which require
individuals to match semantically associated items [114].
Cognitive tests of visuoperceptual, visuoconstructive,
and visuospatial ability are often problematic for patients
with DLB in the relatively early stages [115], and both DLB
and PDD patients have greater visuospatial dysfunction than
AD patients [84,93,111,115–117]. Cormack et al [116]
found that constructional performance was associated with
global cognitive impairment in PDD and AD, whereas in
DLB it was only linked with perceptual and praxis ability.
Visuospatial difficulty in PDD has been attributed to deficits
in planning and sequencing of responses that are analogous
to their motor impairment [118]. FTD groups perform better
on visuospatial tasks than AD groups [119], and their poor
performance on construction tests frequently results from
amotivation, poor planning, and impulsivity [91,106].
Behavioral and personality alterations, particularly apathy,
appetite changes, disinhibition, and stereotypic behaviors,
can help distinguish FTD from AD and VaD [120 –125],
and Kertesz et al [106] found behavioral quantification to be
more sensitive than cognitive testing in differentiating FTD
and AD groups. Patients presenting with semantic dementia
or PNFA develop behavioral changes similar to those seen
in FTD-bv, although later in the course of the illness
[106,126,127]. In later stages of AD, aggression and disin-
hibited behaviors can also become prominent features
[97,128,129]. Whereas AD patients typically have delusions
[130], it is visual hallucinations that are primary features of
DLB [131] and, to a lesser extent, PDD [132–134]. Depres-
sion is also found to a greater extent in PDD than AD [111],
whereas when matched in severity, aberrant motor behavior,
agitation, disinhibition, irritability, euphoria, and apathy are
more severe in AD [133]. The neuropsychiatric features of
AD and VaD are generally similar, and it has been proposed
that the increased prevalence of depression and anxiety
occasionally reported in VaD [135–138] might be due to
mixed pathology in AD and VaD clinical samples and to the
heterogeneity of VaD patients studied [139].
Tables 5 and 6 summarize evidence in support of diag-
nostic differentiation. At comparable dementia severity,
VaD patients exhibit greater deficits in executive function-
ing than tests of verbal learning and memory, whereas AD
patients exhibit the reverse pattern [87,101,137]. VaD pa-
tients often present with memory and executive difficulty,
but their performance shows a range of intact and impaired
skills. This variable cognitive profile of VaD patients is
considered a reflection of the location and extent of pathol-
ogy, making a stereotypic neuropsychological pattern less
likely [140]. Similar patchy and inconsistent performance
can be seen in FTD-bv, although this seems due to impul-
sivity and cursory or amotivational responding. On the
CDT, which calls on multiple cognitive skills [141], plan-
ning, visuospatial, and conceptual errors predominate in
DLB, PDD, and AD groups [142–144], whereas early FTD
patients can perform better than AD patients [119]. The
MMSE, probably the most widely used screening test, is not
sensitive in early FTD [145,146]. Along with behavioral
measures, a double dissociation with FTD patients perform-
ing poorly on tasks requiring inference of mental states,
thoughts, and feelings of others (theory of mind) and AD
performing poorly on tests of episodic memory further aids
differential diagnosis [147], although theory-of-mind tasks
have yet to be designed so as to be easily incorporated into
NPT batteries in clinical practice. Because PDD and DLB
share overlapping clinical symptomatology [83,132] and
neuropathology [148], it is not surprising that both have
similar cognitive and neuropsychiatric features of executive
dysfunction, visuospatial deficits, fluctuating cognition, and
hallucinations during the course of the disease. Fluctuations
in attention and alertness can result in mixed performance in
tests of sustained attention such as trail making, letter can-
cellations, digit span, fluency, and others.
In summary, the utility of NPT in the differential diagnosis
of dementia is variable and stage-dependent. Increased sen-
sitivity is often accompanied by decreased specificity. Pa-
tients cannot be easily fitted in diagnostic categories on the
basis of neuropsychological performance alone. Although
there is currently no consensus on the type of specific tests
that should be used in NPT batteries, differential diagnosis
requires that there be comprehensive testing of memory,
language, visuospatial, and executive abilities.
5. Neuropsychological assessment and
management decisions
Beyond diagnostic issues, comprehensive neuropsycho-
logical assessment can make important contributions to
management decisions in MCI and dementia. Although sys-
tematic studies are lacking, this role of neuropsychology is
well-recognized in clinical practice both from the standpoint
of testing in general [149] and from the standpoint of as-
sessment of dementia specifically [150]. Neuropsychologi-
cal assessment focuses on an individualized approach to the
determination of preserved as well as impaired cognitive
abilities that is not dependent on the use of population-based
cut points [149]. The information collected in the assess-
ment might serve in estimating the functional impact of the
cognitive disorder and the management of functional dis-
ability [151]. For persons still in the workforce, it might
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
play a major role in vocational decision making. Planning
for cognitive rehabilitation interventions in AD and other
forms of dementia also requires a comprehensive under-
standing of retained cognitive abilities on the part of the
affected individual [152].
6. Recommendations from the Third Canadian
Consensus Conference on Diagnosis and Treatment
of Dementia
6.1. Recommendations for brief cognitive tests
1. The MoCA and the DemTect are more sensitive to
MCI than the MMSE. Their use is recommended
when MCI is suspected. There is insufficient evi-
dence to recommend one test over the other (Grade
B, Level 2— because replication in different set-
tings, particularly the general practice setting, is
required).
2. There is insufficient evidence to recommend the de-
layed three-trial MMSE adaptation, the STMS, and
the SSST for the detection of MCI (Grade C, Level 2).
3. A range of brief cognitive tests, including the
MoCA, DemTect, BNA, GPCOG, and the 7MS,
might be more accurate than the MMSE in discrim-
inating dementia from normal. There is insufficient
evidence to recommend one test over the others
(Grade B, Level 2— because replication studies are
needed).
4. The distinction between MCI and AD is important
and is currently made on the basis of clinical
assessment of cognition and function (Grade A,
Level 3).
5. Brief cognitive tests might aid in the assessment of
this distinction between MCI and AD. The DemTect
and the STMS can be recommended because they
have established cut points for both MCI and AD
(Grade B, Level 2).
6.2. Recommendations for neuropsychological testing
and assessment
6. The diagnosis and differential diagnosis of dementia is
currently a clinically integrative one. Neuropsycholog-
ical assessment alone cannot be used for this purpose
and should be used selectively in clinical settings
(Grade B, Level 2).
7. Neuropsychological assessment may aid in:
(A) Addressing the distinction between normal aging,
MCI-CIND, and early dementia (Grade B, Level 2);
(B) Addressing the risk of progression from MCI-
CIND to dementia or AD (Grade B, Level 2);
(C) Differential diagnosis of dementia and other syn-
dromes of cognitive impairment (Grade B, Level 2);
313
(D) Determining whether there has been progression of
cognitive impairment or the development of new
impairment(s) to assist in management (Grade A,
Level 3).
Author Disclosures
Andrew Kertesz has received support from Pfizer Can-
ada (Advisory Board Member), Janssen Ortho Inc.
(Speaker), Novartis (Speaker, Former Advisory Board
Member), and Lundbeck Canada (Speaker).
John Fisk has received support from AstraZenca (Consul-
tant), Bayer (Consultant), Biogen-Idec (Consultant), Bristol-
Myers Squibb (Consultant), Novartis (Consultant), Sanofi-
Aventis (Consultant) and TEVA Neuroscience (Speaker).
Howard H. Feldman has received support from Pfizer
Canada (Grant/Research Support [External], Consultant,
CME Programs), Eisai (Grant/Research Support [External],
Consultant, CME Programs), Janssen Ortho Inc. (Grant/Re-
search Support [External], Consultant, CME Programs), As-
traZeneca (Grant/Research Support [External], Consultant),
Sanofi Synthelabo (Grant/Research Support [External],
Consultant), GlaxoSmithKline (Grant/Research Support
[External], Consultant), Novartis (Consultant, CME Pro-
grams), Eli Lilly (Grant/Research Support [External]), Ser-
vier (Consultant), Myriad (Consultant), Targacept (Consul-
tant), Lundbeck Canada (Consultant), Axonyx (Consultant),
and Forest (CME).
Acknowledgments
The authors gratefully acknowledge numerous sources of
support and funding. Claudia Jacova was funded by a post-
doctoral Canadian Institutes of Health Research/Michael
Smith Foundation for Health Research Training Fellowship
in Neurobiology and Behaviour.
Andrew Kertesz received grant support through the Med-
ical Research Council (MRC) of Canada and the Pharma-
ceutical Manufacturers Association of Canada (PMAC),
MRC PMAC program #14197 (A Canadian Cohort Study of
Cognitive Impairment and Related Dementias [ACCORD]),
and through the National Institutes of Health USA (NIH),
#1 U01-AG24904-01 (Alzheimer’s Disease Neuroimaging
Initiative [ADNI]).
Mervin Blair was funded by grant support from Janssen
and from the Lawson Research Institute.
Howard Feldman received grant support through the
MRC-PMAC, program #14197 (ACCORD), the National
Institutes of Health USA (NIH), #1 U01-AG24904-01
(ADNI), and through the Canadian Institutes of Health
Research (CIHR) operating grant #74580 (Frontotemporal
Dementia with Ubiquinated Inclusions).
314 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
References
[1] The Houston conference on specialty education and training in
clinical neuropsychology. Policy statement. Archives of Clinical
Neuropsychology 1998;157–250.
[2] Mulherin SA, Miller WC. Spectrum bias or spectrum effect? sub-
group variation in diagnostic test evaluation. Ann Intern Med 2002;
137:598 – 602.
[3] Lorentz WJ, Scanlan JM, Borson S. Brief screening tests for de-
mentia. Can J Psychiatry 2002;47:723–33.
[4] Folstein MF, Robins LN, Helzer JE. The Mini-Mental State Exam-
ination. Arch Gen Psychiatry 1983;40:812.
[5] Tombaugh TN, McIntyre NJ. The mini-mental state examination: a
comprehensive review. J Am Geriatr Soc 1992;40:922–35.
[6] Tombaugh TN, McDowell I, Kristjansson B, Hubley AM. Mini-
Mental State Examination (MMSE) and the modified MMSE
(3MS): a psychometric comparison and normative data. Psychol
Assess 1996;8:48 –59.
[7] Lindsay J, Laurin D, Verreault R, Hebert R, Helliwell B, Hill GB,
et al. Risk factors for Alzheimer’s disease: a prospective analysis
from the Canadian Study of Health and Aging. Am J Epidemiol
2002;156:445–53.
[8] Kraemer HC, Moritz DJ, Yesavage J. Adjusting Mini-Mental State
Examination scores for age and educational level to screen for
dementia: correcting bias or reducing validity? Int Psychogeriatr
1998;10:43–51.
[9] Feldman HH, Jacova C. Mild cognitive impairment. Am J Geriatr
Psychiatry 2005;13:645–55.
[10] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG,
Kokmen E. Mild cognitive impairment: clinical characterization
and outcome. Arch Neurol 1999;56:303– 8.
[11] Graham JE, Rockwood K, Beattie BL, Eastwood R, Gauthier S,
Tuokko H, et al. Prevalence and severity of cognitive impairment
with and without dementia in an elderly population. Lancet 1997;
349:1793– 6.
[12] Fisk JD, Merry HR, Rockwood K. Variations in case definition
affect prevalence but not outcomes of mild cognitive impairment.
Neurology 2003;61:1179 – 84.
[13] Feldman H, Levy AR, Hsiung GY, Peters KR, Donald A, Black SE,
et al. A Canadian cohort study of cognitive impairment and related
dementias (ACCORD): study methods and baseline results. Neuro-
epidemiology 2003;22:265–74.
[14] Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL,
DeKosky ST. Practice parameter: early detection of dementia: mild
cognitive impairment (an evidence-based review)—report of the
Quality Standards Subcommittee of the American Academy of Neu-
rology. Neurology 2001;56:1133– 42.
[15] Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, White-
head V, Collin I, et al. The Montreal Cognitive Assessment, MoCA:
a brief screening tool for mild cognitive impairment. J Am Geriatr
Soc 2005;53:695–9.
[16] Kalbe E, Kessler J, Calabrese P, Smith R, Passmore AP, Brand M,
et al. DemTect: a new, sensitive cognitive screening test to support
the diagnosis of mild cognitive impairment and early dementia. Int
J Geriatr Psychiatry 2004;19:136 – 43.
[17] Loewenstein DA, Barker WW, Harwood DG, Luis C, Acevedo A,
Rodriguez I, et al. Utility of a modified Mini-Mental State Exami-
nation with extended delayed recall in screening for mild cognitive
impairment and dementia among community dwelling elders. Int J
Geriatr Psychiatry 2000;15:434 – 40.
[18] Tang-Wai DF, Knopman DS, Geda YE, Edland SD, Smith GE,
Ivnik RJ, et al. Comparison of the short test of mental status and the
mini-mental state examination in mild cognitive impairment. Arch
Neurol 2003;60:1777– 81.
[19] Eibenstein A, Fioretti AB, Simaskou MN, Sucapane P, Mearelli S,
Mina C, et al. Olfactory screening test in mild cognitive impairment.
Neurol Sci 2005;26:156 – 60.
[20] Doty RL, Reyes PF, Gregor T. Presence of both odor identification
and detection deficits in Alzheimer’s disease. Brain Res Bull 1987;
18:597– 600.
[21] Morgan CD, Murphy C. Olfactory event-related potentials in Alz-
heimer’s disease. J Int Neuropsychol Soc 2002;8:753– 63.
[22] Teng EL, Chui HC. The Modified Mini-Mental State (3MS) exam-
ination. J Clin Psychiatry 1987;48:314 – 8.
[23] McDowell I, Hill G, Lindsay J, Helliwell B, Costa L, Beattie BL,
et al. Canadian Study of Health and Aging: study methods and
prevalence of dementia. CMAJ 1994;150:899 –913.
[24] Tuokko H, Hadjistavropoulos T, Miller JA, Beattie BL. The Clock
Test: a sensitive measure to differentiate normal elderly from those
with Alzheimer disease. J Am Geriatr Soc 1992;40:579 – 84.
[25] Bush C, Kozak J, Elmslie T. Screening for cognitive impairment in
the elderly. Can Fam Physician 1997;43:1763– 8.
[26] Shulman KI. Clock-drawing: is it the ideal cognitive screening test?
Int J Geriatr Psychiatry 2000;15:548 – 61.
[27] Morris JC. Clinical dementia rating: a reliable and valid diagnostic
and staging measure for dementia of the Alzheimer type. Int Psy-
chogeriatr 1997;9(Suppl 1):173– 6.
[28] Powlishta KK, Von Dras DD, Stanford A, Carr DB, Tsering C,
Miller JP, et al. The clock drawing test is a poor screen for very mild
dementia. Neurology 2002;59:898 –903.
[29] De Lepeleire J, Heyman J, Baro F, Buntinx F. A combination of
tests for the diagnosis of dementia had a significant diagnostic value.
J Clin Epidemiol 2005;58:217–25.
[30] Stahelin HB, Monsch AU, Spiegel R. Early diagnosis of dementia
via a two-step screening and diagnostic procedure. Int Psychogeriatr
1997;9(Suppl 1):123–30.
[31] Solomon PR, Hirschoff A, Kelly B, Relin M, Brush M, DeVeaux
RD, et al. A 7 minute neurocognitive screening battery highly
sensitive to Alzheimer’s disease. Arch Neurol 1998;55:349 –55.
[32] Solomon PR, Brush M, Calvo V, Adams F, DeVeaux RD, Pendle-
bury WW, et al. Identifying dementia in the primary care practice.
Int Psychogeriatr 2000;12:483–93.
[33] Kuslansky G, Buschke H, Katz M, Sliwinski M, Lipton RB. Screen-
ing for Alzheimer’s disease: the memory impairment screen versus
the conventional three-word memory test. J Am Geriatr Soc 2002;
50:1086 –91.
[34] Brodaty H, Pond D, Kemp NM, Luscombe G, Harding L, Berman
K, et al. The GPCOG: A new screening test for dementia designed
for general practice. J Am Geriatr Soc 2002;50:530 – 4.
[35] Kilada S, Gamaldo A, Grant EA, Moghekar A, Morris JC, O’Brien
RJ. Brief screening tests for the diagnosis of dementia: comparison
with the mini-mental state exam. Alzheimer Dis Assoc Disord 2005;
19:8 –16.
[36] Storey JE, Rowland JT, Basic D, Conforti DA, Dickson HG. The
Rowland Universal Dementia Assessment Scale (RUDAS): a mul-
ticultural cognitive assessment scale. Int Psychogeriatr 2004;16:
13–31.
[37] Darvesh S, Leach L, Black SE, Kaplan E, Freedman M. The
behavioural neurology assessment. Can J Neurol Sci 2005;32:
167–77.
[38] Stuss DT, Meiran N, Guzman DA, Lafleche G, Willmer J. Do long
tests yield a more accurate diagnosis of dementia than short tests? a
comparison of 5 neuropsychological tests. Arch Neurol 1996;53:
1033–9.
[39] Chui H, Zhang Q. Evaluation of dementia: a systematic study of the
usefulness of the American Academy of Neurology’s practice pa-
rameters. Neurology 1997;49:925–35.
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
[40] Chertkow H, Bergman H, Schipper HM, Gauthier S, Bouchard R,
Fontaine S, et al. Assessment of suspected dementia. Can J Neurol
Sci 2001;28(Suppl 1):28 – 41.
[41] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM. Clinical diagnosis of Alzheimer’s disease: report of the
NINCDS-ADRDA Work Group under the auspices of Department
of Health and Human Services Task Force on Alzheimer’s Disease.
Neurology 1984;34:939 – 44.
[42] Levy R. Aging-associated cognitive decline: Working Party of the
International Psychogeriatric Association in collaboration with the
World Health Organization. Int Psychogeriatr 1994;6:63– 8.
[43] Peters KR, Graf P, Hayden S, Feldman H. Neuropsychological
characterization of cognitively-impaired-not-demented (CIND) in-
dividuals: clinical comparison data. Clin Neuropsychol 2004;18:
208 –28.
[44] Peters KR, Graf P, Hayden S, Feldman H. Neuropsychological
subgroups of cognitively-impaired-not-demented (CIND) individu-
als: delineation, reliability, and predictive validity. J Clin Exp Neu-
ropsychol 2005;27:164 – 88.
[45] Lopez OL, Becker JT, Jagust WJ, Fitzpatrick A, Carlson MC,
DeKosky ST, et al. Neuropsychological characteristics of mild cog-
nitive impairment subgroups. J Neurol Neurosurg Psychiatry 2006;
77:159 – 65.
[46] Grundman M, Petersen RC, Ferris SH, Thomas RG, Aisen PS,
Bennett DA, et al. Mild cognitive impairment can be distinguished
from Alzheimer disease and normal aging for clinical trials. Arch
Neurol 2004;61:59 – 66.
[47] Kramer JH, Nelson A, Johnson JK, Yaffe K, Glenn S, Rosen HJ,
et al. Multiple cognitive deficits in amnestic mild cognitive impair-
ment. Dement Geriatr Cogn Disord 2006;22:306 –11.
[48] Collie A, Maruff P, Currie J. Behavioral characterization of mild
cognitive impairment. J Clin Exp Neuropsychol 2002;24:720 –33.
[49] Welsh K, Butters N, Hughes J, Mohs R, Heyman A. Detection of
abnormal memory decline in mild cases of Alzheimer’s disease using
CERAD neuropsychological measures. Arch Neurol 1991;48:278 – 81.
[50] Cahn DA, Salmon DP, Butters N, Wiederholt WC, Corey-Bloom J,
Edelstein SL, et al. Detection of dementia of the Alzheimer type in
a population-based sample: neuropsychological test performance.
J Int Neuropsychol Soc 1995;1:252– 60.
[51] Swearer JM, O’Donnell BF, Kane KJ, Hoople NE, Lavoie M.
Delayed recall in dementia: sensitivity and specificity in patients
with higher than average general intellectual abilities. Neuropsychi-
atry Neuropsychol Behav Neurol 1998;11:200 – 6.
[52] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders, fourth edition, text revision (DSM-IV-TR).
Arlington, VA. American Psychiatric Association; 2000.
[53] Steenhuis RE, Ostbye T. Neuropsychological test performance of
specific diagnostic groups in the Canadian Study of Health and
Aging (CSHA). J Clin Exp Neuropsychol 1995;17:773– 85.
[54] Lambon Ralph MA, Patterson K, Graham N, Dawson K, Hodges JR.
Homogeneity and heterogeneity in mild cognitive impairment and
Alzheimer’s disease: a cross-sectional and longitudinal study of 55
cases. Brain 2003;126(Pt 11):2350 – 62.
[55] Small BJ, Fratiglioni L, Viitanen M, Winblad B, Backman L. The
course of cognitive impairment in preclinical Alzheimer disease:
three- and 6-year follow-up of a population-based sample. Arch
Neurol 2000;57:839 – 44.
[56] Hall CB, Lipton RB, Sliwinski M, Stewart WF. A change point
model for estimating the onset of cognitive decline in preclinical
Alzheimer’s disease. Stat Med 2000;19:1555– 66.
[57] Rubin EH, Storandt M, Miller JP, Kinscherf DA, Grant EA, Morris JC,
et al. A prospective study of cognitive function and onset of dementia
in cognitively healthy elders. Arch Neurol 1998;55:395– 401.
[58] Amieva H, Jacqmin-Gadda H, Orgogozo JM, Carret N, Helmer C,
Letenneur L, et al. The 9 year cognitive decline before dementia of
315
the Alzheimer type: a prospective population-based study. Brain
2005;128(Pt 5):1093–101.
[59] Chen P, Ratcliff G, Belle SH, Cauley JA, DeKosky ST, Ganguli M.
Patterns of cognitive decline in presymptomatic Alzheimer disease:
a prospective community study. Arch Gen Psychiatry 2001;58:
853– 8.
[60] Maruff P, Collie A, Darby D, Weaver-Cargin J, Masters C, Currie J.
Subtle memory decline over 12 months in mild cognitive impair-
ment. Dement Geriatr Cogn Disord 2004;18:342– 8.
[61] Fowler KS, Saling MM, Conway EL, Semple JM, Louis WJ. Paired
associate performance in the early detection of DAT. J Int Neuro-
psychol Soc 2002;8:58 –71.
[62] Bowen J, Teri L, Kukull W, McCormick W, McCurry SM, Larson
EB. Progression to dementia in patients with isolated memory loss.
Lancet 1997;349:763–5.
[63] Tierney MC, Yao C, Kiss A, McDowell I. Neuropsychological tests
accurately predict incident Alzheimer disease after 5 and 10 years.
Neurology 2005;64:1853–9.
[64] Tierney MC, Szalai JP, Snow WG, Fisher RH, Nores A, Nadon
G, et al. Prediction of probable Alzheimer’s disease in memory-im-
paired patients: a prospective longitudinal study. Neurology 1996;46:
661–5.
[65] Arnaiz E, Almkvist O, Ivnik RJ, Tangalos EG, Wahlund LO, Win-
blad B, et al. Mild cognitive impairment: a cross-national compar-
ison. J Neurol Neurosurg Psychiatry 2004;75:1275– 80.
[66] Artero S, Ritchie K. The detection of mild cognitive impairment in
the general practice setting. Aging Ment Health 2003;7:251– 8.
[67] Elias MF, Beiser A, Wolf PA, Au R, White RF, D’Agostino RB.
The preclinical phase of Aalzheimer disease: a 22-year prospective
study of the Framingham Cohort. Arch Neurol 2000;57:808 –13.
[68] Howieson DB, Dame A, Camicioli R, Sexton G, Payami H, Kaye
JA. Cognitive markers preceding Alzheimer’s dementia in the
healthy oldest old. J Am Geriatr Soc 1997;45:584 –9.
[69] Blackwell AD, Sahakian BJ, Vesey R, Semple JM, Robbins TW,
Hodges JR. Detecting dementia: novel neuropsychological markers
of preclinical Alzheimer’s disease. Dement Geriatr Cogn Disord
2004;17:42– 8.
[70] Palmer K, Backman L, Winblad B, Fratiglioni L. Detection of
Alzheimer’s disease and dementia in the preclinical phase: popula-
tion based cohort study. BMJ 2003;326:245.
[71] Devanand DP, Folz M, Gorlyn M, Moeller JR, Stern Y. Question-
able dementia: clinical course and predictors of outcome. J Am
Geriatr Soc 1997;45:321– 8.
[72] Bozoki A, Giordani B, Heidebrink JL, Berent S, Foster NL. Mild
cognitive impairments predict dementia in nondemented elderly
patients with memory loss. Arch Neurol 2001;58:411– 6.
[73] Backman L, Jones S, Berger AK, Laukka EJ, Small BJ. Cognitive
impairment in preclinical Alzheimer’s disease: a meta-analysis.
Neuropsychology 2005;19:520 –31.
[74] Herlitz A, Small BJ, Fratiglioni L, Almkvist O, Viitanen M,
Backman L. Detection of mild dementia in community surveys:
is it possible to increase the accuracy of our diagnostic instru-
ments? Arch Neurol 1997;54:319 –24.
[75] Tian J, Bucks RS, Haworth J, Wilcock G. Neuropsychological
prediction of conversion to dementia from questionable dementia:
statistically significant but not yet clinically useful. J Neurol Neu-
rosurg Psychiatry 2003;74:433– 8.
[76] Fabrigoule C, Rouch I, Taberly A, Letenneur L, Commenges D,
Mazaux JM, et al. Cognitive process in preclinical phase of demen-
tia. Brain 1998;121(Pt 1):135– 41.
[77] Ritchie K, Leibovici D, Ledesert B, Touchon J. A typology of sub-
clinical senescent cognitive disorder. Br J Psychiatry 1996;168:470 – 6.
[78] Ivnik RJ, Smith GE, Cerhan JH, Boeve BF, Tangalos EG, Petersen
RC. Understanding the diagnostic capabilities of cognitive tests.
Clin Neuropsychol 2001;15:114 –24.
316 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
[79] Neary D, Snowden J, Mann D. Frontotemporal dementia. Lancet
Neurol 2005;4:771– 80.
[80] Feldman H, Kertesz A. Diagnosis, classification and natural history
of degenerative dementias. Can J Neurol Sci 2001;28(Suppl 1):
S17–27.
[81] Gustafson L. Frontal lobe degeneration of non-Alzheimer type. II.
Clinical picture and differential diagnosis. Arch Gerontol Geriatr
1987;6:209 –23.
[82] Neary D, Snowden JS, Northen B, Goulding P. Dementia of frontal
lobe type. J Neurol Neurosurg Psychiatry 1988;51:353– 61.
[83] Aarsland D, Litvan I, Salmon D, Galasko D, Wentzel-Larsen T,
Larsen JP. Performance on the dementia rating scale in Parkinson’s
disease with dementia and dementia with Lewy bodies: comparison
with progressive supranuclear palsy and Alzheimer’s disease. J Neu-
rol Neurosurg Psychiatry 2003;74:1215–20.
[84] Ballard CG, Ayre G, O’Brien J, Sahgal A, McKeith IG, Ince PG, et
al. Simple standardised neuropsychological assessments aid in the
differential diagnosis of dementia with Lewy bodies from Alzhei-
mer’s disease and vascular dementia. Dement Geriatr Cogn Disord
1999;10:104 – 8.
[85] Calderon J, Perry RJ, Erzinclioglu SW, Berrios GE, Dening TR,
Hodges JR. Perception, attention, and working memory are dispropor-
tionately impaired in dementia with Lewy bodies compared with Alz-
heimer’s disease. J Neurol Neurosurg Psychiatry 2001;70:157– 64.
[86] Litvan I, Mohr E, Williams J, Gomez C, Chase TN. Differential
memory and executive functions in demented patients with Parkin-
son’s and Alzheimer’s disease. J Neurol Neurosurg Psychiatry 1991;
54:25–9.
[87] Looi JC, Sachdev PS. Differentiation of vascular dementia from AD
on neuropsychological tests. Neurology 1999;53:670 – 8.
[88] Corkin S. Some relationships between global amnesia and the mem-
ory impairment in Alzheimer’s disease. In: Corkin S, Davis K,
Growdon JH, eds. Alzheimer’s disease: a report of progress on
research. New York: Raven Press; 1982. p. 149 – 64.
[89] Karlsson T, Backman L, Herlitz A, Nilsson LG, Winblad B,
Osterlind PO. Memory improvement at different stages of Alz-
heimer’s disease. Neuropsychologia 1989;27:737– 42.
[90] Martin A, Brouwers P, Cox C, Fedio P. On the nature of the
verbal memory deficit in Alzheimer’s disease. Brain Lang 1985;
25:323– 41.
[91] Hodges JR. Frontotemporal dementia (Pick’s disease): clinical fea-
tures and assessment. Neurology 2001;56(Suppl 4):6 –10.
[92] Pasquier F, Grymonprez L, Lebert F, Van der Linden M. Memory
impairment differs in frontotemporal dementia and Alzheimer’s
disease. Neurocase 2001;7:161–71.
[93] Collerton D, Burn D, McKeith I, O’Brien J. Systematic review and
meta-analysis show that dementia with Lewy bodies is a visual-
perceptual and attentional-executive dementia. Dement Geriatr
Cogn Disord 2003;16:229 –37.
[94] Doubleday EK, Snowden JS, Varma AR, Neary D. Qualitative per-
formance characteristics differentiate dementia with Lewy bodies and
Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2002;72:602–7.
[95] Emre M. Dementia associated with Parkinson’s disease. Lancet
Neurol 2003;2:229 –37.
[96] Ballard CG, Aarsland D, McKeith I, O’Brien J, Gray A, Cormack F,
et al. Fluctuations in attention: PD dementia vs DLB with parkin-
sonism. Neurology 2002;59:1714 –20.
[97] Mendez MF, Cummings JL. Dementia: a clinical approach, 3 ed.
Philadelphia: Butterworth-Heinemann (Elsevier); 2003.
[98] Hamilton JM, Salmon DP, Galasko D, Delis DC, Hansen LA,
Masliah E, et al. A comparison of episodic memory deficits in
neuropathologically-confirmed dementia with Lewy bodies and Alz-
heimer’s disease. J Int Neuropsychol Soc 2004;10:689 –97.
[99] Helkala EL, Laulumaa V, Soininen H, Riekkinen PJ. Recall and
recognition memory in patients with Alzheimer’s and Parkinson’s
diseases. Ann Neurol;24:214 –7.
[100] Pillon B, Deweer B, Agid Y, Dubois B. Explicit memory in Alz-
heimer’s, Huntington’s, and Parkinson’s diseases. Arch Neurol
1993;50:374 –9.
[101] Tierney MC, Black SE, Szalai JP, Snow WG, Fisher RH, Nadon
G, et al. Recognition memory and verbal fluency differentiate
probable Alzheimer disease from subcortical ischemic vascular
dementia. Arch Neurol 2001;58:1654 –9.
[102] Branconnier RJ, Cole JO, Spera KF, DeVitt DR. Recall and recog-
nition as diagnostic indices of malignant memory loss in senile
dementia: a Bayesian analysis. Exp Aging Res 1982;8:189 –93.
[103] Davis PE, Mumford SJ. Cued recall and the nature of the memory
disorder in dementia. Br J Psychiatry 1984;144:383– 6.
[104] Chen P, Ratcliff G, Belle SH, Cauley JA, DeKosky ST, Ganguli
M. Cognitive tests that best discriminate between presymptom-
atic AD and those who remain nondemented. Neurology 2000;
55:1847–53.
[105] Grossman M. Frontotemporal dementia: a review. J Int Neuropsy-
chol Soc 2002;8:566 – 83.
[106] Kertesz A, Davidson W, McCabe P, Munoz D. Behavioral quanti-
tation is more sensitive than cognitive testing in frontotemporal
dementia. Alzheimer Dis Assoc Disord 2003;17:223–9.
[107] Faber-Langendoen K, Morris JC, Knesevich JW, LaBarge E, Miller
JP, Berg L. Aphasia in senile dementia of the Alzheimer type. Ann
Neurol 1988;23:365–70.
[108] Kertesz A, Appell J, Fisman M. The dissolution of language in
Alzheimer’s disease. Can J Neurol Sci 1986;13(Suppl):415– 8.
[109] Appell J, Kertesz A, Fisman M. A study of language functioning in
Alzheimer patients. Brain Lang 1982;17:73–91.
[110] Cummings JL, Darkins A, Mendez M, Hill MA, Benson DF. Alz-
heimer’s disease and Parkinson’s disease: comparison of speech and
language alterations. Neurology 1988;38:680 – 4.
[111] Huber SJ, Shuttleworth EC, Freidenberg DL. Neuropsychological
differences between the dementias of Alzheimer’s and Parkinson’s
diseases. Arch Neurol 1989;46:1287–91.
[112] Grossman M, Carvell S, Gollomp S, Stern MB, Vernon G, Hurtig
HI. Sentence comprehension and praxis deficits in Parkinson’s dis-
ease. Neurology 1991;41:1620 – 6.
[113] Grossman M, Carvell S, Stern MB, Gollomp S, Hurtig HI. Sentence
comprehension in Parkinson’s disease: the role of attention and
memory. Brain Lang 1992;42:347– 84.
[114] Howard D, Patterson K. Pyramids and palm trees: a test of semantic
access from pictures and words. St Edmunds, Suffolk, UK: Thames
Valley Test Company; 1992.
[115] Shimomura T, Mori E, Yamashita H, Imamura T, Hirono N, Hashi-
moto M, et al. Cognitive loss in dementia with Lewy bodies and
Alzheimer disease. Arch Neurol 1998;55:1547–52.
[116] Cormack F, Aarsland D, Ballard C, Tovee MJ. Pentagon drawing
and neuropsychological performance in dementia with Lewy bodies,
Alzheimer’s disease, Parkinson’s disease and Parkinson’s disease
with dementia. Int J Geriatr Psychiatry 2004;19:371–7.
[117] Stern Y, Richards M, Sano M, Mayeux R. Comparison of cognitive
changes in patients with Alzheimer’s and Parkinson’s disease. Arch
Neurol 1993;50:1040 –5.
[118] Stern Y, Mayeux R, Rosen J, Ilson J. Perceptual motor dysfunction
in Parkinson’s disease: a deficit in sequential and predictive volun-
tary movement. J Neurol Neurosurg Psychiatry 1983;46:145–51.
[119] Rascovsky K, Salmon DP, Ho GJ, Galasko D, Peavy GM, Hansen
LA, et al. Cognitive profiles differ in autopsy-confirmed frontotem-
poral dementia and AD. Neurology 2002;58:1801– 8.
[120] Bathgate D, Snowden JS, Varma A, Blackshaw A, Neary D. Be-
haviour in frontotemporal dementia, Alzheimer’s disease and vas-
cular dementia. Acta Neurol Scand 2001;103:367–78.
 C. Jacova et al. / Alzheimer’s & Dementia 3 (2007) 299 –317
[121] Bozeat S, Gregory CA, Ralph MA, Hodges JR. Which neuropsy-
chiatric and behavioural features distinguish frontal and temporal
variants of frontotemporal dementia from Alzheimer’s disease?
J Neurol Neurosurg Psychiatry 2000;69:178 – 86.
[122] Kertesz A, Davidson W, Fox H. Frontal behavioral inventory: di-
agnostic criteria for frontal lobe dementia. Can J Neurol Sci 1997;
24:29 –36.
[123] Kertesz A, Nadkarni N, Davidson W, Thomas AW. The Frontal
Behavioral Inventory in the differential diagnosis of frontotemporal
dementia. J Int Neuropsychol Soc 2000;6:460 – 8.
[124] Levy ML, Miller BL, Cummings JL, Fairbanks LA, Craig A. Alz-
heimer disease and frontotemporal dementias: behavioral distinc-
tions. Arch Neurol 1996;53:687–90.
[125] Miller BL, Ikonte C, Ponton M, Levy M, Boone K, Darby A, et al. A
study of the Lund-Manchester research criteria for frontotemporal de-
mentia: clinical and single-photon emission CT correlations. Neurology
1997;48:937– 42.
[126] Davies RR, Hodges JR, Kril JJ, Patterson K, Halliday GM, Xuereb
JH. The pathological basis of semantic dementia. Brain 2005;128(Pt
9):1984 –95.
[127] Marczinski CA, Davidson W, Kertesz A. A longitudinal study of
behavior in frontotemporal dementia and primary progressive apha-
sia. Cogn Behav Neurol 2004;17:185–90.
[128] Devanand DP. The interrelations between psychosis, behavioral
disturbance, and depression in Alzheimer disease. Alzheimer Dis
Assoc Disord 1999;13(Suppl 2):S3– 8.
[129] Mendez MF, Martin RJ, Smyth KA, Whitehouse PJ. Psychiatric
symptoms associated with Alzheimer’s disease. J Neuropsychiatry
Clin Neurosci 1990;2:28 –33.
[130] Paulsen JS, Salmon DP, Thal LJ, Romero R, Weisstein-Jenkins C,
Galasko D, et al. Incidence of and risk factors for hallucinations and
delusions in patients with probable AD. Neurology 2000;54:1965–71.
[131] Ballard C, Holmes C, McKeith I, Neill D, O’Brien J, Cairns N, et al.
Psychiatric morbidity in dementia with Lewy bodies: a prospective
clinical and neuropathological comparative study with Alzheimer’s
disease. Am J Psychiatry 1999;156:1039 – 45.
[132] Aarsland D, Ballard C, Larsen JP, McKeith I. A comparative study
of psychiatric symptoms in dementia with Lewy bodies and Parkin-
son’s disease with and without dementia. Int J Geriatr Psychiatry
2001;16:528 –36.
[133] Aarsland D, Cummings JL, Larsen JP. Neuropsychiatric differences
between Parkinson’s disease with dementia and Alzheimer’s dis-
ease. Int J Geriatr Psychiatry 2001;16:184 –91.
[134] Fenelon G, Mahieux F, Huon R, Ziegler M. Hallucinations in Par-
kinson’s disease: prevalence, phenomenology and risk factors. Brain
2000;123(Pt 4):733– 45.
[135] Ballard C, Neill D, O’Brien J, McKeith IG, Ince P, Perry R. Anxiety,
depression and psychosis in vascular dementia: prevalence and as-
sociations. J Affect Disord 2000;59:97–106.
[136] Cummings JL, Miller B, Hill MA, Neshkes R. Neuropsychiatric
aspects of multi-infarct dementia and dementia of the Alzheimer
type. Arch Neurol 1987;44:389 –93.
[137] Starkstein SE, Sabe L, Vazquez S, Teson A, Petracca G, Chemer-
inski E, et al. Neuropsychological, psychiatric, and cerebral blood
flow findings in vascular dementia and Alzheimer’s disease. Stroke
1996;27:408 –14.
[138] Sultzer DL, Levin HS, Mahler ME, High WM, Cummings JL. A
comparison of psychiatric symptoms in vascular dementia and Alz-
heimer’s disease. Am J Psychiatry 1993;150:1806 –12.
[139] Srikanth S, Nagaraja AV, Ratnavalli E. Neuropsychiatric symptoms
in dementia-frequency, relationship to dementia severity and com-
parison in Alzheimer’s disease, vascular dementia and frontotem-
poral dementia. J Neurol Sci 2005;236:43– 8.
[140] Rao R, Howard R. Vascular dementia: dead or alive? Int J Geriatr
Psychiatry 1998;13:277– 84.
317
[141] Freedman M, Leach L, Kaplan E, Shulman KI. Clock drawing. a
neuropsychological analysis. New York: Oxford University Press;
1994.
[142] Cahn-Weiner DA, Williams K, Grace J, Tremont G, Westervelt H,
Stern RA. Discrimination of dementia with Lewy bodies from Alz-
heimer disease and Parkinson disease using the clock drawing test.
Cogn Behav Neurol 2003;16:85–92.
[143] Rouleau I, Salmon DP, Butters N, Kennedy C, McGuire K. Quan-
titative and qualitative analyses of clock drawings in Alzheimer’s
and Huntington’s disease. Brain Cogn 1992;18:70 – 87.
[144] Rouleau I, Salmon DP, Butters N. Longitudinal analysis of clock
drawing in Alzheimer’s disease patients. Brain Cogn 1996;31:17–34.
[145] Gregory CA, Hodges JR. Clinical features of frontal lobe dementia
in comparison to Alzheimer’s disease. J Neural Transm Suppl 1996;
47:103–23.
[146] Miller BL, Cummings JL, Villanueva-Meyer J, et al. Frontal lobe
degeneration: clinical, neuropsychological, and SPECT characteris-
tics. Neurology 1991;41:1374 – 82.
[147] Gregory C, Lough S, Stone V, Erzinclioglu S, Martin L, Baron-Cohen
S, et al. Theory of mind in patients with frontal variant frontotemporal
dementia and Alzheimer’s disease: theoretical and practical implica-
tions. Brain 2002;125(Pt 4):752– 64.
[148] Colosimo C, Hughes AJ, Kilford L, Lees AJ. Lewy body cortical
involvement may not always predict dementia in Parkinson’s dis-
ease. J Neurol Neurosurg Psychiatry 2003;74:852– 6.
[149] Lezak MD, Howieson DB, Loring DW. Neuropsychological Assess-
ment. New York: Oxford University Press; 2004.
[150] Morgan CD, Baade LE. Neuropsychological testing and assessment
scales for dementia of the Alzheimer’s type. Psychiatr Clin North
Am 1997;20:25– 43.
[151] Sano M. Neuropsychological testing in the diagnosis of dementia. J
Geriatr Psychiatry Neurol 2006;19:155–9.
[152] De Vreese LP, Neri M, Fioravanti M, Belloi L, Zanetti O. Memory
rehabilitation in Alzheimer’s disease: a review of progress. Int J
Geriatr Psychiatry 2001;16:794 – 809.
[153] Christensen H, Griffiths K, Mackinnon A, Jacomb P. A quantitative
review of cognitive deficits in depression and Alzheimer-type de-
mentia. J Int Neuropsychol Soc 1997;3:631–51.
[154] Simard M, van RR, Cohen T. A review of the cognitive and behav-
ioral symptoms in dementia with Lewy bodies. J Neuropsychiatry
Clin Neurosci 2000;12:425–50.
[155] Kramer JH, Jurik J, Sha SJ, Rankin KP, Rosen HJ, Johnson JK, et
al. Distinctive neuropsychological patterns in frontotemporal de-
mentia, semantic dementia, and Alzheimer disease. Cogn Behav
Neurol 2003;16:211– 8.
[156] Mathuranath PS, Nestor PJ, Berrios GE, Rakowicz W, Hodges JR.
A brief cognitive test battery to differentiate Alzheimer’s disease
and frontotemporal dementia. Neurology 2000;55:1613–20.
[157] Gorno-Tempini ML, Dronkers NF, Rankin KP, Ogar JM, Phen-
grasamy L, Rosen HJ, et al. Cognition and anatomy in three variants
of primary progressive aphasia. Ann Neurol 2004;55:335– 46.
[158] Hodges JR, Patterson K, Ward R, Garrard P, Bak T, Perry R,
Gregory C. The differentiation of semantic dementia and frontal
lobe dementia (temporal and frontal variants of frontotemporal de-
mentia) from early Alzheimer’s disease: a comparative neuropsy-
chological study. Neuropsychology 1999;13:31– 40.
[159] Sjogren M, Wallin A, Edman A. Symptomatological characteristics
distinguish between frontotemporal dementia and vascular dementia
with a dominant frontal lobe syndrome. Int J Geriatr Psychiatry
1997;12:656 – 61.
[160] Butters N, Wolfe J, Martone M, Granholm E, Cermak LS. Memory
disorders associated with Huntington’s disease: verbal recall, verbal
recognition and procedural memory. Neuropsychologia 1985;23:
729 – 43.
 Alzheimer’s & Dementia 3 (2007) 318 –324

Cobalamin and homocysteine in older adults: Do we need to test

for serum levels in the work-up of dementia?
Angeles Garcia*
Department of Medicine (Geriatrics), Queen’s University, Kingston, Ontario, Canada

Abstract This article reviews available information on the relationship between cobalamin (Cbl), homo-
cysteine (tHcy) and cognitive decline and dementia in older adults with the aim to propose
recommendations as to the need to perform such determinations in the work-up for dementia. The
article includes brief reviews of the magnitude of the problem, effects of the Canadian folic acid
fortification program on the Cbl and tHcy status in older adults, relation between Cbl and tHcy and
cognition and cognitive changes, and available data on treatment trials up to March 2007. Recom-
mendations and levels of evidence were assigned and approved by consensus following the
directives of the CCCDTD3. The review concludes that determination of Cbl (vit B12) levels are
recommended in the work-up for dementia and cognitive decline because of the high prevalence of
Cbl deficits in this population, independently of the possible effects of normalization of Cbl levels
on cognitive function. Even though elevated tHcy is a risk factor for dementia, there is no prooof
that normalization of tHcy levels changes the course of the disease. Clinical trials on that regard are
on going.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Diagnosis; Dementia; Cobalamin; Vitamin B12; Homocysteine; Folate

1. Introduction of dementia more frequent among younger adults. Aspects

The Third Canadian Consensus Conference on the Diag-
nosis and Treatment of Dementia started from the assump-
tion that previously published guidelines advocating a lim-
ited battery of lab tests (complete blood count, electrolytes,
thyroid-stimulating hormone, calcium, glucose) for routine
dementia diagnosis remain valid. This review is aimed at
summarizing the available evidence for the need to perform
cobalamin (Cbl or B12), folate, and homocysteine serum
determinations in patients whose cognitive function is under
investigation. The review and recommendations are cen-
tered on older adults because this is the age group most
affected by dementia, cognitive decline, and the metabolic
changes considered; the review takes a holistic approach to
investigations and treatment of this population. Therefore,
the recommendations might not be applicable to patients
with early-onset familiar Alzheimer’s disease or other types
*Corresponding author. Tel.: (613) 548-7222 x2229; Fax: (613)
544-4017.
E-mail address: garciaa@providencecare.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.003
that might differentiate Canadian population from other
countries’ populations have been emphasized.
Low cobalamin (Cbl or B12), a commonly occurring
metabolic deficiency in the elderly [1,2], has been consid-
ered a “reversible cause of dementia,” but the fact that the
cognitive deficits of many patients with low Cbl and de-
mentia do not improve after the Cbl levels are normalized
has prompted questions about the relationship between the
two and the relevance of Cbl in cognitive function.
The association between Cbl and cognitive function
has recently been analyzed in light of studies involving
total homocysteine (tHcy). Total homocysteine is a sulfur-
containing amino acid that accumulates in states of low
Cbl and/or folic acid [3,4]. Large epidemiologic studies
have found that elevated tHcy is a risk factor for the
development of cerebrovascular disease, dementia of the
Alzheimer’s type (AD), and vascular dementia [5– 8],
although others have failed to confirm the association
between AD and tHcy [9]. High levels of tHcy have been
associated with white matter abnormalities [10,11].
 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
The inverse relationships between Cbl, folic acid, and
tHcy levels have prompted studies on the effects of B-
vitamin treatment (Cbl, folic acid, and B6) on tHcy, demen-
tia, and cognitive decline in the elderly population. Since
the implementation in 1998 of the Canadian program of
folic acid fortification of grain products, there is virtually no
folic acid deficiency in our population [12,13]; therefore,
low Cbl remains the most important factor determining
elevated tHcy in elderly Canadians.
This article will summarize previous investigations on
Cbl, tHcy, and cognitive function as well as the available
results of treatment studies, some of which are still ongoing.
The aim is to analyze the need to perform Cbl, folic acid,
and/or tHcy determinations in the standard work-up for
dementia and cognitive decline in older adults in Canada.
2. Methods
MedLine database was searched for articles on Vitamin
B12 and elderly, Vitamin B12 and cognition, Vitamin B12
treatment, Homocysteine and elderly, Homocysteine and
cognition, Homocysteine treatment, and Folic acid fortifi-
cation. Articles and presentations at various meetings were
reviewed. Personal communications of unpublished results
and ongoing research are also noted when applicable.
The review includes a brief description of background
biologically relevant factors and changes during the last 6
years in Canada (folic acid fortification) and covers two
outcomes: dementia (including AD, mixed and vascular
dementia) and cognitive performance in older adults.
3. Cobalamin deficit in older adults
Human Cbl levels depend on intake and absorption from
animal-derived proteins. Cbl deficiency is an age-related
condition. Epidemiologic studies have shown that the prev-
alence of low Cbl serum concentrations rises with age
[14,15]. With 200 pg/mL as a normal cutoff point, the
incidence of low Cbl serum concentration among the elderly
population ranges from 7% to 16% [1,2,14].
It is relevant to recognize that levels of Cbl might not be
a good marker of Cbl status/function and that methylma-
lonic acid (MMA) appears to be the best indicator of Cbl
function [15,16]. With elevated MMA as indicator of Cbl
deficiency in various elderly populations, the prevalence of
Cbl deficiency can be much higher, up to 40% [17].
Cobalamin deficiency might cause or contribute to mul-
tiple abnormalities in different tissues/organs, including
megaloblastic anemia, gastrointestinal changes, peripheral
neuropathy, subacute combined degeneration, fatigue, de-
lirium, and depression. Traditionally, megaloblastic anemia
has served as a surrogate marker for the clinical investiga-
tion of Cbl deficiency. However, the effects of Cbl defi-
ciency at the hematologic and the neurologic levels are not
parallel; in fact, the concurrent presentation of both hema-
319
tologic and neurologic abnormalities is found in only 41%
of cases [18]. Analysis of cohort populations has shown that
there is no significant relationship between clinical neuro-
logic or hematologic signs or symptoms of Cbl deficiency
and Cbl levels [19,20]; therefore, presence or absence of
those signs and symptoms should not be used to guide
request for serum determinations of Cbl or assume Cbl
status. Moreover, a large percentage of older persons with
Cbl deficiency (determined either by low levels of Cbl or by
elevated MMA) have neither classic hematologic nor neu-
rologic symptoms of Cbl deficit.
4. Cobalamin and cognitive function
Low Cbl is more frequent among older persons with
dementia than among cognitively normal older adults, but
this does not imply that low Cbl is the cause of dementia or
that the dementia is reversible with Cbl treatment. Although
individual patients with low Cbl levels and dementia might
benefit from Cbl treatment, there is no epidemiologic evi-
dence that low Cbl levels per se cause cognitive decline or
increase the risk of dementia among older adults [21] or that
treatment with Cbl improves cognition in patients with
dementia.
There are reports of dementia patients with low Cbl
whose dementia improved after normalization of Cbl levels
[22,23]. Others have failed to find a significant effect of Cbl
treatment among dementia patients with low Cbl, although
a significant improvement was seen among persons with
cognitive impairment, not dementia [24].
There are no large randomized placebo-controlled trials
assessing the effects of Cbl treatment on patients with de-
mentia and low Cbl, although two small, short-term trials
with negative results have been published [25]. A long-
term, appropriately designed randomized placebo control
trial of patients with dementia and low Cbl might be uneth-
ical at this stage because low Cbl can cause multiple pa-
thologies and normalization of Cbl levels is desirable for
many reasons besides improving cognition.
It is unclear whether there is a subgroup of patients with
dementia and low Cbl who might benefit from Cbl treat-
ment, and what are the cognitive characteristics of these
patients. Nevertheless, because of the high prevalence of
low Cbl as comorbidity among older adults, other available
guidelines for dementia recommend determinations of Cbl
levels in the work-up for dementia and treatment if low Cbl
is found [26].
Among cognitively normal older persons, results on the
relationship between Cbl and cognitive decline are available.
Although we did not find that low Cbl affects cognitive decline
in normal older persons [27], two trials have shown improve-
ment of cognitive function with Cbl treatment. In a single blind
trial among cognitively normal older subjects with low Cbl,
1-month placebo followed by 5 months of treatment with Cbl
resulted in improved cognitive performance [28]. In a short-
320 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
term randomized controlled trial in cognitively normal
women of all ages with unknown Cbl levels, B-vitamin
treatment was found to slightly improve memory [29].
It appears that the effects of Cbl on cognitive function
might be related to levels of metabolites such as tHcy and
MMA as a result of intracellular depletion of Cbl, folate, or
both.
5. Homocysteine and aging
Total homocysteine levels also increase with age [4]. The
most important factors determining tHcy concentrations in
older age are Cbl and folate levels and renal function.
Homocysteine is an amino acid predominantly derived from
the methionine pathway where deficits of folate or Cbl will
result in increases of tHcy levels. Other factors such as
smoking and dehydration also increase tHcy levels. Genetic
markers of tHcy levels include methyltetrahydrofolate re-
ductase (MTHFR) polymorphisms (677C¡T). The impact
of this genetic polymorphism in older age is still being
debated. Total homocysteine is higher in men than in
women. As in Cbl deficiency, there are no reliable clinical
signs or symptoms that predict the presence of elevated
tHcy; therefore, there are no clinical indicators of elevated
tHcy to guide performing determinations [20].
Prevalence of elevated tHcy among older adults varies
slightly according to country and folic acid fortification
status. In Canada, levels of tHcy declined shortly after
fortification was implemented in 1998, but they have risen
in subsequent years in spite of continuous increase in intra-
cellular folic acid levels (red blood cell folate). Between
2003 and 2004 the prevalence of older adults older than the
age of 65 years in Southeastern Ontario with elevated tHcy
was 9%, similar to the prevalence found before fortification
(10%) [13].
6. Homocysteine and cognition
A full review of published data on the relationship be-
tween tHcy and dementia and cognitive decline can be
found in recently published reviews [30]. Only the most
important findings of prospective cohort studies only are
summarized here.
The association between tHcy levels and cognitive func-
tion in older adults has been investigated by using two types
of outcomes: (1) dementia and (2) decline in cognitive test
scores (without implying a diagnosis of dementia). The
studies investigating decline of cognitive scores as primary
outcome used a variety of psychometric tests; therefore,
there is no uniformity of results.
In the Framingham cohort study [7], 1,092 subjects cog-
nitively normal at baseline were followed for 8 years. Di-
agnosis of dementia by Diagnostic and Statistical Manual
[of Mental Disorders], Fourth Edition criteria was used as
the primary outcome. Elevated tHcy at baseline was found
to be a risk factor for dementia (both AD and vascular
dementia), with an increase in relative risk of 1.4 (confi-
dence interval, 1.1 to 1.9) per 1 standard deviation increase
in the log-transformed tHcy level (tHcy levels are not nor-
mally distributed). Of note, the study period included pre-
fortification and postfortification years in the U.S. Similar
results were found in a recently published large cohort study
of older Italians with normal cognitive function at baseline
and a follow-up of 4 years [31]. In a study by Luchsinger
et al [9] on Medicare recipients in the U.S., however, no
association between tHcy and AD was found.
Among the studies investigating the relation between
tHcy levels and cognitive scores, results of the Rotterdam
cohort showed an association between tHcy and decline in
scores of executive function and visual memory tests [32].
Another study in the U.S. involving 1,789 persons of Latino
origin older than the age of 60 years (SALSA study) found
inverse relationships between levels of tHcy and scores in a
test of executive function and a test of global cognitive
function [33]. In another large cohort study including 1,241
French people age 60 years or older, Dufouil et al [34]
found an odds ratio of 2.8 (confidence interval, 1.2 to 6.2) of
cognitive decline at 4 years in persons with elevated tHcy
levels. In a cohort of normal elderly of Southeastern Ontario
we found a significant correlation between tHcy levels at
baseline, tHcy changes over time, and decline in a test of
executive function [27]. Another cohort study from the U.S.
involving 321 men only also found an independent relation-
ship between elevated tHcy and cognitive decline [35] in-
cluding memory. Luchsinger et al [9], however, did not find
a relation between tHcy and decline in scores of memory,
whereas other studies have found significant relationship be-
tween tHcy and decline in scores of various cognitive tests
[36 –38]. Elevated tHcy levels also appear to affect frontal-
executive function decline in patients with strokes [39].
Age is an important determinant of the relationship be-
tween tHcy and cognitive decline. In a recently published
study from the Framingham cohort, Elias et al [40] found an
association between tHcy and decline in multiple cognitive
areas only in individuals younger than the age of 60. It
appears that the effects of homocysteine on cognition might
occur earlier than the target age for dementia screening,
indicating that perhaps primary prevention treatment trials
should include younger subjects.
7. Effects of B vitamins on tHcys
Total homocysteine levels can be lowered with B vita-
mins (folic acid, Cbl, and B6).
Deficits of folic acid cause elevated tHcy. In 1998 for-
tification of grain products with folic acid became manda-
tory in Canada. The aim of the fortification program was to
reduce the incidence of neural tube defects in newborns.
Canada, USA, and Chile are the only countries in the world
with national folic acid fortification programs for more than
 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
5 years. This fact might explain some differences encoun-
tered in studies from different countries and over time.
Shortly after fortification was initiated in Canada, tHcy
levels in older adults (and in the population at large) were
significantly lower than levels before fortification [12].
Analysis of tHcy levels in older adults in Southeastern
Ontario during the years thereafter, however, has shown that
levels of tHcy have increased again to the same level seen
in the prefortification period in spite of constantly increas-
ing concentrations of red blood cell folate that are, at
present, above normal in 92% of older adults. By the end of
2004, in spite of fortification, 16% of older adults in our
population of Southeastern Ontario not taking oral Cbl sup-
plements had elevated tHcy (3% had low Cbl levels),
whereas only 2% of those taking oral Cbl supplements had
elevated tHcy (0% had low Cbl levels) [13]. None had low
red blood cell folate. It appears that in the presence of
excess intracellular folic acid, a large percentage of older
adults not taking Cbl supplements in Canada have elevated
tHcy with normal Cbl levels.
The largest majority of older adults absorb oral crystal-
line Cbl because it is not bound to protein, because the Cbl
deficiency is most frequently due to atrophic gastritis. In
patients with low Cbl levels, oral Cbl decreases tHcy in a
dose-dependent manner, with 1,000 ␮g/day being the most
effective dose to normalize tHcy [41]. It appears that ele-
vated tHcy and low Cbl levels in older adults can be cor-
rected with oral Cbl supplements. Smaller doses can be
effective in preventing elevated tHcy in persons with nor-
mal Cbl. Among community-dwelling older persons with
normal Cbl levels taking between 25 and 37.5 ␮g/day of
oral Cbl, we did not find any subjects with elevated tHcy,
whereas 13.2% of those with normal Cbl levels not taking
any Cbl oral supplements had elevated tHcy [42].
The role of vitamin B6 in tHcy levels has not been fully
investigated, but it appears that B6 alone does not signifi-
cantly decrease tHcy levels [43].
8. Effects of lowering tHcys treatment on cognition
and dementia
There are no published large, long-term randomized
placebo-controlled trials of reduction of tHcy levels with
B vitamins and cognition as primary outcome in patients
with elevated tHcy and dementia or in patients with
dementia regardless of their tHcy level.
In a recently published case series of patients with dementia
and elevated tHcy levels, the combination treatment of B
vitamins and N-acetylcysteine (antioxidant) resulted in cogni-
tive improvement in 5 of 7 patients [44]. Successful improve-
ment of executive cognitive function with homocysteine-
lowering treatment in an individual patient has been reported
[45].
The VITAL trial is aimed at analyzing the effects of oral
high dose B vitamins and aspirin on cognitive performance
321
in patients at high risk of dementia [46]. Final results of this
trial are pending. Several large trials in different groups of
patients include cognitive outcomes, but none has preven-
tion or treatment of dementia as a primary outcome. These
include the VITATOPS in which B vitamins are adminis-
tered for secondary prevention of cerebrovascular events
[47]. The HOPE study, recently published, showed negative
results of B vitamins for secondary prevention of cardio-
vascular events [48]. Although a Mini-Mental State Exam-
ination was conducted in a subgroup of the trial population,
those results have not been published yet. Stott et al [49]
recently published results of a 3-month randomized con-
trolled trial (RCT) in patients with vascular disease. No
effects of B vitamins on cognitive function were seen. An
efficacy study from the VISP trial has shown that a sub-
group of participants (those with Cbl levels between 250
and 637 pmol/L) might benefit from B vitamins [50], but it
is unclear whether this beneficial effect would result in
cognitive improvement.
Current RCTs are analyzing the effects of B vitamins as
primary prevention of cognitive decline in older adults. One
such trial [51] conducted in Holland (a country without folic
acid fortification) showed a significant difference in cogni-
tive decline between the placebo and the active treatment
(folic acid), with worse outcomes in the placebo arm, at 3
years. A short (4 months) RCT on primary prevention of
cognitive decline in older adults failed to show significant
effects of treatment on cognitive performance [52]. Other
trials have been completed and are pending publication
(personal communication).
9. Conclusions
1. Low Cbl is a frequent comorbidity among cogni-
tively normal older adults and older patients with
dementia.
2. Low Cbl can be corrected with either oral or paren-
teral Cbl.
3. The direct relationship between low Cbl and cogni-
tive function remains unclear.
4. MMA is a better indicator of Cbl function than Cbl
levels.
5. Treatment with Cbl in patients with low Cbl and
dementia has yielded variable results.
6. There are no large randomized placebo-controlled
trials of Cbl treatment in patients with low Cbl and
dementia, and because of the wide range of abnor-
malities found in Cbl deficiency states, such trials
might be unethical.
7. Treatment of low Cbl in older adults with normal
cognitive function might improve cognitive scores.
8. Elevated tHcy is frequent among older adults, even
in the presence of elevated intracellular folic acid
levels (red blood cell folate).
322 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
9. Elevated tHcy is a risk factor for dementia.
10. Elevated tHcy seems to affect cognitive decline in
older adults.
11. Elevated tHcy can be corrected with B vitamins/Cbl
even in the presence of elevated red blood cell
folate.
12. There are case reports of successful treatment of
elevated tHcy in patients with dementia.
13. There are no published RCTs of B-vitamin treatment
in patients with dementia of the Alzheimer’s type
and elevated tHcy. Such trials are ongoing.
14. Short-term RCTs in patients with vascular dementia
showed no effects of B vitamins on cognition.
15. Results from primary prevention RCTs (in cogni-
tively normal adults with or without elevated tHcy)
are inconclusive. Larger studies are needed.
16. Red blood cell folate/total folate levels are elevated
in the largest majority of older adults after imple-
mentation of the folic acid fortification program.
10. Recommendations reaching consensus at the Third
Canadian Consensus Conference on Diagnosis and
Treatment of Dementia
1. It is recommended that serum Cbl levels be deter-
mined in all older adults suspected of dementia or
cognitive decline (Grade B, Level 2).
2. Older adults found to have low Cbl levels should be
treated with Cbl (either oral or parenteral forms) be-
cause of potential improvement of cognitive function
and the deleterious effects of low Cbl levels on mul-
tiple organ systems, besides the effects on cognition
(Grade B, Level 2).
3. There is currently insufficient evidence to support the
need for serum tHcy levels to be determined in older
adults suspected of dementia or cognitive decline
(Grade C, Level 3).
4. There is currently insufficient evidence that treatment
of elevated serum tHcy levels affects cognition
(Grade C, Level 3).
5. Determination of serum folic acid or red blood cell
folate in older adults in Canada is optional and might
be reserved for patients with celiac disease, inade-
quate diets, or other conditions that prevent them
from ingesting grain products (Grade E, Level 2).
11. Addendum
Since the submission of this paper, two other papers have
been published that are relevant to the topic discussed.
Although their conclusions do not change ours, or the rec-
ommendations, it is relevant to note them. The first [53] is
the results of a 2-year randomized placebo controlled trial
(RCT) on cognitively normal elderly with elevated tHcy at
baseline and a battery of 8 cognitive tests as end-points. The
treatment arm consisted of daily high dose of Folic acid,
B12 and B6. The authors did not find any improvement, but
decline in one test (the Part B Reitan Trails Making Test) in
the active treatment arm compared to placebo, raisings
questions that remain unexplained [54]. The second [55]
publishes the results obtained in a 3-years RCT with high
dose folic acid in 818 cognitively normal subjects ages 50 to
70 with elevated tHcy. The authors found a significant
improvement in the composite scores of memory, informa-
tion processing speed and sensorimotor speed in the active
treatment group.
Author Disclosures
Angeles Garcia has received support from Janssen-Ortho
(Consultant, Advisor), Pfizer (Consultant, Advisor), Novartis
(Consultant, Advisor), and Lundbeck (Consultant, Advisor).
Acknowledgments
This work was supported by The Ontario Mental Health
Foundation and Queen’s University.
References
[1] Lindenbaum J, Rosenberg IH, Wilson PWF, Stabler SP, Allen RH.
Prevalence of cobalamin deficiency in the Framingham elderly pop-
ulation. Am J Clin Nutr 1994;60:2–11.
[2] Joosten E, van den Berg A, Riezler R, Naurath HJ, Lindenbaum J,
Stabler SP, Allen RH. Metabolic evidence that deficiencies of B12
(cobalamin), folate, and vitamin B6 occur commonly in elderly peo-
ple. Am J Clin Nutr 1993;58:468 –76.
[3] Stabler SP, Marcell PD, Podell ER, Allen RH, Savage DG, Linder-
baum J. Elevations of total homocysteine in the serum of patients with
cobalamin or folate deficiency detected by capillary gas chromatog-
raphy-mass spectrometry. J Clin Invest 1988;81:466 –74.
[4] Selhub J, Jaques PF, Rosenberg IH, Rogers G, Bowman BA, Gunter
EW, et al. Serum total homocysteine concentrations in the third
national health and nutrition examination survey (1991–1994): pop-
ulation reference ranges and contribution of vitamin status to high
serum concentrations. Ann Intern Med 1999;131:331–9.
[5] The Homocysteine Studies Collaboration. Homocysteine and risk of
ischemic heart disease and stroke: a meta-analysis. JAMA 2002;288:
2015–22.
[6] Boysen G, Brander T, Christensen H, Gideon R, Truelsen T. Homo-
cysteine and risk of recurrent stroke. Stroke 2003;34:1258 – 61.
[7] Seshadri S, Beiser A, Selhub J, et al. Plasma homocysteine as a risk
factor for dementia and Alzheimer’s disease. N Engl J Med 2002;
346:476 – 83.
[8] Wald DS, Law M, Morris JK. Homocysteine and cardiovascular
disease: evidence on causality from a meta-analysis. BMJ 2002;325:
1202– 8.
[9] Luchsinger JA, Tang MX, Shea S, Miller J, Green R, Mayeux R.
Plasma homocysteine levels and risk of Alzheimer disease. Neurol-
ogy 2004;62:1972– 6.
[10] Wright CB, Paik MC, Brown TR, Stabler SP, Allen RH, Sacco RL,
et al. Total homocysteine is associated with white matter hyperinten-
sity volume: the Northern Manhattan Study. Stroke 2005;36:1207–
11.
 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
[11] Vermeer SE, Den Heijer T, Koudstaal PJ, Oudkerk M, Hofman A,
Breteler MMB. Rotterdam Scan Study: incidence and risk factors
of silent brain infarcts in the population-based Rotterdam Scan
Study. Stroke 2003;34:392– 6.
[12] Ray JG, Vermeulen MJ, Boss SC, Cole DEC. Declining rate of folate
insufficiency among adults following increased folic acid food forti-
fication in Canada. Can J Public Health 2002;93:249 –53.
[13] Day A, Zanibbi K, Garcia A. Red blood cell folate, homocysteine,
cobalamin and methylmalonic acid through the folic acid fortification
years. Hematol Rep 2005;1:58.
[14] Pennypacker LC, Allen RH, Kelly JP, Matthews LM, Grigsby J, Kaye
K, et al. High prevalence of cobalamin deficiency in elderly outpa-
tients. J Am Geriatr Soc 1994;40:1197–204.
[15] Clarke R, Evans JG, Schneede J, Nexo E, Bates C, Fletcher A, et al.
Vitamin B12 and folate deficiency in later life. Age Ageing 2004;33:
34 – 41.
[16] Allen RH, Stabler SP, Savage DG, Lindenbaum J. Diagnosis of
cobalamin deficiency I: usefulness of serum methylmalonic acid and
total homocysteine concentrations. Am J Hematol 1990;34:90 – 8.
[17] Andres E, Loukili NH, Noel E, Kaltenbach G, Abdelgheni MB, Perrin
AE, et al. Vitamin B12 (cobalamin) deficiency in elderly patients.
CMAJ 2004;171:251–9.
[18] Healton EB, Savage DG, Brust JCM, Garrett TJ, Lindenbaum J. Neu-
rologic aspects of cobalamin deficiency. Medicine 1991;70:229 – 44.
[19] Bjorkegren K, Svardsudd K. Reported symptoms and clinical findings
in relation to serum cobalamin, folate, methylmalonic acid and total
homocysteine among elderly Swedes: a population-based study. J In-
tern Med 2003;254:343–52.
[20] Garcia A, Haron Y. Predictive values of signs and symptoms in
cobalamin deficiency. Gerontology 2002;42:34.
[21] Crystal HA, Ortof E, Frishman WH, Gruber A, Hershman D, Aronson
M. Serum vitamin B12 levels and incidence of dementia in a healthy
elderly population: a report from the Bronx Longitudinal Aging
Study. J Am Geriatr Soc 1994;42:933– 6.
[22] Cunha UG, Rocha FL, Peixoto JM, Motta MF, Barbosa MT. Vitamin
B12 deficiency and dementia. Int Psychogeriatr 1995;7:85– 8.
[23] Abyad A. Prevalence of vitamin B12 deficiency among demented
patients and cognitive recovery with cobalamin replacement. J Nutr
Health Aging 2002;6:254 – 60.
[24] Eastley R, Wilcock GK, Bucks RS. Vitamin B12 deficiency in de-
mentia and cognitive impairment: the effects of treatment on neuro-
psychological function. Int J Geriatr Psychiatry 2000;15:226 –33.
[25] Malouf R, Areosa Sastre A. Vitamin B12 for cognition. Cochrane
Database Syst Rev 2003.CD004326.
[26] Knopman DS, DeKosky ST, Cummings JL, Chui H, Corey-Bloom J,
Relkin N, et al. Practice parameter: diagnosis of dementia (an evidence-
based review): Report of the Quality Standards Subcommittee of the
American Academy of Neurology. Neurology 2001;56:1143–53.
[27] Garcia A, Haron Y, Pulman K, Hua L, Freedman M. Increases in
homocysteine are related to worsening of Stroop scores in normal
elderly subjects: a prospective follow-up study. J Gerontol A Biol Sci
Med Sci 2004;59:1323–7.
[28] van Asselt DZ, Pasman JW, van Lier HJ, Vingerhoets DM, Poels
PJ, Kuin Y, et al. Cobalamin supplementation improves cognitive
and cerebral function in older, cobalamin-deficient persons. J
Gerontol A Biol Sci Med Sci 2001;56:775–9.
[29] Bryan J, Calvaresi E, Hughes D. Short-term folate, vitamin B-12 or
vitamin B-6 supplementation slightly affects memory performance
but not mood in women of various ages. J Nutr 2002;132:1345–56.
[30] Garcia A, Zanibbi K. Homocysteine and cognitive function in elderly
people. CMAJ 2004;171:897–904.
[31] Ravaglia G, Forti P, Maioli F, Martelli M, Servadei L, Brunetti N, et
al. Homocysteine and folate as risk factors for dementia and Alzhei-
mer disease. Am J Clin Nutr 2005;82:636 – 43.
323
[32] Prins ND, den Heijer T, Hofman A, Koudstaal PJ, Jolles J, Clarke R,
Breteler MMB. Homocysteine and cognitive function in the elderly:
the Rotterdam scan study. Neurology 2002;59:1375– 80.
[33] Miller JW, Green R, Ramos MI, Allen LH, Mungas DM, Jagust WJ,
et al. Homocysteine and cognitive function in the Sacramento Area
Latino Study on Aging. Am J Clin Nutr 2003;78:441–7.
[34] Dufouil C, Alperovitch A, Ducros V, Tzourio C. Homocysteine,
white matter hyperintensities, and cognition in healthy elderly people.
Ann Neurol 2003;53:214 –21.
[35] Tucker KL, Qiao N, Scott T, Rosenberg I, Spiro A 3rd. High homo-
cysteine and low B vitamins predict cognitive decline in aging men:
the Veterans Affairs Normative Aging Study. Am J Clin Nutr 2005;
82:627–35.
[36] Kalmijn S, Launer LJ, Lindemans J, Michiel B, Hofman A, Breteler
MM. Total homocysteine and cognitive decline in a community-
based sample of elderly subjects: the Rotterdam Study. Am J Epide-
miol 1999;150:283–9.
[37] McCaddon A, Hudson P, Davies G, Hughes A, Williams JH, Wilkin-
son C. Homocysteine and cognitive decline in healthy elderly. De-
ment Geriatr Cogn Disord 2001;12:309 –13.
[38] Teunissen CE, Blom AHJ, Van Boxtel PJ, Bosma H, de Bruijn C,
Jolles J, et al. Homocysteine: a marker for cognitive performance? a
longitudinal follow-up study. J Nutr Health Aging 2003;7:153–9.
[39] Sachdev PS, Valenzuela MJ, Brodaty H, Wang XL, Looi J, Lorentz
L, et al. Homocysteine as a risk factor for cognitive impairment in
stroke patients. Dement Geriatr Cogn Disord 2003;15:155– 62.
[40] Elias MF, Sullivan LM, D’Agostino RB, Elias PK, Jacques PF,
Selhub J, et al. Homocysteine and cognitive performance in the
Framingham offspring study: age is important. Am J Epidemiol
2005;162:644 –53.
[41] Eussen SPM, de Groot LCPG, Clarke R, Schneede J, Ueland PM,
Hoefnagels WHL, et al. Oral cyanocobalamin supplementation in
older people with vitamin B12 deficiency. Arch Intern Med 2005;
165:1167–72.
[42] Garcia A, Evans L, Paris-Pombo A, Freedman, M. Is low dose oral
cobalamin enough to normalize cobalamin levels and function in the
elderly? J Am Geriatr Soc 2002:50:1401– 4.
[43] Malouf R, Grimley Evans J. The effect of vitamin B6 on cognition.
Cochrane Database Syst Rev 2003.CD004393.
[44] McCaddon A. Homocysteine and cognitive impairment: a case series
in a general practice setting. Nutr J 2006;5:6 –13.
[45] Boxer AL, Kramer JH, Johnston K, Goldman J, Finley R, Miller
BL. Executive dysfunction in hyperhomocystinemia responds to
homocysteine-lowering treatment. Neurology 2005;64:1431– 4.
[46] VITAL collaborative group: effects of vitamins and aspirin on mark-
ers of platelet activation, oxidative stress and homocysteine in people
at high risk for dementia. J Intern Med 2003;254:67–75.
[47] VITATOPS Trial Study Group. The VITATOPS (Vitamins to Pre-
vent Stroke) Trial: rationale and design of an international, large,
simple, randomized trial of homocysteine-lowering multivitamin
therapy in patients with recent transient ischaemic attack or stroke.
Cerebrovasc Dis 2002;13:120 – 6.
[48] Lonn E, Yusuf S, Arnold MJ, Sheridan P, Pogue J, Micks M, et al.
Heart Outcomes Prevention Evaluation (HOPE) 2 investigators: ho-
mocysteine lowering with folic acid and B vitamins in vascular
disease. N Engl J Med 2006;354:1567–77.
[49] Stott DJ, MacIntosh G, Lowe GD, Rumley A, McMahon AD, Lang-
horne P, et al. Randomized controlled trial of homocysteine-lowering
vitamin treatment in elderly patients with vascular disease. Am J Clin
Nutr 2005;82:1320 – 6.
[50] Spence JD, Bang H, Chambless LE, Stampher MJ. Vitamin Interven-
tion For Stroke Prevention trial: an efficacy analysis. Stroke 2005;
36:2404 –9.
324 A. Garcia / Alzheimer’s & Dementia 3 (2007) 318 –324
[51] Durga J, van Boxtel MPJ, Schouten EG, Kok FJ, Jolles J, Katan MB,
et al. The effect of 3-year folic acid supplementation on cognitive
function, a randomized controlled trial. Hematol Rep 2005;1:1.
[52] Lewerin C, Matousek M, Steen G, Johansson B, Steen B, Nilsson-
Ehle H. Significant correlations of plasma homocysteine and serum
methylmalonic acid with movement and cognitive performance in
elderly subjects but no improvement from short-term vitamin therapy:
a placebo-controlled randomized study. Am J Clin Nutr 2005;81:
1155– 62.
[53] McMahon JA, Green TJ, Skeaff CM, Knight RG, Mann JI, Williams
SM. A controlled trial of homocysteine lowering and cognitive per-
formance. N Engl J Med 2006;354:2764 –72.
[54] Clarke R. Vitamin B12, folic acid, and the prevention of dementia.
N Engl J Med. 2006;354:2817–9.
[55] Durga J, van Boxtel MPJ, Schouten EG, Kok FJ, Jolles J, Katan MB,
et al. Effect of 3-year folic acid supplementation on cognitive func-
tion in older adults in the FACIT trial: a randomised, double blind,
controlled trial. Lancet 2007;369:208 –16.
 Alzheimer’s & Dementia 3 (2007) 325–332

The role of biologic markers in the diagnosis of Alzheimer’s disease

Hyman M. Schipper*
Centre for Neurotranslational Research and Bloomfield Centre for Research in Aging, Lady Davis Institute for Medical Research, Sir Mortimer B.
Davis-Jewish General Hospital, McGill University, Montreal, Quebec, Canada; Department of Neurology and Neurosurgery, McGill University,
Montreal, Quebec, Canada

1. Introduction
The objective of this report is to establish consensus
criteria for the use of biologic markers in the diagnosis of
Alzheimer’s disease (AD). To address this issue, we will
first provide background information concerning the signif-
icance of biologic AD markers, differences between genetic
markers, genetic modifiers, and biologic markers, criteria
for an ideal biologic marker, strengths and weaknesses of
biologic markers currently available, and promising future
research developments in AD biomarker discovery. Specific
recommendations to primary care physicians and specialists
will follow.
role to play in the management of patients with the far more
common sporadic form of the illness. The presence of the
apolipoprotein E (APOE) e4 allele, a genetic modifier, is a
strong risk factor for the development of sporadic AD and
predicts earlier onset of clinical dementia in AD patients.
However, the presence of the e4 allele does not confer AD,
and its absence does not exclude this disorder. Thus, testing
for the e4 allele cannot be used as a diagnostic marker of
sporadic AD. Unlike the genetic markers and modifiers, true
biologic markers of AD are “state” indicators that inform of
the presence or absence of AD at the time of measurement
and are thus appropriate for use as diagnostic modalities for
sporadic (and possibly familial) forms of this condition.

2. What is a biologic marker?

A biologic marker of disease is a measurable change in
the physical constitution of an organism that indicates the
presence of the disease. Current biologic indices under in-
vestigation for the early diagnosis of AD include chemical
markers detected in body fluids and brain volume or activity
measurements derived from neuroimaging modalities such
as magnetic resonance imaging (MRI) or positron emission
tomography (PET) of relevant brain regions. Neuroimaging
modalities are expensive, labor-intensive, and not univer-
sally available, prompting a search for reliable chemical
biomarkers and other practical neurodiagnostic modalities.
Chemical markers of AD can be conceptualized within three
general categories: (1) genetic markers, (2) genetic modifiers,
and (3) biologic markers. Genetic markers such as mutant
forms of amyloid precursor protein (APP), presenilin-1, and
presenilin-2 are excellent for predicting disease in rare kin-
dreds with familial AD but are not useful in tracking disease
progression or efficacy of therapeutic intervention in these
subjects. Moreover, these genetic markers have little or no
*Corresponding author. Tel.: 514-340-8260; Fax: 514-340-8925.
E-mail address: hyman.schipper@mcgill.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.015
3. How would a biologic marker assist in the
management of AD?
The advent of a biologic marker that reliably indicates
the presence of AD and differentiates the latter from other
dementing conditions would represent a major achievement
in the management of this common neurodegenerative dis-
order. Specific benefits accruing from successful implemen-
tation of such a marker in clinical practice would include
(1) more rapid and accurate diagnosis of sporadic AD;
(2) ability to diagnose AD in patients with major affective
disorders, clouded sensorium, depressed level of conscious-
ness, and other clinical conditions that often preclude es-
tablishment of a dementia diagnosis by conventional means;
(3) possible monitoring of the progression and severity of
AD and responses to therapeutic interventions; (4) potential
prognostication of mild cognitive impairment (MCI) sub-
jects at highest risk for early conversion to incipient AD;
(5) assembly of homogeneous patient cohorts for therapeu-
tic clinical trials; and (6) considerable cost savings related to
curtailment of ancillary biochemical and imaging modalities
currently used to exclude other causes of dementia.
326 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
4. What are the characteristics of an ideal biologic
marker for use in AD?
The quest for the development of reliable AD biomarkers
worldwide is guided, in large measure, by principles set
forth in a Consensus Report on Molecular and Biochemical
Markers of AD sponsored by the Alzheimer’s Association
(US) and the National Institute on Aging (Neurobiology of
Aging 1998;19:107– 67). In this landmark report it was
suggested an ideal biologic marker of AD fulfill the follow-
ing criteria [1]:
(1) Reflect a fundamental aspect of central nervous sys-
tem (CNS) pathophysiology in AD (plausibility)
(2) Indicate the actual presence of AD and not merely
increased risk
(3) Exhibit high sensitivity and specificity (in the range
of 80% or better for each)
(4) Be efficacious in early or pre-clinical AD (e.g. MCI)
(5) Monitor disease severity or rate of progression
(6) Indicate efficacy of therapeutic intervention
(7) Be noninvasive, inexpensive, and readily available.
In the aftermath of this consensus report, it was also deemed
desirable that (8) the efficacy of the putative biomarker is
corroborated by at least one other independent laboratory, and
that its accuracy (Criterion 3) is demonstrated in discriminating
AD not only from cognitively healthy controls but from pa-
tients with various non-AD dementias [2].
5. What biologic markers are currently available for
the diagnosis of AD?
5.1. CSF A␤1-42
Increased plasma ␤-amyloid1-42 (A␤1-42) levels have
been documented in several kindreds with familial AD [3],
but these families represent a very small subset of the total
AD population. Measurements of total A␤ peptide, A␤1-40,
or soluble APP␣/␤ concentrations in cerebrospinal fluid
(CSF) or blood have not proved useful in the diagnosis of
sporadic AD [4 – 8]. Amyloid fragments are plausible AD
biomarkers because they directly reflect a known patho-
physiologic process in AD brain (senile plaque formation).
Numerous laboratories have demonstrated abnormally low
levels of the amyloid peptide fragment A␤1-42 in the CSF of
sporadic AD patients and individuals with MCI [5]. A
meta-analysis performed by Sunderland et al [9] involving
18 studies of CSF A␤1-42 as an AD diagnostic disclosed an
effect size of 1.56 (95% confidence interval [CI]: 1.43
to 1.69). In 2003 Blennow and Hampel [4] reviewed CSF
A␤1-42 data from 13 studies (⬃600 AD and 450 control
subjects) that used the Innogenetics enzyme-linked immu-
nosorbent assay (Gent, Belgium). They calculated an over-
all sensitivity and specificity of 80% and 90%, respectively,
for differentiating AD from healthy cognitive aging on the
basis of this assay. However, the specificity of this test
remains considerably lower when differentiating AD from
control groups with other causes of dementia. For example,
in a multi-center study involving 150 AD, 100 normal
elderly controls (NEC), and 79 cases of non-AD dementia
[10], the specificity of CSF A␤1-42 in differentiating AD
from NEC and patients with other dementias was 81% and
59%, respectively. This is because CSF A␤1-42 might be
decreased in other degenerative and nondegenerative CNS
disorders including Lewy body dementia (LBD) [11–13],
Creutzfeldt-Jakob disease (CJD) [12,14], multisystem atro-
phy [15], and amyotrophic lateral sclerosis (ALS) [16]. The
advent of mass spectrometry techniques has led to the recent
identification of novel amyloid peptide fragments in AD
CSF that might prove to be of diagnostic significance [17].
5.2. CSF total tau
Increased concentrations of total tau protein (t-tau) have
been consistently observed in AD CSF. A meta-analysis
performed by Sunderland et al [9] involving 35 studies of
CSF tau as an AD diagnostic disclosed an effect size of 1.31
(95% CI, 1.23 to 1.39). In 2003 Blennow and Hampel [4]
reviewed CSF t-tau data from 41 studies (more than 4,000
AD and control subjects) that used either of the two most
commonly available t-tau assays, the Innogenetics or
Athena enzyme-linked immunosorbent assay (Athena Neu-
rosciences, Worcester, MA). Overall, the sensitivity and
specificity of these assays for the diagnosis of AD were in
the range of 80% and 90%, respectively (similar to those of
CSF A␤1-42). Although CSF t-tau distinguished AD patients
from neurologically healthy controls with high accuracy,
the marker performs considerably less well in discriminat-
ing AD from other dementias, with reported specificities of
57% for suspected non-AD dementias [10] and 69% for
autopsy-proven cases [12]. High levels of CSF t-tau might
also predict progression of cognitive decline in MCI, espe-
cially in patients without extensive periventricular white
matter lesions [18]. CSF t-tau appears to be a general
marker for neuronal destruction in a wide range of degen-
erative and nondegenerative CNS conditions and is there-
fore fairly nonspecific for AD. In addition to AD, elevated
CSF t-tau levels might occur in vascular dementia [19],
frontotemporal dementia (FTD) [20], CJD, and (transiently)
in acute ischemic stroke [21]. In a recent study, 34% of
subjects with FTD exhibited significantly low levels of CSF
tau, a finding not observed in the AD cohort [22]. CSF t-tau
values in vascular dementia [23] and LBD [11] are reportedly
intermediate between those of NEC and subjects with AD.
5.3. CSF phospho-tau
By using antibodies against various phosphorylated
epitopes of tau (p-tau), a number of laboratories have doc-
umented significant increases in levels of hyperphosphory-
lated tau in AD CSF relative to NEC and, more importantly,
patients with other dementing and nondementing neurologic
 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
disorders [4,24,25]. Thus, in contrast to t-tau, augmented
CSF p-tau levels might discriminate AD from FTD [26,27],
vascular dementia [28], LBD [29], Parkinson’s disease (PD)
[30], ALS, acute stroke [31], major depression [30], and
schizophrenia [32]. CSF levels of threonine 181 p-tau have
recently been reported to be elevated in sporadic and variant
CJD [33], despite an earlier report to the contrary [34].
These phospho-tau isoforms are plausible AD biomarkers
because they directly reflect a known pathophysiologic pro-
cess in AD brain (neurofibrillary tangle formation). They
are relatively early markers of the disease inasmuch as their
concentrations in CSF are significantly elevated in incipient
AD [35] and MCI [36,37]. On the other hand, CSF p-tau
might not be useful in monitoring disease progression in AD
(possibly as a result of dilutional factors) unless combined
with MRI measurements of hippocampal atrophy in these
patients [38].
5.4. CSF A␤1-42 and p-tau combined
When measured together, CSF A␤1-42 and phospho-tau
exhibit sensitivities for detecting AD when present and
specificities for excluding other dementing disorders in the
respectable range of 80% to 90% [39]. At a prevalence rate
of 45%, the positive and negative predictive values of the
combined test are 90% and 95%, respectively [5]. In addi-
tion, these markers identify AD in early stages (eg, MCI),
reflect the underlying pathophysiology of the disease (see
above), and are relatively inexpensive. It has been argued
that CSF A␤1-42 reflects the stage of the disease (with levels
declining progressively as a function of disease duration),
whereas t-tau and p-tau reflect the intensity of the disease
process (with higher CSF levels connoting more rapid pro-
gression) [40]. Also, the degree of CSF A␤1-42 suppression
and tau elevation might correlate with the APOE e4 allele
burden [41], although whether and to what extent the ge-
netic data should be considered in the analysis of the CSF
biomarker levels remain unclear. Importantly, elevation of
CSF tau and suppressed CSF A␤1-42 in MCI might prog-
nosticate for incipient conversion to AD with sensitivities/
specificities in the range of 83% to 90% [42,43]. This
marker combination showed efficacy in identifying “non-
progressors” in “mixed” (amnestic and non-amnestic) MCI
during a 3-year median follow-up period [44] and might be
helpful in discriminating pre-AD MCI from anxiety/depression
in the elderly [45]. We are aware of no studies implicating
CSF A␤1-42 and tau measurements as useful indices of
therapeutic efficacy in AD.
5.4.1. Caveats
(1) In the vast majority of these AD biomarker studies,
receiver operating characteristic curves (plotting the
relationship between sensitivities and specificities)
were determined solely on the basis of clinical diag-
noses without autopsy confirmation, thereby limiting
the validity of the data. Although prospective AD
327
biomarker studies are deemed more valuable than
retrospective analyses, the former are less likely to
entail neuropathologic diagnoses [46].
(2) CSF examination by lumbar puncture (spinal tap) is
relatively invasive and therefore generally not consid-
ered suitable for mass screening of elderly individuals
with AD risk factors or mild memory impairment. Yet,
in a recent study of 342 AD, MCI, and NEC subjects
who underwent 428 research lumbar punctures, there
was a very low rate of adverse effects (eg, post–lumbar
puncture headaches in 0.93%), and the procedure
was well-tolerated (low anxiety and pain ratings in
visual analog scales) by all patient groups [47].
(3) The immunoassay techniques used to measure CSF
A␤1-42 and tau are not trivial, and CSF specimens
usually must be shipped to specialized laboratories
adept at performing these immunoassays in a reliable
and reproducible fashion.
(4) The cost per sample of the combined CSF A␤1-42
and tau assays at Athena Neurosciences is US$1130
to insurance companies and $765 to MDs. The cost
of an AD biomarker might not be prohibitive if
it obviates the need for additional testing (eg,
neuroimaging).
5.5. Urine AD7C-neuronal thread protein
Measurement of elevated AD7C–neuronal thread protein
(NTP) in urine, an assay commercialized by Nymox Cor-
poration (Montreal, Quebec, Canada), purportedly differen-
tiates AD from control subjects with high sensitivity and
specificity. The protein induces neuritic sprouting and apo-
ptosis in transfected neuronal cells [48] and might nonspe-
cifically reflect AD pathogenesis. However, measurement of
urinary AD7C levels requires extensive protein purification
before customized immunoassay, and the NYMOX findings
await confirmation from independent laboratories. In July
2005 the Food & Drug Administration denied approval of
the urinary AD7C-NTP measurement as an AD neurodi-
agnostic, citing excessive overlap of values between di-
agnostic groups, lack of autopsy confirmation, and ques-
tionable clinical utility (http://www.fdaadvisorycommittee.
com/FDC/AdvisoryCommittee/Committees/Device⫹Panels/
071505_Alzheimertest/071505_AlzheimAlertR.htm). The test
remains available to physicians through Nymox’s Clinical
Laboratories Improvement Act– certified laboratory in New
Jersey.
6. Are there any promising AD biomarkers currently
under development?
6.1. Serum p97 protein (melanotransferrin)
At the time of the Second Canadian Consensus Confer-
ence on Diagnosis and Treatment of Dementia [49], there
was considerable enthusiasm regarding the potential use of
328 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
elevated serum levels of the iron-binding protein p97 (mela-
notransferrin) for the diagnosis of sporadic AD [50,51].
Unfortunately, the interest has dissipated on recognition that
the immune-based measurements of this circulating protein
were probably artifactual [52].
6.2. Platelet APP ratios
Blood platelets contain many of the same amyloid-
processing enzymes present in brain and generate small
amounts of A␤ from APP. Several laboratories have docu-
mented reductions in the 130-kd APP isoform:110-kd APP
isoform ratios in AD platelets relative to NEC and non-AD
dementias [53–55]. In a study of platelet APP ratios involv-
ing 85 probable AD patients and 95 healthy and neurologic
controls, Padovani et al [56] successfully diagnosed AD
with 88.2% sensitivity and 89.4% specificity. Additional
studies provided evidence that low APP 130:110 ratios
correlate with the severity of AD dementia [57,58] and
predict conversion to AD in subjects with MCI [59,60].
Intriguingly, there are at least three reports of APP ratio
“normalization” in AD patients treated with cholinesterase
inhibitors, implicating platelet APP as a possible reporter
molecule of effective therapeutic intervention (Section 4,
Criterion #6) [61– 63]. On the negative side, platelet APP
isoform ratios are determined by analyses of band densities
obtained on Western blotting, a labor-intensive and rela-
tively expensive technique that resists accurate protein
quantification, inter-laboratory standardization, and high-
throughput processing.
6.3. CSF and urine F2-isoprostanes
Oxidative molecular damage has been implicated in
the pathophysiology of numerous aging-related neurode-
generative disorders [64,65]. Several indices of oxidative
substrate damage, such as 8-hydroxy-2-deoxyguanosine and
4-hydroxynonenal, might be augmented in AD CSF relative
to control values. However, many oxidative metabolites are
either nonspecific for AD pathology or difficult to measure
because they are present in trace amounts, chemically un-
stable, or require elaborate technology [66,67]. Elevated
concentrations of F2-isoprostanes, specific indices of lipid
peroxidation, have been reported in the CSF, plasma, and/or
urine of patients with sporadic AD [68,69] and MCI [38,70]
relative to control values. In a recent (but relatively small)
study, inclusion of CSF isoprostane data facilitated dis-
crimination of probable AD from NEC and non-AD de-
mentia above and beyond that conferred by measure-
ments of A␤1-42/tau alone [71].
6.3.1. Caveats
(1) A laboratory at the University of Washington was
able to reproduce the data of Pratico et al [70] in
CSF, but not in blood or urine [72].
(2) F2-isoprostane levels might be elevated in the CSF of
other potentially dementing neurologic disorders,
including Huntington disease, CJD, and multiple
sclerosis [73].
(3) A clinical test based on CSF or urinary isoprostane
measurements might be difficult to standardize and
make readily available because quantification of these
lipid products requires complex gas chromatography/
mass spectrometry techniques using internal standards
that are often difficult to procure.
6.4. Plasma heme oxygenase-1 suppressor factor
In 2000 our laboratory reported that suppressed levels of
heme oxygenase-1 (HO-1) mRNA (measured by Northern
blot) in peripheral blood mononuclear cells distinguish spo-
radic AD from NEC and patients with nonAD dementias,
with a sensitivity and specificity of 88% and 75%, respec-
tively [74]. These findings have been replicated by an in-
dependent laboratory with a sensitive reverse transcriptase
polymerase chain reaction technique [75]. However, HO-1
mRNA procured from AD blood cells was subsequently
shown to display chemical instability [76], confounding its
development as a clinical neurodiagnostic. More recently, a
plasma heme oxygenase-1 suppressor (HOS) factor was
identified, characterized, and observed to be significantly
elevated in patients with early AD relative to subjects with
PD and NEC. MCI subjects exhibited plasma HOS levels
that were intermediate between NEC and AD values [77].
Unlike HO-1 mRNA, the HOS protein (␣1-antitrypsin) is
chemically stable and is undergoing further testing as a
putative peripheral AD biomarker. HO-1 and the HOS pro-
tein reflect AD pathophysiology in so far as HO-1 protein is
augmented and co-localizes to senile plaques and neurofi-
brillary tangles in AD-affected neural tissues [78].
6.5. Biospectroscopy
Biospectroscopy is an optical-based analytic technology
that provides a simultaneous and integrated measure of all
metabolic activities taking place in a biologic sample or
field and then uses the accruing spectral data to create a
profile of molecular biomarkers. Advantages of biospectros-
copy include simplicity of use, short analysis time, low-cost
instrumentation, simultaneous ascertainment of multiple
metabolites/biomarkers in a single specimen, and opportu-
nity for noninvasive, real-time in vivo monitoring both
remotely and at bedside. In preliminary studies, biospectro-
scopic analyses in the near-infrared performed on small
volumes of human plasma revealed metabolomic signatures
that distinguished early sporadic AD from PD and NEC
with high sensitivity and specificity (Schipper HM and
Burns D, manuscript in preparation).
6.6. Other putative biomarkers
An extensive array of endogenous substances detected in
CSF and/or blood have been touted as possible AD biomar-
kers. The latter include various acute phase reactants, cyto-
 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
kines, oxidative stress markers, advanced glycation end
products, ubiquitin-proteasomal constituents, trophic fac-
tors, complement components, glial and synaptic proteins,
cholesterol metabolites and lipoproteins, as well as transi-
tion metals [2,39,66,67,79 – 85]. In many of the published
reports, the chemical species in question were evaluated for
their possible role in the pathophysiology of AD, with
deliberations concerning their biomarker potential arising
post facto [46,86]. Mere determination that a given com-
pound is abnormally elevated or suppressed in AD body
fluids is insufficient to implicate the substance as a viable
biologic marker of the disease in the absence of other
criteria listed in Section 4.
7. Conclusions
(1) AD is a public health problem of epidemic propor-
tions for which current diagnostic and therapeutic
modalities remain inadequate.
(2) The advent of a biologic marker that differentiates
early, sporadic AD from normal aging and other
dementing illnesses would represent a significant
achievement in the evaluation and management of
this debilitating neurodegenerative disorder. An ac-
curate, minimally invasive biologic marker of early
sporadic AD would serve the public interest by fa-
cilitating patient and family counseling, optimizing
stratification of subgroups for enrollment in clinical
drug trials, and interpretation of treatment outcome
measures. The development of a biomarker that dif-
ferentiates “malignant” MCI cases at high risk for
conversion to AD from neuropsychologically iden-
tical cases destined to manifest “benign” aging-
related memory loss would be a particularly signif-
icant accomplishment. On theoretical grounds,
biomarkers might also prove useful in situations in
which concurrent medical conditions preclude or
confound cognitive and neuropsychological testing.
Examples of the latter include patients with major
depression, delirium, suppressed consciousness, or
who are otherwise uncooperative for detailed cogni-
tive testing.
(3) Although several candidate biomarkers of sporadic
AD have been identified and commercialized, none
currently fulfills criteria for an ideal test.
(4) Decreased A␤1-42 and increased phospho-tau protein
levels in the CSF are, at present, the most accurate
and reproducible chemical neurodiagnostics of early
sporadic AD. These biomarkers also show promise
as prognosticators in subjects with MCI. However,
CSF evaluation by lumbar puncture remains imprac-
tical for mass screening of elderly individuals with
memory impairment or AD risk factors.
329
(5) Platelet APP isoform ratios, plasma and urinary F2-
isoprostane levels, systemic alterations in HOS/
HO-1 profiles, and blood biospectrosopy might ful-
fill several key criteria for an ideal biologic marker
of early sporadic AD. However, further experimen-
tation will be required before these candidate bi-
omarkers can be considered for routine clinical use.
Similarly, all AD biomarker candidates emerging
from the application of mass spectrometry and other
proteomic techniques [87,88] will require rigorous
clinical evaluation for their suitability as bonafide
diagnostic tools.
(6) The pathology of AD is highly complex, and it is
therefore likely that combinations of individual bi-
omarkers will provide more accurate diagnostic and
prognostic information than any single marker mea-
sured in isolation (akin to use of multiple biochem-
ical indices to characterize connective tissue disease,
liver failure, or cardiac ischemia).
(7) In October 2004 the National Institute on Aging
announced the Alzheimer Disease Neuroimaging
Initiative (ADNI), a large, multi-institutional pro-
spective study proposing to delineate imaging and
chemical biomarkers in 800 rigorously ascertained
subjects with normal cognition, MCI, and AD. This
initiative represents the most concerted effort yet to
develop accurate diagnostic, prognostic, and surro-
gate markers of sporadic AD [89] and might identify
presymptomatic (pre-MCI) individuals at risk for
this condition [90].
8. Recommendations reaching consensus at the Third
Canadian Consensus Conference on Diagnosis and
Treatment of Dementia
8.1. To primary care physicians
(1) Biologic markers for the diagnosis of AD should not,
at this juncture, be included in the battery of tests
routinely used by primary care physicians to evalu-
ate subjects with memory loss (Grade C, Level 3).
Consideration for such specialized testing in an in-
dividual case should prompt referral of the patient to
a specialist engaged in dementia evaluations or a
memory clinic.
8.2. To specialists
(2) Although highly desirable, there currently exist no
blood- or urine-based AD diagnostics that can be
unequivocally endorsed for the routine evaluation of
memory loss in the elderly (Grade C, Level 3). The
noninvasiveness of such tests, if and when they be-
come available, would be suitable for mass screening
330 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
of subjects with memory loss presenting to special-
ists in their offices and memory clinics.
(3) Because of their relative invasiveness and availabil-
ity of other fairly accurate diagnostic modalities
(clinical, neuropsychological, and neuroimaging),
CSF biomarkers should not be routinely performed
in all subjects undergoing evaluation for memory
loss (Grade D, Level 2).
(4) CSF biomarkers might be considered in cases in
which there are atypical features and diagnostic con-
fusion. CSF biomarkers might be useful in differen-
tiating frontal variants of AD from FTD (Grade B,
Level 2).
(5) When a decision to obtain CSF biomarkers is made,
combined A␤1-42 and p-tau concentrations should be
measured by validated enzyme-linked immunosor-
bent assay (Grade A, Level 1). It might be best to
convey the CSF samples to a centralized facility
(commercial or academic) with a track record in
generating high quality, reproducible data.
(6) CSF biomarker data in isolation are insufficient to
diagnose or exclude AD (Grade C, Level 3). They
should be interpreted in light of clinical, neuropsy-
chological, other laboratory, and neuroimaging data
available for the individual under investigation.
Author Disclosures
Hyman M. Schipper has received support from Capiron
Pharmaceuticals Inc. (Consultant), Osta Biotechnologies
Inc. (Consultant), Molecular Biometrics, and LLC (Found-
ing Scientist, Equity Holder).
Acknowledgments
Hyman M. Schipper’s laboratory is supported by the
Canadian Institutes of Health Research.
References
[1] Klunk WE. Biological markers of Alzheimer’s disease. Neurobiol
Aging 1998;19:145–7.
[2] Frank RA, Galasko D, Hampel H, Hardy J, de Leon MJ, Mehta PD,
et al. Biological markers for therapeutic trials in Alzheimer’s disease:
proceedings of the biological markers working group—NIA initiative
on neuroimaging in Alzheimer’s disease. Neurobiol Aging 2003;24:
521–36.
[3] Iwatsubo T. Amyloid beta protein in plasma as a diagnostic marker
for Alzheimer’s disease. Neurobiol Aging 1998;19:161–3.
[4] Blennow K, Hampel H. CSF markers for incipient Alzheimer’s dis-
ease. Lancet Neurol 2003;2:605–13.
[5] Shoji M. Cerebrospinal fluid Abeta40 and Abeta42: natural course
and clinical usefulness. Front Biosci 2002;7:997–1006.
[6] Olsson A, Hoglund K, Sjogren M, Andreasen N, Minthon L, Lannfelt
L, et al. Measurement of alpha- and beta-secretase cleaved amyloid
precursor protein in cerebrospinal fluid from Alzheimer patients. Exp
Neurol 2003;183:74 – 80.
[7] Mehta PD, Pirttila T, Mehta SP, Sersen EA, Aisen PS, Wisniewski
HM. Plasma and cerebrospinal fluid levels of amyloid beta proteins
1-40 and 1-42 in Alzheimer disease. Arch Neurol 2000;57:100 –5.
[8] Tamaoka A, Fukushima T, Sawamura N, Ishikawa K, Oguni E,
Komatsuzaki Y, et al. Amyloid beta protein in plasma from patients
with sporadic Alzheimer’s disease. J Neurol Sci 1996;141:65– 8.
[9] Sunderland T, Linker G, Mirza N, Putnam KT, Friedman DL, Kim-
mel LH, et al. Decreased beta-amyloid1-42 and increased tau levels in
cerebrospinal fluid of patients with Alzheimer disease. JAMA 2003;
289:2094 –103.
[10] Hulstaert F, Blennow K, Ivanoiu A, Schoonderwaldt HC, Riemen-
schneider M, De Deyn PP, et al. Improved discrimination of AD
patients using beta-amyloid(1-42) and tau levels in CSF. Neurology
1999;52:1555– 62.
[11] Mollenhauer B, Cepek L, Bibl M, Wiltfang J, Schulz-Schaeffer WJ,
Ciesielczyk B, et al. Tau protein, Abeta42 and S-100B protein in
cerebrospinal fluid of patients with dementia with Lewy bodies.
Dement Geriatr Cogn Disord 2005;19:164 –70.
[12] Clark CM, Xie S, Chittams J, Ewbank D, Peskind E, Galasko D,
et al. Cerebrospinal fluid tau and beta-amyloid: how well do these
biomarkers reflect autopsy-confirmed dementia diagnoses? Arch
Neurol 2003;60:1696 –702.
[13] Gomez-Tortosa E, Gonzalo I, Fanjul S, Sainz MJ, Cantarero S,
Cemillan C, et al. Cerebrospinal fluid markers in dementia with lewy
bodies compared with Alzheimer disease. Arch Neurol 2003;60:
1218 –22.
[14] Otto M, Esselmann H, Schulz-Shaeffer W, Neumann M, Schroter A,
Ratzka P, et al. Decreased beta-amyloid1-42 in cerebrospinal fluid of
patients with Creutzfeldt-Jakob disease. Neurology 2000;54:1099 –
102.
[15] Holmberg B, Johnels B, Blennow K, Rosengren L. Cerebrospinal
fluid Abeta42 is reduced in multiple system atrophy but normal in
Parkinson’s disease and progressive supranuclear palsy. Mov Disord
2003;18:186 –90.
[16] Sjogren M, Andreasen N, Blennow K. Advances in the detection of
Alzheimer’s disease-use of cerebrospinal fluid biomarkers. Clin Chim
Acta 2003;332:1–10.
[17] Lewczuk P, Esselmann H, Groemer TW, Bibl M, Maler JM, Steinacker
P, et al. Amyloid beta peptides in cerebrospinal fluid as profiled with
surface enhanced laser desorption/ionization time-of-flight mass spec-
trometry: evidence of novel biomarkers in Alzheimer’s disease. Biol
Psychiatry 2004;55:524 –30.
[18] Maruyama M, Matsui T, Tanji H, Nemoto M, Tomita N, Ootsuki M,
et al. Cerebrospinal fluid tau protein and periventricular white matter
lesions in patients with mild cognitive impairment: implications for 2
major pathways. Arch Neurol 2004;61:716 –20.
[19] Andreasen N, Minthon L, Clarberg A, Davidsson P, Gottfries J,
Vanmechelen E, et al. Sensitivity, specificity, and stability of CSF-tau
in AD in a community-based patient sample. Neurology 1999;53:
1488 –94.
[20] Green AJ, Harvey RJ, Thompson EJ, Rossor MN. Increased tau in the
cerebrospinal fluid of patients with frontotemporal dementia and
Alzheimer’s disease. Neurosci Lett 1999;259:133–5.
[21] Hesse C, Rosengren L, Vanmechelen E, Vanderstichele H, Jensen C,
Davidsson P, et al. Cerebrospinal fluid markers for Alzheimer’s
disease evaluated after acute ischemic stroke. J Alzheimers Dis 2000;
2:199 –206.
[22] Grossman M, Farmer J, Leight S, Work M, Moore P, Van Deerlin V, et
al. Cerebrospinal fluid profile in frontotemporal dementia and Alzhei-
mer’s disease. Ann Neurol 2005;57:721–9.
[23] Leszek J, Malyszczak K, Janicka B, Kiejna A, Wiak A. Total tau in
cerebrospinal fluid differentiates Alzheimer’s disease from vascular
dementia. Med Sci Monit 2003;9:CR484 – 8.
[24] Hampel H, Buerger K, Zinkowski R, Teipel SJ, Goernitz A, Andreasen
N, et al. Measurement of phosphorylated tau epitopes in the differential
 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
diagnosis of Alzheimer disease: a comparative cerebrospinal fluid study.
Arch Gen Psychiatry 2004;61:95–102.
[25] Lewczuk P, Esselmann H, Bibl M, Beck G, Maler JM, Otto M, et al.
Tau protein phosphorylated at threonine 181 in CSF as a neurochem-
ical biomarker in Alzheimer’s disease: original data and review of the
literature. J Mol Neurosci 2004;23:115–22.
[26] Vanmechelen E, Vanderstichele H, Davidsson P, Van Kerschaver E,
Van Der Perre B, Sjogren M, et al. Quantification of tau phosphor-
ylated at threonine 181 in human cerebrospinal fluid: a sandwich
ELISA with a synthetic phosphopeptide for standardization. Neurosci
Lett 2000;285:49 –52.
[27] Schoonenboom NS, Pijnenburg YA, Mulder C, Rosso SM, Van Elk
EJ, Van Kamp GJ, et al. Amyloid beta(1-42) and phosphorylated tau
in CSF as markers for early-onset Alzheimer disease. Neurology
2004;62:1580 – 4.
[28] Kurz A, Riemenschneider M, Wallin A. Potential biological markers for
cerebrovascular disease. Int Psychogeriatr 2003;15(Suppl 1):89 –97.
[29] Parnetti L, Lanari A, Amici S, Gallai V, Vanmechelen E, Hulstaert F.
CSF phosphorylated tau is a possible marker for discriminating Alz-
heimer’s disease from dementia with Lewy bodies: Phospho-Tau
International Study Group. Neurol Sci 2001;22:77– 8.
[30] Blennow K, Wallin A, Agren H, Spenger C, Siegfried J, Van-
mechelen E. Tau protein in cerebrospinal fluid: a biochemical marker
for axonal degeneration in Alzheimer disease? Mol Chem Neuro-
pathol 1995;26:231– 45.
[31] Hesse C, Rosengren L, Andreasen N, Davidsson P, Vanderstichele H,
Vanmechelen E, et al. Transient increase in total tau but not phospho-
tau in human cerebrospinal fluid after acute stroke. Neurosci Lett
2001;297:187–90.
[32] Schonknecht P, Hempel A, Hunt A, Seidl U, Volkmann M, Pantel J,
et al. Cerebrospinal fluid tau protein levels in schizophrenia. Eur Arch
Psychiatry Clin Neurosci 2003;253:100 –2.
[33] Goodall CA, Head MW, Everington D, Ironside JW, Knight RS,
Green AJ. Raised CSF phospho-tau concentrations in variant
Creutzfeldt-Jakob disease: diagnostic and pathological implications.
J Neurol Neurosurg Psychiatry 2006;77:89 –91.
[34] Riemenschneider M, Wagenpfeil S, Vanderstichele H, Otto M,
Wiltfang J, Kretzschmar H, et al. Phospho-tau/total tau ratio in
cerebrospinal fluid discriminates Creutzfeldt-Jakob disease from
other dementias. Mol Psychiatry 2003;8:343–7.
[35] Schonknecht P, Pantel J, Hunt A, Volkmann M, Buerger K, Hampel
H, et al. Levels of total tau and tau protein phosphorylated at threo-
nine 181 in patients with incipient and manifest Alzheimer’s disease.
Neurosci Lett 2003;339:172– 4.
[36] Buerger K, Teipel SJ, Zinkowski R, Blennow K, Arai H, Engel R, et
al. CSF tau protein phosphorylated at threonine 231 correlates with
cognitive decline in MCI subjects. Neurology 2002;59:627–9.
[37] Arai H, Ishiguro K, Ohno H, Moriyama M, Itoh N, Okamura N, et al.
CSF phosphorylated tau protein and mild cognitive impairment: a
prospective study. Exp Neurol 2000;166:201–3.
[38] de Leon MJ, DeSanti S, Zinkowski R, Mehta PD, Pratico D, Segal S,
et al. MRI and CSF studies in the early diagnosis of Alzheimer’s
disease. J Intern Med 2004;256:205–23.
[39] Blennow K. Cerebrospinal fluid protein biomarkers for Alzheimer’s
disease. NeuroRx 2004;1:213–25.
[40] Wahlund LO, Blennow K. Cerebrospinal fluid biomarkers for disease
stage and intensity in cognitively impaired patients. Neurosci Lett
2003;339:99 –102.
[41] Ganzer S, Arlt S, Schoder V, Buhmann C, Mandelkow EM, Finckh
U, et al. CSF-tau, CSF-Abeta1-42, ApoE-genotype and clinical pa-
rameters in the diagnosis of Alzheimer’s disease: combination of
CSF-tau and MMSE yields highest sensitivity and specificity. J Neu-
ral Transm 2003;110:1149 – 60.
[42] Riemenschneider M, Lautenschlager N, Wagenpfeil S, Diehl J,
Drzezga A, Kurz A. Cerebrospinal fluid tau and beta-amyloid 42
331
proteins identify Alzheimer disease in subjects with mild cognitive
impairment. Arch Neurol 2002;59:1729 –34.
[43] Hampel H, Teipel SJ, Fuchsberger T, Andreasen N, Wiltfang J, Otto
M, et al. Value of CSF beta-amyloid1-42 and tau as predictors of
Alzheimer’s disease in patients with mild cognitive impairment. Mol
Psychiatry 2004;9:705–10.
[44] Herukka SK, Hallikainen M, Soininen H, Pirttila T. CSF Abeta42 and
tau or phosphorylated tau and prediction of progressive mild cogni-
tive impairment. Neurology 2005;64:1294 –7.
[45] Ivanoiu A, Sindic CJ. Cerebrospinal fluid TAU protein and amyloid
beta42 in mild cognitive impairment: prediction of progression to
Alzheimer’s disease and correlation with the neuropsychological ex-
amination. Neurocase 2005;11:32–9.
[46] Klunk WE. Blood and CSF biomarkers for AD revisited: what’s new,
what’s good, and is this where we should be looking? Neurobiol
Aging 2002;23:517–9; discussion 521–2.
[47] Peskind ER, Riekse R, Quinn JF, Kaye J, Clark CM, Farlow MR, et
al. Safety and acceptability of the research lumbar puncture. Alzhei-
mer Dis Assoc Disord 2005;19:220 –5.
[48] De La Monte SM, Wands JR. The AD7c-NTP neuronal thread protein
biomarker for detecting Alzheimer’s disease. J Alzheimers Dis 2001;
3:345–53.
[49] Chertkow H, Bergman H, Schipper HM, Gauthier S, Bouchard R,
Fontaine S, et al. Assessment of suspected dementia. Can J Neurol Sci
2001;28(Suppl 1):S28 – 41.
[50] Kennard ML, Feldman H, Yamada T, Jefferies WA. Serum levels of
the iron binding protein p97 are elevated in Alzheimer’s disease. Nat
Med 1996;2:1230 –5.
[51] Kim DK, Seo MY, Lim SW, Kim S, Kim JW, Carroll BJ, et al. Serum
melanotransferrin, p97 as a biochemical marker of Alzheimer’s dis-
ease. Neuropsychopharmacology 2001;25:84 –90.
[52] Desrosiers RR, Bertrand Y, Nguyen QT, Demeule M, Gabathuler R,
Kennard ML, et al. Expression of melanotransferrin isoforms in
human serum: relevance to Alzheimer’s disease. Biochem J 2003;
374(Pt 2):463–71.
[53] Di Luca M, Pastorino L, Bianchetti A, Perez J, Vignolo LA, Lenzi
GL, et al. Differential level of platelet amyloid beta precursor protein
isoforms: an early marker for Alzheimer disease. Arch Neurol 1998;
55:1195–200.
[54] Di Luca M, Pastorino L, Cattabeni F, Zanardi R, Scarone S, Racagni
G, et al. Abnormal pattern of platelet APP isoforms in Alzheimer
disease and Down syndrome. Arch Neurol 1996;53:1162– 6.
[55] Rosenberg RN, Baskin F, Fosmire JA, Risser R, Adams P, Svetlik D,
et al. Altered amyloid protein processing in platelets of patients with
Alzheimer disease. Arch Neurol 1997;54:139 – 44.
[56] Padovani A, Pastorino L, Borroni B, Colciaghi F, Rozzini L, Mon-
astero R, et al. Amyloid precursor protein in platelets: a peripheral
marker for the diagnosis of sporadic AD. Neurology 2001;
57:2243– 8.
[57] Baskin F, Rosenberg RN, Iyer L, Hynan L, Cullum CM. Platelet APP
isoform ratios correlate with declining cognition in AD. Neurology
2000;54:1907–9.
[58] Borroni B, Colciaghi F, Archetti S, Marcello E, Caimi L, Di Luca M, et
al. Predicting cognitive decline in Alzheimer disease. Role of platelet
amyloid precursor protein. Alzheimer Dis Assoc Disord 2004;18:
32– 4.
[59] Borroni B, Colciaghi F, Caltagirone C, Rozzini L, Broglio L, Cat-
tabeni F, et al. Platelet amyloid precursor protein abnormalities in
mild cognitive impairment predict conversion to dementia of Alzhei-
mer type: a 2-year follow-up study. Arch Neurol 2003;60:1740 – 4.
[60] Padovani A, Borroni B, Colciaghi F, Pettenati C, Cottini E, Agosti C,
et al. Abnormalities in the pattern of platelet amyloid precursor
protein forms in patients with mild cognitive impairment and Alz-
heimer disease. Arch Neurol 2002;59:71–5.
332 H.M. Schipper / Alzheimer’s & Dementia 3 (2007) 325–332
[61] Borroni B, Colciaghi F, Pastorino L, Pettenati C, Cottini E, Rozzini
L, et al. Amyloid precursor protein in platelets of patients with
Alzheimer disease: effect of acetylcholinesterase inhibitor treatment.
Arch Neurol 2001;58:442– 6.
[62] Liu HC, Chi CW, Ko SY, Wang HC, Hong CJ, Lin KN, et al.
Cholinesterase inhibitor affects the amyloid precursor protein iso-
forms in patients with Alzheimer’s disease. Dement Geriatr Cogn
Disord 2005;19:345– 8.
[63] Zimmermann M, Borroni B, Cattabeni F, Padovani A, Di Luca M.
Cholinesterase inhibitors influence APP metabolism in Alzheimer
disease patients. Neurobiol Dis 2005;19:237– 42.
[64] Schipper HM. Brain iron deposition and the free radical-mitochondrial
theory of ageing. Ageing Res Rev 2004;3:265–301.
[65] Markesbery WR, Carney JM. Oxidative alterations in Alzheimer’s
disease. Brain Pathol 1999;9:133– 46.
[66] Galasko D. Biological markers and the treatment of Alzheimer’s
disease. J Mol Neurosci 2001;17:119 –25.
[67] Maccioni RB, Lavados M, Maccioni CB, Mendoza-Naranjo A. Bio-
logical markers of Alzheimer’s disease and mild cognitive impair-
ment. Curr Alzheimer Res 2004;1:307–14.
[68] Tuppo EE, Forman LJ, Spur BW, Chan-Ting RE, Chopra A, Cavalieri
TA. Sign of lipid peroxidation as measured in the urine of patients
with probable Alzheimer’s disease. Brain Res Bull 2001;54:565– 8.
[69] Pratico D, V MYL, Trojanowski JQ, Rokach J, Fitzgerald GA. In-
creased F2-isoprostanes in Alzheimer’s disease: evidence for en-
hanced lipid peroxidation in vivo. Faseb J 1998;12:1777– 83.
[70] Pratico D, Clark CM, Liun F, Rokach J, Lee VY, Trojanowski JQ.
Increase of brain oxidative stress in mild cognitive impairment: a
possible predictor of Alzheimer disease. Arch Neurol 2002;59:972– 6.
[71] Montine TJ, Kaye JA, Montine KS, McFarland L, Morrow JD, Quinn
JF. Cerebrospinal fluid abeta42, tau, and f2-isoprostane concentra-
tions in patients with Alzheimer disease, other dementias, and in
age-matched controls. Arch Pathol Lab Med 2001;125:510 –2.
[72] Montine TJ, Quinn JF, Milatovic D, Silbert LC, Dang T, Sanchez S,
et al. Peripheral F2-isoprostanes and F4-neuroprostanes are not in-
creased in Alzheimer’s disease. Ann Neurol 2002;52:175–9.
[73] Greco A, Minghetti L, Levi G. Isoprostanes, novel markers of oxi-
dative injury, help understanding the pathogenesis of neurodegenera-
tive diseases. Neurochem Res 2000;25:1357– 64.
[74] Schipper HM, Chertkow H, Mehindate K, Frankel D, Melmed C,
Bergman H. Evaluation of heme oxygenase-1 as a systemic biological
marker of sporadic AD. Neurology 2000;54:1297–304.
[75] Ishizuka K, Kimura T, Yoshitake J, Akaike T, Shono M, Takamatsu
J, et al. Possible assessment for antioxidant capacity in Alzheimer’s
disease by measuring lymphocyte heme oxygenase-1 expression with
real-time RT-PCR. Ann N Y Acad Sci 2002;977:173– 8.
[76] Schipper HM, Mawal Y, Chertkow H, Bergman H. Blood HO-1
mRNA in AD and MCI. Neurology 2001;57:2142–3.
[77] Maes OC, Kravitz S, Mawal Y, Su H, Liberman A, Mehindate K, et
al. Characterization of alpha(1)-antitrypsin as a heme oxygenase-1
suppressor in Alzheimer plasma. Neurobiol Dis 2006;24:89 –100.
[78] Schipper HM, Cissé S, Stopa EG. Expression of heme oxygenase-1 in
the senescent and Alzheimer-diseased brain. Ann Neurol 1995;37:
758 – 68.
[79] Small DH. Alzheimer’s disease biomarkers: their value in diagnosis
and clinical trials. Front Biosci 2002;7:986 – 8.
[80] Irizarry MC. Biomarkers of Alzheimer disease in plasma. NeuroRx
2004;1:226 –34.
[81] Davidsson P, Sjogren M. The use of proteomics in biomarker dis-
covery in neurodegenerative diseases. Dis Markers 2005;21:81–92.
[82] Ho L, Sharma N, Blackman L, Festa E, Reddy G, Pasinetti GM. From
proteomics to biomarker discovery in Alzheimer’s disease. Brain Res
Brain Res Rev 2005;48:360 –9.
[83] Wada H. Analyses of serum concentrations of apolipoproteins in the
demented elderly. Intern Med 2000;39:220 –2.
[84] Meli M, Perier C, Ferron C, Parssegny F, Denis C, Gonthier R, et al.
Serum pentosidine as an indicator of Alzheimer’s disease. J Alzhei-
mers Dis 2002;4:93– 6.
[85] Mulder C, Scheltens P, Visser JJ, van Kamp GJ, Schutgens RB.
Genetic and biochemical markers for Alzheimer’s disease: recent
developments. Ann Clin Biochem 2000;37(Pt 5):593– 607.
[86] Padovani A, Borroni B, Di Luca M. Advances on biological markers
in early diagnosis of Alzheimer disease. Adv Clin Chem 2005;39:
107–29.
[87] Choe LH, Dutt MJ, Relkin N, Lee KH. Studies of potential cerebro-
spinal fluid molecular markers for Alzheimer’s disease. Electrophore-
sis 2002;23:2247–51.
[88] Yu H-L, Chertkow HM, Bergman H, Schipper HM. Aberrant profiles
of native and oxidized glycoproteins in Alzheimer plasma. Proteom-
ics 2003;3:2240 – 8.
[89] Trojanowski JQ, Lee M-Y. Commentary on “Diagnosis of Alzhei-
mer’s disease: Two decades of progress”: Alzheimer’s disease
research—new opportunities for early diagnosis and drug discovery.
Alz Dement 2005;1:119 –20.
[90] Morris JC, Kimberly A, Quaid K, Holtzman DM, Kantarci K, Kaye
J, et al. Role of biomarkers in studies of presymptomatic Alzheimer’s
disease. Alz Dement 2005;1:145–51.
 Alzheimer’s & Dementia 3 (2007) 333–335

Structural neuroimaging in the diagnosis of dementia

Tiffany Chow*
The Rotman Research Institute Baycrest Centre for Geriatric Care; Department of Medicine, Neurology Division, University of Toronto, Toronto,
Ontario, Canada

Abstract This article reviews recent use of structural imaging in the diagnosis of dementia and presents
evidence-based recommendations approved at the meeting of the Third Consensus Conference on
the Diagnosis and Treatment of Dementia (CCCDTD3) held in Montreal in March 2006. Although
the consensus group concurs that it is possible to specify circumstances in which structural imaging
has a role to rule out pathology, it also has a role in ruling in concomitant cerebrovascular disease
for patient management.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Alzheimer’s disease; Vascular dementia; MRI; CT; Structural imaging

The decision to exclude structural neuroimaging from a
dementia evaluation must weigh several opposing factors:
the physician’s confidence that the history and neurologic
examination have sufficiently ruled out the likelihood of
finding a neurosurgical lesion that would reverse the pa-
tient’s cognitive complaints, the availability of the imaging
procedure, the need to identify comorbidity, and the practice
setting (clinical vs research). Primary care providers and
specialists will perceive this set of considerations differ-
ently. Indicators to increase the yield from structural imag-
ing while decreasing the number of studies ordered and
minimizing the number of treatable dementia cases missed
might be useful to all physicians (Appendix 1) [1– 4]. Such
indicators would also decrease the costs of dementia work-
ups.
Search criteria for the indications listed below included
original articles or clinical trials examining the utility of
structural magnetic resonance imaging (MRI) or computed
tomography (CT). Utility could be reported in terms of
benefit to patient outcome or cost/benefit analyses. All but
two [5,6] of the 27 references used for this review pertain to
patients diagnosed through specialty clinics. Most studies
are from research cohorts; four were from clinical cohorts
[2,7–9]. Although most of the evidence discovered for this
*Corresponding author. Tel.: (416) 785-2500 x3459; Fax: (416) 785-
2862.
E-mail address: tchow@rotman-baycrest.on.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.016
review was derived from neither primary care nor research
environments exclusively, some indications for imaging
have implications for any clinical setting, whereas other
indications are currently mandated for research purposes
only.
The following indications could not be separated into
applications for primary care versus research setting or CT
of head versus MRI. Justifications for ordering a diagnostic
CT of head or MRI of brain in the past focused on the
importance of ruling out etiologies of dementia that could
be reversed with a neurosurgical procedure (normal pres-
sure hydrocephalus, subdural hematoma, and neoplasm).
One cost analysis calculated a work-up individualized to
course of illness that offers imaging only to those subjects
younger than age 65 years [6]. Other studies have since
shown with Class II and III evidence that only a small
number of cases with neurosurgical lesions have dementia
as the sole presenting symptom, and that even fewer of
those cases enjoy significant resolution of the cognitive
impairment after treatment [2,5–10]. In this sense, there is
fair evidence to recommend against structural neuroimaging
to rule out a neurosurgical lesion during a work-up for
dementia (Grade D, Level 2). This recommendation did not
reach consensus at the CCCDTD3 meeting, however.
Although four studies report Class II evidence that atro-
phy as measured on CT [11] or MRI [12–14] distinguishes
Alzheimer’s disease (AD) from nondemented healthy el-
derly, neuroimaging has seldom changed the initial clinical
334 T. Chow / Alzheimer’s & Dementia 3 (2007) 333–335
diagnosis of AD. Instead, there is fair evidence to support
use of structural neuroimaging to rule in concomitant cere-
brovascular disease that can affect patient management
(Grade B, Level 2). Two studies provide Class II evidence
that CT and MRI scans identify cerebrovascular disease that
was not suspected clinically, resulting in the addition of
stroke risk management to the care plan [1,15–17].
Within a research setting, there is fair evidence to sup-
port the use of structural neuroimaging to predict conver-
sion of mild cognitive impairment (MCI) to AD and pro-
gression through stages of AD (Grade B, Level 2). Three
labs have replicated Class II evidence to support volumetric
analyses of entorhinal cortex and hippocampus to identify
those patients with MCI who are most likely to progress to
AD within several years [18 –20]. There is also fair evi-
dence to support the use of structural neuroimaging to track
the progression of AD in clinical trials, especially if the
morphometry is combined with neuropsychological testing
(Grade B, Level 2) [12,21–23].
1. Recomendations supported at the Third
CCCDTD meeting
1. There is fair evidence to support the selective use of
CT or MRI scanning in the work-up for dementia per
1999 guidelines (Grade B, Level 2).
2. There is fair evidence to support use of structural
neuroimaging to rule in concomitant cerebrovascular
disease that can affect patient management (Grade B,
Level 2).
3. There is fair evidence to support the use of structural
neuroimaging to track the progression of AD in clin-
ical trials, especially if the morphometry is combined
with neuropsychological testing (Grade B, Level 2).
Author Disclosures
Tiffany Chow has received support from with Novartis
(Speaker’s Bureau), Janssen Ortho Inc. (Consultant), and
CanCog Technologies (Consultant).
Acknowledgments
Tiffany Chow is supported by NIA grant F32
AG022802; the University of Toronto Dean’s Fund for New
Faculty (#457494), and an endowment to the Sam and Ida
Ross Memory Clinic.
References
[1] Chui H, Zhang Q. Evaluation of dementia: a systematic study of the
usefulness of the American Academy of Neurology’s practice param-
eters. Neurology 1997;49:925–35.
[2] Freter SBH, Gold S, Chertkow H, Clarfield AM. Prevalence of
potentially reversible dementias and actual reversibility in a memory
clinic cohort. CMAJ 1998;159:657– 62.
[3] Gifford DR, Holloway RG, Vickrey BG. Systematic review of clin-
ical prediction rules for neuroimaging in the evaluation of dementia.
Arch Intern Med 2000;160:2855– 62.
[4] Martin DC, Miller J, Kapoor W, Karpf M, Boller F. Clinical predic-
tion rules for computed tomographic scanning in senile dementia.
Arch Intern Med 1987;147:77– 80.
[5] Alexander EM, Buchner DM, Cain KC, Larson EB. Do surgical brain
lesions present as isolated dementia? a population-based study. J Am
Geriatr Soc 1995;43:138 – 43.
[6] Foster GR, Scott DA, Payne S. The use of CT scanning in demen-
tia: a systematic review. Int J Technol Assess Health Care 1999;
15:406 –23.
[7] Walstra GJM, Teunisse S, van Gool HvC. Reversible dementia in
elderly patients referred to a memory clinic. J Neurol 1997;244:
17–22.
[8] Clarfield AM. The decreasing prevalence of reversible dementias: an
updated meta-analysis. Arch Intern Med 2003;163:2219 –29.
[9] Chong MS, Sahadevan S. An evidence-based clinical approach to the
diagnosis of dementia. Ann Acad Med Singapore 2003;32:740 – 48.
[10] Simon D. Routine imaging in suspected dementia: diagnostic and
therapeutic yield. Clin Res 1992;40:565A.
[11] Jobst KA, Barnetson LP, Shepstone BJ. Accurate prediction of his-
tologically confirmed Alzheimer’s disease and the differential diag-
nosis of dementia: the use of NINCDS-ADRDA and DSM-III-R
criteria, SPECT, X-ray CT, and APO E4 medial temporal lobe
dementias—the Oxford Project to Investigate Memory and Aging. Int
Psychogeriatr 1997;(Suppl 1):191–222; discussion 247–52.
[12] Zakzanis KK. Quantitative evidence for neuroanatomic and neuro-
psychological markers in dementia of the Alzheimer’s type. J Clin
Exp Neuropsychol 1998;20:259 – 69.
[13] Frisoni GB, Weiss C, Geroldi C, Bianchetti A, Trabucchi M. Linear
measures of atrophy in mild Alzheimer disease. AJNR Am J Neuro-
radiol 1996;17:913–23.
[14] Laakso MP, Soininen H, Partanen K, Lehtovirta M, Hallikanen M,
Hanninen T, HMHTHEVPRP. MRI of the hippocampus in Alzheimer’s
disease: sensitivity, specificity, and analysis of the incorrectly classified
subjects. Neurobiol Aging 1998;19:23–31.
[15] Massoud F, Devi G, Moroney JT, Stern Y, Lawton A, Bell K, et al.
The role of routine laboratory studies and neuroimaging in the diag-
nosis of dementia: a clinicopathological study. J Am Geriatr Soc
2000;48:1204 –10.
[16] Wilcock G, Mobius HJ, Stoffler A, MMM 500 group. A double-blind,
placebo-controlled multicentre study of memantine in mild to mod-
erate vascular dementia (MMM500). Int Clin Psychopharmacol 2002;
17:297–305.
[17] Lopez OL, Kuller LH, Becker JT, Jagust WJ, DeKosky ST, Fitz-
patrick A, et al. Classification of vascular dementia in the Cardiovas-
cular Health Study Cognition Study. Neurology 2005;64:1539 – 47.
[18] Killiany RJ, Gomez-Isla T, Moss MB, Kikinis R, Jolesz F, Tanzi R,
et al. MRI measures of entorhinal cortex vs hippocampus in preclin-
ical AD. Neurology 2002;58:1188 –96.
[19] Rusinek H, Frid D, Tsui WH, Tarshish CY, Convit A, de Leon MJ.
Regional brain atrophy rate predicts future cognitive decline: 6-year
longitudinal MR imaging study of normal aging. Radiology 2003;
229:691– 6.
[20] Jack CR Jr, Shiung MM, Weigand SD, O’Brien PC, Gunter JL, Boeve
BF, et al. Brain atrophy rates predict subsequent clinical conversion
in normal elderly and amnestic MCI. Neurology 2005;65:1227–31.
[21] Frisoni GB, Scheltens P, Galluzzi S, Nobili FM, Fox NC, Robert PH,
et al. Neuroimaging tools to rate regional atrophy, subcortical cere-
brovascular disease, and regional cerebral blood flow and metabo-
 T. Chow / Alzheimer’s & Dementia 3 (2007) 333–335
lism: consensus paper of the EADC. J Neurol Neurosurg Psychiatry
2003;74:1371– 81.
[22] Cardenas VA, Hardin D, Ezekiel F, Weber P, Jagust WJ, Chui HC,
et al. Comparison of methods for measuring longitudinal brain
change in cognitive impairment and dementia. Neurobiol Aging
2003;24:537– 44.
[23] Zamrini E, De Santi S, Tolar M. Imaging is superior to cognitive
testing for early diagnosis of Alzheimer’s disease. Neurobiol Aging
2004;25:685–91.
[24] Organizing Committee CCCotAoD. Assessing dementia: the Cana-
dian Consensus. CMAJ 1991;144:851–3.
Appendix 1. CCCAD indicators [24]
A cranial CT scan is recommended if one or more of the
following criteria are present:
Age younger than 60 years
335
Rapid (eg, during 1 to 2 months) unexplained decline
in cognition or function
“Short” duration of dementia (less than 2 years)
Recent/significant head trauma
Unexplained neurologic symptoms (eg, new onset of
severe headache or seizures)
History of cancer (especially in sites and types that
metastasize to the brain)
Use of anticoagulants or history of a bleeding disorder
History of urinary incontinence and gait disorder early
in the course of dementia (as might be found in
normal pressure hydrocephalus)
Any new localizing sign (eg, hemiparesis or a Babin-
ski’s reflex)
Unusual or atypical cognitive symptoms or presenta-
tion (eg, progressive aphasia)
Gait disturbance
 Alzheimer’s & Dementia 3 (2007) 336 –340
Functional neuroimaging in the diagnosis of dementia
Michael Borrie*
Division of Geriatric Medicine, St. Joseph’s Health Care London-Parkwood Site; Professor, Department of Medicine, University of Western Ontario,
London, Ontario, Canada
The functional neuroimaging modalities of fluoro-2-
deoxy-D-glucose positron emission tomography (FDG-PET),
single photon emission computed tomography (SPECT),
functional magnetic resonance imaging (fMRI), magnetic
resonance spectroscopy (MRS), and their roles in diagnos-
ing dementia and predicting the progression from mild cog-
nitive impairment (MCI) to dementia and tracking its pro-
gression have recently been reviewed [1– 4]. The question
asked by this evidence-based review was “Is there sufficient
evidence to recommend the use of a functional neuroimag-
ing technique by family physicians or specialists in Canada
to make or differentiate a diagnosis of dementia in people
presenting with cognitive impairment?”
1. Methods
The literature review used PubMed and EMBASE data-
bases including articles in English and human research from
January 1996 to February 2006.Search terms included de-
mentia, mild cognitive impairment, PET, SPECT, fMRI,
and MRS. A separate search used the same parameters but
restricted it to review articles to identify recent evidence-
based reviews.
2. FDG-PET
Reduction in brain glucose metabolism (MRglc) cor-
rected for atrophy, in the parietotemporal, frontal and
posterior cingulate cortices, is the typical pattern in Alz-
heimer’s disease (AD). High resolution PET scanning,
coregistered with MRI, has resulted in reports of reductions
in MRglc in more specific regions of interest (ROIs). The
hippocampus, known to be affected early in AD, has reduc-
tions in MRglc in patients with mild cognitive impairment
(MCI) compared with normal elderly [5–7].
*Corresponding author. Tel.: (519) 685-4021; Fax: 519-685-4093.
E-mail address: michael.borrie@sjhc.london.on.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.004
The decline from normal cognition to MCI has been
predicted by baseline entorhinal cortex MRglc with 84%
accuracy, 83% sensitivity, and 85% specificity [5]. With
ROIs, it is possible to find significant differences when
comparing normal controls with patients with MCI or AD,
but in this study there were no significant differences be-
tween MCI and AD [7]. However, in studying MRglc re-
ductions in ROIs in normal elderly and in subjects with MCI
and AD, De Santi et al [6] reported a diagnostic accuracy of
85% (80% MCI, 92% AD).
There are fewer studies that follow the decline from MCI
to AD, with the numbers of subjects in these studies ranging
from 17 to 52 [2]. Overall, the parietotemporal area meta-
bolic reductions predicted an overall accuracy of 86%, sen-
sitivity and specificity ranging from 75% to 100%.
A recent meta-analysis of PET articles, published be-
tween 1989 and 2003, found 15 articles that reported
sensitivity and specificity that met the study inclusion/
exclusion criteria. The summary sensitivity of PET was
86% confidence (95% confidence interval [CI], 76% to
93%), and the summary specificity was 86% (95% CI, 72%
to 93%) [8]. The limitations included use of an imperfect
reference standard, verification bias, and variability in the
choice of the disease spectrum. This limited generalizability
to a clinical setting. The authors recommended further re-
search on the use of PET in the diagnosis of AD, focusing
on these limitations in clinical settings.
A four-point visual rating scale to evaluate the presence
and severity of medial temporal lobe (MTL) hypometabo-
lism on FDG-PET scans has been evaluated in normal
controls and subjects with MCI and AD. The visual MTL
ratings were compared with MRglc data from ROIs in the
MRI-coregistered PET scans of all subjects as well as stan-
dard rating of neocortical evaluation. There was high intra-
rater and inter-rater reliabilities between the MTL and cor-
tical ratings (P ⫽ .001). MTL ratings were also highly
correlated with ROI MRglc. The diagnostic accuracy or
cortical ratings alone was improved by combining MTL
 M. Borrie / Alzheimer’s & Dementia 3 (2007) 336 –340
with cortical ratings, correctly identifying 100% of AD,
77% of MCI, and 85% of normal elderly. If these obser-
vations are confirmed in other studies, this scale has
promise as a clinically useful instrument in patients with
MCI undergoing PET scans [9].
The cost, availability, and, to a lesser extent, radiation
exposure are three present barriers for PET scanning to be
used clinically in Canada. The approval in 2004 in the U.S.
by Medicare and Medicaid for the use of PET, in cases that
remain uncertain after standard clinical evaluation for de-
mentia, will increase the clinical use and reporting of its
diagnostic utility.
Ligand molecules that cross the blood-brain barrier and
combine with amyloid are being developed and tested for
use with PET scanning to quantify in vivo amyloid depos-
ited in human brains [10 –12]. If these molecules are equally
successful with SPECT in defining amyloid load, the bar-
riers of cost and availability of PET might be removed.
A recent review of the role of neuroimaging in MCI [4]
concluded that although neuroimaging modalities had po-
tential to become valuable tools for the early diagnosis of
AD, larger prospective longitudinal studies are needed. The
Alzheimer’s Disease Neuroimaging Initiative (ADNI), be-
gun in the United States and Canada in 2005, is one study
that will establish standardization of neuroimaging with
MRI at 1.5 T and 3 T and PET and might realize this
potential.
3. SPECT
SPECT uses radiopharmacologic perfusion agents to
image blood flow patterns in the brain. 99mTechnetium-
hexamethyl-propylenamine oxime (99mTc-HMPAO) is the
most commonly used agent, which is lipophilic and quickly
crosses the blood-brain barrier into the neurons and glia and
is metabolized into a “trapped” lipophobic compound.
Clearance of 99mTc-HMPAO from the vascular pool al-
lows the imaging of brain tissue, which has a symmetrical
pattern with a gray to white matter flow pattern of 4:1
[13]. AD has a characteristic pattern with perfusion def-
icits in the parietal and temporal lobes [14]. A recent
systematic review examined the added value of SPECT to
discriminate between clinically defined AD, vascular de-
mentia (VD), frontotemporal dementia (FTD), and other
non-dementia comparison groups [15]. Randomized con-
trolled trials have not been conducted because of the use
of ionizing radiation, leaving case-control and cohort
studies as the next best alternative [15].
SPECT discriminated between AD, VD, FTD, and non-
dementia patient comparison (PC) groups. The pooled
weighted sensitivities ranged from 65.7% (AD versus PC)
to 71.5% (AD versus FTD). Specificity was lowest overall
for AD versus VD at 75.9%, increasing to 79.1% for AD
versus PC [15]. Data from two studies that compared clin-
ical, SPECT, and histopathologic verification [16,17] were
337
pooled [15]. The weighted sensitivity for brain SPECT
against neuropathology was 74%, with a weighted specific-
ity of 91%. This finding suggests that clinical criteria are
more sensitive than brain SPECT (81% versus 74%), and
brain SPECT has a higher specificity than clinical criteria
(91% versus 70%) [15]. These are conservative estimates
because, in practice, SPECT scans are not usually inter-
preted without clinical information. The authors concluded
that in dementia patients who present with possible comor-
bid symptoms of vascular disease or who do not fit clinical
criteria, SPECT scanning might help rule in AD when the
SPECT scan is positive and rule it out when the SPECT
scan is negative for AD [15].
One SPECT 2-year longitudinal study in MCI subjects
reported that the initial findings in the hippocampal-
amygdaloid complex, the posterior cingulate, the anterior
thalamus, and the caudal portion of the anterior cingulate
predicted progression to dementia in 78% of those who
progress to AD [4,18].
There have been few SPECT studies reporting cross-
sectional differences between MCI and controls [19,20].
Some SPECT studies that used conventional clinical routine
techniques, found computed tomography or MRI had better
discrimination between normal controls, MCI, and dementia
than SPECT [21] and in discriminating between progressive
compared with stable MCI [19,22].
4. fMRI
Functional MRI scans visualize the location of activity in
the brain in response to defined sensory stimulation or cogni-
tive tasks compared with scans taken at rest. Areas of increased
intensity on the scans correspond to the brain areas activated by
the stimulus. Neuronal activity is measured indirectly through
the blood oxygenation level dependent (BOLD) hemodynamic
response to the defined stimulus [23].
Recently published fMRI studies have examined the re-
sponse in people with AD [24], MCI [25], MCI compared
with mild AD [26 –28], FTD versus early AD [29], MCI and
a cholinesterase inhibitor [30], AD and a cholinesterase
inhibitor [31], and normal older adults with or without
APOE e4 allelle [32]. In each of these studies, the number
of subjects was small (range, 7 to 41 subjects per group) and
recruited from university centers or tertiary referral centers.
In the one study of normal older adults, the subjects were
recruited by print advertisements [32].
Only one study had a longitudinal follow-up, mean time
2.5 years, to compare those with MCI who “declined” and
those who remained stable [25]. In this study, activation in
the hippocampus and parahippocampal gyrus (PHG) corre-
lated with better memory performance. However, paradox-
ically those who “progressed” on the clinical dementia rat-
ing scale activated the PHG to a greater extent. This was
interpreted as a possible compensatory response to an un-
derlying progressive AD pathology. In a normal older adult
338 M. Borrie / Alzheimer’s & Dementia 3 (2007) 336 –340
study comparing those with and without the APOE e4 al-
lelle, those with the e4 allelle had a greater magnitude and
extent of BOLD brain response during the picture learning
stimulus compared with those matched with the e3 allelle.
The result suggested that the APOE e4 genotype directly
influences the brain’s response and was consistent with the
hypothesis of a compensatory mechanism in those at genetic
risk of AD [32].
These fMRI studies are providing insights into how brain
ROIs respond to stimuli in response to different degrees of
impairment and assist in hypothesis generation. They are
not applicable as a diagnostic tool to primary or specialist
practice but will continue to advance research understand-
ing of brain function.
5. MRS
MRS is an application of MR imaging that can measure
the concentration of cerebral metabolites of the brain in
vivo. Although various nuclei can be used, the hydrogen
proton 1H and phosphorus 31P are the nuclei that are in high
enough concentration to measure routinely. 1H is most fre-
quently used in MRS for cognitive research because it
provides the most information on metabolic markers of
neurons, myelin, and glia and also has higher sensitivity and
spatial resolution [33–36]. Compounds with a concentration
of at least 0.5 to 1.0 mmol that are commonly measured
with the standard 1.5 T strength magnet include N-acetyl
aspartate (NAA), myoinositol (MI), choline, and creatine.
NAA, an amino acid present only in the central nervous
system neurons of adults, reflects neuronal density or neu-
ronal viability. It is reduced in a number of disease states,
including degenerative central nervous system diseases.
MI is a component of membrane lipids and is considered
a glial cell marker. Choline is thought to reflect cellular
density. Creatine is an osmolyte and has been used as a
“constant” internal reference, although its constancy has
been questioned [34].
Higher magnetic field MR systems at 3 and 4 T strength
(even up to 7 T) have been used for human studies, provid-
ing better signal/noise and contrast/noise ratios and im-
proved resolution [33,37]. Most studies in AD have shown
a reduction in NAA.
Studies of brain regions affected early in AD have shown
reductions in NAA [37– 41], whereas MI has been shown to
increase [37,38,40,42]. Also, correlations between NAA
measured in the MTL and cognitive measures have been
reported in AD [41,43]. NAA and MI ratios show differ-
ences between normal elderly controls, subjects with MCI,
and AD with NAA/MI decreased in AD, with MCI mean
levels falling between controls and AD [37,44].
An MCI longitudinal study, mean period of 3 years,
measuring occipital cortex NAA/creatine ratios predicted
dementia at 100% sensitivity and 75% specificity (area
under the curve, 0.91; 95% CI, 0.80 to 0.97) [45]. Surpris-
ingly in this study, measurements from the hippocampal
cortex were not predictive. These findings need to be du-
plicated.
In cognitively intact subjects older than 85 years, MRS
has shown that subjects with small hippocampal volumes
have a significantly lower NAA/MI ratio in parietal and
temporal lobes compared with the other subjects. This study
did not report a longitudinal component; therefore, it is
unclear whether these subjects are at greater risk of devel-
oping dementia [46].
MRS holds considerable promise as a modality for dif-
ferentiating between MCI and AD and predicting those who
might progress from MCI to AD. Future standardization of
MRS variables and neuroanatomic regions of study will
allow comparability between studies.
6. Summary
Of the four functional neuroimaging modalities re-
viewed, SPECT is the most available to family physicians
and specialists. It might provide added value in patients with
comorbid symptoms of vascular disease or who do not fit
the clinical criteria for dementia. Amyloid ligand binding in
SPECT is still speculative. PET scanning for dementia di-
agnosis is not available for clinical use in Canada except to
a very limited degree, but its utility clinically in the United
States will be followed with interest. It has also been
encouraged for use in clinical trials. MRS and fMRI
continue in the research arena and hold considerable
promise in being able to predict progression from MCI to
dementia, determine response to research compounds,
and are useful in hypothesis generation. In combination
with structural MRI and neuropsychological tests, func-
tional neuroimaging holds promise for an earlier diagno-
sis of cognitive impairment.
7. Recommendations approved at the 3rd
CCCDTD Conference
1. There is fair evidence that functional imaging with
PET or SPECT scanning might assist specialists in the
differential diagnosis of dementia, particularly those
with questionable early stage dementia or those with
FTD. There is variability across centers, with requi-
site expertise in these modalities that needs to be
taken into account in determining utility (Grade B,
Level 2).
2. fMRI and MRS scanning are not recommended for
use by family physicians or specialists to make or
differentiate a diagnosis of dementia in people pre-
senting with cognitive impairment. They remain very
promising research tools (Grade D, Level 3).
 M. Borrie / Alzheimer’s & Dementia 3 (2007) 336 –340
Author Disclosures
Michael Borrie has received support from Pfizer Canada
(Medical Advisory Board Member, Speaker, Clinical Trial
Investigator), Janssen Ortho Inc. (Medical Advisory Board
Member, Speaker, Clinical Trial Investigator), Novartis
(Medical Advisory Board Member, Speaker, Clinical Trial
Investigator), Lundbeck Canada (Medical Advisory Board
Member, Speaker, Clinical Trial Investigator), Neurochem
(Clinical Trials Investigator), Sanofi Aventis (Clinical Tri-
als Investigator), GlaxoSmithKline (Clinical Trials Investi-
gator), ONO (Clinical Trials Investigator), Myriad (Clinical
Trials Investigator), and HMR (Clinical Trials Investigator).
Acknowledgments
The author would like to acknowledge the contributions
to the writing of this paper of Matthew Smith, Research
Coordinator; Howard Chertkow, MD FRCP(C); and Judy
McCallum, Administrative Assistant.
References
[1] Petrella JR, Coleman RE, Doraiswamy PM. Neuroimaging and early
diagnosis of Alzheimer disease: A look to the future1. Radiology
2003;226:315–36.
[2] Mosconi L. Brain glucose metabolism in the early and specific diag-
nosis of Alzheimer’s disease. FDG-PET studies in MCI and AD. Eur
J Nucl Med Mol Imaging 2005;32;486 –510.
[3] Zakzanis KK, Graham SJ, Campbell Z. A meta-analysis of structural
and functional brain imaging in dementia of the Alzheimer’s type: a
neuroimaging profile. Neuropsychology Review 2003;13;1.
[4] Wolf H, Jelic V, Gertz HJ, Nordberg A, Julin P, Wahlund LO. A
critical discussion of the role of neuroimaging in mild cognitive
impairment. Acta Neurol Scand Suppl 2003;107:52–76.
[5] de Leon MJ, Convit A, Wolf OT, Tarshish CY, DeSanti S, Rusinek
H, et al. Prediction of cognitive decline in normal elderly subjects
with 2-[18F]fluoro-2-deoxy-D-glucose/positron-emission tomography
(FDG/PET). Proc Natl Acad Sci U S A 2001;98:10966 –71.
[6] De Santi S, de Leon MJ, Rusinek H, Convit A, Tarshish CY, Roche
A, et al. Hippocampal formation glucose metabolism and volume
losses in MCI and AD. Neurobiol Aging 2001;22:529 –39.
[7] Nestor PJ, Fryer TD, Smielewski P, Hodges JR. Limbic hypometab-
olism in Alzheimer’s disease and mild cognitive impairment. Ann
Neurol 2003;54:343–51.
[8] Patwardhan MB, McCrory DC, Matchar DB, Samsa GP, Rutschmann
OT. Alzheimer disease: Operating characteristics of PET - A meta-
analysis. Radiology 2004;231:73– 80.
[9] Mosconi L, De Santi S, Li Y, Li J, Zhan J, Tsui WH, et al. Visual
rating of medial temporal lobe metabolism in mild cognitive impair-
ment and Alzheimer’s disease using FDG-PET. Eur J Nucl Med Mol
Imaging 2006;33:210 –21.
[10] Nordberg A. PET imaging of amyloid in Alzheimer’s disease. Lancet
Neurol 2004;3:519 –27.
[11] Verhoeff N, Wilson AA, Takeshita S, Trop L, Mussey D, Singh K, et al.
In-Vivo imaging of Alzheimer disease ␤-Amyloid with [11C]SB-13
PET. AM J Geriatr Psychiatry 2004;12:584 –95.
[12] Wang Y, Klunk WE, Debnath ML, Huang GF, Holt DP, Shao L, et
al. Development of a PET/SPECT agent for amyloid imaging in
Alzheimer’s disease. J Mol Neurosci. 2004;24:55– 62.
339
[13] Morano GN and Seibyl JP. Technical overview of brain SPECT
imaging: Improving acquisition and processing data. J Nucl Med
Technol 2003;31:191–5.
[14] Small GW, Rabins PV, Barry PP, Buckholtz NS, DeKosky ST, Ferris
SH, et al. Diagnosis and treatment of Alzheimer disease and related
disorders: Consensus statement of the American Association for Ge-
riatric Psychiatry, the Alzheimer’s Association, and the American
Geriatrics Society. JAMA 1997;278:1363–71.
[15] Dougall NJ, Bruggink S, Ebmeier KP. Systematic review of the
diagnostic accuracy of 99mTc-HMPAO-SPECT in dementia. Am J
Geriatr Psychiatry 2004;12:554 –70.
[16] Bonte FJ, Weiner MF, Bigio E, White CL. Brain blood flow in the
dementias: SPECT with histopathologic correlation in 54 patients.
Radiology 1997;202:793–7.
[17] Jagust W, Thisted R, Devous MD Sr, Van Heertum R, Mayberg H,
Jobst K, et al. SPECT perfusion imaging in the diagnosis of
Alzheimer’s disease: a clinical-pathologic study. Neurology 2001;
56:950 – 6.
[18] Johnson KA, Jones K, Holman BL, Becker JA, Spiers PA, Satlin A,
et al. Preclinical prediction of Alzheimer’s disease using SPECT.
Neurology 1998;50:1563–71.
[19] Kogure D, Matsuda H, Ohnishi T, Asada T, Uno M, Kunihiro T, et al.
Longitudinal evaluation of early Alzheimer’s disease using brain
perfusion SPECT. J Nucl Med 2000;41:1155– 62.
[20] Celsis P, Agniel A, Cardebat D, Demonet JF, Ousset PJ, Puel M.
Age-related cognitive decline: A clinical entity? A longitudinal study
of cerebral blood flow and memory performance. J Neurol Neurosurg
Psychiatry 1997;62:601– 8.
[21] Scheltens P, Launer LJ, Barkhof F, Weinstein HC, Jonker C. The
diagnostic value of magnetic resonance imaging and technetium 99m-
HMPAO single-photon-emission computed tomography for the diag-
nosis of Alzheimer disease in a community-dwelling elderly popula-
tion. Alzheimer Dis Assoc Disord 1997;11:63–70.
[22] McKelvey R, Bergman H, Stern J, Rush C, Zahirney G, Chertkow H.
Lack of prognostic significance of SPECT abnormalities in non-
demented elderly subjects with memory loss. Can J Neurol Sci 1999;
26:23– 8.
[23] Smith SM. Overview of fMRI analysis. Br J Radiol 2004;77:S167–
S75.
[24] Golby A, Silverberg G, Race E, Gabrieli S, O’Shea J, Knierim K, et
al. Memory encoding in Alzheimer’s disease: an fMRI study of
explicit and implicit memory. Brain 2005;128:773– 87.
[25] Dickerson BC, Salat DH, Bates JF, Mtiya M, Killiany RJ, Greve DN,
et al. Medical temporal lobe function and structure in mild cognitive
impairment. Ann Neurol 2004;56:27–35.
[26] Rombouts SARB, Barkhof F, Goekoop R, Stam CJ, Scheltens P.
Altered resting state networks in mild cognitive impairment and mild
Alzheimer’s disease: An fMRI study. Hum Brain Mapp 2005;26:
231–9.
[27] Yetkin FZ, Rosenberg RN, Weiner MF, Purdy PD, Cullum CM.
FMRI of working memory in patients with mild cognitive impairment
and probably Alzheimer’s disease. Eur Radiol 2006;16:193–206.
[28] Rombouts SARB, Goekoop R, Stam CJ, Barkhof F, Scheltens P.
Delayed rather than decreased BOLD response as a marker for early
Alzheimer’s disease. NeuroImage 2005;26:1078 – 85.
[29] Rombouts SARB, van Swieten JC, Pijnenburg YAL, Goekoop R,
Barkhof F, Scheltens P. Loss of frontal fMRI activation in early
frontotemporal dementia compared to early AD. Neurology 2003;60:
1904 – 8.
[30] Goekoop R, Rombouts SARB, Jonker C, Hibbel A, Knol DL, Truyen
L, et al. Challenging the cholinergic system in mild cognitive impair-
ment: a pharmacological fMRI study. NeuroImage 2004;23:1450 –9.
[31] Kircher TTJ, Erb M, Grodd W, Leube DT. Cortical activation during
cholinesterase-inhibitor treatment in Alzheimer disease. Preliminary
340 M. Borrie / Alzheimer’s & Dementia 3 (2007) 336 –340
findings from a pharmaco-fMRI study. Am J Geriatr Psychiatry
2005;13:1006 –13.
[32] Bondi MW, Houston WS, Eyler LT, Brown GG. fMRI evidence of
compensatory mechanisms in older adults at genetic risk for Alzhei-
mer disease. Neurology 2005;64:501– 8.
[33] Di Costanzo A, Trojsi F, Tosetti M, Giannatempo GM, Nemore F,
Piccirillo M, et al. High-field proton MRS of human brain. Eur J
Radiol 2003;48:146 –53.
[34] Ross AJ and Sachdev PS. Magnetic resonance spectroscopy in cog-
nitive research. Brain Research Reviews 2004;44:83–102.
[35] Gujar SK, Maheshwari S, Bjorkman-Burtscher I, Sundgren PC. Mag-
netic resonance spectroscopy. J Neuro-Ophthalmol 2005;25:217–26.
[36] Mason GF. Magnetic resonance spectroscopy for studies of neuro-
transmission in vivo. Psychopharmacol Bull 2003;37:26 – 40.
[37] Kantarci K, Reynolds G, Petersen RC, Boeve BF, Knopman DS,
Edland SD, et al. Proton MR spectroscopy in mild cognitive impair-
ment and Alzheimer disease: Comparison of 1.5 and 3 T. Am J
Neuroradiol 2003;24:843–9.
[38] Parnetti L, Tarducci R, Presciutti O. Proton magnetic resonance
spectroscopy can differentiate Alzheimer’s disease from normal ag-
ing. Mech. Ageing Dev 1997;97:9 –14.
[39] Frederick BB, Satlin A, Yurgelun-Todd DA, Renshaw PF. In vivo
proton magnetic resonance spectroscopy of Alzheimer’s disease in
the parietal and temporal lobes. Biol Psychiatry 1997;42:147–50.
[40] Kantarci K, Jack CR, Jr. Xu YC, Campeau NG, O’Brien PC, Smith
GE, et al. Regional metabolic patterns in mild cognitive impairment
and Alzheimer’s disease: a 1H MRS study. Neurology 2000;55:
210 –7.
[41] Jessen F, Block W, Traber F, Keller E, Flacke S, Papassotiropoulos
A, et al. Proton MR spectroscopy detects a relative decrease of
N-acetyl aspartate in the medial temporal lobe of patients with AD.
Neurology 2000;55:684 – 8.
[42] Rose SE, de Zubicaray GI, Greig I, Wang D, Galloway GJ, Chalk JB,
et al. A 1H MRS study of probable Alzheimer’s disease and normal
aging: implications for longitudinal monitoring of dementia progres-
sion. Magn Reson Imaging 1999;17:291–9.
[43] Chantal S, Labelle M, Bouchard RW, Braun CM, Boulanger Y.
Correlation of regional proton magnetic resonance spectroscopic met-
abolic changes with cognitive deficits in mild Alzheimer disease.
Arch Neurol 2002;59:955– 62.
[44] Catani M, Cherubini A, Howard R, Tarducci R, Pelliccioli GP,
Piccirilli M, et al. 1H-MR spectroscopy differentiates mild cognitive
impairment from normal brain aging. NeuroReport 2001;12:2315–7.
[45] Modrego PJ, Fayed N, Pina MA. Conversion from mild cognitive
impairment to probable Alzheimer’s disease predicted by brain mag-
netic resonance spectroscopy. Am J Psychiatry 2005;162:667–75.
[46] Spencer DC, Zitzelberger T, Spielman D, Kaye J. MRS in relation to
hippocampal volume in the oldest old. Neurology 2003;60:1194 – 6.
 Alzheimer’s & Dementia 3 (2007) 341–347

General risk factors for dementia: A systematic evidence review

Christopher Pattersona,*, John Feightnerb, Angeles Garciac, Chris MacKnightd
a
Division of Geriatric Medicine, Department of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada
b
Department of Family Medicine, Faculty of Medicine, University of Western Ontario, London, Ontario, Canada
c
Department of Medicine (Geriatrics), Queen’s University, Kingston, Ontario, Canada
d
Division of Geriatric Medicine, Dalhousie University, Halifax, Nova Scotia, Canada

Abstract This review identifies and quantifies general (ie, nongenetic) risk factors for all-cause dementia,
Alzheimer’s disease, and vascular dementia specifically.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Alzheimer’s disease; Vascular dementia; Risk factors

1. Background reason, the present systematic review focuses solely on

longitudinal cohort studies.
Since the Canadian Consensus Conference on Demen-
We posed the question “What modifiable risk factors are
tia was convened in 1999, many advances have been
associated with dementia, Alzheimer’s disease, or vascular
made in the understanding of dementing disorders. One
dementia?” A comprehensive electronic literature search
area that was considered worthy of further investigation
was performed on Medline and EMBASE from 1966 to
was that of risk factors for dementia, because knowledge
December 2005. The search string is listed in Appendix 1.
of these risk factors has direct relevance to the primary
This search identified 3,424 articles. It was possible to reject
prevention of these disorders. The topic of preventing
1,705 of these as obviously irrelevant. The remaining 1,719
dementia was previously addressed at the Canadian Con-
references with their abstracts were reviewed by a single
sensus Conference on Dementia [1], and an update on
primary prevention appears as a companion piece to the reader (C.P.), who identified possibly relevant articles and
present article [2]. set aside the remainder. These latter references and abstracts
were then reviewed by three other readers (J.F., A.G., C.M.)
to ensure that no possibly relevant articles had been dis-
2. Methods carded. The following criteria were used for relevance: (1)
longitudinal cohort studies; (2) a population broadly repre-
The identification and relative importance of risk factors
sentative of Canadian demographics; (3) dementia, Alzhei-
are best addressed through longitudinal cohort studies. Al-
mer’s disease, or vascular dementia as outcome; and (4)
though case-control studies can be used to explore risk
general risk factors identified (eg, hypertension, educational
factors for disease, significant differences have been ob-
status, occupation, chemical exposure.) Articles addressing
served when conclusions of case-controlled studies are
genetic risk factors were excluded.
compared with longitudinal studies. An example of differ-
After assembling all possibly relevant publications, the
ent conclusions is found with tobacco, where case-control
full articles were obtained, and each was assessed indepen-
studies have suggested that tobacco is associated with a
dently by two reviewers for quality. Additional articles from
lower risk of dementia, whereas numerous longitudinal
personal files and those identified by scrutiny of bibliogra-
studies have shown that the risk is increased [3]. For this
phies of retrieved articles were added to the pool for quality
assessment. After consulting several key references on ap-
praisal of risk factor literature, the following criteria were
*Corresponding author. Tel.: (905) 521-7931; Fax: (905) 521-7905. chosen to determine whether an article was good, fair, poor,
E-mail address: pattec@hhsc.ca or ineligible.
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.001
342 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
2.1. Population characteristics
Inclusion criteria were defined, including ages, locations,
dates; exclusion of dementia at inception; and population
broadly representative of Canadian demographics.
2.2. Follow-up
All participants were accounted for at end of follow-up,
without excessive “preventable” losses. No differences in
follow-up were found between those developing dementia
and those with no dementia.
2.3. Exposures (risk factors)
List of risk factors was considered. Exposures were mea-
sured in the same way for those with and without dementia.
2.4. Outcomes
Dementia and subtypes were defined by standardized
criteria. Ascertainment of outcome was independent of ex-
posure status.
2.5. Analysis
Important confounders were included in analysis (age,
sex, education at minimum). Statistical analysis was appro-
priate (ie, multiple logistic regression). Specific measures of
risk are stated (eg, relative risk).
The overall quality rating was determined as good (all
criteria fulfilled adequately); fair (most criteria fulfilled ad-
equately, no fatal flaw); or poor (fatal flaw or multiple minor
inadequacies). Each article was reviewed for quality inde-
pendently by two readers, and a consensus was reached if
disagreement occurred. Articles graded as good or fair were
then submitted for data abstraction by two independent
reviewers.
3. Results
Risk factors are either immutable (eg, age, gender, eth-
nicity) or potentially modifiable. This article deals only with
potentially modifiable risk factors. A total of 60 articles of
single longitudinal studies were identified to be of good or
fair quality. In addition, eight systematic reviews of good or
fair quality also addressed relevant risk factors. Table 1
summarizes risk factors and their relative risk (RR) for
dementia of different types. Other measures of risk (eg,
odds ratio [OR], hazard ratio [HR]) are indicated where
necessary.
4. Potentially modifiable risk factors
4.1. Blood pressure
The relationship between blood pressure and subsequent
development of dementia is somewhat confusing. Both
higher and lower systolic (SBP) and diastolic blood pressure
(DBP) levels have been implicated in increased risks of
subsequent dementia. For example, from the Kungsholmen
project, SBP in excess of 180 mm Hg is associated with an
increased RR of all-cause dementia (ACD) of 1.6 and for
Alzheimer’s disease (AD) of 1.5 after 6 years [4]. However,
in the same study population, a SBP of less than 140 mm Hg
was also associated with increased risk of ACD (RR, 1.9)
and AD (RR, 2.2) [5]. From the combined results of the
Rotterdam and Gothenberg H-70 studies in individuals who
are taking antihypertensive agents, each increase of 10
mm Hg in SBP appears to lower the RR of ACD (0.93) and
AD (0.89) [6]. A longitudinal study in Finland reported an
increased relative risk of AD of 2.3 in the long term after
elevated SBP [7]. However, a decline in SBP of 15 or more
mm Hg appears to predate the onset of both ACD and AD,
according to data from the Kungsholmen project [8]. From
the same study group a DBP of less than 65 mm Hg is also
associated with a RR of 1.54 for ACD and 1.7 for AD, as
does a pulse pressure of less than 70 mm Hg [9]. Thus, it
would appear that increased SBP increases the risk of sub-
sequent ACD, AD, and vascular dementia (VaD) in women,
but so apparently does reduced systolic, diastolic, and pulse
pressure below normal levels, overall a somewhat contra-
dictory and confusing situation.
4.2. Diabetes mellitus
Several studies have linked the presence of type 2 dia-
betes mellitus to subsequent development of dementia.
Analysis of data from the Canadian Study of Health and
Aging (CSHA) concluded that diabetes increased the risk of
ACD and AD, although the increased risks did not reach
statistical significance. For VaD, however, the RR was 2.03
[10]. Diabetes in midlife was shown to increase the subse-
quent risk of ACD (OR, 2.83) in a stratified sample of civil
servants from Tel Aviv [11]. From the Kungsholmen Study
in Sweden, Xu et al [12] concluded that there was an
increased risk of ACD and VaD (HR, 1.5 and 2.6, respec-
tively), although the increased risk for AD did not reach
statistical significance. Data collected within a large ran-
domized controlled trial of raloxifene suggested an in-
creased risk of ACD in those with diabetes mellitus, but the
increased OR of 2.38 did not reach statistical significance
[13]. Thus, there is fair evidence that the presence of type 2
diabetes mellitus increases the risk of ACD and VaD.
4.3. Stroke
The presence of clinical strokes or of silent infarctions on
neuroimaging increases the risk of subsequent dementia.
The temporal relationship between stroke and onset of de-
mentia is a diagnostic criterion for VaD. Longitudinal stud-
ies have shown that stroke increases the subsequent risk of
ACD (RR, 2.4) [14] and AD (RR, 1.83) [15]. The presence
of silent brain infarctions visible on magnetic resonance
imaging scanning is associated with an increased risk of
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
Table1
Risk factors for dementia: RR (95% CI) [references]
ACD
Systolic hypertension 1.6 (1.1 to 2.2) [4]
DBP ⬍ 65 mm Hg 1.5 (1.0 to 2.1) [4]
PP ⬎ 84 mm Hg 1.3 (0.9 to 1.7) [9]
PP ⬍ 70 mm Hg 1.4 (1.0 to 1.9) [9]
SBP decline ⱖ 15 mm Hg 3.1 (1.5 to 6.3) [8]
SBP ⬍ 140 mm Hg 1.9 (1.2 to 3.2) [5]
Diabetes mellitus, type 2 1.26 (0.9 to 1.76) [10]
2.38 (0.92 to 6.12) [13]
2.83 (1.4 to 5.71) [11]
Stroke 2.4 (1.6 to 3.7) [14]
Silent infarcts on neuroimaging 2.17 (1.04 to 4.54) [18]
HR, 2.26 (1.09 to 4.70) [16]
Serum cholesterol 1 3.1 (1.5 to 6.3) [19]
Serum homocysteine 1 2.08 (1.31 to 3.30) [20]
Serum thyroid-stimulating hormone 2
Sex hormones: higher total estrogen 1.45 (1.08 to 1.94) [22]
level (women)
Sex hormones: higher free testosterone 0.74 (0.57 to 0.96) [23]
index (men)
Depression 1.87 (1.09 to 3.20) [26]
Men: OR, 3.5 (1.9 to 6.5)
Women: OR, 1.2 (0.7 to 2.0) [24]
High fat intake 2.4 (1.1 to 5.4) [27]
Fish consumption ⬎ once/week 0.4 (0.2 to 0.9) [27]
0.73 (0.52 to 1.03) [28]
Vitamin E: lowest tertile serum level 2.54 (1.06 to 6.10) [29]
Moderate wine consumption, 250 to OR, 0.19 (.05 to 0.66) [30]
500 mL/day 0.56 (0.36 to 0.92) [28]
Activity: highest level, physical 0.58 (0.41 to 0.83) [33]
0.63 (0.40 to 0.98) [32]
Activity: highest level, mental, daily 0.59 (0.37 to 0.96) [34]
Smoking
Occupation Manual work: 1.6 (1.0 to 2.5) [49]
Exposure to toxins (pesticides,
fertilizers, fumigants), defoliants
Head injury With LOC: 1.14 (0.84 to 1.56) [37]
Moderate: HR, 2.39 (1.24 to 4.58) [38]
Severe: HR, 4.48 (2.09 to 9.63) [38]
Education ⬎ 15 vs ⬍ 12 0.64 (0.4 to 1.0) [48]
Statin drugs 1.19 (0.82 to 1.75) [41]
HR, 1.19 (0.53 to 2.34) [40]
All lipid-lowering drugs
NSAIDs 0.51 (0.37 to 0.70) [44]
Benzodiazepines Ever used: OR, 1.7 (1.2 to 2.4)
Former use: OR, 2.3 (1.2 to 4.5) [45]
Vaccination
Abbreviation: LOC, loss of consciousness.
343
AD VAD
RR, 1.5 (1.0 to 2.3) [4]; 2.05 (1.2 to 3.53) [25],
OR, 2.3 (1.0 to 5.5) [7] women only
1.7 (1.1 to 2.4) [4]
1.4 (1.0 to 2.0) [9]
1.7 (1.2 to 2.3) [9]
3.1 (1.3 to 7.0) [8]
2.2 (1.2 to 3.8) [5]
1.3 (0.83 to 2.03) [10] 2.03 (1.15 to 3.57) [10]
1.3 (0.9 to 2.1) [12] HR, 2.6 (1.2 to 6.1) [12]
1.83 (1.14 to 2.95) [15] Required for diagnosis
2.1 (1.0 to 4.4) [7]
3.1 (1.2 to 8.5) [19]
2.11 (1.19 to 3.76) [20]
3.5 (1.1 to 11.5) [21]
1.30 (0.88 to 1.99) [22]
Men: OR, 4.2 (2.1 to 8.8) OR, 2.41 (1.22 to 4.52) [25]
Women: OR, 1.0 (0.5 to 1.9) [24]
0.3 (0.1 to 0.9) [27] Shellfish: OR, 0.45
(.22 to .88) [25]
2.10 (0.81 to 5.54) [29]
0.53 (0.3 to 0.95) [28]
0.5 (0.28 to 0.90) [32] OR, 0.46 (0.25 to 0.82),
0.55 (0.34 to 0.88) [33] women [25]
0.69 (0.5 to 0.96) [31]
1.10 (0.94 to 1.29) [37]
1.99 (1.33 to 2.98) [3]
Manual work: 1.4 (0.9 to 2.1) [49]
4.35 (1.05 to 17.90) [36] OR, 2.05 (1.03 to 3.85) [25]
With LOC: 1.02 (0.68 to 1.15) [37]
Moderate: HR, 2.32 (1.04 to 5.1) [38]
Severe: HR, 4.51 (1.77 to 11.47) [38]
0.48 (0.27 to 0.84) [48]
0.82 (0.46 to 1.46) [41]
HR, 1.19 (0.35 to 2.96) [40]
OR, 0.26 (0.08 to 0.88) [39], age less
than 80 y
0.42 (0.26 to 0.66) [42]
Any vaccination: 0.40
(0.17 to 0.96) [36]
Influenza: 0.75 (0.54 to 1.04)
Poliomyelitis: 0.6 (.37 to .99)
Diphtheria or tetanus: 0.41
(.27 to .62) [47]
344 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
ACD (HR, 2.26) [16]. From a prospective study of stroke
patients in hospital in Lille, France, in which prestroke
cognitive function was assessed with the IQCODE [17],
the presence of silent infarctions increased the risk of sub-
sequent ACD (RR, 2.7) [18]. Thus, there is compelling
evidence that the presence of strokes increases the risk of
both ACD and VaD.
4.4. Serum cholesterol
Midlife elevation in serum cholesterol has been associ-
ated with increased risk of subsequent AD (OR, 2.1) [7] and
(OR, 3.1) [19]. The risk of ACD was also elevated (RR, 3.1)
with higher levels of cholesterol [19]. There appears to be
good evidence that elevated serum cholesterol is associated
with an increased risk of ACD and AD.
4.5. Hyperhomocysteinemia
A single prospective study has established an association
between serum homocysteine levels in excess of 15 ␮mol/L
and ACD (RR, 2.08) and AD (RR, 2.11; 95% confidence
interval [CI], 1.19 to 3.76) [20].
4.6. Hyperthyroidism
A single study showed an association between a de-
pressed level of serum thyroid-stimulating hormone and
subsequent development of AD [21].
4.7. Sex hormone levels
In women, levels of total estrogen are associated with a
higher risk of ACD (RR, 1.45) [22], whereas in men, higher
levels of free testosterone index are associated with lower
risk of AD (RR, 0.74) [23].
4.8. Depression
The presence of depressive symptoms was associated
with subsequent development of dementia only in men in
the PAQUID Study. For men, the OR of developing ACD
was 3.5, and the RR for AD was 4.3. In women there was
no statistically significant association [24]. For VaD, data
from the CSHA indicated that the presence of depression
increased the risk of VaD (OR, 2.41) [25]. A systematic
evidence review concluded that depression is associated
with an RR of 1.87 for ACD [26].
4.9. Lifestyle factors
4.9.1. Diet
A high intake of total fat (⬎ 85.5 g/day) increases the RR
for ACD by 2.4 compared with a fat intake of ⬍ 75.5 g/day
[27], but in this longitudinal population study from the
Netherlands, increased amounts of saturated fat and choles-
terol were not established as definite risk factors. The con-
sumption of fish greater than 18.5 g/day compared with less
than 3.0 g/day lowered the risk of both ACD (RR, 0.4) and
AD (RR, 0.3). Other studies have also suggested the benefit
of eating fish regularly. In the PAQUID longitudinal study,
weekly consumption of fish was associated with an RR of
0.73 for ACD, although this was of marginal significance
[28]. Regular consumption of shellfish was also associated
with a reduced risk of VaD in the CSHA, with an OR of
0.45 [25]. There appears to be consistent epidemiologic
evidence that regular consumption of fish and seafood is
associated with reduced risk of dementia. A single study
reported an association between the lowest tertile of serum
vitamin E levels and elevated risk of both ACD and AD, but
statistical significance was not reached [29].
4.9.2. Wine
The PAQUID study in France established that the con-
sumption of moderate amounts of red wine (250 to 500
mL/day) was associated with a RR of 0.56 for ACD and a
RR of 0.53 for AD [28]. In the same study population,
moderate consumption, compared with other degrees of
usage, was associated with a lower risk of ACD (RR, 0.19)
[30]. Furthermore, in the CSHA population, consumption of
wine at least weekly was associated with a reduced risk of
AD (OR, 0.49), although this lost statistical significance
when decedents were included in the analysis [31].
4.9.3. Activity
There is evidence that regular physical activity is asso-
ciated with a reduced risk of dementia. In the CSHA, reg-
ular physical activity was associated with an OR of 0.69 for
AD [31], and when this was quantified, those with the
highest level of physical activity had a reduced risk of ACD
(OR, 0.63) and of AD (OR, 0.5) [32]. In the Cardiovascular
Health Study from the USA, when those expending the
highest quartile of energy expenditure were compared with
the lowest, the RR was 0.58 for ACD and 0.55 for AD [33].
Regular exercise was also associated with RR of VaD in the
CSHA; in women only, OR for VaD was 0.46 [25].
Daily mental activities were associated with a RR of 0.59
for ACD in the Kungsholmen Study [34]. In the Washington
Heights District of New York, a longitudinal study found an
association between high leisure activities and decreased
incidence of ACD, with a RR of 0.62 [35]. It is thus
plausible that maintaining a high level of mental activity
might be associated with reduced subsequent incidence of
ACD.
4.9.4. Smoking
A systematic review [3] compared the association of
smoking with dementia in case-controlled and prospective
cohort studies. Although there appeared to be a protective
effect in the case-controlled studies, the prospective cohort
studies established an increased risk for AD (RR, 1.99) in
studies that described the number of smokers at baseline [3].
This should settle, once and for all, the positive association
between smoking and AD.
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
4.9.5. Exposure to toxins
Evidence from CSHA established an association be-
tween exposure to pesticides or fertilizers and VaD (OR,
2.05) [25]. Exposure to defoliants and fumigants was asso-
ciated with an increased risk of AD (RR, 4.35) in a longi-
tudinal study from Manitoba, Canada [36].
4.9.6. Head trauma
Four population-based samples comprised the EURODEM
Study. In this study head trauma with unconsciousness was
associated with an increased RR of both ACD (1.14) and AD
(1.02); however, neither of these reached statistical signifi-
cance [37]. In the CSHA, prior head injury with or without
loss of consciousness did not show any increased risk of AD
(OR, 0.87; 95% CI, 0.56 to 1.36) [31]. However, in a
long-term study of U.S. servicemen who had been injured
during World War II, moderate head injury was associated
with an HR of 2.39 for ACD and 2.32 for AD [38]. For
severe injury the HR for ACD was 4.48 and 4.51 for AD. A
conclusive association between head injury and subsequent
dementia remains to be established.
4.10. Drug exposure
4.10.1. Lipid-lowering agents
Although a nested case-control study within the CSHA
suggested that in individuals younger than 80 years of age,
use of all types of lipid-lowering agents reduced the risk of
AD, with an OR of 0.26 at 5 years [39], other studies have
not confirmed this finding. In the Cache County Study, the
use of statin-type drugs was associated with a nonsignificant
increased HR of 1.19 for both ACD (95% CI, 0.53 to 2.34)
and AD (95% CI, 0.53 to 2.96) at 3 years [40]. There were
similar findings in an American community-based study in
which use of statin agents was associated with a nonsignif-
icant increased HR of 1.19 (95% CI, 0.83 to 1.75) for ACD
but a nonsignificant reduced HR of 0.82 (95% CI, 0.46 to
1.46) for AD [41].
4.10.2. Nonsteroidal anti-inflammatory drugs
Individual studies such as the Cache County popula-
tion have concluded that exposure to nonsteroidal anti-
inflammatory drugs (NSAIDs) is associated with reduced
risk of AD (RR, 0.42) [42]. A systematic review of six
longitudinal studies in 2003 concluded that there was a
nonsignificant reduction in RR of 0.84 (95% CI, 0.54
to1.05) for AD [43]. However, a more recent and compre-
hensive systematic review of 25 studies, both cohort and
case-controlled, concluded that NSAID exposure was asso-
ciated with a significantly reduced risk of both prevalent
(RR, 0.51) and incident (RR, 0.79) cases of ACD [44].
4.10.3. Benzodiazepines
In a French population-based study, exposure to benzo-
diazepine drugs at any time was associated with an in-
creased risk of ACD (OR, 1.7), and for former use the OR
was 2.3 [45]. However, a systematic review of six prospec-
345
tive studies concluded that increased risk of cognitive de-
cline was reported in three, decreased in two, and un-
changed in one [46].
4.10.4. Vaccinations
Prior inoculations against poliomyelitis, diphtheria, or
tetanus and influenza are all associated with a lower risk of
AD [47]. In addition, inoculations of any kind appear to
reduce AD risk [36].
4.11. Education
Longer periods of education have been associated with
reduced risk of AD, and this relationship has been con-
firmed. When compared with those receiving less than 12
years of education, people engaged in more than 15 years of
education had a reduced risk of both ACD (RR, 0.64) and
AD (RR, 0.48) [48].
4.12. Occupation
Data from the Kungsholmen project in Sweden sug-
gested that manual labor increased the risk of both ACD
(RR, 1.6) and AD (RR, 1.4), although the latter just missed
statistical significance [49]. Other studies have not con-
firmed this relationship.
5. Conclusion
This systematic review has explored the current state of
evidence concerning risk factors for ACD, AD, and VaD,
which are presented in Table 1. Although for many of these
risk factors the evidence remains inconclusive or contradic-
tory, for others an emerging picture of consensus is appear-
ing. It should be noted that with rare exceptions, there are no
data on the effects of multiple risk factors in the same
individual. One exception is a Finnish study in which the
combined risk of systolic hypertension (OR, 2.3) and hy-
percholesterolemia (OR, 2.1) increases the OR for AD to
3.5 [7]. Research priorities should focus on remediable risk
factors that have the highest chance of being corrected.
These include such risk factors as hypertension, hypercho-
lesterolemia, inadequate physical and mental activity, and
dietary factors. Attempts to limit the noxious effects of
smoking and head injury are justifiable for their own sake
and might also have benefits in the cognitive domain.
Author Disclosures
Angeles Garcia has received support from Janssen-Ortho
(Consultant, Advisor), Pfizer (Consultant, Advisor), Novartis
(Consultant, Advisor), and Lundbeck (Consultant, Advisor).
Chris MacKnight has received support from Alzheimer
Society of Nova Scotia (Board Member), Janssen-Ortho
(Speaker, Trial Investigator), Pfizer Canada (Speaker, Trial
Investigator), Novartis (Speaker, Trial Investigator), Lund-
beck Canada (Speaker, Trial Investigator), Voyager Phar-
346 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
maceuticals (Trial Investigator), Myriad (Trial Investiga-
tor), and Neurochem (Trial Investigator).
Acknowledgments
During preparation of this article, Christopher Patterson
received financial support from the Institute of Advanced
Studies, University of Bologna, Bologna, Italy.
Chris MacKnight is supported from a CIHR New Inves-
tigator Award.
References
[1] Black SE, Patterson C, Feightner J. Preventing dementia. Can J Neu-
rol Sci 2001;28(Suppl 1):S56 – 66.
[2] Patterson C, Feightner JW, Garcia A, MacKnight CR. Primary pre-
vention of dementia. Alzheimers Dement 2007;3:.
[3] Almeida OP, Hulse GK, Lawrence D, Flicker L. Smoking as a risk
factor for Alzheimer’s disease; contrasting evidence from a system-
atic review of case-control and cohort studies. Addiction 2002;1997:
15–28.
[4] Qui C, von Strauss E, Fastbom J, Winblad B, Fratiglioni L. Low
blood pressure and risk of dementia in the Kungsholmen Project: a
6-year follow-up study. Arch Neurol 2003;60:223– 8.
[5] Guo C, Viitanen M, Winblad B, Fratiglioni L. Low blood pressure
and incidence of dementia in a very old sample: dependent on initial
cognition. J Am Geriatr Soc 1999;47:723– 6.
[6] Ruitenberg A, Skoog I, Ott A, Aevarsson O, Witteman JCM, Lernfelt
B, et al. Blood pressure and risk of dementia: results from the
Rotterdam Study and the Tothenburg H-70 Study. Dement Geriatr
Cogn Disord 2001;12:33–9.
[7] Kivipelto M, Helkala E-L, Laakso MP, Hänninen T, Hallikainen M,
Alhainen K, et al. Midlife vascular risk factors and Alzheimer’s
disease in later life: longitudinal, population based study. BMJ 2001;
322:1447–51.
[8] Qui C, von Strauss E, Winblad B, Fratiglioni L. Decline in blood
pressure over time and risk of dementia: a longitudinal study from the
Kungsholmen project. Stroke 2004;35:1810 –5.
[9] Qui C, Winblad B, Wiitanen M, Fratiglioni L. Pulse pressure and risk of
Alzheimer disease in persons aged 75 years and older: a community-
based, longitudinal study. Stroke 2003;34:594 –9.
[10] MacKnight C, Rockwood K, Awalt E, McDowell I. Diabetes mellitus
and the risk of dementia, Alzheimer’s disease and vascular cognitive
impairment in the Canadian Study of Health and Aging. Dement
Geriatr Cogn Disord 2002;14:77– 83.
[11] Schnaider Beeri M, Goldbourt U, Silverman JM, Noy S, Schmeidler
J, Ravona-Springer R, et al. Diabetes mellitus in midlife and the risk
of dementia three decades later. Neurology 2004;63:1902–7.
[12] Xu WL, Qiu CX, Wahlin A, Winblad B, Fratiglioni L. Diabetes
mellitus and risk of dementia in the Kungsholmen project: a 6-year
follow-up study. Neurology 2004;63:1181– 6.
[13] Yaffe K, Blackwell T, Kanaya AM, Davidowitz N, Barrett-Connor E,
Krueger K. Diabetes, impaired fasting glucose, and development of
cognitive impairment in older women. Neurology 2004;63:658 – 63.
[14] Ivan CS, Seshadri S, Beiser A, Au R, Kase CS, Kelly-Hayes M, et al.
Dementia after stroke: The Framingham Study. Stroke 2004;35:1264 – 8.
[15] Honig LS, Tang M-X, Albert S, Costa R, Luchsinger J, Manley J, et
al. Stroke and the risk of Alzheimer disease. Arch Neurol 2003;60:
1707–12.
[16] Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ,
Breteler MMB. Silent brain infarcts and the risk of dementia and
cognitive decline. N Engl J Med 2003;348:1215–22.
[17] Jorm AF, Scott R, Cullen JS, MacKinnon AJ. Performance of the
Informant Questionnaire on Cognitive Decline in the Elderly
(IQCODE) as a screening test for dementia. Psychol Med 1991;21:
785–90.
[18] Henon H, Durieu I, Guerouaou D, Lebert F, Pasquier F, Leys D.
Poststroke dementia: incidence and relationship to prestroke cogni-
tive decline. Neurology 2001;57:1216 –22.
[19] Notkola I-L, Sulkava R, Pekkanen J, Erkinjuntti T, Ehnholm C,
Kivinen P, et al. Serum total cholesterol, apolipoprotein E ␧4 allele,
and Alzheimer’s disease. Neuroepidemiology 1998;17:14 –20.
[20] Ravaglia G, Forti P, Maioli F, Martelli M, Servadei L, Brunetti N,
et al. Homocysteine and folate as risk factors for dementia and
Alzheimer disease. Am J Clin Nutr 2005;82:636 – 43.
[21] Kalmijn S, Mehta KM, Pols HAP, Hofman A, Drexhage HA, Breteler
MMB. Subclinical hyperthyroidism and the risk of dementia: the
Rotterdam Study. Clin Epidemiol 2000;53:733–7.
[22] Geerlings MI, Launer LJ, de Jong FH, Ruitenberg A, Stijnen T, van
Swieten JC, et al. Endogenous estradiol and risk of dementia in
women and men: the Rotterdam Study. Ann Neurol 2003;53:607–15.
[23] Moffat SD, Zonderman AB, Metter EJ, Kawas C, Blackman MR,
Harman SM, et al. Free testosterone and risk for Alzheimer disease in
older men. Neurology 2004;62:188 –93.
[24] Fuhrer R, Cufouil C, Dartigues JF, for the Paquid Study. Exploring
sex differences in the relationship between depressive symptoms and
dementia incidence: prospective results from the PAQUID Study.
J Am Geriatr Soc 2003;51:1055– 63.
[25] Hébert R, Lindsay J, Verreault R, Rockwood K, Hill G, Dubois M-F.
Vascular dementia: incidence and risk factors in the Canadian Study
of Health and Aging. Stroke 2000;31:1487–93.
[26] Jorm AF. History of depression as a risk factor for dementia: an
updated review. Aust N Z J Psychiatry 2001;35:776 – 81.
[27] Kalmijn S, Launer LJ, Ott A, Witteman JCM, Hofman A, Breteler
MMB. Dietary fat intake and the risk of incident dementia in the
Rotterdam study. Ann Neurol 1997;42:776 – 82.
[28] Larrieu S, Letenneur L, Helmer C, Dartigues JF, Barberger-Gateau P.
Nutritional factors and risk of incident dementia in the PAQUID
longitudinal cohort. J Nutr Health Aging 2004;8:150 – 4.
[29] Helmer C, Peuchant E, Letenneur L, Bourdel-Marchasson I, Larrieu
S, Dartiques JF, et al. Association between antioxidant nutritional
indicators and the incidence of dementia: results from the PAQUID
prospective cohort study. Eur J Clin Nutr 2003;57:1555– 61.
[30] Orgogozo J-M, Dartigues J-F, Lafont S, Letenneur L, Commenges D,
Salamon R, et al. Wine consumption and dementia in the elderly: a
prospective community study in the Bordeaux area. Rev Neurol
1997;153:3, 185–92.
[31] Lindsay J, Laurin D, Verreault R, Hébert R, Helliwell B, Hill GB,
et al. Risk factors for Alzheimer’s disease: a prospective analysis
from the Canadian Study of Health and Aging. Am J Epidemiol
2002;156:445–53.
[32] Laurin D, Verreault R, Lindsay J, MacPherson K, Rockwood K.
Physical activity and risk of cognitive impairment and dementia in
elderly persons. Arch Neurol 2001;58:498 –504.
[33] Podewills LJ, Guallar E, Kuller LH, Fried LP, Lopez OL, Carlson M,
et al. Physical activity, APOE genotype and dementia risk: findings
from the Cardiovascular Health Cognition Study. Am J Epidemiol
2005;161:639 –51.
[34] Wang H-X, Karp A, Winblad B, Fratiglioni L. Late-life engagement
in social and leisure activities is associated with a decreased risk of
dementia: a longitudinal study from the Kungsholmen project. Am J
Epidemiol 2002;155:1081–7.
[35] Scarmeas N, Levy G, Tang M-X, Manly J, Stern Y. Influence of
leisure activity on the incidence of Alzheimer’s disease. Neurology
2001;57:2236 – 42.
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 341–347
[36] Tyas LS, Manfreda J, Strain LA, Montgomery PR. Risk factors for
Alzheimer’s disease: a population-based, longitudinal study in Mani-
toba, Canada. Int J Epidemiol 2001;30:590 –7.
[37] Launer LJ, Andersen K, Dewey ME, Letenneur L, Ott A, Amaducci
LA, et al. Rates and risk factors for dementia and Alzheimer’s
disease: results from EURODEM pooled analyses. Neurology 1999;
52:78 – 84.
[38] Plassman BL, Havlik RJ, Steffens DC, Helms MJ, Newman TN,
Drosdick D, et al. Documented head injury in early adulthood and
risk of Alzheimer’s disease and other dementias. Neurology 2000;55:
1158 – 66.
[39] Rockwood K, Kirkland S, Hogan DB, MacKnight C, Merry H,
Verreault R, et al. Use of lipid-lowering agents, indication bias, and
the risk of dementia in community-dwelling elderly people. Arch
Neurol 2002;59:223–7.
[40] Zandi PP, Sparks L, Khachaturian AS, Tschanz J, Norton M, Stein-
berg M, et al. Do statins reduce risk of incident dementia and Alz-
heimer disease? Arch Gen Psychiatry 2005;62:217–24.
[41] Li G, Higdon R, Kukull W, Peskind E, Van Valen Moore K,
Tsuang D, et al. Statin therapy and risk of dementia in the elderly:
a community-based prospective cohort study. Neurology 2004;63:
1624 – 8.
[42] Zandi PP, Anthony JC, Hayden KM, Mehta K, Mayer L, Breitner
JCS, et al. Reduced incidence of AD with NSAID but not H2 receptor
antagonists: the Cache County Study. Neurology 2002;59:880 – 6.
[43] Etminan M, Gill S, Samii A. Effect of non-steroidal anti-inflamma-
tory drugs on risk of Alzheimer’s disease: systematic review and
meta-analysis of observational studies. BMJ 2003;327:128.
[44] de Craen AJM, Gussekloo J, Vrijsen B, Westendorp RGJ. Meta-
analysis of nonsteroidal antiinflmmatory drug use and risk of demen-
tia. Am J Epidemiol 2005;161:114 –20.
[45] Lagnaoui R, Bégaud B, Moore N, Chaslerie A, Fourrier A, Letenneur
L, et al. Benzodiazepine use and risk of dementia: a nested case-
control study. J Clin Epidemiol 2002;55:314 – 8.
[46] Verdoux H, Lagnaoui R, Begaud B. Is benzodiazepine use a risk
factor for cognitive decline and dementia? a literature review of
epidemiological studies. Psychol Med 2005;35:307–15.
[47] Verreault R, Laurin D, Lindsay J, De Serres G. Past exposure to
vaccines and subsequent risk of Alzheimer’s disease. CMAJ 2001;
165:1495– 8.
[48] Kukull WA, Higdon R, Bowen JD, McCormich WC, Teri L, Schel-
lenberg GD, et al. Dementia and Alzheimer disease incidence: a
prospective cohort study. Arch Neurol 2002;59:1737– 46.
347
[49] Qiu C, Karp A, von Strauss E, Winblad B, Fratiglioni L, Bellander T.
Lifetime principal occupation and risk of Alzheimer’s disease in the
Kungsholmen Project. Am J Ind Med 2003;43:204 –11.
Appendix 1: Search strategies
1966 to December 2005
RF & Dementia or AD AND
((Relative [title/abstract] AND risk* [title/abstract]) OR
(relative risk [Text Word]) OR risks [Text word] OR Cohort
studies [MeSH: no exp] OR (cohort [Title/Abstract] AND
stud* [Title/Abstract]) Limits – Abstracts, Review, Human
1966 to December 2005
((“risk factors” [MeSH Terms] OR risk factors {Text
Word]) AND (“dementia” [MeSH Terms] OR dementia
[Text Word]) OR (((“alzheimer disease” [TIAB] NOT Med-
line [SB]) OR “alzheimer disease” [MeSH Terms] OR alz-
heimer [Text Word]) AND s [All Fields] AND (“disease”
[MeSH Terms] OR disease [Text Word])) NOT (“therapy”
[Subheading] OR (“therapeutics” [TIAB] NOT Medline
[SB]) OR “Therapeutics” [MeSH Terms] OR therapy [Text
Word])) AND systematic [sb] AND ((“humans” [TIAB]
NOT Medline [SB]) or “humans” [MeSH Terms] OR Hu-
man [Text Word])
1996 to 2006
(Prevention of Alzheimer*) AND systematic [sb] OR
(prevention of dementia) AND systematic [sb] OR (“risk
factors AND (dementia OR alzheimer*)) AND systematic
[sb] All fields, publication date from 1996 to 2006, core
clinical journals OR (”risk factors AND (dementia OR Alz-
heimer*)) AND systematic [sb]
 Alzheimer’s & Dementia 3 (2007) 348 –354

Primary prevention of dementia

Christopher Pattersona,*, John Feightnerb, Angeles Garciac, Chris MacKnightd
a
Division of Geriatric Medicine, Department of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada
b
Department of Family Medicine, Faculty of Medicine, University of Western Ontario, London, Ontario, Canada
c
Department of Medicine (Geriatrics), Queen’s University, Kingston, Ontario, Canada
d
Division of Geriatric Medicine, Dalhousie University, Halifax, Nova Scotia, Canada

Abstract The purpose of this article is to recommend strategies to practicing physicians for the prevention
of dementia in people without preexisting cognitive deficits.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Alzheimer’s disease; Vascular dementia; Primary prevention

1. Background
In preparation for this article, a systematic evidence
review was undertaken to determine the risk factors for
all-cause dementia (ACD), Alzheimer’s disease (AD), or
vascular dementia (VaD) [1]. This review involved litera-
ture from 1996 until December 2005; 1996 was chosen as
the initial date because literature up until 1998 had been
reviewed in preparation for an article by Black et al [2] on
“Preventing Dementia” for the 2nd Canadian Consensus
Conference on Dementia. The new review addresses poten-
tially modifiable risk factors. Genetic factors are considered
in another article in this series [3].
Risk factors that are potentially remediable are listed in
Table 1 and form the background for this article on primary
prevention. The relative risk (RR) for ACD, AD, or VaD for
each potentially remediable risk factor is included in Table 1.
The table shows statistically significant RR values without
95% confidence intervals, which are detailed in the risk
factor article [1]. The presence of a potentially remediable
risk factor does not imply that modification of that risk
factor will influence the subsequent development of demen-
tia. Only randomized controlled trials (RCTs) are consid-
ered to provide strong evidence for efficacy of risk factor
modification (Tables 2 and 3).
*Corresponding author. Tel.: (905) 521-7931; Fax: (905) 521-7905.
E-mail address: pattec@hhsc.ca
2. Methods
Studies addressing prevention of dementia were identi-
fied from three sources. First, from the original search for
risk factors performed in December 2005 (total of 3,424
articles with abstracts); second, from a targeted search for
RCTs and systematic evidence reviews in February 2006
(see Appendix 1 for search strategy); and third, articles
gleaned from bibliographies of retrieved articles and authors
files.
Quality assessment of RCTs was performed with criteria
of the Canadian Task Force on Preventive Health Care and
resulted in designation of good (all criteria fulfilled); fair
(minor flaws only, no fatal flaw), and poor (fatal flaw or
multiple minor flaws) [4]. For systematic reviews, criteria
described by Hunt and McKibbon [5] were used for quality
assessment.
Where conflicting evidence exists, priority is given to
studies (1) with higher levels of evidence, (2) with better
quality, and (3) most contemporary. For example, in the
case of estrogens, where several systematic evidence re-
views of epidemiologic and case-controlled trials suggested
that estrogens were protective against dementia (Level 2
evidence), conclusions are superseded by a large RCT in-
dicating an increased risk of dementia (Level 1 evidence)
[6]. We found very few RCTs of interventions to prevent
dementia on which to base firm conclusions, and maintain-
ing this rigorous standard of evidence leads to a large
number of “C” grade recommendations, where there is in-
sufficient or contradictory evidence of efficacy.

1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.005
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354 349

Table 1
Potentially modifiable risk factors for dementia: RR is shown except where alternative risk measures are used: hazards ratio (HR); odds ratio (OR)
Illnesses and Conditions ACD AD VaD

Hypertension 1.6 1.5 to 2.3 (OR) 2.05

Diabetes mellitus 2.83 2.03 2.6 (HR)
Stroke 2.4 1.83 Required for diagnosis
Silent infarctions on neuroimaging 2.17 to 2.26 (HR) 1.83
Reduced thyroid-stimulating hormone 3.5
Hypercholesterolemia 3.1 2.1 to 3.1
Hyperhomocysteinemia 2.08 2.11
Depression 1.87 4.2 (men) 2.41 (OR)
Lifestyle
High fat diet 2.4
High omega 3 fatty acid./fish in diet 0.4 0.3 0.45 (OR)
Moderate consumption of wine 0.19 (OR) to 0.56 0.53
Activity, highest level, physical 0.58 to 0.63 0.5 to 0.69 0.46 (OR)
Activity, higher level, mental 0.59
Smoking 2.2 1.99 to 2.3
Occupation (manual work) 1.6
Exposure to toxins 4.35 2.05 (OR)
Medications
3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) 0.33 to 1.19 (HR) 0.82 to 1.19 (HR)
All lipid-lowering agents 0.26 (OR)
NSAIDs 0.51 0.42
Lowest tertile serum vitamin E level 2.54 2.10
Benzodiazepines 1.7 to 2.3 (OR)
Vaccinations 0.4
Antihypertensives 0.93 0.89
Miscellaneous
Head injuries (serious) 2.39 to 4.48 (HR) 2.32 to 4.51 (HR)
Education (⬎15 y vs ⬍12 y) 0.64 0.48

Abbreviations: ACD, any cause dementia; AD, Alzheimer’s disease; VaD, vascular dementia.

3. Hypertension though it is tempting to assume that a reduced incidence of

The association between elevated systolic blood pressure
and stroke is well-established. However, the relationship
between hypertension and dementia is somewhat less clear.
Both elevated systolic blood pressure and depressed dia-
stolic blood pressure in middle age are positively correlated
with the later development of dementia [1]. In the years
preceding the onset of dementia, systolic blood pressure
might fall [1]. Furthermore, not all epidemiologic studies
confirm the positive association between elevated systolic
pressure and subsequent development of dementia. Al-
stroke resulting from antihypertensive treatment will trans-
late into a lower risk of dementia, this is not borne out by
evidence from all RCTs.
In 1998 the results of the Systolic Hypertension in Eu-
rope (SYST-Eur) trial were published [7]. Three thousand
one hundred sixty-two patients older than the age of 60
years with seated systolic blood pressures of 160 to 219 mm
Hg and diastolic blood pressures below 95 mm Hg were
randomized to receive nitrendipine 10 to 40 mg/day with the
addition of enalapril and/or hydrochlorothiazide titrated to

Table 2
Quality assessment for individual RCTs
Reference Adequacy of Comparison of Placebo Groups Treated All Subjects Dementia Outcome Analysis Quality
Randomization and Treatment Groups Equally Accounted for Assessed Blindly Appropriate
at Conclusion

Forette et al [7], 1998 Yes Equal Yes Yes Yes Yes Good
PROGRESS [8], 2003 Yes Equal Yes Yes Yes Yes Good
Shepherd et al [14], 2002 Yes Equal Yes Yes Yes Yes Good
Shumaker et al [6], 2004 Yes More in placebo group Yes Yes Yes Yes Good
aged ⬎75 y; more in
estrogen group were
hypertensive
350 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354

Table 3
Results of individual RCTs
Reference Population N Intervention Duration Results
(years)

Forette et al [7], 1998 Men and women ⬎60 y; seated
systolic blood pressure 160
to 219 mm Hg; diastolic
blood pressure ⬍95 mm Hg;
dementia excluded at
baseline
PROGRESS [8], 2003 Men and women with transient
ischemic attack or stroke
within 5 y
Shepherd et al [14], 2002 Men and women aged 70 to
82 y with increased risk or
preexisting vascular disease
Shumaker et al [6], 2004 Post menopausal women 65 to
75 y prior hysterectomy
Post menopausal women 65 to
75 y no prior hysterectomy
Pooled
2418 1238 nitrendipine 10–40 mg/day ⫾ 2.0 Treated group had lower risk of
enalapril 5–20 mg/day ⫾ ACD; 7.7–3.8/1000 patient-y;
hydrochlorothiazide 12.5–25 mg/ 50% reduction; P ⫽ .05
day 1180 placebo
6105 3051 perindopril ⫾ indapamide 3.9 12% lower risk of ACD (⫺8%
3054 placebo to 28%); 34% lower risk of
dementia with recurrent
stroke (3% to 55%)
5804 2888 prescribed pravastatin 40 mg/ 3.2 No significant difference in
day 2912 placebo cognitive function between
treatment and placebo groups
2947 1464 estrogen alone 5.2 HR for probable ACD
1483 placebo 1.4 (0.83, 2.66)
4532 2229 estrogen & progestin 4.0 HR for probable ACD
2.05 (1.21, 3.48)
7479 Pooled 4.5 HR for probable ACD
1.76 (1.19, 2.60)

reduce the systolic blood pressure by at least 20 mm Hg to
reach a value below 150 mm Hg or to matching placebo.
These individuals were free of dementia at baseline and
were subsequently monitored with the Mini-Mental State
Examination (MMSE), with further diagnostic tests to reach
a definitive diagnosis of dementia by using the Diagnostic
and Statistical Manual [of Mental Disorders], Third Edi-
tion, Revised criteria if the score fell below 23. After a
median follow-up of 2 years, the incidence of dementia was
reduced by 50% from 7.7 to 3.8 cases per 1,000 patient
years. The RR of dementia was 0.47 (95% confidence in-
terval [CI], 0.28 to 0.78), and interestingly, most were
diagnosed with probable AD. In absolute terms, the number
of older hypertensive patients needed to treat for 5 years to
prevent one case of dementia is approximately 53 [(NNT(5) ⫽
53]. Five other RCTs of blood pressure lowering have exam-
ined dementia as an outcome. Only one other, the
PROGRESS Study, showed a RR reduction for dementia of
0.89 (95% CI, 0.74 to 1.07), which was not statistically
significant (P ⫽ .2) [8].
A meta-analysis of four trials (PROGRESS, SCOPE,
SHEP, SySt-Eur) concluded that the RR reduction for demen-
tia was 0.80 (95% CI, 0.63 to 1.02), just missing statistical
significance [9]. It should be noted that the PROGRESS and
HOPE trials enrolled patients with established cardiovascular
and cerebrovascular disease, so they cannot be considered as
truly primary prevention studies. A further systematic evidence
review protocol has been registered with the Cochrane Data-
base of Systematic Reviews by members of the Cochrane
Dementia and Cognitive Improvement Group [10]. The date of
completion of this review is uncertain.
Although there are compelling reasons to treat systolic
hypertension in older individuals (eg, reducing the risk of
subsequent stroke), the influence on dementia reduction,
although encouraging, is not definitive.
4. Diabetes mellitus
There is evidence from epidemiologic studies that the
presence of diabetes is associated with subsequent cognitive
decline, specifically the development of ACD, AD, and
VaD [1]. Attempts to determine whether treatment of type 2
diabetes mellitus affects the development of cognitive im-
pairment or dementia have been addressed in a systematic
review performed in the Cochrane Database [11]. Five
RCTs were identified in which different treatments for type
2 diabetes were compared, and in which measures of cog-
nitive function were made at entry and after treatment. In
one trial the comparison of cognitive function was not
blinded, in three studies cognitive function was not appro-
priately evaluated, and the fifth did not report on baseline
cognitive function. The Cochrane analysis was that “there is
no convincing evidence relating type or intensity of diabetic
treatment to the prevention or management of cognitive
impairment in type 2 diabetes” [11].
5. Stroke
Stroke is associated with increased risk of ACD, AD, and
VaD. Although a comprehensive discussion of the literature
regarding stroke prevention is beyond the scope of this
article, suffice it to say that there is an abundance of liter-
ature addressing the topic of primary prevention of stroke.
The consensus statement from the National Stroke Associ-
ation is a systematic review of fair quality [12]. These
guidelines offer recommendations for treatment of hyper-
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354
tension; acetylsalicylic acid (ASA) and statins after myo-
cardial infarction; warfarin or ASA for nonvalvular atrial
fibrillation; consideration of carotid endarterectomy for
asymptomatic stenosis greater than 60%; and attention to
lifestyle factors (diet, smoking cessation, regular exercise,
moderate alcohol use, and a healthy diet). Notwithstanding
the efficacy of these interventions to reduce the risk of
subsequent stroke, a known risk factor for dementia, it
remains to be seen whether such measures will influence the
incidence of dementia.
6. Hyperlipidemia
Several epidemiologic studies have established a rela-
tionship between hyperlipidemia and subsequent develop-
ment of dementia. A systematic evidence review of the topic
of statins for the prevention of AD was performed for the
Cochrane Database of Systematic Reviews [13]. The au-
thors concluded that in the absence of RCTs there was “no
good evidence to recommend statins for reducing the risk of
AD”. Subsequent to this review, an RCT (PROSPER) was
published in 2002 [14]. In this study, 5,804 men and women
with a history or risk factors for vascular disease were
randomized to receive either pravastatin 40 mg per day or
placebo. After follow-up of 3.2 years, various vascular end
points were considered, but there was no significant effect
on cognitive function or disability. However, this was not a
primary prevention trial. Not withstanding the epidemio-
logic evidence, there is at present no clear rationale for the
prescription of statin drugs to lower the risk of AD.
7. Hyperhomocysteinemia
Epidemiologic evidence links the presence of hyperho-
mocysteinemia to an increased risk of ACD and AD. The
literature search failed to identify any randomized con-
trolled treatment trials aimed at lowering serum homocys-
teine levels with an outcome of dementia or AD.
8. Atrial fibrillation
Atrial fibrillation is recognized as a risk factor for stroke
and has been identified in epidemiologic studies as a risk
factor for dementia. Although numerous studies indicate
that treatment with anticoagulants or ASA reduces the risk
of stroke, the literature search did not identify any studies
with dementia as an outcome.
In a meta-analysis of 16 RCTs, Hart et al [15] concluded
that adjusted dose warfarin reduced the risk of stroke by
62% (range, 48% to 72%), whereas ASA (6 trials) reduced
the risk by 22% (range, 2% to 38%). Another meta-analysis
of three trials of ASA for nonvalvular atrial fibrillation
concluded that the risk of all stroke was reduced by 30%
(odds ratio [OR], 0.7; 95% CI, 0.47 to 1.07); ischemic
stroke (OR, 0.70; 95% CI, 0.46 to 1.07), and disabling fatal
351
stroke (OR, 0.86; 95% CI, 0.50 to 1.49). Only the combined
outcome of stroke, myocardial infarction, or vascular death
was significantly reduced (OR, 0.71; 95% CI, 0.51 to 0.97)
[16]. Although it is possible that antithrombotic treatment
for atrial fibrillation might result in a lower risk of dementia,
it remains unproven.
9. Lifestyle factors
9.1. Diet
Both total fat and reduced levels of omega 3 fatty acids
have been linked to increased risk of dementia in epidemi-
ologic studies [1]. Two systematic evidence reviews have
addressed the issues of omega 3 fatty acids for the preven-
tion of dementia [17,18]. Both concluded that there were no
RCTs on which to base a recommendation. Similarly, no
RCTs were identified by the search strategy used here for other
dietary advice in relation to the prevention of dementia.
9.2. Alcohol
Epidemiologic evidence has established that a moderate
consumption of wine (250 to 500 mL/day) reduces the risk
of dementia [1]. However, the literature search did not
identify any RCTs addressing the topic of any type of
alcohol for the prevention of dementia. Advice about alco-
hol consumption should be tempered by the known risks of
excessive use in terms of neurologic and other organ dam-
age as well as increased risk of injury from intoxication.
9.3. Activity, physical or mental
Although higher levels of physical activity have been
linked to lower subsequent rates of dementia, the search
failed to identify any RCTs of exercise with dementia as
an outcome. Similarly, no RCTs of mental activity were
identified.
9.4. Tobacco
Although case-controlled studies had suggested that to-
bacco smoking reduced the risk of dementia, longitudinal
cohort studies have identified a significant increased risk of
ACD and AD for tobacco smokers [1]. Even in the absence
of RCTs, this epidemiologic evidence further adds to the
many other reasons to discourage this habit.
9.5. Specific medications
9.5.1. Nonsteroidal anti-inflammatory drugs
Several recent systematic evidence reviews have ad-
dressed the topic of nonsteroidal anti-inflammatory drugs
(NSAIDs) on subsequent risk of AD and other dementias
[19,20]. Although two of these [19,20] concluded that those
consuming NSAIDs had a lower risk of ACD and AD, the
third [21] did not. There have been no prospective RCTs of
NSAIDs in this context. Not withstanding the epidemio-
logic evidence, there is insufficient evidence at present to
352 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354
recommend for or against prescribing any NSAID for the
specific intention of reducing the risk of ACD or AD.
9.5.2. Vitamin supplementation
A large prospective observational study has suggested
that combined use of vitamin E (400 units daily) and vita-
min C (500 mg daily) for at least 3 years was associated
with a reduction in the incidence of AD [23]. Other studies
have shown no such reduction, and high doses of vitamin E
(ⱖ400 units daily) have been associated with increased
mortality from all causes. A recent systematic evidence
review concluded that in the absence of prospective RCTs,
there is no justification for consuming antioxidant vitamins
C and E for the intention of reducing the subsequent risk of
AD [22].
9.5.3. Estrogens
Systematic reviews of case-control and epidemiologic
evidence have suggested that consumption of estrogens
alone or in conjunction with progestins by perimenopausal
or postmenopausal women has been associated with a re-
duced risk of AD and ACD. However, the situation changed
dramatically with publication of the results from the Women’s
Health Initiative Memory Study, a large RCT of estrogen and
progestin replacement in more than 7,000 women aged 65 to
79 years [6]. Women assigned to estrogens alone had an
overall hazard ratio (HR) for probable dementia of 1.49
(95% CI, 0.83 to 2.66) compared with placebo, whereas
those assigned to estrogen plus progestin showed a HR of
2.05 (95% CI, 1.21 to 3.48). Importantly, after participants
with baseline modified MMSE (3MS) scores at or below the
education-adjusted screening cut point were excluded, the
HR was 1.77 (95% CI, 0.74 to 4.23; P ⫽ .20) in the estrogen
alone group and 2.19 (95% CI, 1.25 to 3.84) in the pooled
group [6]. The daily doses used were 0.625 mg of conju-
gated equine estrogen and 2.5 mg of medroxyprogesterone
acetate. In absolute terms, treatment of about 111 women
for 5 years would result in one additional case of dementia.
Thus, there is no justification for prescribing estrogens or
estrogen plus progestin to postmenopausal women with the
intention of reducing the risk of dementia.
10. Miscellaneous factors
10.1. Head injury
Although there has been some controversy from the
literature concerning the contribution of head injury to sub-
sequent development of dementia, it is now considered
well-established that serious head injury with loss of con-
sciousness increases the risk of subsequent AD [1].
Although there have been no RCTs on prevention of
dementia by reduction of head injuries, prudence would
dictate that measures to reduce serious head injuries (eg,
wearing of helmets during contact sports, bicycling, skate-
boarding) including legislation on wearing helmets should
be considered.
10.2. Education
The relationship between education and subsequent de-
velopment of dementia might be explained by innate ability
(those possessing more innate intelligence might progress
further in formal education and might be less likely to
develop dementia), or that the acquisition of more formal
education somehow promotes improved resilience to factors
that predispose to dementia, or that some other factor might
have a common mechanism affecting the ability to pursue
formal education yet reducing the risk of dementia. Regard-
less of the mechanism and in spite of the appeal of advocacy
for increasing standards of education, there is no evidence
that this strategy would lower the risk of dementia.
10.3. Occupational exposure to toxins
Longitudinal cohort studies have established a relation-
ship between exposure to environmental toxins and subse-
quent development of dementia, specifically pesticides and
fumigants [1]. Although the absence of RCTs prevents a
firm recommendation, it would seem prudent to minimize
exposure to such chemicals by the appropriate use of pro-
tective clothing and respiratory protection.
11. Discussion
A large amount of epidemiologic evidence is accumulat-
ing for many risk factors for the subsequent development of
ACD, AD, and VaD. However, when these risk factors are
critically evaluated in the context of studies aimed at reduc-
ing the incidence of dementia, the quantity of Level 1
(evidence from at least one well-designed RCT) is pitifully
small. Very few recommendations can be based on highest
quality evidence. Furthermore, it is unlikely that RCTs will
ever be conducted on many of the risk factors identified in
this review; thus recommendations must necessarily be
based on less stringent types of evidence. The research
community is encouraged to focus on studies to determine
whether physical or mental activity can reduce the risk of
subsequent dementia; whether treatments aimed at hyper-
lipidemia, hyperhomocysteinemia, and glycemic control are
beneficial; and whether any type of NSAID, combination of
vitamins, alternatives to estrogens (eg, selective estrogen
receptor modifiers), or lipid-lowering agents might be ef-
fective in primary prevention of dementia.
12. Recommendations for the primary prevention of
dementia (ACD, AD, VaD)
1. There is good evidence to treat systolic hyperten-
sion (⬎160 mm Hg) in older individuals. In ad-
dition to reducing the risk of stroke, the incidence
 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354
of dementia might be reduced. The target blood
pressure should be 140 mm Hg or less (Grade A,
Level 1).
2. Although ASA and statin medications after myocar-
dial infarction, antithrombotic treatment for nonval-
vular atrial fibrillation, and correction of carotid ar-
tery stenosis ⬎60% have been shown to reduce the
risk of stroke, there is insufficient evidence to rec-
ommend for or against these measures for the spe-
cific purpose of primary prevention of dementia
(Grade C, Level 1).
3. Although there are many reasons for treating type 2
diabetes, hyperlipidemia, and hyperhomocysteine-
mia, there is insufficient evidence to recommend
treatment of these conditions for the specific pur-
pose of reducing the risk of dementia (Grade C,
Level 2).
4. There is insufficient evidence to recommend for or
against the prescription of NSAIDs for the sole purpose
of reducing the risk of dementia (Grade C, Level 2).
5. There is good evidence to avoid the use of estrogens
alone or together with progestins for the sole pur-
pose of reducing the risk of dementia (Grade E,
Level 1).
6. Although there is insufficient evidence to make a
firm recommendation, physicians might advocate for
strategies, including legislation, to reduce the risk of
serious head injuries (Grade C, Level 2).
7. Although there is insufficient evidence to make a
firm recommendation, physicians might advise their
patients about, and advocate for, appropriate pro-
tective clothing during administration of pesti-
cides, fumigants, fertilizers, and defoliants (Grade
C, Level 2).
8. There is insufficient evidence to recommend for or
against supplementation with vitamins E or C for the
prevention of dementia (Grade C, Level 2). High
dose vitamin E (ⱖ400 units/day) is associated with
excess mortality and should not be recommended
(Grade E, Level 1).
9. Although recommendations might be made on other
grounds (such as part of a healthy lifestyle), there is
insufficient evidence to recommend for or against
higher levels of physical or mental activity for the
specific purpose of reducing the incidence of demen-
tia (Grade C, Level 2).
10. Although there is insufficient evidence to make a
firm recommendation for the primary prevention of
dementia, physicians might advocate for appropriate
levels of education and strategies to retain students
in appropriate learning environments (Grade C,
Level 2).
11. Although there is insufficient evidence to make a
firm recommendation for the primary prevention of
dementia, physicians might choose to advise their
353
patients about the potential advantages of increased
consumption of fish, reduced consumption of dietary
fat, and moderate consumption of wine (Grade C,
Level 2).
Author Disclosures
Angeles Garcia has received support from Janssen-Ortho
(Consultant, Advisor), Pfizer (Consultant, Advisor), Novar-
tis (Consultant, Advisor), and Lundbeck (Consultant, Advi-
sor).
Chris MacKnight has received support from Alzheimer
Society of Nova Scotia (Board Member), Janssen Ortho Inc.
(Speaker, Trial Investigator), Pfizer Canada (Speaker, Trial
Investigator), Novartis (Speaker, Trial Investigator), Lund-
beck Canada (Speaker, Trial Investigator), Voyager Phar-
maceuticals (Trial Investigator), Myriad (Trial Investiga-
tor), and Neurochem (Trial Investigator).
Acknowledgments
Chris MacKnight is supported from a CIHR New Inves-
tigator Award.
During preparation of this article, Christopher Patterson
received financial support from the Institute of Advanced
Studies, University of Bologna, Bologna, Italy.
References
[1] Patterson C, Feightner JW, Garcia A, MacKnight CR. General risk
factors for dementia. a systematic evidence review. Alzheimers De-
ment 2007;3.
[2] Black SE, Patterson C, Feightner J. Preventing dementia. Can J Neu-
rol Sci 2001;28(Suppl 1):S56 – 66.
[3] Hsuing G-YR, Sadovnick AD. Genetics and dementia. risk factors,
diagnosis and management. Alzheimers Dement 2007;3.
[4] Woolf SH, Battista RN, Anderson GM, et al. Assessing the clinical
effectiveness of preventive manoeuvres: analytic principles and sys-
tematic methods in reviewing evidence and developing clinical prac-
tice recommendations. J Clin Epidemiol 1990;43:891–905.
[5] Hunt DL, McKibbon KA. Locating and appraising systematic re-
views. In: Mulrow C, Cook D, eds. Systematic reviews: synthesis of
best evidence for health care decisions. Philadelphia: American Col-
lege of Physicians; 1998. p. 13–22.
[6] Shumaker SA, Legault C, Kuller L, Rapp SR, Thal L, Lane DS, et al.
Conjugated equine estrogens and incidence of probable dementia and
mild cognitive impairment in postmenopausal women: Women’s
Health Initiative Memory Study. JAMA 2004;291:2947–58.
[7] Forette F, Seux M-L, Staessen JA, Thijs L, Birkenhäger WH, Babar-
skiene MR, et al. Prevention of dementia in randomized double-blind
placebo-controlled Systolic Hypertension in Europe (Syst-Eur) trial.
Lancet 1998;352:1347–51.
[8] The PROGRESS Collaborative Group. Effects of blood pressure
lowering with perindopril and indapamide therapy on dementia and
cognitive decline in patients with cerebrovascular disease. Arch In-
tern Med 2003;163:1069 –75.
[9] Feigin V, Ratnasabapathy Y, Anderson C. Does blood pressure low-
ering treatment prevent dementia or cognitive decline in patients with
cardiovascular and cerebrovascular disease? J Neurol Sci 2005;229 –
230:151–5.
354 C. Patterson et al. / Alzheimer’s & Dementia 3 (2007) 348 –354
[10] McGuinness B, Todd S, Passmore P, Bullock R. The effects of blood
pressure lowering on development of cognitive impairment and de-
mentia in patients without apparent prior cerebrovascular disease.
Cochrane Database Syst Rev 2006;(1).
[11] Areosa Sastre A, Grimley Evans J. Effect of the treatment of Type II
diabetes mellitus on the development of cognitive impairment and
dementia. Cochrane Database Syst Rev 2006;1.
[12] Gorelick PB, Sacco RL, Smith DB, et al. Prevention of a first stroke:
a review of guidelines and a multidisciplinary consensus statement
from the National Stroke Association. JAMA 1999;281:1112–20.
[13] Scott HD, Laake K. Statins for the prevention of Alzheimer’s disease.
Cochrane Database Syst Rev 2006;1: accession number: 00075320-
100000000-02141.
[14] Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe
SM, et al. Pravastatin in elderly individuals at risk of vascular disease
(PROSPER): a randomized controlled trial. Lancet 2002;360:1623–30.
[15] Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic
therapy to prevent stroke in patients with atrial fibrillation: a meta-
analysis. Ann Intern Med 1999;131:492–501.
[16] Aguilar MI, Hart R. Oral anticoagulants for preventing stroke in
patients with non-valvular atrial fibrillation and no previous history of
stroke or transient ischemic attacks. Cochrane Database Syst Rev
2006;1: accession number: 00075320-100000000-01342.
[17] Lim W, Gammack J, Van Niekerk J, Dangour A. Omega 3 fatty acid
for the prevention of dementia. Cochrane Database Syst Rev 2006:
CD005379.
[18] MacLean CH, Issa AM, Newberry SJ, Mojica WA, Morton SC,
Garland RH, et al. Effects of omega-3 fatty acids on cognitive
function with aging, dementia, and neurological diseases. Evidence
Report/Technology Assessment; 2005; number 114:13– 6.
[19] Etminan M, Gill S, Samii A. Effect of non-steroidal anti-inflammatory
drugs on risk of Alzheimer’s disease: systematic review and meta-
analysis of observational studies. BMJ 2003;327:128.
[20] Szekely CA, Thorne JE, Zandi PP, Ek M, Messias E, Breitner JC,
et al. Nonsteroidal anti-inflammatory drugs for the prevention of
Alzheimer’s disease: a systematic review. Neuroepidemiology
2004;23:159 – 69.
[21] de Craen AJM, Gussekloo J, Vrijsen B, Westendorp RGJ. Meta-
analysis of nonsteroidal antiinflmmatory drug use and risk of demen-
tia. Am J Epidemiol 2005;161:114 –20.
[22] Boothby LA, Doering PL. Vitamin C and vitamin E for Alzheimer’s
disease. Ann Pharmacother 2005;39:2073– 80.
[23] Zandi PP, Anthony JC, Khachaturian AS, Stone SV, Gustafson D,
Tschanz JT, et al. Reduced risk of Alzheimer disease in users of
antioxidant vitamin supplements: the Cache County Study. Arch
Neurol 2004;61:82– 8.
Appendix 1
Search strategies for systematic reviews of RCTs on
prevention of dementia or Alzheimer’s disease
Medline: used clinical queries function of Medline to
search for systematic reviews (dementia AND prevent*)
AND systematic [sb] 126 articles; (Alzheimer* AND pre-
vent* AND systematic (sb) 62 articles; (Alzheimer* OR
dementia) AND prevent* Field: All Fields, Limits: Publi-
cation Date from 1996 to 2006, Randomized Controlled
Trial, Human 122 articles
Also searched: google scholar: NHS database;
Total, 310 articles
Articles were considered ineligible if (1) secondary pre-
vention for those with dementia or AD; (2) human immu-
nodeficiency virus–related dementia; (3) articles with neu-
rochemical focus; (4) articles not containing original or new
data.
For targeted search of individual preventive interven-
tions, the following search strategy was used:
Cochrane Database of Systematic Reviews (up to
2006);
ACP Journal Club (1991 to February 2006);
Database of Abstracts of Review of Effects (DARE)
(up to 2006);
Cochrane Central Register of Controlled Trials
(CCRCT) (up to 2006)
 Alzheimer’s & Dementia 3 (2007) 355–384

Management of mild to moderate Alzheimer’s disease and dementia

David B. Hogana,*, Peter Baileyb, Anne Carswellc, Barry Clarked, Carole Cohene,
Dorothy Forbesf, Malcolm Man-Son-Hingg, Krista Lanctôth, Debra Morgani, Lilian Thorpej
a
Departments of Medicine and Clinical Neuroscience, University of Calgary; Calgary, Alberta, Canada
b
Department of Medicine (Neurology), Dalhousie University, Halifax, Nova Scotia, Canada
c
School of Occupational Therapy, Dalhousie University, Halifax, Nova Scotia, Canada
d
Department of Family Medicine, Dalhousie University, Halifax, Nova Scotia, Canada
e
Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada
f
School of Nursing, Faculty of Health Sciences, University of Western Ontario, London, Ontario, Canada
g
Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada
h
Departments of Psychiatry and Pharmacology, University of Toronto, Toronto, Ontario, Canada
i
Institute of Agricultural Rural and Environmental Health, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
j
Department of Psychiatry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Abstract The authors were charged with making a series of evidence-based recommendations that would
provide concrete advice on all aspects of the management of mild to moderate stages of dementia
and Alzheimer’s disease. The recommendations were primarily targeted to primary care physicians
practicing in Canada. The assigned topic area did not include either the assessment of a patient with
suspected dementia or the prevention of Alzheimer’s disease and other dementias. An extensive
examination of the available literature was conducted. Explicit criteria for grading the strength of
recommendations and the level of evidence supporting them were used. The 28 evidence-based
recommendations agreed on are presented in this article.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Alzheimer’s disease; Vascular dementia; Mild-to Moderate; Disease management

1. Introduction State Examination (MMSE) score between 10 –11 and

24 –26 (inclusive). This would translate to a Global De-
This article deals with the management of Alzheimer’s
terioration Scale stage of 4-6 and/or a Clinical Dementia
disease (AD) and other forms of dementia. The working
Rating score of 1–2.
group was asked to produce a summary of the available
In this document we will not deal with the prevention of
evidence on this topic. On the basis of this review we were
dementia and AD. Also, other than for AD we will make
to develop a series of recommendations that would provide
few recommendations for the management of specific types
concrete advice on all aspects of therapy along with sug-
of dementia (eg, dementia with Lewy bodies [DLB]). We
gestions for future developments. The primary target audi-
will not address the issue of the cost-effectiveness of the
ence for these recommendations would be primary care
various interventions described for patients with mild to
physicians.
moderate AD and dementia. A fiercely debated issue at this
The definition of mild to moderate AD and dementia
time is whether any of the available medications for mild to
was left up to the authors of the articles reviewed. Typ-
moderate AD are cost-effective. Our assigned topic area
ically they defined it as a patient meeting criteria for a
overlapped with the severe dementia working group be-
diagnosis of AD and/or dementia who had a Mini-Mental
cause a number of the studies we examined dealt with
moderate to severe stages of the condition. We would advise
*Corresponding author. Tel.: 403-220-4578; Fax: 403-283-6151. the reader to review the background article of this working
E-mail address: dhogan@ucalgary.ca group as well.
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.006
356 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
2. Methods
Our working group was charged with the responsibility
of addressing the management of mild to moderate AD and
dementia. The chair (D.H.) and the initial members of the
working group (P.B., C.C., M.H., L.T.) were selected by the
Steering Committee of the Third Canadian Consensus Con-
ference on Diagnosis and Treatment of Dementia (3rd
CCCDTD). In a series of teleconferences the assigned task
was divided into a number of subsections, with delegation
of lead responsibility to members of the working group. To
ensure that we had the range of required talent and exper-
tise, additional members of the working group were re-
cruited (A.C., B.C., D.F., and K.L.). The involved subsection
leads identified, discussed, and resolved in mutually acceptable
manner areas of potential overlap. During our teleconferences
we clarified how we would approach the task and dealt col-
lectively with any concerns expressed by working group mem-
bers. Minutes were kept of these teleconferences and circulated
to all members of the working group and the Chair of the 3rd
CCCDTD Steering Committee.
The initial literature searches were conducted by an in-
formation specialist hired by the 3rd CCCDTD. PubMed
and Embase databases were used. The strategy and major
keywords used in the searches were “dementia” OR “Alz-
heimer’s disease” AND “mild” OR “moderate” AND “ther-
apy” OR “treatment.” Secondary search terms included
(listed alphabetically) “affective disorder,” “agitation,” “an-
tidepressants,” “anti-inflammatory drugs,” “antioxidants,”
“anxiety,” “anxiolytics” OR “tranquilizers,” “behaviour,”
“cholinesterase inhibitors,” “care-giver,” “counseling,” “de-
pression,” “disinhibition,” “discontinue,” “education,” “en-
vironment,” “ginkgo,” “hormones,” “hypnotics” OR “sleep
medications,” “maintain,” “memantine,” “metabolic en-
hancers,” “neurotrophic agents,” “nootropics,” “rehabilita-
tion,” “selective serotonin reuptake inhibitors” OR “sleep.”
The search was limited to articles written in English, dealing
with human research, and published since January 1, 1996.
A total of 1,615 articles were identified. Six hundred
sixty-one of the articles were eliminated after examination
of the title. The titles, authors, and abstracts of the remain-
ing 954 were distributed to all working group members.
Each was responsible for selecting articles for detailed re-
view by their subsection, abstracting data from the selected
articles, synthesizing the available information, and devel-
oping draft recommendations. Full texts of articles selected
by working group members were provided by the informa-
tion specialist. On the request of working group members,
additional searches were conducted by the information spe-
cialist. The search strategies for them were developed in
consultation with the individuals making the request. Work-
ing group members were also encouraged to use their own
files and to search the reference lists of selected articles for
additional relevant articles. Working group members were
told that although they could use articles that summarize the
research literature (eg, meta-analyses, systematic reviews,
consensus statements, clinical practice guidelines), there
would be some areas in which they would have to perform
a primary review of the pertinent literature. They were also
asked to focus their energies on what they thought were the
key areas within their subsections and to draft recommen-
dations that would be both important and feasible for a
primary care physician.
The draft recommendations developed by working group
members and a first draft of the background article were
distributed by the Chair to all working group members for
review, discussion, and modification. After this process the
background article with recommendations was submitted to
the Steering Committee for posting on the Web page of
the 3rd CCCDTD. Feedback received was discussed by the
working group, and final modifications were made to the
recommendations before their presentation at the consensus
meeting on March 10, 2006. All of the recommendations
presented in this article achieved consensus (80% plus ap-
proval by participants of the 3rd CCCDTD).
The quality of the literature (levels of evidence) was
graded by using the following system adapted from Cana-
dian Task Force on Preventive Health Care [1]:
I. Evidence from at least one properly randomized
controlled trial (RCT).
II-1. Evidence from well-designed controlled trials with-
out randomization.
II-2. Evidence from well-designed cohort or case-
control analytic studies, preferably from more than
one center or research group.
II-3. Evidence from comparisons between times or
places with or without the intervention. Dramatic
results in uncontrolled experiments are included in
this category.
III. Opinions of respected authorities, on the basis of
clinical experience, descriptive studies, or reports
of expert committees.
The strength of the recommendations (Grade of Recom-
mendation) was graded by using the following system
adapted from the Canadian Task Force on Preventive Health
Care [1,2]:
A: There is good evidence to support this maneuver.
B: There is fair evidence to support this maneuver.
C: The existing published evidence is either conflicting
or insufficient and does not allow one to recommend
for or against this maneuver; however, a recommen-
dation might be made on other grounds.
D: There is fair evidence to recommend against this
maneuver.
E: There is good evidence to recommend against this
maneuver.
Our recommendations were based on the best available
evidence. We preferentially used rigorously done system-
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
atic reviews of the current literature. When a primary ex-
amination of the literature was done, we sought to base our
conclusions on statistically and clinically significant find-
ings from high quality RCTs.
The conclusions of the 1998 Canadian Consensus
Conference on Dementia (CCCD) that were considered rel-
evant to our assigned area and were still supported by the
members of the working group are also included in this
document [1].
3. Results: Management of mild to moderate
Alzheimer’s disease and dementia
3.1. General measures
A number of the conclusions reached at the CCCD [1]
were believed to be both relevant to our assigned topic area
and received the support of the members of the working
group after minor modifications. Three are presented below,
and others will be appear later in the document.
3.2. Recommendation 1
Most patients with dementia can be assessed and man-
aged adequately by their primary care physicians. However,
to assist them in meeting the needs of patients and their
caregivers, it is recommended that (1) all patients with
dementia and their families who consent be referred to the
local chapter of the Alzheimer Society (eg, First Link pro-
gram where available); and (2) primary care physicians
should be aware of the resources available for the care of
those with dementia in their community (eg, support groups,
adult day programs) and to make appropriate referrals to
them (Grade B, Level III).
3.3. Recommendation 2
The referral/consultation process is essential to the de-
livery of high quality health care. In the care of a patient
with mild to moderate dementia, reasons to consider referral
to a geriatrician, geriatric psychiatrist, neurologist, or other
health care professional (eg, neuropsychologist, nurse,
nurse practitioner, occupational therapist, physical therapist,
psychologist, social worker) with the appropriate knowl-
edge and expertise in dementia care would include (1)
continuing uncertainty about the diagnosis after initial as-
sessment and follow-up; (2) request by the patient or the
family for another opinion; (3) presence of significant de-
pression, especially if there is no response to treatment; (4)
treatment problems or failure with specific medications for
AD; (5) need for additional help in patient management (eg,
behavioral problems, functional impairments) or caregiver
support; (6) genetic counseling when indicated; and (7) if
the patient and/or family express interest in either diagnostic
or therapeutic research studies that are being carried out by
the recipient of the consult request (Grade B, Level III).
357
3.4. Recommendation 3
The care and management of patients with dementia
from specific cultural groups should take into account the
risk of isolation, the importance of culturally appropriate
services, and issues that arise in providing caregiver support
(Grade B, Level III).
The presence of a preexisting dementia is the risk factor
most strongly associated with the development of delirium
in older hospitalized patients [3]. A multicomponent inter-
vention to prevent delirium (ie, orienting communication,
therapeutic activities, sleep enhancement strategies, exer-
cise and mobilization, provision of vision and hearing aids,
oral repletion of dehydration) has been shown to decrease
the likelihood of delirium developing in older hospitalized
patients at increased risk [4]. Once it occurs, the manage-
ment of delirium in a patient with dementia remains empir-
ical, with no evidence from recent studies to support
changes from current practices [5].
Comorbidities are both common and costly in patients
with AD and dementia [6,7]. Appropriate therapy of their
comorbidities is an important component of the care pro-
vided to these patients. There is evidence that patients with
dementia are less likely to be offered recommended therapy
for other conditions [8]. Also, poor control of comorbidities
might accelerate the rate of progression in AD. With dia-
betes as an example, hyperglycemia itself can induce cog-
nitive changes by disrupting glucose metabolism with the
formation of abnormal glycosylation products and the de-
velopment of microvascular changes [9]. Hyperinsulinemia
is a risk factor for accelerated cognitive decline [10]. Insulin
increases the release of beta-amyloid and interferes with its
degradation by competing for the insulin-degrading enzyme
that metabolizes both insulin and beta-amyloid in the central
nervous system [11]. Optimal diabetic management might
slow down the rate of further decline in patients with ex-
isting AD, but this requires confirmation [12]. Community-
based studies have shown that the presence of comorbidities
predicts a higher mortality rate in patients with AD [13–15].
The presence of AD and dementia will affect the manage-
ment of other chronic conditions. A unique feature of the
care of a demented patient is reduced reliance on patient
self-care and a concomitant increase in the effort to provide
caregiver support and education.
3.5. Recommendation 4: Recommendations with regards
to the general medical care of a patient with mild to
moderate dementia
A. Patients with mild to moderate dementia when hos-
pitalized should be identified as being at increased
risk for delirium. They should be offered multicom-
ponent interventions including orienting communica-
tion, therapeutic activities, sleep enhancement strat-
egies, exercise and mobilization, provision of vision
358 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
and hearing aids, and/or oral repletion of dehydration
to decrease their risk of developing delirium (Grade
B, Level II-1).
B. Comorbidities of patients with mild to moderate AD
should be appropriately managed (Grade B, Level
III).
C. The management of other chronic medical conditions
might have to be modified in the setting of a demen-
tia. In general there should be less reliance on patient
self-care and a concomitant increase in the role
played by caregivers (Grade B, Level III).
Ensuring medication adherence can be a significant chal-
lenge in the care of an individual with dementia because
cognitive factors limit the ability of patients to self-medicate
[16]. Caregivers might have to become involved in medi-
cation management. The use of compliance aids would be
another option. For example, computer telephony systems
might improve medication adherence [17].
Studies have shown that older adults with probable de-
mentia are more likely to be taking anticholinergics than
matched controls [18]. A partial listing of medications with
anticholinergic activity is given in Table 1. Use of medica-
tions with anticholinergic effects can worsen the cognitive
status of individuals with AD and dementia [19 –23]. Older
persons taking anticholinergic medications can manifest
significant deficits in cognitive functioning and be classified
as having mild cognitive impairment [24]. Although some
drugs (eg, amitriptyline, benztropine) are well-known for
their anticholinergic effects, numerous other medications
possess mild anticholinergic properties. By themselves
these latter agents are unlikely to lead to significant clinical
symptoms, but the additive effects of multiple anticholin-
ergic medications taken simultaneously might result in ad-
verse effects. Another concern with their use is that they
might blunt the effects of cholinesterase inhibitors because
anticholinergics directly oppose the therapeutic effect of
these medications. The concurrent use of anticholinergics
and cholinesterase inhibitors is reportedly common [25]. A
small cohort study found that concomitant therapy with
anticholinergics was associated with worse outcomes in a
group of demented individuals being treated with donepezil
[26]. Another small study looked at cognitive and behav-
ioral status both on and off incontinence medications that
had anticholinergic effects. The subjects exhibited better
performance on specific measures of cognition and behavior
when off these medications. A significant, inverse relation-
ship was found between mental status and anticholinergic
level [27].
3.6. Recommendation 5: Recommendations about the
use of medications in the setting of a mild to
moderate dementia
A. Determination of how medications are being con-
sumed and identification of any problems/concerns
with medication management, including poor adher-
ence, should be done on all patients with mild to
moderate dementia. If problems are detected, in par-
ticular with adherence, the use of compliance aids or
the assumption of medication management by an-
other party will be necessary. The effectiveness of
any alterations in medication management will have
to be assessed (Grade B, Level III).
B. Even when the patient is safely self-managing their
medications, there should be planning for the in-
volvement of a third party in the management of
medications for all patients with a progressive de-
mentia because this will eventually become necessary
in nearly all (Grade B, Level III).
C. The use of medications with anticholinergic effects
should be minimized in persons with AD (Grade D,
Level III).
The presence of dementia does not in itself mean that
patients lack capacity to make decisions about themselves
and their estate. AD and other neurodegenerative dementias
are progressive conditions, however, that at some point will
likely rob patients of their mental capacity to consent to
treatment, consent to participate in a research study, look
after their estate, and/or make decisions about other aspects
of their life. Questions about capacity are likely to occur
during the mild and moderate stages of AD.
3.7. Recommendation 6: Ethicolegal recommendations
A. Although each case should be considered individu-
ally, in general the diagnosis of dementia should be
disclosed to the patient and family. This process
should include a discussion of prognosis, diagnostic
uncertainty, advance planning, driving issues, treat-
ment options, support groups, and future plans
(Grade B, Level III).
B. Primary care physicians should be aware of the per-
tinent laws in their jurisdiction about informed con-
sent, the assessment of capacity, the identification of
a surrogate decision maker, and the responsibilities of
physicians in these matters (Grade B, Level III).
C. While patients with AD retain capacity, they should
be encouraged to update their will and to enact both
an advance directive and an enduring power of attor-
ney (Grade B, Level III).
4. Nonpharmacologic interventions for cognitive and
functional limitations
4.1. Cognitive training/cognitive rehabilitation
Cognitive training is defined as guided practice on a set of
standard tasks designed to reflect specific cognitive function
such as memory or attention. Cognitive rehabilitation is an
individualized approach to helping people with cognitive im-
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384 359

Table 1 Table 1
Select medications with anticholinergic activity [18,280,281] Continued
Antiarrhythmic Glycopyrrolate
Disopyramide Hyoscine butylbromide
Antidiarrheal Hyoscyamine
Diphenoxylate/atropine Methscopolamine bromide
Antiemetics/antivertigo Oxybutynin
Cyclizine Pinaverium bromide
Dimenhydrinate Propantheline
Meclizine Tolterodine
Scopolamine Muscle relaxants
Trimethobenzamide Cyclobenzaprine
Antihistamines, either single or combination products containing Orphenadrine
Azatadine Opioid
Brompheniramine Meperidine
Carbinoxamine Tricyclic antidepressants
Chlorpheniramine Amitriptyline
Clemastine Amoxapine
Cyproheptadine Clomipramine
Dexbrompheniramine Doxepin
Dexchlorpheniramine Imipramine
Dimenhydrinate Nortriptyline
Diphenhydramine Protriptyline
Doxylamine Trimipramine
Hydroxyzine
Phenindamine
Promethazine
Trimeprazine pairment during which the person and family member deter-
Triprolidine mine personally relevant goals and then devise strategies for
Antiparkinsonian
addressing them to improve function in everyday contexts.
Benztropine
Biperiden Three small-sample RCTs examined the effect of cognitive
Ethopropazine training on memory tasks and functional performance in per-
Orphenadrine sons with dementia also taking cholinesterase inhibitors [28 –
Procyclidine 30]. The results suggest that performance on instrumental ac-
Trihexyphenidyl
tivities of daily living (IADL) were modestly enhanced.
Antipsychotics
Chloropromazine Overall there were no significant differences in functional
Clozapine performance between the groups at the end of the study period
Flupenthixol or on follow-up. The modest improvements seen on specific
Fluphenazine memory tasks were not sustained or generalizable to other
Loxapine
tasks.
Mesoridazine
Methotrimeprazine The only systematic review of cognitive training con-
Olanzapine cluded that because of a limited number of small-sample
Pericyazine RCTs and their equivocal results, it was not possible to draw
Pimozide any firm conclusions about the effectiveness of these inter-
Prochlorperazine
ventions on cognitive skills [31]. The data suggest that there
Promazine
Promethazine was a training effect on the specific skills being trained
Thioflupromazine (splinter skills) but little or no generalization to other cog-
Thioproperazine nitive abilities or to functional performance.
Thioridazine There was some evidence that an errorless learning par-
Thiothixene
adigm and/or consistent practice of usual daily activities
Zuclopenthioxol
Bronchodilators (procedural memory training) worked in a cognitive reha-
Atropine bilitation program to improve functional performance [32–
Ipratropium 35]. There were gains in functional performance that were
Gastrointestinal/genitourinary antispasmodics, either single or maintained during a period of 2 years. However, none of the
combination products containing
studies were RCTs.
Belladonna alkaloids
Clidinium bromide 4.2. Environment
Dicyclomine
Dicycloverine There is one high quality RCT, some pilot projects, and
Flavoxate
two systematic reviews of the literature examining the im-
360 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
pact of environmental interventions on functional perfor-
mance. The RCT examining the effectiveness of a home-
environmental intervention was conducted to examine,
among other things, its effect on the daily functioning of
persons with dementia living at home. The program con-
sisted of individualized environmental modifications (con-
sistency, structured) and aides (bath seats, visual cues). The
results indicated a modest but statistically significant effect
on IADL 3 and 6months after intervention [36], with trends
still present at 12 months [37].
Two systematic reviews of environmental impact on
functional performance were undertaken. Day et al [38]
reported that discrete individualized design features such as
organization of space, simplicity, and structure that pro-
moted orientation, problem-solving, memory, and mobility
had positive impacts on functional performance. A review
of six RCTs that met inclusion criteria [39] suggested that
environmental modification in the form of assistive devices,
aides, and adaptations together with individualized occupa-
tional therapy impacted positively on functional perfor-
mance of persons with cognitive impairment still living at
home.
4.3. Exercise
An RCT of 153 persons [40] examined whether a
3-month home-based exercise program together with
strategies for caregivers to manage behavior improved
functional independence in those with mild to moderate
AD. The results indicated that the exercise group had
significant increases on measures of physical functioning
and ADL. This trend persisted at 1 year after interven-
tion. A systematic review of the literature supports the
notion that repetitive, consistent training of tasks by
using procedural memory improves motor tasks and func-
tional performance [41]. A meta-analysis of 2,020 sub-
jects in 30 trials to determine whether exercises are
beneficial for people with mild to moderate dementia
suggested that exercise programs increased strength, fit-
ness, functional performance (ie, ADL, IADL), cognitive
function, and positive behavior [42].
4.4. Occupational therapy
One experimental study compared a structured occupa-
tional therapy program of caregiver strategies, environmen-
tal modifications, and community-based assistance to care-
givers receiving a written report [43]. They found that for
persons with moderate dementia, self-care improvements
noted after intervention were maintained at 3 months. How-
ever, a systematic review [39] concluded that there is in-
sufficient evidence about the effectiveness of individualized
occupational therapy programs, and further research was
required.
4.5. Other therapies
One small RCT (14 subjects) reported that transcutane-
ous electrical nerve simulation (TENS) appeared to have a
statistically beneficial effect on ADL that was present 6
weeks after intervention [44]. A systematic review of TENS
suggests that it produced short-term improvements in spe-
cific neuropsychological tests, but that the effectiveness of
this intervention was stage-dependent, did not appear to
impact functional performance, and was not maintained
beyond 3 weeks [45]. A systematic review of psychosocial
interventions for persons with mild or moderate dementia
[46] reported that reality orientation was not effective in
improving functional performance.
4.6. Recommendation 7: Recommendations for
nonpharmacologic interventions for the management of
the cognitive and functional limitations arising from mild
to moderate AD
The available research on nonpharmacologic interven-
tions for functional performance in persons with mild or
moderate dementia is limited. There are few well-designed
studies on which to base any firm conclusions. There is
promise for individualized exercise programs, individual-
ized environmental modifications, and individualized prac-
tice of ADL and IADL tasks. Future research, however, is
required to provide the evidence needed to reach solid
conclusions about the effectiveness of these interventions.
A. There is insufficient research evidence to come to any
firm conclusions about the effectiveness of cognitive
training/cognitive rehabilitation in improving and/or
maintaining cognitive and/or functional performance
in persons with mild or moderate dementia (Grade C,
Level 1).
B. Further research is required to be able to conclude
that cognitive training/cognitive rehabilitation is ef-
fective in improving cognitive and/or functional per-
formance in persons with mild or moderate dementia
(Grade B, Level II-1).
C. Although there is some indication of a beneficial
impact on IADL and ADL, there is insufficient evi-
dence to make firm conclusions about the effective-
ness of environmental interventions in promoting
functional performance in persons with mild or mod-
erate dementia (Grade C, Level 1).
D. There is good evidence to indicate that individualized
exercise programs have an impact on functional per-
formance in persons with mild or moderate dementia
(Grade A, Level 1).
E. For other nonpharmacologic therapeutic interventions
there is insufficient evidence to allow any conclusions
being made about their efficacy in improving or
maintaining functional performance in persons with
mild or moderate dementia (Grade C, Level 1).
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
5. Pharmacotherapy for cognitive and
functional impairments
5.1. Retained general recommendations for the
pharmacotherapy of dementia initially proposed by the
CCCD [1] (with minor amendments)
5.1.1. Recommendation 8
Primary care physicians should be able to administer and
interpret brief measures of functional activities and cogni-
tive abilities or refer to health care professionals with the
required knowledge and expertise (Grade B, Level III).
5.1.2. Recommendation 9
After treatment has been started, patients should be re-
assessed regularly by the appropriate health care profes-
sional involved in their care (Grade B, Level III).
5.1.3. Recommendation 10
Records should be kept such that stabilization, improve-
ment, or persisting deterioration in treated patients will be
determinable (Grade B, Level III).
5.1.4. Recommendation 11
In monitoring the response to therapy of patients with
dementia, the input of caregivers (where available) should
be sought. They can provide information on the patient’s
cognition, behavior, and social and daily functioning (Grade
B, Level III).
5.1.5. Recommendation 12
If the attending primary care physician is unable to
perform the assessments required to gauge response to ther-
apy, referral to another health care professional with knowl-
edge and expertise in dementia care (eg, other physician,
nurse, occupational therapist) or a service (eg, memory
clinic) who is willing to perform such assessments is ad-
vised (Grade B, Level III).
5.1.6. Recommendation 13
Primary care physicians should be able to communicate
appropriate information concerning dementia, including re-
alistic treatment expectations, to their patients and their
families (Grade B, Level III).
Most clinicians view the cholinesterase inhibitors as
first-line treatment for mild to moderate stages of AD [47].
There are three cholinesterase inhibitors available in Can-
ada: donepezil, galantamine, and rivastigmine. Although
rivastigmine and galantamine have additional modes of ac-
tion, they all inhibit the breakdown of acetylcholine by
blocking the enzyme acetylcholinesterase.
Cochrane reviews of each of the agents are available [48 –
50] as well as other meta-analyses [51–53] and qualitative
systematic reviews [54 –57]. Cognitive, functional, and global
outcomes have been measured in these trials. The primary
cognitive test in most studies has been the cognitive section of
the Alzheimer’s Disease Assessment Scale (ADAS-Cog). This
instrument consists of 11 individual items (spoken language,
361
comprehension of spoken language, recall of test instructions,
word finding, following commands, naming objects, construc-
tion, ideational praxis, orientation, word recall, and word rec-
ognition) and is scored out of 70 (higher scores indicate greater
impairment). MMSE scores are also often reported as a sec-
ondary cognitive outcome. Activities of daily living have been
evaluated with a variety of instruments such as the Progressive
Deterioration Scale (PDS), the Disability Assessment for De-
mentia (DAD), and the Bristol Activities of Daily Living Scale
(BADLS). Global assessment has usually been done with the
Clinician’s Interview-Based Impression of Change with care-
giver input (CIBIC-Plus). Patients are assessed at baseline. At
subsequent assessments they are graded on a 1 to 7 scale
relative to this baseline assessment, with one indicating very
much improved, four no change, and seven very much worse.
The RCTs of these agents have shown consistent, albeit mod-
est, benefits of treatment in cognition, activities of daily living,
and global clinical state. Systematic reviews including those
with severe dementia find similar results to those including
only trials of those with mild to moderate impairment. The
methodologic limitations of the published studies (eg, report-
ing more than one outcome without statistical correction for
multiple comparisons, absence of final outcome measures on
subjects who had withdrawn) were specifically noted in a
systematic review of the cholinesterase inhibitors. Because of
the modest benefits seen and the methodologic limitations of
the studies, the authors questioned the utility of these agents
[57]. Other commentators have concluded, however, that
the likely effect of the methodologic concerns does not
invalidate the findings of the studies reviewed and that
these agents are modestly efficacious for the treatment of
mild to moderate AD.
Comparison of the relative efficacy and tolerability of the
cholinesterase inhibitors across clinical trials is not appropriate
because of differences in the baseline characteristics of the
subjects, the efficacy assessments done, and how subjects were
assessed for adverse effects. The available trials that directly
compared one cholinesterase inhibitor to another have meth-
odologic limitations and/or show no significant differences in
their primary outcome measures [47,58,59].
The most common side effects encountered with the
cholinesterase inhibitors in the RCTs of these agents were
gastrointestinal (eg, anorexia, nausea, vomiting, diarrhea)
[47–50,60]. They are more likely to occur at the commence-
ment of therapy or when the dose of the agent is increased.
These side effects are dose-related and tend to be transient.
Gastrointestinal side effects in the RCTs were more com-
mon with rivastigmine. Slower titration and ensuring riv-
astigmine is taken with food appear to decrease the risk of
gastrointestinal side effects. Weight loss did occur during
the RCTs of all three agents, but a follow-up of patients with
AD treated with cholinesterase inhibitors found that there
was not an increased risk for long-term weight loss (com-
pared with patients with AD not receiving a cholinesterase
inhibitor) and might in fact be a protective factor [61].
362 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
A variety of other adverse effects can occur. Dizziness
has been reported with all three agents. If disabling, dose
reduction would be a reasonable initial approach. Syncope,
while rare, has been associated with the use of these agents.
In a small descriptive study, noninvasive evaluation suc-
cessfully identified the probable cause of syncope in most
patients with AD treated with donepezil [62]. Therapeutic
options would include stopping the agent or pacemaker
implantation [63]. Donepezil has been associated with sleep
disturbances, vivid dreams/nightmares, and hypnopompic
hallucinations [60]. The Cochrane meta-analysis of donepe-
zil confirmed a dose-related increased odds ratio for insom-
nia [47]. Rivastigmine and galantamine are less likely to
cause sleep disturbances. Management options for this
problem would include changing the timing and/or dose of
donepezil or switching to another cholinesterase inhibitor.
A prescribing cascade involves the misinterpretation of
an adverse reaction to one drug followed by the prescription
of potentially inappropriate second drug to deal with this
adverse effect. The use of cholinesterase inhibitors is asso-
ciated with an increased risk of receiving an anticholinergic
drug to manage urinary incontinence [64]. The use of an
anticholinergic drug in this setting might represent a clini-
cally important prescribing cascade. Another potential ex-
ample of this would be the higher use of hypnotics in
patients with AD treated with donepezil [65].
A number of studies have shown that it is possible to
switch from one cholinesterase inhibitor to another [66 –72].
Generally in these studies patients abruptly discontinued the
first cholinesterase inhibitor and started taking the second
agent the following day at the usual starting dose followed
by up-titration at the usual rate for the new agent. Unfortu-
nately, these studies do not tell us when we should switch.
Commonly cited reasons for switching include unsatisfac-
tory response to the first agent, intolerable side effects to the
first agent, and request of the patient and/or caregiver.
Switching from donepezil to memantine was well-tolerated
in a study sponsored by Lundbeck, whether it was done
abruptly (donepezil discontinued abruptly with memantine
up-titrated to 20 mg/d during a period of 3 weeks) or
gradually (donepezil dropped from 10 mg/d to 5 mg/d for 2
weeks before stopping with memantine up-titrated to 20
mg/d during a period of 3 weeks) [73]. Please note that
switching from one cholinesterase inhibitor to another cho-
linesterase inhibitor or memantine can lead to deterioration
in the status of a patient. Patients and families should be
informed of this possibility before a switch is made.
5.2. Recommendation 14: Recommendations regarding the
use of cholinesterase inhibitors
A. All three cholinesterase inhibitors available in Canada
are modestly efficacious for mild to moderate AD. They
are all viable treatment options for most patients with
mild to moderate AD (Grade A, Level I).
B. Although all three cholinesterase inhibitors available
in Canada have efficacy for mild to moderate AD,
equivalency has not been established in direct com-
parisons. Selection of which agent to be used will be
based on adverse effect profile, ease of use, familiar-
ity, and beliefs about the importance of the differ-
ences between the agents in their pharmacokinetics
and other mechanisms of action (Grade B, Level I).
C. All physicians prescribing these agents should be
aware of the contraindications and precautions with
the use of cholinesterase inhibitors (Grade B, Level
III).
D. If adverse effects occur with a cholinesterase inhib-
itor, the agent should either be discontinued (if the
side effects are judged to be disabling and/or danger-
ous), or the dose of the agent should be decreased,
with an option to retry the higher dose after 2 to 4
weeks if the lower dose is tolerated (if the side effects
are judged to be minor in severity) (Grade B, Level
III).
E. If nausea and/or vomiting occur with the use of a
cholinesterase inhibitor, review how the medication
is being taken (eg, dose, frequency, with or without
food, evidence of an unintentional overdose) and con-
sider modifying the prescription (eg, lower dose),
responsibility for administration (eg, caregiver taking
over from the patient), the directions given to the
patient (eg, with food), or stopping the agent. Al-
though antiemetics can be used for nausea and/or
vomiting, a number of them (eg, dimenhydrinate,
prochlorperazine) have anticholinergic properties that
can lead to adverse cognitive effects (Grade B, Level
III).
F. Clinicians should consider the possible contributing
role of cholinesterase inhibitors in new-onset or wors-
ening medical presentations and the potential risk of
co-prescribing cholinesterase inhibitors and other
drugs to patients with dementia (Grade B, Level II-2).
G. Patients can be switched from one cholinesterase
inhibitor to another. A decision to make a switch is
based on the judgment of the prescribing physician
and the patient (or their proxy) about the relative
benefits and risks of making a change in the patient’s
pharmacotherapy (Grade B, Level III).
H. Patients can be switched from a cholinesterase inhib-
itor to memantine (note: see recommendation 15B).
The decision of when to make a switch is based on
the judgment of the prescribing physician and the
patient (or their proxy) (Grade B, Level III).
Memantine is a low to moderate affinity, uncompetitive
antagonist to glutamate N-methyl-D-aspartate (NMDA) re-
ceptors and might prevent excitatory neurotoxicity in de-
mentia. Published studies show a small beneficial effect of
memantine in moderate to severe AD [74]. It has not been
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
shown to be effective for mild stages of AD. Although the
one published study showed efficacy on cognitive, behav-
ioral, and global outcomes, two additional unpublished
studies of memantine failed to show statistically significant
benefits for patients with mild AD [74,75].
The NMDA receptor antagonist memantine and the cho-
linesterase inhibitors have different mechanisms of action.
It seems reasonable to assume that an additive effect might
be achieved from combination therapy. One RCT did show
additional benefit when memantine was added to chronic
donepezil therapy in patients with moderate to severe AD
[76]. An unpublished study of memantine failed to show
any statistically significant benefit when it was given to
participants with mild to moderate AD who were on a stable
dose of a cholinesterase inhibitor [75].
5.3. Recommendation 15: Recommendations regarding the
use of memantine
A. Memantine is an option for patients with moderate to
severe stages of AD (Grade B, Level I). Its use in
mild stages of AD is not recommended (Grade D,
Level I).
B. Combination therapy of a cholinesterase inhibitor and
memantine is rational (because the medications have
different mechanisms of action), appears to be safe,
and might lead to additional benefits for patients with
moderate to severe AD. This would be an option for
patients with AD of a moderate severity (Grade B,
Level I).
An area of clinical uncertainty is when to stop a cho-
linesterase inhibitor (or any other agent being given for
AD). There is general acceptance that these decisions
should be individualized and based on the balance between
benefits and harm for the patient. This has become partic-
ularly contentious as a result of studies suggesting that
interrupting therapy for prolonged periods of time (eg, 6
weeks) can result in the loss of treatment benefits that
cannot be recaptured [77].
5.4. Recommendation 16: Recommendations about when
medications for the treatment of cognitive and functional
manifestations of AD should be discontinued
A. The patient and/or their proxy decision maker decide
to stop;
B. The patient refuses to take the medication;
C. The patient is sufficiently nonadherent with the med-
ication that continued prescription of it would be
useless, and it is not possible to establish a system for
the administration of the medication to rectify the
problem;
D. There is no response to therapy after a reasonable
trial;
E. The patient experiences intolerable side effects;
363
F. The comorbidities of the patient make continued use
of the agent either unacceptably risky or futile (eg,
terminally ill); or
G. The patient’s dementia progresses to a stage where
there is no significant benefit from continued therapy
(Grade B, Level III).
5.5. Recommendation 17
After stopping therapy for AD, patients should be care-
fully monitored, and if there is evidence of a significant
decline in their cognitive status, functional abilities, or the
development/worsening of behavioral challenges, consider-
ation should be given to reinstating the therapy (Grade B,
Level III).
Both basic and clinical studies have examined whether
antioxidants might be beneficial for individuals with AD
[78,79]. It remains unclear whether reactive oxidative spe-
cies are a cause or are a consequence of AD pathology. A
number of studies have examined the natural antioxidants
found in foods, vitamins E and C, and carotenes. In vitro
studies have shown that vitamin E can decrease both
amyloid-induced lipid peroxidation and oxidative stress
while suppressing the inflammatory signaling cascade
[79,80]. Vitamin C can block the creation of nitrosamines
through the reduction of nitrates and might also affect cat-
echolamine synthesis. Carotenes can modify lipid peroxida-
tion [80]. Cohort studies examining the relationship be-
tween antioxidant intake and the likelihood of subsequent
dementia have shown equivocal results [81– 83]. Uncer-
tainty persists as to the optimal agent, timing, duration,
dosage, and mode of intake [79]. There has been one high
quality RCT that examined the utility of high dose vitamin
E (2,000 IU/d) in subjects with moderate AD [84]. Although
this study suggested benefit, a recent RCT for subjects with
mild cognitive impairment found none [85]. An older Co-
chrane review concluded there was insufficient evidence of
efficacy of vitamin E in the treatment of AD to justify its use
[86]. Recent reports of a higher mortality rate among those
treated with high doses of vitamin E (400⫹ IU/d) supple-
mentation have cast even further doubt on the advisability
of using this antioxidant [87]. Treatment with idebenone, a
synthetic analog of coenzyme Q10 that is an antioxidant, did
not slow cognitive decline in a 52-week RCT of 536 sub-
jects with AD [88].
A number of studies have examined the potential role of
vitamin supplementation. Homocysteine levels can increase
with a deficiency of any of vitamins B6, B12, and/or folic
acid [89]. In the Framingham study the incidence of both
cerebrovascular disease and AD was increased if homocys-
teine levels were greater than 14 ␮mol/ L [90]. Other stud-
ies, however, have failed to support the association between
higher homocysteine levels and AD [91,92]. Individuals
with lower levels of folate and/or vitamin B12 have been
found to have a higher risk of developing dementia in some
364 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[93] but not all studies [94]. Several trials of vitamin B12 for
AD have shown inconsistent results [95,96]. Cochrane re-
views for vitamins B1 (for AD), B6 (for cognition), B12 (for
cognition), and folic acid (for cognition and dementia) have
been done. No conclusions could be drawn from the B1
studies examined [97]. No methodically adequate B6 trials
that involved people with dementia were found [98]. Two
studies of people with dementia and low serum B12 levels
were examined, but no statistically significant evidence of
cognitive benefits with B12 supplementation was found [99].
No benefits on any measure of cognition or mood from folic
acid with or without vitamin B12 supplementation were seen
in patients with a dementia [100]. Although more studies are
needed, the routine administration of any of these vitamins
to individuals with a dementia who do not have a docu-
mented deficiency state cannot be endorsed at the present
time.
Ginkgo biloba is an ancient Chinese herbal preparation
that is commonly used for the treatment of cognitive im-
pairment and dementia. Ginkgo is one of the five top pre-
scribed products in Germany, and in North America it is the
top-selling herbal remedy [101]. A review of the published
studies of ginkgo concluded there was a small but statisti-
cally significant benefit seen with ginkgo compared with the
placebo arm. Although the typical benefit seen was consis-
tently less than that obtained with the cholinesterase inhib-
itors, ginkgo was better tolerated [102]. A Cochrane review
of Ginkgo biloba for cognitive impairment and dementia
concluded that the agent appeared both safe and promising.
Concerns were expressed, however, about the methods used
in a number of the earlier trials. The more methodologically
sound trials showed inconsistent results [103]. Fifteen pub-
lished case reports have described a temporal association
between use of ginkgo and a bleeding event [104]. In 13 of
the case reports other identified risk factors for bleeding
were present. Patients using ginkgo, particularly those with
known bleeding risks (eg, concurrent use of warfarin or
antithrombotics), should be counseled about a possible in-
crease in bleeding risk.
A number of lines of evidence suggest that there are
increased levels of inflammatory mediators (eg, interleukins
1 and 6, tumor necrosis factor, complement) in the brains of
those with AD [105,106]. Microglia cells are activated
around amyloid plaques [107]. There is epidemiologic evi-
dence that patients on anti-inflammatory agents have a re-
duced incidence of AD [108,109]. The RCTs that use anti-
inflammatory drugs (eg, naproxen) as therapy for AD,
however, have been unsuccessful to date [110]. Initial trials
with nonsteroidal anti-inflammatory drugs (NSAIDs) were
plagued with high dropout rates [111]. A systematic review
concluded that indomethacin cannot be recommended for
the treatment of mild to moderate AD [112]. Selective
cyclooxygenase 2 inhibitors were better tolerated, but the
trials of both rofecoxib and celecoxib were negative
[111,113]. There are mechanisms other than cyclooxygen-
ase inhibition that might contribute to the potential benefits
of NSAIDs in the prevention or management of AD. Some
(ie, sulindac, indomethacin, flurbiprofen. ibuprofen) but not
all NSAIDs have been found to affect beta-amyloid depo-
sition and metabolism. It is possible that the negative
NSAID trials to date might have occurred because the
wrong NSAID was tested. For example, no RCTs have been
completed that used ibuprofen [114]. A study of prednisone
(20 mg for 1 month and then 10 mg for 1 year) in 132
patients failed to show any significant benefit [115]. An
RCT of hydroxychloroquine was also negative [116].
The large RCTs of inhibitors of the 3-hydroxy-3-
methylglutaryl coenzyme A (HMG-CoA) reductase enzyme
(ie, “statins”) have had primary cardiovascular outcomes,
although some of the studies have included secondary cog-
nitive outcomes. The MRC/BHF Heart Protection Trial of
simvastatin found no beneficial impact with active treatment
on the likelihood of either cognitive decline or being diag-
nosed with a dementia [117]. The PROSPER trial of prav-
astatin also could not detect any cognitive benefit with
active therapy [118]. On the other hand, a 12-month
placebo-controlled pilot trial of atorvastatin 60 mg in 71
patients with mild to moderate AD showed statistically
significant improvement on the ADAS-Cog at 6 months and
a positive trend at 12 months [119]. Additional trials are
ongoing.
Estrogens have a number of potentially beneficial effects
for women with dementia. A Cochrane review examined the
effect of hormone replacement therapy on cognition in
women with a dementia. Five double-blind RCTs that in-
cluded 210 women were evaluated in detail. Short-lived and
clinically insignificant positive effects with conjugated
equine estrogens (CEEs) were found on the MMSE (CEE,
0.625 mg/d only), Trail-Making Test-B (CEE, 0.625 mg/d
only), and digit span backwards (CEE, 1.25 mg/d only).
Cued delayed recall of a word list was positively affected after
2 months of transdermal diestradiol. Control subjects did sig-
nificantly better on delayed recall (1 month), finger tapping (12
months), and on the Clinical Dementia Rating scale. The pos-
itive effects seen were believed to be possibly from random
effects caused by multiple analyses. After correction for mul-
tiple testing only the short-term effect of transdermal estrogen
remained statistically significant [120].
Androgens (the primary androgen is testosterone; it can
be metabolized to the more potent androgen, dihydrotestos-
terone) can influence brain function directly through inter-
actions with androgen receptors or indirectly through estra-
diol (testosterone is converted to estradiol by aromatase)
[121]. In vitro and animal studies indicate that androgen
depletion is associated with higher brain levels of beta-
amyloid, hyperphosphorylation of tau protein, and de-
creased neuronal survival after exposure to a toxin [121–
123]. Studies of healthy older men suggest that therapy with
testosterone has a weak and inconsistent association with
better visuospatial and memory scores on testing [122].
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
Some but not all studies have shown an association between
reduced testosterone levels and a diagnosis of AD [124].
Two small intervention studies that included subjects with
AD have been done. In one study 10 hypogonadal nursing
home patients with AD were randomized to either intramus-
cular testosterone enanthate 200 mg every 2 weeks or pla-
cebo [125]. Unblinded assessments at 3, 6, and 9 months
showed improvements on the ADAS-Cog, MMSE, and the
Clock Drawing Test. One patient became aggressive and
developed hypersexual behaviour. No other problems were
noted. The second study was a randomized, double-blind,
placebo-controlled 6-week trial that examined the effects of
weekly intramuscular injections of 100 mg of testosterone
enanthate on subjects with mild cognitive impairment or
AD (a total of 15 subjects with AD were enrolled) [126].
Improvements in spatial memory/ability and verbal memory
were seen with testosterone therapy. No adverse effects
were encountered. Both groups of researchers believed that
additional studies were required.
A large number of other agents with diverse mechanisms
of action have been tested as potential therapies for AD. A
partial listing (in alphabetical order) of these agents would
include acetyl-L-carnitine; active or passive beta-amyloid
immunization; Alzhemed; aniracetam; BMY21,502; be-
sipiridine; cerebrolysin; clioquinol; cytidinediphosphocho-
line (CDP-choline); D-cycloserine; DGAVP; dehydroepi-
androsterone; doxycycline and rifampin; erythropoietin;
extract of Melissa officinalis; garlic; gamma-secretase in-
hibitor; growth hormone releasing hormone; huperzine A;
hydergine; ispronicline; lethicin; lithium carbonate; melato-
nin; milacemide; neramexane; nicergoline; nicotine; nimo-
dipine; paclitaxel; phosphatidyl serine; physostigmine;
piracetam; propentofylline; rosiglitazone; selegiline;
velnacrine; and vinpocetine. The studies of these agents
have been either negative or inconclusive. None have
been approved for the treatment of AD in Canada.
5.6. Recommendation 18: Recommendations with regard
to supplements, herbal preparations, and other
medications for the cognitive and functional
manifestations of AD and dementia
A. High-dose (ie, 400⫹ IU/day) vitamin E supplemen-
tation is not recommended for the treatment of AD
(Grade E, Level I).
B. The use of the synthetic antioxidant idebenone is not
recommended for the treatment of AD (Grade E,
Level I).
C. The administration of vitamin B1, B6, B12, and/or
folic acid supplements to persons with AD who are
not deficient in these vitamins is not recommended
(Grade D, Level III).
D. There is insufficient evidence to allow for a recommen-
dation either for or against the use of ginkgo biloba in
365
the treatment of dementia. Further methodologically
sound trials are required (Grade C, Level I).
E. The use of an anti-inflammatory drug is not recom-
mended for the treatment of the cognitive, functional,
or behavioral manifestations of a dementia (Grade D,
Level I).
F. The use of a HMG-CoA reductase enzyme inhibitor is
not recommended for the treatment of the cognitive,
functional, or behavioral manifestations of a dementia
(Grade D, Level III).
G. Hormone replacement therapy (estrogens combined
with a progestagen) or estrogen replacement therapy
(estrogen alone) is not recommended for the cognitive
impairments of women with AD (Grade D, Level I).
H. There is insufficient evidence to recommend the use
of androgens (eg, testosterone) to treat AD in men
(Grade C, Level I).
I. There is negative, inconclusive, or conflicting evi-
dence for a number of other agents proposed as
potential therapies for the cognitive and behavioral
manifestations of AD. Their use cannot be recom-
mended at this time (Grade C or D, Levels I to III,
varies between agents).
6. Nonpharmacologic and pharmacologic therapy of
behavioral and mood disturbances
Neuropsychiatric symptoms of dementia, also known as
behavioral and psychological symptoms of dementia
(BPSD), occur in the majority of people with dementia over
the course of the disease [127]. Disease severity affects the
prevalence of psychiatric symptoms. Major depression oc-
curs more often in mild to moderate cognitive impairment,
whereas most other symptoms are considered more com-
mon with greater dementia severity [128,129]. However,
noncognitive behavioral symptoms do not correlate well
with each other. In some longitudinal studies they have not
shown progressive worsening with time but have appeared
episodically [130]. In mild AD, some symptoms are com-
mon. For example, Lopez et al [129] found that of those
with mild AD, 60% had anxiety, 55% had lack of energy,
50.5% had anhedonia, 49% had agitation, 39% had irrita-
bility, and 25.5% had delusions or hallucinations. In early
dementia psychotic symptoms (especially visual hallucina-
tions) are relatively uncommon with AD, but they occur
more frequently with DLB.
The typical rating scales used for psychiatric disorders
become progressively more difficult to use as the severity of
dementia increases. For example, whereas in mild AD in-
struments such as the Geriatric Depression Scale [131]
might still be valid in screening for depression, by the later
stages psychometric factors such as internal consistency
worsen [132], necessitating the use of alternate instruments.
Instruments available for rating behavioral and mood symp-
366 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
toms in more advanced dementia are addressed in more
detail in the section on severe dementia.
6.1. Recommendation 19
Assessment of patients with mild to moderate AD should
include measures of behavior and other neuropsychiatric
symptoms (Grade B, Level III).
6.2. Recommendation 20
The management of BPSD should include a careful doc-
umentation of behaviors and identification of target symp-
toms, a search for potential triggers or precipitants, record-
ing of the consequences of the behavior, an evaluation to
rule out treatable or contributory causes, and consideration
of the safety of the patient, their caregiver, and others in
their environment (Grade B, Level III).
Few data are available for the nonpharmacologic treat-
ment of depression in dementia. One controlled study [133]
suggests that behavioral treatments, including those empha-
sizing pleasant events (for the patient) and those emphasiz-
ing caregiver problem solving, might decrease depressive
symptoms in this group. A Cochrane review [134] exam-
ined high quality trials of antidepressants for depressive
disorders in patients with dementia. Eight studies met their
stringent inclusion criteria. The mean MMSE score before
treatment in the studies ranged between 15 and 23. The
authors concluded there was only weak support for the
efficacy of antidepressants in treating depression in patients
with dementia, but they noted that only two studies used
selective serotonin reuptake inhibitors (SSRIs) and sample
sizes were small.
Although the treatment of depression in demented people
might not improve cognition [135], there is good evidence
that it might improve the quality of life of patients and their
caregivers [136]. Therefore, most review articles [137]
strongly recommend treatment of depression in people with
AD. Antidepressants with significant anticholinergic activ-
ity (eg, tertiary amine tricyclics such as amitriptyline, imip-
ramine, trimipramine, and doxepin) are likely to worsen
cognition [138] and should generally be avoided. The SSRIs
generally have less anticholinergic activity (note: paroxetine
has significantly more anticholinergic activity than other
SSRIs) [139], although they do have other side effects that
are of concern in frail older people (ie, gastrointestinal
symptoms, weight loss, sleep problems, and hyponatremia)
[140]. They have been associated with falls and hip frac-
tures [141], possibly related to mechanisms that differ from
those causing falls in tricyclics [142]. The choice of a spe-
cific antidepressant therefore has to consider both its particular
advantages and potential adverse effects, individualizing the
choice for each patient. Once a decision has been made to use
an antidepressant, of the classes of agents currently available,
SSRIs would be appropriate for first-line treatment of depres-
sion in patients with AD. Other classes of nontricyclic antide-
pressants might also be appropriate for first-line treatment of
depression [143]. Although data are sparse, electroconvulsive
treatment can be useful for patients with depression who have
not responded to antidepressants or who cannot tolerate
pharmacotherapy [144].
6.3. Recommendation 21: Recommendations with regard
to the management of depressive symptoms in the setting
of mild to moderate dementia
A. Because depressive syndromes are frequent in pa-
tients with dementia, physicians should consider di-
agnosing depression when patients present with the
subacute development (eg, weeks, rather than months
or years) of symptoms characteristic of depression
such as behavioral symptoms, weight and sleep
changes, sadness, crying, suicidal statements, or ex-
cessive guilt (Grade B, Level III).
B. Depressive symptoms that are not part of a major
affective disorder, severe dysthmia, or severe emo-
tional lability should initially be treated nonpharma-
cologically (Grade B, Level III).
C. If the patient had an inadequate response to the non-
pharmacologic interventions or has a major affective
disorder, severe dysthymia, or severe emotional la-
bility, a trial of an antidepressant should be consid-
ered (Grade B. Level III).
D. If an antidepressant is prescribed to a person with
AD, the preferred choice would be an agent with
minimal anticholinergic activity, such as an SSRI
(Grade B, Level III).
The prevalence of sleep disturbance and complaints is
high in older persons including those with dementia. Nor-
mal age-associated changes in sleep include less deep sleep
and less rapid eye movement (REM) sleep and worse sleep
efficiency (defined as the amount of time sleeping over the
amount of time spent in bed). The major sleep difficulty
encountered in older individuals is a decreased ability to
maintain sleep. A variety of factors such as medical and
psychiatric illnesses, changes in the timing and consolida-
tion of sleep (from changes in the endogenous circadian
rhythm), medications, the presence of other sleep disorders
(eg, periodic limb movements in sleep [PLMS], REM sleep
behavior disorder [RBD], sleep-disordered breathing), en-
vironmental factors (eg, noise and light), and poor sleep
habits can all contribute to a decline in the quantity and/or
quality of sleep [145]. Sleep disturbances can affect up to
44% of community-dwelling persons with dementia [146].
For family caregivers, nocturnal disturbances such as being
awakened at night by care recipients wandering or getting
out of bed repeatedly are disturbing aspects of care and a
major risk factor for institutionalization [147].
McCurry et al [148,149] evaluated the effectiveness of
the Nighttime Insomnia Treatment and Education for Alz-
heimer Disease (NITE-AD) program in improving sleep.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
Caregivers in the treatment group received specific recom-
mendations about setting up and implementing a sleep pro-
gram for their care recipient and training in behavior man-
agement skills. Care recipients were also instructed to walk
daily and increase daytime light exposure with the use of a
light box. Control subjects received general dementia edu-
cation and caregiver support. After 2 months of treatment,
care recipients exhibited significantly greater reductions in
number of nighttime awakenings, total time awake at night,
and depression. At 6-month follow-up, treatment gains were
maintained, and additional significant improvements in du-
ration of night awakenings emerged. Clinicians who recom-
mend these interventions should be aware that many
caregivers need active assistance with setting up and imple-
menting a sleep program. Simply providing caregivers with
education is often insufficient. A strength of this study is the
combination of approaches (sleep program, walking, and
bright light) that individually might influence behavioral
problems.
Although the treatment of insomnia in the older patients
with the newer sedative hypnotics has been shown to be
effective and safe [145], a recent and very comprehensive
meta-analysis has found that the magnitude of effect was
small, and that the increased risk of adverse effects was of
concern [150]. These adverse effects include ataxia, mem-
ory loss, and falls, as well as impairment in driving ability.
There is also a risk of dependency and accidental overuse.
There are no randomized clinical trials of sedative hypnotic
medications for sleep disturbance in AD [151], but these
agents are widely clinically used. Their short-term use can
be justified in situations in which sleep disturbance is par-
ticularly problematic. The pharmacotherapy of specific
sleep disturbances such as PLMS and restless legs syn-
drome are supported by a greater evidence base and might
include other pharmacologic interventions such as dopa-
mine agonists.
RBD is manifested by vivid, often frightening dreams
during REM sleep but without atonia. Patients “act out their
dreams” with vocalization, flailing of their limbs, and/or
moving around the bed. The history of RBD is obtained
from the patient’s bed partner. RBD is frequently associated
with synucleinopathies including Parkinson’s disease and
DLB. It is considered a suggestive feature of DLB [152].
REM sleep reportedly has a cholinergic basis [153]. Al-
though this is an area requiring further study, treatment
options for RBD in the setting of DLB would include
a clonazepam, cholinesterase inhibitor, melatonin, and
quetiapine [152,154 –157].
6.4. Recommendation 22: Recommendations with
regard to sleep problems in the setting of a mild to
moderate dementia
A. Patients with AD experiencing sleep problems should
first undergo a careful assessment for medical ill-
367
nesses (including pain), psychiatric illnesses (espe-
cially depression), potentially contributing medica-
tions, environmental factors, and/or poor sleep habits
(eg, daytime naps) that might be adversely affecting
sleep. Any identified secondary cause should be man-
aged (Grade B, Level III).
B. The presence of an RBD in the setting of a dementia
would be suggestive of DLB and related conditions.
Treatment options would include clonazepam (Grade
B, Level II-2).
C. Nonpharmacologic approaches to sleep disturbances
can be effective for patients with AD, but a combi-
nation of these approaches will likely be required
(Grade B, Level I).
D. When considered clinically necessary, pharmaco-
logic interventions for insomnia, including short- to
intermediate-acting benzodiazepines and related
agents, can be used at the lowest effective doses and
for the shortest possible time (Grade B, Level III).
Several systematic reviews of approaches that address be-
havioral problems associated with dementia also have been
conducted [158 –161], with the most comprehensive review
conducted by Livingston et al [162]. Specific types of psycho-
social education for caregivers about managing behavioral and
psychological symptoms were effective treatments whose ben-
efits lasted for months. Music therapy, Snoezelen, and possibly
sensory stimulation were beneficial during the treatment ses-
sion but had no long-term effects [162].
Several Cochrane Reviews have examined the effect of a
variety of interventions in managing the symptoms of de-
mentia. They concluded that although some of the interven-
tions studied look promising, they remain unproven. All of
the following Cochrane Reviews included subjects with
moderate to severe dementia and who resided in long-term
care facilities. A review of reminiscence therapy included
four trials that revealed significant improvement in cogni-
tion and mood 4 to 6 weeks after the treatment [163]. A
review of validation therapy included three studies that
demonstrated no significant differences between validation
and social contact or between validation and usual therapy
[164]. Five trials were included in a review on bright light
therapy [165]. This review revealed insufficient evidence of
the effectiveness of bright light in managing sleep, behav-
ior, cognitive, or mood disturbances associated with demen-
tia. A review to assess whether music therapy can diminish
behavioral and cognitive problems or improve social and
emotional functioning included five trials. However, there
was insufficient evidence to draw any useful conclusions
[166]. A review that assessed the efficacy of aroma therapy
as an intervention for persons with dementia concluded that
the one small trial provided insufficient evidence [167]. A
review of Snoezelen (multi-sensory stimulation) included
two trials that demonstrated some short-term benefits in
promoting adaptive behaviors during and immediately after
368 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
participation in the sessions. However, the carryover and
longer-term effects of Snoezelen were not evident [168]. A
review of the effectiveness of a range of massage and touch
therapies offered to persons with dementia is currently be-
ing conducted [169].
The lack of demonstrated effectiveness in all of these
reviews does not mean that the interventions are ineffective,
but rather that there is a lack of sufficient evidence to
demonstrate effectiveness. More experimental studies with
larger samples and improved quality are needed to further
examine the efficacy of these interventions with different
types and severity of dementia. Combining subjects with a
variety of dementias and at various stages of their illness
does not contribute to our understanding of the efficacy of
these interventions.
Most trials of cholinesterase inhibitors have been done
on those with a mild to moderate severity of their dementia.
Many of these trials have included caregiver-rated scales
that measure behavioral and psychiatric symptoms such as
the Neuropsychiatric Inventory (NPI) [170], the Cornell
Depression Rating Scale [171], the Behavioral Rating Scale
for Dementia [172], the noncognitive subscale of the ADAS
[173], the Behavioral Pathology in Alzheimer’s Disease
Rating Scale [174], the Gottries-Brane-Steen scale (GBS)
[175], and the Cohen-Mansfield Agitation Inventory [176].
Trinh et al [177] performed a meta-analysis that exam-
ined the efficacy of cholinesterase inhibitors in the treatment
of the neuropsychiatric symptoms and the functional im-
pairments that arise with AD. They reviewed 29 parallel
group or crossover randomized, double-blind, placebo-
controlled trials of patients with mild to moderate probable
AD. Patients randomized to cholinesterase inhibitors im-
proved on average 1.72 points on the NPI and 0.03 points on
the ADAS-noncog compared with those on placebo. There
were no significant differences between the various cho-
linesterase inhibitors. The authors concluded that there was
a modest improvement of neuropsychiatric symptoms with
the use of cholinesterase inhibitors.
Cummings et al [178] used a caregiver mailout survey to
compare 84 patients taking donepezil with 248 patients not
on it. They found that those taking donepezil were signifi-
cantly less likely to be threatening, destroy property, and
talk loudly. Although the treatment groups were not ran-
domly assigned, supportive evidence for donepezil having a
positive effect on difficult behaviors was the decreased rate
of use of sedatives in those on the agent. Winblad et al [179]
studied 286 patients with mild to moderate AD by using the
GBS scale (which includes various behavioral and emo-
tional items) and the NPI. There were nonsignificant im-
provements on the emotional/behavioral items of the GBS
and on the NPI seen in those patients receiving donepezil
compared with those allocated to the placebo group.
Holmes et al [180] examined the efficacy of donepezil in the
treatment of neuropsychiatric symptoms in AD by using the
NPI. After an initial open trial of donepezil, 134 patients
were randomized to either placebo or 10 mg of donepezil
and assessed at further intervals. Patients randomized to
donepezil after the open-label segment had significantly
more improvement on the NPI as well as on the NPI-distress
measure compared with the placebo group. Discordant re-
sults were obtained in the AD2000 study [181]. In this
randomized, placebo-controlled, double-blind trial the ef-
fects of donepezil on 565 community-residing patients with
mild to moderate AD were studied. The NPI was used. The
study design was complicated, and as a result, its outcomes
are difficult to evaluate. Although the authors concluded
that there was no significant improvement on any of their
cognitive, functional, or psychiatric measures, it is unclear
what conclusions can be drawn from this study. A Cochrane
review [49] of donepezil for AD evaluated the available
data on the effects of treatment on behavior (as measured by
the NPI). The authors concluded that there was a significant
benefit seen with donepezil. Compared with the placebo
arm, there was a mean difference of 6.20 at 6 weeks and 3.2
at 24 weeks favoring active therapy on the intent to treat
(ITT)–last observation carried forward (LOCF) analysis. An
article by Seltzer et al [182] was not included in the Co-
chrane review because they used an apathy scale [183] as
their behavioral measure. This trial was a double-blind,
24-week, placebo-controlled study of donepezil in 153 pa-
tients with early AD. The active treatment group scored
better on the apathy scale, although this was not statistically
significant.
Rösler al [184] used the CIBIC-Plus to measure the
effects of rivastigmine treatment compared with placebo in
98 patients with mild to moderate AD. This was a 26-week
randomized, placebo-controlled trial followed by an open-
label extension period. A variety of behavioral and psychi-
atric measures derived from the CIBIC-Plus showed sig-
nificant improvements with rivastigmine compared with
placebo. Finkel [185] published a meta-analysis of three
6-month, placebo-controlled trials of rivastigmine in mild to
moderate AD. He found that patients on rivastigmine with
neuropsychiatric symptoms at baseline exhibited significant
improvements with paranoid and delusional thoughts as
well as in aggression. Patients who did not have significant
neuropsychiatric symptoms at study entry were less likely to
see their emergence if treated with rivastigimine, but this
was not statistically significant for hallucinations.
Herrmann et al [186] performed a post hoc analysis of
pooled data from three large trials (2,033 subjects) of ga-
lantamine treatment for mild to moderate AD. They found
that compared with those on placebo, treatment with galan-
tamine was associated with a better total NPI score and
better scores on specific NPI items that included agitation/
aggression, anxiety, disinhibition, and aberrant motor be-
havior. Similar beneficial results were found in a priori
defined symptom clusters 1 (delusions, hallucinations), 3
(disinhibition, elation, aberrant motor behavior), and 4 (hal-
lucinations, anxiety, apathy, aberrant motor behavior). The
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
authors believed that the cluster of hallucinations, anxiety,
apathy, and aberrant motor behaviors might represent a
specific group of cholinergic-responsive behavioral symp-
toms. A Cochrane review of galantamine for AD and mild
cognitive impairment examined 10 trials with a total of
6,805 subjects [50]. Treatment effects on the NPI were
reported at 3 months in one trial and at 6 months in three
trials. There was no statistically significant treatment effect
at 3 months for the 24 to 32 mg per day dosage (OC and
ITT), but at 6 months there was a significant treatment effect
(OC and ITT) for 16 mg per day.
Areosa Sastre et al [187] performed a Cochrane review
of the use of memantine for dementia. They identified only
one reported trial for mild to moderate AD [188]. In this
trial, NPI scores at 24 weeks were significantly better (3.5
points) if the patients received memantine compared with
those allocated to the placebo arm. However, there were two
completed but unpublished studies that examined subjects
with mild to moderate AD. Their specific behavioral results
are not available for detailed evaluation. This makes it
difficult to make any conclusions at this time.
RCT data on the effects of cholinesterase inhibitors and
memantine on the BPSD come principally from secondary
outcome measures of trials primarily designed to measure
cognitive and global outcomes in subjects with a low fre-
quency and severity of BPSD. With few exceptions, these
studies were not designed to look at BPSD as the primary
outcome. The available data have to be interpreted with this
in mind. The benefits seen in the studies done to date might
not apply to patient populations with more severe baseline
neuropsychiatric abnormalities.
Antipsychotics, antidepressants, anticonvulsants, and
other medications have also been used for the treatment of
neuropsychiatric symptoms of dementia. However, most of
the trials of these agents have been conducted in more
severely impaired populations (see section on severe AD for
information and recommendations on specific agents).
Patients with psychotic features and a mild dementia
might be prescribed an antipsychotic. Patients with DLB
exhibit an abnormal sensitivity to antipsychotics, and they
should be avoided if possible [152]. The presence of visual
hallucinations and visuospatial/constructional dysfunction
(eg., problems copying the intersecting pentagons on the
MMSE) early in the course of the dementia would be
suggestive of DLB [189]. In this autopsy-confirmed study,
the positive predictive value for DLB was 83% for visual
hallucinations, whereas the lack of visuospatial impairment
had the best negative predictive value (90%). Treatment
options for visual hallucinations in the setting of DLB
would include cholinesterase inhibitors and/or a cautious
trial of an atypical neuroleptic [152]. Of the available atyp-
ical neuroleptics, quetiapine would appear to be an attrac-
tive choice, but controlled trials are needed [152,157].
369
6.5. Recommendation 23: Recommendations with regard
to the management of BPSD in the setting of a mild to
moderate dementia
A. Nonpharmacologic treatment of BPSD should be
considered first. Nonpharmacologic interventions are
often used in combination with pharmacotherapy
(Grade C, Level I).
B. Although there is insufficient evidence regarding the
effectiveness of the interventions to strongly advo-
cate for their routine use in the management of
BPSD, some persons with dementia might benefit
from the following: music, Snoezelen (multi-sensory
stimulation), bright light therapy, reminiscence ther-
apy, validation therapy, aroma therapy, and massage
and touch therapy (Grade C, Level II-3).
C. Pharmacotherapy for BPSD should be initiated only
after consideration, and usually a trial where appro-
priate, of nonpharmacologic interventions (Grade B,
Level III).
D. The presence of visual hallucinations in the setting of
mild dementia would suggest that the patient has
DLB. Patients with DLB are abnormally sensitive to
antipsychotics. If pharmacotherapy is required for
the visual hallucinations, a cholinesterase inhibitor
should be tried first if possible. If acute symptom
control is required or the cholinesterase inhibitor is
ineffective, a cautious trial of an atypical antipsy-
chotic (eg, very low dose quetiapine) can be at-
tempted (Grade B, Level II-2).
E. Medications for BPSD should normally be initiated at
a low starting dose and then subsequently titrated
carefully on the basis of the patient’s response and the
presence of adverse effects (Grade B, Level III).
F. There should be periodic attempts to taper and with-
draw medications after a period of 3 months of be-
havioral stability (Grade B, Level III).
G. Patients who have mild to moderate AD and neuro-
psychiatric symptoms can be considered for a trial of
a cholinesterase inhibitor and/or memantine for these
symptoms (Grade B, Level III).
H. Treatment of BPSD with cholinesterase inhibitors or
memantine should persist until clinical benefits can
no longer be demonstrated (Grade B, Level III).
Behavioral disturbances have been reported to occur in
63% of community-dwelling persons with dementia [190].
These disturbances increase distress for those with demen-
tia, increase the strain for caregivers, and might be poten-
tially dangerous for the care recipient, caregivers, and oth-
ers. Those with behavioral disturbances enter long-term
care facilities nearly 2 years earlier than those without
[191]. Dementia-related behavioral disturbances that most
frequently occur in the home setting are wandering, general
restlessness, agitation, and uncooperativeness [192].
370 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
In 1994, the Canadian Study on Health and Aging
(CSHA) [193] reported that half of those diagnosed with
dementia live in the community and are cared for by family
and friends who must deal with the long-term and disabling
behavioral problems associated with the care recipient’s
dementia. In 2000, the proportion of persons with dementia
who live at home had risen to 80% [194]. Although com-
munity services have been shown to be useful in prolonging
independence and quality of life of those with dementia, the
CSHA revealed that community services are underutilized.
Sixty-nine percent of spouse caregivers and 46% of care-
givers who were sons or daughters used no services, and
only 3% to 5% of caregivers used three or more services
[195]. Only 38% of Canadians with dementia received
home care services in 2000/2001 [196]. Several barriers to
using home care services have been identified: cultural and
ethnic factors, a reluctance to use formal services until
absolutely necessary, a perception that caregiving is a fa-
milial responsibility, acknowledging their family member
has dementia and they are unable to manage, lack of aware-
ness, acceptability and accessibility of services (eg, distance
especially in rural areas, costs, inadequate training of ser-
vice providers), and challenges in service delivery (care
often needed 24 hours a day) [197–199].
Although community-based programs have primarily fo-
cused on caregivers [200], in this section we will highlight
community-based programs and interventions that aim to
manage the behavioral disturbances associated with mild to
moderate dementia. Interventions that are of benefit to care-
givers are described elsewhere. Included studies in this
review had to meet the following criteria: (1) the majority of
subjects were diagnosed with AD or related dementia, (2)
the majority of subjects were classified as having mild to
moderate dementia (defined as MMSE score of 肁10 or a
Global Deterioration Scale score of 3 to 5 [201], and (3) the
sample size was at least 10. The term caregivers refers to
unpaid caregivers who are usually family or friends of the
care recipient and who provide support and assistance in the
activities of daily living and instrumental activities.
Case management involves an assessment of client and
caregiver needs and the development, implementation, and
monitoring of a care plan that can maintain a client safely in
the community. The care plan typically involves the ar-
rangement and coordination of a number of in-home and
community-based services such as housekeeping, personal
care, and respite services to supplement care already re-
ceived from family and friends. Case management activity
has been shown to be reactive and focused on dealing with
the consequences of the behavioral problems rather than
addressing the management of the behavior [192]. A dem-
onstration project randomly assigned individuals with de-
mentia and their caregivers to a treatment group that re-
ceived community-care services (eg, homemaking, personal
care, companion services, and adult day care) and case
management or to a control group that received regular
medical care [202,203]. The study investigated whether
increased access to community care and case management
led to a reduction in health care use and expenditures. Only
a tendency toward reduced health care expenditures was
observed in the treatment group. Case management did not
result in increased access to health care or to the prevention
of conditions that might increase expenditures. Perhaps case
manager involvement produces short-term increases in ex-
penditures, and a longer time period (greater than 3 years) is
required to demonstrate the benefits. The demonstration
study was also not designed to promote collaboration be-
tween the case manager and other health care practitioners
in identifying and managing high-risk people. Further re-
search is necessary to demonstrate and evaluate the role
of case managers in managing behavioral problems of
community-dwelling persons with dementia.
Adult day care programs provide supervised, structured
activities for the care recipients during the day, enabling
their caregivers to rest or tend to other responsibilities.
Several studies have evaluated the effect on behavior prob-
lems of persons with dementia. One study demonstrated that
behavioral difficulties such as wandering, agitation, and
anger were not found to predict use of adult day care
programs [204]. Several studies revealed that combined
family support (patient and family supported by one pro-
fessional staff member) while participating in an adult day
care program was more effective in decreasing behavior
problems (including inactivity and nonsocial behaviour),
improving mood, improving caregiver’s level of confidence,
and delayed nursing home placement when compared with
those who received psychogeriatric day care only. The sup-
port offered to caregivers included informational, practical,
emotional, and social support [205–207].
Support groups for caregivers are widespread in health
and voluntary organizations. However, the majority of ex-
perimental studies have been unable to demonstrate any
significant effect in decreasing caregiver burden and stress
or on patient behavioral and psychological symptoms [208 –
210]. Hébert et al [211] addressed some of the limitations of
these previous studies by offering the group program during
a longer period (15 two-hour weekly sessions), incorporat-
ing a specific theoretical framework, and focusing on the
management of behavioral problems and the reactions they
created. This experimental study demonstrated a 14% de-
crease in reactions to the behavioral problems of the care
receivers compared with a 5% decrease in the control group.
The frequency of behavioral problems also decreased. Fo-
cusing on immediate, tangible outcomes that are expected to
change as a result of the intervention (such as behavioral
problems) rather than more long-term, global outcomes
(such as preventing institutionalization and enhancing well-
being) are important lessons learned from this study.
More expensive home-based programs have also dem-
onstrated a significant effect in managing the disturbing
behaviors associated with dementia. In a Finnish study
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
Eloniemi-Sulkava et al [212] compared a 2-year program
that consisted of systematic, comprehensive support by a
dementia family care coordinator (nurse) with conventional
care. The coordinator addressed health problems of the care
recipient and caregiver, behavioral and psychological symp-
toms of the care recipient including restlessness, anger,
delusions, aggression, and depression, and caregiver stress
and burden. Educational courses were offered annually. The
support of the coordinator deferred placement in long-term
institutionalized care, especially for persons with severe
dementia. The authors recommend that the intervention be
targeted especially at persons with behavioral problems that
threaten their ability to continue to remain in their home.
A less intensive psychoeducational intervention, based
on the Progressively Lowered Stress Threshold Model
[213], was offered in home during two 2-hour sessions to
teach caregivers how to manage behavioral problems (eg,
structured routine, environment modifications, decreased
environmental stimuli) [214]. This study was conducted in
five states in the United States. The intervention had a
positive impact during a period of 12 months on the fre-
quency of and response to problem behaviors among spou-
sal caregivers, and nonspousal caregivers reported a reduc-
tion in the frequency of memory/behavioral problems. An
experimental study conducted in Taiwan demonstrated that
a two-session in-home caregiver training program and tele-
phone consultations every 2 weeks for 3 months decreased
physically nonaggressive behavior, verbally aggressive and
nonaggressive behavior (but not physical aggression), and
improved caregivers’ self-efficacy for managing these be-
haviors. The subjects’ level of severity of dementia is not
reported [215]. Other similar behavioral intervention studies
[216] have not been able to demonstrate a significant effect
in managing disturbing behaviors, possibly as a result of
their small sample sizes and short duration of the programs.
The remaining retrieved community-based studies did
not meet our inclusion criteria because the sample sizes
were less than 10 [217,218], or the subjects’ MMSE scores
were less than 10 [219].
6.6. Recommendation 24: Recommendations concerning
community-based programs for the management of
behavioral disturbances
For the following community-based programs for the
management of behavioral disturbances, there is limited
high quality evidence regarding effectiveness. The recom-
mendations are based on one to two RCTs for each program.
A. Adult day care (greater involvement of the caregiver
might decrease problem behaviors in the care recip-
ient) (Grade B, Level II-2).
B. Support groups that focus on the management of
behavioral problems and extend for a period of sev-
eral months (Grade B, Level I).
371
C. In-home systematic, comprehensive support by a
health care provider with advanced training in de-
mentia care during an extended period (ie, couple of
years) (Grade B, Level I).
D. In-home psychoeducational intervention that teaches
caregivers how to manage behavioral problems
(Grade B, Level I).
7. Driving
In North America, driving a motor vehicle is an impor-
tant aspect of modern culture that has almost become an
activity of daily living. The ability to drive is often needed
to maintain independent mobility [220]. It is also an expres-
sion of autonomy and independence and contributes to
many important aspects of a person’s quality of life, includ-
ing maintenance of family/social ties and participation in
recreational activities. For all individuals, including those
with physical, mental, or functional disabilities, the ability
to drive permits the continuation of independent living.
Loss of driver licensure for an older person can have a
direct health impact, with increases in depressive symptom-
atology [221]. Increased loneliness, social isolation, and
stress on family and friends have been linked to the loss of
the ability to drive in older persons [222–225], with this
impact tending to be greater for those in rural areas [226].
Even for those in urban areas, public transportation systems
do not adequately replace the mobility and freedom of
operating one’s own motor vehicle [227]. Older drivers
often feel angry and frustrated with those who deem them
unfit to drive, straining personal and professional relation-
ships [228]. However, as much as it is desirable to promote
personal independence, the safety implications of driving
from an individual and societal perspective require careful
consideration.
Given that many forms of dementia are progressive in
nature, and unlike other medical conditions, persons with
dementia often lack the insight to curtail their driving ex-
posure in the face of an increased crash risk, some authors
have advocated for the suspension of driving privileges in
all persons who are diagnosed with dementia [229]. How-
ever, some studies have shown that some older persons in
the early stage of dementia are able to safely operate a
motor vehicle [230,231]. Research and public policy have
been primarily focused on cognitive and visuoperceptual
deficits as they relate to driving [232–235]. Cognitive def-
icits affecting driving include memory impairment, poor
sequencing skills, impaired insight and judgment, apraxia,
and slowed processing time [236 –243]. Visuoperceptual
deficits are an important subset of cognitive skills directly
related to driving ability [244 –246].
7.1. Dementia and driving risk
Table 2 summarizes controlled studies that have exam-
ined driving risk in patients with dementia. A total of 25
372 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384

Table 2
Summary of controlled studies determining motor vehicle crash risk in persons with dementia [229 –232,282–302]
Study (Author, Date) Methodology Outcome Measure Used Main Findings

Waller et al [300], Case control: 99 AD patients; 495 unmatched State driving record No difference in crash rates
1993 controls
Cooper et al [284], Case control: 165 dementia patients; 165 age-, State driving record Dementia patients 2⫻ more likely to
1993 sex matched have had a crash
Tuokko et al [298], Case control: 249 AD patients; 249 age-, sex State driving record AD patients 2.5⫻ more likely to have
1995 matched controls had a crash
Trobe et al [230], 1996 Case control: 143 AD patients; 715 age-, sex State driving record No difference in crash rates
matched controls
Carr et al [231], 2000 Case control: 63 AD patients; 58 unmatched State driving record No difference in crash rates
controls
Drachman et al [287], Case control: 130 dementia patients; 112 Caregiver reported crashes Dementia patients 2.5⫻ more likely to
1993 unmatched controls have had a crash
Friedland et al [229], Case control: 30 AD patients; 20 age-matched Caregiver reported crash past 5 y AD patients 8⫻ more likely to have had
1988 controls a crash
Zuin et al [302], 2002 Case control: 56 dementia patients; 31 Caregiver reported crashes More frequent crashes in dementia
unmatched controls patients (P ⫽ .12)
Ott et al [294], 2003 Case control: 27 mild dementia patients; 40 Caregiver reported driving ability More frequent crashes in dementia
unmatched controls patients
Hunt [291], 1989 Case control: 12 questionable dementia patients; On road performance More mild dementia patients failed test
14 mild dementia patients 13 age matched (P ⬍ .05)
controls
Hunt [292], 1993 Case control: 12 very mild dementia patients; On road performance 5/13 mild failed test; all controls and
13 mild dementia patients; 13 unmatched very mild passed
controls
Rebok et al [295], Case control: 10 AD patients; 12 unmatched On road performance Worse performance in AD patients
1994 controls
Fitten et al [232], 1995 Case control: 25 mild dementia patients; 24 On road performance Worse performance in dementia patients
age-matched controls
Cushman [286], 1996 Case control: 32 early AD patients; 91 On road performance More AD patients failed test
unmatched controls
Hunt et al [293], 1997 Case control: 36 very mild AD patients; 29 On road performance Worse performance in AD patients
mild AD patients; 58 unmatched controls
Wald [299], 1998 Case control: 112 dementia patients; 50 On road performance Worse performance in dementia patients
unmatched controls
Duchek et al [288], Case control: 49 very mild AD patients; 29 On road performance Worse performance in AD patients
1998 mild AD patients: 58 unmatched controls
Bieliauskas et al [282], Case control: 9 AD patients; 9 age matched On road performance More driving errors in AD patients
1998 controls
Duchek et al [289], Case control: 21 very mild AD patients; 29 On road performance More rapid decline in driving skills in
2003 mild AD patients; 58 unmatched controls dementia patients
Whelihan [301], 2005 Case control: 23 mild dementia patients; 23 On road performance Worse performance in dementia patients
age-matched controls
Harvey [290], 1995 Case control: 13 dementia patients; 125 Driving simulator performance Worse performance in dementia patients
unmatched controls
Rizzo et al [296], 1997 Case control: 21 AD patients; 18 unmatched Driving simulator performance Higher crash rate in AD patients
controls (P ⫽ .02)
Cox et al [285], 1998 Case control: 29 AD patients; 21 age-matched Driving simulator performance Worse performance in AD patients
controls
Rizzo et al [297], 2001 Case control: 18 AD patients; 12 unmatched Driving simulator performance Higher crash rate in AD patients
controls (P ⬍ .05)
Carr et al [283], 1998 Case control: 70 dementia patients; 667 Traffic sign recognition test Worse performance in dementia patients
unmatched controls

studies were identified, all using a case-control design. In-
creased driving risk for patients with dementia was found in
two of five studies with state driving records as their out-
come, four of four with caregiver report of collisions/
driving ability, ten of ten with on-road driving performance,
four of four with driving simulator performance, and one of
one with the ability to recognize traffic signs. Similar results
were found for studies that examined patients with AD only.
All studies that used a form of driver performance (on-road
or driving simulator performance) as their outcome measure
found that drivers with dementia performed worse than
control subjects. However, when using state crash records,
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
the most objective and arguably the most relevant measure
of driving risk, increased driving risk for patients with
dementia was not consistently found. Collision rates as
recorded on state driving records might be a relatively
insensitive measure of driving risk and thus have limited
ability to detect differences in outcomes between drivers
with and without dementia. However, crash rates might be
the most important outcome measures because they are the
events of most concern to society, placing individual drivers
at most risk. It is possible that as the driving skills of
patients with dementia deteriorate, patients, their families,
and health professionals undertake measures to restrict driv-
ing exposure, thus reducing crash risk. No studies were
found examining the driving performance of persons with
mild cognitive impairment.
7.2. Magnitude of driving risk
It is difficult to precisely estimate the magnitude of
driving risk for persons with dementia. Of the two of five
studies that found a positive association between crashes
recorded on the state driving record and dementia, persons
with dementia had a 2 to 2.5 times increased risk. However,
factors such as severity of dementia and level of driving
exposure were not factored into these estimates. The latter is
important because persons with dementia drive fewer miles
per year compared with age- and sex-matched controls.
When tested on road and in driving simulators, it is very
clear that persons with dementia are poorer drivers com-
pared with those with normal cognition. This poorer perfor-
mance might not translate into consistently higher collision
rates per miles driven because of self, family, or authority
imposed driving restriction.
7.3. Cognitive functions and driving
Numerous studies have examined the association of spe-
cific neuropsychological tests and driving risk [247]. Brief
tests of general cognitive functioning such as the MMSE
[248] have shown an inconsistent relationship with driving
risk. Similar results were also found for studies specifically
examining tests of attention and concentration, visuospatial
skills, memory, executive functioning, and language. In
each of these domains, there were a few studies finding
positive associations between specific tests and driving risk,
but these results have not been confirmed in other studies.
There is no single brief cognitive test that has sufficient
validity, reliability, sensitivity, and/or specificity to be con-
sidered a robust tool in identifying older drivers with de-
mentia who are cognitively unfit to drive or need further
testing. Further research is necessary to determine whether
a combination of brief tests can meet this goal.
7.4. Assessment and follow-up
In many provinces and territories, physicians and other
clinicians are legally mandated to report drivers whom they
373
believe have medical conditions that might/will impact on
their ability to drive to their respective Ministries of Trans-
portation. Given that not all drivers with dementia (espe-
cially those with mild dementia) have higher crash rates
than drivers without dementia, a diagnosis of dementia is
not sufficient in itself to lead to automatic revocation of the
driver’s licenses. Rather, as recommended by a number of
organizations and groups [249,250], determination of the
functional driving abilities at the individual level is the
fairest and most appropriate method of assessing fitness to
drive in persons with mild dementia.
Given that many causes of dementia are progressive in
nature, most persons with dementia will eventually need to
give up driving. To lessen the impact of transition to non-
driving, planning for this inevitability should take place as
soon as the diagnosis of dementia is made. This planning
could include the development of alternative transportation
options and participation in driver cessation support groups.
For those with a progressive dementia initially deemed safe
to drive, progressive deterioration in driving skills can be
expected. Studies show that persons with mild dementia
who are initially deemed safe to drive are often found to be
unsafe 6 to 12 months later [251,252].
7.5. Compensation methods
Potential methods of compensating for the decreased
driver performance and increased crash risk in drivers with
dementia include the following.
7.5.1. Retraining/education programs
No studies were found assessing the efficacy of
retraining/education programs on improving the driving
performance in persons with dementia. Because persons
with dementia have underlying progressive memory and
cognitive deficits and often difficulties with insight and
judgment, attempts to upgrade their driving skills is not a
reasonable option.
7.5.2. Use of co-pilots
No studies were found assessing the efficacy of having
other persons accompany drivers with dementia with re-
spect to reducing their driving risk. Because many crashes
occur in a split second without time to give instructions to
drivers, this method of compensation would seem to be
ineffective.
7.5.3. Use of on-board navigation and crash
warning systems
No studies were found that assessed the efficacy of these
systems on improving the driving risk in persons with de-
mentia. Because the information processing of most persons
with dementia is impaired, these technologies are unlikely
to compensate for the driving deficiencies that drivers with
dementia demonstrate.
374 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
7.5.4. Restricted licensing
Although some studies have shown that, in general, re-
stricted licenses reduce crash rates, no studies were found
that specifically examined whether the granting of restricted
or conditional licensing to drivers with dementia reduces
their driving risk. Although many drivers with dementia
perform adequately in routine predicable situations, they do
not perform as well in situations that are less predictable,
precisely the time when many crashes occur. Therefore, it is
unlikely that the use of restricted/conditional licenses sig-
nificantly reduces crash risk in persons with dementia.
7.6. Recommendation 25: Recommendations with regard
to driving a motor vehicle and individuals with a mild to
moderate dementia
A. Clinicians should counsel persons with a progressive
dementia (and their families) that giving up driving will
be an inevitable consequence of their disease. Strategies
to ease this transition should occur early in the clinical
course of the disease (Grade B, Level II-2).
B. No single brief cognitive test (eg, MMSE) or combi-
nation of brief cognitive tests has sufficient sensitiv-
ity or specificity to be used as a sole determinant of
driving ability. Abnormalities on cognitive tests such
as the MMSE, clock drawing, and Trails B should
result in further in-depth testing of driving ability
(Grade B, Level III).
C. Driving is contraindicated in persons who, for cog-
nitive reasons, have an inability to independently
perform multiple instrumental activities of daily liv-
ing (eg, medication management, banking, shopping,
telephone use, cooking) or any of the basic activities
of daily living (eg, toileting, dressing) (Grade B,
Level III).
D. The driving ability of persons with earlier stages of
dementia should be tested on an individual basis
(Grade B, Level III).
E. A health professional– based comprehensive off-road
and on-road driving evaluation is the fairest method
of individual testing (Grade B, Level III).
F. In places where comprehensive off-road and on-road
driving evaluations are not available, clinicians must
rely on their own judgment (Grade B, Level III).
G. For persons deemed safe to drive, reassessment of
their ability to drive should take place every 6 to 12
months or sooner if indicated (Grade B, Level III).
H. Compensatory strategies are not appropriate for those
deemed unsafe to drive (Grade B, Level III).
8. Support of caregivers
In Canada about half of the individuals with dementia
reside in the community, and more than 90% are cared for
by family and friends [193,195]. Previous research has
characterized the negative consequences of dementia care-
giving or caregiver burden. Depression and anxiety are of
particular concern. All individuals caring for someone with
a chronic illness have been found to have increased rates of
psychiatric morbidity, but those caring for someone with
dementia are even more at risk, with rates reportedly as high
as 50% [253].
The primary theoretical model explaining the specific
effects of this burden is the stress/health model that explains
how the stress of caregiving is translated into psychiatric
and physical morbidity [254]. Factors associated with sig-
nificant caregiver burden and/or stress have been well-
described. These include factors related to the dementia
patient and those related to the caregiver themselves. Prob-
lem behaviors of the dementia patient such as depression
and aggression are the most important predictors of care-
giver burden. Less predictive of burden is the amount of
assistance with ADL provided to the patient by the care-
giver. Important caregiver factors that have been found to be
associated with caregiver burden include female gender,
low income, low life satisfaction, poor self-esteem and
self-assessed competence, and lack of social support. Lim-
ited research has been done to understand the positive as-
pects of caregiving. These might be of particular importance
because caregivers who could identify at least one positive
aspect of caregiving are less likely to report burden or
depression [255]. A thorough understanding of the negative
and positive aspects of caregiving is important not only to
reduce caregiver burden but also to reduce potentially neg-
ative outcomes for individuals with dementia. For example,
caregiving burden has been shown to be a risk factor for
nursing home placement [256].
Many intervention studies have been designed to reduce
caregiver burden and delay the negative outcomes of burden
such as nursing home placement. These interventions target
patient and caregiver factors that have been associated with
burden. These interventions include those designed to re-
duce the behavioral and psychological problems of the de-
mentia patient or the amount of assistance needed by the
demented patient. They also include strategies aimed pri-
marily at the caregiver to increase their knowledge about the
disease, provide additional resources and supports, reduce
anxiety and depression, or alter caregiver behavior.
Interventions have been delivered in a variety of modal-
ities either singly or in combination. Intervention formats
include those that are group based, individually based, tech-
nology based (telephone or computer), or service configu-
rations [257]. Broadly speaking, the interventions focus on
education, psychotherapy, or provision of services. Educa-
tion can include providing information about dementia or
skills training to change the caregivers’ interaction with the
dementia patient. Psychotherapy can include psychological
support for the caregiver, assistance to develop their social
support network, and self-care strategies. Services might
include access to respite care, day care, or a case manager.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
The most clinically relevant effects of these interventions
have been reductions in rates of caregiver depression and
delays in nursing home placement, although some might
argue that prolonging community care might not always be
a desirable outcome. Other positive effects include care-
giver satisfaction with services, improved self-appraisal of
caregiver coping skills and knowledge, and improved rela-
tionships with the care recipients. Common features of
successful interventions include (1) a continuing relation-
ship between the caregiver and helper over time, (2) a
variety of flexible interventions offered that can meet the
varied needs of caregivers, (3) psycho-education or skills
training that teaches caregivers to change their interactions
with patients with dementia, and (4) involvement of the
caregiver and patient in the intervention. Interventions that
appear unsuccessful include short educational programs that
only enhance knowledge about dementia, support groups
alone, and service configurations such as case management
or brief interventions that do not include long-term contact
with the caregiver [161,254,256 –258]. A Cochrane review
on respite care for people with dementia and their caregivers
found no evidence of either benefit or adverse effects from
the use of respite services. The authors believed their results
should be treated with caution because they might reflect the
dearth of high quality studies rather than any true lack of
benefit [259].
There is often a major discrepancy between the elaborate
programs that the literature tells us are effective and what is
locally available. Although these latter services might not
decrease caregiver burden, they can improve caregiver sat-
isfaction. Long-term involvement with these programs
tends to be more beneficial.
The studies examined were often limited by small sam-
ple sizes, short follow-up times, and the many caregiver and
patient outcomes that were studied. The heterogeneity of the
caregiver experience is clearly a factor limiting the effec-
tiveness of these interventions or the generalizability of
many studies. Targeting caregivers who meet specific cri-
teria such as scoring above a certain level for measures of
depression might improve outcomes in the future. Schultz
and Martire [254] recommended the assessment of care-
giver risk in five domains. Interventions can then be targeted
to those areas identified as problematic for individual care-
givers. The risk areas are safety, self-care and preventative
health behaviors, caregiver support (informational, instru-
mental, and emotional), depression and distress, and prob-
lem behaviors of care recipients.
A number of the RCTs of drug therapy for mild to
moderate AD have included caregiver burden as a second-
ary outcome measure. A recent systematic review of the
effect of cholinesterase inhibitors on burden and active time
use of caregivers of persons with AD concluded that they
had a small beneficial effect on both [260].
There have been a few guidelines and position state-
ments on dementia care published since the CCCD. The
375
American Academy of Neurology (AAN) published guide-
lines in 2001 that included two recommendations of rele-
vance for family caregivers [261]. One recommendation
was that short-term programs directed toward educating
caregivers about AD be offered to improve caregiver satis-
faction. In addition, the AAN recommended that intensive
long-term education and support services (when available)
be offered to caregivers to delay time to nursing home
placement. Listed as practice options (evidence supported
by expert opinion, case series, and studies with historical
controls only) were use of computer networks to provide
education and support to caregivers, telephone support pro-
grams, and respite services including adult day care.
The American Association of Geriatric Psychiatry issued
a statement on family caregivers of dementia patients in
2001 [262]. It emphasized the importance of caregivers in
the treatment of patients with dementia and recommended
that caregiver counseling be a reimbursable covered service.
The American Medical Association has also encouraged a
physician/caregiver/patient partnership in dementia care.
They have advocated that physicians monitor caregiver
functioning and have developed Web-based materials and
information to assist physicians in addressing caregiver
needs [263]. A caregiver self-assessment tool is available to
assist in identifying caregivers at risk of adverse health
outcomes. Caregivers who report high levels of distress in
completing this questionnaire are encouraged to get a
check-up from their physician, obtain respite from caregiv-
ing, and consider joining a support group. Physicians caring
for these caregivers are encouraged to discuss the need for
counseling and other interventions and to refer them to
available community services.
Financial hardships can be encountered in caring for a
person with dementia at home. Government plans might
provide a degree of financial relief. For example, in Canada
most patients with a dementia would be eligible for a dis-
ability tax credit. Physicians and other qualified practitio-
ners can alert caregivers of this potential tax benefit and can
help them in making an application by filling out Part B of
the Disability Tax Credit Certificate (Form T2201, available
at www.cra.gc.ca/forms). For details of this program please
refer to Information Concerning People With Disabilities
(guide RC4064) or visit the Canada Revenue Agency Web
site (www.cra.gc.ca/disability).
8.1. Recommendation 26: Recommendations with regard
to caregivers
A. The clinician should acknowledge the important role
played by the caregiver in dementia care. The clini-
cian should work with caregivers and families on an
ongoing basis and schedule regular appointments for
patients and caregivers together and alone (Grade B,
Level III).
376 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
B. The clinician should inquire about caregiver informa-
tion and support needs, provide education to patients
and families about dementia, and assist in recruiting
other family members and formal community ser-
vices to share the caregiving role. If available, refer
patients to specialized dementia services (eg, Alzhei-
mer Society, community-based dementia programs,
memory clinics) that offer comprehensive treatment
programs including caregiver support, education, and
training (Grade A, Level 1).
C. The clinician should inquire about caregiver health
(both physical and psychiatric), offer treatment for
these problems (including individual psychotherapy
or medications as indicated), and refer to appropriate
specialists (Grade B, Level III).
D. The clinician should enquire about problem behav-
iors of the dementia patient and the effect these
behaviours are having on the caregiver. If these are
causing significant caregiver distress, refer the care-
giver and patient to specialized dementia services
that can offer treatment to the patient and assist the
caregiver in modifying their interactions with the
patient (Grade A, Level 1).
E. Pharmacotherapy for AD can decrease caregiver bur-
den and the time required of caregivers to support the
care recipient. It should be considered as a means to
help support caregivers (Grade B, Level I).
F. Future studies of medications for the treatment of AD
and dementia should examine the impact of these
agents on caregiver burden and the time required to
support the care recipient. There is a need to ensure
consistency in the measurement of these outcomes
(Grade B, Level III).
9. Training needs and system issues
To implement the recommendations for the management of
the mild and moderate stages of AD, we need to ensure that
clinicians are adequately trained, needed resources and ser-
vices are available, and that we explore different models of
care. Please note that the literature referenced below originates
almost entirely outside Canada, and caution must be exercised
in extrapolating the findings reported to our country.
Dementia presents unique challenges to the clinician
[264]. Many health care providers remain uninformed about
AD [265,266]. Specific educational needs among primary
care physicians would include knowledge about local diag-
nostic and support services, development of assessment and
communication skills, management of behavioral problems,
and the coordination of support services [267]. A particular
attitudinal barrier is the skepticism of many primary care
practitioners about the effectiveness of interventions for
AD [268].
Studies on the uptake of dementia guidelines by physi-
cians have generally been disappointing, but the likelihood
of this occurring does appear to be modifiable [269 –271].
Guideline adherence is poor without specific interventions.
Common obstacles identified by Canadian physicians
to meeting the health care needs of their community
include inadequate time, inadequate remuneration, and
lack of accessible community resources [272]. We were
struck by the difference between what the literature tells
us works and what is available in our communities. The
availability of required community-based resources will be
critical in improving the care provided to those with demen-
tia and their families. There is some evidence that using
in-home help services earlier might delay institutionaliza-
tion [273].
The favored model for chronic disease management is
the delivery of services by multidisciplinary teams who
collaboratively educate, counsel, and empower patients with
self-care techniques to manage their chronic diseases. Sup-
ported by customized treatment plans and the multidisci-
plinary team, patients are charged with undertaking neces-
sary lifestyle and behavioral changes to manage their
condition responsibly. Chronic disease management is de-
pendent on the promotion of patient self-management and
clinician adherence to evidence-based guidelines [274].
This approach would require modification if used for de-
mentia. There would have to be reduced reliance on patient
self-care and a concomitant increased effort on caregiver
support and education. A chronic disease management ap-
proach, however, could address the need for comprehen-
sive, evidence-based management of the person with de-
mentia during the course of their illness. This could permit
the efficient use of the entire continuum of resources in-
cluding community-based and facility-based continuing
care [275]. Studies suggest that a more systematic approach
to the management of dementia can improve satisfaction
levels and enhance adherence to guidelines [276]. Whether
this approach for dementia improves patient outcomes,
however, remains unknown at the present time.
Shared care models are being explored as a way to deal
with chronic medical conditions. It can be defined as shared
responsibility, enhanced information exchange, continuing
medical education, and explicit clinical guidelines between
specialty services and primary care practitioners [277]. Al-
though it is being explored as a way to provide dementia
care, obstacles to effective collaboration include therapeutic
nihilism, risk reduction or avoidance, concerns about com-
petency, and limited access to required resources [278].
Another barrier to shared care models for dementia is that a
number of primary care physicians believe that dementia
care should be dealt with by specialists and not themselves
[267]. Reimbursement issues are another barrier to the de-
livery of high quality dementia care by physicians [279].
The complex, time-consuming aspects of dementia care are
not adequately reimbursed by fee-for-service payments.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
9.1. Recommendation 27: Recommendations with regards
to education
A. All clinicians caring for patients with mild to moderate
AD have to acquire the core knowledge and skills
required to manage this condition (Grade B, Level III).
B. A multifaceted educational program should be imple-
mented to promote adoption of the recommendations of
the 3rd CCCDTD by practitioners (Grade B, Level I).
9.2. Recommendation 28: Recommendations with regards
to the organization and funding of care for those
with a dementia
A. Every community should examine the services locally
available for the management of those with a dementia,
assess their adequacy, and implement plans to deal with
identified deficiencies (Grade C, Level III).
B. There is a need to modify the prevailing model of
chronic disease management (ie, less reliance on pro-
motion of patient self-management coupled with
greater caregiver involvement) for dementia. The ef-
ficacy and efficiency of modified chronic disease
management for dementia should be explored (Grade
C, Level III).
C. Shared care models for the management of patients
with mild to moderate AD and dementia should be
developed and evaluated. This will require the accep-
tance of joint responsibility on the part of primary
care practitioners and specialty services in delivering
care to patients with dementia (Grade C, Level III).
D. Dementia care must be adequately funded and reim-
bursed. Inadequate remuneration should not be a bar-
rier to the delivery of good dementia care (Grade C,
Level III).
Author Disclosures
David Hogan has been a site principle investigator in
studies supported by Janssen Ortho, Neurochem, Novartis
and Pfizer Canada within the last five years and given
presentations sponsored by Janssen Ortho, Merck Frosst,
Novartis and Pfizer within the last five years.
Peter Bailey has had support from pharmaceutical firms
(as a Speaker) and the C5R (President).
Krista Lanctôt has had support from Pfizer Canada (Con-
sultant, Speaker, Research Support), Abbott Laboratories
(consultant, research support), Janssen Ortho Inc. (Consul-
tant, Research Support), and Lundbeck Canada (Research
Support).
Lilian Thorpe has received support for research, advisory
boards and/or presentations from the following: Astra Zen-
eca, Bristol-Meyers Squibb, Eli Lilly, Glaxo SmithKline,
Janssen Ortho, Lundbeck, Novartis, Organon, Pfizer and
Wyeth.
377
Acknowledgments
The authors gratefully acknowledge numerous sources of
support and funding.
David Hogan is supported by the Brenda Strafford Foun-
dation Chair in Geriatric Medicine, University of Calgary.
Dorothy Forbes receives salary support as a CIHR New
Investigator.
Krista Lanctôt was supported by an Ontario Mental
Health Foundation New Investigator Fellowship.
Debra Morgan receives support from a CIHR New
Emerging Team grant Strategies to Improve the Care of
Persons with Dementia in Rural and Remote Areas funded
by CIHR (Institute of Aging, Institute of Health Services
and Policy Research, Rural and Northern Health Initiative)
and partners (Saskatchewan Health Research Foundation,
Alzheimer Society of Saskatchewan, and University of
Saskatchewan).
References
[1] Patterson CJS, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus
Conference on Dementia. CMAJ 1999;160(Suppl):S1–15.
[2] Canadian Task Force on Preventive Health Care. New grades for
recommendations from the Canadian Task Force on Preventive
Health Care. CMAJ 2003;169:207– 8.
[3] Elie M, Cole MG, Primeau FJ, Bellavance F. Delirium risk factors
in elderly hospitalized patients. J Gen Intern Med 1998;13:204 –12.
[4] Inouye SK, Bogardus ST, Charpentier PA, Leo-Summers L, Acam-
pora D, Holford TR, et al. A multicomponent intervention to prevent
delirium in hospitalized patients. N Engl J Med 1999;340:669 –76.
[5] Britton A, Russell R. Multidisciplinary team interventions for de-
lirium in patients with chronic cognitive impairment. Cochrane
Database Syst Rev 2004:CD000395.
[6] Hill JW, Futterman R, Duttagupta S, Mastey V, Lloyd JR, Fillit H.
Alzheimer’s disease and related dementias increase costs of co-
morbidities in managed Medicare. Neurology 2002;58:62–70.
[7] Schubert CC, Boustani M, Callahan CM, Perkins AJ, Carney CP,
Fox C, et al. Comorbidity profile of dementia patients in primary
care: are they sicker? J Am Geriatr Soc 2006;54:104 –9.
[8] Winkelmayer WC, Fischer MA, Schneeweiss S, Wang PS, Levin R,
Avorn J. Underuse of ACE inhibitors and angiotensin II receptor
blockers in elderly patients with diabetes. Am J Kidney Dis 2005;
46:1080 –7.
[9] Biessels GJ, van der Heide LP, Kamal A, Bleys RLAW, Gispen
WH. Ageing and diabetes: implications for brain function. Eur
J Pharmacol 2002;441:1–14.
[10] Kalmijn S, Feskens EJM, Launer LJ, Stijnen T, Kromhout D. Glu-
cose intolerance, hyperinsulinaemia and cognitive function in a
general population of elderly men. Diabetologia 1995;38:1096 –102.
[11] Cook DG, Leverenz JB, McMillan PJ, Kulstad JJ, Ericksen S, Roth
RA, et al. Reduced hippocampal insulin-degrading enzyme in late-
onset Alzheimer’s disease is associated with the apolipoprotein
E-epsilon4 allele. Am J Pathol 2003;162:313–9.
[12] Messier C, Gagnon M. Glucose regulation and cognitive functions:
relation to Alzheimer’s disease and diabetes. Behav Brain Res 1996;
75:1–11.
[13] Martin DC, Miller JK, Kapoor W, Arena VC, Boller F. A controlled
study of survival with dementia. Arch Neurol 1987;44:1122– 6.
378 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[14] Jagger C, Clarke M, Stone A. Predictors of survival with Alzhei-
mer’s disease: a community-based study. Psychol Med 1995;25:
171–7.
[15] Agüero-Torres H, Fratglioni L, Guo Z, Viitanen M, Winblad B.
Prognostic factors in very old demented adults: a seven-year
follow-up from a population-based survey in Stockholm. J Am
Geriatr Soc 1998;46:444 –52.
[16] Pereles L, Romonko L, Murzyn T, Hogan D, Silvius J, Stokes E,
et al. Evaluation of a self-medication program. J Am Geriatr Soc
1996;44:161–5.
[17] Biem HJ, Turnell RW, D’Arcy C. Computer telephony: automated
calls for medical care. Clin Invest Med 2003;26:259 – 68.
[18] Roe CM, Anderson MJ, Spivack B. Use of anticholinergic medica-
tions by older adults with dementia. J Am Geriatr Soc 2002;50:
836 – 42.
[19] Sunderland T, Tariot PN, Newhouse PA. Differential responsivity of
mood, behavior, and cognition to cholinergic agents in elderly neu-
ropsychiatric populations. Brain Res 1988;472:371– 89.
[20] Thienhaus OJ, Allen A, Bennett JA, et al. Anticholinergic serum
levels and cognitive performance. Eur Arch Psychiatry Clin Neuro-
sci 1990;240:28 –33.
[21] Kitano S, Morimoto S, Fukuo, et al. Circulating suppressing factor
for the muscarinic acetylcholine receptor in patients with senile
dementia of the Alzheimer type. Gerontology 1992;39(Suppl 1):
24 – 8.
[22] Mulsant BH, Gharabawi GM, Bossie CA, Mao L, Martinez RA,
Tune LA, et al. Correlates of anticholinergic activity in patients with
dementia and psychosis treated with risperidone or olanzapine.
J Clin Psychiatry 2004;65:1708 –14.
[23] Chew ML, Mulsant BH, Pollock BG. Serum anticholinergic activity
and cognition in patients with moderate-to-severe dementia. Am J
Geriatr Psychiatry 2005;13:535– 8.
[24] Ancelin ML, Artero S, Porter F, Dupuy A-M, Touchon J, Ritchie K.
Non-degenerative mild cognitive impairment in elderly people and
use of anticholinergic drugs: longitudinal cohort study. BMJ 2006;
332:455–9.
[25] Carnahan RM, Lund BC, Perry PJ, Chrischilles EA. The concurrent
use of anticholinergics and cholinesterase inhibitors: rare event or
common practice? J Am Geriatr Soc 2004;52:2082–7.
[26] Lu CJ, Tune LE. Chronic exposure to anticholinergic medications
adversely affects the course of Alzheimer disease. Am J Geriatr
Psychiatry 2003;11:458 – 61.
[27] Jewart RD, Green J, Lu CJ, Cellar J, Tune LE. Cognitive, behav-
ioral, and physiological changes in Alzheimer disease patients as a
function of incontinence medications. Am J Geriatr Psychiatry 2005;
13:324 – 8.
[28] Cahn-Weiner DA, Malloy PF, Rebok GW, Ott BR. Results of a
randomized placebo-controlled study of memory training for mildly
impaired Alzheimer’s disease patients. Applied Neuropsychology
2003;10:215–23.
[29] Lowenstein DA, Acevedo A, Czaja SJ, Duara R. Cognitive rehabil-
itation of mildly impaired Alzheimer Disease patients on cholines-
terase inhibitors. Am J Geriatr Psychiatry 2004;12:395– 402.
[30] Bottino CM, Carvalho IA, Alvarez AM, Avila R, Zukauskas PR,
Bustamante SE, et al. Cognitive rehabilitation combined with drug
treatment in Alzheimer’s disease patients: a pilot study. Clin Rehabil
2005;19:861–9.
[31] Clare L, Woods RT, Moniz Cook ED, Orrell M, Spector A. Cog-
nitive rehabilitation and cognitive training for early-stage Alzhei-
mer’s disease and vascular dementia. Cochrane Database Syst Rev
2003:CD003260.
[32] DeVreese LP, Neri M, Fiorvante M, Belloi L, Zanetti O. Memory
rehabilitation in Alzheimer’s disease: a review of progress. Int J
Geriatr Psychiatry 2001;16:794 – 809.
[33] Clare L, Roth I, Wilson B, Carter G, Hodges J. Relearning face-
name associations in early Alzheimer’s disease. Neuropsychology
2002;16:538 – 47.
[34] Farina E, Fioravanti R, Chiavari L, Imbornone E, Alberoni M,
Pomati G, et al. Comparing two programs of cognitive training in
Alzheimer’s diseases: a pilot study. Acta Neurologica Scandinavica
2002;105:365–71.
[35] Clare L, Wilson B, Carter G, Hodges J. Cognitive rehabilitation as
a component of early intervention in Alzheimer’s disease: a single
case study. Aging Ment Health 2003;7:15–21.
[36] Gitlin L, Corcoran M, Winter L, Boyce A, Hauck W. A randomized,
controlled trial of a home environmental intervention: effect of
efficacy and upset in caregivers and on daily function of persons
with dementia. Gerontologist 2001;41:4 –14.
[37] Gitlin LN, Hauck WW, Dennis MP, Winter L. Maintenance of
effects of the home environment skill-building program for family
caregivers and individuals with Alzheimer disease and related dis-
orders. J Gerontol Series A Biol Sci Med Sci 2005;60:368 –74.
[38] Day K, Carreon D, Stump C. The therapeutic design of environ-
ments for people with dementia: a review of the empirical research.
Gerontologist 2000;40:397– 413.
[39] Steyltjens EMJ, Dekker J, Bouter LM, Jellma S, Bakker EB, van den
Ende CHM. Occupational therapy for community dwelling elderly
people: as systematic review. Age Aging 2004;33:453– 60.
[40] Teri L, Gibbons LE, McCurry SM, Logsdon RG, Buchner DM,
Barlow WE, et al. Exercise plus behavioural management in patients
with Alzheimer disease: a randomized controlled trial. JAMA 2003;
290:2015–22.
[41] Patterson JT, Wessel J. Strategies for retraining functional move-
ment in persons with Alzheimer disease: a review. Physiotherapy
Canada 2002;54:274 – 80.
[42] Heyn P, Abreu BC, Ottenbacher JJ. The effects of exercise training
on elderly persons with cognitive impairment and dementia: a meta-
analysis. Arch Phys Med Rehabil 2004;85:1694 –704.
[43] Dooley NR, Hinojosa J. Improving quality of life for persons with
Alzheimer disease and their family caregivers: brief occupational
therapy intervention. Am J Occup Ther 2004;58:561–9.
[44] Scherder E, Bouma A, Steen AM. Effects of isolated transcutaneous
electrical nerve stimulation on memory and affective behaviour in
patients with probable Alzheimer’s disease. Biol Psychiatry 1998;
43:417–24.
[45] Cameron M, Lonergan E, Lee H. Transcutaneous electrical nerve
stimulation (TENS) for dementia. Cochrane Database Syst Rev
2003:CD004032.
[46] Bates J, Boote J, Beverley C. Psychosocial interventions for people
with milder dementing illness: a systematic review. J Adv Nurs
2004;45:644 –58.
[47] Birks J. Cholinesterase inhibitors for Alzheimer’s disease. Cochrane
Database Syst Rev 2006:CD005593.
[48] Birks J, Grimley Evans J, Iakovidou V, Tsolaki M. Rivastigmine for
Alzheimer’s disease. Cochrane Database Syst Rev 2000:CD001191.
[49] Birks J, Harvey R. Donepezil for dementia due to Alzheimer’s
disease. Cochrane Database Syst Rev 2006:CD001190.
[50] Loy C, Schneider L. Galantamine for Alzheimer’s disease and
mild cognitive impairment. Cochrane Database Syst Rev 2006:
CD001747.
[51] Lanctôt KL, Herrmann N, Yau KK, Khan LR, Liu BA, LouLou
MM, et al. Efficacy and safety of cholinesterase inhibitors in Alz-
heimer’s disease: a meta-analysis. CMAJ 2003;169:557– 64.
[52] Ritchie CW, Ames D, Clayton T, Lai R. Meta-analysis of random-
ized trials of the efficacy and safety of donepezil, galantamine, and
rivastigmine for the treatment of Alzheimer disease. Am J Geriatr
Psychiatry 2004;12:358 – 69.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[53] Rockwood K. Size of the treatment effect on cognition of cholines-
terase inhibition in Alzheimer’s disease. J Neurol Neurosurg Psy-
chiatry 2004;75:677– 85.
[54] Clegg A, Bryant J, Nicholson T, McIntyre L, De Broe S, Gerard K,
et al. Clinical and cost-effectiveness of donepezil, rivastigmine, and
galantamine for Alzheimer’s disease: a systematic review. Int J
Technol Assess Health Care 2002;18:497–507.
[55] Wolfson C, Oremus M, Shukla V, Momoli F, Demers L, Perrault A,
et al. Donepezil and rivastigmine in the treatment of Alzheimer’s
disease: a best-evidence synthesis of the published data on their
efficacy and cost-effectiveness. Clin Ther 2002;24:862– 86.
[56] Thompson S, Lanctôt KL, Herrmann N. The benefits and risks
associated with cholinesterase inhibitor therapy in Alzheimer’s dis-
ease. Expert Opin Drug Saf 2004;3:425– 40.
[57] Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt HP, van den
Bussche H. Cholinesterase inhibitors for patients with Alzheimer’s
disease: systematic review of randomised clinical trials. BMJ 2005;
331:321–7.
[58] Hogan DB, Goldlist B, Naglie G, Patterson C. Comparison studies
of cholinesterase inhibitors for Alzheimer’s disease. Lancet Neurol
2004;3:622– 6.
[59] Bullock R, Touchon J, Bergman H, Gambina G, He Y, Rapatz G,
et al. Rivastigmine and donepezil treatment in moderate to moder-
ately-severe Alzheimer’s disease over a 2-year period. Curr Med
Res Opin 2005;21:1317–27.
[60] Hogan DB, Patterson C. Progress in clinical neurosciences: treat-
ment of Alzheimer’s disease and other dementias—review and com-
parison of the cholinesterase inhibitors. Can J Neurol Sci 2002;29:
306 –14.
[61] Gillette-Guyonnet S, Cortes F, Cantet C, Vellas B. Long-term cho-
linergic treatment is not associated with greater risk of weight loss
during Alzheimer’s disease: data from the French REAL.FR cohort.
J Nutr Health Aging 2005;9:69 –73.
[62] Bordier P, Lanusse S, Garrigue S, Reynard C, Robert F, Gencel L,
et al. Causes of syncope in patients with Alzheimer’s disease treated
with donepezil. Drugs Aging 2005;22:687–94.
[63] Bordier P, Garrigue S, Barold SS, Bressolles N, Lanusse S, Clement
J. Significance of syncope in patients with Alzheimer’s disease
treated with cholinesterase inhibitors. Europace 2003;5:429 –31.
[64] Gill SS, Mamdani M, Naglie G, Streiner DL, Bronskill SE, Kopp A,
et al. A prescribing cascade involving cholinesterase inhibitors and
anticholinergic drugs. Arch Intern Med 2005;165:808 –13.
[65] Stahl SM, Markowitz JS, Gutterman EM, Papadopoulos G. Co-use
of donepezil and hypnotics among Alzheimer’s disease patients
living in the community. J Clin Psychiatry 2003;64:466 –72.
[66] Auriacombe S, Pere JJ, Loria-Kanza Y, Vellas B. Efficacy and
safety of rivastigmine in patients with Alzheimer’s disease who
failed to benefit from treatment with donepezil. Curr Med Res Opin
2002;18:129 –38.
[67] Emre M: Switching cholinesterase inhibitors in patients with Alz-
heimer’s disease. Int J Clin Pract Suppl 2002;127:64 –72.
[68] Gauthier S, Emre M, Farlow MR, Bullock R, Grossberg GT, Potkin
SG. Strategies for continued successful treatment of Alzheimer’s
disease: switching cholinesterase inhibitors. Curr Med Res Opin
2003;19:707–14.
[69] Alwaahhabi FK. Successfully switching acetylcholinesterase inhib-
itor therapy in probable Lewy body dementia. J Psychopharmacol
2005;19:214 – 6.
[70] Sadowsky CH, Farlow MR, Atkinson L, Steadman J, Koumaras B,
Chen M, et al. Switching from donepezil to rivastigmine is well
tolerated: results of an open-label safety and tolerability study. Prim
Care Companion J Clin Psychiatry 2005;7:43– 8.
[71] Wilkinson DG, Howe I. Switching from donepezil to galantamine: a
double-blind study of two wash-out periods. Int J Geriatr Psychiatry
2005;20:489 –91.
379
[72] Dantoine T, Auriacombe S, Sarazin M, Becker H, Pere J-J,
Bourdeix I. Rivastigmine monotherapy and combination therapy
with memantine in patients with moderately severe Alzheimer’s
disease who failed to benefit from previous cholinesterase inhib-
itor treatment. Int J Clin Pract 2006;60:110 – 8.
[73] Waldemar G, Kørner A, Hyvärinen M, Wetterberg P, Lehto H,
Josiassen MK. Tolerability of switching from donepezil to meman-
tine treatment in patients with moderate to severe Alzheimer’s
disease. Available at: http://www.lundbeck.com/investor/Pipeline/
ClinicalData/TOLERABILITY%20OF%20SWITCHING%20FROM%
20DONEPEZIL%20TO%20MEMANTINE%20TREATMENT%20
IN%20PATIENTS%20WITH%20MODERATE%20TO%20SEVERE
%20ALZHEIMER02B9S%20DISEASE%20Dajs%20%
20%20ICMPE%202005.pdf. Accessed July 2, 2007.
[74] Areosa Satre A, Sherriff F, McShane R. Memantine for dementia.
Cochrane Database Syst Rev 2005:CD003154.
[75] Alzheimer’s Association fact sheet on “Memantine (Namenda)”.
Updated January 10, 2006.
[76] Tariot PN, Farlow MR, Grossberg GT, Graham SM, McDonald S,
Gergel I, et al. Memantine treatment in patients with moderate to
severe Alzheimer disease already receiving donepezil: a randomized
controlled trial. JAMA 2004;291:317 Accessed 2 July, 2007.24.
[77] Doody RS, Geldmacher DS, Gordon B, Perdomo CA, Pratt RD.
Open-label, multicenter, phase 3 extension study of the safety and
efficacy of donepezil in patients with Alzheimer’s disease. Arch
Neurol 2001;58:427–3.
[78] Markesbery WR. Oxidative stress hypothesis in Alzheimer’s dis-
ease. Free Radic Biol Med 1997;23:134 – 47.
[79] Luchsinger JA, Mayeux R. Dietary factors and Alzheimer’s disease.
Lancet Neurol 2004;3:579 – 87.
[80] Butterfield DA, Castegna A, Drake J, Scapagnini G, Calabrese V.
Vitamin E and neurodegenerative disorders associated with oxida-
tive stress. Nutr Neurosci 2002;5:229 –39.
[81] Masaki KH, Losonczy KG, Izmirlian G, Foley DJ, Ross GW, Petro-
vitch H, et al. Association of vitamin E and C supplement use with
cognitive function and dementia in elderly men. Neurology 2000;
54:1265–72.
[82] Engelhart MJ, Geerlings MI, Ruitenberg A, Van Swieten JC,
Hofman A, Witteman JCM, et al. Dietary intake of antioxidants and
risk of Alzheimer disease. JAMA 2002;287:3223–9.
[83] Grodstein F, Chen J, Willett WC. High-dose antioxidant supplements
and cognitive function in community-dwelling elderly women. Am J
Clin Nutr 2003;77:975– 84.
[84] Sano M, Ernesto C, Thomas RG, Klauber MR, Schafer K, Grund-
man M, et al. A controlled trial of selegiline, alpha-tocopherol, or
both as treatment for Alzheimer’s disease. N Engl J Med 1997;336:
1216 –22.
[85] Petersen RC, Thomas RG, Grundman M, Bennett D, Doody R,
Ferris S, et al. Vitamin E and donepezil for the treatment of mild
cognitive impairment. N Engl J Med 2005;352:2379 – 88.
[86] Tabet N, Birks J, Grimley Evans J, Orrel M, Spector A. Vitamin E
for Alzheimer’s disease. Cochrane Database Syst Rev 2000:
CD002854.
[87] Miller ER, Pastor-Barriuso R, Dalai D, Riemersma RA, Appel LJ,
Guallar E. Meta-analysis: high-dosage vitamin E supplementation
may increase all-cause mortality. Ann Intern Med 2005;142:37– 46.
[88] Thal LJ, Grundman M, Berg J, Ernstrom K, Margolin R, Pfeiffer E,
et al. Idebenone treatment fails to slow cognitive decline in Alzhei-
mer’s disease. Neurology 2003;61:1498 –502.
[89] Leboeuf R. Homocysteine and Alzheimer’s disease. J Am Diet
Assoc 2003;103:304 –7.
[90] Seshadri S, Beiser A, Selhub J, Jacques PF, Rosenberg IH,
d’Agostino RB, et al. Plasma homocysteine as a risk factor for
dementia and Alzheimer’s disease. N Engl J Med 2002;346:476 – 83.
380 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[91] Kalmijn S, Launer LJ, Lindemans J, Bots ML, Hofman A, Breteler
MMB. Total homocysteine and cognitive decline in a community-
based sample of elderly subjects: the Rotterdam study. Am J Epi-
demiol 1999;50:283–9.
[92] Luchsinger JA, Tang MX, Shea S, Miller J, Green R, Mayeux R.
Plasma homocysteine levels and risk of Alzheimer disease. Neurol-
ogy 2004;62:1972– 6.
[93] Wang HX, Wahlin A, Basun H, Fastbom J, Winblad B, Fratiglioni
L. Vitamin B12 and folate in relation to the development of Alz-
heimer’s disease. Neurology 1991;56:1188 –94.
[94] Crystal HA, Ortof E, Frishman WH, Gruber A, Hershman D, Aron-
son M. Serum vitamin B12 levels and incidence of dementia in a
healthy elderly population: a report from the Bronx longitudinal
aging study. J Am Geriatr Soc 1994;42:933– 6.
[95] Nilsson K, Gustafson L, Hultberg B. Improvement of cognitive
functions after cobalamin/folate supplementation in elderly patients
with dementia and elevated plasma homocysteine. Int J Geriatr
Psychiatry 2001;16:609 –14.
[96] Kwok T, Tang C, Woo J, Lai WK, Law LK, Pang CP. Randomized
trial of the effect of supplementation on the cognitive function of
older people with subnormal cobalamin levels. Int J Geriatr Psychi-
atry 1998;13:611– 6.
[97] Rodriguez-Martin JL, Qizilbash N, Lopez-Arrieta JM. Thiamine for
Alzheimer’s disease. Cochrane Database Syst Rev 2001.CD001498.
[98] Malouf R, Grimley Evans J. Vitamin B6 for cognition. Cochrane
Database Syst Rev 2003.CD004393.
[99] Malouf R, Areosa Sastre A. Vitamin B12 for cognition. Cochrane
Database Syst Rev 2003.CD004326.
[100] Malouf R, Grimley Evans J, Aerosa Sastre A. Folic acid with or
without vitamin B12 for cognition and dementia. Cochrane Database
Syst Rev 2003.CD004514.
[101] Watkins RW. Herbal therapeutics: the top 12 selling remedies.
Emerg Med 2002;34:12–9.
[102] Kurz A, Van Baelen B. Ginkgo biloba compared with vholinesterase
inhibitors in the treatment of dementia: a review based on meta-
analyses by the Cochrane Collaboration. Dement Geriatr Cogn Dis-
ord 2006;18:217–25.
[103] Birks J, Gimley Evans J. Ginkgo biloba for cognitive impairment
and dementia. Cochrane Database Syst Rev 2002.CD003120.
[104] Bent S, Goldberg H, Padula A, Avins AL. Spontaneous bleeding
associated with ginkgo biloba: a case report and systematic review
of the literature: a case report and systematic review of the literature.
J Gen Intern Med 2005;20:657– 61.
[105] Vandenabeele P, Fiers W. Is amyloidogenesis during Alzheimer’s
disease due to an IL-1/IL-6 mediated acute phase response in the
brain? Immunology Today 1991;12:217–9.
[106] Aisen PS, Davis KL. Inflammatory mechanisms in Alzheimer’s
disease: implications for therapy. Am J Psychiatry 1994;151:
1105–13.
[107] Giulian D. Microglia and the immune pathology of Alzheimer dis-
ease. Am J Hum Genet 1999;65:13– 8.
[108] McGeer PL, Schulzer M, McGeer EG. Arthritis and anti-
inflammatory agents as possible protective factors for Alzheimer’s
disease: a review of 17 epidemiologic studies. Neurology 1996;47:
425–32.
[109] Wolfson C, Perrault A, Moride Y, Esdaile JM, Abenhaim L, Momoli
F. A case-control analysis of nonsteroidal anti-inflammatory drugs
and Alzheimer’s disease: are they protective? Neuroepidemiology
2002;21:81– 6.
[110] Launer LJ. Nonsteroidal anti-inflammatory drugs and Alzheimer
disease: what next? JAMA 2003;289:2865–7.
[111] Aisen PS. The potential of anti-inflammatory drugs for the treatment
of Alzheimer’s disease. Lancet Neurol 2002;1:279 – 84.
[112] Tabet N, Feldman H. Indomethacin for Alzheimer’s disease. Co-
chrane Database Syst Rev 2002.CD003673.
[113] Firuzi O, Praticò D. Coxibs and Alzheimer’s disease: should they
stay or should they go? Ann Neurol 2006;59:219 –28.
[114] Tabet N, Feldman H. Ibuprofen for Alzheimer’s disease. Cochrane
Database Syst Rev 2003.CD004031.
[115] Aisen PS, Davis KL, Berg JD, Schafer K, Campbell K, Thomas RG,
et al. A randomized controlled trial of prednisone in Alzheimer’s
disease. Neurology 2000;54:588 –93.
[116] van Gool WA, Weinstein HC, Scheltens PK, Walstra GJ. Effect of
hydroxychloroquine on progression of dementia in early Alzhei-
mer’s disease: an 18-month randomised, double-blind, placebo-
controlled study. Lancet 2001;358:455– 60.
[117] Heart Protection Study Collaborative Group. MRC/BHF Heart Pro-
tection Study of cholesterol lowering with simvastatin in 20,536
high-risk individuals: a randomized placebo-controlled trial. Lancet
2002;360:7–12.
[118] Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM,
Cobbe SM, et al. Pravastatin in elderly individuals at risk of vascular
disease (PROSPER): a randomised controlled trial. Lancet 2002;
360:1623–30.
[119] Sparks DL, Sabbagh MN, Connor DJ, Lopez J, Launer LJ, Browne
P, et al. Atorvastatin for the treatment of mild to moderate Alzhei-
mer disease: preliminary results. Arch Neurol 2005;62:753–7.
[120] Hogervorst E, Yaffe K, Richards M, Huppert F. Hormone replace-
ment therapy to maintain cognitive function in women with demen-
tia. Cochrane Database Syst Rev 2002.CD003799.
[121] Henderson VW, Hogervorst E. Testosterone and Alzheimer disease:
is it men’s turn now? Neurology 2004;62:170 –1.
[122] Almeida OP, Flicker L. Testosterone and dementia: too much ado
about too little data. J Br Menopause Soc 2003;9:107–10.
[123] Ramsden M, Shin TM, Pike CJ. Androgens modulate neuronal
vulnerability to kainite lesion. Neuroscience 2003;122:573– 8.
[124] Moffat SD, Zonderman AB, Metter EJ, Kawas C, Blackman MR,
Harman SM, Resnick SM. Free testosterone and risk for Alzheimer
disease in older men. Neurology 2004;62:188 –93.
[125] Tan RS, Pu SJ. A pilot study on the effects of testosterone in
hypogonadal aging male patients with Alzheimer’s disease. Aging
Male 2003;6:13–7.
[126] Cherrier MM, Matsumoto AM, Amory JK, Asthana S, Bremner W,
Peskind ER, et al. Testosterone improves spatial memory in men
with Alzheimer disease and mild cognitive impairment. Neurology
2005;64:2063– 8.
[127] Lyketsos CG, Lopez O, Jones B, Fitzpatrick AL, Breitner J,
DeKosky S. Prevalence of neuropsychiatric symptoms in dementia
and mild cognitive impairment: results from the cardiovascular
health study. JAMA 2002;288:1475– 83.
[128] Mega MS, Cummings JL, Fiorello T, Gornbein J. The spectrum of
behavioral changes in Alzheimer’s disease. Neurology 1996;46:
130 –5.
[129] Lopez OL, Becker JT, Sweet RA, Klunk W, Kaufer DI, Saxton J,
et al. Psychiatric symptoms vary with the severity of dementia in
probable Alzheimer’s disease. J Neuropsychiatry Clin Neurosci
2003;15:346 –53.
[130] Marin DB, Green CR, Schmeidler J, Harvey PD, Lawlor BA, Ryan
TM, et al. Noncognitive disturbances in Alzheimer’s disease: fre-
quency, longitudinal course, and relationship to cognitive symp-
toms. J Am Geriatr Soc 1997;45:1331– 8.
[131] Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al.
Development and validation of a geriatric depression screening
scale: a preliminary report. J Psychiatr Res 1982– 83;17:37– 49.
[132] Muller-Thomsen T, Arlt S, Mann U, Mass R, Ganzer S. Detecting
depression in Alzheimer’s disease: evaluation of four different
scales. Arch Clin Neuropsychol 2005;20:271– 6.
[133] Teri L, Logsdon RG, Uomoto J, McCurry SM. Behavioral treatment
of depression in dementia patients: a controlled clinical trial. J
Gerontol B Psychol Sci Soc Sci 1997;52:P159 – 66.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[134] Bains J, Birks JS, Dening TR. The efficacy of antidepressants in the
treatment of depression in dementia. Cochrane Database Syst Rev
2002.CD003944.
[135] Munro CA, Brandt J, Sheppard JM, Steele CD, Samus QM,
Steinberg M, et al. Cognitive response to pharmacological treatment
for depression in Alzheimer disease: secondary outcomes from the
depression in Alzheimer’s disease study (DIADS). Am J Geriatr
Psychiatry 2004;12:491– 8.
[136] Moretti R, Torre P, Antonello RM, Cazzato G, Bava A. Depression
and Alzheimer’s disease: symptom or comorbidity? Am J Alzhei-
mers Dis Other Demen 2002;17:338 – 44.
[137] Lyketsos CG, Olin J. Depression in Alzheimer’s disease: overview
and treatment. Biol Psychiatry 2002;52:243–52.
[138] Lechevallier-Michel N, Molimard M, Dartigues JF, Fabrigoule C,
Fourrier-Reglat A. Drugs with anticholinergic properties and cog-
nitive performance in the elderly: results from the PAQUID Study.
Br J Clin Pharmacol 2005;59:143–51.
[139] Fujishiro J, Imanishi T, Onozawa K, Tsushima M. Comparison of
the anticholinergic effects of the serotonergic antidepressants, par-
oxetine, fluvoxamine and clomipramine. Eur J Pharmacol 2002;454:
183– 8.
[140] Herrmann N. Use of SSRIs in the elderly: obvious benefits but
unappreciated risks. Can J Clin Pharmacol 2000;7:91–5.
[141] Liu B, Anderson G, Mittmann N, To T, Axcell T, Shear N. Use of
selective serotonin-reuptake inhibitors of tricyclic antidepressants
and risk of hip fractures in elderly people. Lancet 1998;351:1303–7.
[142] Pacher P, Ungvari Z. Selective serotonin-reuptake inhibitor antide-
pressants increase the risk of falls and hip fractures in elderly people
by inhibiting cardiovascular ion channels. Med Hypotheses 2001;
57:469 –71.
[143] American Geriatrics Society and American Association for Geriatric
Psychiatry. Consensus statement on improving the quality of mental
health care in U.S. nursing homes: management of depression and
behavioral symptoms associated with dementia. J Am Geriatr Soc
2003;51:1287–98.
[144] van der Wurff FB, Stek ML, Hoogendijk WJG, Beekman ATF. The
efficacy and safety of ECT in depressed older adults: a literature
review. In J Geriatr Psychiatry 2003;18:894 –904.
[145] Ancoli-Israel S, Ayalon L. Diagnosis and treatment of sleep disor-
ders in older adults. Am J Geriatr Psychiatry 2006;14:95–103.
[146] Carpenter BD, Strauss ME, Patterson MB. Sleep disturbances in
community-dwelling patients with Alzheimer’s disease. Clin Ger-
ontol 1995;16:35– 49.
[147] Gaugler JE, Edwards AB, Femia EE, et al: Predictors of institution-
alization of cognitively impaired elders: family help and the timing
of placement. J Gerontol B Psychol Sci Soc Sci 2000;55B:P247–55.
[148] McCurry SM, Gibbons LE, Logsdon RG, et al. Training caregivers
to change the sleep hygiene practices of patients with dementia: the
NITE-AD project. J Am Geriatr Soc 2003;51:455– 60.
[149] McCurry SM, Gibbons LE, Logsdon RG, et al. Nighttime insomnia
treatment and education for Alzheimer’s disease: a randomized,
controlled trial. J Am Geriatr Soc 2005;53:793– 802.
[150] Glass J, Lanctôt KL, Herrmann N, Sproule BA, Busto UE. Sedative
hypnotics in older people with insomnia: meta-analysis of risks and
benefits. BMJ 2005;331:1169.
[151] Bliwise DL. Sleep disorders in Alzheimer’s disease and other de-
mentias. Clin Cornerstone 2004;6(Suppl 1A):S16 –28.
[152] McKeith IG, Dickson DW, Lowe J, Emre M, O’Brien JT, Feldman
H, et al. Diagnosis and management of dementia with Lewy bodies:
third report of the DLB consortium. Neurology 2005;65:1863–72.
[153] Steriade M. Basic mechanisms of sleep generation. Neurology 1992;
42(Suppl 6):9 –18.
[154] Ringman JM, Simmons JH. Treatment of REM sleep behavior
disorder with donepezil: a report of three cases. Neurology 2000;
55:870 –1.
381
[155] Reading PJ, Luce AK, McKeith IG. Rivastigmine in the treatment of
parkinsonian psychosis and cognitive impairment: preliminary find-
ings from an open trial. Mov Disord 2001;16:1171– 4.
[156] Boeve BF, Silber MH, Ferman TJ. REM sleep behavior disorder in
Parkinson’s disease and dementia with Lewy bodies. J Geriatr Psy-
chiatry Neurol 2004;17:146 –57.
[157] Poewe W. Treatment of dementia with Lewy bodies and Parkinson’s
disease dementia. Mov Disord 2005;20(Suppl 12):S77– 82.
[158] Forbes DA. Strategies for managing the behavioral symptomatology
associated with dementia of the Alzheimer type: a systematic over-
view. Can J Nurs Res 1998;30:67– 86.
[159] Opie J, Rosewarne R, O’Connor DW. The efficacy of psychosocial
approaches to behaviour disorders in dementia: a systematic litera-
ture review. Aust N Z J Psychiatry 1999;33:789 –99.
[160] McMinn B, Draper B. Vocally disruptive behaviour in dementia:
development of an evidence based practice guideline. Aging Ment
Health 2005;9:16 –24.
[161] Verkaik R, van Weert JCM, Francke AL. The effects of psychoso-
cial methods on depressed, aggressive and apathetic behaviors of
people with dementia: a systematic review. Int J Geriatr Psychiatry
2005;20:301– 4.
[162] Livingston G, Johnston K, Katona C, Paton J, Lyketsos CG; Old
Age Task Force of the World Federation of Biological Psychiatry.
Systematic review of psychological approaches to the management
of neuropsychiatric symptoms of dementia. Am J Psychiatry 2005;
162:1996 –2021.
[163] Woods B, Spector A, Jones C, et al. Reminiscence therapy for
dementia. Cochrane Database Syst Rev 2005:CD001120.
[164] Neal M, Briggs M. Validation therapy for dementia. Cochrane
Database Syst Rev 2003.CD001394.
[165] Forbes D, Morgan DG, Bangma J, et al. Light therapy for managing
sleep, behaviour, and mood disturbances in dementia. Cochrane
Database Syst Rev 2004:CD003946.
[166] Vink AC, Birks JS, Bruinsma MS, et al. Music therapy for people
with dementia. Cochrane Database Syst Rev 2003:CD003477.
[167] Thorgrimsen L, Spector A, Wiles A, et al. Aroma therapy for
dementia. Cochrane Database Syst Rev 2003:CD003150.
[168] Chung JCC, Lai CKY. Snoezelen for dementia. Cochrane Database
Syst Rev 2002.CD003152.
[169] Viggo Hansen N, Jørgensen T, Ørtenblad L. Massage and touch for
dementia. Cochrane Database Syst Rev 2004.CD004989.
[170] Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi
DA, Gornbein J. The Neuropsychiatric Inventory: comprehensive
assessment of psychopathology in dementia. Neurology 1994;44:
2308 –14.
[171] Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. Cornell
Scale for Depression in Dementia. Biol Psychiatry 1988;23:271– 84.
[172] Tariot PN, Mack JL, Patterson MB, Edland SD, Weiner MF, Fil-
lenbaum G, et al: The Behavior Rating Scale for Dementia of the
Consortium to Establish a Registry for Alzheimer’s Disease: the
Behavioral Pathology Committee of the Consortium to Establish a
Registry for Alzheimer’s Disease. Am J Psychiatry 1995;152:
1349 –57.
[173] Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzhei-
mer’s disease. Am J Psychiatry 1984;141:1356 – 64.
[174] Reisberg B, Borenstein J, Salob SP, Ferris SH, Franssen E, Geor-
gotas A. Behavioral symptoms in Alzheimer’s disease: phenome-
nology and treatment. J Clin Psychiatry 1987;48(Suppl):9 –15.
[175] Gottfries CG, Brane G, Steen G. A new rating scale for dementia
syndromes. Gerontology 1982;28(Suppl 2):20 –31.
[176] Cohen-Mansfield J, Billig N. Agitated behaviors in the elderly. I. A
conceptual review. J Am Geriatr Soc 1986;34:711–21.
[177] Trinh NH, Hoblyn J, Mohanty S, Yaffe K. Efficacy of cholinesterase
inhibitors in the treatment of neuropsychiatric symptoms and func-
382 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
tional impairment in Alzheimer disease: a meta-analysis. JAMA
2003;289:210 – 6.
[178] Cummings JL, Donohue JA, Brooks RL. The relationship between
donepezil and behavioral disturbances in patients with Alzheimer’s
disease. Am J Geriatr Psychiatry 2000;8:134 – 40.
[179] Winblad B, Engedal K, Soininen H, Verhey F, Waldemar G, Wimo
A, et al. A 1-year, randomized, placebo-controlled study of done-
pezil in patients with mild to moderate AD. Neurology 2001;57:
489 –95.
[180] Holmes C, Wilkinson D, Dean C, Vethanayagam S, Olivieri S,
Langley A, et al. The efficacy of donepezil in the treatment of
neuropsychiatric symptoms in Alzheimer disease. Neurology 2004;
63:214 –9.
[181] Courtney C, Farrell D, Gray R, Hills R, Lynch L, Sellwood E, et al.
Long-term donepezil treatment in 565 patients with Alzheimer’s
disease (AD2000): randomised double-blind trial. Lancet 2004;363:
2105–15.
[182] Seltzer B, Zolnouni P, Nunez M, Goldman R, Kumar D, Ieni J, et al.
Efficacy of donepezil in early-stage Alzheimer disease: a random-
ized placebo-controlled trial. Arch Neurol 2004;61:1852– 6.
[183] Starkstein SE, Mayberg HS, Preziosi TJ, Andrezejewski P, Lei-
guarda R, Robinson RG. Reliability, validity, and clinical correlates
of apathy in Parkinson’s disease. J Neuropsychiatry Clin Neurosci
1992;4:134 –9.
[184] Rösler M, Retz W, Retz-Junginger P, Dennler HJ. Effects of two-
year treatment with the cholinesterase inhibitor rivastigmine on
behavioural symptoms in Alzheimer’s disease. Behav Neurol 1998;
11:211– 6.
[185] Finkel SI. Effects of rivastigmine on behavioral and psychological
symptoms of dementia in Alzheimer’s disease. Clin Ther 2004;26:
980 –90.
[186] Herrmann N, Rabheru K, Wang J, Binder C. Galantamine treatment
of problematic behavior in Alzheimer disease: post-hoc analysis of
pooled data from three large trials. Am J Geriatr Psychiatry 2005;
13:527–34.
[187] Areosa SA, Sherriff F, McShane R. Memantine for dementia. Co-
chrane Database Syst Rev 2005.CD003154.
[188] Peskind ER, Potkin SG, Pomara N, McDonald S, Xie Y, Gergel I.
Memantine monotherapy is effective and safe for the treatment of
mild to moderate Alzheimer’s disease. a randomized controlled trial.
American Association for Geriatric Psychiatry, 17th Annual Meet-
ing, Baltimore, Maryland, February 21–24, 2004.
[189] Tiraboschi P, Salmon DP, Hansen LA, Hofstetter RC, Thal LJ,
Corey-Bloom J. What best differentiates Lewy body from Alzhei-
mer’s disease in early stage dementia? Brain 2006;129:729 –35.
[190] Cohen-Mansfield J, Werner P, Watson V, et al. Agitation among
elderly persons at adult day-care centres: the experiences of relatives
and staff members. Int Psychogeriatr 1995;7:447–58.
[191] Phillips VL, Diwan S. The incremental effect of dementia-related
problem behaviors on the time to nursing home placement in poor,
frail, demented older people. J Am Geriatr Soc 2003;51:188 –93.
[192] Diwan S, Phillips VL. Agitation and dementia-related problem be-
haviors and case management in long-term care. Int Psychogeriatr
2001;13:5–21.
[193] Canadian Study of Health and Aging Working Group. Canadian
Study of Health and Aging. study methods and prevalence of de-
mentia. CMAJ 1994;150:899 –913.
[194] Czaja SJ, Eisdorfer C, Schulz R. Future directions in caregiving:
implications for intervention research. In: Schulz R, ed. Handbook
of dementia caregiving intervention research. New York: Springer
Publishing Company; 2000. p. 283–319.
[195] Canadian Study of Health and Aging Working Group. Patterns of
caring for people with dementia in Canada. Can J Aging 1994;13:
470 – 87.
[196] Forbes DA, Debra D, Janzen BL. Rural and urban Canadians with
dementia: use of health care services. Can J Aging 2006;25:321–30.
[197] Edwards DF, Baum CM, Meisel M, et al. Home-based multidisci-
plinary diagnosis and treatment of inner-city elders with dementia.
Gerontologist 1999;39:483– 8.
[198] Morgan D, Semchuk K, Stewart N, et al. Rural families caring for a
relative with dementia: barriers to use of formal services. Soc Sci
Med 2002;55:51– 64.
[199] Smale B, Dupuis S. A literature review. caregivers of persons with
dementia—roles, experiences, supports, and coping. Murray Alzhei-
mer Research and Education Program, University of Waterloo;
2004.
[200] Zarit SH, Gaugler JE, Jarrott SE. Useful services for families:
research findings and directions. Int J Geriatr Psychiatry 1999;14:
165–77.
[201] Gauthier S. Advances in the pharmacotherapy of Alzheimer’s dis-
ease. CMAJ 2002;166:616 –23.
[202] Newcomer R, Miller R, Clay T, et al. Effects of the Medicare
Alzheimer’s disease demonstration on Medicare expenditures.
Health Care Financ Rev 1999;20:45– 65.
[203] Newcomer R, Spitalny M, et al. Effects of the Medicare Alzheimer’s
Disease Demonstration on the use of community-based services.
Health Serv Res 1999;34:645– 67.
[204] Douglass C, Visconti C. Factors influencing the use of adult day care
by individuals with Alzheimer’s disease: a multivariate examination
of the California Alzheimer’s Disease Diagnostic and Treatment
Center Program. Home Health Care Serv Q 1998;17:53– 69.
[205] Droes RM, Breebaart E, Ettema T, et al. Effect of integrated family
support versus day care only on behavior and mood of patients with
dementia. Int Psychogeriatr 2000;12:99 –115.
[206] Droes RM, Breebaart E, Meiland FJM, et al. Effect of Meeting
Centres Support Program on feelings of competence of family carers
and delay of institutionalization of people with dementia. Aging
Ment Health 2004;8:201–11.
[207] Droes RM, Meiland F, et al. Effect of combined support for people
with dementia and carers versus regular day care on behaviour and
mood of persons with dementia: results from a multi-centre imple-
mentation study. Int J Geriatr Psychiatry 2004;19:673– 84.
[208] Berger G, Bernhardt T, Schramm U, et al. No effects of a combi-
nation of caregivers support group and memory training/music ther-
apy in dementia patients from a memory clinic population. Int J
Geriatr Psychiatry 2004;19:223–31.
[209] Gendron CE, Poitras LR, Dastoor DP, et al. Cognitive-behavior
group intervention for spousal caregivers: findings and clinical con-
siderations. Clin Gerontol 1996;17:3–19.
[210] Haley WE, Brown SL, Levine EG. Experimental evaluation of the
effectiveness of group intervention for dementia caregivers. Geron-
tologist 1987;27:376 – 82.
[211] Hébert R, Levesque L, Vezina J, et al. Efficacy of a psychoeducative
group program for caregivers of demented persons living at home: a
randomized controlled trial. J Gerontol B Psychol Sci Soc Sci
2003;58:S58 – 67.
[212] Eloniemi-Sulkava U, Notkola IL, Hentinen M, et al. Effects of
supporting community-living demented patients and their caregiv-
ers: a randomized trial. J Am Geriatr Soc 2001;49:1282–7.
[213] Hall GR, Buckwalter KC. Progressively lowered stress threshold: a
conceptual model for care of adults with Alzheimer’s disease. Arch
Psychiatr Nurs 1987;1:399 – 406.
[214] Gerdner LA, Buckwalter KC, Reed D. Impact of a psychoeduca-
tional intervention on caregiver response to behavioral problems.
Nurs Res 2002;51:363–74.
[215] Huang HL, Shyu YI, Chen MC, et al. A pilot study on a home-based
caregiver training program for improving caregiver self-efficacy and
decreasing the behavioral problems of elders with dementia in Tai-
wan. Int J Geriatr Psychiatry 2003;18:337– 45.
 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
[216] Gormley N, Lyons D, Howard R. Behavioural management of
aggression in dementia: a randomized controlled trial. Age Ageing
2001;30:141–5.
[217] Bourgeois MS, Burgio LD, Schulz R, et al. Modifying repetitive
verbalizations of community-dwelling patients with AD. Gerontol-
ogist 1997;37:30 –9.
[218] Rowe M, Alfred D. The effectiveness of slow-stroke massage in
diffusing agitated behaviors in individuals with Alzheimer’s disease.
J Gerontol Nurs 1999;25:22–34.
[219] Baker R, Bell S, Baker E, et al. A randomized controlled trial of the
effects of multi-sensory stimulation (MSS) for people with demen-
tia. Br J Clin Psychol 2001;40(Pt 1):81–96.
[220] Keplinger FS. The elderly driver: who should continue to drive?
Phys Med Rehabil 1998;12:147–55.
[221] Marottoli RA, Mendes de Leon CF, Glass TA, et al. Driving ces-
sation and increased depressive symptoms: prospective evidence
from the New Haven EPESE. J Am Geriatr Soc 1997;45:202– 6.
[222] Marottoli RA, Glass TA, Williams CS, Cooney LMJ, Berkman LF.
Consequences of driving cessation: decreased out-of-home activity
levels. J Gerontol 2000;55:S334 – 40.
[223] Johnson JE. Urban older adults and the forfeiture of a driver’s
license. J Gerontol Nurs 1999;12:12– 8.
[224] Bonnel WB. Giving up the car: older women’s losses and experi-
ences. J Psychosoc Nurs Mental Health Services 1999;37:10 –5.
[225] O’Neill D. Predicting and coping with the consequences of stopping
driving. Alzheimer Dis Assoc Disord 1997;11(Suppl 1):70 –2.
[226] Johnson JE. Older rural adults and the decision to stop driving: the
influence of family and friends. J Community Health Nurs 1998;15:
105–16.
[227] Rosenbloom S. Transportation needs of the elderly population. Clin
Geriatr Med 1993;9:297–310.
[228] Marshall S, Gilbert N. Saskatchewan physicians’ knowledge and
attitudes regarding assessment of medical fitness to drive. Can Med
Assoc J 1999;160:1701– 4.
[229] Friedland R, Koss E, Kumar A, et al: Motor vehicle crashes in
dementia of the Alzheimer type. Ann Neurol 1988;24:782– 6.
[230] Trobe JD, Waller PF, Cook-Flannagan CA, Teshima SM, Bieliaus-
kas LA. Crashes and violations among drivers with Alzheimer
disease. Arch Neurol 1996;53:411– 6.
[231] Carr DB, Duchek JMJ. Characteristics of motor vehicle crashes of
drivers with dementia of the Alzheimer type. J Am Geriatr Soc
2000;48:18 –22.
[232] Fitten LJ, Perryman KM, Wilkinson CJ, Little RJ, Burns MM,
Pachana N, et al. Alzheimer and vascular dementias and driving: a
prospective road and laboratory study. JAMA 1995;273:1360 –5.
[233] Galski T, Ehle HT, Bruno RL. An assessment of measures to predict
the outcome of driving evaluations in patients with cerebral damage.
Am J Occup Ther 1990;44:709 –13.
[234] Korteling JE, Kaptein MA. Neuropsychological driving fitness tests
for brain-damaged subjects. Arch Phys Med Rehabil 1996;77:
138 – 46.
[235] Odenheimer GL, Beaudet M, Jette AM, Albert MS, Grande L,
Minaker KL. Performance-based driving evaluation of the elderly
driver: safety, reliability, and validity. J Gerontol 1994;49:M153–9.
[236] Fox GK, Bashford GM, Caust SL. Identifying safe versus unsafe
drivers following brain impairment: the Coorabel Programme. Dis-
abil Rehabil 1992;14:140 –5.
[237] van Zomeren AH, Brouwer WH, Rothengatter JA, Snoek JW. Fit-
ness to drive a car after recovery from a severe head injury. Arch
Phys Med Rehabil 1988;69:90 – 6.
[238] Brooke MM, Questad KA, Patterson DR, Valois TA. Driving eval-
uation after traumatic brain injury. Am J Phys Med Rehabil 1992;
71:177– 82.
[239] Mazer BL, Korner-Bitensky NA, Sofer S. Predicting ability to drive
after stroke. Arch Phys Med Rehabil 1998;79:743–50.
383
[240] Gouvier WD, Maxfield MW, Schweitzer JR, et al. Psychometric
prediction of driving performance among the disabled. Arch Phys
Med Rehabil 1989;70:745–50.
[241] Kumar R, Powell B, Tani N, Naliboff B, Metter EJ. Perceptual
dysfunction in hemiplegia and automobile driving. Gerontology
1991;31:807–10.
[242] Uc E, Rizzo M, Anderson SW, Shi Q, Dawson JD. Driver landmark
and traffic sign identification in early Alzheimer’s disease. J Neurol
Neurosurg Psychiatr 2005;76:764 – 8.
[243] Galski T, Bruno RL, Ehle HT. Driving after cerebral damage: a
model with implications for evaluation. Am J Occup Ther 1992;46:
324 –32.
[244] Simms B. Perception and driving: theory and practice. Br J Occup
Ther 1985;48:363– 6.
[245] Sundet K, Goffeng L, Hofft E. To drive or not to drive: neuropsy-
chological assessment for driver’s license among stroke patients.
Scand J Psychol 1995;36:47–58.
[246] Sivak M, Olson P, Kewman DG, Won H, Henson DL. Driving and
perceptual/cognitive skills: behavioral consequences of brain dam-
age. Arch Phys Med Rehabil 1981;62:476 – 83.
[247] Molnar FJ, Patel A, Marshall SC, Man-Son-Hing M, Wilson KG.
Clinical utility of office-based cognitive predictors of fitness to drive
in persons with dementia: a systematic review. J Am Geriatr Soc
2006;54:1809 –24.
[248] Folstein JF, Folstein SE, McHugh PR. Mini-mental State: a practical
method for grading the cognitive state of patients for the clinician.
J Psychiatr Res 1975;12:189 –98.
[249] Dubinsky RM, Stein AC, Lyons K. Practice parameter: risk of
driving in Alzheimer’s disease. Neurology 2000;54:2205–11.
[250] Australian Society for Geriatric Medicine. Dementia and driving.
Accessed February 10, 2006. Available at: http://www.asgm.org.au/
pdfdocs/position_statements/PositionStatementNo11.pdf. Accessed
February 10, 2006.
[251] Hunt L, Morris JC, Edwards D, Wilson BS. Driving performance in
persons with mild senile dementia of the Alzheimer type. J Am
Geriatr Soc 1993;41:747–52.
[252] Fox GK, Bowden SC, Bashford GM, Smith DS. Alzheimer’s disease
and driving: prediction and assessment of driving performance.
J Am Geriatr Soc 1997;45:949 –53.
[253] Schulz R, O’Brien AT, Bookwala J, Fleissner K. Psychiatric and
physical morbidity effects of dementia caregiving: prevalence, cor-
relates and causes. Gerontologist 1995;35:771–91.
[254] Schulz R, Martire LM. Family caregiving of persons with dementia.
Am J Geriatr Psychiatry 2004;12:240 –9.
[255] Hébert R, Dubois MF, Wolfson C, Chambers L, Cohen C. Factors
associated with long-term institutionalization of older people with
dementia: data from the Canadian Study of Health and Aging. J
Gerontol Biol Sci Med Sci 2001;56:M693–9.
[256] Pusey H, Ricards D. A systematic review of the effectiveness of
psychosocial interventions for carers of people with dementia. Ag-
ing Ment Health 2001;5:107–19.
[257] Schulz R, O’Brien A, Czaja A, Ory M, Norris R, Martire LM, et al.
Dementia caregiver intervention research: In search of clinical sig-
nificance. Gerontologist 2002;42:589 – 602.
[258] Brodaty H, Green A, Koschera A. Meta-analysis of psychosocial
intervention for caregivers of people with dementia. J Am Geriatr
Soc 2003;51:657– 64.
[259] Lee H, Cameron M. Respite care for people with dementia and their
carers. Cochrane Database Syst Rev 2004.CD004396.
[260] Lingler JH, Martire LM, Schulz R. Caregiver-specific outcomes in
antidementia clinical drug trials: a systematic review and meta-
analysis. J Am Geriatr Soc 2005;53:983–90.
[261] Doody RS, Stevens JC, Beck C, Dubinsky RM, Kaye JA, Gwyther
L, et al. Practice parameter: management of dementia (an evidence-
384 D.B. Hogan et al. / Alzheimer’s & Dementia 3 (2007) 355–384
based review—report of the Quality Standards Subcommittee of the
American Academy of Neurology. Neurology 2001;56:1154 – 66.
[262] American Association of Geriatric Psychiatry. Family and caregiver
counseling in dementia: medical necessity. Available at: http://
www.aagponline.org/prf/position_dementia.asp. Accessed February
12, 2006.
[263] American Medical Association. Caregiver health. Available at:
http://www.ama-assn.org/ama/pub/category/5099.html. Accessed July
2, 2007.
[264] Foster NL. Barriers to treatment: the unique challenges for physi-
cians providing dementia care. J Geriatr Psychiatry Neurol 2001;14:
188 –98.
[265] Barrett JJ, Haley WE, Harrell LE, Powers RE. Knowledge about
Alzheimer disease among primary care physicians, psychologists,
nurses, and social workers. Alzheimer Dis Assoc Disord 1997;11:
99 –106.
[266] Pucci E, Angeleri F, Borsetti G, Brizioli E, Cartechini E, Giuliana G,
et al. General practitioners facing dementia: are they fully prepared?
Neurol Sci 2004;24:384 –9.
[267] Turner S, Iliffe S, Downs M, Wilcock J, Bryans M, Levin E, et al.
General practitioners’ knowledge, confidence and attitudes in the
diagnosis and management of dementia. Age Ageing 2004;33:
461–7.
[268] Olafsdottir M, Foldevi M, Marcusson J. Dementia in primary care:
why the low detection rate? Scand J Prim Health Care 2001;19:
194 – 8.
[269] Gifford DR, Holloway RG, Frankel MR, Albright CL, Meyerson R,
Griggs RC, et al. Improving adherence to guidelines through edu-
cation and opinion leaders: a randomized, controlled trail. Ann
Intern Med 1999;131:237– 46.
[270] Waldorff FB, Almind G, Makela M, Moller S, Waldemar G. Im-
plementation of a clinical dementia guideline: a controlled study on
the effect of a multifaceted strategy. Scand J Prim Health Care
2003;21:142–7.
[271] Cairns R, Evans J, Prince M. The impact of the NICE guidelines on
the diagnosis and treatment of Alzheimer’s disease among general
medical inpatients. Int J Geriatr Psychiatry 2004;19:800 –2.
[272] Pereles I, Russell ML. Needs for CME in geriatrics. Part 2: physi-
cian priorities and perceptions of community representatives. Can
Fam Physician 1996;42:632– 40.
[273] Gaugler JE, Kane RL, Kane RA, Newcomer R. Early community-
based service utilization and its effects on institutionalization in
dementia caregiving. Gerontologist 2005;45:177– 85.
[274] Wolff JL, Boult C. Moving beyond round pegs and square holes:
restructuring Medicare to improve chronic care. Ann Intern Med
2005;143:439 – 45.
[275] Fillit HM. Managed care decisions in Alzheimer disease. Am J
Managed Care 2000;6(Suppl):S1149 – 6.
[276] Cherry DL, Vickrey BG, Schwankovsky L, Heck E, Plauchm M,
Yep R. Interventions to improve quality of care: the Kaiser
Permanente-Alzheimer’s Association Dementia Care Project.
Am J Manag Care 2004;10:553– 60.
[277] Rockman P, Salach L, Gotlib D, Cord M, Turner T. Shared mental
health care: model for supporting and mentoring family physicians.
Am Fam Physician 2004;50:397– 402.
[278] Iliffe S, Wilcock J, Haworth D. Obstacles to shared care for patients
with dementia: a qualitative study. Fam Pract 2006;23:353– 62.
[279] Fillit H, Geldmacher DS, Welter RT, Maslow K, Fraser M. Opti-
mizing coding and reimbursement to improve management of Alz-
heimer’s disease and related dementias. J Am Geriatr Soc 2002;50:
1871– 8.
[280] Mintzer J, Burns A. Anticholinergic side-effects of drugs in elderly
people. J R Soc Med 2000;93:457– 62.
[281] Defilippi JL, Crismon ML. Drug interactions with cholinesterase
inhibitors. Drugs Aging 2003;20:437– 44.
[282] Bieliauskas LA, Roper BR, Trobe J, Green P, Lacy M. Cognitive
measures, driving safety and Alzheimer’s disease. Clin Neuropsy-
chol 1998;12:206 –12.
[283] Carr DB, Labarge E, Dunnigan K, Storandt M. Differentiating
drivers with dementia of the Azheimer type from healthy older
persons with a traffic sign recognition test. J Gerontol 1998;53:
M135–9.
[284] Cooper PJ, Tallman K, Tuokko H, Beattie BL. Vehicle crash in-
volvement and cognitive deficit in older drivers. J Safety Res 1993;
24:9 –17.
[285] Cox DJ, Quillian WC, Thorndike FP, Kovachev BP, Hanna G.
Evaluating driving performance of outpatients with Alzheimer Dis-
ease. J Am Board Fam Pract 1998;11:264 –71.
[286] Cushman L. Cognitive capacity and concurrent driving performance
in older drivers. IATSS Research 1996;20:38 – 45.
[287] Drachman DA, Swearer JM. Driving and Alzheimer’s disease: the
risk of crashes. Neurology 1993;43:2448 –56.
[288] Duchek JM, Hunt L, Ball K, Buckles V, Morris JC. Attention and
driving performance in Alzheimer’s disease. J Gerontol 1998;53:
130 – 41.
[289] Duchek JM, Carr DB, Hunt L, et al. Longitudinal driving perfor-
mance in early stage dementia of the Alzheimer type. J Am Geriatr
Soc 2003;51:1342–7.
[290] Harvey R, Fraser D, Bonner D, Warnes A, Warrington E, Rossor M.
Dementia and driving: results of a semi-realistic simulator study. Int
J Geriatr Psychiatr 1995;10:859 – 64.
[291] Hunt L. Dementia and road test performance. VTI Rapport 372A
1989.
[292] Hunt LA, Morris JC, Edwards D, Wilson BS. Driving performance
in persons with mild senile dementia of the Alzheimer type. J Am
Geriatr Soc 1993;41:747–52.
[293] Hunt LA, Murphy CF, Carr D, Duchek JM, Buckles V, Morris JC.
Reliability of the Washington University Road Test: a performance-
based assessment for drivers with dementia of the Alzheimer type.
Arch Neurol 1997;54:707–12.
[294] Ott BR, Heindel WC, Whelihan WM, Caron MD, Piatt AL, DiCarlo
MA. Maze test performance and reported driving ability in early
dementia. J Geriatr Psychiatr Neurol 2003;16:151–5.
[295] Rebok GW, Keyl PM, Blaustein MJ, Tune L. The effects of Alz-
heimer disease on driving-related abilities. Alzheimer Dis Assoc
Disord 1994;8:228 – 40.
[296] Rizzo M, Reinach S, McGehee D, Dawson J. Simulated car crashes
and crash predictors in drivers with Alzheimer disease. Arch Neurol
1997;54:545–51.
[297] Rizzo M, McGehee D, Dawson J, Anderson SN. Simulated car
crashes at intersection in drivers with Alzheimer disease. Alzheimer
Dis Assoc Disord 2001;15:10 –20.
[298] Tuokko H, Tallman K, Beattie BL, Cooper P, Weir J. An examina-
tion of driving records in a dementia clinic. J Gerontol 1995;50:
S173– 81.
[299] Wald JL. Neuropsychological impairment and driving performance
in dementia. PhD thesis, University of Alberta. 1998.
[300] Waller PF, Trobe JD, Olson PL. Crash characteristics associated
with early Alzheimer disease. Proc Assoc Advance Automotive Med
1993;1:157–73.
[301] Whelihan WM, Dicarlo MA, Paul RH. The relationship of neuro-
psychological functioning to driving competence in older persons
with early cognitive decline. Arch Clin Neuropsychol 2005;20:
217–28.
[302] Zuin D, Ortiz H, Boromei D, Lopez OL. Motor vehicle crashes and
abnormal driving behaviors in patients with dementia in Mendoza,
Argentina. Eur J Neurol 2002;9:29 –34.
 Alzheimer’s & Dementia 3 (2007) 385–397

Clinical practice guidelines for severe Alzheimer’s disease

Nathan Herrmanna,*, Serge Gauthierb, Paul G. Lysyc
a
Department of Psychiatry, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada
b
Alzheimer’s Disease Research Unit, McGill Center for Studies in Aging, and Department of Neurology, McGill University, Montreal, Quebec, Canada
c
Department of Family Medicine, McGill University, Montreal, Quebec, Canada

Abstract Background: Although severe Alzheimer’s disease (AD) represents a prevalent, serious, and
costly public health problem, few practice guidelines exist to help physicians manage this disorder.
Methods: A search of English language medical databases was performed from 1996 to the present
for articles pertaining to the management of AD. The focus of this review was on studies that
included patients with severe disease. Studies were assessed by considering the subjects, trial design,
analysis, and results. Recommendations were based on the best available evidence.
Results: Severe AD can be defined and diagnosed reliably by using measures of cognition,
function, behavior, and global staging. Specific assessments would also include medical status,
safety issues, and the health status of the caregiver. Disease-specific management would include
treatment with cholinesterase inhibitors and/or memantine. Treatment of neuropsychiatric symptoms
begins with nonpharmacologic behavioral and environmental approaches. Severe agitation, aggres-
sion, and psychosis that are potentially dangerous to the patient, caregiver, and others in the
environment can be treated with atypical antipsychotics, with consideration of their increased risk
of cerebrovascular adverse events and mortality. All pharmacologic approaches require careful
monitoring and regular periodic reassessments to determine whether ongoing treatment is necessary.
Conclusions: Evidence-based guidelines for the management of severe AD have the potential to
improve the quality of life for the patient and their caregiver. More randomized controlled trials
aimed specifically at this phase of illness are still urgently required.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Alzheimer’s disease; Severe dementia; Therapy; Practice guidelines

1. Introduction often considered a hallmark of the more severe stages of the

Alzheimer’s disease (AD) is characterized by a progressive
impairment of cognition, function, and behavior. Although
numerous reviews and clinical practice guidelines (CPGs) for
AD exist, most have focused on mild to moderate illness. Only
recently has focus shifted to more advanced stages of the
illness [1–5]. The Canadian Study of Health and Aging
(CSHA) estimated that 50% of individuals diagnosed with AD
were in the moderate to severe stages of the illness, and 90%
of residents in long-term care facilities were in moderate to
severe stages [6]. Disease severity increases with age [7] and
negatively affects survival [8]. With progression of the disease,
patients with AD become totally dependent on caregivers,
*Corresponding author. Tel.: 416-480-6133; Fax: 416-480-6022.
E-mail address: n.herrmann@utoronto.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.007
illness. This dependence leads to increased caregiver time,
increased caregiver burden and stress, and increased risk of
institutionalization [2]. The costs of caring for patients with
AD rise dramatically from mild to moderate to severe illness
[9], and the severe stages of the illness might account for up to
70% to 80% of the total costs [10].
The significant prevalence, disability, burden, and cost as-
sociated with severe AD clearly make this an important soci-
etal health problem. Development of CPGs for this stage of
illness was, therefore, recommended.
2. Methods
This article represents the background for CPGs for
severe AD as part of the Third Canadian Consensus Con-
ference on Diagnosis and Treatment of Dementia. A liter-
386 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
ature review was performed with PubMed and Embase
databases that included articles in English language, human
research, and publications from 1996 to the present. Specific
search terms included the following: dementia or Alzhei-
mer’s disease, severe, therapy or treatment, cholinesterase
inhibitors, antioxidants, hormones, anti-inflammatory drugs,
neurotropics, metabolic enhancers, neurotropic agents, me-
mantine, gingko, hypnotics, antidepressants, anxiolytics, se-
lective serotonin reuptake inhibitors, sleep, depression, anx-
iety, agitation, disinhibition, affective disorders, counseling,
education, caregiver environment, behavior, and rehabilita-
tion. We attempt to provide recommendations on the defi-
nition, assessment, and management of severe AD by using
an evidence-based approach and focusing particularly on
studies published since 1998 (the date of the last Canadian
Consensus Conference on Dementia). Recommendations
(Table 1) were rated on the basis of rules of evidence
developed by the Canadian Task Force on the periodic
health examination [11].
3. Results
3.1. Definition
There are no current widely accepted consensus criteria
for severe AD. In the only published CPGs for severe AD
[5], moderately severe AD was defined as a Mini-Mental
State Examination score (MMSE) [12] of 10 to 15, severe
AD 3 to 9, and very severe AD ⬍3. These authors noted that
although these cognitive function scores generally correlate
with increasing dependence and neuropsychiatric symp-
toms, this definition requires validation and clearly does not
encompass a rating of the functional capacity, behavior, and
medical status of the patient. Several other reviews have
also suggested that MMSE scores ⬍10 represent severe AD
[1,13–15].
Another approach is to define severe AD as a stage in
which the patient becomes dependent on a caregiver for
their survival. This approach encompasses not only cogni-
tive decline but also functional impairment and forms the
basis of several commonly used global staging instruments.
For example, the Global Deterioration Scale (GDS) [16]
defines moderately severe dementia as the stage in which
the patient can no longer survive without some assistance,
severe dementia as the stage when patients are entirely
dependent on a caregiver for survival, and very severe
dementia as the stage in which basic psychomotor and
verbal skills are lost and “the brain appears to no longer be
able to tell the body what to do.” Similarly, the Clinical
Dementia Rating Scale (CDR) [17,18] suggests that for
moderate and severe stages, there is no pretense of inde-
pendent function outside the home. In their stages, both
scales include clinical descriptors related to cognition (eg,
memory, orientation, language), function, and physical
characteristics (eg, incontinence, bedridden), whereas the
Table 1
Recommendations for the management of severe AD (levels of evidence
appear in parentheses)
1. Severe AD can be defined as the stage in which the patient
becomes totally dependent on a caregiver for survival. This will
typically correspond to MMSE ⬍10 and GDS 6 to7
(Grade B, Level 2).
2. Patients with severe AD should be assessed at least every 4 months
or if treated with pharmacotherapy at least every 3 months
(Grade C, Level 3).
3. Assessment should include cognition (eg MMSE) function behavior
medical status nutrition safety and caregiver health
(Grade B, Level 3).
4. The goals for management are to improve the quality of life for
patient and caregivers maintain optimal function and provide
maximum comfort (Grade B, Level 3).
5. Medical management includes treatment of intercurrent medical
conditions (eg infections parkinsonian symptoms seizures pressure
ulcers) ameliorating pain improving nutritional status and
optimizing sensory function (Grade B, Level 3).
6. Patients with severe AD can be treated with ChEIs memantine or
the combination. Expected benefits would include modest
improvements or slower decline in cognition function and behavior
(Grade A, Level 1).
7. Treatment with ChEIs and/or memantine should persist until
clinical benefit can no longer be demonstrated. Treatment should
not be discontinued simply because of institutionalization
(Grade C, Level 3).
8. The management of BPSD should begin with appropriate
assessments diagnosis and identification of target symptoms and
consideration of safety of the patient their caregiver and others in
their environment (Grade B, Level 3).
9. Nonpharmacologic treatments should be initiated first. Approaches
that might be useful for severe AD include behavioral management
for depression and caregivers/staff education programs for a variety
of behaviors. Music and multi-sensory intervention (Snoezelen) are
useful during treatment sessions but longer-term benefits have not
been demonstrated (Grade B, Level 1).
10. Pharmacologic interventions should be initiated concurrently with
nonpharmacologic approaches in the presence of severe depression
psychosis or aggression that puts the patient or others at risk of
harm (Grade B, Level 3).
11. Pharmacologic interventions for BPSD should be initiated at the
lowest doses titrated slowly and monitored for effectiveness and
safety (Grade B, Level 3).
12. Attempts to taper and withdraw medications for BPSD after a
period of 3 months of behavioral stability should occur in a
standardized fashion (Grade A, Level 1).
13. Risperidone and olanzapine can be used for severe agitation
aggression and psychosis. The potential benefit of all antipsychotics
must be weighed against the potential risks such as cerebrovascular
adverse events and mortality (Grade A, Level 1).
14. There is insufficient evidence to recommend for or against the use
of trazodone in the management of nonpsychotic agitated patients
(Grade C, Level 3).
15. Benzodiazepines should be used only for short periods as prn
agents (Grade B, Level 1).
16. Selective serotonin reuptake inhibitors can be used for the
treatment of severe depression (Grade B, Level 3).
17. If BPSD fails to improve after appropriate nonpharmacologic and
pharmacologic interventions refer to a specialty service
(Grade B, Level 3).
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
GDS also includes descriptors of behavior (eg, delusions,
agitation).
Because global staging is far more holistic, it would
theoretically be the preferred approach for defining severe
dementia. The advantages of defining severe dementia as an
MMSE score ⬍10 is the fact that this instrument is struc-
tured and already well-known to Canadian physicians, and
its use is already incorporated in drug benefit formulary
requirements for most provinces. Furthermore, studies have
suggested that MMSE scores are surprisingly strongly cor-
related with global staging scores [19]. If clinicians should
decide to rely on the MMSE scores for the definition of
severe AD, they must be aware of the limitations of this
scale’s use in this population, described in the following
section.
3.2. Assessment
Assessing patients with severe dementia requires a multi-
factorial approach that would include cognition, function,
behavior, medical status, safety, and medicolegal issues, as
well as caregiver status. Patients with severe AD live 3 to 4
[8] years during which numerous medical and psychosocial
crises might arise. It is, therefore, recommended that pa-
tients in this stage be assessed at least every 4 months.
3.2.1. Cognitive assessment
It is widely recognized the MMSE suffers from a floor
effect and is relatively insensitive to change in severe AD
when scores drop below 10 [15], and it can no longer be
used when language is severely affected [1]. The Severe
Impairment Battery (SIB) is a reliable and well-validated
cognitive assessment instrument that was developed specif-
ically for patients with severe dementia [20]. It examines a
broader range of cognitive functions and is better able to
grade variations of cognition in patients whose MMSE
score is ⬍15 [21]. Unfortunately, the SIB requires 20 to 30
minutes to administer, making it impractical for primary
care settings. A shorter version of the SIB has recently been
developed, although its use in clinical settings requires
further evaluation [22].
With the caveats noted above, the MMSE is therefore the
recommended cognitive assessment instrument at the
present time. It should be noted that clinical trials in mod-
erate to severe AD have incorporated the MMSE, with
benefits being noted in some of these studies [23–26].
3.2.2. Functional assessment
Because loss of independent function is a hallmark of
severe AD, functional scales that are heavily weighted to-
ward instrumental activities of daily living have limited
utility. The Alzheimer’s Disease Cooperative Study–Activi-
ties of Daily Living Inventory (ADCS-ADL) [27] is a well-
validated functional assessment scale that has been modified
for use in moderate to severe AD. The ADCS-ADLsev [28]
uses a subset of 19 items from the original version and has
already been used in clinical trials of moderate to severe
387
AD. Although not designed specifically for moderate to
severe AD, the Disability Assessment for Dementia (DAD)
[29], a 40-item functional inventory, has also been used
successfully in a clinical trial of moderately severe AD.
Both of these scales are fairly lengthy; therefore they would
be unsuitable for use in primary care settings. In contrast,
the ADL Index evaluates only six basic ADLs assessed for
degree of assistance required (none, some, or complete)
[30]. This scale has been recommended by previous severe
AD guidelines [5]. Finally, it must be noted that assessment
of ADL function in severe AD will need to consider the
context and environment in which the assessment occurs;
many patients, especially those in long-term care, will not
have the opportunity to demonstrate competence in many
areas, because their caregivers might assist or perform these
to facilitate care.
3.2.3. Global assessment
As noted previously, global assessment or staging can be
used for defining severe AD. The potential advantages of
global assessment include avoiding the floor effects of other
instruments, assessing longitudinal change and the effects
of treatment, and helping predict clinical course for care
planning [31].
Among the most commonly used global assessment
scales in clinical research are the GDS and CDR. The CDR
uses a semistructured interview that requires 30 to 40 min-
utes to administer and evaluates cognition (memory, orien-
tation, judgment/problem solving), community affairs,
home and hobbies, and personal care [17,18]. Because of
the time required to administer and the need for a semi-
structured interview, it is not suitable for use in primary care
and clinical practice. The GDS is a seven-stage instrument
in which stage 6 represents severe and stage 7 represents
very severe dementia [16]. The scale provides multiple
clinical anchors including cognition, function, and behavior.
It has excellent inter-rater and test/retest reliability [32] and
content validity [33] and correlates significantly with the
MMSE [19].
3.2.4. Behavioral assessment
The neuropsychiatric symptoms of dementia, also known
as the behavioral and psychological symptoms of dementia
(BPSD), become increasingly prevalent and disturbing in
severe AD [34]. Agitation and aggressive behaviors in-
crease in prevalence with increased disease severity and
occur in up to two thirds of patients with severe AD.
Aberrant motor behaviors such as restlessness, pacing, and
wandering are also common and particularly prevalent and
disturbing in residents of long-term care who remain am-
bulatory [1,34,35]. Depression, apathy, anxiety, delusions,
and hallucinations remain common in severe AD as well
[1]. Monitoring and assessment of BPSD are essential in
severe AD because these behaviors not only cause suffering
for the patient but are also associated with stress, burden,
and increased rates of depression in the caregiver [36 –38].
388 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
They are associated with more rapid cognitive and func-
tional decline and increased mortality [39 – 43] and are sig-
nificant risk factors for institutionalization [44,45].
A variety of rating scales for BPSD exist that can be used
in both long-term care patients and community-dwelling
individuals [46,47]. The Neuropsychiatric Inventory (NPI)
[48] measures the frequency and severity of 12 behaviors:
delusions, hallucinations, agitation, depression/dysphoria,
anxiety, euphoria/elation, apathy/indifference, disinhibition,
irritability/lability, aberrant motor behaviors, sleep and ap-
petite/eating disturbances. Other useful easily administered
scales include Cornell Depression Rating Scale [49] and the
Cohen-Mansfield Agitation Inventory [50]. The use of de-
pression rating scales has been shown to increase the sen-
sitivity of depression diagnosis in nursing home residents
[51], and systematic use of structured rating scales can also
reduce agitated aggressive behaviors in hospitalized demen-
tia patients [52].
3.2.5. Medical assessment
The medical assessments typically recommended by
CPGs for mild to moderate dementia often focus on reveal-
ing potentially reversible causes of dementia. In contrast,
the medical assessment of severe dementia should focus on
identifying medical causes of excess morbidity specifically
associated with this stage of illness. Any acute change in
cognition, function, or behavior should alert the clinician to
an intercurrent medical problem causing delirium and
should lead to appropriate physical and laboratory exami-
nations. The value, if any, of neuroimaging in severe de-
mentia has not been demonstrated.
Severe AD is often associated with progressive gait in-
stability, risk of falls, and parkinsonian symptoms. Myoc-
lonus and seizures can also occur [1]. Loss of ambulation
leading to being bedridden is associated with the develop-
ment of problems such as contractures and pressure ulcers
[53,54]. Continence needs to be monitored because incon-
tinence is associated with the decision to institutionalize
[55,56]. Weight and nutritional status should also be mon-
itored. Potential problems with constipation should be con-
sidered. Swallowing difficulties might appear in very severe
dementia, increasing the risk of aspiration, malnutrition, and
dehydration.
3.2.6. Other assessments
It is essential to assess the safety of the patient and their
caregiver. This requires review of issues such as aggression,
wandering, neglect, and abuse. Caregiver stress and health,
both physical and emotional, require ongoing monitoring
because they ultimately determine whether and when insti-
tutionalization is necessary.
3.3. Management
3.3.1. General management principles
The goal of management of severe AD is to improve the
quality of life for the patient and caregiver. This goal im-
plies therapeutic activism (ie, avoiding nihilism) and in-
volving caregivers not only in decision making but also in
therapeutic maneuvers as well. Although maintenance of
function, providing maximum comfort, and prolonging life
have been recognized as major goals [57], the latter might
not necessarily be a goal for end-stage disease. Palliative
and end-of-life care have been reviewed elsewhere [58 – 60]
and will be discussed in other guidelines in this series.
Medical management includes treating intercurrent medical
conditions such as infections, seizures, parkinsonian symp-
toms, and pressure ulcers. The detection of pain can be
challenging, although treatment is essential and evidence
suggests that pain is currently undertreated in these patients
[61]. Malnutrition can be treated with nutritional supple-
ments and extra care during meals [62]. Optimizing hearing
and vision can be important [63]. The clinician should
ensure caregivers are provided with education and support
including linkages to the local Alzheimer Society Chapter.
Home-care and day programs should be used. Finally, help-
ing caregivers around decisions regarding institutionaliza-
tion is an important function at this stage of disease.
3.4. Disease-specific therapies
Despite the increased attention, there are remarkably few
randomized controlled trials (RCTs) that have focused on
cognitive enhancement or disease modification for severe
AD. This is a concern because data from treatment trials of
mild to moderate AD might not necessarily translate to
severe illness, given emerging data that suggest differences
in neuropathology and neurochemistry at different stages.
For example, stage-specific changes in neuritic plaques and
neurofibrillary tangles have been documented [64,65] as
well as changes in neurotransmitters such as catecholamines
[66,67], ␥-aminobutyric acid [68], and glutamate [69]. Per-
haps the best example of stage-specific neurotransmitter
changes is the cholinergic system, in which cholinergic
markers, such as choline acetyltransferase, do not decline
significantly until later stages of AD, and milder stages are
characterized by persevered choline acetyltransferase
[70,71]. These stage-specific changes suggest that treat-
ments studied in milder stages of AD might not necessarily
be effective in severe AD.
3.4.1. Cholinesterase inhibitors
There are three cholinesterase inhibitors (ChEIs) ap-
proved for use in Canada for the treatment of mild to
moderate AD: donepezil, galantamine, and rivastigmine.
These agents have demonstrated consistent, significant, but
modest improvements in cognition, global assessment, and
function in patients with MMSEs that range from 10 to 26
[72]. There is also evidence of benefit on neuropsychiatric
symptoms from these trials [73] and from one trial designed
specifically with behavior as the primary outcome measure
[74]. It is interesting to note that post hoc analyses of some
of the mild to moderate AD ChEI trials suggested greater
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
benefit in patients with moderate compared with mild illness
[75–77].
There are four published double-blind randomized pla-
cebo-controlled trials of ChEIs that included patients with
severe AD. Donepezil was studied for 24 weeks in 290
patients with moderate to severe AD whose average base-
line MMSE score was 12, and average baseline SIB scores
were 79 to 80 [23]. Inclusion criteria for this study were
MMSE scores 5 to 17, although 50% of enrolled patients
scored 5 to 12 on baseline MMSE, and 86% of these were
rated as severe dementia on global staging. Treatment with
donepezil was associated with statistically significant ben-
efits on cognition (as measured by both SIB and MMSE),
global outcome, function, and behavior (as measured by the
NPI). Similarly, in a post hoc analysis of the patients in this
trial with severe dementia (MMSE 5 to 12, n ⫽ 145),
donepezil treatment demonstrated statistically significant
benefits on global, cognition, functional, and behavioral
measures [78]. With respect to neuropsychiatric symptoms,
there was a 5.5-point treatment difference in favor of done-
pezil (P ⫽ .006), with significant improvement noted in
depression/dysphoria, anxiety, and apathy.
In another randomized double-blind placebo-controlled
24-week trial of donepezil, 208 nursing home residents were
enrolled, with inclusion criteria that included baseline
MMSE scores of 5 to 26 and at least one neuropsychiatric
symptom on the NPI that occurred several times per week
[24]. The average MMSE score was 14, although only 22%
to 26% of the sample had severe dementia. The primary
outcome measure, neuropsychiatric symptoms scores on the
NPI, was not affected by treatment, although secondary
measures of ADL function and disease stage improved
significantly with donepezil compared with placebo. Suba-
nalysis of the patients with severe AD was not reported.
In a randomized placebo-controlled study with exclu-
sively severe AD patients, rivastigmine was compared with
quetiapine in 93 patients in a 26-week trial [79]. The inclu-
sion criteria for this relatively small study focused on the
presence of agitation, irritability, and/or aberrant motor be-
haviors in institutionalized patients. The resultant average
baseline SIB score was 62, making this one of the more
cognitively impaired patient samples to date. Although
there were no statistically significant differences at 6 or 26
weeks between rivastigmine and placebo, placebo-treated
patients (n ⫽ 19) improved on average by 3.3 points on the
SIB compared with 3.1-point worsening for patients on
rivastigmine (n ⫽ 15). Besides the very small sample size,
placebo-treated patients had a 10-point average higher SIB
score at baseline compared with the rivastigmine patients,
possibly biasing these results.
Finally, in the most recent double-blind RCT, AD pa-
tients with MMSE scores of 1 to 10 living in nursing homes
were treated with either donepezil or placebo. After 6
months of treatment, donepezil-treated patients improved
statistically more on cognitive measures and declined less
389
on ADLs compared with placebo [80]. There was no sig-
nificant benefit noted on behavior in this study. Although
the authors argued that their results suggest that even in
institutionalized AD patients with severe dementia, donepe-
zil improves well-being and can facilitate caregiving, others
have suggested that it is still unclear how clinically mean-
ingful these results are [81].
Several post hoc analyses of rivastigmine- and galantamine-
treated patients with MMSE scores ⱕ14 have been reported,
suggesting improved cognition and functional benefits com-
pared with placebo [77,82,83]. Similarly in a randomized con-
trolled head-to-head study of rivastigmine and donepezil, 43%
of the total 994 patients included had MMSE scores of 10 to 14
at baseline [84]. The declines in SIB and MMSE scores were
similar with both drugs and significantly smaller than would be
expected from patients in untreated naturalistic studies. None
of these studies, however, would have included patients de-
fined as severe by MMSE scores ⬍10. In a 52-week open-label
study of 173 nursing home residents with moderately severe
AD, the effects of rivastigmine were recorded [85]. Average
baseline MMSE was 9.2. After 52 weeks the average change
from baseline was an improvement of 0.1 point, and a dose-
dependent effect was noted (patients on 12 mg per day im-
proved significantly by 1.4 points; P ⫽ .021). Furthermore,
patients in the study experienced improvement in most behav-
iors measured by the NPI, and 40% of patients who had been
treated with psychotropics for BPSD discontinued or reduced
the dosage of their medication by week 52.
In summary, although use of a ChEI in severe AD is
supported by at least one well-designed RCT, not all results
are positive, and more confirmatory studies in patients ex-
clusively with severe AD are required.
3.4.2. Memantine
The best studied treatment for severe AD is memantine,
a noncompetitive N-methyl-D-aspartate receptor antagonist.
Three RCTs have been published to date. In a 12-week
randomized double-blind placebo-controlled trial, meman-
tine was studied in 166 patients with MMSE ⬍10 and GDS
5 to 7 [86]. Average MMSE scores were 6, with 96% of
patients being staged as severe or very severe, although only
about half the samples were diagnosed with AD. Global
outcomes, behavior, and function all improved significantly
with memantine. In a larger 28-week double-blind con-
trolled trial, 252 patients with moderate to severe AD were
treated with either memantine or placebo. Inclusion criteria
included MMSE scores of 3 to 14 and GDS 5 or 6 [25].
Average baseline MMSE score was 8, with 56% staged as
moderate to severe (GDS, 6). Global outcome demonstrated
a trend in favor of memantine (P ⫽ .06), with statistically
significant benefits noted on function and cognition. These
two studies have been analyzed with a number needed to
treat (NNT) analysis [87]. A responder analysis from the
study by Winblad and Poritis [86] was statistically signifi-
cant with NNTs of three and four. In the study by Reisberg
390 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
et al [25], NNT for global improvement or stabilization and
one secondary outcome (ADL or SIB) was 6 (P ⫽ .004;
95% confidence interval, 4 to 12). Interestingly, the num-
bers needed to harm for adverse events were similar to
placebo except for agitation, in which placebo-treated pa-
tients actually had a significantly higher risk. These authors
concluded that on the basis of the low NNTs, a medium
effect size, and an extremely benign safety profile, meman-
tine’s benefit in moderate to severe AD was of similar
magnitude to that of the ChEIs in mild to moderate AD [87].
In a third study, 404 patients with moderate to severe AD
were randomized to either memantine or placebo for 24
weeks [26]. Inclusion criteria included MMSE 5 to 14 and
having received a stable dose of donepezil for at least 6
months before randomization. Average baseline MMSE in
these patients was 10. All outcome measures including
cognition (SIB), global function, and behavior (NPI) statis-
tically favored memantine.
In a post hoc analysis of the studies by Reisberg et al [25]
and Tariot et al [26], behavioral improvements favored
memantine in both studies and reached statistical signifi-
cance in the latter [88]. In both studies, agitation/aggression
improved significantly with memantine therapy, perhaps
suggesting a slightly different profile of behavioral benefits
compared with the ChEIs.
In summary, use of memantine in severe AD is supported
by at least one high quality RCT as monotherapy or in
combination with a ChEI.
3.4.3. Selegiline/␣-tocopherol
In a double-blind placebo-controlled trial, 341 moder-
ately severe AD patients were treated with ␣-tocopherol
(vitamin E), selegiline, or the combination for up to 2 years
[89]. Inclusion criteria were based on a CDR of 2 (moder-
ately severe), with baseline MMSEs averaging 11 to 13. The
primary outcome measure was time to occurrence of any of
death, institutionalization, loss of two or three basic ADLs,
or conversion to a CDR stage of 3 (severe dementia). There
were no differences in the primary outcomes, although be-
cause of a failed randomization, baseline MMSEs differed
significantly among the treatment groups. After correcting
statistically for baseline differences in MMSE, it appeared
there was a statistically significant delay in outcome for
selegiline, ␣-tocopherol, and the combination. There were
no significant differences, however, in cognitive measures.
Because of the need for statistical adjustment and the lack of
benefit on cognitive measures, treatment with selegiline or
vitamin E was not recommended in the previous Canadian
Consensus Guidelines [90]. Subsequently, enthusiasm for
treatment with vitamin E has waned even further because of
a large negative trial of vitamin E in mild cognitive impair-
ment [91] as well as concerns that doses of vitamin E in
excess of 400 IU per day are associated with increased
mortality [92].
3.4.4. Summary
There is evidence from large well-designed randomized
placebo-controlled trials to recommend the use of ChEIs,
memantine, and the combination for the treatment of severe
AD. Treatment benefits appear to include modest improve-
ments in cognition, function, and behavior. None of the
studies were longer than 6 months, leaving questions about
duration of benefits. It does appear clear, however, that
admission to a long-term care facility does not in itself
justify treatment discontinuation. The RCTs reviewed
above included patients with MMSEs as low as 3 to 5, and
although recommendations to discontinue treatment below
these scores might not be unreasonable at the present, it is
still possible that benefits might persist at this stage. We
prefer to recommend that treatment persist until clinical
benefit can no longer be demonstrated, and/or the patient
has reached a very severe stage of AD (GDS, 7).
3.5. Management of BPSD
The management of BPSD begins with appropriate as-
sessment and diagnosis, the identification of target symp-
toms, and consideration of the safety of the patient, their
caregiver, and others in their environment. After optimizing
medical conditions [93,94] and concomitant medications
[95,96] as well as improving sensory impairment [63,97–
99], specific treatment proceeds to nonpharmacologic inter-
ventions such as caregiver education, environmental, and
behavioral approaches. If behaviors persist and are suffi-
ciently disturbing to patient or caregiver or endanger safety,
pharmacologic interventions can be instituted with appro-
priate monitoring of effectiveness and safety.
3.5.1. Nonpharmacologic interventions
Nonpharmacologic interventions include environmental
(eg, wandering tracks, bright light, aromatherapy), behav-
ioral (eg, differential reinforcement), sensory (eg, music,
Snoezelen), physical activity (eg, walking), social contact
(eg, simulated presence), psychotherapeutic (eg, reminis-
cence therapy), and caregiver/staff training interventions. A
number of systematic reviews have critically appraised the
evidence for these interventions [100 –103]. Unfortunately,
none of these reviews focused on severe AD, and although
some of the studies reviewed included patients with severe
AD, most included patients with a variety of dementia
diagnoses and illness severity. The most comprehensive
review was recently published by the Old Age Task Force
of the World Federation of Biological Psychiatry [104].
After identifying 162 studies that met inclusion criteria,
only nine studies were judged to be of high quality (level 1
evidence). These authors concluded that only behavioral
management, specific types of caregiver and staff education,
and possibly cognitive stimulation appear to have lasting
benefits. Unfortunately, a closer evaluation of the level I
studies identified by these authors suggests their application
to severe AD is questionable. For example, in the highest
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
quality study of behavioral treatment of depression in de-
mentia [105], average MMSE scores were 16.5, whereas the
highest quality study of cognitive stimulation therapy en-
rolled patients with average MMSE scores of 14.4 and an
average CDR of 1.4 [106]. Music therapy (no high quality
RCTs) and Snoezelen (one high quality RCT) were also
rated as potentially useful, although benefits have only been
documented during treatment sessions and not in the longer
term. Not included in this review was use of light therapy
and aromatherapy, although these have been examined in
Cochrane Reviews. These reviews concluded that there is
insufficient evidence to assess the value of light therapy (on
the basis of five published studies) [107], and more studies
were required to assess the effectiveness of aromatherapy
(on the basis of two studies) [108]. With respect to aroma-
therapy, a high quality double-blind placebo-controlled trial
with Melissa (lemon balm) included institutionalized pa-
tients with severe dementia only (CDR, 3) [109]. Treatment
with active aromatherapy significantly improved scores on
agitation, aggression, irritability, and aberrant motor behav-
iors, with benefits occurring during the first week of treat-
ment. Although limitations of the study included the method
of randomization and blinding, perhaps the major limitation
to aromatherapy is the increasing move to scent-free work-
ing environments, which might make this intervention im-
practical in many long-term care institutions.
Nonpharmacologic interventions clearly require more re-
search particularly for patients with severe AD. In spite of
the lack of high quality rigorous RCTs, organizations such
as the American Geriatric Society and the American Asso-
ciation for Geriatric Psychiatry have recently recommended
that after the assessment and treatment of associated med-
ical conditions, the initial treatment of BPSD in the absence
of psychosis or immediate danger to self or others should be
nonpharmacologic approaches [110].
3.5.2. Pharmacologic interventions
Before initiation of pharmacotherapy, a patient’s medical
status must be reviewed to ensure avoidance of allergies and
drug-drug interactions as well as specific underlying comor-
bidities that might increase the possibility of adverse drug
reactions (eg, parkinsonism). It is important to identify
target symptoms that are more likely to respond to pharma-
cologic interventions. BPSDs such as agitation, aggression,
psychosis, apathy, and depression are likely to respond to
pharmacotherapy, whereas wandering, perseverative shout-
ing, and some sexually inappropriate behaviors are unlikely
to respond. Pharmacotherapy might be considered first line
in the presence of severe depression, psychosis, and aggres-
sion [110]. All forms of pharmacotherapy must be moni-
tored for effectiveness and adverse events. If effective and
well-tolerated, attempts to taper and discontinue these med-
ications should occur in a standardized fashion. Four ran-
domized double-blind placebo-controlled trials with institu-
tionalized severe dementia patients demonstrated that most
391
patients can be withdrawn from antipsychotics without ex-
acerbating behavior [111–114]. Finally, before initiating
therapy with other forms of pharmacotherapy for BPSD,
clinicians should consider use of the disease-specific treat-
ments, because treatment with ChEIs might improve apathy,
depression, anxiety, and psychosis, whereas memantine
might improve agitation and aggression as noted previously.
3.5.2.1. Antipsychotics
Antipsychotics potentially improve delusions, hallucina-
tions, aggression, agitation, and sleep disturbance in AD
patients [115]. Meta-analyses of the RCTs of typical anti-
psychotics such as haloperidol suggest response rates
⬎60% with benefits 18% to 26% greater than placebo
[116,117]. In these studies, however, the efficacy rate was
approximately equivalent to the side effect rate, and the
risks of developing parkinsonism and tardive dyskinesia are
significant even when treated with low doses of typical
antipsychotics [118,119]. There is also evidence that links
the use of typical antipsychotics with increased cognitive
decline in AD [120]. As a result, previous reviews and
guidelines have preferentially recommended atypical anti-
psychotics for treatment of agitation, aggression, and psy-
chosis [121–124], although most of these recommendations
were published before safety warnings regarding the risk of
cerebrovascular adverse events (CVAEs) and mortality. It
is, therefore, important to briefly review their efficacy and
safety before making recommendations about their use.
More than 2,000 patients, mostly with severe dementia,
were included in the published RCTs of risperidone [125–
128], olanzapine [129,130], and quetiapine [79]. For exam-
ple, in studies with risperidone, the average MMSE scores
ranged from 5 to 9. The studies with risperidone and olan-
zapine demonstrated efficacy greater than placebo, with
reasonable tolerability of doses of risperidone approxi-
mately 1 mg per day and olanzapine 5 to 10 mg per day. In
the only published RCT of quetiapine [79], treatment with
100 mg per day was no more effective than placebo and
significantly worsened cognition. This is in contrast with a
large unpublished trial of quetiapine that suggested efficacy
greater than placebo at 200 mg per day and no significant
changes in cognitive function [131].
After reviewing data from the published and unpublished
RCTs of atypical antipsychotics for BPSD, health regula-
tory agencies including Health Canada issued warnings
about increased CVAEs and mortality in drug-treated pa-
tients relative to placebo-treated patients. In 11 RCTs with
olanzapine and risperidone, 48 of 2,187 (2.2%) drug-treated
patients experienced CVAEs compared with 10 of 1,190
(0.8%) placebo-treated patients, resulting in a relative risk
of 2.7 (95% confidence interval, 1.4 to 5.3) [132]. Similarly,
increased mortality was noted with risperidone (4/1,009, 4%
versus placebo 22/712; 3.1%), quetiapine (20/365; 5.51%
versus placebo 7/217; 3.2%), and olanzapine (42/1,184;
3.5% versus placebo 7/478; 1.5%) [133].
392 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
With respect to CVAEs, the increased incidence might
be accounted for by nonspecific vascular events rather
than completed strokes [132], and large observational
database studies have failed to confirm an increased risk
of stroke compared with typical antipsychotics and even
untreated dementia patients [134 –137]. In terms of mor-
tality data, a meta-analysis of 15 BPSD RCTs compared
3,353 drug-treated patients with 1,757 placebo-treated
patients [138]. There were 118 (3.5%) deaths with atyp-
ical antipsychotics compared with 40 (2.3%) deaths with
placebo and no evidence that risk varied by drug, sever-
ity, or diagnosis. Furthermore, in two studies that in-
cluded haloperidol arms, a similar magnitude of risk was
noted. In contrast, four observational studies have found
no risk of increased mortality associated with atypical
antipsychotic use for BPSD [139 –142].
In summary, efficacy of the atypical antipsychotics, ris-
peridone and olanzapine, is supported by well-designed
RCTs, although these data are tempered by increased
CVAEs and mortality. Although observational studies fail
to confirm these risks, these potentially serious adverse
events must be considered if their use is necessary. There is
no rationale for using typical antipsychotics because their
use is clearly associated with increased extra-pyramidal
symptoms, probably similar rates of CVAEs and mortality,
as well as worsened cognition.
3.5.2.2. Antidepressants
A number of well-designed RCTs suggest that treatment
with antidepressants is useful for depression in patients with
mild to moderate AD [143]. For example, in one of the most
rigorous studies published to date, AD patients with major
depression were randomized to treatment with sertraline or
placebo [144]. Although treatment with sertraline demon-
strated significant overall benefit, inclusion criterion was
MMSE ⬎10, and baseline MMSE scores were 16.3 to 17.5.
Specific criteria for diagnosis of depression in AD have
been proposed [145], although it is not clear how valid or
reliable they are in patients with severe AD. Clinicians
should avoid the use of tricyclic antidepressants as a result
of anticholinergic effects and concerns about worsening
cognition.
Antidepressants have also been used for the treatment of
agitation in AD. In a small randomized double-blind con-
trolled trial, 85 patients with average MMSEs of 6.4 to 9.9
were randomized to citalopram, perphenazine, or placebo
[146]. Only citalopram-treated patients experienced statis-
tically more improvement compared with placebo, with
benefits noted on agitation, aggression, psychosis, lability/
tension, and retardation.
Finally, there have been three RCTs published that used
trazodone for the treatment of agitation and aggression
[147–149]. One of the studies was a small double-blind
placebo-controlled crossover trial comparing trazodone
with buspirone [147]. Although trazodone was significantly
more effective than buspirone and placebo, this study in-
cluded patients with mild-moderate dementia (GDS aver-
age, 4.4). The second study was a randomized trial com-
paring trazodone with haloperidol in moderate to severe
dementia [148] in which both medications appeared equally
efficacious. Only one of these trials was a double-blind
placebo-controlled parallel group design comparing traz-
odone, haloperidol, and behavior therapy, which included
patients with average MMSEs of 12 to 14 [149]. Trazodone
treatment was no better than placebo, and the most frequent
reason for dropouts with trazodone was actually an increase
in agitation. It is important to note that this was a negative
study, and none of the interventions were significantly bet-
ter than placebo; however, a number of methodologic issues
have been raised that question the interpretation and con-
clusions of this study [150]. Although previous recommen-
dations included the use of trazodone for management of
agitation and sleep disturbance [90], a recent meta-analysis
of the use of trazodone for BPSD concluded that there was
insufficient evidence to recommend its use for BPSD at the
present time [151].
In summary, although evidence from RCTs supports the
use of antidepressants, especially selective serotonin re-
uptake inhibitors, for the treatment of depression in mild to
moderate AD, similar studies have not been completed in
patients with severe AD. The use of citalopram for the
treatment of agitation appears promising but requires repli-
cation in a larger, longer-term study. At the present time
there is insufficient evidence to recommend for or against
the use of trazodone for the treatment of BPSD in severe
AD.
3.5.2.3. Anticonvulsants
There have been four published RCTs with carbamaz-
epine and four RCTs with valproate for the treatment of
BPSD in severe dementia. The trials of carbamazepine
[152–155] demonstrated significant benefit compared with
placebo, although concerns about carbamazepine’s tolera-
bility and potential for drug interactions have limited its use.
None of the valproate studies [156 –159] demonstrated ben-
efit on primary outcome measures compared with placebo,
and higher doses are associated with unacceptable side
effects, such as excessive sedation [160].
In summary, there are RCT data to support the use of
carbamazepine for agitation and aggression, although side
effects and drug interactions probably relegate its use to a
second line agent. There is no evidence to support the use of
valproate for BPSD.
3.5.2.4. Other pharmacologic approaches
There is only one recent RCT examining the use of
benzodiazepines for BPSD, an RCT comparing intramuscu-
lar lorazepam with olanzapine and placebo during a 24-hour
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
period [161]. Both treatments were better than placebo, with
no differences in adverse events including sedation. Use of
benzodiazepines should be limited to short-term, prn (as
circumstances may require) use for behavioral emergencies
or sedation for procedures. Because of their well-known
adverse effects including excessive sedation, falls, and cog-
nitive impairment [162], regular, daily use should be
avoided. There are a number of small RCTs with variable
results investigating the use of beta-blockers, such as pro-
pranolol and pindolol [163–166]. Other RCTs for BPSD
have included buspirone [147,167], conjugated estrogens
[168 –170], dronabinol [171], long-acting opioids [172], and
melatonin [173,174]. There is insufficient evidence to sup-
port the use of any of these interventions at the present time.
4. Conclusions
Severe AD represents a huge societal burden as a result
of its prevalence, costs, and the suffering incurred by pa-
tients and their caregivers. This review has emphasized the
increased attention this phase of illness has received re-
cently, including the recognition that it requires specific
CPGs. The recommended interventions have the potential to
improve the quality of life for both patient and caregiver and
are supported by a reasonable amount of RCT data. More
research is still required to operationalize the definition of
severe AD and develop clinically useful measures of cog-
nition. More RCTs are required to elaborate the use of
ChEIs and memantine and better define the length of useful
therapy as well as the greatest severity of illness in which
benefit can be elucidated. More effective nonpharmacologic
approaches must be developed and examined with rigorous
RCTs. Safer and effective pharmacotherapy for BPSD must
also be developed. Ultimately, effective disease-modifying
therapy for mild AD should hopefully preclude the need to
treat this phase of illness.
Author Disclosures
Nathan Herrmann has received research support, hono-
raria, and/or consultant fees from Lundbeck, Janssen Ortho
Inc, Pfizer, Novartis, and Eli Lilly.
Serge Gauthier has been an investigator and/or consul-
tant for Lundbeck, Pfizer, and Merz.
Paul G. Lysy has nothing to disclose.
References
[1] Boller F, Verny M, Hugonot-Diener L, Saxton J. Clinical features
and assessment of severe dementia: a review. Eur J Neurol 2002;9:
125–36.
[2] Tariot PN. Medical management of advanced dementia. J Am Geri-
atr Soc 2003;51:S305–13.
[3] Voisin T, Reynish E, Portet F, Feldman H, Vellas B. What are the
treatment options for patients with severe Alzheimer’s disease? CNS
Drugs 2004;18:575– 83.
393
[4] Farlow MR. Moderate to severe Alzheimer disease: definition and
clinical relevance. Neurology 2005;65:S1– 4.
[5] Vellas B, Gauthier S, Allain H, Andrieu S, Aquino JP, Berrut G, et
al. Consensus statement on dementia of Alzheimer type in the severe
stage. J Nutr Health Aging 2005;9:330 – 8.
[6] Canadian Study of Health and Aging. Study methods and prevalence
of dementia. CMAJ 1994;150:899 –913.
[7] Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzhei-
mer disease in the US population: prevalence estimates using the
2000 census. Arch Neurol 2003;60:1119 –22.
[8] Larson EB, Shadlen MF, Wang L, McCormick WC, Bowen JD, Teri
L, et al. Survival after initial diagnosis of Alzheimer disease. Ann
Intern Med 2004;140:501–9.
[9] Hux MJ, O’Brien BJ, Iskedjian M, Goeree R, Gagnon M, Gauthier
S. Relation between severity of Alzheimer’s disease and costs of
caring. CMAJ 1998;159:457– 65.
[10] Wimo A, Ljunggren G, Winblad B. Costs of dementia and dementia
care: a review. Int J Geriatr Psychiatry 1997;12:841–56.
[11] Woolf SH, Battista RN, Anderson GM, Logan AG, Wang E. As-
sessing the clinical effectiveness of preventive maneuvers: analytic
principles and systematic methods in reviewing evidence and de-
veloping clinical practice recommendations—a report by the Cana-
dian Task Force on the Periodic Health Examination. J Clin Epide-
miol 1990;43:891–905.
[12] Folstein MF, Folstein SE, McHugh PR. ”Mini-mental state”: a
practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 1975;12:189 –98.
[13] Reisberg B, Sclan G, Franssen EH, Klugger A, Ferris S. Dementia
staging in chronic care populations. Alzheimer Disease Related
Disorders 1944;8:S188 –205.
[14] Ferris SH, Yan B. Differential diagnosis and clinical assessment of
patients with severe Alzheimer disease. Alzheimer Dis Assoc Disord
2003;17(Suppl 3):S92–5.
[15] Feldman HH, Woodward M. The staging and assessment of mod-
erate to severe Alzheimer disease. Neurology 2005;65:S10 –7.
[16] Reisberg B, Ferris SH, de Leon MJ, Crook T. The Global Deterio-
ration Scale for assessment of primary degenerative dementia. Am J
Psychiatry 1982;139:1136 –9.
[17] Berg L. Clinical Dementia Rating (CDR). Psychopharmacol Bull
1988;24:637–9.
[18] Morris JC. Clinical dementia rating: a reliable and valid diagnostic
and staging measure for dementia of the Alzheimer type. Int Psy-
chogeriatr 1997;9(Suppl 1):173– 8.
[19] Solomon PR, Adams FA, Groccia ME, DeVeaux R, Growdon JH,
Pendlebury WW. Correlational analysis of five commonly used
measures of mental status/functional abilities in patients with Alz-
heimer disease. Alzheimer Dis Assoc Disord 1999;13:147–50.
[20] Saxton J, McGonigle-Gibson G, Swihart A, Miller VJ, Boller F.
Assessment of the severely impaired patient: description and vali-
dation of a new neuropsychological test battery. Psychol Assess-
ment: J Consulting Clin Psychology 1990;2:2.
[21] Panisset M, Roudier M, Saxton J, Boller F. Severe impairment
battery: a neuropsychological test for severely demented patients.
Arch Neurol 1994;51:41–5.
[22] Saxton J, Kastango KB, Hugonot-Diener L, Boller F, Verny M,
Sarles CE, et al. Development of a short form of the Severe Impair-
ment Battery. Am J Geriatr Psychiatry 2005;13:999 –1005.
[23] Feldman H, Gauthier S, Hecker J, Vellas B, Subbiah P, Whalen E.
A 24-week, randomized, double-blind study of donepezil in mod-
erate to severe Alzheimer’s disease. Neurology 2001;57:613–20.
[24] Tariot PN, Cummings JL, Katz IR, Mintzer J, Perdomo CA,
Schwam EM, et al. A randomized, double-blind, placebo-controlled
study of the efficacy and safety of donepezil in patients with Alz-
heimer’s disease in the nursing home setting. J Am Geriatr Soc
2001;49:1590 –9.
394 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
[25] Reisberg B, Doody R, Stoffler A, Schmitt F, Ferris S, Mobius H.
Memantine in moderate-to-severe Alzheimer’s disease. N Engl
J Med 2003;348:1333– 41.
[26] Tariot PN, Farlow MR, Grossberg GT, Graham SM, McDonald S,
Gergel I. Memantine treatment in patients with moderate to severe
Alzheimer disease already receiving donepezil: a randomized con-
trolled trial. JAMA 2004;291:317–24.
[27] Galasko D, Bennett D, Sano M, Ernesto C, Thomas R, Grundman
M, et al. An inventory to assess activities of daily living for clinical
trials in Alzheimer’s disease: the Alzheimer’s Disease Cooperative
Study. Alzheimer Dis Assoc Disord 1997;11(Suppl 2): S33–9.
[28] Galansko DR, Schmitt FA, Jin S. Detailed assessment of cognition
and activities of daily living in moderate to severe Alzheimer’s
disease. Neurobiol Aging 2000;21:S168.
[29] Gelinas I, Gauthier L, McIntyre M, Gauthier S. Development of a
functional measure for persons with Alzheimer’s disease: the dis-
ability assessment for dementia. Am J Occup Ther 1999;53:471– 81.
[30] Katz S, Akpom CA. 12: Index of ADL. Med Care 1976;14:116 – 8.
[31] Cohen-Mansfield J, Reisberg B, Bonnema J, Berg L, Dastoor DP,
Pfeffer RI, et al. Staging methods for the assessment of dementia:
perspectives. J Clin Psychiatry 1996;57:190 – 8.
[32] Reisberg B, Franssen EH, Bobinski M, Auer S, Monteiro I, Boksay
I, et al. Overview of methodologic issues for pharmacologic trials in
mild, moderate, and severe Alzheimer’s disease. Int Psychogeriatr
1996;8:159 –93.
[33] Overall JE, Scott J, Rhoades HM, Lesser J. Empirical scaling of the
stages of cognitive decline in senile dementia. J Geriatr Psychiatry
Neurol 1990;3:212–20.
[34] Lopez OL, Becker JT, Sweet RA, Klunk W, Kaufer DI, Saxton J, et
al. Psychiatric symptoms vary with the severity of dementia in
probable Alzheimer’s disease. J Neuropsychiatry Clin Neurosci
2003;15:346 –53.
[35] Wood S, Cummings JL, Hsu MA, Barclay T, Wheatley MV,
Yarema KT, et al. The use of the neuropsychiatric inventory in
nursing home residents: characterization and measurement. Am J
Geriatr Psychiatry 2000;8:75– 83.
[36] Clyburn LD, Stones MJ, Hadjistavropoulos T, Tuokko H. Predicting
caregiver burden and depression in Alzheimer’s disease. J Gerontol
B Psychol Sci Soc Sci 2000;55:S2–13.
[37] Coen RF, Swanwick GR, O’Boyle CA, Coakley D. Behaviour
disturbance and other predictors of carer burden in Alzheimer’s
disease. Int J Geriatr Psychiatry 1997;12:331– 6.
[38] Donaldson C, Tarrier N, Burns A. Determinants of carer stress in
Alzheimer’s disease. Int J Geriatr Psychiatry 1998;13:248 –56.
[39] Miller TP, Tinklenberg JR, Brooks JO, 3rd, Fenn HH, Yesavage JA.
Selected psychiatric symptoms associated with rate of cognitive
decline in patients with Alzheimer’s disease. J Geriatr Psychiatry
Neurol 1993;6:235– 8.
[40] Moritz DJ, Fox PJ, Luscombe FA, Kraemer HC. Neurological and
psychiatric predictors of mortality in patients with Alzheimer dis-
ease in California. Arch Neurol 1997;54:878 – 85.
[41] Walsh JS, Welch HG, Larson EB. Survival of outpatients with
Alzheimer-type dementia. Ann Intern Med 1990;113:429 –34.
[42] Lopez OL, Wisniewski SR, Becker JT, Boller F, DeKosky ST.
Psychiatric medication and abnormal behavior as predictors of pro-
gression in probable Alzheimer disease. Arch Neurol 1999;
56:1266 –72.
[43] Mortimer JA, Ebbitt B, Jun SP, Finch MD. Predictors of cognitive
and functional progression in patients with probable Alzheimer’s
disease. Neurology 1992;42:1689 –96.
[44] Ryden MB. Aggressive behavior in persons with dementia who live
in the community. Alzheimer Dis Assoc Disord 1988;2:342–55.
[45] Chenoweth B, Spencer B. Dementia: the experience of family care-
givers. Gerontologist 1986;26:267–72.
[46] Weiner MF, Koss E, Wild KV, Folks DG, Tariot P, Luszczynska H,
et al. Measures of psychiatric symptoms in Alzheimer patients: a
review. Alzheimer Dis Assoc Disord 1996;10:20 –30.
[47] Cummings JL. Neuropsychiatric and behavioral alterations and their
management in moderate to severe Alzheimer disease. Neurology
2005;65:S18 –24.
[48] Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi
DA, Gornbein J. The Neuropsychiatric Inventory: comprehensive
assessment of psychopathology in dementia. Neurology 1994;44:
2308 –14.
[49] Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. Cornell
Scale for Depression in Dementia. Biol Psychiatry 1988;23:271– 84.
[50] Cohen-Mansfield J, Billig N. Agitated behaviors in the elderly. I: a
conceptual review. J Am Geriatr Soc 1986;34:711–21.
[51] Teresi J, Abrams R, Holmes D, Ramirez M, Eimicke J. Prevalence
of depression and depression recognition in nursing homes. Soc
Psychiatry Psychiatr Epidemiol 2001;36:613–20.
[52] Nilsson K, Palmstierna T, Wistedt B. Aggressive behavior in hos-
pitalized psychogeriatric patients. Acta Psychiatr Scand 1988;78:
172–5.
[53] Souren LE, Franssen EH, Reisberg B. Contractures and loss of
function in patients with Alzheimer’s disease. J Am Geriatr Soc
1995;43:650 –5.
[54] Auer SR, Sclan SG, Yaffee RA, Reisberg B. The neglected half of
Alzheimer disease: cognitive and functional concomitants of severe
dementia. J Am Geriatr Soc 1994;42:1266 –72.
[55] O’Donnell BF, Drachman DA, Barnes HJ, Peterson KE, Swearer
JM, Lew RA. Incontinence and troublesome behaviors predict in-
stitutionalization in dementia. J Geriatr Psychiatry Neurol 1992;5:
45–52.
[56] Thomas P, Ingrand P, Lalloue F, Hazif-Thomas C, Billon R, Vieban
F, et al. Reasons of informal caregivers for institutionalizing demen-
tia patients previously living at home: the Pixel study. Int J Geriatr
Psychiatry 2004;19:127–35.
[57] Gillick M, Berkman S, Cullen L. A patient-centered approach to
advance medical planning in the nursing home. J Am Geriatr Soc
1999;47:227–30.
[58] Sampson EL, Ritchie CW, Lai R, Raven PW, Blanchard MR. A
systematic review of the scientific evidence for the efficacy of a
palliative care approach in advanced dementia. Int Psychogeriatr
2005;17:31– 40.
[59] Lyness JM. End-of-life care: issues relevant to the geriatric psychi-
atrist. Am J Geriatr Psychiatry 2004;12:457–72.
[60] Sachs GA, Shega JW, Cox-Hayley D. Barriers to excellent end-of-
life care for patients with dementia. J Gen Intern Med 2004;19:
1057– 63.
[61] Scherder E, Oosterman J, Swaab D, Herr K, Ooms M, Ribbe M, et
al. Recent developments in pain in dementia. BMJ 2005;330:461– 4.
[62] Magri F, Borza A, del Vecchio S, Chytiris S, Cuzzoni G, Busconi L,
et al. Nutritional assessment of demented patients: a descriptive
study. Aging Clin Exp Res 2003;15:148 –53.
[63] Allen NH, Burns A, Newton V, Hickson F, Ramsden R, Rogers J, et
al. The effects of improving hearing in dementia. Age Ageing
2003;32:189 –93.
[64] Tiraboschi P, Hansen LA, Thal LJ, Corey-Bloom J. The importance
of neuritic plaques and tangles to the development and evolution of
AD. Neurology 2004;62:1984 –9.
[65] Di Patre PL, Read SL, Cummings JL, Tomiyasu U, Vartavarian LM,
Secor DL, et al. Progression of clinical deterioration and patholog-
ical changes in patients with Alzheimer disease evaluated at biopsy
and autopsy. Arch Neurol 1999;56:1254 – 61.
[66] Lanctot KL, Herrmann N, Mazzotta P. Role of serotonin in the
behavioral and psychological symptoms of dementia. J Neuropsy-
chiatry Clin Neurosci 2001;13:5–21.
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
[67] Herrmann N, Lanctot KL, Khan LR. The role of norepinephrine in
the behavioral and psychological symptoms of dementia. J Neuro-
psychiatry Clin Neurosci 2004;16:261–76.
[68] Lanctot KL, Herrmann N, Mazzotta P, Khan LR, Ingber N.
GABAergic function in Alzheimer’s disease: evidence for dysfunc-
tion and potential as a therapeutic target for the treatment of behav-
ioural and psychological symptoms of dementia. Can J Psychiatry
2004;49:439 –53.
[69] Cacabelos R, Takeda M, Winblad B. The glutamatergic system and
neurodegeneration in dementia: preventive strategies in Alzheimer’s
disease. Int J Geriatr Psychiatry 1999;14:3– 47.
[70] Davis KL, Mohs RC, Marin D, Purohit DP, Perl DP, Lantz M, et al.
Cholinergic markers in elderly patients with early signs of Alzhei-
mer disease. JAMA 1999;281:1401– 6.
[71] DeKosky ST, Ikonomovic MD, Styren SD, Beckett L, Wisniewski
S, Bennett DA, et al. Upregulation of choline acetyltransferase
activity in hippocampus and frontal cortex of elderly subjects with
mild cognitive impairment. Ann Neurol 2002;51:145–55.
[72] Lanctot KL, Herrmann N, Yau KK, Khan LR, Liu BA, LouLou
MM, et al. Efficacy and safety of cholinesterase inhibitors in Alz-
heimer’s disease: a meta-analysis. CMAJ 2003;169:557– 64.
[73] Trinh NH, Hoblyn J, Mohanty S, Yaffe K. Efficacy of cholinesterase
inhibitors in the treatment of neuropsychiatric symptoms and func-
tional impairment in Alzheimer disease: a meta-analysis. JAMA
2003;289:210 – 6.
[74] Holmes C, Wilkinson D, Dean C, Vethanayagam S, Olivieri S,
Langley A, et al. The efficacy of donepezil in the treatment of
neuropsychiatric symptoms in Alzheimer disease. Neurology 2004;
63:214 –9.
[75] Schneider LS, Anand R, Farlow MR. Systematic review of the
efficacy of rivastigmine for patients with Alzheimer’s disease. In-
ternational Journal of Geriatric Psychopharmacology 1998;1:S26 –
34.
[76] Wilkinson DG, Hock C, Farlow M, van Baelen B, Schwalen S.
Galantamine provides broad benefits in patients with ’advanced
moderate’ Alzheimer’s disease (MMSE ⬍ or ⫽ 12) for up to six
months. Int J Clin Pract 2002;56:509 –14.
[77] Blesa R, Davidson M, Kurz A, Reichman W, van Baelen B,
Schwalen S. Galantamine provides sustained benefits in patients
with ‘advanced moderate’ Alzheimer’s disease for at least 12
months. Dement Geriatr Cogn Disord 2003;15:79 – 87.
[78] Feldman H, Gauthier S, Hecker J, Vellas B, Xu Y, Leni JR, et al.
Efficacy and safety of donepezil in patients with more severe Alz-
heimer’s disease: a subgroup analysis from a randomized, placebo-
controlled trial. Int J Geriatr Psychiatry 2005;20:559 – 69.
[79] Ballard C, Margallo-Lana M, Juszczak E, Douglas S, Swann A,
Thomas A, et al. Quetiapine and rivastigmine and cognitive decline
in Alzheimer’s disease: randomised double blind placebo controlled
trial. BMJ 2005;330:874.
[80] Winblad B, Kilander L, Eriksson S, Minthon L, Batsman S, Wet-
terholm AL, et al. Donepezil in patients with severe Alzheimer’s
disease: double-blind, parallel-group, placebo-controlled study.
Lancet 2006;367:1057– 65.
[81] Hogan DB. Donepezil for severe Alzheimer’s disease. Lancet 2006;
367:1031–2.
[82] Doraiswamy PM, Krishnan KR, Anand R, Sohn H, Danyluk J,
Hartman RD, et al. Long-term effects of rivastigmine in moderately
severe Alzheimer’s disease: does early initiation of therapy offer
sustained benefits? Prog Neuropsychopharmacol Biol Psychiatry
2002;26:705–12.
[83] Burns A, Spiegel R, Quarg P. Efficacy of rivastigmine in subjects
with moderately severe Alzheimer’s disease. Int J Geriatr Psychiatry
2004;19:243–9.
[84] Bullock R, Touchon J, Bergman H, Gambina G, He Y, Rapatz G, et
al. Rivastigmine and donepezil treatment in moderate to moderately-
395
severe Alzheimer’s disease over a 2-year period. Curr Med Res Opin
2005;21:1317–27.
[85] Edwards K, Koumaras B, Chen M, Gunay I, Mirski D. Long-term
effects of rivastigmine treatment on the need for psychotropic med-
ications in nursing home patients with Alzheimer’s disease. Clin
Drug Invest 2005;25:507–15.
[86] Winblad B, Poritis N. Memantine in severe dementia: results of the
9M-Best Study (benefit and efficacy in severely demented patients
during treatment with memantine). Int J Geriatr Psychiatry 1999;14:
135– 46.
[87] Livingston G, Katona C. The place of memantine in the treatment of
Alzheimer’s disease: a number needed to treat analysis. Int J Geriatr
Psychiatry 2004;19:919 –25.
[88] Gauthier S, Wirth Y, Mobius HJ. Effects of memantine on behav-
ioural symptoms in Alzheimer’s disease patients: an analysis of the
Neuropsychiatric Inventory (NPI) data of two randomised, con-
trolled studies. Int J Geriatr Psychiatry 2005;20:459 – 64.
[89] Sano M, Ernesto C, Thomas RG, Klauber MR, Schafer K, Grund-
man M, et al. A controlled trial of selegiline, alpha-tocopherol, or
both as treatment for Alzheimer’s disease: the Alzheimer’s Disease
Cooperative Study. N Engl J Med 1997;336:1216 –22.
[90] Patterson CJ, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus
Conference on Dementia. CMAJ 1999;160:S1–15.
[91] Petersen RC, Thomas RG, Grundman M, Bennett D, Doody R,
Ferris S, et al. Vitamin E and donepezil for the treatment of mild
cognitive impairment. N Engl J Med 2005;352:2379 – 88.
[92] Miller ER, 3rd, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel
LJ, Guallar E. Meta-analysis: high-dosage vitamin E supplementa-
tion may increase all-cause mortality. Ann Intern Med 2005;142:
37– 46.
[93] Kiely DK, Morris JN, Algase DL. Resident characteristics associ-
ated with wandering in nursing homes. Int J Geriatr Psychiatry
2000;15:1013–20.
[94] Cohen-Mansfield J, Billig N, Lipson S, Rosenthan AS, Pawlson LG.
Medical correlates of agitation in nursing home residents. Geron-
tology 1990;36:150 – 8.
[95] Learoyd BM. Psychotropic drugs and the elderly patient. Med J Aust
1972;1:1131–3.
[96] Lanctot KL, Bowles SK, Herrmann N, Best TS, Naranjo CA. Drugs
mimicking dementia: dementia symptoms associated with psycho-
tropic drugs in institutionalised cognitively impaired patients. CNS
Drugs 2000;14:381–90.
[97] Corbin SL, Eastwood MR. Sensory deficits and mental disorders of
old age: causal or coincidental associations? Psychol Med 1986;16:
251– 6.
[98] Horowitz A. The relationship between vision impairment and the
assessment of disruptive behaviors among nursing home residents.
Gerontologist 1997;37:620 – 8.
[99] Chapman FM, Dickinson J, McKeith I, Ballard C. Association
among visual hallucinations, visual acuity, and specific eye pathol-
ogies in Alzheimer’s disease: treatment implications. Am J Psychi-
atry 1999;156:1983–5.
[100] Cohen-Mansfield J. Nonpharmacologic interventions for inappropri-
ate behaviors in dementia: a review, summary, and critique. Am J
Geriatr Psychiatry 2001;9:361– 81.
[101] Snowden M, Sato K, Roy-Byrne P. Assessment and treatment of
nursing home residents with depression or behavioral symptoms
associated with dementia: a review of the literature. J Am Geriatr
Soc 2003;51:1305–17.
[102] Opie J, Rosewarne R, O’Connor DW. The efficacy of psychosocial
approaches to behaviour disorders in dementia: a systematic litera-
ture review. Aust N Z J Psychiatry 1999;33:789 –99.
396 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
[103] Verkaik R, van Weert JC, Francke AL. The effects of psychosocial
methods on depressed, aggressive and apathetic behaviors of people
with dementia: a systematic review. Int J Geriatr Psychiatry 2005;
20:301–14.
[104] Livingston G, Johnston K, Katona C, Paton J, Lyketsos CG. Sys-
tematic review of psychological approaches to the management of
neuropsychiatric symptoms of dementia. Am J Psychiatry 2005;162:
1996 –2021.
[105] Teri L. Behavioral treatment of dementia: an overview. Semin Clin
Neuropsychiatry 1997;2:100 –1.
[106] Spector A, Thorgrimsen L, Woods B, Royan L, Davies S, Butter-
worth M, et al. Efficacy of an evidence-based cognitive stimulation
therapy programme for people with dementia: randomised con-
trolled trial. Br J Psychiatry 2003;183:248 –54.
[107] Forbes D, Morgan DG, Bangma J, Peacock S, Pelletier N, Adamson
J. Light therapy for managing sleep, behaviour, and mood distur-
bances in dementia. Cochrane Database Syst Rev 2005;2:
CD003946.
[108] Thorgrimsen L, Spector A, Wiles A, Orrell M. Aroma therapy for
dementia. Cochrane Database Syst Rev 2003;CD003150.
[109] Ballard CG, O’Brien JT, Reichelt K, Perry EK. Aromatherapy as a
safe and effective treatment for the management of agitation in
severe dementia: the results of a double-blind, placebo-controlled
trial with Melissa. J Clin Psychiatry 2002;63:553– 8.
[110] Consensus statement on improving the quality of mental health care
in U.S. nursing homes: management of depression and behavioral
symptoms associated with dementia. J Am Geriatr Soc 2003;51:
1287–98.
[111] Cohen-Mansfield J, Lipson S, Werner P, Billig N, Taylor L, Woos-
ley R. Withdrawal of haloperidol, thioridazine, and lorazepam in the
nursing home: a controlled, double-blind study. Arch Intern Med
1999;159:1733– 40.
[112] van Reekum R, Clarke D, Conn D, Herrmann N, Eryavec G, Cohen
T, et al. A randomized, placebo-controlled trial of the discontinua-
tion of long-term antipsychotics in dementia. Int Psychogeriatr
2002;14:197–210.
[113] Ballard CG, Thomas A, Fossey J, Lee L, Jacoby R, Lana MM, et al.
A 3-month, randomized, placebo-controlled, neuroleptic discontin-
uation study in 100 people with dementia: the neuropsychiatric
inventory median cutoff is a predictor of clinical outcome. J Clin
Psychiatry 2004;65:114 –9.
[114] Ruths S, Straand J, Nygaard HA, Bjorvatn B, Pallesen S. Effect of
antipsychotic withdrawal on behavior and sleep/wake activity in
nursing home residents with dementia: a randomized, placebo-con-
trolled, double-blinded study: the Bergen District Nursing Home
Study. J Am Geriatr Soc 2004;52:1737– 43.
[115] Rabinowitz J, Katz IR, De Deyn PP, Brodaty H, Greenspan A,
Davidson M. Behavioral and psychological symptoms in patients
with dementia as a target for pharmacotherapy with risperidone.
J Clin Psychiatry 2004;65:1329 –34.
[116] Schneider LS, Pollock VELyness SA. A metaanalysis of controlled
trials of neuroleptic treatment in dementia. J Am Geriatr Soc 1990;
38:553– 63.
[117] Lanctot KL, Best TS, Mittmann N, Liu BA, Oh PI, Einarson TR, et
al. Efficacy and safety of neuroleptics in behavioral disorders asso-
ciated with dementia. J Clin Psychiatry 1998;59:550 – 63.
[118] Sweet RA. Neuroleptic-induced parkinsonism in elderly patients
diagnosed with psychotic major depression and dementia of the
Alzheimer type. Am J Geriatr Psychiatry 1996;4:311–9.
[119] Caligiuri MP, Rockwell E, Jeste DV. Extrapyramidal side effects in
patients with Alzheimer’s disease treated with low-dose neuroleptic
medication. Am J Geriatr Psychiatry 1998;6:75– 82.
[120] McShane R, Keene J, Gedling K, Fairburn C, Jacoby R, Hope T. Do
neuroleptic drugs hasten cognitive decline in dementia? prospective
study with necropsy follow up. BMJ 1997;314:266 –70.
[121] Alexopoulos GS, Silver JM, Kahn DA, Frances A, Carpenter D. The
expert consensus guideline series: treatment of agitation in older
persons with dementia. Postgrad Med 1998;26:1– 88.
[122] Herrmann N. Recommendations for the management of behavioral
and psychological symptoms of dementia. Can J Neurol Sci 2001;
28(Suppl 1): S96 –107.
[123] Sink KM, Holden KF, Yaffe K. Pharmacological treatment of neu-
ropsychiatric symptoms of dementia: a review of the evidence.
JAMA 2005;293:596 – 608.
[124] Alexopoulos GS, Streim J, Carpenter D, Docherty JP. Using anti-
psychotic agents in older patients. J Clin Psychiatry 2004;65(Suppl
2): 5–104.
[125] Katz IR, Jeste DV, Mintzer JE, Clyde C, Napolitano J, Brecher M.
Comparison of risperidone and placebo for psychosis and behavioral
disturbances associated with dementia: a randomized, double-blind
trial—Risperidone Study Group. J Clin Psychiatry 1999;60:107–15.
[126] De Deyn PP, Rabheru K, Rasmussen A, Bocksberger JP, Dautzen-
berg PL, Eriksson S, et al. A randomized trial of risperidone,
placebo, and haloperidol for behavioral symptoms of dementia.
Neurology 1999;53:946 –55.
[127] Chan WC, Lam LC, Choy CN, Leung VP, Li SW, Chiu HF. A
double-blind randomised comparison of risperidone and haloperidol in
the treatment of behavioural and psychological symptoms in Chinese
dementia patients. Int J Geriatr Psychiatry 2001;16:1156 – 62.
[128] Brodaty H, Ames D, Snowdon J, Woodward M, Kirwan J, Clarnette
R, et al. A randomized placebo-controlled trial of risperidone for the
treatment of aggression, agitation, and psychosis of dementia. J Clin
Psychiatry 2003;64:134 – 43.
[129] Street JS, Clark WS, Gannon KS, Cummings JL, Bymaster FP,
Tamura RN, et al. Olanzapine treatment of psychotic and behavioral
symptoms in patients with Alzheimer disease in nursing care facil-
ities: a double-blind, randomized, placebo-controlled trial—the
HGEU Study Group. Arch Gen Psychiatry 2000;57:968 –76.
[130] De Deyn PP, Carrasco MM, Deberdt W, Jeandel C, Hay DP, Feld-
man PD, et al. Olanzapine versus placebo in the treatment of psy-
chosis with or without associated behavioral disturbances in patients
with Alzheimer’s disease. Int J Geriatr Psychiatry 2004;19:115–26.
[131] Zhong KX, Tariot PN, Mintzer J, Minkwitz MC, Devine NA.
Quetiapine to treat agitation in dementia: a randomized, double-
blind, placebo-controlled study. Curr Alzheimer Res 2007;4:81–93.
[132] Herrmann N, Lanctot KL. Do atypical antipsychotics cause stroke?
CNS Drugs 2005;19:91–103.
[133] Health Canada. Health Canada endorsed important safety informa-
tion on atypical antipsychotic drugs and dementia. Available at:
http://www.hc-sc.gc.ca/dhp-mps/medeff/advisories-avis/prof/atyp-
antipsycho_hpc_e.html. Accessed May 12, 2005.
[134] Liperoti R, Gambassi G, Lapane KL, Chiang C, Pedone C, Mor V,
et al. Cerebrovascular events among elderly nursing home patients
treated with conventional or atypical antipsychotics. J Clin Psychi-
atry 2005;66:1090 – 6.
[135] Layton D, Harris S, Wilton LV, Shakir SA. Comparison of incidence
rates of cerebrovascular accidents and transient ischaemic attacks in
observational cohort studies of patients prescribed risperidone, quetia-
pine or olanzapine in general practice in England including patients
with dementia. J Psychopharmacol 2005;19:473– 82.
[136] Herrmann N, Mamdani M, Lanctot KL. Atypical antipsychotics and risk of
cerebrovascular accidents. Am J Psychiatry 2004;161:1113–5.
[137] Gill SS, Rochon PA, Herrmann N, Lee PE, Sykora K, Gunraj N, et
al. Atypical antipsychotic drugs and risk of ischaemic stroke: pop-
ulation based retrospective cohort study. BMJ 2005;330:445.
[138] Schneider LS, Dagerman K, SInsel P. Risk of death with atypical
antipsychotic drug treatment for dementia: meta-analysis of random-
ized placebo-controlled trials. JAMA 2005;294:1934 – 43.
 N. Herrmann et al. / Alzheimer’s & Dementia 3 (2007) 385–397
[139] Raivio MM, Laurila JV, Strandberg TE, Tilvis RS, Pitkala KH.
Psychotropic medication and stroke outcome. Am J Psychiatry
2005;162:1027– 8.
[140] Nasrallah HA, White T, Nasrallah AT. Lower mortality in geriatric
patients receiving risperidone and olanzapine versus haloperidol:
preliminary analysis of retrospective data. Am J Geriatr Psychiatry
2004;12:437–9.
[141] Suh GH, Shah A. Effect of antipsychotics on mortality in elderly
patients with dementia: a 1-year prospective study in a nursing
home. Int Psychogeriatr 2005;17:429 – 41.
[142] Wang PS, Schneeweiss S, Avorn J, Fischer MA, Mogun H, Solomon
DH, et al. Risk of death in elderly users of conventional vs. atypical
antipsychotic medications. N Engl J Med 2005;353:2335– 41.
[143] Lyketsos CG, Olin J. Depression in Alzheimer’s disease: overview
and treatment. Biol Psychiatry 2002;52:243–52.
[144] Lyketsos CG, DelCampo L, Steinberg M, Miles Q, Steele CD,
Munro C, et al. Treating depression in Alzheimer disease: efficacy
and safety of sertraline therapy, and the benefits of depression
reduction: the DIADS. Arch Gen Psychiatry 2003;60:737– 46.
[145] Olin JT, Schneider LS, Katz IR, Meyers BS, Alexopoulos GS,
Breitner JC, et al. Provisional diagnostic criteria for depression of
Alzheimer disease. Am J Geriatr Psychiatry 2002;10:125– 8.
[146] Pollock BG, Mulsant BH, Rosen J, Sweet RA, Mazumdar S, Bha-
rucha A, et al. Comparison of citalopram, perphenazine, and placebo
for the acute treatment of psychosis and behavioral disturbances in
hospitalized, demented patients. Am J Psychiatry 2002;159:460 –5.
[147] Lawlor BA, Radcliffe J, Molchan SE, Martinez RA, Hill JL, Sun-
derland T. A pilot placebo-controlled study of trazodone and buspi-
rone in Alzheimer’s disease. Int J Geriatr Psychiatry 1994;9:55–9.
[148] Sultzer DL, Gray KF, Gunay I, Berisford MA, Mahler ME. A
double-blind comparison of trazodone and haloperidol for treatment
of agitation in patients with dementia. Am J Geriatr Psychiatry
1997;5:60 –9.
[149] Teri L, Logsdon RG, Peskind E, Raskind M, Weiner MF, Tracten-
berg RE, et al. Treatment of agitation in AD: a randomized, placebo-
controlled clinical trial. Neurology 2000;55:1271– 8.
[150] Herrmann N, Black SE. Behavioral disturbances in dementia: will
the real treatment please stand up? Neurology 2000;55:1247– 8.
[151] Martinon-Torres G, Fioravanti M, Grimley Evans J. Trazodone for
agitation in dementia. Cochrane Database Syst Rev 2004;
CD004990.
[152] Tariot PN, Erb R, Leibovici A, Podgorski CA, Cox C, Asnis J, et al.
Carbamazepine treatment of agitation in nursing home patients with
dementia: a preliminary study. J Am Geriatr Soc 1994;42:1160 – 6.
[153] Cooney C, Mortimer A, Smith A, Newton K, Wrigley M. Carbam-
azepine use in aggressive behaviour associated with senile dementia.
Int J Geriatr Psychiatry 1996;11:901–5.
[154] Tariot PN, Erb R, Podgorski CA, Cox C, Patel S, Jakimovich L, et
al. Efficacy and tolerability of carbamazepine for agitation and
aggression in dementia. Am J Psychiatry 1998;155:54 – 61.
[155] Olin JT, Fox LS, Pawluczyk S, Taggart NA, Schneider L. A pilot
randomized trial of carbamazepine for behavioral symptoms in treat-
ment-resistant outpatients with Alzheimer disease. Am J Geriatr
Psychiatry 2001;9:400 –5.
[156] Porsteinsson AP, Tariot PN, Erb R, Cox C, Smith E, Jakimovich L,
et al. Placebo-controlled study of divalproex sodium for agitation in
dementia. Am J Geriatr Psychiatry 2001;9:58 – 66.
[157] Sival RC, Haffmans PM, Jansen PA, Duursma SA, Eikelenboom P.
Sodium valproate in the treatment of aggressive behavior in patients
with dementia: a randomized placebo controlled clinical trial. Int J
Geriatr Psychiatry 2002;17:579 – 85.
397
[158] Tariot PN, Schneider LS, Mintzer JE, Cutler AJ. Safety and toler-
ability of divalproex sodium in treatment of signs and symptoms of
mania in elderly patients with dementia: results of a double-blind
placebo-controlled trial. Curr Ther Res Clin Exp 2001;62:51– 67.
[159] Tariot PN, Raman R, Jakimovich L, Schneider L, Porteinsson A,
Thomas R, et al. Divalproex sodium in nursing home residents with
possible or probable Alzheimer disease complicated by agitation: a
randomized, controlled trial. Am J Geriatr Psychiatry 2005;
13:942–9.
[160] Lonergan ET, Luxenberg J. Valproate preparations for agitation in
dementia. Cochrane Database Syst Rev 2005;3.
[161] Meehan KM, Wang H, David SR, Nisivoccia JR, Jones B, Beasley
SM, Jr, et al. Comparison of rapidly acting intramuscular olanzap-
ine, lorazepam, and placebo: a double-blind, randomized study in
acutely agitated patients with dementia. Neuropsychopharmacology
2002;26:494 –504.
[162] Stern RG, Duffelmeyer ME, Zemishlani Z, Davidson M. The use of
benzodiazepines in the management of behavioral symptoms in
demented patients. Psychiatr Clin North Am 1991;14:375– 84.
[163] Greendyke RM, Kanter DR, Schuster DB, Verstreate S, Wootton J.
Propranolol treatment of assaultive patients with organic brain dis-
ease: a double-blind crossover, placebo-controlled study. J Nerv
Ment Dis 1986;174:290 – 4.
[164] Greendyke RM, Kanter DR. Therapeutic effects of pindolol on
behavioral disturbances associated with organic brain disease: a
double-blind study. J Clin Psychiatry 1986;47:423– 6.
[165] Herrmann N, Lanctot KL, Eryavec G, Khan LR. Noradrenergic
activity is associated with response to pindolol in aggressive Alz-
heimer’s disease patients. J Psychopharmacol 2004;18:215–20.
[166] Peskind ER, Tsuang DW, Bonner LT, Pascualy M, Riekse RG,
Snowden MB, et al. Propranolol for disruptive behaviors in nursing
home residents with probable or possible Alzheimer disease: a
placebo-controlled study. Alzheimer Dis Assoc Disord 2005;19:
23– 8.
[167] Cantillon M, Brunswick R, Molina D, Bahro M. Buspirone vs.
haloperidol: a double-blind trial for agitation in nursing home pop-
ulation with Alzheimer’s disease. Am J Geriatr Psychiatry 1996;4:
263–7.
[168] Kyomen HH, Hennen J, Gottlieb GL, Wei JY. Estrogen therapy and
noncognitive psychiatric signs and symptoms in elderly patients
with dementia. Am J Psychiatry 2002;159:1225–7.
[169] Kyomen HH, Satlin A, Hennen J, Wei JY. Estrogen therapy and
aggressive behavior in elderly patients with moderate-to-severe de-
mentia: results from a short-term, randomized, double-blind trial.
Am J Geriatr Psychiatry 1999;7:339 – 48.
[170] Hall KA, Keks NA, O’Connor DW. Transdermal estrogen patches
for aggressive behavior in male patients with dementia: a random-
ized, controlled trial. Int Psychogeriatr 2005;17:165–78.
[171] Volicer L, Stelly M, Morris J, McLaughlin J, Volicer BJ. Effects of
dronabinol on anorexia and disturbed behavior in patients with
Alzheimer’s disease. Int J Geriatr Psychiatry 1997;12:913–9.
[172] Manfredi PL, Breuer B, Wallenstein S, Stegmann M, Bottomley G,
Libow L. Opioid treatment for agitation in patients with advanced
dementia. Int J Geriatr Psychiatry 2003;18:700 –5.
[173] Serfaty M, Kennell-Webb S, Warner J, Blizard R, Raven P. Double
blind randomised placebo controlled trial of low dose melatonin for
sleep disorders in dementia. Int J Geriatr Psychiatry 2002;17:
1120 –7.
[174] Singer C, Tractenberg RE, Kaye J, Schafer K, Gamst A, Grundman
M, et al. A multicenter, placebo-controlled trial of melatonin for
sleep disturbance in Alzheimer’s disease. Sleep 2003;26:893–901.
 Alzheimer’s & Dementia 3 (2007) 398 – 403

Management of dementia with a cerebrovascular component

Christian Boctia,*, Sandra Blackb, Chris Frankc
a
Division of Neurology, Department of Medicine, Université de Montréal, Montréal, Quebec, Canada
b
Division of Neurology, Department of Medicine, University of Toronto, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
c
Program for Care of the Elderly, Department of Medicine, Queen’s University Providence Continuing Care Center, Kingston, Ontario, Canada

Abstract Background: Cerebrovascular disease (CVD) has been reconceptualized as an important factor in
cognitive decline and dementia. The concept of vascular cognitive impairment (VCI) has been
introduced to capture the whole spectrum of cognitive decline associated with CVD, and to
emphasize the possibility of early intervention on vascular risk factors.
Methods: Intervention studies in prevention and symptomatic therapy of dementia associated with
CVD are reviewed.
Results: We present recommendations based on the review of current literature. There is some
evidence that treating hypertension can prevent cognitive decline. Galantamine has shown some
benefit in the symptomatic treatment of possible Alzheimer’s disease with CVD and donepezil has
shown some benefits in vascular dementia. Other pharmacologic interventions lack sufficient
empirical support.
Conclusions: A more pro-active stance towards cognitive decline of vascular etiology is warranted,
with some evidence supporting both prevention and symptomatic therapy.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Vascular dementia; Vascular cognitive impairment; Alzheimer’s disease; Guidelines; Prevention; Therapy

1. Introduction cholinesterase inhibitors have been carried out in dementia

Since the last Canadian Consensus Conference on De-
mentia [1], cerebrovascular disease has been reconceptual-
ized as a major factor in cognitive decline and dementia in
the elderly [2– 4]. Some of the most important concepts are
(1) virtually all traditional risk factors for cerebrovascular
disease also are risk factors for Alzheimer’s disease [5,6];
(2) subcortical ischemic lesions such as lacunar infarcts
greatly influence the clinical syndrome of dementia [7–9];
and (3) the most common pathologic substrate of dementia
in population-based autopsy series is combined cerebrovas-
cular disease and Alzheimer’s pathology [10,11]. In addi-
tion to these epidemiologic and pathologic interactions,
some evidence supports a cholinergic deficit in vascular
dementia [12], presumably through vascular damage to the
cholinergic pathways [13,14]. Clinical trials with two of the
*Corresponding author. Tel.: 514-252-3528; Fax: 514-252-3529.
E-mail address: cbocti@gmail.com
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.009
associated with cerebrovascular disease [15–17].
2. Definitions
These recent concepts are just beginning to generate new
approaches not only to vascular dementia (VaD) but also to
cognitive decline at any degree of severity, including the
“brain-at-risk” and predementia stages in which the most
efficient preventive strategies could be implemented [18].
The concept of vascular cognitive impairment (VCI) has
been introduced to emphasize the high prevalence of cog-
nitive impairment with a vascular component and foster a
more proactive stance toward cognitive decline. VCI en-
compasses all degrees of severity of cognitive deficits re-
sulting from vascular disease: vascular cognitive impair-
ment, no dementia (V-CIND); VaD, which is itself
heterogeneous; and mixed Alzheimer’s disease and cerebro-
vascular disease (AD-CVD) [4,19,20]. An alternative um-
brella term to VCI has been proposed to further clarify the
same concept, vascular cognitive disorders [21]. Here the
 C. Bocti et al. / Alzheimer’s & Dementia 3 (2007) 398 – 403
term VCI is equivalent to V-CIND or vascular mild cogni-
tive impairment, and the other two main subtypes remain
VaD and AD-CVD. In both of these schemes, the inclusion
of the mildest form of cognitive impairment will hopefully
enable both patients and health professionals to optimize
preventive strategies before the stage of VaD is attained.
Many of the propositions just presented have been for-
mulated in response to the consensus-derived National In-
stitute of Neurological Disorders and Stroke–Association In-
ternationale pour la Recherche et l’Enseignement (NINDS–
AIREN) VaD criteria [22]. These criteria are the most
commonly used and include three concepts: (1) evidence of
dementia, (2) evidence of vascular disease of the brain, and
(3) evidence of a temporal link between (1) and (2). They
have been used successfully in clinical trials to identify
patients who evolve differently than AD patients [16,17].
However, many critiques are well-founded and justify the
current effort to better define the spectrum of VaD. One
fundamental limitation is that the definition is modeled on
the cognitive profile of AD, in which memory is promi-
nently affected. The most common cognitive deficit of VaD,
in contrast, is frontal/executive dysfunction [23,24]. Most
clinical tools developed for AD do not capture this type of
cognitive deficit [25], with the consequence that the contri-
bution of vascular disease is often clinically unrecognized.
A possible solution to that problem is the recent publication
of a neuropsychology protocol that is specifically aimed at
screening patients with VCI, including a 5-minute battery
with a real potential for widespread clinical use [26]. An-
other limitation of the VaD criteria comes from the imaging
criteria; they are quite complex to apply [27] and do not
appear to be specific in unselected stroke populations [28].
There is a need for prospective, detailed information about
the clinical, cognitive, imaging, and functional aspects of
VaD to develop new criteria. There are no universally ac-
cepted criteria for V-CIND. The criteria derived from the
Canadian Study of Health and Aging have been shown to
have important clinical implications; rates of death or insti-
tutionalization were similar to those of AD subjects after 5
years [29]. Other criteria have been proposed by Italian
investigators for subcortical vascular MCI, and the clinical
outcomes have similarly been far from benign [30]. These
criteria are modified from research criteria for subcortical
vascular dementia proposed in 2000, defining a more ho-
mogeneous subgroup of patients within the VaD category
[31]. A major strength of these criteria is the adequate
emphasis on executive dysfunction as the major cognitive
domain affected. As for the AD-CVD category or mixed
dementia, there is an important need for more elaborate and
valid criteria. The original AD criteria provide the “possible
AD with cerebrovascular disease” category [32], but the
relative contribution of each pathologic process is a chal-
lenge to establish in most cases [33].
399
3. Methods
The current review considered all randomized controlled
trials of interventions to improve VaD. A preliminary list of
potential articles was obtained through an exhaustive En-
glish language literature review on PubMed and Embase
with the terms “Vascular dementia OR Multi-infarct demen-
tia OR Vascular cognitive impairment OR Vascular cogni-
tive disorders OR Subcortical Ischemic Vascular Dementia”
and “management OR treatment OR prevention OR cholin-
ergic agents”. The list was reduced to 29 articles on the
basis of methodologic quality and non-overlap with other
reports. We included here only those that dealt with drugs
easily available in Canada.
A standardized analysis of the quality of the evidence
was undertaken, integrating a rating of the quality of the
report with the Jadad scale [34] and such quantitative as-
pects as statistical significance and magnitude of effects
(naturalstandard.com) [35].
The following categories will be utilized to grade the
levels of evidence.
(I) Evidence obtained from at least one properly ran-
domized controlled trial.
(II-1) Evidence obtained from well-designed controlled tri-
als without randomization.
(II-2) Evidence obtained from well-designed cohort or case
control analytic studies preferably from more than
one center or research group.
(II-3) Evidence obtained from comparisons between times
or places with or without the intervention. Dramatic
results in uncontrolled experiments are included in
this category.
(III) Opinions of respected authorities based on clinical
experience, descriptive studies, or reports of expert
committees [36].
Recommendations were graded as follows:
(A) There is good evidence to support this maneuver.
(B) There is fair evidence to support this maneuver.
(C) There is insufficient evidence to recommend for or
against this maneuver, but recommendations might
be made on other grounds.
(D) There is a fair evidence to recommend against this
procedure.
(E) There is good evidence to recommend against this
procedure.
4. Nonpharmacologic interventions
Few studies have been published on this topic. A
Cochrane database review included six randomized con-
trolled trials of cognitive training in dementia [37]. Only
one of these articles included patients with vascular demen-
tias [38]. No benefit was reported in any of the main out-
400 C. Bocti et al. / Alzheimer’s & Dementia 3 (2007) 398 – 403
comes for any of these studies (cognitive, global, functional,
and behavioral outcomes).
4.1. Recommendation 1
4.1.1. Use of nonpharmacologic interventions
There is currently insufficient evidence to recommend
the use of cognitive training for vascular dementia (Grade
C, Level 2).
5. Pharmacologic therapy in vascular dementia
5.1. Aspirin
Although the use of aspirin for prevention of recurrent
stroke is supported by considerable evidence [39], its use in
vascular dementia is not. According to a Canadian survey,
the clinical practice of prescribing aspirin as an antiplatelet
in the context of VaD is very common (86% of respondents)
[40]. There are no clinical trials supporting potential bene-
fits of this intervention [41]. This information needs to be
weighted against the potential risk of hemorrhage, which
might be increased in some patients under consideration here;
cerebral microbleeds as detected by gradient-echo magnetic
resonance imaging are very prevalent in subjects with subcor-
tical ischemic damage. Microbleeds could be a marker of
increased hemorrhagic risk, although a recent review failed to
find compelling evidence to modify clinical practice [42].
5.2. Recommendation 2
5.2.1. Antiplatelet therapy with aspirin
There is currently no evidence to support the use of
aspirin to specifically treat dementia associated with cere-
brovascular disease (Grade C, Level 3). Aspirin or other
antiplatelet therapies should be used for prevention of recur-
rent ischemic stroke in appropriate patients [43] (Grade A,
Level 1).
6. Blood pressure reduction
In recent years there has been increasing interest in blood
pressure lowering as a means to prevent dementia or protect
against further cognitive decline. The evidence supporting
such a practice is included here. Because hypertension is a
major risk factor for stroke and stroke can be associated
with dementia, it is biologically plausible that reducing
hypertension will reduce dementia and cognitive decline.
Although there are promising results from two of four large
randomized controlled trials of antihypertensive drugs
(Syst-Eur and Progress) [44,45], two others failed to show a
benefit for cognitive outcomes in the treatment arm (SHEP,
SCOPE) [46,47]. Methodologic issues limit the conclusions
that can be drawn specifically for the prevention of demen-
tia. Although numbers of subjects were large in the two
positive trials, cognitive testing was relatively crude, and
primary end points were noncognitive. In addition, event
rates were low, resulting in very wide confidence intervals,
and a meta-analysis of pooled data from the four aforemen-
tioned trials failed to reach significance [48].
6.1. Recommendation 3
6.1.1. Investigations for vascular risk factors
It is recommended that vascular risk factors are identified
in all patients with vascular cognitive impairment (Grade C,
Level 3).
6.2. Recommendation 4
6.2.1. Treating hypertension
There is some evidence that treating hypertension might
prevent further cognitive decline associated with cerebro-
vascular disease. There is no compelling evidence that one
class of agent is superior to another; calcium channel block-
ers or angiotensin-converting enzyme inhibitors might be
considered (Grade B, Level 1).
Treatment for hypertension should be implemented for
other reasons, including the prevention of recurrent stroke
(Grade A, Level 1).
7. Symptomatic treatment
Several molecules have been the object of a controlled
trial in symptomatic treatment of VaD. Nimodipine, me-
mantine, donepezil, and galantamine all had randomized
controlled trials supporting some benefits in VaD.
7.1. Nimodipine
Several trials have been conducted with nimodipine in
both degenerative and vascular dementias. A recent
Cochrane review included a total of 14 trials of nimodipine,
of which seven reported sufficient data to be included in the
meta-analysis [49]. The total number of subjects included
was 2,492, the typical trial duration was 12 weeks, and the
dosage of nimodipine was 90 mg/day. Results favored treat-
ment over placebo for global scales and cognitive outcomes
in treatment completers with “cerebrovascular dementia”.
The drug was well-tolerated, with remarkably low dropout
rates in most studies. The benefits were of small magnitude,
and no benefit could be demonstrated for behavioral or
functional outcomes in any of the trials. The strength of the
conclusions is undermined by the lack of data for fully half
of the identified reports.
7.2. Recommendation 5
7.2.1. Nimodipine in vascular dementia
There is insufficient evidence for or against the use of
nimodipine for VaD (Grade C, Level 1).
8. Cholinesterase inhibitors
One large-scale (n ⫽ 592), high quality (Jadad scale ⫽ 5)
trial with galantamine in a mixed population of patients with
 C. Bocti et al. / Alzheimer’s & Dementia 3 (2007) 398 – 403
probable VaD (n ⫽ 252) and possible AD with CVD (n ⫽
295) showed overall positive results on cognitive, func-
tional, and behavioral measures. The effect size was small
(0.3 for cognitive outcome), commensurate with that ob-
served in trials of AD. Dropout rates as a result of gastro-
intestinal adverse events were significantly higher with the
titration dosage used in that study. The overall effect was
driven by the AD-CVD group, because the results did not
reach statistical significance for the probable VaD group.
This trial was not powered to draw conclusions on the
probable VaD subgroup [14].
Two identical large-scale double-blind randomized con-
trolled trials (Jadad scale ⫽ 5) have been conducted with
donepezil in vascular dementia, with a total sample size of
1,219 subjects [16,17]. Outcomes have been statistically
significant for cognition in both trials, with small effect size
(Alzheimer’s Disease Assessment Scale, cognitive portion
and Mini-Mental State Examination). For global scales and
functional outcomes, there were inconsistencies between
both dosages used, 5 mg or 10 mg daily. There was benefit
on the Clinician’s Interview-Based Impression of Change
(CIBIC)-plus for both dosages. On another measure, the
Clinical Dementia Rating-Sum of Boxes, the lower 5-mg
dosage did not produce significant results, but the10-mg
dose did. The same situation prevailed for the functional
outcome, the Alzheimer’s Disease Functional Assessment
and Change Scale. The subscore for Instrumental Activities
of Daily Living, however, was significant for both doses.
Tolerability was apparently better for the lower dosage, with a
differential dropout rate for the 10-mg dose. It was suggested
that the lower dosage might be preferable in this population.
There is no high quality trial for rivastigmine in vas-
cular dementia. A post hoc analysis that compared AD
patients with and without vascular risk factors showed a
somewhat increased magnitude of benefit in the first
group [50]. An open uncontrolled trial of rivastigmine
showed benefits of small magnitude in cognitive func-
tions and behavior [51].
8.1. Recommendation 6
8.1.1. Use of cholinesterase inhibitors in dementia caused
by AD-CVD
There is fair evidence of benefits of small magnitude for
galantamine in cognitive, functional, behavioral, and global
measures in AD with CVD. Galantamine can be considered
a treatment option for mixed AD with CVD (Grade B,
Level 1).
8.2. Recommendation 7
8.2.1. Use of cholinesterase inhibitors in
probable/possible vascular dementia with the NINDS-
AIREN diagnostic criteria
There is insufficient evidence for or against the use of
galantamine (Grade C, Level 1).
401
There is fair evidence of benefits of small magnitude for
donepezil in cognitive and global outcomes, with less robust
benefits on functional measures. Donepezil can be considered
a treatment option for vascular dementia (Grade B, Level 1).
9. Memantine
Memantine, an N-methyl-D-aspartate receptor antago-
nist, has been the object of two trials in vascular dementia
with a total sample of 900 patients [52,53], and a further
trial included a subgroup of VaD patients, but the results are
not reported separately [54]. The quality of the two reports
reviewed here varies from good to excellent (Jadad scale, 4
and 5). Primary outcomes have been favorable to treatment
only for cognitive measures in both trials, with small effect
size in both (0.22). The global scales failed to capture any
significant benefit (Clinical Global Impression of Change,
CIBIC-plus). The drug was well-tolerated.
9.1. Recommendation 8
9.1.1. Use of memantine
There is some evidence of small magnitude of cognitive
benefit that is not captured in global measures for patients
with VaD. There is insufficient information to recommend
memantine for the treatment of vascular dementia (Grade C,
Level 1).
Author Disclosures
Christian Bocti has received support from Janssen Ortho
Inc. (Speaker, Consultant), Novartis (Speaker) and Pfizer
Canada (Speaker).
Sandra Black has received support from Eisai (Clinical
investigation, CME lecturer, Ad hoc consultant), Pfizer
(Clinical investigation, CME lecturer, Ad hoc consultant),
Janssen-Ortho (Clinical investigation, CME lecturer, Ad
hoc consultant), Novartis (Clinical investigation, CME lec-
turer, Ad hoc consultant), Lundbeck (Ad hoc consultants,
CME lecturer), Sanofi-Aventis (Trial investigator), and
Myriad (Trial investigator, Consultant—Ad hoc).
Chris Frank has nothing to disclose.
Acknowledgments
Christian Bocti has been supported by a scholarship from
the Alzheimer Society of Canada, and an operating grant
from Canadian Institutes for Health Research (CIHR).
Sandra Black’s work has been supported by operating
grants from the CIHR, Alzheimer Association US and Alz-
heimer Society of Canada. She receives salary support from
the Department of Medicine, University of Toronto, and
Sunnybrook Research Institute.
402 C. Bocti et al. / Alzheimer’s & Dementia 3 (2007) 398 – 403
References
[1] Patterson CJ, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus Con-
ference on Dementia. 1999;160(12 Suppl):S1–15.
[2] Bowler JV, Steenhuis R, Hachinski V. Conceptual background to
vascular cognitive impairment. Alzheimer Dis Assoc Disord 1999;
13(Suppl 3): S30 –7.
[3] Roman GC. Vascular dementia may be the most common form of
dementia in the elderly. J Neurol Sci 2002;203-204:7–10.
[4] O’Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni
L, et al. Vascular cognitive impairment. Lancet Neurol. 2003;2:89 –
98.
[5] Decarli C. Vascular factors in dementia: an overview. J Neurol Sci
2004;226:19 –23.
[6] Honig LS, Kukull W, Mayeux R. Atherosclerosis and AD: analysis of
data from the US National Alzheimer’s Coordinating Center. Neu-
rology 2005;64:494 –500.
[7] Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA,
Markesbery WR. Brain infarction and the clinical expression of
Alzheimer disease: the Nun Study. JAMA 1997;277:813–7.
[8] Esiri MM, Nagy Z, Smith MZ, Barnetson L, Smith AD. Cerebrovas-
cular disease and threshold for dementia in the early stages of Alz-
heimer’s disease. Lancet 1999;354:919 –20.
[9] Petrovitch H, Ross GW, Steinhorn SC, Abbott RD, Markesbery W,
Davis D, et al. AD lesions and infarcts in demented and non-
demented Japanese-American men. Ann Neurol 2005;57:98 –103.
[10] Lim A, Tsuang D, Kukull W, Nochlin D, Leverenz J, McCormick W,
et al. Clinico-neuropathological correlation of Alzheimer’s disease in
a community-based case series. J Am Geriatr Soc 1999;47:564 –9.
[11] Neuropathology Group of the Medical Research Council Cognitive
Function and Ageing Study (MRC CFAS). Pathological correlates of
late-onset dementia in a multicentre, community-based population in
England and Wales. Lancet 2001;357:169 –75.
[12] Tohgi H, Abe T, Kimura M, Saheki M, Takahashi S. Cerebrospinal
fluid acetylcholine and choline in vascular dementia of Binswanger
and multiple small infarct types as compared with Alzheimer-type
dementia. J Neural Transm 1996;103:1211–20.
[13] Tomimoto H, Ohtani R, Shibata M, Nakamura N, Ihara M. Loss of
cholinergic pathways in vascular dementia of the Binswanger type.
Dement Geriatr Cogn Disord 2005;19:282– 8.
[14] Bocti C, Swartz RH, Gao FQ, Sahlas DJ, Behl P, Black SE. A new
visual rating scale to assess strategic white matter hyperintensities
within cholinergic pathways in dementia. Stroke 2005;36:2126 –31.
[15] Erkinjuntti T, Kurz A, Gauthier S, Bullock R, Lilienfeld S, Damaraju
CV. Efficacy of galantamine in probable vascular dementia and Alz-
heimer’s disease combined with cerebrovascular disease: a random-
ised trial. Lancet 2002;359:1283–90.
[16] Black S, Roman GC, Geldmacher DS, Salloway S, Hecker J, Burns
A, et al. Efficacy and tolerability of donepezil in vascular dementia:
positive results of a 24-week, multicenter, international, randomized,
placebo-controlled clinical trial. Stroke 2003;34:2323–30.
[17] Wilkinson D, Doody R, Helme R, Taubman K, Mintzer J, Kertesz A,
et al. Donepezil in vascular dementia: a randomized, placebo-
controlled study. Neurology 2003;61:479 – 86.
[18] Hachinski V. Preventable senility: a call for action against the vas-
cular dementias. Lancet 1992;340:645– 8.
[19] Hachinski V. Vascular dementia: a radical redefinition. Dementia.
1994;5:130 –2.
[20] Rockwood K, Davis H, MacKnight C, Vandorpe R, Gauthier S,
Guzman A, et al. The Consortium to Investigate Vascular Impairment
of Cognition: methods and first findings. Can J Neurol Sci 2003;30:
237– 43.
[21] Roman GC, Sachdev P, Royall DR, Bullock RA, Orgogozo JM,
Lopez-Pousa S, et al. Vascular cognitive disorder: a new diagnostic
category updating vascular cognitive impairment and vascular de-
mentia. J Neurol Sci 200415;226:81–7.
[22] Roman GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC,
Garcia JH, et al. Vascular dementia: diagnostic criteria for research
studies—report of the NINDS-AIREN International Workshop. Neu-
rology 1993;43:250 – 60.
[23] Kertesz A, Clydesdale S. Neuropsychological deficits in vascular
dementia vs Alzheimer’s disease: frontal lobe deficits prominent in
vascular dementia. Arch Neurol 1994;51:1226 –31.
[24] Sachdev PS, Brodaty H, Valenzuela MJ, Lorentz L, Looi JC, Wen W,
et al. The neuropsychological profile of vascular cognitive impair-
ment in stroke and TIA patients. Neurology 2004;62:912–9.
[25] O’Sullivan M, Morris RG, Markus HS. Brief cognitive assessment for
patients with cerebral small vessel disease. J Neurol Neurosurg Psy-
chiatry 2005;76:1140 –5.
[26] Hachinski V, Iadecola C, Petersen RC, Breteler MM, Nyenhuis DL,
Black SE, et al. National Institute of Neurological Disorders and
Stroke-Canadian Stroke Network vascular cognitive impairment har-
monization standards. Stroke 2006;37:2220 – 41.
[27] van Straaten EC, Scheltens P, Knol DL, van Buchem MA, van Dijk
EJ, Hofman PA, et al. Operational definitions for the NINDS-AIREN
criteria for vascular dementia: an interobserver study. Stroke 2003;
34:1907–12.
[28] Ballard CG, Burton EJ, Barber R, Stephens S, Kenny RA, Kalaria RN,
et al. NINDS AIREN neuroimaging criteria do not distinguish stroke
patients with and without dementia. Neurology 2004;63:983– 8.
[29] Rockwood K, Wentzel C, Hachinski V, Hogan DB, MacKnight C,
McDowell I. Prevalence and outcomes of vascular cognitive impair-
ment: Vascular Cognitive Impairment Investigators of the Canadian
Study of Health and Aging. Neurology 2000;54:447–51.
[30] Frisoni GB, Galluzzi S, Bresciani L, Zanetti O, Geroldi C. Mild
cognitive impairment with subcortical vascular features: clinical char-
acteristics and outcome. J Neurol 2002;249:1423–32.
[31] Erkinjuntti T, Inzitari D, Pantoni L, Wallin A, Scheltens P, Rockwood
K, et al. Research criteria for subcortical vascular dementia in clinical
trials. J Neural Transm Suppl 2000;59:23–30.
[32] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM. Clinical diagnosis of Alzheimer’s disease: report of the
NINCDS-ADRDA Work Group under the auspices of Department of
Health and Human Services Task Force on Alzheimer’s Disease.
Neurology 1984;34:939 – 44.
[33] Lopez OL, Kuller LH, Becker JT, Jagust WJ, DeKosky ST, Fitz-
patrick A, et al. Classification of vascular dementia in the Cardiovas-
cular Health Study Cognition Study. Neurology 2005;64:1539 – 47.
[34] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Ga-
vaghan DJ, et al. Assessing the quality of reports of randomized
clinical trials: is blinding necessary? Control Clin Trials 1996;17:
1–12.
[35] Available at: http://www.naturalstandard.com/explanation_column-
s.html. Accessed February 5, 2006.
[36] Woolf SH, Battista RN, Anderson GM, Logan AG, Wang E. Assess-
ing the clinical effectiveness of preventive maneuvers: analytic prin-
ciples and systematic methods in reviewing evidence and developing
clinical practice recommendations—a report by the Canadian Task
Force on the Periodic Health Examination. J Clin Epidemiol 1990;
43:891–905.
[37] Clare L, Woods RT, Moniz Cook ED, Orrell M, Spector A. Cognitive
rehabilitation and cognitive training for early-stage Alzheimer’s dis-
ease and vascular dementia. Cochrane Database Syst Rev 2003(4):
CD003260.
[38] Quayhagen MP, Quayhagen M, Corbeil RR, Hendrix RC, Jackson JE,
Snyder L, et al. Coping with dementia: evaluation of four nonphar-
macologic interventions. Int Psychogeriatr 2000;12:249 – 65.
 C. Bocti et al. / Alzheimer’s & Dementia 3 (2007) 398 – 403
[39] Antiplatelet Trialists’ Collaboration. Collaborative overview of trials
of antiplatelet therapy. I: prevention of death, myocardial infarction,
and stroke by prolonged antiplatelet therapy in various categories of
patients. Br Med J 1994;308:81–106.
[40] Molnar FJ, Man Song Hing M, St John P, Brymer C, Rockwood K,
Hachinski V. Subcortical vascular dementia: survey of treatment
patterns and research considerations. Can J Neurol Sci 1998;25:
320 – 4.
[41] Williams PS, Spector A, Orrell M, Rands G. Aspirin for vascular
dementia. Cochrane Database Syst Rev 2000(2):CD001296.
[42] Koennecke HC. Cerebral microbleeds on MRI: prevalence, associa-
tions, and potential clinical implications. Neurology 2006;66:165–71.
[43] Goldstein LB, Adams R, Alberts MJ, Appel LJ, Brass LM, Bushnell
CD, et al. Primary prevention of ischemic stroke: a guideline from the
American Heart Association/American Stroke Association Stroke
Council: cosponsored by the Atherosclerotic Peripheral Vascular Dis-
ease Interdisciplinary Working Group; Cardiovascular Nursing Coun-
cil; Clinical Cardiology Council; Nutrition, Physical Activity, and
Metabolism Council; and the Quality of Care and Outcomes Research
Interdisciplinary Working Group: the American Academy of Neurol-
ogy affirms the value of this guideline. Stoke 2006;37:1583– 633.
[44] Forette F, Seux ML, Staessen JA, Thijs L, Birkenhager WH, Babar-
skiene MR, et al. Prevention of dementia in randomised double-blind
placebo-controlled systolic hypertension in Europe (Syst-Eur) trial.
Lancet 1998;352:1347–51.
[45] Tzourio C, Anderson C, Chapman N, Woodward M, Neal B,
MacMahon S, et al. Effects of blood pressure lowering with
perindopril and indapamide therapy on dementia and cognitive
decline in patients with cerebrovascular disease. Arch Intern Med
2003;163:1069 –75.
[46] Di Bari M, Pahor M, Franse LV, Shorr RI, Wan JY, Ferrucci L, et al.
Dementia and disability outcomes in large hypertension trials: lessons
403
learned from the systolic hypertension in the elderly program (SHEP)
trial. Am J Epidemiol 2001;153:72– 8.
[47] Lithell H, Hansson L, Skoog I, Elmfeldt D, Hofman A, Olofsson B,
et al. The study on cognition and prognosis in the elderly (SCOPE):
principal results of a randomized double-blind intervention trial.
J Hypertens 2003;21:875– 86.
[48] Feigin V, Ratnasabapathy Y, Anderson C. Does blood pressure low-
ering treatment prevent dementia or cognitive decline in patients with
cardiovascular and cerebrovascular disease? J Neurol Sci 2005;229-
230:151–5.
[49] Lopez-Arrieta JM, Birks J. Nimodipine for primary degenerative,
mixed and vascular dementia. Cochrane Database Syst Rev 2002(3):
CD000147.
[50] Kumar V, Anand R, Messina J, Hartman R, Veach J. An efficacy and
safety analysis of Exelon in Alzheimer’s disease patients with con-
current vascular risk factors. Eur J Neurol 2000;7:159 – 69.
[51] Moretti R, Torre P, Antonello RM, Cazzato G, Bava A. Rivastigmine
in subcortical vascular dementia: a randomized, controlled, open
12-month study in 208 patients. Am J Alzheimers Dis Other Demen
2003;18:265–72.
[52] Orgogozo JM, Rigaud AS, Stoffler A, Mobius HJ, Forette F. Efficacy
and safety of memantine in patients with mild to moderate vascular
dementia: a randomized, placebo-controlled trial (MMM 300). Stroke
2002;33:1834 –9.
[53] Wilcock G, Mobius HJ, Stoffler A. A double-blind, placebo-
controlled multicentre study of memantine in mild to moderate vas-
cular dementia (MMM500). Int Clin Psychopharmacol 2002;17:297–
305.
[54] Winblad B, Poritis N. Memantine in severe dementia: results of the
9M-Best Study (Benefit and efficacy in severely demented patients
during treatment with memantine). Int J Geriatr Psychiatry 1999;14:
135– 46.
 Alzheimer’s & Dementia 3 (2007) 404 – 410

Disclosure of the diagnosis of dementia

John D. Fisk *, B. Lynn Beattieb, Martha Donnellyc, Anna Byszewskid, Frank J. Molnare
a,
a
QE II Health Sciences Centre; Departments of Psychiatry, Medicine, and Psychology, Dalhousie University, Halifax, Nova Scotia, Canada
b
Division of Geriatric Medicine, Department of Medicine, UBC, Vancouver, British Columbia, Canada
c
Division of Community Geriatrics, Department of Family Practice, and Division of Geriatric Psychiatry, Department of Psychiatry, UBC; and
Geriatric Psychiatry Outreach Team, Vancouver Hospital, Vancouver, British Columbia, Canada
d
Department of Medicine, University of Ottawa; Geriatric Day Hospital, Ottawa Hospital; and Ottawa Health Research Institute, Ottawa,
Ontario, Canada
e
Division of Geriatric Medicine, Department of Internal Medicine, Ottawa Hospital; Department of Medicine, University of Ottawa; Ottawa Health
Research Institute; and Elisabeth Bruyere Research Institute, Ottawa, Ontario, Canada

Abstract Most ethical guidelines strongly promote disclosure of a diagnosis of dementia to the affected
individual, based on the principle of autonomy. Nevertheless, codes of medical ethics allow for
various interpretations of this issue and surveys of clinical practice illustrate that such disclosure is
by no means the rule. We argue that diagnostic disclosure for persons with dementia must be
considered a process that begins when cognitive impairment is first suspected and that evolves over
time as information is obtained. Whenever possible and appropriate, this process should involve not
only the affected individual but also their family and/or other current or potential future care
providers. Once a diagnosis is established it should be disclosed in a manner consistent with the
expressed wishes of the patient, using an individualized patient-centered approach that maintains
the individual’s personal integrity. Diagnostic disclosure of dementia is a process that may require
additional time as well as follow-up or referral to other specialists. We recommend that a progres-
sive disclosure process be employed to address issues including: remaining diagnostic uncertainty,
treatment options, future plans, financial planning, assigning power of attorney, wills and “living
wills”, driving privileges and the need for eventual driving cessation, available support services, and
potential research participation. The potential for adverse psychological consequences to diagnostic
disclosure must be assessed and these should be addressed through education and support of the
patient and their family/caregivers throughout the diagnostic disclosure process. At present, few data
are available regarding patients’ perspectives on the diagnostic disclosure process and its conse-
quences. This limitation and the apparent discrepancies in physician and caregiver opinions about
the disclosure process, make it incumbent upon health care professionals to evaluate the diagnostic
disclosure process within their practice.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Alzheimer’s disease; Dementia; Diagnosis; Ethics

Although various ethical guidelines for medical care
exist, all with common underlying principles, few have
specific recommendations regarding dementia. In 1997 the
Alzheimer Society of Canada (ASC) released its Ethical
Guidelines [1] document, which represented the results of a
2-year consultative process that was based on a “discourse
ethics” approach previously used in developing the Fairhill
*Corresponding author. Tel.: 902-473-2581; Fax: 902-473-7166.
E-mail address: John.Fisk@cdha.nshealth.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.008
Guidelines [2]. This approach used focus group discussions
that included caregivers, individuals with mild Alzheimer’s
disease, and professionals (eg, physicians, nurses, lawyers,
ethicists, and administrators) but also extended this consul-
tation process by disseminating the information nationally
for review and comment [3]. Both the Fairhill Guidelines
and the ASC Ethical Guidelines addressed issues important
in clinical practice and everyday life for persons with Alz-
heimer’s disease, such as diagnostic disclosure, driving
privileges, autonomy in decision making, maintaining qual-
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
ity of life, appropriate use of restraints, and end-of-life care.
The ASC Ethical Guidelines, however, also addressed ge-
netic testing and participation in research. In 1999 the Ca-
nadian Consensus Conference on Dementia (CCCD) ac-
knowledged the diversity and scope of ethical issues
addressed by the ASC guidelines and chose to focus on
discussion of diagnostic disclosures and driving privileges
[4]. Revision of the ASC Ethical Guidelines through an-
other consultative process was undertaken in 2001–2002 in
response to increased awareness, earlier diagnosis, the con-
ceptualization of mild cognitive impairment (MCI) [5], and
the availability of symptomatic treatments. For the current
revision of the CCCD, the topics discussed in the original
document were reconsidered with a view to both updating
them and adding new relevant issues. First, it was decided
that because of its complexity, the issue of driving needed to
be considered as a separate topic with discussions not lim-
ited to that of ethical issues. Second, it was determined that
the issue of diagnostic disclosure warranted a review in light
of new developments and, in particular, because of the
importance of diagnostic disclosure as the starting point
from which treatment and management decisions follow.
For this review the PubMed and Embase databases were
searched for articles with the keywords “Dementia OR
Alzheimer’s disease AND ethics AND diagnosis.” For dis-
cussion, preference was given to publications between 1996
and 2006.
As described by Maguire [6], “Clinicians who diagnose
patients with dementia are faced with a number of ethical
dilemmas: They must be truthful, yet do no harm; they must
respect patients’ autonomy, yet consider the concerns of
those who live with and care for those patients.” Most
ethical guidelines dealing with diagnostic disclosure to
those with dementia and their family/caregivers, including
those recommendations put forth in the previous CCCD,
strongly promote the disclosure of a diagnosis of dementia
to the affected individual, on the basis of the principle of
autonomy [4]. This viewpoint is not universally held, how-
ever, and codes of medical ethics in general allow for
various interpretations of the issue of diagnostic disclosure
[7]. Two relatively recent systematic reviews have been
undertaken by groups in the United States [8] and in the
United Kingdom [9]. Both point out the need for continued
discussion, research, practice guideline refinement, and ed-
ucation of professionals and the public about this issue. The
need for a Canadian perspective on these issues is rein-
forced by the finding that only 4% of the articles discussing
diagnostic disclosure were published in Canada [8].
We have little information about whether available prac-
tice guidelines are known by practitioners and how practi-
tioners’ experiences with dementia influence their accep-
tance and use of such guidelines. A review of studies of
general practitioners’ practices in the United States regard-
ing diagnostic disclosure concluded that approximately half
routinely withhold disclosure to the patient [8]. A survey of
405
general practitioners in the United Kingdom identified com-
munication about diagnosis as one of the “main difficulties”
experienced by general practitioners when dealing with pa-
tients with dementia [10]. Those who reported greater dif-
ficulty with the diagnosis and management of dementia
were also less likely to endorse open communications strat-
egies for diagnostic disclosure [10]. This same association
between difficulty in establishing a diagnosis and disclosing
the diagnosis was also reported in an American sample of
primary care physicians [11].
Although guidelines favoring diagnostic disclosure are
based on the principle of patient autonomy, the arguments
that have typically been put forth against diagnostic disclo-
sure are based largely on the principle of non-maleficence
(ie, the obligation to avoid harm). Justification for the em-
phasis on this latter principle has typically relied on (1) the
lack of absolute diagnostic certainty from clinical informa-
tion, (2) the absence of effective treatments of progressive
dementia, (3) the questionable ability of persons with more
advanced disease to understand the implications of the di-
agnosis, and (4) the potential for adverse psychological
responses to diagnostic disclosure as their justification [12].
Diagnostic certainty, although never 100%, has im-
proved considerably through the development of diagnostic
criteria that now span a range of less prevalent forms of
dementia in addition to Alzheimer’s disease [13]. These
criteria are unlikely to be used routinely outside specialty
referral clinics, and lack of familiarity in their use among
primary care physicians might contribute to uncertainty in
dementia diagnosis. Nevertheless, lack of access to a diag-
nostic opinion from a specialty clinic and lack of certainty
about a suspected diagnosis do not negate the responsibility
of primary care physicians to provide “an open, honest
presentation of information as it is perceived and known”
[12]. Moreover, as recommended by the previous CCCD,
the topic of diagnostic uncertainty itself should be discussed
within the context of diagnostic disclosure (Recommenda-
tion 21) [4]. In recent years, MCI has become increasingly
well-defined to the point where it can be considered as a
diagnostic entity itself rather than just a risk factor for
dementia [14], making it increasingly important that the
same considerations for diagnostic disclosure apply to this
condition. However, aside from diagnostic uncertainly, dis-
closure of a diagnosis of MCI is also complicated by prog-
nostic uncertainty [15]. Explaining uncertainty to patients
and family members is never a simple matter, as was evi-
dent in a qualitative study of the psychosocial impact of
disclosing a diagnosis of dementia to patients and their
family/caregivers by Smith and Beattie [16]. They found
that family/caregivers’ understanding of the clinician’s un-
certainty about the diagnosis was subject to their precon-
ceived ideas about the source of their loved ones’ cognitive
problems. Such findings illustrate the importance of estab-
lishing a clear understanding of the perspectives of both the
patient and their family/caregivers to accurately convey
406 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
both the diagnostic opinion and the uncertainty that might
surround this opinion.
The diagnosis of dementia presents a framework within
which target symptoms and symptomatic treatment options
can be discussed. Although the absence of effective disease-
modifying treatments had previously been raised as a po-
tential argument against diagnostic disclosure, such argu-
ments were rejected in the ASC Ethical Guidelines [1] even
before the widespread availability of cholinesterase inhibi-
tors for the treatment of Alzheimer’s disease. Smith and
Beattie [16] noted that “disclosure facilitated patient care
for all families,” and it is important that care not be equated
with pharmacologic treatment alone. Even in the absence of
approved symptomatic pharmacologic treatments for a
given diagnosis, treatment of specific symptoms of demen-
tia might involve either pharmacologic or behavioral inter-
ventions. Prevention strategies for future problems can be
discussed meaningfully with patients and their family/care-
givers once a diagnosis has been disclosed. Even though
approved pharmacologic treatments for Alzheimer’s disease
remain symptomatic and treatments for other forms of de-
mentias are limited, diagnostic disclosure can open a dia-
logue among the patient, their family/caregivers, and health
care professionals about a broad range of important issues.
These include, but are not limited to, financial planning,
assigning power of attorney, wills and “living wills”, driv-
ing privileges, research participation, and the availability
and access to formal and informal community services.
Without an effective process of diagnostic disclosure, this
dialogue is not possible.
The previous CCCD cited “severe dementia where un-
derstanding of the diagnosis is unlikely” as a possible ex-
ception to diagnostic disclosure to the patient (Recommen-
dation 21) [4]. Although this has historically been described
throughout discussions of diagnostic disclosure, one hopes
that a diagnosis will usually be made long before a severe
stage is reached. Regardless, for some persons with demen-
tia, profound lack of insight is an early feature and is not a
reflection of severe disease [17]. Although it can be argued
that the principle of autonomy no longer applies to persons
who are incapable of understanding what is being disclosed
to them, obtaining a clear understanding of the level of
insight of persons with dementia is challenging [18]. More-
over, an inability to fully appreciate the implications of a
diagnosis need not preclude the possibility of deriving any
benefit from diagnostic disclosure. The opinion expressed
more than a decade ago that the individual patient has a
“moral and legal right . . . to receive a specific diagnosis
unless he or she waives it” [2] asserts the principle of
autonomy in ethical decision making and challenges clini-
cians to base their decisions on more than personal conve-
nience and their presumptions about the patients’ potential
to benefit from knowledge of their diagnosis. The difficulty
faced by clinicians in disclosing a diagnosis, and in dealing
with the psychosocial consequences of this disclosure, does
not absolve them of the responsibility to respect patients’
autonomy. As described by Marzanski [7], “The truth may
be neither fully understood nor remembered by the patient
and difficult for the doctor, but neither of these problems
should remove the obligation to be honest and truthful”.
Although the availability of clinical diagnostic criteria
for improved early and differential diagnosis, as well as
symptomatic treatments for Alzheimer’s disease, has ad-
dressed some of the arguments against diagnostic disclo-
sure, the potential for adverse psychological consequences
remains an important consideration and is frequently cited
by general practitioners, among others, as the rationale for
failing to disclose a diagnosis [19]. However, it is important
to note that concerns about adverse psychological reactions
are most often raised by family/caregivers rather than the
patients themselves [20]. Surveys of nonimpaired elderly,
faced with the hypothetical situation of having dementia,
have consistently found that the majority indicate that they
would want to be told of their diagnosis [21,22]. This was
found to be true even for family/caregivers who themselves
did not want their impaired family member to be told of the
diagnosis of dementia [20]. Despite the views favoring
diagnostic disclosure to patients that have been expressed in
most guidelines, including the previous CCCD [4], surveys
of patients and their family/caregivers have typically found
that family/caregivers are told of the diagnosis more often
than the affected individual [23] and that family/caregivers
frequently indicate that they do not want the affected indi-
vidual to be told [24]. If one respects the principle of the
autonomy of the individual, however, decisions regarding
diagnostic disclosure should be based on an accurate ap-
praisal of the patient’s prior wishes rather than the caregiv-
er’s current wishes.
Among the possible negative consequences of diagnostic
disclosure, the potential for suicide is often cited as a reason
to consider withholding the diagnosis [4]. Suicide among
persons with dementia appears to be rare, however, with
reports limited to only a few cases of individuals with
relatively early stage Alzheimer’s disease who were de-
scribed as having preserved insight into their disabilities
[25,26]. Varied approaches to this issue have so far failed to
establish an association between increased suicide risk and
the disclosure of a diagnosis of dementia. A study of family
members attending a community dementia support group
reported that suicidal thoughts were expressed by only two
of 28 patients who were told of their diagnosis, and that
neither had acted on these [23]. Another study found less
than a 1% prevalence of suicidal thoughts in a sample of
221 demented patients as well as an association between
clinical depression and a stated “wish to die” that did not
appear different from that seen in general population sur-
veys of the elderly [27]. A survey of nondemented patients
attending an outpatient clinic found only three of 182 re-
spondents to a questionnaire about diagnostic disclosure
indicated that they would consider suicide if told of a
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
diagnosis of Alzheimer’s disease [28]. Finally, the findings
of a retrospective study of autopsy cases suggest that it
might be those individuals with Alzheimer’s disease who
are undiagnosed who are at greater risk for committing
suicide than the general population [29]. To date, there
remains limited data regarding the adverse psychological
reactions to receiving a diagnosis of dementia. The need for
better understanding of the psychological consequences of a
disclosure of a diagnosis of dementia is clear as is the need
to constantly update such knowledge as new developments
in diagnosis and treatment occur. To fully understand the
psychological consequences of diagnostic disclosure, we
must consider not just the short-term consequences; in
which shock, anger, denial, and depression might be more
common, but also the long-term consequences in which the
benefits of disclosure on psychological well-being might
become more evident [8].
The issue is not one of whether to disclose the diag-
nosis of dementia or not but rather how and when to do
so [30]. The consensus opinion of the previous CCCD [4]
and subsequent commentaries [31] favors an individual-
ized approach to diagnostic disclosure that is sensitive to
each patient’s unique situation and that also involves
family/caregivers in the process. Although it might well
be the case that those who more fully appreciate the
implications of the diagnosis are the family/caregivers of
the patient, a patient-centered approach to diagnostic
disclosure might help family/caregivers appreciate that
their loved one is respected as an individual, regardless of
their medical condition. The process of diagnostic dis-
closure might itself serve as an opportunity to appropri-
ately model communication strategies for the patient and
their family/caregivers. However, despite suggestions as
to how meetings with patients and their family/caregivers
might take place, there remains no research demonstrat-
ing the relative advantages of individual versus joint
disclosure to patients and their family/caregivers.
The increasingly multicultural nature of North America
and the absence of research that identifies how ethnicity or
belief in cultural idioms might affect the process of disclo-
sure of the diagnosis of dementia also pose challenges for
clinicians. Studies suggest that ethnic minority seniors
might be less familiar than Anglo-European Americans with
the biomedical model of disease, instead normalizing de-
mentia as part of the aging process [32,33]. However, in one
study, fully 54% of caregivers, including 41% of Anglo-
European Americans, believed in mixed models of causa-
tion that combined folk explanations and biomedical
elements [34].
In the previous CCCD [4], Recommendation 21 stated
that the diagnosis of dementia should be disclosed to the
patient and to their family, and that in addition to diagnostic
uncertainty, the discussions should include the topics of
prognosis, advance planning, treatment options, support
groups, and future plans. Other recommendations on this
407
issue arising from the ASC guidelines [1] include ensuring
that sufficient time is available for the disclosure process
and considering the possibility of follow-up sessions to
address unresolved issues. Lack of follow-up appointments
and limited discussion of information about community-
based services and treatment options were indeed noted as
sources of dissatisfaction in the diagnostic disclosure pro-
cess for family/caregivers in a qualitative study conducted
at U.S. Alzheimer’s Disease Centers [35]. Providing ade-
quate time for the detailed discussion of assessment results
helps ensure that patients and family members appreciate
that the diagnostic process has been thorough, and even if
the diagnostic opinions are discrepant from their own, they
seem more likely to accept them as valid [16]. Providing
written general educational materials has also been recom-
mended [30], although details that are specific for the indi-
vidual might be helpful as well. For example, one study
found that regardless of the severity or type of dementia,
patients generally had a positive response to reading their
unedited clinic notes [36].
Early diagnosis might be the key to resolving much of
the debate surrounding the ethics of diagnostic disclosure.
Early diagnosis clearly limits arguments against disclosure
that are based on a presumed failure to understand the
diagnosis, its implications, or its uncertainty. Because care-
givers seem more likely to express preferences for lack of
diagnostic disclosure when the patient is more severely
affected, it seems likely that family/caregivers will be more
supportive of diagnostic disclosure to the patient if the
diagnosis is made earlier [24,35]. However, important dis-
crepancies between physicians’ and family/caregivers’
opinions about the disclosure process have also been de-
scribed. Connell et al [35] found that although physicians
often believed that the disclosure process had gone well;
many family/caregivers reported that the disclosure was too
direct and insensitive. They concluded that family/caregiv-
ers “would have liked to have the diagnosis disclosed in a
compassionate manner and to include the patient in the
office visit in a manner that preserves their dignity and a
sense of hope” [35].
Much as establishing the diagnosis of dementia requires
a process, so too does diagnostic disclosure. Disclosure in a
manner that is compassionate for the patient and their fam-
ily/caregiver will vary with individual circumstances, but
the need to provide information about the diagnosis along
with information about what needs to be done next cannot
be communicated effectively in a single office appointment.
Follow-up sessions must be mandatory, although this can
take a variety of forms, potentially including meetings with
the clinician responsible for the diagnostic investigations in
concert with the family physician. This longitudinal process
of disclosure seems crucial for effective communication and
education of the patient and their family/caregivers.
Establishment of guidelines for diagnostic disclosure of
dementia is made difficult by the lack of methodologically
408 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
adequate systematic study of this issue. Most quantitative
studies to date have focused on attitudes and practice rather
than the outcomes of disclosure [9]. Few studies have ex-
amined the perspectives of patients themselves; instead they
focus on those of physicians and family/caregivers [8,9]. A
recent qualitative study of the perspectives of patients with
dementia and their caregivers’ attending a Canadian day hos-
pital program supported full disclosure involving a gradual
process [37,38]. This study highlighted the need to provide
hope in the face of a difficult diagnosis and both patients and
caregivers expressed the wish for detailed information and
referral to community supports. Though limited, the body of
literature relevant to this topic allows for some recommenda-
tions for clinical practice as a guide for health care profession-
als’ behavior during the diagnostic disclosure process.
1. Summary
(1) Diagnostic disclosure must be recognized as a pro-
gressive process that begins as soon as the possibility
of cognitive impairment is suspected and that in-
volves education and discussion throughout the in-
vestigative phase of the diagnosis. During the initial
investigations of complaints or suspicions of cogni-
tive loss, the patient’s understanding and attitudes
about cognitive loss and dementia, as well as those
of their family members/caregivers, should be estab-
lished. Diagnostic investigations, including the pas-
sage of time when necessary, provide opportunities
to discuss diagnostic uncertainty and to address spe-
cific fears, concerns, and preconceptions/misconcep-
tions that patients and their family/caregivers might
have regarding dementia. Health care professionals
should explore with patients and family/caregivers
what their beliefs are about the causes of dementia
and how they wish to be informed about the biomed-
ical model. They should be educated as to how the
process of making a diagnosis can help long-term
planning by both the patient and family/caregiver.
Those involved in the diagnostic disclosure should
establish who should be involved (eg, the patient
only, the patient and their spouse, the patient and
other family members). Although the autonomy of
the patient must be the primary consideration, it is
essential that those involved in the investigative pro-
cess encourage the involvement of relevant family
members and/or other current or potential future care
providers in the investigative process and the even-
tual diagnostic disclosure.
(2) Once a diagnosis is established, this must be dis-
closed to the patient and their family/caregivers in a
manner that is consistent with the expressed wishes
of the patient, with a patient-centered approach that
maintains the personal integrity of the individual. An
individualized approach to disclosure is recom-
mended to take into account variations in the needs
of the patient and family/caregivers that reflect the
nature and severity of the patient’s cognitive prob-
lems, their background knowledge and expectations,
and the cultural factors that can influence the under-
standing and acceptance of the diagnosis. The pres-
ence of comorbid medical and mental health issues
that might influence the diagnostic disclosure pro-
cess should also be considered. These factors might
require that additional time, follow-up appointments,
or referral to other specialists be used. Those pro-
viding the diagnosis must demonstrate empathy and
respect and instill a sense of hope. Although some
patients prefer a very direct approach, others require
a more gentle explanation of their condition. The
progressive disclosure process should also be used to
address remaining diagnostic uncertainty, treatment
options, future plans, financial planning, assigning
power of attorney, wills and “living wills”, driving
privileges, available support services, and potential
research participation if appropriate. If physicians
lack knowledge of these issues, they must be pre-
pared to discuss this openly with the patient and to
make appropriate referrals. Disclosure should take a
patient-centered approach that maintains the per-
sonal integrity of the individual and instills a sense
of hope. The immediate psychological impact of the
diagnosis should be assessed.
(3) The potential for adverse psychological conse-
quences must be assessed and should be addressed
by education and support of the patient and family/
caregivers throughout the diagnostic disclosure pro-
cess. Reactions to a diagnosis of dementia can in-
clude anger, denial, shock, and grief, and health
professionals need to be prepared to deal with these
difficult but often normal stages of emotional reac-
tions to a life-altering diagnosis. Potential negative
reactions may be avoided or moderated through ed-
ucational interventions early in the diagnostic pro-
cess. The benefits of knowing the diagnosis can be
highlighted early in the investigative process, and the
roles of specific health professionals in the diagnos-
tic process as well as their potential roles in later
management issues can be clarified.
(4) Follow-up plans for the patient and their family/
caregivers must be made and discussed at the time of
diagnostic disclosure. This follow-up should be tai-
lored to the individual needs of the patient and their
family/caregiver and should include education and
linkage to available community supports. Follow-up
appointments to discuss the diagnosis must be made
available to the patient and their family/caregivers,
including the possibility of separate appointments if
appropriate or requested. If patients have chosen to
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
not yet involve family/caregivers in the diagnostic
process, they should be strongly encouraged to do so
for follow-up appointments. Follow-up appoint-
ments might involve referral by the primary care
physician to other clinicians for specific services or
might involve referral back to the primary care phy-
sician for ongoing management in cases in which the
diagnosis is disclosed in a specialized clinic setting.
More specific information, such as written summa-
ries, that is valued by patients and family/caregivers
can be provided through follow-up sessions. Data are
lacking that reflect patients’ perspectives on the di-
agnostic disclosure process, the consequences of di-
agnostic disclosure, and in particular the long-term
consequences. Because the limited information
available points to discrepancies in physician and
caregiver opinions about the disclosure process, it is
incumbent on health care professionals to evaluate
this process within the context of their practice. Re-
gardless of who discloses the diagnosis, everyone
involved in the care of the patient has a responsibil-
ity to ensure that a plan for follow-up appointments
is established and carried out.
2. Recommendations Approved at the Third CCCDTD
Conference
(1) The process of diagnostic disclosure for persons with
cognitive impairment or dementia must begin as
soon as the possibility of cognitive impairment is
suspected (Grade A, Level 3).
(2) Both the diagnosis of dementia and the disclosure of
the diagnosis must be considered processes that pro-
vide opportunities for education and discussion
(Grade A, Level 3).
(3) The potential for adverse psychological conse-
quences must be assessed and addressed through
education of the patient and family/caregivers
(Grade B, Level 3).
(4) Once a diagnosis is established, this must be dis-
closed to the patient and their family/caregivers in a
manner that is consistent with the expressed wishes
of the patient (Grade B, Level 3).
(5) Follow-up plans must be made and discussed at the
time of diagnostic disclosure (Grade A, Level 3).
Author Disclosures
John Fisk has received support from AstraZenca (Consul-
tant), Bayer (Consultant), Biogen-Idec (Consultant), Bristol-
Myers Squibb (Consultant), Novartis (Consultant), Sanofi-
Aventis (Consultant) and TEVA Neuroscience (Speaker).
Lynn Beattie has received support from Novartis (Med-
ical Advisory Board Member, Speaker), Pfizer Canada
409
(Speaker, Clinical Trials), Janssen Ortho Inc. (Medical Ad-
visory Board, Speaker, Clinical Trials).
References
[1] Fisk JD, Sadovnik AD, Cohen CA, Gauthier S, Dossetor JB, Eberhart
A, et al. Ethical guidelines of the Alzheimer Society of Canada. Can
J Neurol Sci 1998;25:242– 8.
[2] Post SG, Whitehouse PJ. Fairhill guidelines on ethics of the care of
people with Alzheimer’s disease: a clinical summary. J Am Geriatr
Soc 1995;43:1423–9.
[3] Cohen CA, Whitehouse PJ, Post SG, Gauthier S, Eberhart A, LeDuc
L. Ethical issues in Alzheimer’s disease: the experience of a national
Alzheimer Society Task Force. Alzheimer Dis Assoc Disord 1999;
13:66 –70.
[4] Patterson CJS, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus Con-
ference on Dementia. CMAJ 1999;160(Suppl):S1–15.
[5] Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL,
DeKosky ST. Practice parameter: early detection of dementia—mild
cognitive impairment (an evidence-based review). Neurology 2001;
56:1133– 42.
[6] Maguire CP. Telling the diagnosis of dementia: consider each patient
individually. Int Psychogeriatrics 2002;14:123– 6.
[7] Marzanski M. Would you like to know what is wrong with you? on
telling the truth to patients with dementia. Journal of Medical Ethics
2000;26:108 –13.
[8] Carpenter B, Dave J. Disclosing a dementia diagnosis: a review of
opinion and practice, and a proposed research agenda. Gerontologist
2004;44:149 –58.
[9] Bamford C, Lamont S, Eccles M, Robinson L, May C, Bond J.
Disclosing a diagnosis of dementia: a systematic review. Int J Geri-
atric Psychiatry 2004;19:151– 69.
[10] Turner S, Iliffe S, Downs M, Wilcock J, Bryans M, Levin E, et al.
General practitioner’s knowledge, confidence and attitudes in then
diagnosis and management of dementia. Age Ageing 2004;33:461–7.
[11] Cody M, Beck C, Shue VM, Pope S. Reported practices of primary
care physicians in the diagnosis and management of dementia. Age-
ing Ment Health 2002;6:72– 6.
[12] Drickamer MA, Lachs MS. Should patients with Alzheimer’s disease
be told their diagnosis? N Engl J Med 1992;326:947–51.
[13] Lopez OL, Litvan A, Catt KE, Stowe R, Klunk W, Kaufer DI, et al.
Accuracy of four clinical diagnostic criteria for the diagnosis of
neurodegenerative dementias. Neurology 1999;53:1292–9.
[14] Petersen RC. Mild cognitive impairment as a diagnostic entity. J In-
tern Med 2004;256:183–94.
[15] Fisk JD, Rockwood K. Outcomes of incident mild cognitive impair-
ment in relation to case definition. J Neurol Neurosurg Psychiatry
2005;76:1175–7.
[16] Smith AP, Beattie BL. Disclosing a diagnosis of Alzheimer’s disease:
patient and family experiences. Can J Neurol Sci 2001;28(Suppl
1):S67–71.
[17] Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S, et
al. Frontotemporal lobar degeneration: a consensus on clinical diag-
nostic criteria. Neurology 1998;51:1546 –54.
[18] Mullen R, Howard R, David A, Levy R. Insight in Alzheimer’s
disease. Int J Geriatr Psychiat 1996;11:645–51.
[19] DeLepeleire J, Butinx F, Aertgeerts B. Disclosing the diagnosis of
dementia: the performance of Flemish general practitioners. Int Psy-
chogeriatrics 2004;16:421– 8.
[20] Maguire CP, Kirby M, Coen RE, Coakley D, Lawlor BA, O’Neill D.
Family members’ attitudes toward telling the patient with Alzhei-
mer’s disease their diagnosis. Br Med J 1996;313:529 –30.
410 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 404 – 410
[21] Erde EL, Nadal EC, Scholl TO. On truth telling and the diagnosis of
Alzheimer’s disease. J Fam Pract 1988;26:401– 6.
[22] Holroyd S, Snustad DG, Chalifoux ZL. Attitudes of older adults on
being told the diagnosis of Alzheimer’s disease. J Am Geriatr Soc
1996;44:400 –3.
[23] Holroyd S, Turnbull Q, Wolf AM. What are patients and their fam-
ilies told about the diagnosis of dementia? results of a family survey.
Int J Geriatr Psychiatry 2002;17:218 –21.
[24] Fahy M, Wald C, Walker Z, Livingston G. Secrets and lies: the
dilemma of disclosing the diagnosis to an adult with dementia. Age
Ageing 2003;32:439 – 41.
[25] Ferris SH, Hofeldt GT, Carbone G, Masciandaro P, Troetel WM,
Imbimbo BP. Suicide in two patients with a diagnosis of probable
Alzheimer disease. Alzheimer Dis Assoc Disord 1999;13:88 –90.
[26] Vega U, Kishikawa, Ricanatti E, Friedland RP. Suicide and Alzhei-
mer disease. Am J Geriatric Psychiatry 2002;10:484 –5.
[27] Draper B, Maccuspie-Moore C, Brodaty H. Suicidal ideation and the
‘wish to die’ in dementia patients: the role of depression. Age Ageing
1998;27:503–7.
[28] Turnbull Q, Wolf AMD, Holroyd S. Attitudes of elderly subjects
toward “truth telling” for the diagnosis of Alzheimer’s disease. J
Geriatr Psychiatry Neurol 2003;16:90 –3.
[29] Rubio A, Vestner AL, Stewart JM, Forbes NT, Conwell Y, Cox C.
Suicide and Alzheimer’s pathology in the elderly: a case-control
study. Biol Psychiatry 2001;49:137– 45.
[30] Pinner G, Bouman WP. To tell or not to tell: on disclosing the
diagnosis of dementia. Int Psychogeriatrics 2002;14:127–37.
[31] Pinner G, Bouman WP. What should we tell people about dementia?
Advances in Psychiatric Treatment 2003;9:335– 41.
[32] Henderson JN, Gutierrez-Mayka M. Ethnocultural themes in caregiv-
ing to Alzheimer’s disease patients in Hispanic families. Clin Ger-
ontol 1992;11:59 –74.
[33] Braun KL, Browne CV. Perceptions of dementia, caregiving and help
seeking among Asian and Pacific Islander Americans. Health Soc
Work 1998;l23:262–73.
[34] Hinton L, Franz CE, Teo G, Levkoff SE. Conceptions of dementia in
a multiethnic sample of family caregivers. J Am Geriatr Soc 2005;
53:1405–10.
[35] Connell CM, Boise L, Stuckey JC, Holmes SB, Hudson ML. Atti-
tudes toward the diagnosis and disclosure of dementia among family
caregivers and primary care physicians. Gerontologist 2004;44:
500 –7.
[36] Jha A, Tabet N, Orrell M. To tell or not to tell: comparison of older
patients’ reaction to their diagnosis of dementia and depression. Int J
Geriatric Psychiatry 2001;16:87– 885.
[37] Aminzadeh F, Byszewski AM, Molnar FJ, Eisner M. Emotional
impact of dementia diagnosis: exploring persons with dementia and
caregivers’ perspectives. Aging Ment Health 2007;11:281–90.
[38] Byszewski A, Molnar F, Aminzadeh F, Eisner M, Gardezi F, Bassett
R. Dementia diagnosis disclosure: a study of patient and caregiver
perspectives. Alzheimer Dis Assoc Disord 2007;21:107–14.
 Alzheimer’s & Dementia 3 (2007) 411– 417

Ethical considerations for decision making for treatment

and research participation
John D. Fiska,*, B. Lynn Beattieb, Martha Donnellyc
a
QE II Health Sciences Centre; Department of Psychiatry, Department of Medicine, and Department of Psychology, Dalhousie University, Halifax, Nova
Scotia, Canada
b
Department of Medicine, Division of Geriatric Medicine, University of British Columbia, and the BC Network for Aging Research, Vancouver, British
Columbia, Canada
c
Division of Community Geriatrics, Department of Family Practice, and Division of Geriatric Psychiatry, Department of Psychiatry, University of
British Columbia, and Geriatric Psychiatry Outreach Team, Vancouver Hospital, Vancouver, British Columbia, Canada

Abstract Here we review issues of patient decision-making and consent to treatment and research by
persons with cognitive impairment and dementia. Clinicians and researchers must recognize their
primary duty to care for the individual and must clearly distinguish their role as a clinician and/or
researcher. Distinctions between standard care and research must be clearly understood by everyone,
as must the clinician’s role in each. Both actual and perceived conflicts of interest must be avoided.
At present there is insufficient evidence to recommend specific methods for determining compe-
tency for decision-making, but a diagnosis of cognitive impairment or dementia does not preclude
such competence. Competency is not a unitary or static construct and must be considered as the
ability to make an informed decision about participation in the particular context of the specific
treatment or study. Clinicians and researchers should consider consent as a process involving both
the patient with cognitive impairment and his or her family/caregiver, particularly given the
potential that competency for decision-making will change over time. As the availability of advance
directives remains limited, clinicians and researchers must make efforts to ensure that decisions
made by proxies are based on the prior attitudes and values of the patient.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Alzheimer’s disease; Dementia; Decision-making; Competency; Ethics

For the current revision of the Canadian Consensus Con-
ference on Dementia (CCCD), the emergence of improved
methods for early diagnosis of dementia as well as the
availability of approved symptomatic treatments for Alzhei-
mer’s disease meant that the issues of patient decision
making for treatment and research required discussion. The
ethical considerations surrounding these issues had not been
included in previous CCCD guidelines [1]. Not only did
these issues seem of greater current priority, but also the
ethical issues of decision making seemed to represent a
logical extension of the discussions and recommendations
surrounding the issue of diagnostic disclosure. Both have
the issue of patient autonomy at their center. For this re-
*Corresponding author. Tel.: 902-473-2581; Fax: 902-473-7166.
E-mail address: John.Fisk@cdha.nshealth.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.08.001
view, the PubMed and Embase databases were searched for
articles with the keywords “Dementia OR Alzheimer’s dis-
ease AND ethics AND competency.” For discussion, pref-
erence was given to publications between 1996 and 2006.
We also examined Canadian and international guidelines on
the ethics of research, with particular attention to issues
relevant to the participation of individuals with cognitive
impairment and dementia in research.
We review and discuss issues on consent to treatment in
clinical practice and consent to research participation by
persons with cognitive impairment and dementia. We also
review studies that have examined patient decision making
for treatment and the participation of persons with dementia
in research. Although attempts have been made to draw
distinctions between the ethics of clinical care and the ethics
of the conduct of research [2], this view is strongly opposed
412 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417
by many [3] and is inconsistent with most national and
international ethical guidelines for research [4,5]. Thus
many of the same principles apply, regardless of whether
treatment or research participation is the issue of discussion.
For this reason, we have chosen to discuss these issues
together in a single document and have combined the dis-
cussions and recommendations regarding treatment decision
making and research participation decision making.
Although there might not be clear distinctions between
the ethical issues that clinicians and scientists must con-
sider, there are nonetheless clear distinctions between the
roles of clinician and scientist. Specifically, appointments,
investigations, and data collection that represent manag-
ment of an individual’s health care must be clearly distin-
guished from similar activities that represent the conduct of
a clinical trial for the research participant. Clinicians need to
be keenly aware of the ethical issues, such as conflicts of
interest, that they are likely to face when they conduct
clinical trials. They must also be aware of the values and
beliefs on which their own decision making is based. How-
ever, although research involving patients with cognitive
impairment and dementia presents its own particular chal-
lenges, the ethical considerations themselves are not unique
to dementia.
In 1997 the Alzheimer Society of Canada’s ethical
guidelines document [6,7] addressed both autonomy in de-
cision making and participation in research. Even though
the lack of available symptomatic treatment in Canada
at that time limits their current applicability to some extent,
these guidelines illustrate an important point. Specifically,
they illustrate the differences that can emerge between gen-
eral international guidelines on the ethical conduct of re-
search and guidelines that are developed by individuals who
are affected by specific medical conditions and their repre-
sentatives. Individuals affected by specific medical condi-
tions and their advocates, while continuing to point out
the importance of protecting the rights of vulnerable indi-
viduals, generally place greater emphasis on ensuring that
the potential to benefit from participation in research is not
denied to specific groups of individuals. This inclusive
stance toward research participation by persons with de-
mentia has been articulated by High et al [8]: “To deny
persons access to research participation out of fear of ex-
ploitation of specific groups of persons is to avoid rather
than accept and practice ethical responsibility.”
The principles that best describe ethical decision making
regarding research participation are those articulated in the
Belmont Report, produced by the US National Commission
for the Protection of Human Subjects of Biomedical and
Behavioural Research [9]. These include the principles of
respect for persons, beneficence (ie, the obligation to do no
harm and to maximize the potential for benefit while min-
imizing the potential for harm), and justice. These princi-
ples, in turn, are applied to clinical research through con-
sideration of informed consent, assessment of risks and
benefits, and the selection of subjects. In Canada, research
on human subjects conducted at institutions that receive
support for federal funding bodies must be conducted in
accordance with guidelines established jointly by the Cana-
dian Institutes for Health Research, the Social Sciences and
Humanities Research Council, and the National Science and
Engineering Research Council [10] that are largely based on
these same ethical principles. The same principles are
present in most international regulatory research guidelines
such as the World Medical Association’s Declaration of
Helsinki [11], the Ethical Guidelines for Biomedical Re-
search of the Council for International Organisations of
Medical Sciences (CIOMS) [5], and the Council of Eu-
rope’s Convention on Human Rights and Biomedicine [12].
Internationally accepted guidelines for the conduct of clin-
ical trials are published in the International Conference on
Harmonisation of Technical Requirements for Registration
of Pharmaceuticals for Human Use Harmonised Tripartite
Guideline (ICH) [13]. Unlike the ethical guidelines for
research, these guidelines, often referred to as the “good
clinical practice” (ICH-GCP) guidelines, were produced as
a joint regulatory and industry project “to facilitate the
adoption of new or improved technical research and devel-
opment approaches which update or replace current prac-
tices, where these permit a more economical use of human,
animal and material resources, without compromising
safety” [13]. Given their origins and mandate, it is perhaps
not surprising that despite recognizing the need to address
issues specific to the conduct of clinical trials in elderly
populations and reference to research on Alzheimer’s dis-
ease [14], ICH-GCP guidelines provide no insight into the
specific ethical issues that arise in those contexts. As with
all guidelines, the challenge becomes how to find the right
balance when attempting to apply the general principles to
the specific ethical dilemmas that arise.
Perhaps the most obvious ethical dilemma associated
with clinical trials for cognitive impairment and dementia is
balancing respect for the autonomy of the individual with
the protection of vulnerable persons. Most ethical guide-
lines regarding research focus on the individual in terms of
the analysis of risk/benefit ratios and the consent process.
Guidelines developed for persons with Alzheimer’s disease,
however, also acknowledge the need to consider the conse-
quences of research participation on families and the im-
portance of involving the caregiver/family in the consent
process [8,15]. Even for individuals with Alzheimer’s dis-
ease who are competent to provide informed consent, con-
sent might be required of their caregiver/family. Typically it
is they who will be expected to bring the person to sched-
uled appointments; observe the subject for adverse events,
including those that should be reported immediately to the
research team, as well as those of less urgent nature; ensure
that the subject takes medications as required; observe and
report possible treatment benefits; relay to the investigators
the subject’s continued assent to participation; and take part
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417
in evaluating treatment efficacy, such as reporting changes
in emotional burden or time spent on tasks for care.
Most guidelines promote the view that “the principle of
respect for human dignity entails high ethical obligations to
vulnerable populations” [10]. However, most also agree
with the concept that the “protection should be proportion-
ate to the risk involved, with the least protection required
when research involves minimal risk” [16]. This consider-
ation of the limits of the acceptable risk/benefit ratio is
central to the evaluation of all clinical trials [17]. However,
it becomes particularly difficult when evaluating clinical
trials for patients with cognitive impairment and dementia
because of the potential for uncertainty about the subject’s
capacity for autonomous and informed decision making. A
decade ago, High et al [8] noted “no clear consensus exists
either in the literature or in regulatory guidelines as to what
constitutes an acceptable degree of risk when cognitively
impaired persons are involved in research.” This remains
the case today.
The concept of minimal risk has been used in most
international ethical guidelines for research, although it has
been described in various ways [17]. Usually, this is meant
to ensure that “the probability and magnitude of harm or
discomfort anticipated in the research are not greater in and
of themselves than those ordinarily encountered in daily life
or during the performance of routine physical or psycho-
logical examinations or tests” [18]. Such is rarely the case in
clinical trials of new treatments for dementia, however, and
although it might be argued that research that does not
involve “risk of harm beyond a minor increment over min-
imal” [19] is acceptable when dealing with cognitively
impaired individuals, such terms are open to the criticism of
being poorly defined and allowing wide interpretation. Also
contributing to difficulty in establishing consensus on the
issue of risk is the use of varied ethical guidelines for ethics
review. In Canada, research on human subjects conducted at
institutions that receive support for federal funding bodies
must be conducted in accordance with the Tricouncil Policy
Statement [10], but research conducted outside such insti-
tutions is governed by a patchwork of laws and voluntary
industry guidelines such as the ICH-GCP guidelines [13].
Both international regulatory bodies and organizations with
the mandate of patient advocacy typically defer decisions of
ethical acceptability of clinical trials to individual ethical
review boards. The difficulties in coming to terms with
the issue of risk, even for an organization such as the
Alzheimer’s Association in the USA, are evident in their
statement:
“There is considerable disagreement in the ethics liter-
ature and in regulatory and statutory language about
definitions of risk. In addition to the two polar defini-
tions, ‘minimal risk’ and ‘greater than minimal risk,’
there are also various gradations, such as ‘minor increase
over minimal risk.’ The Alzheimer’s Association has
chosen not to draw these finer distinctions in its position
413
statement. However, this should not be construed to
mean that the Association necessarily opposes involve-
ment of decidedly impaired individuals in all nonthera-
peutic research involving minor increments over mini-
mal risk. The degree of risk is a judgment call frequently
and appropriately made by IRBs [institutional review
boards]” [20].
Although it has been argued that research ethics review
boards spend a disproportionate amount of time addressing
the issue of informed consent at the expense of addressing
the analysis of risk [17], free and fully informed consent is
both central to the ethical conduct of research and particu-
larly problematic for research in dementia. Most ethical
guidelines take a relatively protectionist stance with regard
to avoiding the potential for exploitation of individuals who
are not competent to provide informed consent, but vari-
ability does exist among various guidelines [21]. A diagno-
sis of dementia or other forms of cognitive impairment does
not preclude competence to provide informed consent for
participation in clinical trials [22] or in nontherapeutic re-
search, which contains minimal risk. Improved diagnostic
methods also mean that Alzheimer’s disease and other de-
mentias are detected at earlier stages in which many cogni-
tive and functional abilities are relatively preserved. The
conceptualization of mild cognitive impairment (MCI) [23]
as a diagnostic entity that warrants clinical therapeutic and
prevention trials reinforces this trend.
Competence to consent to either treatment or research
participation is not a simple matter of the stage of dementia
or severity of cognitive impairment for an individual. In
part, this is because competence is not a unitary construct.
One can be competent in some aspects of one’s life (eg,
competent to consent to research) without being competent
in others (eg, competent to drive a motor vehicle) [24]. As
the risk/benefit ratio and other aspects of a given treatment
or study vary, so do the requirements for competence. This
concept of a “sliding-scale” of competency [25] is most
typically considered in the context of research participation,
as acknowledged in the Tri-Council Policy Statement:
“Competence to participate in research, then, is not an
all-or-nothing decision. It requires that they be compe-
tent to make an informed decision about participation
in particular research. Competence is neither a global
condition nor a static one; it may be temporary or
permanent” [10].
Although much of the discussion regarding the ethical
issues of decision making has been focused on the topic of
research participation, most of the research on the topic
of competency has focused on consent to treatment. The use
of standardized methods to determine competency to pro-
vide informed consent has been promoted by some, but
there remains no clear consensus on the best standardized
assessment methods to address this [24,26]. Moreover,
given the evidence of inconsistencies in expert clinical judg-
ments on the issue of competence for mildly affected pa-
414 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417
tients [27], the question can be raised as to what can be
considered a gold standard for evaluating such methods. For
the primary care physician with limited time and access to
formal competency assessment methodologies, it would be
advantageous if standard mental status screening instru-
ments could be used to determine consent capacity as well
as for diagnostic screening purposes. There is limited evi-
dence to suggest that screening instruments can provide
some guidance as to the probability of competency to con-
sent to treatment [28,29] and to research [30], but the
complexity of the issues precludes any simple means of
establishing competency in a specific treatment or research
setting with certainty [31]. Clearly, more impaired individ-
uals are less likely to be competent to make informed
decisions, but beyond this, the issues become much more
complex.
Research on competency for the provision of informed
consent in a research context has been less common than for
consent to treatment. Although Kim et al [32] have reported
that even mild Alzheimer’s disease has significant effects on
competency to consent to research, this issue remains an
important topic for further research. Even if standardized
assessments of competency to participate in clinical or re-
search decision making were to become available, signifi-
cant practical, social, cultural, and legislative variations are
still likely to limit their application. The absence of clear
and consistent legislation regarding competency for deci-
sion making and the important role of proxy decision mak-
ing, particularly in the context of research participation,
have been recognized for some time [25,33]. Across Can-
ada, legislation regarding substitute decision making for the
participation of incompetent individuals in research is often
unclear and varies among provinces and territories [34].
For both practical and ethical reasons in most situations,
clinicians, researchers, and research ethics boards have typ-
ically relied on a family member of the person with demen-
tia to provide “third party authorization” for decision mak-
ing, even when there has been no legal determination that
the patient lacks competence to provide informed consent.
In some cases, proxy consent is aided by the availability of
advance directives for research participation that specifi-
cally identify proxy decision makers. But despite the pro-
motion of the use of advance directives by some [19],
concerns about this have been raised because of their lim-
ited availability [35]. Even when consent from a legally
authorized third party is given, the ongoing assent of the
individual subject is almost invariably necessary.
Despite the common use of proxies in the decision-
making process, reservations about proxy decision making
for participation by cognitively impaired subjects in re-
search have been expressed for some time [36]. More
recently, the factors that influence caregivers’ decisions
regarding research participation have themselves become
the source of study. Many of the considerations that have
arisen are an extension of those raised previously in
the context of diagnostic disclosure. First and foremost,
informed proxy decision making by the family/caregiver
requires that they be informed about dementia. Unfortu-
nately, the limited numbers of studies to date that have
examined caregiver knowledge about Alzheimer’s disease
have found relatively poor understanding of issues such as
causes, symptoms, and drug treatments [37]. This highlights
the need for education of patients and their family/caregivers in
the diagnostic disclosure process before the possibility of treat-
ment and/or research participation is raised. Clinicians must
also recognize that, as with other demands of providing care
to persons with dementia, the process of providing proxy
consent can itself be burdensome, particularly when the
risks of clinical trial participation are not fully appreciated
[38]. Factors affecting proxy decision making were exam-
ined by Karlawish et al [39] in a study of caregivers’ views
about the acceptance of risk associated with potential treat-
ments for Alzheimer’s disease. They noted that caregivers’
tolerance for risk of death associated with a hypothetical
treatment was influenced by demographic factors and that
working adult children who were caring for early-stage
patients were more willing to accept that risk than spouses
of such patients. For the clinician, these views point out the
need to recognize that although proxies are expected to
represent the patient, they are not without their own biases
when it comes to decision making. It also points out the
importance of ensuring that decisions made by proxies re-
garding treatment and research participation reflect the prior
attitudes and values of the patient rather than those of the
proxy.
A study by Connell et al [40] of caregiver attitudes
toward participation in longitudinal, nontherapeutic re-
search is also important in that it points out the possible
presence of expectations of improved access to clinical care
through research participation. Although the direct implica-
tions of this American study to a Canadian context can be
questioned, it also pointed out the existence of ethnic/racial
differences in attitudes toward research, presenting barriers
to equitable research participation that are likely to gener-
alize beyond the USA. As was discussed in the context of
diagnostic disclosure, we have very limited information
regarding the potential influences of ethnic and cultural
issues with regard to treatment and research participation
decision making among patients with dementia and their
family/caregivers. This is an important consideration with
regard to the ethical principle of justice and the equal
distribution of the risks and benefits. The possibility of
systematic differences in expected benefits of research par-
ticipation reinforces the need to draw a clear distinction
between research participation and clinical care for both the
patient and their family/caregiver. Although improved mon-
itoring and access to health care professionals might be the
case in some clinical trials, the assumption that receiving
treatment through participation in a clinical trial results in
improved outcome over similar usual care has not been
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417 415

supported [41]. It is also the case that participation in studies ing for persons with dementia has its own set of challenges that
might reduce rather than improve access to care, for exam- require careful consideration. The existence of social and
ple, when subjects are enrolled into the placebo arm of a cultural barriers to the implementation of existing ethical
clinical trial of a new treatment rather than being started on guidelines for clinical practice must be acknowledged, and
an established therapy. the extent to which similar barriers exist in the implemen-
Research directly related to proxy decisions about enroll- tation of guidelines for the ethical conduct of research is
ment in clinical trials has been relatively rare. Those studies itself an important topic that warrants research [43].
that have been conducted suggest that although proxies
actively involve the research subjects in the decision- 1. Summary
making process [38,42], they do not always think that
their decisions are consistent with those of the research
(1) At present there is insufficient evidence to recom-
subject [15]. For the clinician and researchers, it is impor-
mend a specific standardized method for determining
tant to recognize that the interests of the person with de-
the competency of persons with dementia for deci-
mentia and those of a substitute decision maker might differ.
sion making for treatment or research. A diagnosis of
Thus, clinicians have an obligation to determine, to the best
dementia, or other forms of cognitive impairment,
of their ability, that the decisions regarding treatment and
does not preclude competence to provide informed
research participation are guided by the individual patient’s
consent for treatment or for participation in research,
wishes and, if this is not clearly known, that it has been
and competency must be considered as the ability to
made with the individual patient’s best interests in mind
make an informed decision about participation in the
[8,10,15].
particular context of the specific treatment or study.
As new potential treatments for dementia become avail-
(2) Even in the absence of a legal determination of
able for study and use, decision making on the part of
competency, clinicians and researchers should con-
patients and their family/caregivers will become increas-
sider obtaining consent as a process involving both
ingly complex and will require even greater educational
the patient with dementia and their family/caregiver.
efforts and systematic evaluation. Because changes in com-
(3) The distinctions between the clinician’s role in the
petency are to be expected over time in most of the demen-
management of the individual’s health care and his/
tias, there is also clear need for ongoing monitoring and
her potential role in the conduct of research must be
reaffirmation of consent/assent. Indeed, this could easily be
clearly understood by everyone, as must the proce-
considered an important requirement even when compe-
dures that represent standard care and research. Con-
tency of the subject can be assured. Thus, for all research, it
flicts of interest, both actual and perceived, must be
is reasonable to expect that the procedures to evaluate the
avoided. If conflicts arise in a research context, the
ability of the potential subject to understand the nature of the
opinion of a physician other than the research phy-
research, the consequences of participation (ie, potential risks
sician might be necessary to ensure that decisions
and benefits), and alternative choices are described [7,15].
regarding treatment and continued research partici-
In summary, it is important that the clinicians and re-
pation are made in the best interests of the patient.
searchers are aware of their primary duty to care for the
(4) The potential that competency for decision making
individual and to clearly distinguish their roles as a clinician
will change over time must be recognized. This
and/or researcher. They must also be aware of the values
might lead to a change from one of obtaining the
and beliefs that influence their own ethical decision making.
patient’s ongoing consent to one of obtaining ongo-
The ethical principles of autonomy, beneficence, and justice
ing assent. Assent is usually required.
represent a balance. In research, it is the principles of
(5) To the best of their ability, clinicians and researchers
autonomy (respect for persons) and beneficence (do no
must ensure that the decisions made by proxies re-
harm/do good) that might be most obviously in conflict, but
garding treatment and research are based on the prior
this is not unique to research, and this conflict will likely
attitudes and values of the patient. This might be
become even more apparent as new treatments for dementia
assisted by the availability of advance directives.
become available and/or as the risks of current treatments
become better understood. Finding balance requires the ethical
analysis of risk and the process of informed consent. However, 2. Recommendations approved at the Third
considerations of informed consent in dementia invariably Canadian Consensus Conference on Diagnosis and
come up against the issue of competency. It is important to Treatment of Dementia
recognize that competency is not a unitary or static construct,
and the variability of existing legislation and current ethical (1) Provision of the best standard of care for the patient
guidelines on this issue might be of limited help to the indi- must always remain the priority (Grade A, Level 3).
vidual clinician or researcher dealing with the individual pa- (2) Drawing a clear distinction between research partic-
tient and a specific set of circumstances. Proxy decision mak- ipation and clinical care is essential for both the
416 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417
patient and their family/caregiver. The distinctions
between the clinician’s role in the management of
the individual’s health care and his/her potential role
in the conduct of research must be clearly under-
stood by everyone, as must the procedures that rep-
resent standard care and research. In research set-
tings, the availability of a physician other than the
research physician to provide general care is recom-
mended to ensure that decisions regarding treatment
are made in the best interests of the patient (Grade A,
Level 3).
(3) A diagnosis of dementia, or other forms of cognitive
impairment, does not preclude competence to pro-
vide informed consent, whether it is for treatment
decisions, for participation in clinical trials, or for
participation in nontherapeutic research. Compe-
tency must be considered as the ability to make an
informed decision about participation in the partic-
ular context of the specific treatment or research
study (Grade A, Level 3).
(4) For studies, it is reasonable to expect that the pro-
cedures that will be used to evaluate the ability of the
potential subject to understand the nature of the
research, the consequences of participation (ie, po-
tential risks and benefits), and alternative choices are
described. However, at present there is insufficient
evidence available to recommend the use of a spe-
cific standardized method for determining compe-
tency for decision making either for treatment or
research (Grade B, Level 3).
(5) Even in the absence of a legal determination of the
competency of the patient with cognitive impairment
or dementia, it is important that the clinician and
researcher consider the consent process as one that
should involve both the patient and their family/
caregiver for treatment and research decision mak-
ing. In research settings, research ethics boards
might explicitly require that consent/assent be ob-
tained from both parties (Grade B, Level 3).
(6) The potential that competency for treatment and re-
search decision making will change over time must
be recognized. This might lead to a change from one
of obtaining the patient’s ongoing consent to one of
obtaining ongoing assent. Assent is almost invari-
ably required, and the decision to discontinue treat-
ment, whether it is therapy or research, must always
be an option (Grade A, Level 3).
(7) To the best of their ability, clinicians and researchers
must ensure that the decisions made by proxies re-
garding treatment and research are based on the prior
attitudes and values of the patient. Proxies have a
responsibility to represent the patient, and all parties
must recognize the challenges of doing so (Grade A,
Level 3).
Author Disclosures
John Fisk has received support from AstraZenca (Con-
sultant), Bayer (Consultant), Biogen-Idec (Consultant),
Bristol-Myers Squibb (Consultant), Novartis (Consultant),
Sanofi-Aventis (Consultant) and TEVA Neuroscience
(Speaker).
Lynn Beattie has received support from Novartis (Med-
ical Advisory Board Member, Speaker), Pfizer Canada
(Speaker, Clinical Trials), Janssen Ortho Inc. (Medical Ad-
visory Board, Speaker, Clinical Trials).
References
[1] Patterson CJS, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus Con-
ference on Dementia. CMAJ 1999;160(Suppl):S1–15.
[2] Miller FG, Brody H. What makes placebo-controlled trials unethical?
Am J Bioeth 2002;2:3–9.
[3] Lemmens T, Miller PB. Avoiding a Jekyll-and-Hyde approach to the
ethics of clinical research and practice. Am J Bioeth 2002;2:14 –7.
[4] American Medical Association. Code of medical ethics. Available at:
www.ama-assn.org/ama/pub/category/8292.html. Accessed: Septem-
ber 6, 2007.
[5] Council for International Organizations of Medical Sciences. Interna-
tional ethical guidelines for biomedical research involving human sub-
jects, 2002. Available at: www.cioms.ch/frame_guidelines_nov_2002.
htm. Accessed: September 6, 2007.
[6] Alzheimer Society of Canada. Ethical guidelines, 2003. Available at:
www.alzheimer.ca/english/care/ethics-intro.htm.
[7] Fisk JD, Sadovnik AD, Cohen CA, Gauthier S, Dossetor JB, Eberhart
A, et al. Ethical guidelines of the Alzheimer Society of Canada. Can
J Neurol Sci 1998;25:242– 8.
[8] High DM, Whitehouse PJ, Post SJ, Berg L. Guidelines for addressing
ethical and legal issues in Alzheimer disease research: a position
paper. Alzheimer Dis Assoc Disord 1994;8:66 –74.
[9] National Commission for the Protection of Human Subjects of Bio-
medical and Behavioural Research, the Belmont Report, ethical prin-
ciples and guidelines for the protection of human subjects of research,
1979. Available at: http://www.hhs.gov/ohrp/humansubjects/guidance/
belmont.htm. Accessed: September 6, 2007.
[10] Tri-Council policy statement: ethical conduct for research involving
humans, 1998 (with 2000, 2002 updates).Available at: http://pre.
ethics.gc.ca/english/policystatement/policystatement.cfm.
[11] World Medical Association Declaration of Helsinki, ethical principles
for medical research involving human subjects. Available at: www.
wma.net/e/policy/b3.htm. Accessed: September 6, 2007.
[12] Council of Europe, Convention for the Protection of Human Rights
and Dignity of the Human Being with Regard to the Application of
Biology and Medicine: Convention on Human Rights and Biomedi-
cine, Oviedo, 4.IV. 1997. Available at: http://conventions.coe.int/
treaty/en/Treaties/Html/164.htm. Accessed: September 6, 2007.
[13] International Conferences on Harmonization of Technical Require-
ments for Registration of Pharmaceuticals for Human Use (ICH).
Revised ICH Terms of Reference. Available at: http://www.ich.org/
cache/compo/276-254-1.html. Accessed: September 6, 2007.
[14] International Conference on Harmonization of Technical Require-
ments for Registration of Pharmaceuticals for Human Use, 1993. ICH
Harmonised Tripartite Guideline, Studies in Support of Special Pop-
ulations: Geriatrics E7. Available at: http://www.ich.org/cache/
compo/276-254-1.html. Accessed: September 6, 2007.
 J.D. Fisk et al. / Alzheimer’s & Dementia 3 (2007) 411– 417
[15] Karlawish JHT. Research involving cognitively impaired adults.
N Engl J Med 2003;348:1389 –92.
[16] Office of Human Subjects Research, information sheet 7, National In-
stitutes of Health. Available at: http://ohsr.od.nih.gov/info/sheet3.html.
[17] Weijer C. The ethical analysis of risk. J Law Med Ethics 2000;28:
344 – 61.
[18] Department of Health and Human Services. Code of federal regula-
tions title 45 part 46: regulations for the protection of human subjects,
section 46.102.i, 2001. Available at: http://www.hhs.gov/ohrp/
humansubjects/guidance/45cfr46.htm. Accessed: September 6, 2007.
[19] Keyserlingk EW, Glass K, Kogan S, Gauthier S. Proposed guidelines
for the participation of persons with dementia as research subjects.
Perspect Biol Med 1995;38:319 – 62.
[20] Alzheimer’s Association. Ethical Issues in Dementia Research (with
special emphasis on “informed consent”). Available at: http://www.
alz.org/national/documents/statements_ethicalissues.pdf. Accessed:
September 6, 2007.
[21] Wendler D, Prasad K. Core safe guards for clinical research with
adults who are unable to consent. Ann J Intern Med 2001;135:
514 –23.
[22] Karlawish JHT, Casarett DJ, James BD. Alzheimer’s disease patients’
and caregivers’ capacity, competency, and reasons to enroll in an
early-phase Alzheimer’s disease clinical trial. J Am Geriatr Soc
2002;50:2019 –24.
[23] Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL,
DeKosky ST, et al. Practice parameter: early detection of dementia:
mild cognitive impairment (an evidence-based review). Neurology
2001;56:1133– 42.
[24] Marson DC. Loss of competency in Alzheimer’s disease: conceptual
and psychometric approaches. Int J Law Ethics 2001;24:267– 83.
[25] High DM. Research with Alzheimer’s disease subjects: informed
consent and proxy decision-making. J Am Geriatr Soc 1992;40:
950 –7.
[26] Moye J, Karel M, Azar AR, Gurrera RJ. Capacity to consent to
treatment: empirical comparison of the instruments in older adults
with and without dementia. Gerontologist 2004;44:166 –75.
[27] Marson DC, McInturff B, Hawkins L, Bartolucci A, Harrell LE.
Consistency of physician judgements of capacity to consent in Alz-
heimer’s disease. J Am Geriatr Soc 1997;45:453–7.
[28] Etchells E, Darzins P, Silberferfeld M, Singer PA, McKenny J, Naglie
G, et al. Assessment of patient capacity to consent to treatment. J Gen
Intern Med 1999;14:27–34.
[29] Karlawish JHT, Casarett DJ, James BD, Xie SX, Kim SYH. The
ability of persons with Alzheimer’s disease (AD) to make a decision
about taking an AD treatment. Neurology 2005;64:1514 –9.
417
[30] Kim SYH, Caine ED. Utility and limits of the Mini Mental State
Examination in evaluating consent capacity in Alzheimer’s disease.
Psychiatric Services 2002;3:1322– 4.
[31] McQuillan MP, Tariot P. Who cand say yes (or no) to a physician: and
how does the physician know they can? Neurology 2005;64:1494 –5.
[32] Kim SY, Caine ED, Currier GW, Leibovici A, Ryan JM. Assessing
the competence of persons with Alzheimer’s disease in providing
informed consent for participation in research. Am J Psychiatry 2001;
158:712–7.
[33] The Ethics and Humanities Subcommittees of the American Acad-
emy of Neurology. Ethical issues in clinical research in neurology.
advancing knowledge and protecting human research subjects. Neu-
rology 1998;50:592–5.
[34] Alzheimer Society of Canada. Tough issues: ethical guidelines of
the Alzheimer Society of Canada. Toronto: Alzheimer Society of
Canada; 1997.
[35] Sachs GA, Stocking CB, Stern R, et al. Ethical aspects of dementia
research: informed consent and proxy consent. Clin Res 1994;42:
403–12.
[36] Candilis PJ, Wesley RW, Wichman A. A survey of researchers using
a consent policy for cognitively impaired human research subjects.
IRB 1993;15:1– 4.
[37] Werner P. Correlates of family caregivers knowledge about Alzhei-
mer’s disease. Int J Geriatr Psychiatry 2001;16:32– 8.
[38] Sugarman J, Cain C, Wallace R, Welsh-Bohmer KA. How proxies
make decisions about research for patients with Alzheimer’s disease.
J Am Geriatr Soc 2001;49:1110 –9.
[39] Karlawish JHT, Klocinski JL, Merz J, Clark CM, Asch DA. Care-
givers’ preferences for the treatment of patients with Alzheimer’s
disease. Neurology 2000;55:1008 –14.
[40] Connell CM, Shaw BA, Holmes SB, Foster N. Caregiver’s attitudes
toward their family members’ participation in Alzheimer disease
research: implications for recruitment and retention. Alzheimer Dis
Assoc Disord 2001;15:137– 45.
[41] Vist GE, Hagen KB, Devereaux PJ, Bryant D, Kristoffersen DT,
Oxman AD. Systematic review to determine whether participation in
a trial influences outcome BMJ 2005;330:1175– 81.
[42] Karlawish JHT, Casarett D, Klocinski J, Sankar P. How do AD
patients and their caregivers decide whether to enroll in a clinical
trial? Neurology 2001;56:789 –92.
[43] Graham C, Holmes C, Lindesay J. Clinical involvement in anti-
dementia drug trials: why bother? Int J Geriatr Psychiatry 1999;14:
257– 60.
 Alzheimer’s & Dementia 3 (2007) 418 – 427

Genetics and dementia: Risk factors, diagnosis, and management

Ging-Yuek Robin Hsiunga,b,c,*, A. Dessa Sadovnickb,d
a
Division of Neurology, University of British Columbia, Vancouver, BC, Canada
b
Brain Research Centre, University of British Columbia, Vancouver, BC, Canada
c
St. Paul’s Hospital, Providence Health Care Center, Vancouver, British Columbia, Canada
d
Department of Medical Genetics, University of British Columbia, Vancouver, BC, Canada

Abstract New developments in molecular genetics have improved our understanding on a number of
neurodegenerative dementias considerably, especially Alzheimer’s disease and frontotemporal de-
mentia. However, this explosion of information can be overwhelming to clinicians, making it
difficult to integrate into regular clinical practice. In this article, we briefly reviewed our current
understanding regarding causative genetic mutations and genetic risk factors on the major forms of
dementia, which provided the background information for discussion in the third Canadian Con-
sensus Conference on Diagnosis and Treatment of Dementia. The principles of genetic counselling
were applied. Guidelines and recommendations on the application of genetics in the assessment,
diagnosis, and management of patients and families with dementia were summarized.
© 2007 The Alzheimer’s Association. All rights reserved.

Keywords: Genetic counselling; Alzheimer’s disease; Frontotemporal dementia; APOE; PSEN1; PSEN2; MAPT; PGRN

1. Background
Advances in genetics, especially with completion of the
human genome project, have improved our understanding
of the pathogenesis of many diseases, including Alzhei-
mer’s disease (AD) and other types of dementia [1– 4]. With
the increase in public awareness, dementia patients and
family members as well as the general public now often
request genetic testing as part of their assessment of risk of
dementia, even in the absence of treatments known to slow
or prevent disease progression [5–7]. It is thus expected that
these requests will increase over time when such treatments
become realities. The clinician must understand the role of
genetic risk factors to provide accurate information to their
patients and to refer them to genetics clinics when appro-
priate. Canadian sites can be found at http://ccmg.medical.org/
clinical.html (accessed March 3, 2006). Here we review the
current literature on the genetic risk factors for dementia,
summarize the current knowledge, and provide evidence-based
recommendations to clinicians.
*Corresponding author. Tel.: (604) 682-2344 x63459; Fax: (604) 822-7191.
E-mail address: hsiung@interchange.ubc.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.010
2. Methods
We initially retrieved 1721 articles by a PubMed search
on genetic risk factors for AD and other dementias from
January 1996 up to and including December 2005. Relevant
articles before 1996 have been summarized elsewhere in
meta-analyses and reviews [8,9]. We further conducted a
review of all the genes listed in the ALZGENE website
(accessed February 14, 2006) [10] and articles from bibli-
ographies of selected studies. Authors’ files were also
scanned. For genetic association studies, we only consid-
ered those with large sample sizes (⬎300 cases and ⬎300
controls), with well-defined clinical diagnostic criteria and
with findings consistently replicated in four or more inde-
pendent samples. Because we came to the conclusion that
APOE is the only genetic risk factor that has been consis-
tently replicated to date, we further examined studies on
APOE and its interaction with other risk factors in relation
to AD. We also looked at single gene mutations studies,
because within specific families they are known to be causal
and thus represent risk factors for unaffected biologic rela-
tives of demented individuals carrying the specific muta-
tion.
 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427 419

Table 1
Genes confirmed to be associated with dementia mentioned in this article
Disease Chromosome Gene Penetrance Frequency of Mutation Pathology
Mutation in Families

EOAD 21q21 APP Full penetrance About 18% of familial Missense mutations around Increases A␤42
early-onset AD A␤ portion of APP production from
APP processing
EOAD 14q24.3 PSEN1 Full penetrance Represent up to 78% Mostly missense mutations Promotes cleavage at
of familial early- ␥-secretase site and
onset AD increases A␤
production and
accumulation
EOAD 1q31.42 PSEN2 Incomplete penetrance Rare, ⬃4% of familial Missense mutations Promotes cleavage at
early-onset AD ␤-secretase site
leading to increased
A␤ production and
accumulation
CADASIL 19p13 NOTCH3 Full penetrance 400 families described 95% are missense Functional
worldwide, but mutations consequences of
probably under- mutation still
recognized unclear; probably
affects protein
conformation of
Notch3 product
FTDP-17 17 MAPT Full penetrance 10% to 30% of Missense and splice site Increase ratio of 4-
familial FTD, 5% mutations repeat tau to 3-
of all FTD repeat tau
FTLD-U 17 PGRN High penetrance 23% of familial FTD, Missense, splice site, and Leads to null
5% of all FTD promoter region mutations with
mutations haploinsufficiency

Disease Chromosome Gene Penetrance Allele frequency in Mutation Pathology

population

LOAD 19q13.2 APOE Risk factor, but e2, ⬃10%, e4 genotype, heterozygous Acts as a molecular
neither necessary e3, ⬃75%, increases odds by ⬃3 chaperone to
nor sufficient to e4, ⬃15% and homozygous by ⬃9 promote A␤
cause disease accumulation and
senile plaque
formation

Abbreviations: EOAD, early-onset AD; LOAD, late-onset AD; FTDP-17, FTD with parkinsonism linked to chromosome 17.

3. Results peptide is released, which can undergo further conforma-

3.1. Causative mutations for dementia
3.1.1. Alzheimer’s disease
Through linkage analysis of affected pedigrees (posi-
tional cloning), mutations in three genes have been found to
cause early-onset familial AD (Table 1) [1,11]. The first
gene identified was the amyloid precursor protein gene
(APP) on chromosome 21 [12–16]. Although the physio-
logic function of the amyloid precursor protein (APP) re-
mains unclear, the post-translational processing of APP is
clearly involved in the pathogenesis of AD [17,18]. APP is
a membrane bound protein that can undergo a series of
endoproteolytic cleavages by enzymes known as secretases.
When cleaved by ␣-secretase in the middle of the A␤
domain within APP, a soluble fragment of the protein is
released, and there is no accumulation of the peptide;
whereas when it is cleaved at the ␤ and ␥ sites, the A␤
tional change into an insoluble form that aggregates in
senile plaques [19 –21]. APP mutations that are causal for
early-onset familial AD all promote cleavage at the ␤ or ␥
sites, leading to an overproduction of the A␤ peptide[20].
Two other genes with mutations that cause early-onset fa-
milial AD are presenilin 1 (PSEN1) and presenilin 2
(PSEN2), which are located on chromosomes 14 and 1,
respectively [22,23]. Current evidence suggests that both
PSEN1 and PSEN2 play an important role in the ␥-secretase
complex, and mutations in these genes lead to an excessive
cleavage at the ␥ site leading to excessive production and
accumulation of A␤ [17,19]. Taken together, the three
causal genes identified to date for AD provide strong sup-
port for the amyloid cascade hypothesis, in which the ac-
cumulation of A␤ peptide into senile neuritic plaques is
central to the pathogenesis of AD.
To date, although well-documented families exist in
420 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
which late-onset AD (onset after age 65 years) appears to be
inherited in an autosomal dominant manner, no causal mu-
tation(s) has been identified.
3.1.2. Frontotemporal dementia
The use of the term frontotemporal dementia (FTD) has
been inconsistent and is still evolving [4,24 –28]. To clarify,
here we use FTD to refer to the clinical presentation of the
dementing illness and frontotemporal lobar degeneration
(FTLD) to denote the pathologic diagnosis of the disease.
FTD is a clinical syndrome of dementia characterized by
changes in behavior, personality, and social conduct, with
relative preservation of memory function in the early stages
[4]. Incidence estimates of FTD range from 3.3 per 100,000
for 50- to 59-year-olds to 8.9 per 100,000 for 60- to 69-
year-olds [29], representing approximately 5% of all de-
mentias presenting to specialized Canadian clinics [30].
Pathologically, FTLD is associated with circumscribed de-
generation of the prefrontal and anterior temporal lobes with
pathologic changes that are distinctly different from AD
type pathology. The pathology of FTLD is quite varied and
includes (1) the classic presentation of Pick’s disease with
Pick’s bodies, (2) neurofibrillary tangles with tau immuno-
staining, (3) achromatic tau-positive balloon neurons as
seen in corticobasal degeneration, (4) intraneuronal and
intranuclear inclusions that are ubiquitin-positive but tau-
negative (FTLD-U), and (5) dementia lacking distinct his-
topathology (DLDH) [25,26,31].
The clinical presentation of FTLD is also highly variable
and might comprise (1) FTD, exhibiting early decline in
social interpersonal conduct, loss of insight, emotional
blunting, and mental inflexibility; (2) nonfluent progressive
aphasia (also known as primary progressive aphasia), asso-
ciated with early loss of speech and language function; and
(3) semantic dementia, showing loss of ability to name and
understand words and to recognize faces and objects. The
term frontal variant of FTD (FTD-fv) has been used to
specifically refer to the behavioral syndrome of FTD (FTD-
bv) and temporal variant of FTD (FTD-tv) to denote the
clinical presentation of semantic dementia (26,27,32,33). In
some FTD patients, there is coexisting motor neuron dis-
ease, and these cases have been termed FTD-MND or FTD-
ALS. FTD might also present clinically with movement
disorders such as parkinsonism, corticobasal degeneration,
and progressive supranuclear palsy.
In a prevalence study of FTD from the Netherlands, 43%
have a positive family history [34]. In 1998, mutations in
the microtubule-associated protein tau gene (MAPT) on
chromosome 17 were found to be causative for a familial
form of FTD associated with parkinsonism (hence the name
FTDP-17) [35–37]. About 10% to 30% of familial FTD
have a mutation in tau [38,39]. Tau is a phosphoprotein
expressed in both central and peripheral nervous system and
is found predominantly in neuronal axons. The major phys-
iologic function of tau is to bind to and stabilize microtu-
bules. Tau also likely plays a role in microtubule-dependent
axonal transport [40]. Mutations in MAPT causing FTDP-17
generally reduce the ability of tau to interact with microtu-
bules, leading to faulty microtubule assembly [25,40].
Nonetheless, the exact mechanism of how tau causes neu-
rodegeneration remains unclear and is an area of intense
research.
Recently, causative mutations in the progranulin gene
(PGRN) have been identified for an autosomal dominant form
of familial FTLD associated with ubiquitinated neuronal in-
tranuclear inclusions (FTLD-U) [41]. Mutations in PGRN are
responsible for 23% of familial FTD and represent about 5% of
all FTD cases encountered in a clinical setting [42]. PGRN is
a secreted growth factor involved in the regulation of multiple
processes including development, wound repair, and inflam-
mation [43]. Elevated PGRN expression in activated microglia
is found in a number of neurodegenerative diseases including
Creutzfeldt-Jakob disease, amyotrophic lateral sclerosis, and
Alzheimer’s disease, suggesting that PGRN might play an
important role in maintenance of neuronal survival. To date, all
identified PGRN mutations leading to disease development are
due to null mutations, which likely result in haploinsufficiency
[41,44 – 46].
Mutations in other genes, including CHMP2B and VCP,
have also been found in familial FTD, but these appear to be
very rare cases [47,48]. Other genetic loci with significant
linkage to FTD are actively being investigated.
3.1.3. Cerebral autosomal dominant arteriopathy with
subcortical infarcts and leukoencephalopathy
Cerebral autosomal dominant arteriopathy with subcor-
tical infarcts and leukoencephalopathy (CADASIL) is cur-
rently the only known inherited form of vascular dementia
[49]. The clinical spectrum includes recurrent transient isch-
emic attacks, migraine headaches, psychiatric disorders, and
progressive dementia with associated deep white matter
lesions visible on brain imaging. Although initially thought
to be a rare disorder, CADASIL is likely underdiagnosed
because the number of families being identified is continu-
ously increasing [34]. The pathologic hallmark is accumu-
lation of electron dense granules in the media of cerebral
arterioles. To date, more than 100 different mutations in the
NOTCH3 gene on chromosome 19p13 have been reported
in CADASIL patients of white and Asian descent [49 –52].
About 95% are missense point mutations, and the others are
small deletions and one splice site mutation [3]. However,
the mechanism by which these mutations become patho-
genic remains unknown and warrants ongoing research.
3.1.4. Others
There are a number of other genetic diseases associated
with progressive dementia such as Huntington’s disease, some
forms of spinocerebellar ataxia, hereditary amyloid angiopa-
thy, fragile X syndrome, and familial prion diseases. Discus-
sion of each of these diseases is beyond the scope of this
 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
review. However, the principles behind genetic counseling and
testing in these disorders are similar to those discussed here.
4. Genetic services for patients at risk for a causal
mutation for a dementing illness
Causative mutations refer to single gene mutations lead-
ing to expression of a disease. In dementias, most currently
known causative mutations are autosomal dominant
[2,10,53,54], and affected individuals are characterized by
an early-onset dementia (usually younger than age of 60
years). Thus, biologic relatives of affected individuals with
a causal mutation are at risk for the mutation on the basis of
the genetic closeness (DNA sharing) of the relative to the
affected individual.
Currently, predictive genetic testing (PGT) is possible
only for dementia families in whom a known causal muta-
tion has been identified (ie, a specific mutation has been
found in at least one affected family member). PGT is an
option for unaffected biologic relatives of the affected in-
dividual with the causal mutation [55]. For example, em-
pirically, a full brother or sister of an affected individual
would have a 50% (1 in 2) chance of having the mutation,
whereas a half-sibling would have a 25% (1 in 4) chance
and a first cousin a 12.5% risk (1 in 8). In these rare families
(constituting less than 3% of all dementias), a specific ge-
netic change (mutation) travels through the family with the
disease. These specific mutations can be shared by more
than one family or can be specific to a single family. In
families in whom a causal mutation is suspected but not
known, the options are to enroll in a research study or to pay
for testing by commercial laboratories in the USA in which
screens are available for some of the more common muta-
tions.
The two most common mutations in AD (APP and
PSEN1) show virtually 100% penetrance, meaning that an
individual with the family-specific inherited mutation will
almost always develop the disease if he/she lives past the
usual age of disease onset for that family [56,57]. Pen-
etrance of PSEN2 (a rare AD mutation on chromosome 1)
and tau mutations (for FTDP-17) is more variable, with
reports of an octogenarian carrier remaining cognitively
intact while passing on the disease to affected children
[22,58]. However, in contrast to the situation for Hunting-
ton’s disease [12], but similar to that for breast cancer [13],
an individual for whom PGT for familial dementia is an
option must fully understand that the absence of a family-
specific mutation does not “protect” that individual from
AD or another dementia as a result of causes other than the
family-specific mutation. In other words, the absence of the
mutation simply restores the level of risk to that of the
general population (age and ethnically matched), which for
a common disorder such as AD is intrinsically high.
In several prospective studies of patients receiving PGT
for dementias, no negative emotional effects were observed
421
in up to 30 months of follow-up [6]. Nevertheless, all
individuals eligible for PGT should be referred to specialists
in this field. These services are available at Canadian Col-
lege of Medical Genetics’ accredited clinics listed on http://
ccmg.medical.org/clinical.html, and laboratories where mo-
lecular genetic tests are performed are listed on http://
ccmg.medical.org/molecular.html.
In principle, prenatal genetic diagnosis is an option for
families showing an autosomal dominant dementia with an
identified mutation, but this is not standard practice in Can-
ada. Such concern is not unique to dementias but is faced by
prospective parents for many other late-onset genetic disor-
ders such as familial breast cancer and familial hyperlipid-
emia. Individual cases might need to be discussed with the
appropriate medical ethics board [55]. Although there is
agreement that a young person carrying a family-specific
genetic mutation needs genetic counseling about the risks of
passing that mutation to a child, there is great concern
among the genetics community about terminating a preg-
nancy when the disease onset will probably not occur until
an older age. It is hoped that the progress in understanding
the molecular pathogenesis of dementias might realistically
be expected to lead to effective treatment or ideally preven-
tion within the next few decades. For this reason the present
discussion will exclude prenatal genetic diagnosis.
5. Genetic counseling for unaffected individuals at risk
for familial dementia, but no family-specific causal
mutation has been identified
For families with no known mutation but showing clin-
ical presentation of an early-onset dementia with a family
history compatible with an autosomal dominant mode of
inheritance, genetic risk counseling must follow the predic-
tions of the autosomal dominant model for the specific
family structure (eg, twin of an affected individual, off-
spring of an affected parent, niece of an affected uncle
whose mother died well before the age of onset in the
family, etc). It is critical to obtain as much documentation as
possible on reportedly affected individuals to determine the
“best estimate” clinical diagnosis. In large families, this
might be very labor intensive, and the clinician might opt to
refer the family to a specialized center, http://ccmg.medi-
cal.org/clinical.html. It is not uncommon to find that al-
though families are initially believed to have familial AD,
on follow-up they have heterogeneous etiologies with re-
spect to dementia (eg, alcohol-related dementia, Parkinson’s
disease, depression). Such findings would significantly alter
the risk estimates for genetic counseling [55,59].
On rare occasions an unaffected individual might still
want to be tested even though no DNA is available from an
affected family member. This type of genetic testing is not
offered as a service in Canada but might be available com-
mercially. Individuals in these situations might opt for test-
ing through private companies, where one can screen for the
422 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
most common PSEN1 mutations. However, this is not en-
couraged, because proper counseling before and after the
tests and assistance with interpretation of test results are not
often in place.
Because it is believed that novel genes for familial de-
mentias, including late-onset AD, will be identified in the
future, families might have the opportunity to bank tissue
material (through autopsy) and DNA (through a blood sam-
ple or a buccal smear) from affected individuals for the
potential future benefit of other family members [57,58].
This must of course be done with appropriate ethics ap-
proval and signed informed consents.
6. Genetic susceptibility factor in AD
In the previous section we discussed causal genetic mu-
tations. In contrast, the identification of a genetic suscepti-
bility factor increases the likelihood (risk) of an individual
developing the disease within his/her lifetime, but there is
also a good chance that this will not occur. Thus, although
the proportion of individuals who develop the disease is
greater among those who have the specific genetic risk
factor compared with those who do not, it is still possible for
an individual to possess the risk factor but never develop the
disease (low sensitivity), whereas some people who do not
possess the risk factor might still develop the disease (low
specificity).
The literature review described above identified more
than 1,000 publications related to the study of genetic risk
factors for AD, with 308 genes implicated [10]. However, to
date, only one gene, APOE, has been consistently confirmed
as a genetic susceptibility factor for AD. There are three
common polymorphisms in APOE, with e3 occurring in
about 75% of chromosomes in white populations, with e2 in
8% and e4 in 15% [60]. The association of the e4 allele in
APOE with late-onset AD was first reported in 1993
[61,62]. There is an apparent gene-dosing effect according
to the actual genotype; the odds of developing AD for
carriers of one APOE e4 allele is 3.2 (95% confidence
interval [CI], 2.9 to 3.5) compared with e3/e3 individuals,
whereas the odds for homozygous e4/e4 carriers is 11.6
(95% CI, 8.9 to 15.4) [8,63]. In a meta-analysis with 5,930
patients and 8,607 controls, the e2 allele was shown to be
associated with a lower risk of AD (odds ratio, 0.6; 95% CI,
0.5 to 0.8 in people with e2/e3 genotype in reference to
e3/e3) [9]. Although APOE as a genetic risk factor for AD
has been demonstrated in populations of various ethnicity,
the strength of association appears to be weaker among
African Americans and Hispanics compared with whites,
whereas it is stronger among Japanese [9,64 – 69]. In addi-
tion, APOE exhibits interactions with age and gender, with
the highest odds occurring between ages 55 and 65 years,
and these odds are greater for female ⑀4 carriers compared
with male. How APOE e4 actually affects the pathogenesis
of AD remains an area of intense investigation, although
current evidence suggests that it is a molecular chaperone
that helps promote the deposition of A␤ into amyloid
plaques [70].
Although the APOE e4 allele has been confirmed to be a
genetic susceptibility risk for AD, its use as a diagnostic tool
has been examined in several studies. The consensus was
that because of its low sensitivity and specificity, it should
not be used for general screening of patients at risk or for
diagnosis of AD [71–73]. On the other hand, several pro-
spective cohort studies suggest that it might significantly
add to a predictive model in the assessment of patients
categorized as mild cognitive impairment (MCI) or cogni-
tively impaired not demented (CIND) whose risk of pro-
gression to dementia is high [69,71,74,75]. There are also
three randomized controlled trials demonstrating that APOE
genotype can affect therapeutic response to cholinesterase
inhibitors [76 –78]. Therefore, APOE genotype might be
used, with appropriate counseling, to stratify patients at risk
of progression to dementia, and it might potentially have an
important role in the assessment of therapeutic interventions
that prevent or delay the progression of AD.
A recent randomized clinical trial conducted to examine
the impact of genetic risk assessment for adult children of
people with AD shows that revealing their APOE genotype
when presenting their risk of dementia did not increase
symptoms of depression or anxiety, whereas the genetic
information helped them make long-term life-planning de-
cisions and motivated their engagement in risk reduction
activities [5]. It must be noted that this study used MSc
trained genetic counselors to convey the information. A
study examining the public perception on predictive testing
of AD found that not only individuals with family history of
AD are interested, and many people without family history
of AD are just as interested in their genetic risks [79].
Clinicians must be prepared to discuss, at least in some
detail, the topics of predictive testing in autosomal domi-
nant families and genetic risk factors such as APOE when
raised in the clinical setting.
7. Conclusion
Although more than 300 genes have been implicated in
the pathogenesis of AD and related dementias, at present
only three (APP, PSEN1, PSEN2) have been confirmed to
be causatively associated with early-onset AD and two
(MAPT and PGRN) with FTD. Causal genes have also been
identified for some diseases in which dementia is a major
component of the phenotype, including CADASIL, Hun-
tington’s disease, and familial Creutzfeldt-Jakob disease.
Genetic testing for diagnostic and predictive purposes is
only available as a clinical service for families in whom a
causal mutation has already been identified in an affected
family member. Predictive testing in minors and prenatal
diagnosis are not within the normal patterns of practice in
Canada, but they are dealt with on a case-by-case basis,
 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
Fig 1. A hypothetical pedigree in which individuals are at risk for inheriting the disease genetic mutation.
often involving local medical ethics committees (http://cc-
mg.medical.org/clinical.html.)
Commercial services for genetic testing are not endorsed
because of the apparent need for longitudinal counseling of
individuals having such tests with respect to both the inter-
pretation of the results and the reaction and adjustment after
revelation of the findings.
For AD the only confirmed genetic susceptibility risk
factor is the APOE e4 allele. Nevertheless, because of its
low specificity and sensitivity, APOE genotyping should not
be used as either a screening or a diagnostic tool. However,
in patients with mild cognitive impairment, research has
shown that APOE genotype might be used to stratify pa-
tients at risk of progression and to predict response to
therapy. Research is ongoing to identify novel genetic fac-
tors important in the risk of developing dementia.
8. Summary of recommendations
8.1. Genetic testing for confirmation of the etiology of a
dementia
8.1.1. Preamble
Genetic mutations have been identified in certain types
of familial dementia, as discussed in this article.
8.1.2. Recommendation 1
Symptomatic patients with a clear family history of a
dementia should be referred to specialized dementia clinics
423
for determination of the best estimate clinical diagnosis
(Grade A, Level 2). Testing for specific mutations might be
arranged by the specialist through accredited genetic centers
(http://ccmg.medical.org/clinical.html). In situations in
which a family-specific mutation has been identified, failure
to find evidence for this mutation in another clearly affected
family member can be interpreted as being (1) phenocopy,
(2) dementia as a result of other causes, (3) non-paternity,
(4) non-maternity (eg, when the patient does not know about
an adoption).
8.2. Predictive genetic testing for asymptomatic at-risk
individuals with an apparent autosomal dominant
inheritance, and a family-specific mutation has been
identified
8.2.1. Preamble
Gene mutations for dementia or neurodegenerative dis-
ease with dementia tend to be inherited in an autosomal
dominant pattern. The majority of mutations identified to
date show high penetrance and expressivity.
8.2.2. Recommendation 2
With appropriate pre-testing and post-testing counseling,
PGT can be offered to at-risk individuals (Grade B, Level
2). Examples are (1) first-degree relatives of an affected
individual with the mutation (eg, children and siblings;
Figure 1, individuals III-2, III-3, and II-6); (2) first cousins
of an affected individual if the common ancestors (parents
424 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
who were siblings) died before the average age of onset of
dementia in the family (eg, individual II-7 in Figure 1); (3)
nieces and nephews of affected individuals whose parent
(sibling of the affected individual) died well before the
average age of onset of dementia in the family (eg, individ-
ual III-4 in Figure 1); (4) PGT in minors is not generally
offered in Canada, but it occasionally might be considered
on a case-by-case basis by the relevant medical ethics com-
mittee(s) (eg, individual III-6 in Figure 1); (5) individuals
who are not at risk for the inherited disease do not require
testing (eg, individual III-1 in Figure 1).
8.2.3. Preamble
De novo mutations can arise in an early-onset dementia
patient with a negative family history.
8.2.4. Recommendation 3
In young persons (60 years or younger) presenting with
an early-onset dementia, it is sometimes worthwhile to test
for the most common mutations on the basis of the best
estimate diagnosis (eg, in early-onset AD, one might test for
the most common mutations in PSEN1, APP) (Grade B,
Level 2). If a mutation is identified, it would have direct
implications for offspring of the individual (if a de novo
mutation is assumed). It would also be important to test
other family members such as parents and siblings for
possible non-penetrance of a mutation.
8.2.5. Preamble
Genetic counseling can be offered on the basis of the
assumptions of an autosomal dominant model of inheritance
in some families in whom affected individuals occur in at
least 2 generations, or there are at least 3 cases in the
collateral generation (siblings, first cousins) with insuffi-
cient information available on the previous generation, even
in the absence of an identified mutation.
8.2.6. Recommendation 4
Careful review of all available documentation (examina-
tion, clinical records, autopsy reports, etc) on reportedly
affected relatives is essential to rule out the heterogeneous
causes of dementia in the elderly, including depression,
alcoholism, vascular dementia. If appropriate after review,
genetic counseling should include the risks associated with
an autosomal dominant mode of inheritance as the upper-
most risk. For the benefit of future generations, it is advis-
able to bank DNA and/or autopsy material from affected
individuals in such families in case novel gene mutations
might be discovered in the future (Grade B, Level 2).
8.3. Prenatal genetic testing
8.3.1. Recommendation 5
Although prenatal testing for a known family mutation
associated with adult-onset dementia is technically possible
(eg, individual III-5 in Figure 1), this is not generally of-
fered in Canada. Currently, there is insufficient evidence to
recommend prenatal testing for dementia (Grade C, Level
3). The ethical implications are complex and need to be
explored further. Requests might be considered on a case-
by-case basis by the relevant medical ethics committee(s).
8.4. Ethical issues in genetic testing
8.4.1. Preamble
Guidelines still need to be developed for genetic testing
in an affected individual who is demented and not cogni-
tively competent to give consent. Such testing probably will
not benefit the individual directly but could impact on future
generations if a known mutation is identified, or if new
mutations are discovered in the future through ongoing
research.
8.4.2. Recommendation 6
After careful genetic counseling with family members,
if it is decided that genetic testing and/or banking of
DNA (or autopsy material) for future studies is in the best
interest of family members, this might be done even
without the consent or assent of the affected individual
who is not cognitively competent, if consent from the
family is given (Grade B, Level 2). However, extreme
care must be taken to minimize any distress the patient
might experience while the sample is being obtained. In
cases in which family members, after extensive counsel-
ing, cannot agree on a plan, the case might have to go
before a medical ethics committee.
8.4.3. Preamble
A relatively rare but certainly real concern with ge-
netic testing is the situation in which testing one family
member can identify an “obligate carrier” who perhaps
does not want this information. For example, in a family
with AD because of PS1 with a mean age of onset of 50
years, testing the daughter could in fact (assuming correct
parentage) identify the unaffected parent as an obligate
carrier (Figure 1, II-6 and III-7). The ethical dilemma is
the right of the daughter to know versus the right of the
parent not to know.
8.4.4. Recommendation 7
Concerned family members should have appropriate pre-
test and post-test counseling available to be able to make
informed decisions (Grade B, Level 2). In cases of conflict
among family members, medical ethics committees might
become involved.
8.5. Genetic susceptibility risk factors
8.5.1. Preamble
APOE e4 increases risk of progression to AD in patients
with MCI/CIND. It also affects response to cholinesterase
inhibitors.
8.5.2. Recommendation 8
Genetic screening with APOE genotype in asymptomatic
individuals in the general population is not recommended
 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
because of the low specificity and sensitivity (Grade E,
Level 2).
8.5.3. Recommendation 9
Genetic testing with APOE genotype is not recom-
mended for the purpose of diagnosing AD because the
positive and negative predictive values are low (Grade E,
Level 2).
8.5.4. Recommendation 10
Although APOE e4 cannot be used as a diagnostic or a
screening test for AD, it has been consistently shown to be
a significant risk factor for progression from MCI/CIND to
AD (Grade B, Level 2). It might be used to stratify MCI/
CIND patients at risk for progression, and it might also be
important in the assessment of therapeutic efficacy in clin-
ical research (Grade B, Level 1).
Note: Recommendations 1 and 10 received only weak
consensus (76% each), whereas recommendations 2 to 9
received strong consensus (greater than 80%).
Author Disclosures
Ging-Yuek Robin Hsiung has received support from
Novartis (Speaker fees), and Pfizer ⫹ Ono (Site Investigator
on clinical trials).
A. Dessa Sadovnick has received support from Biogen
Idec (Speaker, Meeting Participant), Serono (Speaker,
Meeting Participant), Teva Neuroscience (Funding for mul-
tiple sclerosis research, Speaker, Meeting Participant), and
Berlex Canada (Funding for multiple sclerosis research,
Speaker, Meeting Participant).
Acknowledgments
A. Dessa Sadovnick is a recipient of a Michael Smith
Distinguished Scholar Award.
Ging-Yuek Robin Hsiung is supported by funds from the
St. Paul’s Hospital Foundation.
References
[1] Bertram L, Tanzi RE. The current status of Alzheimer’s disease
genetics: what do we tell the patients? Pharmacol Res 2004;50:385–
96.
[2] Pastor P, Goate AM. Molecular genetics of Alzheimer’s disease. Curr
Psychiatry Rep 2004;6:125–33.
[3] Federico A, Bianchi S, Dotti MT. The spectrum of mutations for
CADASIL diagnosis. Neurol Sci 2005;26:117–24.
[4] Neary D, Snowden J, Mann D. Frontotemporal dementia. Lancet
Neurol 2005;4:771– 80.
[5] Roberts JS, Cupples LA, Relkin N, Whitehouse P, Green RC, Group
ftRS. Genetic risk assessment for adult children of people with Alz-
heimer’s disease: the Risk EValuation and Education for ALzhei-
mer’s Disease (REVEAL) Study. J Geriatr Psychiatry Neurol 2005;
18:250 –5.
[6] Molinuevo JL, Pintor L, Peri JM, Lleo A, Oliva R, Marcos T, et al.
Emotional reactions to predictive testing in Alzheimer’s disease and
425
other inherited dementias. Am J Alzheimers Dis Other Demen
2005;20:233– 8.
[7] Evans GA. The Human Genome Project: applications in the diagnosis
and treatment of neurologic disease. Arch Neurol 1998;55:1287–90.
[8] Rubinsztein DC, Easton DF. Apolipoprotein E genetic variation and
Alzheimer’s disease. a meta-analysis. Dement Geriatr Cogn Disord
1999;10:199 –209.
[9] Farrer LA, Cupples LA, Haines JL, Hyman B, Kukull WA, Mayeux
R, et al. Effects of age, sex, and ethnicity on the association between
apolipoprotein E genotype and Alzheimer disease: a meta-analysis—
APOE and Alzheimer Disease Meta Analysis Consortium. JAMA
1997;278:1349 –56.
[10] Bertram L, McQueen M, Mullin K, Blacker D, Tanzi R. The AlzGene
Database: Alzheimer Research Forum. Available at: http://www.alz-
gene.org. Accessed Feb 14, 2006.
[11] Rocchi A, Pellegrini S, Siciliano G, Murri L. Causative and suscep-
tibility genes for Alzheimer’s disease: a review. Brain Res Bull
2003;61:1–24.
[12] Tanzi RE, Hyman BT. Studies of amyloid beta-protein precursor
expression in Alzheimer’s disease. Ann N Y Acad Sci 1991;640:
149 –54.
[13] Selkoe DJ, Podlisny MB, Joachim CL, Vickers EA, Lee G, Fritz LC,
et al. Beta-amyloid precursor protein of Alzheimer disease occurs as
110- to 135-kilodalton membrane-associated proteins in neural and
nonneural tissues. Proc Natl Acad Sci U S A 1988;85:7341–5.
[14] Schellenberg GD, Pericak-Vance MA, Wijsman EM, Moore DK,
Gaskell PC Jr, Yamaoka LA, et al. Linkage analysis of familial
Alzheimer disease, using chromosome 21 markers. Am J Hum Genet
1991;48:563– 83.
[15] Goate A, Chartier-Harlin MC, Mullan M, Brown J, Crawford F,
Fidani L, et al. Segregation of a missense mutation in the amyloid
precursor protein gene with familial Alzheimer’s disease. Nature
1991;349:704 – 6.
[16] Kang J, Lemaire HG, Unterbeck A, Salbaum JM, Masters CL, Grz-
eschik KH, et al. The precursor of Alzheimer’s disease amyloid A4
protein resembles a cell-surface receptor. Nature 1987;325:733– 6.
[17] Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease:
progress and problems on the road to therapeutics. Science 2002;297:
353– 6.
[18] Selkoe DJ, Schenk D. Alzheimer’s disease: molecular understanding
predicts amyloid-based therapeutics. Annu Rev Pharmacol Toxicol
2003;43:545– 84.
[19] Brunkan AL, Goate AM. Presenilin function and gamma-secretase
activity. J Neurochem 2005;93:769 –92.
[20] Selkoe DJ. The cell biology of beta-amyloid precursor protein and
presenilin in Alzheimer’s disease. Trends Cell Biol 1998;8:447–53.
[21] Sudoh S, Kawamura Y, Sato S, Wang R, Saido TC, Oyama F, et al.
Presenilin 1 mutations linked to familial Alzheimer’s disease increase
the intracellular levels of amyloid beta-protein 1-42 and its N-termi-
nally truncated variant(s) which are generated at distinct sites. J Neu-
rochem 1998;71:1535– 43.
[22] Rogaev EI, Sherrington R, Rogaeva EA, Levesque G, Ikeda M, Liang
Y, et al. Familial Alzheimer’s disease in kindreds with missense
mutations in a gene on chromosome 1 related to the Alzheimer’s
disease type 3 gene. Nature 1995;376:775– 8.
[23] Sherrington R, Rogaev EI, Liang Y, Rogaeva EA, Levesque G, Ikeda
M, et al. Cloning of a gene bearing missense mutations in early-onset
familial Alzheimer’s disease. Nature 1995;375:754 – 60.
[24] Bird T, Knopman D, VanSwieten J, Rosso S, Feldman H, Tanabe H,
et al. Epidemiology and genetics of frontotemporal dementia/Pick’s
disease. Ann Neurol 2003;54(Suppl 5):S29 –31.
[25] Munoz DG, Dickson DW, Bergeron C, Mackenzie IR, Delacourte A,
Zhukareva V. The neuropathology and biochemistry of frontotempo-
ral dementia. Ann Neurol 2003;54(Suppl 5):S24 – 8.
426 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
[26] Kertesz A, McMonagle P, Blair M, Davidson W, Munoz DG. The
evolution and pathology of frontotemporal dementia. Brain 2005;
128(Pt 9):1996 –2005.
[27] Kertesz A. Pick complex: an integrative approach to frontotemporal
dementia—primary progressive aphasia, corticobasal degeneration,
and progressive supranuclear palsy. Neurologist 2003;9:311–7.
[28] Kertesz A, Munoz DG, Hillis A. Preferred terminology. Ann Neurol
2003;54)Suppl 5):S3– 6.
[29] Knopman DS, Petersen RC, Edland SD, Cha RH, Rocca WA. The
incidence of frontotemporal lobar degeneration in Rochester, Minne-
sota, 1990 through 1994. Neurology 2004;62:506 – 8.
[30] Feldman H, Levy AR, Hsiung GY, Peters KR, Donald A, Black SE,
et al. A Canadian cohort study of cognitive impairment and related
dementias (ACCORD): study methods and baseline results. Neuro-
epidemiology 2003;22:265–74.
[31] Mackenzie IR, Shi J, Shaw CL, Duplessis D, Neary D, Snowden JS,
et al. Dementia lacking distinctive histology (DLDH) revisited. Acta
Neuropathol (Berl) 2006;112:551–9.
[32] Perry RJ, Hodges JR. Differentiating frontal and temporal variant
frontotemporal dementia from Alzheimer’s disease. Neurology 2000;
54:2277– 84.
[33] Bozeat S, Gregory CA, Ralph MA, Hodges JR. Which neuropsychi-
atric and behavioural features distinguish frontal and temporal vari-
ants of frontotemporal dementia from Alzheimer’s disease? J Neurol
Neurosurg Psychiatry 2000;69:178 – 86.
[34] Rosso SM, Donker Kaat L, Baks T, Joosse M, de Koning I, Pijnen-
burg Y, et al. Frontotemporal dementia in The Netherlands: patient
characteristics and prevalence estimates from a population-based
study. Brain 2003;126(Pt 9):2016 –22.
[35] Hutton M, Lendon CL, Rizzu P, Baker M, Froelich S, Houlden H, et
al. Association of missense and 5’-splice-site mutations in tau with
the inherited dementia FTDP-17. Nature 1998;393:702–5.
[36] Poorkaj P, Bird TD, Wijsman E, Nemens E, Garruto RM, Anderson
L, et al. Tau is a candidate gene for chromosome 17 frontotemporal
dementia. Ann Neurol 1998;43:815–25.
[37] Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti
B. Mutation in the tau gene in familial multiple system tauopathy with
presenile dementia. Proc Natl Acad Sci U S A 1998;95:7737– 41.
[38] Stanford PM, Brooks WS, Teber ET, Hallupp M, McLean C, Halliday
GM, et al. Frequency of tau mutations in familial and sporadic
frontotemporal dementia and other tauopathies. J Neurol 2004;251:
1098 –104.
[39] Poorkaj P, Grossman M, Steinbart E, Payami H, Sadovnick A, Noch-
lin D, et al. Frequency of tau gene mutations in familial and sporadic
cases of non-Alzheimer dementia. Arch Neurol 2001;58:383–7.
[40] Iqbal K, Alonso Adel C, Chen S, Chohan MO, El-Akkad E, Gong
CX, et al. Tau pathology in Alzheimer disease and other tauopathies.
Biochim Biophys Acta 2005;1739:198 –210.
[41] Baker M, Mackenzie IR, Pickering-Brown SM, Gass J, Rademakers
R, Lindholm C, et al. Mutations in progranulin cause tau-negative
frontotemporal dementia linked to chromosome 17. Nature 2006;442:
916 –9.
[42] Gass J, Cannon A, Mackenzie IR, Boeve B, Baker M, Adamson J, et
al. Mutations in progranulin are a major cause of ubiquitin-positive
frontotemporal lobar degeneration. Hum Mol Genet 2006;15:2988 –
3001.
[43] He Z, Bateman A. Progranulin (granulin-epithelin precursor, PC-cell-
derived growth factor, acrogranin) mediates tissue repair and tumor-
igenesis. J Mol Med 2003;81:600 –12.
[44] Masellis M, Momeni P, Meschino W, Heffner R Jr, Elder J, Sato C,
et al. Novel splicing mutation in the progranulin gene causing familial
corticobasal syndrome. Brain 2006;129(Pt 11):3115–23.
[45] Cruts M, Gijselinck I, van der Zee J, Engelborghs S, Wils H, Pirici D,
et al. Null mutations in progranulin cause ubiquitin-positive fronto-
temporal dementia linked to chromosome 17q21. Nature 2006;442:
920 – 4.
[46] Mukherjee O, Pastor P, Cairns NJ, Chakraverty S, Kauwe JS, Shears S,
et al. HDDD2 is a familial frontotemporal lobar degeneration with ubiq-
uitin-positive, tau-negative inclusions caused by a missense mutation in
the signal peptide of progranulin. Ann Neurol 2006;60:314 –22.
[47] Guyant-Marechal L, Laquerriere A, Duyckaerts C, Dumanchin C,
Bou J, Dugny F, et al. Valosin-containing protein gene mutations:
clinical and neuropathologic features. Neurology 2006;67:644 –51.
[48] Momeni P, Rogaeva E, Van Deerlin V, Yuan W, Grafman J, Tierney
M, et al. Genetic variability in CHMP2B and frontotemporal demen-
tia. Neurodegener Dis 2006;3:129 –33.
[49] Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, Mouton P,
et al. Notch3 mutations in cerebral autosomal dominant arteriopathy
with subcortical infarcts and leukoencephalopathy (CADASIL), a
mendelian condition causing stroke and vascular dementia. Ann N Y
Acad Sci 1997;826:213–7.
[50] Suwanwela N, Srikiatkhachorn A, Tangwongchai S, Phanthumchina
K. Mutation of the Notch 3 gene in a Thai cerebral autosomal
dominant arteriopathy with subcortical infarcts and leukoencephalop-
athy family. J Med Assoc Thai 2003;86:178 – 82.
[51] Thomas NJ, Morris CM, Scaravilli F, Johansson J, Rossor M, De
Lange R, et al. Hereditary vascular dementia linked to notch 3
mutations: CADASIL in British families. Ann N Y Acad Sci 2000;
903:293– 8.
[52] Uyama E, Tokunaga M, Suenaga A, Kotorii S, Kamimura K, Taka-
hashi K, et al. Arg133Cys mutation of Notch3 in two unrelated
Japanese families with CADASIL. Intern Med 2000;39:732–7.
[53] Bird TD. Genetic factors in Alzheimer’s disease. N Engl J Med
2005;352:862– 4.
[54] Kalaria RN, Viitanen M, Kalimo H, Dichgans M, Tabira T. The
pathogenesis of CADASIL: an update. J Neurol Sci 2004;226:35–9.
[55] Sadovnick A, editor. Genetic counselling in dementia. Oxford: Black-
well, 2002.
[56] Rogaeva E, Tandon A, St George-Hyslop PH. Genetic markers in the
diagnosis of Alzheimer’s disease. J Alzheimers Dis 2001;3:293–304.
[57] Rogaeva EA, Fafel KC, Song YQ, Medeiros H, Sato C, Liang Y, et
al. Screening for PS1 mutations in a referral-based series of AD cases:
21 novel mutations. Neurology 2001;57:621–5.
[58] Bird TD, Sumi SM, Nemens EJ, Nochlin D, Schellenberg G, Lampe
TH, et al. Phenotypic heterogeneity in familial Alzheimer’s disease:
a study of 24 kindreds. Ann Neurol 1989;25:12–25.
[59] Sadovnick AD. Genetic counselling and genetic screening for Alz-
heimer’s disease and other dementias. Can J Neurol Sci 2001;
28(Suppl 1):S52–5.
[60] Zannis VI, Kardassis D, Zanni EE. Genetic mutations affecting hu-
man lipoproteins, their receptors, and their enzymes. Adv Hum Genet
1993;21:145–319.
[61] Strittmatter WJ, Saunders AM, Schmechel D, Pericak-Vance M,
Enghild J, Salvesen GS, et al. Apolipoprotein E: high-avidity
binding to beta-amyloid and increased frequency of type 4 allele in
late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A
1993;90:1977– 81.
[62] Saunders AM, Roses AD. Apolipoprotein E4 allele frequency, isch-
emic cerebrovascular disease, and Alzheimer’s disease. Stroke 1993;
24:1416 –7.
[63] Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell
PC, Small GW, et al. Gene dose of apolipoprotein E type 4 allele and
the risk of Alzheimer’s disease in late onset families. Science 1993;
261:921–3.
[64] Nishimura AL, Oliveira JR, Mitne-Neto M, Guindalini C, Nitrini
R, Bahia VS, et al. Lack of association between the brain-derived
neurotrophin factor (C-270T) polymorphism and late-onset Alz-
heimer’s disease (LOAD) in Brazilian patients. J Mol Neurosci
2004;22:257– 60.
 G-Y.R. Hsiung and A.D. Sadovnick / Alzheimer’s & Dementia 3 (2007) 418 – 427
[65] Raygani AV, Zahrai M, Doosti M, Javadi E, Rezaei M, Pourmot-
abbed T. Association between apolipoprotein E polymorphism and
Alzheimer disease in Tehran, Iran. Neurosci Lett 2005;375:1– 6.
[66] Lai CL, Tai CT, Lin SR, Lin RT, Yang YH, Liu CK. Apolipoprotein
E in Taiwan Chinese patients with dementia. Dement Geriatr Cogn
Disord 2003;16:208 –11.
[67] Katzman R, Zhang MY, Chen PJ, Gu N, Jiang S, Saitoh T, et al.
Effects of apolipoprotein E on dementia and aging in the Shanghai
Survey of Dementia. Neurology 1997;49:779 – 85.
[68] Tang MX, Stern Y, Marder K, Bell K, Gurland B, Lantigua R, et al.
The APOE-epsilon4 allele and the risk of Alzheimer disease among
African Americans, whites, and Hispanics. JAMA 1998;279:751–5.
[69] Hsiung GY, Sadovnick AD, Feldman H. Apolipoprotein E epsilon4
genotype as a risk factor for cognitive decline and dementia: data
from the Canadian Study of Health and Aging. CMAJ 2004;171:
863–7.
[70] Shobab LA, Hsiung GY, Feldman HH. Cholesterol in Alzheimer’s
disease. Lancet Neurol 2005;4:841–52.
[71] Aggarwal NT, Wilson RS, Beck TL, Bienias JL, Berry-Kravis E,
Bennett DA. The apolipoprotein E epsilon4 allele and incident Alz-
heimer’s disease in persons with mild cognitive impairment. Neuro-
case 2005;11:3–7.
[72] Mayeux R, Saunders AM, Shea S, Mirra S, Evans D, Roses AD, et al.
Utility of the apolipoprotein E genotype in the diagnosis of Alzheimer’s
disease: Alzheimer’s Disease Centers Consortium on Apolipoprotein E
and Alzheimer’s Disease. N Engl J Med 1998;338:506 –11.
427
[73] Tierney MC, Szalai JP, Snow WG, Fisher RH, Tsuda T, Chi H, et al.
A prospective study of the clinical utility of ApoE genotype in the
prediction of outcome in patients with memory impairment. Neurol-
ogy 1996;46:149 –54.
[74] Petersen RC, Smith GE, Ivnik RJ, Tangalos EG, Schaid DJ, Thi-
bodeau SN, et al. Apolipoprotein E status as a predictor of the
development of Alzheimer’s disease in memory-impaired individu-
als. JAMA 1995;273:1274 – 8.
[75] Devanand DP, Pelton GH, Zamora D, Liu X, Tabert MH, Goodkind
M, et al. Predictive utility of apolipoprotein E genotype for Alzheimer
disease in outpatients with mild cognitive impairment. Arch Neurol
2005;62:975– 80.
[76] Petersen RC, Thomas RG, Grundman M, Bennett D, Doody R, Ferris
S, et al. Vitamin E and donepezil for the treatment of mild cognitive
impairment. N Engl J Med 2005;352:2379 – 88.
[77] Bizzarro A, Marra C, Acciarri A, Valenza A, Tiziano FD, Brahe C, et
al. Apolipoprotein E epsilon4 allele differentiates the clinical re-
sponse to donepezil in Alzheimer’s disease. Dement Geriatr Cogn
Disord 2005;20:254 – 61.
[78] Poirier J, Delisle MC, Quirion R, Aubert I, Farlow M, Lahiri D, et al.
Apolipoprotein E4 allele as a predictor of cholinergic deficits and
treatment outcome in Alzheimer disease. Proc Natl Acad Sci U S A
1995;92:12260 – 4.
[79] Cutler SJ, Hodgson LG. To test or not to test: interest in genetic
testing for Alzheimer’s disease among middle-aged adults. Am J
Alzheimers Dis Other Demen 2003;18:9 –20.
 Alzheimer’s & Dementia 3 (2007) 428 – 440

Toward a revision of criteria for the dementias

Kenneth Rockwooda,*, Rémi W Bouchardb, Richard Camiciolic, Gabriel Légerd
a
Department of Medicine (Geriatric Medicine & Neurology), Dalhousie University, Halifax, Nova Scotia, Canada
b
Department of Neurology, Laval University, Québeec, Québec, Canada
c
Department of Medicine (Neurology), University of Alberta, Edmonton, Alberta, Canada
d
Department of Medicine, University of Montreal, Montréal, Québec, Canada

Abstract This article critically considers current diagnostic criteria for dementia and reports recommen-
dations approved by at least 80% of experts attending the Third Canadian Consensus Conference on
the Diagnosis and Treatment of Dementia. There was consensus that many of the features proposed
as essential to a diagnosis of dementia in the 1980s no longer are relevant (for example, the
requirements for memory impairment, electroencephalogram and cerebrospinal fluid studies, age-
specific exclusions). In addition, other syndromes such as frontotemporal dementia have been
recognized and need to inform new dementia criteria. It is also recognized that a diagnosis of
depression need not exclude a dementia diagnosis. Other proposals, such as neuropathology
should be considered as additional evidence and not as a gold standard or that some people with
dementia have prolonged plateaus so that progressive decline need not be a criterion for
Alzheimer’s disease, were more controversial and did not receive similar support. Given the
evidence of the last three decades, there is merit in reconsidering the criteria by which dementia and
Alzheimer’s disease are diagnosed.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: Dementia; Alzheimer’s disease; Diagnostic criteria; Neuropathology; Frontotemporal dementia; Depression

1. Introduction: What can we know with certainty
about dementia diagnosis?
This report constitutes a background article for the Third
Canadian Consensus Conference on the Diagnosis and
Treatment of Dementia (CCCDTD3) held in March 2006.
This article differs from the other 17 articles preceding it in
that the goal was not so much to deliver recommendations
for current physician management of dementing diseases as
to point the way toward revising and improving the diag-
nostic criteria for different forms of dementia being used
around the world. Because much of this knowledge is pro-
posive, there is no possibility for randomized controlled
trials or even clinical-pathologic studies that compare sub-
jects diagnosed according to different criteria. In keeping
with the fashion of evidence grading, these recommenda-
tions would all be rated as Level 2 or 3 evidence and would
*Corresponding author. Tel.: (902) 473-8687; Fax: (902) 473-1050.
E-mail address: kenneth.rockwood@dal.ca
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.07.014
represent Grade B recommendations. Unless otherwise in-
dicated, the recommendations presented here were ap-
proved by at least 80% of the votes cast by the 45 experts
who attended the CCCDTD3, representing the experts on
dementia in Canada from Geriatric Medicine, Neurology,
Psychiatry, Neuropsychology, and Family Practice. Be-
cause these recommendations were all debated and reached
broad consensus within the entire Canadian academic com-
munity, we hope that these proposals will initiate a serious
debate on revision of criteria during the coming years.
As people age, most notice changes in their cognitive
functioning. Some changes are beneficial. Expert learners,
for example, have vast resources of prior information on
which to draw, thus allowing them readily to contextualize
and interpret complex information. But for most people, the
perceptible differences that occur with age are declines in
function, such as less efficient word finding or less efficient
execution of mental calculation.
Until about the late 1960s, the differential diagnosis of
memory-related complaints could operate with reasonable
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
certainty. It was recognized, for example, that although
age-related changes in cognitive functioning were common,
these were not, in themselves, a disease. A clear separation
was made between such age-related benign changes and the
malignant changes of the dementias, especially Alzheimer’s
disease (AD). These differences could be validated against
autopsy, which represented a gold standard for definitive
diagnosis.
Since the late 1960s, tremendous progress has been made
in understanding dementia. Quite apart from progress in
molecular neurobiology, which is likely to impact how we
think about dementia in years to come, whole new dementia
syndromes have been identified (eg, frontotemporal demen-
tia with relative memory sparing) as have new diagnostic
entities, such as dementia with Lewy bodies (DLB), or
human immunodeficiency virus–related dementia.
Some of the progress in understanding cognition has
undermined prior certainties. Importantly, the comforting
notion of the gold standard no longer seems valid. Although
neuropathology has an important contribution to make, the
recognition that people with apparently identical lesions can
differ widely in their cognitive function and that differing
sets of neuropathologic criteria yield differing estimates of
dementias in the same brains each suggest that its role, like
that of other measures, best provides construct (cf. criterion)
validity. Similarly, the sharp line between normal aging and
pathology is not as clear as it once was. It does not seem
plausible that we should chiefly look only for single causes
of dementia; on a population basis, dementias with more
than one cause occur commonly, and our criteria must
reflect this.
In general, there are two approaches to handling concep-
tual uncertainty. One is to clear out all the underbrush and
find a firm ground on which new structures can be built,
such as a new gold standard, and thus new, definitive cri-
teria. A second approach is to recognize that conceptual as
well as instrumental uncertainty is part of this stage of our
endeavors. Gold standards are unlikely to be widely avail-
able, multiple causes of dementia will be a reality, and any
new consensus criteria will need to anticipate revision. In
consequence, it will be necessary to build in methods of
information handling that allow the contribution of individ-
ual items to be evaluated.
Our approach conforms most closely to the second, and
thus we have chosen to recognize uncertainty rather than to
effect a consensus that will assume it away. In what follows,
we will briefly recapitulate the state of the art in dementia
diagnosis, enumerate the chief challenges that new data
have provided, and propose how to proceed in a way that
allows pragmatic clinical decisions to be made without
adopting a false sense of security. This is a philosophical
decision, not an evidence-based one. So too would be the
decision to pursue any other path.
429
2. Where have we come from in dementia diagnosis?
2.1. Overview
There are many ways to define and classify the demen-
tias. For example, 10 years ago, a comprehensive Cana-
dian review of the nosology of dementia noted four types
of classification [1]: (1) histologic/neuropathologic (eg,
DLB); (2) etiologic (eg, vascular, infectious, but also
genetic); (3) brain region involved (eg, the focal atro-
phies); and (4) clinical profile (eg, dementia with motor
signs, early-onset, rapidly progressive).
For the syndromic diagnosis, it is necessary to begin
broadly. Dementia is typically defined as a clinical syn-
drome of cognitive (historically memory is here privileged)
decline that is sufficiently severe to interfere with social or
occupational functioning. Routine clinical practice shows
that these cognitive and functional changes typically are
accompanied by changes in behavior and in personality, but
these have had variable prominence in diagnostic criteria.
Dementia remains a clinical diagnosis, in which labora-
tory or imaging tests as yet provide only supportive roles.
Although some view the development of biomarkers as
essential for advancing the field [2], experience from other
fields shows that unless there is very clear clinical correla-
tion, the effect is to substitute the problem of the seeming
arbitrariness of clinical markers with the irrelevance of
biologic ones [3]. As such, the diagnosis is based on a
careful history. In specialist practice, this is typically done
by using a semistructured interview with an informant, eg,
the Clinical Dementia Rating (CDR) [4] or the Brief Cog-
nitive Rating Scale (BCRS) [5], and a detailed medical and
neurologic examination. The CDR and BCRS each incor-
porate neurocognitive testing, which otherwise can be
added by using instruments such as the Behavioural Neu-
rology Assessment [6]. At the syndromic level, the differ-
ential diagnosis classically includes the cognitive disorders
of delirium and depression, as well as the possibility of
having no cognitive impairment. Since the proposal for
Benign Senescent Forgetfulness [7] the possibility of a
diagnosis that falls short of dementia but cannot be regarded
as normal cognition must also be considered. In the Cana-
dian context, during the last dozen years this would include
terms such as cognitive impairment, no dementia (CIND)
[8 –10] and mild cognitive impairment (MCI) and its sub-
types [11,12] and questionable and very early Alzheimer’s
disease [13].
To construct a differential diagnosis of the cause of the
dementia, the starting point is typically the clinical presen-
tation. In memory clinics, with the predominance of AD, the
usual question is whether the presentation conforms to the
usual pattern seen with that illness. Otherwise, the usual
considerations of the nature of onset, the duration of symp-
toms, the degree of progression, and the presence of medical
and neurologic signs allow characteristic syndromes to be
430 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
identified. As detailed in the last Canadian consensus con-
ference on the assessment of dementia [14], many condi-
tions can confound the diagnosis of dementia including
communication disorders, comorbid psychiatric and medi-
cal illnesses including developmental disorders. In addition,
conditions such as transient global amnesia, subclinical ep-
ileptic syndrome, and medication toxicity challenge clini-
cians to be alert to their possibility when evaluating patients
with cognitive complaints. A careful history, examination,
and longitudinal follow-up will elucidate whether an under-
lying or coexisting dementing process exists, so that rational
investigation and treatment can be carried out.
2.2. Current general criteria for dementia
Even though the DSM IIIR criteria [15] have notionally
been displaced by DSM IV [16] and DSM-IV-TR, the most
recent version [17], each remains in wide use. DSM-IV-TR
requires impairment of memory and at least one of the
following domains: language, praxis, gnosis, executive
functioning. These impairments need to be sufficiently se-
vere to impair social or professional life and must not occur
as a consequence of a delirium or by another medical,
neurologic, or psychiatric condition. Memory impairment,
although present in most people with dementia, is not an
essential requirement; instead, at least two cognitive do-
mains must be impaired, and of course the other criteria
must be met. The presence of dementia without a dominant
memory complaint has been demonstrated in large cohorts
of patients with dementia [18,19].
2.2.1. Recommendations
1. Although memory impairment is an important part of
most dementias, there are some dementias (subcorti-
cal ischemic dementia, primary progressive aphasia,
some other types of frontotemporal dementia [FTD])
in which the requirement for memory impairment
limits the sensitivity of a dementia diagnosis. The
requirement for memory impairment should be dropped
from the criteria for dementia in favor of impairment in
at least two domains of cognitive function.
2. The new onset of a mood disorder or behavioral
disturbances, in the face of changes in cognition,
should be seen as supportive of a diagnosis of
dementia.
3. Alzheimer’s disease
The diagnosis of AD rests on the standard criteria for
dementia being met and typically are seen with insidious
onset and progressive decline. This approach generally is
nonproblematic, especially because memory is a predomi-
nant symptom in AD. The requirement for an insidious
onset should be seen as supportive but not absolute; uncom-
plicated AD can have an acute onset, for example, as with
coming on after a delirium [20,21].
3.1. Recommendation
3. An acute onset should not necessarily exclude a di-
agnosis of AD. (This recommendation received weak
consensus support only.)
The NINCDS-ADRDA [22] criteria for probable AD
have also been used for research purposes worldwide and
remain applicable for clinical diagnosis in practice. They are
summarized as follows: dementia based on clinical exami-
nation and neuropsychological tests, with onset between
ages 40 and 90 years; deficits in two or more areas of
cognition; progressive worsening of memory and other cog-
nitive functions; normal consciousness; supportive features
include progressive deterioration of language, praxis, and
gnosis, impaired activities of daily living and behavior,
positive familial history, normal cerebrospinal fluid (CSF),
nonspecific electroencephalogram (EEG), and progressive
cerebral atrophy with time on neuroimaging. Other possible
features include some plateaus in the course, seizures in late
stage, some neuropsychiatric and extrapyramidal symp-
toms, and normal computed tomography or magnetic reso-
nance imaging (MRI) scans for age. Although CSF and
EEG studies would not be part of a usual evaluation, they
sometimes can be indicated clinically. Neuroimaging that is
normal for age can be seen in established disease, especially
in its early stages. The discussion of EEG and CSF studies
as being “supportive” likely reflects an earlier era in which
AD was believed to be a diagnosis of exclusion; therefore
these tests are no longer appropriate as considerations in
usual practice. Similarly, there is no reason a priori to
exclude patients younger than 40 or older than 90.
3.2. Recommendations
4. There is no need to suggest that routine EEG and CSF
studies help exclude a diagnosis of Alzheimer’s disease.
5. There is no reason a priori to exclude patients younger
than 40 or older than 90.
Although the NINDS-ADRDA criteria are widely used,
when interpreted strictly they select so-called pure AD pa-
tients. In consequence, they exclude AD patients with
mixed disorders. Broader use of the DSM-IV-TR category
of “Dementia due to multiple etiologies” needs to be en-
couraged, with specification of the diseases contributing to
the dementia routinely spelled out.
3.3. Recommendation
6. Broader use of the DSM-IV-TR category of “Demen-
tia due to multiple etiologies” needs to be encouraged,
with specification of the diseases contributing to the
dementia routinely spelled out.
At present, standard diagnostic criteria for dementia do
not adequately recognize focal presentations. Although un-
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
common, focal presentations of dementia occur and can be
recognized in expert hands. One example is progressive visual
impairment [23,24]. At present, standard diagnostic criteria for
dementia do not adequately recognize focal presentations.
3.4. Recommendation
7. Revision of AD criteria should recognize the possi-
bility of focal presentations.
In mild dementia, depression is common [25]. In contrast
to an earlier era in which it was thought to be important to
rule out depression before making a diagnosis of AD, there
is now recognition of symptom overlap, perhaps reflecting
shared pathophysiology.
3.5. Recommendation
8. Depressive features should be recognized as concom-
itant in patients with AD and should not exclude a
diagnosis.
Much of the misery that arises from a diagnosis of AD
has less to do with the patient’s current state than with the
prospect of disease progression. Should meaningful decline
be operationalized in the criteria? It is well-accepted that
patients with vascular dementia can have periods in which
their illness plateaus [26]. This has been less well-accepted
with AD, where the experience of relentless progression is
usual. Indeed, despite occasional suggestions that “the
course of dementia can be progressive, static or remittent”
[27], in neurodegenerative dementias, the idea of progres-
sion of cognitive impairment has been characterized as
“obvious and essential” [28]. Nevertheless, several groups
have observed that a significant minority of people with AD
(about 15%) [29 –31] exhibit comparatively little disease
progression. Two broad possibilities suggest themselves to
account for apparent stability of supposedly progressive
disorders, falsely negative progression as a result of insen-
sitivity of the measures to detect change over time or true
variability in progression [32,33]. Studies of the latter pos-
sibility have generally focused on people with more rapid
progression (eg, as has been suggested in some studies of
DLB in comparison with AD without concomitant DLB
[32]). In short, although dementia is typically considered to
be a progressive illness, relative stabilization in vascular
cognitive impairment (VCI), fluctuation in DLB, and vari-
ability in the rates of progression of AD, with both very
rapid [34] and slowly progressive forms, need to be recog-
nized [35]. Each variation contributes to the nuance that
operational criteria require to capture the variable phenom-
ena encountered in clinical practice [34].
3.6. Recommendation
9. Although most people with AD have a progressive
cognitive decline, periods of relative stability (or of
431
rapid decline) can also occur. (In consequence, the
requirement for continued cognitive decline should be
dropped from the criteria for AD in clinical practice.)
(Note: This recommendation by the authors was con-
sidered controversial and did not reach consensus at
the meeting.).
In 1984 the AD consensus criteria proposed that neuro-
pathology serve as the gold standard in the diagnosis of AD
and other causes of dementia. Much data since then suggest
that such a view is naive. Instead, it is reasonable to view
the status of neuropathology not in providing criterion val-
idation but as another factor to consider, in short, in pro-
viding construct validity. In 1996 Stuss and Levine, citing
work that showed that some elderly people met neuropatho-
logic criteria for dementia but in fact were not demented
[36,37], asked whether neuropathology alone could be con-
sidered as the absolute criterion for the classification of
dementia. The question remains [38]. In addition, although
many studies have demonstrated a good correlation between
clinical severity and Braak’s neuropathologic staging, stud-
ies in very elderly people suggest more that Braak’s staging
represents a broad concept of the evolution of neurofibril-
lary tangles rather than a precise hierarchical model of
cognitive decline abilities near the upper limit of life [39].
No area of scientific inquiry is without controversy, so the
fact that neuropathologists disagree about the neuropatho-
logic diagnosis of dementia does not disqualify it as a useful
measure [40]. Still, although the classic criteria for AD have
a high degree of accuracy (usually reported at ⫾90% [41–
45], this has been disputed when referral and work-up bias
are taken into account [46].
3.7. Recommendation
10. The classification of neuropathology as providing a
definitive diagnosis of AD (or other cause of demen-
tia) should be dropped. Instead, semiquantitative and
quantitative criteria should be applied to pathologic
findings to model neurodegenerative diseases in a
manner that goes beyond simplistic dichotomous
categories. (Note: This recommendation by the au-
thors was considered controversial and did not reach
consensus at the meeting.).
4. Frontotemporal dementia
People who are not dementia specialists could be for-
given for feeling somewhat at sea when getting to grips with
FTD. There is particular heterogeneity in the frontal lobe
dementias, both clinically and neuropathologically, and it
has led to many attempts at classification. None has yet
proved to be entirely satisfactory. A 2001 consensus con-
ference popularized a three-part account [47]. First, the
account proposed an easy to recognize frontolobar degen-
eration, which consists of difficulty in adapting to social
432 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
situations, with disinhibition, impulsivity, loss of insight,
and other aspects of impaired social conduct. In addition,
two language-dominant forms were suggested, primary pro-
gressive aphasia according to the classic description of
Mesulam [48] and semantic dementia. The idea that each
involves a frontal/insular/anterior cingulate circuitry, with a
left lateralized degeneration in the language disorders, is
attractive and has some empirical support [49,50]. Still, the
tripartite approach does not go without many notations. For
example, it does not say enough about the movement dis-
orders that commonly accompany FTD or about the syn-
drome of frontal lobe dementia with motor neuron disease.
Moreover, it is silent on cases in which memory problems are
an early feature, such as very old adults. In such cases, differ-
entiation from atypical AD can be especially challenging.
One classification describes frontotemporal lobar degen-
eration [41] with these core features: insidious onset and
gradual progression, early impairment in social and per-
sonal conduct, early emotional blunting, and loss of insight;
supportive features include examples of behavioral changes,
speech and language impairment, primitive reflexes, and
extrapyramidal sign. Imaging studies are said to show pre-
dominant atrophy in the frontal and/or temporal region,
supported by frontal lobe abnormalities on neuropsycholog-
ical testing. The latter can also show relative sparing of
memory, spatial orientation, and perception.
A Canadian contribution describes a set of disorders that
can be called Pick’s complex [51]. This classification em-
phasizes the overlap that can be seen with the syndromes,
both clinically and pathologically, and also draws attention
to an akinetic rigid syndrome, which can be a common and
disabling part of the evolution of these disorders and which
is not always given its due.
Before these more recent efforts, the Lund and Manches-
ter criteria [52,53] were widely used, but they increasingly
appear to be displaced by more recent classifications. For
the present purposes, the elucidation of the clinical profile of
FTD makes evident the need not to privilege memory over
other cognitive disorders in defining dementia [54]. This
experience reinforces recommendation 1.
The rest of the criteria are variations on the DSM-IV-TR
[17] approach and include the notion of impairment in
social or occupational functioning, a decline compared with
the previous level of functioning, a gradual onset and pro-
gressive decline, and the deficits do not occur only during
delirium and cannot be explained by any other medical,
neurologic, or psychiatric disorder.
It is important to note that primary progressive aphasia
and semantic aphasia (also called semantic dementia) might
initially not meet dementia criteria. Initially those syn-
dromes are primarily aphasic disorders and will progress to
a dementia syndrome, often with a dysexecutive syndrome
and/or apraxia. Considering the initial focality of these
syndromes and the possibility of many underlying etiolo-
gies, specific operationalized criteria for these subgroups are
not recommended for primary clinical practice. Neverthe-
less, many recently published studies regarding the clinical
differential profile of these progressive aphasic disorders
have led to some degree of consensus for the diagnostic
criteria that might be useful for research and clinical trials
[41,46,50,55,56].
4.1. Recommendation
11. Specific frontal temporal lobar degeneration criteria
should be used, including criteria that recognize
aphasic disorders and including primary progressive
aphasia and semantic dementia, when these disor-
ders are considered. At present, no one set of FTD
criteria captures all the variants encountered in prac-
tice. The DSM-IV-TR category of “dementia due to
Pick’s disease” specifically is inadequate.
5. Dementia with Lewy bodies
It is vitally important that physicians have a good work-
ing knowledge of DLB, because it is one dementia in which
common treatments can often be either especially bene-
ficial (eg, the dramatic improvement sometimes seen with
cholinesterase inhibitors [57]) or markedly detrimental
(the dramatic, even fatal worsening sometimes seen with
neuroleptics [58]).
The description of DLB has thus been an essential con-
tribution to modern dementia practice. Still, the initial two
sets of consensus criteria [58,59], while offering good spec-
ificity, were believed to be inadequately sensitive to the
protean manifestations of DLB, so that in late 2005 a third
set of criteria were proposed [60]. These criteria added new
clinical features and reclassified previously supportive fea-
tures into two categories. Features are now separated into
those with a higher frequency of occurrence compared with
other dementia syndromes (suggestive features) or those
with a common occurrence but lower specificity (supportive
features). For example, the former now includes severe
neuroleptic sensitivity and rapid eye movement sleep be-
havior disorder (which can be manifested by occasionally
dramatic verbal and motor outbursts that both often pre-date
the dementia by years). In addition, imaging studies that
suggest reduced striatal dopamine uptake are now seen as
substantially supporting the diagnosis of DLB. Further-
more, new supportive features include early autonomic
dysfunction, characteristic metaiodobenzylguanidine car-
diac scintigraphy, and structural imaging that shows pres-
ervation of the medial temporal lobe structures. The cri-
teria also noted the common difficulty of appropriately
quantifying fluctuation by history alone and suggested ways
to operationalize this common but sometimes clinically
challenging part of the diagnosis of DLB.
The diagnosis of DLB in routine practice often varies
according to the clinical service on which it is made [61].
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
Stereotypically, psychiatrists will more often see the phe-
notype in which hallucinations are common, neurologists
where movement disorders predominate, and geriatricians
where loss of autonomy is especially a problem. Operation-
ally, this challenge is readily dealt with, but the conceptual
issue of how to handle overlap, here between DLB and other
disorders, again must be faced. The promoters of criteria for
any disorder that aim to select “pure” patients for research
purposes need to recognize that this is not inherently more
“scientific” but rather a decision to trade off internal validity
for generalizability. Of note, the presence of a single sug-
gestive feature in the absence of any of the core features
now allows a diagnosis of possible DLB.
A technological approach to DLB diagnosis, which uses
single photon emission computed tomography (SPECT) im-
aging with a dopamine transporter scan (DaTScan) [62], is
gaining popularity. In a multi-center European study, the
accuracy of classification between probable DLB, possible
DLB, and non-DLB (chiefly AD) was between 84% and
87% [63]. Although impressive, the real value added by
such an approach would be its ability to distinguish not
between patients with clinically evident DLB and clinically
evident AD but to help diagnosis in cases that clinically are
rated as possible DLB. In such cases the distribution of scan
results was bimodal, with roughly half having abnormal
scans and half not. In consequence, the role of DaTScans in
aiding usual practice is not yet clear. In terms of neuropsy-
chology, the clinical diagnosis is aided by recognizing that
visuoconstructional tasks best distinguish DLB from AD.
On the ubiquitous Mini-Mental State Examination (MMSE)
[64] it has been suggested that copying a pentagon incor-
rectly is more characteristic of DLB than AD, but it is
somewhat insensitive. In general, the more complex the
visual task, the more sensitive it is for DLB, but of course
specificity is traded off, because AD patients will also have
trouble.
5.1. Recommendation
12. Because of the present uncertainty with criteria for
DLB, use of the broader 2005 consensus criteria will
require additional data to test the hypothesis that
they have better accuracy. Pending new confirma-
tory studies, it is reasonable to continue to use the
1996 consensus criteria in routine clinical practice.
6. Parkinsonian dementia
As noted, parkinsonism is an important aspect of the
clinical diagnosis of DLB. Obviously too, parkinsonism will
be present in the dementia seen with Parkinson’s disease
(PD), which is termed Parkinson’s disease dementia (PDD).
It seems likely that the difference between DLB and PDD
reflects where LBs first accumulate. Patients with halluci-
nations, inattention, and cognitive impairment have them in
433
the cortex; those with striatal LBs first will have PD. For
now, PDD has yet to benefit from validated clinical criteria
for its diagnosis, although it is clear that dementia occurs
commonly in advanced PD [65] and in older patients [66],
so that about 30% to 40% of PD patents, on a population
basis, have dementia [65– 68].
One additional consideration is the overlap between
DLB/PDD and AD. Clinicopathologic reports of the extent
of overlap with AD yield variable estimates [69,70], but
there is no reason to expect that PDD will be exempt from
the mixed pathology seen with other dementias (eg, AD and
vascular contribution) [71–74]. Nevertheless, PDD is a dif-
ferent clinical syndrome in the spectrum of DLB. In the
setting of DLB, it is particularly important to consider how
to deal with the relationship between DLB, AD, and PDD.
It seems reasonable to suggest that patients with predomi-
nantly neocortical LBs will have the classic triad of hallu-
cinations, fluctuation, and parkinsonism, whereas striatal
LBs will give rise to the motor disorder first. In addition, the
more that coincident AD pathology exists, the less likely it
is that the presentation will be classic. There is a greater loss
of nigral DA neurons in PDD versus DLB, reflecting dif-
ferential initial focal involvement of the LBs. The single LB
disease model is probably best for neurobiologic studies,
but the DLB/PDD distinction has advantages for classifying
patients. The 2005 DLB consensus criteria used an arbitrary
1-year window for parkinsonism symptoms to precede cog-
nitive changes in patients to be labeled DLB [60], so that a
delay beyond 1 year would suggest PDD. Note too that
cognitive decline often can be detected with early PD but
well before dementia [75,76]. Because some patients with
PDD [77] as with DLB [78] respond well to cholinesterase
inhibitors, in clinical practice it would seem important not to
withhold a treatment trial on the basis of deficiencies in the
current criteria. A full review of therapy for these conditions
is outside the mandate of this article.
6.1. Recommendation
13. Criteria for PDD are needed and likely are best
incorporated in new criteria that update those for
dementia with LBs.
7. Vascular dementia and vascular
cognitive impairment
Vascular dementia (VaD) occurring with AD is common,
but “pure” vascular dementia appears to be uncommon [79].
A full review of the management of dementia with a cere-
brovascular component is presented in article by Bocti et al
(this issue). In consequence, there are both conceptual and
pragmatic difficulties in trying to portray VaD as a distinct
entity. Of the criteria now in use (the NINDS-AIREN [80]
and ADTCC criteria [81] have been especially influential),
no set is perfect. In practice, most are insensitive [82], and
434 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
different sets of criteria give differing estimates of who has
VaD [82– 86]. Vascular lesions are found in many dementia
patients, including those with otherwise classic AD, where
they are often detected only by routine neuroimaging
[87]. On the other hand, patients with only vascular
pathology as the cause of their dementia have been un-
common in many series [44,87]. Most patients have
mixed pathology, so that patients with vascular lesions
commonly have evidence of other neurodegenerative dis-
orders [38,71,72]. It is also important to note that many
patients with positive imaging do not have a stroke his-
tory [82,88] and vice versa [89].
The NINDS-AIREN criteria for VaD are problematic,
because they adhere to a multiple infarction model and also
privilege memory impairment among cognitive deficits,
when in fact executive dysfunction often predominates [90].
Although the requirement for a temporal relationship be-
tween dementia and stroke makes sense for cases of post-
stroke dementia, it does not apply to the cases (which in
memory clinics typically will be the majority [82]) in which
cognitive deterioration might be slowly progressive rather
than stepwise [91]. Moreover, it does not acknowledge the
common incidence of silent infarctions [92].
Against this background, the proposal to expand the
concept to one of VCI makes sense [93]. Still, the broad
concept has meant that subtypes are necessary for clinical
recognition. Three subtypes in particular are recognized:
VCI–no dementia (VCI-ND), a subcortical vascular demen-
tia with a prominent dysexecutive profile and white matter
changes on neuroimaging, and VaD with multiple or single
strategic infarcts. These profiles appear to be readily recog-
nizable in memory clinic practices [82,89].
8. Recommendation
14. The proposal to develop new criteria for VCI has
merit. The criteria should move away from the mod-
els of post-stroke dementia and multi-infarct demen-
tia. Moreover, they should emphasize opportunities
for prevention of impairment that arise as a conse-
quence of potentially modifiable cerebrovascular
disease.
9. Normal pressure hydrocephalus
Normal pressure hydrocephalus (NPH) is not a particu-
larly common cause of dementia, but as a potentially treat-
able syndrome, it has special importance. Classically, it is
characterized as dementia seen with gait impairment and
with urinary urgency or incontinence [94]. NPH can be
idiopathic or secondary, usually as a result of hemorrhage
or other brain injury. It remains relatively understudied,
making the development of evidence-based guidelines
difficult [95].
Recent consensus guidelines published in 2005 focused
on clinical documentation of one or more of the character-
istic symptoms of idiopathic NPH in combination with a
brain imaging study demonstrating nonobstructive ventric-
ular enlargement disproportionate to cerebral atrophy. It
was suggested that probable, possible, and unlikely NPH
classifications be used. Adjunct tests, involving transient
removal of CSF via lumbar puncture or lumbar drain, were
advocated for the dual purpose of adding to diagnostic
certainty and assisting in prognostication about response to
treatment [95].
A common clinical scenario is the patient referred for
evaluation for NPH not on the basis of the characteristic
clinical triad but on the basis of the neuroimaging that has
been interpreted as compatible with it and the suggestion
that NPH be ruled out. The presence of central atrophy,
associated white matter changes, the extent of sulcal efface-
ment and the pattern of the gyri in relation to these features
each offer clues from the scan in the clinically nonevident
case. Such patients require careful follow-up.
Typically, the differential diagnosis of NPH includes
obstructive hydrocephalus, multiple systems atrophy (asso-
ciated with ataxia and incontinence), vascular gait impair-
ment, and AD with gait impairment. Where the clinical triad
is the starting point, the finding of hydrocephalus on imag-
ing is predictive of shunt responders [96]. Other clinical
features that suggest a good response include a shorter
duration of illness, lack of prominent cortical cognitive
deficits, and good response to a removal of CSF (by external
or internal lumbar drainage or by single or repeated lumbar
puncture) [97]. As with other dementias, mixed pathologies
are not uncommon. Thus neuropsychological demonstration
of anomia or marked medial temporal lobe atrophy on MRI
might suggest AD and a poorer potential for response to
shunting. Although longer duration of cognitive impairment
is a predictor of poor response, patients might respond
despite the presence of negative prognostic indicators [98].
Ancillary tests are less helpful in predicting response. In
particular, intracranial pressure monitoring and ventricular
cisternography are not recommended routinely [95].
9.1. Recommendation
15. The diagnosis of dementia caused by NPH should
use the recent 2005 consensus criteria, recognizing
their limitations. In uncertain cases referral to a
specialized clinic is recommended.
10. Dementias associated with infectious diseases
Several infectious diseases might be associated with a
progressive dementia. Human immunodeficiency virus pro-
duces a so-called subcortical dementia that can sometimes
occur in the absence of a history of known human immu-
nodeficiency virus or acquired immune deficiency syn-
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
drome [99]. Neurosyphilis, although now rare, is potentially
treatable [100]. Acute viral encephalitis can leave patients
with dementia and can sometimes present in an indolent
fashion. Clues to acute and subacute central nervous system
infection include headache, fever, seizures, systemic com-
plaints, infection of peripheral tissue, rapid progression,
focal neurologic features (in the setting of abscesses with
parenchymal involvement), and travel to endemic areas.
Immunosuppression predisposes to infections in general,
and specific disorders, such as progressive multifocal leu-
koencephalopathy, are important to consider in this setting.
Worldwide tuberculosis and neurocysticercosis are common
infections that have a predilection for the central nervous
system. Lyme disease is an infection that should be consid-
ered in individuals with appropriate symptoms, such as a
rash and polyarthritis, from an endemic area. Creutzfeldt-
Jakob disease (CJD), Whipple’s disease, and other infec-
tions associated with dementia generally offer clues on the
clinical examination as noted above. Other infections have
abnormalities on CSF examination, including elevated pro-
tein, pleocytosis, and evidence for microorganisms on ex-
amination or culture. The polymerase chain reaction is use-
ful for amplifying genomic material for specific infections,
including herpes simplex and Whipple’s disease.
10.1. Recommendation
16. In young patients or those with atypical presenta-
tions, systemic features, or relevant risk factors, in-
fectious causes should be borne in mind in the dif-
ferential diagnosis of dementia. Focal cognitive
deficits raise this possibility, even when they do not
meet traditional dementia criteria. Such features
should prompt specialist referral.
11. Creutzfeldt-Jakob disease
CJD is a rapidly progressive dementia that can only be
definitively diagnosed by a brain tissue examination that
shows prion proteins with associated spongiform changes.
CJD can be sporadic or familial (associated with mutations
in gene coding for the prion protein) or transmissible. Trans-
missible forms of CJD include iatrogenic and variant CJD.
Criteria for probable CJD have been proposed (http://www.
eurocjd.ed.ac.uk/def.htm) and include typical EEG features
with at least two of the following clinical features: myoclonus,
visual or cerebellar signs, pyramidal/extrapyramidal signs, or
akinetic mutism. The clinical diagnosis can also be made
with a history of a rapidly progressive dementia with dura-
tion of less than 2 years and at least two of the above clinical
features, with a positive 14-3-3 test suggested in recent
iterations of the criteria. The clinical criteria might be in-
sensitive in early cases [101] that have had pathologic
confirmation with accepted criteria [102,103]. The utility of
the 14-3-3 test has come into question because of its vari-
435
able sensitivity and lack of specificity [104]. Diffusion-
weighted MRI might have superior sensitivity compared
with EEG and superior specificity compared with CSF ex-
amination [105]. In general, sporadic CJD is inexorably
progressive, resulting in death within a year; however, the
course of sporadic CJD can be considerably longer [106]. A
long duration is common in the autosomal dominant prion
disease, Gerstmann-Straussler-Scheinker syndrome. Variant
CJD is a progressive neuropsychiatric disorder ultimately
leading to ataxia, dementia, and myoclonus (or chorea)
without the typical EEG appearance of CJD or the propor-
tion with elevation in the 14-3-3 protein [107]. Young-onset
sporadic CJD cases share features of a long neuropsychiat-
ric prodrome with variant CJD [108].
Spongiform changes are also found in variant CJD,
where they are found in the pulvinar in 70% of cases [107].
It has been proposed that MRI be included among diagnos-
tic criteria, given reasonable sensitivity and specificity in
some series [109]. Periodic sharp complexes on EEG
changes are characteristic but can occur transiently or late in
the course of disease and are thus insensitive [110]. False
positives include cases of DLB, AD, and VaD.
11.1. Recommendation
17. Documentation of rapid progression should raise the
possibility of CJD and suggests prompt referral.
MRI should be done in the appropriate clinical
setting.
12. Other dementias with specific etiologies
Nondegenerative dementias caused by acquired systemic
or metabolic derangements and genetic conditions can be
diagnosed if they otherwise meet DSM-IVTR criteria and
are often associated with a specific etiology. Genetic con-
ditions are rare, tend to occur in younger individuals, and
are often inherited in an autosomal recessive fashion. Hun-
tington’s disease is common among young-onset dementias
and is associated with an autosomal dominant family his-
tory. Cerebral autosomal dominant arteriopathy with sub-
cortical infarcts and leukoencephalopathy (CADASIL) is an
example of an autosomal dominant VaD that illustrates the
overlap between genetic and vascular conditions [111].
Acquired conditions of acute onset might be consistent
with vascular causes of dementia, including cerebrovascular
disease [112], vasculitis, and amyloid angiopathy (with or
without hemorrhage) [112,113]. Nonvascular causes can
also give rise to an acute-onset dementia [20]. Although
delirium can occur in the absence of dementia, they com-
monly co-occur; dementia is a risk factor for delirium and
vice versa [114,115]. Rapid progression of symptoms can
be seen in AD and DLB [116,117]. Dementia associated
with motor neuron disease can also have a rapid course
[118]. Dramatic fluctuations in the level of consciousness
436 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
can mimic an acute stroke-like onset and are characteristic
of DLB and can be seen in PDD [119]. Other syndromes
that should be considered in the differential diagnosis of a
rapidly progressive dementia include viral encephalitis,
paraneoplastic (limbic) encephalitis, central nervous system
cancer, Hashimoto’s encephalitis, other autoimmune disor-
ders (including anti-phospholipid syndrome, systemic lupus
erythematosus, sarcoidosis, and non-vasculitic autoimmune
meningoencephalitis) [120,121], infections, and metabolic
disorders. Neuroimaging and other specific studies can be
helpful in identifying ischemic changes and in evaluating
other brain lesions.
Other dementias for which broad experience is lacking,
which have variably valid diagnostic criteria, include argyr-
ophilic grain disease [122], neurofilament inclusion disease
[123], late-life hippocampal sclerosis [124 –126], corticoba-
salganglionic degeneration, progressive supranuclear palsy,
Whipple’s disease [127], Huntington’s disease, pantothe-
nate kinase–associated neurodegeneration [128], autoim-
mune channelopathies [129], leukodystrophies [130], and
superficial siderosis [131].
13. Concluding comments
Some of what is needed to advance the diagnosis of
dementia is consensus not on a final set of criteria but on
recording enough information that will allow future criteria
to be developed on the basis of data—to yield “evidence-
based not eminence-based” recommendations. In Canada
this has been the rationale behind the ACCORD [19] and
CIVIC [82] studies.
Similarly, clinical pathologic correlation studies should
examine the pathologic markers as quantitative traits that
might have a dose-response and topographic-response rela-
tionship to various aspects of the dementia syndrome. For
example, many disputes about neuropathologic markers
now cast them as either/or [132], so that the possibility of
their use in the aggregate, for example, as an index variable
[133], has been less well pursued.
Our emphasis is on refining criteria as we go and on
anticipating future developments that will change the way
we think about cognitive disorders. This strikes us as ap-
propriate to the optimism inspired by the last 10 years of
progress.
Author Disclosures
Kenneth Rockwood has received support from Pfizer
Canada (Consultant, Medical Advisory Board Member,
Speaker), Janssen Ortho Inc. (Consultant, Medical Advisory
Board Member, Research Grant Holder), Novartis (Speaker,
Consultant), Lundbeck Canada (Advisory Board Member),
Merck Frosst (Speaker), CIHR (Research and Career Award
Holder), Alzheimer Society of Canada (Research Grant
Holder), Dalhousie Medical Research Foundation (Career
Award Holder), and the Nova Scotia Department of Health
(Consultant).
Rémi W. Bouchard has received support from Pfizer
Canada (Consultant, Advisory Board Member, Speaker,
Lecturer, Clinical Trials), Novartis (Consultant, Advisory
Board Member, Speaker, Lecturer, Clinical Trials), Janssen
Ortho Inc. (Consultant, Advisory Board Member, Speaker,
Lecturer, Clinical Trials), and Lundbeck Canada (Consul-
tant, Advisory Board Member, Speaker, Lecturer, Clinical
Trials).
Richard Camicioli has received support from Novartis
(Grant, Speaker), Pfizer (Grant, Medical Advisory Board
Member, Speaker), Janssen-Ortho (Medical Advisory
Board Member, Speaker), and GlaxoSmithKline (Medical
Advisory Board Member).
Gabriel Léger has received support from Janssen Ortho
Inc. (Meeting Participant), Novartis (Meeting Participant,
Speaker), and Pfizer Canada (Meeting Participant, Speaker).
Acknowledgments
Kenneth Rockwood is supported by an Investigator
Award from the Canadian Institutes of Health Research
(CIHR) and the Dalhousie Medical Research Foundation as
the Kathryn Allen Weldon Professor of Alzheimer’s Re-
search. He receives research support from the CIHR and
from the Alzheimer Society of Canada.
Richard Camicioli is supported by an operating grant
from the CIHR.
References
[1] Stuss DT, Levine B. The dementias: nosological and clinical factors
related to diagnosis. Brain Cogn 1996;31:99 –113.
[2] Wiltfang J, Lewczuk P, Riederer P, Grunblatt E, Hock C, Scheltens
P, et al. Consensus paper of the WFSBP Task Force on Biological
Markers of Dementia: the role of CSF and blood analysis in the early
and differential diagnosis of dementia. World J Biol Psychiatry
2005;6:69 – 84.
[3] Rothwell PM. External validity of randomised controlled trials: “to
whom do the results of this trial apply?” Lancet 2005;365:82–93.
[4] Morris JC. The Clinical Dementia Rating (CDR): current version
and scoring rules. Neurology 1993;43:2412– 4.
[5] Reisberg B, Ferris SH. Brief Cognitive Rating Scale (BCRS). Psy-
chopharmacol Bull 1988;24:629 –36.
[6] Darvesh S, Leach L, Black SE, Kaplan E, Freedman M. The behav-
ioural neurology assessment. Can J Neurol Sci 2005;32:167–77.
[7] Kral VA. Senescent forgetfulness: benign and malignant. Can Med
Assoc J 1962;86:257– 60.
[8] Ebly EM, Hogan DB, Parhad I. Cognitive impairment in the non-
demented elderly: results from the Canadian Study of Health and
Aging. Arch Neurol 1995;52:612–9.
[9] Graham JE, Rockwood K, Beattie BL, Eastwood R, Gauthier S,
Tuokko H, et al. Prevalence and severity of cognitive impairment
with and without dementia in an elderly population. Lancet 1997;
349:1793– 6.
[10] Tuokko H, Frerichs R, Graham J, Rockwood K, Kristjansson B, Fisk
J, et al. Five-year follow-up of cognitive impairment with no dementia.
Arch Neurol 2003;60:577– 82.
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
[11] Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV,
et al. Current concepts in mild cognitive impairment. Arch Neurol
2001;58:1985–92.
[12] Winblad B, Palmer K, Kivipelto M, Jelic V, Fratiglioni L, Wahlund
LO, et al. Mild cognitive impairment– beyond controversies, to-
wards a consensus: report of the International Working Group on
Mild Cognitive Impairment. J Intern Med 2004;256:240 – 6.
[13] Morris JC, Storandt M, Miller JP, McKeel DW, Price JL, Rubin EH,
et al. Mild cognitive impairment represents early-stage Alzheimer
disease. Arch Neurol 2001;58:397– 405.
[14] Patterson CJ, Gauthier S, Bergman H, Cohen CA, Feightner JW,
Feldman H, et al. The recognition, assessment and management of
dementing disorders: conclusions from the Canadian Consensus
Conference on Dementia. CMAJ 1999;160:S1–15.
[15] American Psychiatric Association, Committee on Nomenclature and
Statistics. Diagnostic and statistical manual of mental disorders, 3rd
ed, revised. Washington, DC: American Psychiatric Association;
1987.
[16] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders. 4th ed. Washington, DC: American Psychiatric
Association; 1994.
[17] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders (IV-TR). Washington, DC: American Psychiat-
ric Association; 2000.
[18] Lopez OL, Becker JT, Kaufer DI, Hamilton RL, Sweet RA, Klunk
W, et al. Research evaluation and diagnosis of probable Alzheimer’s
disease over the last two decades: pt 1. Neurology 2000;55:1854 –
62.
[19] Feldman H, Levy AR, Hsiung GY, Peters KR, Donald A, Black SE,
et al. A Canadian cohort study of cognitive impairment and related
dementias (ACCORD): study methods and baseline results. Neuro-
epidemiology 2003;22:265–74.
[20] King P, Davichand P, Rockwood K. Dementia of acute onset in the
Canadian Study of Health and Aging. Int Psychogeriatr 2005;17:
451–9.
[21] King P, Wang Y, Rockwood K. Cognitive impairment of acute
onset. Am J Geriatr Psychiatry 2006;14:893– 6.
[22] McKhann G, Drachman D, Folstein M, Katzman R, Price D,
Stadlan EM. Clinical diagnosis of Alzheimer’s disease: Rrport of
the NINCDS-ADRDA Work Group under the auspices of De-
partment of Health and Human Services Task Force on Alzhei-
mer’s Disease. Neurology 1984;34:939 – 44.
[23] Renner JA, Burns JM, Hou CE, McKeel DW Jr, Storandt M, Morris
JC. Progressive posterior cortical dysfunction: a clinicopathologic
series. Neurology 2004;63:1175– 80.
[24] Tang-Wai DF, Graff-Radford NR, Boeve BF, Dickson DW, Parisi
JE, Crook R, et al. Clinical, genetic and neuropathological charac-
teristics of posterior cortical atrophy. Neurology 2004;63:1168 –74.
[25] Visser PJ, Verhey FR, Ponds RW, Kester A, Jolles J. Distinction
between preclinical Alzheimer’s disease and depression. J Am Geri-
atr Soc 2000;48:479 – 84.
[26] Hachinski VC, Iliff LD, Zilhka E, Du Boulay GH, McAllister VL,
Marshall J, et al. Cerebral blood flow in dementia. Arch Neurol
1975;32:632–7.
[27] Gustafson L. What is dementia? Acta Neurol Scand Suppl 1996;
168:22– 4.
[28] Graeber M, Muller U. Dementia with Lewy bodies: disease concept
and genetics. Neurogenetics 2003;4:157– 62.
[29] Bowler JV, Munoz DG, Merskey H, Hachinski V. Factors affecting
the age of onset and rate of progression of Alzheimer’s disease.
J Neurol Neurosurg Psychiatry 1998;65:184 –90.
[30] Perrault A, Wolfson C, Egan M, Rockwood K, Hogan DB. Prog-
nostic factors for functional independence in older adults with mild
dementia: results from the Canadian Study of Health and Aging.
Alzheimer Dis Assoc Disord 2002;16:239 – 47.
437
[31] Holmes C, Lovestone S. Long-term cognitive and functional decline
in late onset Alzheimer’s disease: therapeutic implications. Age
Ageing 2003;32:200 – 4.
[32] Galasko DR, Gould RL, Abramson IS, Salmon DP. Measuring
cognitive change in a cohort of patients with Alzheimer’s disease.
Stat Med 2000;19:1421–32.
[33] Ashford JW, Schmitt FA. Modeling the time-course of Alzheimer
dementia. Curr Psychiatry Rep 2001;3:20 – 8.
[34] Soto ME, Gillette-Guyonnet S, Vellas B; The REAL.FR group.
Rapid cognitive decline: searching for a definition and predictive
factors among elderly with Alzheimer’s disease. J Nutr Health
Aging 2005;9:158 – 61.
[35] Fisk JD, Rockwood K. Outcomes of incident mild cognitive impair-
ment in relation to case definition. J Neurol Neurosurg Psychiatry
2005;76:1175–7.
[36] Katzman R, Terry R, DeTeresa R, Brown T, Davies P, Fuld P, et al.
Clinical, pathological, and neurochemical changes in dementia: A
subgroup with preserved mental status and numerous neocortical
plaques. Ann Neurol 1988;23:138 – 44.
[37] Crystal H, Dickson D, Fuld P, Masur D, Scott R, Mehler M, et al.
Clinico-pathologic studies in dementia: nondemented subjects with
pathologically confirmed Alzheimer’s disease. Neurology 1988;38:
1682– 87.
[38] Neuropathology Group. Medical Research Council Cognitive Func-
tion and Aging Study: athological correlates of late-onset dementia
in a multicentre, community-based population in England and
Wales—Neuropathology Group of the Medical Research Council
Cognitive Function and Ageing Study (MRC CFAS). Lancet 2001;
357:169 –75.
[39] Gold G, Bouras C, Kövari E, Canuto A, Glaría BG, Malky A, et al.
Clinical validity of Braak neuropathological staging in the oldest-
old. Acta Neuropathol (Berl) 2000;99(5):579 – 82.
[40] Nagy Z, Esiri MM, Hindley NJ, Joachim C, Morris JH, King EM,
et al. Accuracy of clinical operational diagnostic criteria for Alz-
heimer’s disease in relation to different pathological diagnostic
protocols. Dement Geriatr Cogn Disord 1998;9:219 –26.
[41] Blacker D, Albert MS, Bassett SS, Go RC, Harrell LE, Folstein MF.
Reliability and validity of NINCDS-ADRDA criteria for Alzhei-
mer’s disease: the National Institute of Mental Health Genetics
Initiative. Arch Neurol 1994;51:1198 –204.
[42] Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S,
et al. Frontotemporal lobar degeneration: a consensus on clinical
diagnostic criteria. Neurology 1998;51:1546 –54.
[43] Rockwood K, Howard K, Thomas VS, Mallery L, MacKnight C,
Sangalang V, et al. Retrospective diagnosis of dementia using an
informant interview based on the Brief Cognitive Rating Scale. Int
Psychogeriatr 1998;10:53– 60.
[44] Holmes C, Cairns N, Lantos P, Mann A. Validity of current clinical
criteria for Alzheimer’s disease, vascular dementia and dementia
with Lewy bodies. Br J Psychiatry 1999;174:45–50.
[45] Knopman DS, DeKosky ST, Cummings JL, Chui H, Corey-Bloom
J, Relkin N, et al. Practice parameter: diagnosis of dementia (an
evidence-based review)—report of the Quality Standards Subcom-
mittee of the American Academy of Neurology, Neurology 2001;
56:1143–53.
[46] Bowler JV, Munoz DG, Merskey H, Hachinski V. Fallacies in the
pathological confirmation of the diagnosis of Alzheimer’s disease.
J Neurol Neurosurg Psychiatry 1998;64:18 –24.
[47] McKhann GM, Albert MS, Grossman M, Miller B, Dickson D,
Trojanowski JQ, et al. Clinical and pathological diagnosis of fron-
totemporal dementia: report of the Work Group on Frontotemporal
Dementia and Pick’s Disease. Arch Neurol 2001;58:1803–9.
[48] Mesulam MM. Primary progressive aphasia: differentiation from
Alzheimer’s disease. Ann Neurol 1987;22:533– 4.
438 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
[49] Seeley WW, Bauer AM, Miller BL, Gorno-Tempini ML, Kramer
JH, Weiner M, et al. The natural history of temporal variant fron-
totemporal dementia. Neurology 2005;64:1384 –90.
[50] Rosen HJ, Gorno-Tempini ML, Goldman WP, Perry RJ, Schuff N,
Weiner M, et al. Patterns of brain atrophy in frontotemporal demen-
tia and semantic dementia. Neurology 2002;58:198 –208.
[51] Kertesz A, Munoz DG. Primary progressive aphasia and Pick com-
plex. J Neurol Sci 2003;206:97–107.
[52] The Lund Manchester Groups. Clinical and neuropathological crir-
teria for frontotemporal dementia. J Neurol Neurosurg Psychiatry
1994;57:416 – 8.
[53] Miller BL, Ikonte C, Ponton M, Levy M, Boone K, Darby A, et al.
A study of the Lund-Manchester research criteria for frontotemporal
dementia: clinical and single-photon emission CT correlations. Neu-
rology 1997;46:937– 42.
[54] Pasquier F, Petit H. Frontotemporal dementia: its rediscovery. Eur
Neurol 1997;38:1– 6.
[55] Grosman M. Frontotemporal dementia: a review. J Int Neuropsychol
Soc 2002;8:566 – 83.
[56] Gorno-Tempini ML, Dronkers NF, Rankin KP, Ogar JM, Phen-
grasamy L, Rosen HJ, et al. Cognition and anatomy in three variants
of primary progressive aphasia. Ann Neurol 2004;55:335– 46.
[57] Shea C, MacKnight C, Rockwood K. Donepezil for treatment of
dementia with Lewy bodies: a case series of nine patients. Int
Psychogeriatr 1998;10:229 –38.
[58] McKeith IG, Perry RH, Fairbairn AF, Jabeen S, Perry EK. Opera-
tional criteria for senile dementia of Lewy body type (SDLT).
Psychol Med 1992;22:911–22.
[59] McKeith IG, Galasko D, Kosaka K, Perry EK, Dickson DW, Hansen
LA, et al. Consensus guidelines for the clinical and pathologic diag-
nosis of dementia with Lewy bodies (DLB): report of the consortium on
DLB international workshop. Neurology 1996;47:1113–24.
[60] McKeith IG, Dickson DW, Lowe J, Emre M, O’Brien JT, Feldman
H, et al. Diagnosis and management of dementia with Lewy bodies:
third report of the DLB consortium. Neurology 2005;65:1863–72.
[61] Lopez OL, Litvan I, Catt KE, Stowe R, Klunk W, Kaufer DI, et al.
Accuracy of four diagnostic criteria for the diagnosis of neurode-
generative dementias. Neurology 1999;53:1292–9.
[62] Soret M, Koulibaly PM, Darcourt J, Buvat I. Partial volume effect
correction in SPECT for striatal uptake measurements in patients
with neurodegenerative diseases. impact upon patient classification.
Eur J Nucl Med Mol Imaging 2006 [Epub ahead of print].
[63] O’Brien JT, Colloby S, Fenwick J, Williams ED, Firbank M, Burn
D, et al. Dopamine transporter loss visualized with FP-CIT SPECT
in the differential diagnosis of dementia with Lewy bodies. Arch
Neurol 2004;61:919 –25.
[64] Folstein MF, Folstein SE, McHugh PR.“Mini-mental state”: A prac-
tical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 1975;12:189 –98.
[65] Korczyn AD. The complex nosological concept of vascular demen-
tia. J Neurol Sci 2002;203-204:3– 6.
[66] Aarsland D, Tandberg E, Larsen JP, Cummings JL. Frequency of
dementia in Parkinson’s disease. Arch Neurol 1996;53:538 – 42.
[67] Aarsland D, Andersen K, Larsen JP, Lolk A, Nielsen H, Kragh-
Sorensen P. Risk of dementia in Parkinson’s disease: a community-
based, prospective study. Neurology 2001;56:730 – 6.
[68] Emre M. Dementia associated with Parkinson’s disease. Lancet
Neurol 2003;2:229 –37.
[69] Boller F, Mizutani T, Roessmann U, Gambetti P. Parkinson disease,
dementia, and Alzheimer’s disease: clinicopathological correlations.
Ann Neurol 1980;7:329 –35.
[70] Aarsland D, Perry R, Brown A, Larsen JP, Ballard C. Neuropathol-
ogy of dementia in Parkinson’s disease: a prospective, community-
based study. Ann Neurol 2005;58:773– 6.
[71] Erisi MM, Wilcock GK, Morris JH. Neuropathological assessment
of the lesions of significance in vascular dementia. J Neurol Neu-
rosurg Psychiatry 1997;63:749 –53.
[72] Tomlinson BE, Blessed G, Roth M. Observations on the brain of
demented old people. J Neurol Sci 1970;11:205– 42.
[73] Hansen LA, Samuel W. Criteria for Alzheimer’s disease and the
nosology of dementia with Lewy bodies. Neurology 1997;48:
126 –32.
[74] Hughes AJ, Daniel SE, Blankson S, Lees AJ. A clinicopathologic
study of 100 cases of Parkinson’s disease. Arch Neurol 1993;50:
140 – 8.
[75] Merdes AR, Hansen LA, Jeste DV, Galasko D, Hofstetter CR, Ho
GJ, et al. Influence of Alzheimer pathology on clinical diagnostic
accuracy in dementia with Lewy bodies. Neurology 2003;60:1586 –
90.
[76] Srikanth S, Nagaraja AV, Ratnavalli E. Neuropsychiatric symptoms
in dementia-frequency, relationship to dementia severity and com-
parison in Alzheimer’s disease, vascular dementia and frontotem-
poral dementia. J Neurol Sci 2005;236:43– 8.
[77] Emre M, Aarsland D, Albanese A, Byrne EJ, Deuschl G, De Deyn
PP, et al. Rivastigmine for dementia associated with Parkinson’s
disease. N Engl J Med. 2004;351:2509 –18.
[78] McKeith I, Del Ser T, Spano P, Emre M, Wesnes K, Anand R, et al.
Efficacy of rivastigmine in dementia with Lewy bodies: a random-
ised, double-blind, placebo-controlled international study. Lancet
2000;16;356:2031– 6.
[79] O’Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni
L, et al. Vascular cognitive impairment. Lancet Neurol 2003;2:89 –
98.
[80] Roman G, Tatemichi TK, Erkinjuntti T, the NINDS-AIREN Work-
group. Vascular dementia: diagnostic criteria for research studies—
report of the NINDS-AIREN International workshop. Neurology
1993;43:250 – 60.
[81] Chui HC, Victoroff JI, Margolin D, Jagust W, Shankle R, Katzman
R. Criteria for the diagnosis of ischemic vascular dementia proposed
by the State of California Alzheimer’s Disease Diagnostic and
Treatment Centers. Neurology 1992;42:473– 80.
[82] Rockwood K, Davis H, MacKnight C, Vandorpe R, Gauthier S,
Guzman A. The Consortium to Investigate Vascular Impairment of
Cognition: methods and first findings. Can J Neurol Sci 2003;30:
237– 43.
[83] Lopez OL, Kuller LH, Becker JT, Jagust WJ, DeKosky ST, Fitz-
patrick A, et al. Classification of vascular dementia in the Cardiovas-
cular Health Study Cognition Study. Neurology 2005;64:1539 – 47.
[84] Verhey FR, Lodder J, Rozendaal N, Jolles J. Comparison of seven
sets of criteria used for the diagnosis of vascular dementia. Neuro-
epidemiology 1996;15:166 –72.
[85] Pohjasvaara T, Mantyla R, Ylikoski R, Kaste M, Erkinjuntti T.
Comparison of different clinical criteria (DSM-III, ADDTC, ICD-10,
NINDS-AIREN, DSM-IV) for the diagnosis of vascular dementia:
National Institute of Neurological Disorders and Stroke-Association
Internationale pour la Recherche et l’Enseignement en Neurosciences.
Stroke 2000;31:2952–7.
[86] Gold G, Bouras C, Canuto A, Bergallo MF, Herrmann FR, Hof PR,
et al. Clinicopathological validation study of four sets of clinical
criteria for vascular dementia. Am J Psychiatry 2002;159:82–7.
[87] Hulette C, Nochlin D, McKeel D, Morris JC, Mirra SS, Sumi SM,
et al. Clinical-neuropathologic findings in multi-infarct dementia: a
report of six autopsied cases. Neurology 1997;48:668 –72.
[88] Knopman DS, Rocca WA, Cha RH, Edland SD, Kokmen E. Inci-
dence of vascular dementia in Rochester, Minn, 1985–1989. Arch
Neurol 2002;59:1605–10.
[89] Rockwood K, Black SE, Song X, Hogan DB, Gauthier S, MacKnight
C, et al. Clinical and radiographic subtypes of vascular cognitive
 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440
impairment in a clinic-based cohort study. J Neurol Sci
2006;15:7–14.
[90] Erkinjuntti T, Inzitari D, Pantoni L, Wallin A, Scheltens P, Rock-
wood K, et al. Research criteria for subcortical vascular dementia in
clinical trials. J Neural Transm Suppl 2000;59:23–30.
[91] Prins ND, van Dijk EJ, den Heijer T, Vermeer SE, Jolles J, Koud-
staal PJ, et al. Cerebral small-vessel disease and decline in infor-
mation processing speed, executive function and memory. Brain
2005;128(Pt 9):2034 – 41.
[92] Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ,
Breteler MM. Silent brain infarcts and the risk of dementia and
cognitive decline. N Engl J Med 2003;348:1215–22.
[93] Hachinski V. Preventable senility: a call for action against the
vascular dementias. Lancet. 1992;340:645– 8.
[94] Hebb AO, Cusimano MD. Idiopathic normal pressure hydrocepha-
lus: a systematic review of diagnosis and outcome. Neurosurgery
2001;49:1166 – 86.
[95] Relkin N, Marmarou A, Klinge P, Black PM. Diagnosing idiopathic
normal pressure hydrocephalus. Neurosurgery 2005;57(Suppl):S4 –16.
[96] Vanneste J, Augustijn P, Tan WF, Dirven C. Shunting normal
pressure hydrocephalus: the predictive value of combined clinical
and CT data. J Neurol Neurosurg Psychiatry 1993;56:251– 6.
[97] Gallia GL, Rigamonti D, Williams MA. The diagnosis and treatment
of idiopathic normal pressure hydrocephalus. Nat Clin Pract Neurol
2006;2:375– 81.
[98] Poca MA, Mataró M, Matarín M, Arikan F, Junqué C, Sahuquillo J.
Good outcome in patients with normal-pressure hydrocephalus and
factors indicating poor prognosis. J Neurosurg 2005;103:455– 63.
[99] McArthur JC, Brew BJ, Nath A. Neurological complications of HIV
infection. Lancet Neurol 2005;4:543–55.
[100] Timmermans M, Carr J. Neurosyphilis in the modern era. J Neurol
Neurosurg Psychiatry 2004;75:1727–30.
[101] Rabinovici GD, Wang PN, Levin J, Cook L, Pravdin M, Davis J,
et al. First symptom in sporadic Creutzfeldt-Jakob disease. Neurol-
ogy 2006;66:286 –7.
[102] Masters CL, Harris JO, Gajdusek DC, Gibbs CJ Jr, Bernoulli C,
Asher DM. Creutzfeldt–Jakob disease: patterns of worldwide occur-
rence and the significance of familial and sporadic clustering. Ann
Neurol 1979;5:177– 88.
[103] WHO Global Surveillance. Diagnosis and therapy of human trans-
missible spongiform encephalopathies: report of a WHO consulta-
tion. In: World Health Organization: emerging and other communi-
cable diseases, surveillance and control. Geneva, Switzerland:
World Health Organization; 1998.
[104] Geschwind MD, Martindale J, Miller D, DeArmond SJ, Uyehara-
Lock J, Gaskin D, et al. Challenging the clinical utility of the 14-33
protein for the diagnosis of sporadic Creutzfeldt-Jakob disease. Arch
Neurol 2003;60:813– 6.
[105] Tschampa HJ, Kallenberg K, Urbach H, Meissner B, Nicolay C,
Kretzschmar HA, et al. MRI in the diagnosis of sporadic
Creutzfeldt-Jakob disease: a study on inter-observer agreement.
Brain 2005;128:2026 –33.
[106] Brown P, Rodgers-Johnson P, Cathala F, Gibbs CJ Jr, Gajdusek DC.
Creutzfeldt-Jakob disease of long duration: clinicopathological char-
acteristics, transmissibility, and differential diagnosis. Ann Neurol
1984;16:295–304.
[107] Will RG, Zeidler M, Stewart GE, Macleod MA, Ironside JW,
Cousens SN, et al. Diagnosis of new variant Creutzfeldt-Jakob
disease. Ann Neurol 2000;47:575– 82.
[108] Boesenberg C, Schulz-Schaeffer WJ, Meissner B, Kallenberg K,
Bartl M, Heinemann U, et al. Clinical course in young patients with
sporadic Creutzfeldt-Jakob disease. Ann Neurol 2005;58:533– 43.
[109] Young GS, Geschwind MD, Fischbein NJ, Martindale JL, Henry
RG, Liu S, et al. Diffusion-weighted and fluid-attenuated inver-
sion recovery imaging in Creutzfeldt-Jakob disease: high sensi-
439
tivity and specificity for diagnosis. Am J Neuroradiol 2005;
26:1551– 62.
[110] Steinhoff BJ, Zerr I, Glatting M, Schulz-Schaeffer W, Poser S,
Kretzschmar HA. Diagnostic value of periodic complexes in
Creutzfeldt-Jakob disease. Ann Neurol 2004;56:702– 8.
[111] Dong Y, Hassan A, Zhang Z, Huber D, Dalageorgou C, Markus HS.
Yield of screening for CADASIL mutations in lacunar stroke and
leukoaraiosis. Stroke 2003;34:203–5.
[112] Greenberg SM, Vonsattel JP, Stakes JW, Gruber M, Finklestein SP.
The clinical spectrum of cerebral amyloid angiopathy: presentations
without lobar hemorrhage. Neurology 1993;43:2073–9.
[113] Harkness KA, Coles A, Pohl U, Xuereb JH, Baron JC, Lennox GG.
Rapidly reversible dementia in cerebral amyloid inflammatory vas-
culopathy. Eur J Neurol 2004;11:59 – 62.
[114] McCusker J, Cole M, Dendukuri N, Belzile E, Primeau F. Delirium
in older medical inpatients and subsequent cognitive and functional
status: a prospective study. CMAJ 2001;165:575– 83.
[115] Rockwood K, Cosway S, Carver D, Jarrett P, Stadnyk K, Fisk J.
The risk of dementia and death after delirium. Age Ageing
1999;28:551– 6.
[116] Tschampa HJ, Neumann M, Zerr I, Henkel K, Schröter A, Schulz-
Schaeffer WJ, et al. Patients with Alzheimer’s disease and dementia
with Lewy bodies mistaken for Creutzfeldt-Jakob disease. J Neurol
Neurosurg Psychiatry 2001;71:33–9.
[117] Reinwald S, Westner IM, Niedermaier N. Rapidly progressive Alz-
heimer’s disease mimicking Creutzfeldt-Jakob disease. J Neurol
2004;251:1020 –2.
[118] Catani M, Piccirilli M, Geloso MC, Cherubini A, Finali G, Pellic-
cioli G, et al. Rapidly progressive aphasic dementia with motor
neuron disease: a distinctive clinical entity. Dement Geriatr Cogn
Disord 2004;17:21– 8.
[119] Ballard CG, Aarsland D, McKeith I, O’Brien J, Gray A, Cormack F,
et al. Fluctuations in attention: PD dementia vs DLB with parkin-
sonism. Neurology 2002;59:1714 –20.
[120] Caselli RJ, Boeve BF, Scheithauer BW, O’Duffy JD, Hunder GG.
Nonvasculitic autoimmune inflammatory meningoencephalitis
(NAIM): a reversible form of encephalopathy. Neurology 1999;53:
1579 – 81.
[121] Gómez-Puerta JA, Cervera R, Calvo LM, Gómez-Ansón B, Espi-
nosa G, Claver G, et al. Dementia associated with the antiphospho-
lipid syndrome: clinical and radiological characteristics of 30 pa-
tients. Rheumatology (Oxford) 2005;44:95–9.
[122] Tolnay M, Clavaguera F. Argyrophilic grain disease: a late-onset
dementia with distinctive features among tauopathies. Neuropathol-
ogy 2004;24:269 – 83.
[123] Cairns NJ, Grossman M, Arnold SE, Burn DJ, Jaros E, Perry RH,
et al. Clinical and neuropathologic variation in neuronal intermedi-
ate filament inclusion disease. Neurology 2004;63:1376 – 84.
[124] Blass DM, Hatanpaa KJ, Brandt J, Rao V, Steinberg M, Troncoso
JC, et al. Dementia in hippocampal sclerosis resembles frontotem-
poral dementia more than Alzheimer disease. Neurology 2004;63:
492–7.
[125] Hatanpaa KJ, Blass DM, Pletnikova O, Crain BJ, Bigio EH, Hedreen
JC, et al. Most cases of dementia with hippocampal sclerosis may
represent frontotemporal dementia. Neurology 2004;63:538 – 42.
[126] Lippa CF, Dickson DW. Hippocampal sclerosis dementia: expand-
ing the phenotypes of frontotemporal dementias? Neurology 2004;
63:414 –5.
[127] Louis ED, Lynch T, Kaufmann P, Fahn S, Odel J. Diagnostic
guidelines in central nervous system whipple’s disease. Ann Neurol
1996;40:561– 8.
[128] Houlden H, Baker M, McGowan E, Lewis P, Hutton M, Crook R, et
al. Variant Alzheimer’s disease with spastic paraparesis: clinical
characterization. Neurology 2000;54:1103–9.
440 K. Rockwood et al. / Alzheimer’s & Dementia 3 (2007) 428 – 440

[129] Graves TD, Hanna MG. Neurological channelopathies. Postgrad
Med J 2005;81:20 –32.
[130] Lyon G, Fattal-Valevski A, Kolodny EH. Leukodystrophies: clinical
and genetic aspects. Top Magn Reson Imaging. 2006;17:219 – 42.
[131] Kumar N. Superficial siderosis: associations and therapeutic impli-
cations. Arch Neurol 2007;64:491– 6.
[132] Gold G, Kövari E, Corte G, Herrmann FR, Canuto A, Bussière T, et al.
Clinical validity of A beta-protein deposition staging in brain aging and
Alzheimer disease. J Neuropathol Exp Neurol 2001;60:946 –52.
[133] Song X, Mitnitski A, Rockwood K. Index variables for studying
outcomes in vascular cognitive impairment. Neuroepidemiology
2005;25:196 –204.
 Alzheimer’s & Dementia 3 (2007) 441– 443
Perspectives
Commentary on “The Third Canadian Consensus Conference on the
Diagnosis and Treatment of Dementia:”
Clinical guidelines are not enough—System-wide, population-based
programs are needed to improve the care of patients with Alzheimer’s
disease and related dementias
Howard Fillit*
Alzheimer’s Drug Discovery Foundation and the Institute for the Study of Aging, New York, New York
This volume truly represents a timely, expert, well-
organized synopsis of recent developments in the treatment
and care of patients with Alzheimer’s disease, particularly
from the Canadian perspective, but based on worldwide
data. The purpose of this effort is to create a consensus of
recommendations on relevant issues based on new knowl-
edge. The consensus process is rigorous and clearly defined,
and has a strong base of evidence. The papers and recom-
mendations presented here cover many critical and relevant
areas of recent advances in the field. Relevant clinical topics
range from early detection to treatment and management of
care. Ultimately, 215 recommendations were considered,
and 149 recommendations were approved with strong con-
sensus by the Canadian academic community involved in
clinical research and care. I will not attempt to comment
specifically on these many recommendations that were so
carefully and thoughtfully evaluated by my esteemed col-
leagues, though in general, I certainly agree with them.
Rather, my perspective will address the real value of this
extensive, expensive, and time-consuming work. It is criti-
cally important to note that, quoting Dr. Chertkow in his
paper, “Translating, disseminating, and implementing our
recommendations received particular emphasis.” An imple-
mentation committee was created to address these “knowl-
edge translation” issues, with a specific work plan that
included publications, tool kits, and practice aids for distri-
bution to Canadian family practitioners, course material,
and “assessment of the impact of . . . recommendations on
*Corresponding author.
E-mail address: hfillit@alzdiscovery.org
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.09.003
clinical practice and patient outcomes.” Chertkow states,
“In conclusion, the CCCDTD3 was a time-consuming but
rewarding enterprise for all those who participated. Only
time will tell, though, if we will improve both the care
provided by Canadian physicians and the outcomes of pa-
tients with a dementia.”
We are all happy that the committee had a “rewarding”
experience. But demonstrating the impact of these recom-
mendations on clinical care and outcomes is the most im-
portant, difficult, and complex task. I have not practiced
medicine in Canada, but I have practiced geriatric medicine
in the United States for almost 30 years. I have also had the
opportunity to practice “population-based” medicine in the
United States through my tenure as a corporate medical
director for a national Medicare managed-care organization
that served over 125,000 elderly individuals throughout the
country. From these experiences at bedside and in the ad-
ministrative health system office, I know that while the goal
to improve the quality of care for dementia is certainly
laudable and critical, there are many, many challenges along
the way.
In the trenches, physicians throughout the world are
under stress. In the United States, the average primary-care
office visit is quite brief, generally less than 10 minutes, and
little time is available to provide extensive counseling in our
national health system for the elderly (called Medicare,
which provides extensive and comprehensive health bene-
fits, including prescription drug coverage, to over 95% of
the population). One-hour visits for counseling are not well-
paid. Primary-care physicians, who see the majority of de-
mented persons in the United States (and in most other parts
of the world) [1], are ill-equipped to provide cognitive
442 H. Fillit / Alzheimer’s & Dementia 3 (2007) 441– 443
assessments, and have little experience with dementia pa-
tients, dementia care management strategies, and the use of
antidementia treatments. From a “back of the envelope”
perspective, in a typical internal-medicine or family practice
(as opposed to an academic neurology clinic), about 50% of
the patients are elderly. If 5% of these patients over 65 have
dementia, that amounts to 2.5% of the total population that
the physician sees. If the physician has a practice of 1,500
active patients, that amounts to 750 elderly, or about 20
dementia patients in the entire practice. If each dementia
patient was seen four times a year, the physician would have
a total of 80 dementia visits a year, or a little more than one
patient with dementia per week: not really quite enough to
get the doctor’s attention, or focus on issues such as care-
management strategies and quality-of-care measures.
The same arithmetic holds true for health systems and
society, but here we must address not only clinical and
quality issues, but also cost issues. To get the attention of
the national systems which ultimately pay for care and have
a societal perspective in the distribution of finite resources
and effort with multiple diseases competing for attention,
and to get these systems to invest in improving the baseline
of care for a particular condition, we must be able to dem-
onstrate to these organizations that the dementias are prev-
alent and costly, and that something proven effective can be
done to improve care. Getting the attention of health sys-
tems throughout the world has been difficult with regard to
dementia for a variety of reasons, including under-reporting,
undercoding, underdiagnosis [2]. Also, payers often lack an
understanding of the impact of dementias on individuals and
society. They do not see the dementias as traditional med-
ical illnesses, and the direct relationship between “losing
your keys” and costly hospitalizations and other direct med-
ical costs is not always self-evident [3,4]. Thus, at least in
the United States, Medicare certainly “knows” about the
dementias, and as much as we in the field think it is so
important, compared with congestive heart failure and dia-
betes, Alzheimer’s disease is not really on the radar of
Medicare.
Studies show that changing the way that care is provided
for any disease by the distribution or publication of educa-
tional materials and consensus opinions is not enough [5].
Guidelines for the elderly present many challenges, espe-
cially to busy physicians [6]. System-wide, programmatic
interventions that change the way the system of care oper-
ates with regard to the particular disease state are needed
[7–9]. One approach has involved screening, collaborative
care, and care-management, disease-management interven-
tions [10 –17]. Another involves systematic approaches to
quality improvement [18], such as Medicare’s QIC pro-
grams (http://www.ahqa.org/pub/media/159_766_2687.cfm).
Few of these, to my knowledge, include programs related to
dementia. A third recent approach is pay-for-performance
[19], which is being explored and implemented in the
United States, the United Kingdom, and elsewhere. Thus,
the Centers for Medicare and Medicaid Services in the
United States have adopted many principles of managed
care into their fee-for-service program because the lessons
learned from managed care have been valuable as an ex-
perimental laboratory to improve our health system with
regard to specific diseases in a population-based manner.
I emphasize this because the individual physician alone
cannot be the target for improving the quality of care for
dementia patients. Individual physicians, or even groups of
physicians, do not have the resources, incentives, position in
the health system, or capacity to change the system of care.
The distribution of guidelines, recommendations, practice
aids, and “kits” to physicians has relatively little impact if
not linked to an interventional program that can change
physician or patient behavior by adding value and resources
to the health system itself [20 –22]. These programs often
offer additional system-wide services to physicians that
can enable and ensure quality of care and adherence to
guideline-recommended care. To implement effectively the
149 recommendations for dementia presented in this vol-
ume in clinical practice is a great challenge, particularly
because many of them do not have convincing “grade A”
recommendations that would warrant the financial and re-
source investment in the programs, systems, and services
necessary to ensure their adoption and adherence.
From the recommendations of the Canadians, as well as
of others, we probably know what we need to do to improve
care for Alzheimer’s disease [8,9,23–25], and we have made
some progress in changing the system. We have developed
telephonic approaches to cognitive assessment [26,27] and
Alzheimer’s disease management programs [11–16]. In the
United States and elsewhere, nonprofit organizations such
as the Alzheimer’s Association (http://www.alz.org/we_can_
help_24_7_helpline.asp) and the Alzheimer’s Foundation of
America (AFA) provide system-wide telephonic services
directly to patients to promote early detection, support groups,
and telephonic care management. The AFA also conducts a
community-based memory screening program. Some dem-
onstration programs and research reports demonstrated that
Alzheimer’s disease care management can have an impact
on outcomes of importance and relevance (http://www.
nationalmemoryscreening.org/). The United States Veteran’s
Administration health system, one of the largest health systems
in the world, has also adopted certain aspects of assessment
and care management for Alzheimer’s disease (http://www1.
va.gov/geriatricsshg/docs/VADementiaActUpdate1_05.doc).
Why is all this important in terms of this volume of
papers? To really change clinical practice and promote
quality of care requires more than just the publication of
well-written academic guidelines and the consensus opin-
ions of experts. Apart from academic specialists, the busy
practicing physicians in the community, their patients, and
the medical managers of health systems will not have the
time, focus, interest, incentives, or mechanisms to read this
entire volume, let alone consider, remember, and utilize the
 H. Fillit / Alzheimer’s & Dementia 3 (2007) 441– 443
149 recommendations during their business day and imple-
ment them. Health systems, including national health sys-
tems, managed-care organizations, and other payers, will
not adopt these guidelines, promote system-wide changes,
and implement programs without clear demonstrations that
there will be societal or population benefits. This is good
thing, because before we commit large sums of money and
resources to improve the health of populations with Alzhei-
mer’s disease and related dementias, we need to be sure
about what we are doing, and how we will do it.
Therefore I applaud the extensive efforts of my esteemed
Canadian neighbors to begin the process of improving the
population-wide health of patients with dementias in North
America. The Canadians have a unique and effective
national health system that should be able to implement
system-wide changes that, together with greater knowledge
on the part of physicians, can make a difference, if there is
the will and the money. Their efforts could serve as a model
for health systems throughout the world, both public and
private, to make changes that would have real effects on the
quality, efficiency, effectiveness, and cost of care for Alz-
heimer’s disease and related dementias. They have set the
challenge for themselves, but to paraphrase Dr. Chertkow
again, only time will tell if the appropriate commitment of
resources and innovations will systematically improve both
the care provided by Canadian physicians and the outcomes
of patients with a dementia in Canada.
References
[1] Wilkinson D, Stave C, Keohane D, Vincenzino O. The role of general
practitioners in the diagnosis and treatment of Alzheimer’s disease: a
multinational survey. J Int Med Res 2004;32:149 –59.
[2] Fillit H, Geldmacher DS, Welter RT, Maslow K, Fraser M. Optimiz-
ing coding and reimbursement to improve management of Alzhei-
mer’s disease and related dementias. J Am Geriatr Soc 2002;50:
1871– 8.
[3] Wilkinson D. Is there a double standard when it comes to dementia
care? Int J Clin Pract 2005;146(Suppl).3–7.
[4] Fillit H, Hill JW, Futterman R. Health care utilization and costs of
Alzheimer’s disease: the role of co-morbid conditions, disease stage,
and pharmacotherapy. Fam Med 2002;34:528 –35.
[5] Cochrane LJ, Olson CA, Murray S, Dupuis M, Tooman T, Hayes S.
Gaps between knowing and doing: understanding and assessing the
barriers to optimal health care. J Contin Educ Health Prof 2007;27:
94 –102.
[6] Tinetti ME, Bogardus ST Jr, Agostini JV. Potential pitfalls of disease-
specific guidelines for patients with multiple conditions. N Engl
J Med 2004;351:2870 – 4.
[7] Epping-Jordan JE, Pruitt SD, Bengoa R, Wagner EH. Improving the
quality of health care for chronic conditions. Qual Saf Health Care
2004;13:299 –305.
[8] Fillit H, Knopman D, Cummings J, Appel F. Opportunities for im-
proving managed care for individuals with dementia: part 2—a
framework for care. Am J Manag Care 1999;5:317–24.
443
[9] Fillit H, Knopman D, Cummings J, Appel F. Opportunities for im-
proving managed care for individuals with dementia: part 1—the
issues. Am J Manag Care 1999;5:309 –15.
[10] Boustani M, Callahan CM, Unverzagt FW, et al. Implementing a
screening and diagnosis program for dementia in primary care. J Gen
Intern Med 2005;20:572–7.
[11] Charlson M, Charlson RE, Briggs W, Hollenberg J. Can disease
management target patients most likely to generate high costs? The
impact of comorbidity. J Gen Intern Med 2007;22:464 –9.
[12] Dancha L, Martini EM. Using data in disease management. Oppor-
tunities and challenges. Med Group Manage J 1998;45:20 – 8.
[13] Iraqi A. Disease management intervention on quality and outcomes of
dementia care. Ann Intern Med 2007;147:69.
[14] Cherry DL, Vickrey BG, Schwankovsky L, Heck E, Plauchm M, Yep
R. Interventions to improve quality of care: the Kaiser Permanente-
Alzheimer’s Association Dementia Care Project. Am J Manag Care
2004;10:553– 60.
[15] Clark PA, Bass DM, Looman WJ, McCarthy CA, Eckert S. Outcomes
for patients with dementia from the Cleveland Alzheimer’s Managed
Care Demonstration. Aging Ment Health 2004;8:40 –51.
[16] Mittelman MS, Ferris SH, Shulman E, Steinberg G, Levin B. A
family intervention to delay nursing home placement of patients with
Alzheimer disease. A randomized controlled trial. JAMA 1996;276:
1725–31.
[17] Callahan CM, Boustani MA, Unverzagt FW, et al. Effectiveness of
collaborative care for older adults with Alzheimer disease in primary
care: a randomized controlled trial. JAMA 2006;295:2148 –57.
[18] Diamond LH. Local implementation of clinical practice guidelines
and continuous quality improvement: challenges and opportunities.
Semin Dial 2000;13:364 – 8.
[19] Boyd CM, Darer J, Boult C, Fried LP, Boult L, Wu AW. Clinical
practice guidelines and quality of care for older patients with multiple
comorbid diseases: implications for pay for performance. JAMA
2005;294:716 –24.
[20] Gandhi TK, Sequist TD, Poon EG, Karson AS, Murff H, Fairchild
D,et al. Primary care clinician attitudes towards electronic clinical
reminders and clinical practice guidelines. AMIA Annu Symp Proc
2003;848.
[21] Gerhardt WE, Schoettker PJ, Donovan EF, Kotagal UR, Muething
SE. Putting evidence-based clinical practice guidelines into practice:
an academic pediatric center’s experience. Jt Comm J Qual Patient
Saf 2007;33:226 –35.
[22] Green CJ, Fortin P, Maclure M, Macgregor A, Robinson S. Informa-
tion system support as a critical success factor for chronic disease
management: necessary but not sufficient. Int J Med Inform 2006;75:
818 –28.
[23] Cummings JL, Frank JC, Cherry D, et al. Guidelines for managing
Alzheimer’s disease: part I. Assessment. Am Fam Physician 2002;
65:2263–72.
[24] Cummings JL, Frank JC, Cherry D, et al. Guidelines for managing
Alzheimer’s disease: part II. Treatment. Am Fam Physician 2002;65:
2525–34.
[25] Doody RS, Stevens JC, Beck C, et al. Practice parameter: manage-
ment of dementia (an evidence-based review). Report of the Quality
Standards Subcommittee of the American Academy of Neurology.
Neurology 2001;56:1154 – 66.
[26] Hill J, McVay JM, Walter-Ginzburg A, et al. Validation of a brief
screen for cognitive impairment (BSCI) administered by telephone
for use in the Medicare population. Dis Manag 2005;8:223–34.
[27] de Jager CA, Budge MM, Clarke R. Utility of TICS-M for the
assessment of cognitive function in older adults. Int J Geriatr Psy-
chiatry 2003;18:318 –24.
 Alzheimer’s & Dementia 3 (2007) 444 – 445

Commentary on “The Third Canadian Consensus Conference on the

Diagnosis and Treatment of Dementia”
Bruno Vellasa,*, Emma Reynisha, Philippe Robertb
a
Alzheimer’s Disease Research and Clinical Center, INSERM U 558, Gerontopole, Toulouse, France
b
Memory and Alzheimer’s Disease Center, CM2R CHU, University of Nice, Nice, France

Abstract A comprehensive and useful initiative has been performed by our Canadian colleagues. The subjects
presented are under active discussion in many countries, with the goal of direct applications in daily
clinical practice. The review provided by this special issue of the Journal will therefore be very
useful for all. In this commentary we will not discuss all papers, but will underline only some
important points concerning mild cognitive impairment and the diagnosis and management of
Alzheimer’s disease.
© 2007 The Alzheimer’s Association. All rights reserved.
Keywords: ●●●

1. Mild cognitive impairment loss. To increase our knowledge, research on cognitive

In the future, the place for mild cognitive impairment
(MCI) will be found between the predementia stage of the
disease [1] and the population targeted in the primary pre-
vention of Alzheimer’s disease (AD). The biological and
pathophysiological process of Alzheimer’s disease starts
decades before clinical symptoms appear, and large preven-
tive trials are presently planned or underway for patients in
the pre-MCI stage of cognitive decline. However, MCI
patients are commonly seen in our memory clinics and need
specific assessment, follow-up, and intervention. Certainly a
comprehensive multidomain intervention for a healthy life,
including physical and cognitive exercise, nutrition, social
activities, control of vascular and metabolic risk factors, and
control of hearing and visual impairments, seems to make
sense, and will shortly be tested in both the primary and
secondary prevention domains. In this field, the information
indicating that physical training may be protective against
cognitive decline appears robust enough for clinical recom-
mendations. There is evidence that physicians and therapists
should promote activity at an intensity level that is adapted
to a person’s overall physical capacities, as part of a healthy
lifestyle for older individuals with and without memory
*Corresponding author. Tel.: 33 (5) 61 77 76 49; Fax: 33 (5) 61 49 71 09.
E-mail address: brunovellas@aol.com
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.09.001
intervention, physical training, or psychological interven-
tions is needed, and must be performed in different coun-
tries to take into account potential cultural differences.
2. Diagnosis of Alzheimer’ disease
The diagnosis of AD is generally easy to make clinically
in mild to moderate stages of the disease, with its typical
forms of progressive memory loss and decline in the activ-
ities of daily living. The more extensive investigations in-
volving complex neuropsychological testing, neuroimaging,
and cerebrospinal fluid biomarkers should perhaps be used
initially for more atypical cases. However, surrogate mark-
ers, including biomarkers, will be very useful for therapeu-
tic research, as mentioned recently [2].
Recommendations concerning the revision of criteria for
the diagnosis of AD perhaps need further discussion. Rock-
wood et al. in particular indicated the importance of depres-
sion in mild dementia, where it is common. In contrast to an
earlier era, when it was felt important to rule out depression
before making a diagnosis of AD, there is now recognition
of symptom overlap, perhaps reflecting a shared pathophys-
iology. In addition, it is important to underline that apathy is
even more frequent than depressive symptoms in MCI and
mild AD [3,4], and these symptoms are frequently falsely
 B. Vellas et al. / Alzheimer’s & Dementia 3 (2007) 444 – 445
regarded as depressive symptoms, potentially leading to
psychotropic misuse.
3. The management of dementia
The management of dementia, from very early to severe
stages of AD, is still a difficult area. Recommendations and
possible complications are well-documented, but further
information is required about continuing care: how fre-
quently should patients be followed, what assessment of
associated clinical disorders should be made and how
should they be managed, when should treatment be contin-
ued or withdrawn, and what is the most appropriate man-
agement of endstage disease? These issues all require some
thought.
The Plasa Study (see www.clinical trial.gov), a 2-year
randomized, clinical study on a specific care plan versus
usual care, incorporates a complete neuropsychological and
geriatric assessment every 6 months in a specialized mem-
ory clinic among 60 hospitals in France. This periodicity
must not be assumed to be fixed, but is adaptable in frail,
elderly patients, in those with the aggressive form of the
disease, and in those with a milder evolution of AD. Weight
loss and mobility disorders occur often in Alzheimer’s pa-
tients, and must be taken into account in the care plan, as
must behavioral and psychological symptoms of dementia
(BPSDs). Instauration and cessation of treatment are still
difficult questions to resolve in each individual situation.
There is an obvious need to focus major research efforts on
the biological basis, diagnosis, and treatment of early AD. It
is also our duty to take care of patients who are or shortly
will be at the most severe stage of the disease, and who will
probably not be able to benefit from new treatments. Even
if there are no current, widely accepted consensus criteria
for severe AD, it is important to underline, in accordance
with the European Dementia Consensus, that the dementia
syndrome is inadequately defined by criteria which only
include the domains of cognition. Severe dementia encom-
445
passes a range of disabilities in each of three domains, i.e.,
cognition, behavior, and function. In particular, BPSD treat-
ment is a key point in the management of severe AD.
Herrmann and Gauthier, in their recommendation 9, indi-
cate that “approaches that may be useful for severe AD
include behavioural management for depression, and
caregivers/staff education programs for a variety of behav-
iors.” Taking into account that the most problematic BPSDs
are agitation, aggressiveness, oppositional behavior, and
psychotic disturbances, it is also important to dedicate spe-
cific guidelines for their management [5].
Finally, here as well as in the guidelines published in the
last decade, it is indicated that for BPSDs, nonpharmaco-
logical treatment should be initiated first. Protocols involv-
ing such interventions are more complex methodologically
than those for pharmacological trials, but we feel that it is
now time to go beyond good intentions, and to develop
more systematic research in this field. The consensus
reached by our Canadians colleagues is a very good exam-
ple for all of us, and represents an important achievement in
the field of AD.
References
[1] Dubois B, Feldman HH, Jacova C, et al. Research criteria for the
diagnosis of Alzheimer’s disease: revising the NINCDS-ADRDA cri-
teria. Lancet Neurol 2007;6:734 – 46.
[2] Lovestone S. Biomarkers in Alzheimer’s disease. J Nutr Health Aging
2006;10:118 –22.
[3] Robert PH, Berr C, Volteau M, et al. Apathy in patients with mild
cognitive impairment and the risk of developing dementia of Alzhei-
mer’s disease: a one-year follow-up study. Clin Neurol Neurosurg
2006;108:733– 6.
[4] Robert PH, Berr C, Volteau M, et al. Neuropsychological performance
in mild cognitive impairment with and without apathy. Dement Geriatr
Cogn Disord 2006;21:192–7.
[5] Benoit M, Robert PH, Staccini P, et al. One-year longitudinal evalu-
ation of neuropsychiatric symptoms in Alzheimer’s disease. The
REAL.FR Study. J Nutr Health Aging 2005;9:95–9.
 Alzheimer’s & Dementia 3 (2007) 446 – 448

Commentary on “The Third Canadian Consensus Conference on the

Diagnosis and Treatment of Dementia”—An appraisal
Paul S. Aisen*
●●●

Since the second series of consensus documents from
Canadian Alzheimer’s disease (AD) experts was published
in 1999, the field has undergone major changes in the areas
of genetics, risk and prevention, prodromal stages, diagno-
sis, and management. The methodology utilized by this
Canadian group is well-constructed, and has yielded a new
collection of papers that represents a major resource to
students, general practitioners, specialists, and research sci-
entists. It was a large effort, but the payoff is substantial.
Most important, though, will be the efforts to assure optimal
translation of the recommendations into clinical practice.
I offer comments on a few specific areas within this very
broad project.
1. There continue to be many areas of confusion sur-
rounding the interpretation of observational studies
of modifiable risk factors of dementia (Patterson et
al.), such as blood pressure, cholesterol, smoking,
and depression. This probably reflects the very dif-
ficult issue of bias in nonrandomized studies; it is a
major problem with no ready solution.
2. With regard to the genetics of susceptibility to AD
(Hsiung and Sadovick), the apolipoprotein E geno-
type is unquestionably important, but it remains un-
certain whether any other genetic risk factor will
have much impact.
3. The relationship between homocysteine, B vitamins,
and cognitive impairment and dementia (Garcia) is
given a balanced review. Whereas homocysteine el-
evation is linked to both cerebrovascular disease
and AD, there is insufficient evidence to justify
testing for or treating homocysteine in the older
population with or without cognitive dysfunction.
We recently presented the results of a randomized,
controlled trial of high-dose B vitamin supple-
*Corresponding author.
E-mail address: psa@georgetown.edu
1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2007.09.002
mentation in mild-to-moderate AD (conducted by
the Alzheimer’s Disease Cooperative Study), with
overall negative results [1].
4. The discussion of various diagnostic labels for con-
ditions between the extremes of normal cognition
and dementia (Chertkow et al.) is very clear and
useful. Though the consensus group did not favor
Cognitive Impairment, No Dementia (CIND), or
Mild Cognitive Impairment (MCI) as optimal terms,
outside of Canada, and particularly within the realm
of therapeutic trials, MCI prevails. But controversy
continues, ranging from the issue of whether amnes-
tic forms of MCI should simply be referred to as
very mild AD [2], to the optimal memory test to
define the category. Most likely, useful criteria in
primary-care practice will differ from those in ther-
apeutic trials that must serve the purpose of identi-
fying a cohort with a predictable rate of progression.
There is reason for some optimism that this area will
be clarified by the application of biomarker studies.
Thus, an individual with objectively confirmed epi-
sodic memory impairment that by history is progres-
sive, coupled with a marker of amyloid accumula-
tion such as low cerebrospinal fluid abeta [3] or an
increased signal on amyloid positron emission to-
mography imaging [4], is likely to have an early
stage of AD (whether it is referred to as amnestic
MCI or very early AD).
5. Few randomized prevention trials have aimed to reduce
the risk of AD (Patterson et al.). This serves as a reflection
of the cost and time necessary to conduct such trials, as
well as uncertainty regarding the interventions that should
be tested. The results of the interrupted Alzheimer’s Dis-
ease Anti-inflammatory Prevention Trial suggest that non-
steroidal, anti-inflammatory drugs may, if anything, in-
crease the risk of cognitive decline and dementia [5],
and the Women’s Health Initiative indicated an in-
creased risk in women assigned to take hormones
 P.S. Aisen / Alzheimer’s & Dementia 3 (2007) 446 – 448
[6]. Results of the Ginkgo Evaluation of Memory
Trial [7], assessing the preventive effect of ginkgo
biloba, are expected soon.
6. While important controversies continue to surround
reimbursement issues, a consensus has developed
regarding optimal pharmacotherapy for the primary
cognitive symptoms of AD. Hogan et al. provide a
particularly thorough update on treatment, covering
primary and secondary symptoms, and important is-
sues such as driving safety and caregiver support. Cho-
linesterase inhibitors are modestly effective, and are
generally appropriate therapy for mild to moderate AD.
As reviewed by Herman and Gauthier, donepezil and
presumably other cholinesterase inhibitors are also use-
ful in the treatment of severe AD. Memantine, alone or
in combination with a cholinesterase inhibitor, is
modestly effective for moderate to severe AD.
While Hogan et al. conclude that vitamin E is not
recommended in the treatment of AD, this is open to
question. The only randomized, controlled trial of
vitamin E in AD showed benefits [8], and this must
be weighed against evidence from meta-analyses
and randomized trials in other populations that sug-
gest the possibility of risk of congestive heart failure
[9] or death [10]. In the case of vitamin E, recom-
mendations should be individualized (considering,
for example, cardiac risk profiles). Evidence regard-
ing the efficacy of nonpharmacological approaches,
such as cognitive training, is still insufficient to
support recommendations.
7. The review of the literature by Bocti et al. suggests
that, in addition to attention to vascular risk factors,
patients with dementia and cerebrovascular disease
benefit to some degree from galantamine and per-
haps other cholinesterase inhibitors. Evidence for
the benefits of Memantine is weaker.
8. Fisk et al. review the difficult issues surrounding
discussions of the diagnosis of dementia with pa-
tients at various stages of impairment, and with their
families. They reasonably advocate progressive
disclosure during the process of evaluation, and an
individualized approach that considers the needs,
wishes, and expectations of patients, and they
emphasize the need for adequate time and, of
course, empathy and respect.
9. The review of ethical issues surrounding consent-to-
treatment and research participation (Fisk et al.) pro-
vides an excellent summary of the complexity of
capacity determination, the role of proxy decision-
making, and caregiver/family involvement in this
difficult process. Principles of autonomy and benef-
icence must guide the consent process in this most
vulnerable of populations. It is also essential to con-
sider that therapy for dementia, including severe
stages, cannot be optimized without research partic-
447
ipation. To preclude research in consideration of an
incomplete capacity to consent may slow or halt
therapeutic advances.
10. The consensus recommendations on the diagnosis of
dementia (Rockwell et al., and Robillard) recognize
the varied presentations in terms of cognitive do-
mains as well as time course, the limited value of
some ancillary procedures such as electroencepha-
lograms and routine cerebrospinal fluid studies, and
the frequent contributions of multiple etiologies. Neu-
ropsychological assessment (Jacova et al.) should be
used selectively to clarify cognitive impairment in MCI
and dementia. Although a huge number of such tests
were described and demonstrated to be useful, a few
straightforward tests that are brief and require little
training to administer and interpret may be sufficient
in most clinical settings: the delayed paragraph re-
call test, the MMSE, and clock-drawing. In some
cases, referral for comprehensive neuropsychologi-
cal testing may be appropriate. Although dementia
and AD will continue to be clinical diagnoses rely-
ing primarily on historical information provided by
knowledgeable informants, in coming years, there
may be increased attention to the role of amyloid-
related biomarkers, from biochemical assessments in
cerebrospinal fluid to amyloid imaging, in the diag-
nosis of AD.
This methodically researched and carefully written
series of review papers represents a highly useful sum-
mary of the state of the art in AD diagnosis and treat-
ment. All papers in this series offer valuable, practical
guidance to clinicians. Huge progress has been made
since the first series from this group, and the field now
appears to be on the verge of additional major therapeutic
advances. The authors of this series have done a substan-
tial service to primary care and to specialty clinicians in
providing guidance today, and laying the groundwork for
advances to come.
References
[1] Aisen PS, Jin S, Thomas RG, Sano M, Diaz-Arrastia R, Thal L.
ADCS Homocysteine Trial. Alzheimers Dementia 2007;3(Suppl.):
S199.
[2] Morris JC. Mild cognitive impairment is early-stage Alzheimer dis-
ease: time to revise diagnostic criteria. Arch Neurol 2006;63:15– 6.
[3] Fagan AM, Roe CM, Xiong C, Mintun MA, Morris JC, Holtzman
DM. Cerebrospinal fluid tau/beta-amyloid(42) ratio as a prediction of
cognitive decline in nondemented older adults. Arch Neurol 2007;64:
343–9.
[4] Forsberg A, Engler H, Almkvist O, et al. PET imaging of amyloid
deposition in patients with mild cognitive impairment. Neurobiol
Aging 2007.
[5] Group AR, Lyketsos CG, Breitner JC, et al. Naproxen and celecoxib
do not prevent AD in early results from a randomized controlled trial.
Neurology 2007;68:1800 – 8.
448 P.S. Aisen / Alzheimer’s & Dementia 3 (2007) 446 – 448
[6] Shumaker SA, Legault C, Thal L, et al. Estrogen plus progestin and
the incidence of dementia and mild cognitive impairment in
postmenopausal women: the Women’s Health Initiative Mem-
ory Study: a randomized controlled trial. JAMA 2003;289:
2651– 62.
[7] DeKosky ST, Fitzpatrick A, Ives DG, et al. The Ginkgo Evaluation of
Memory (GEM) Study: design and baseline data of a randomized trial
of ginkgo biloba extract in prevention of dementia. Contemp Clin
Trials 2006;27:238 –53.
[8] Sano M, Ernesto C, Thomas RG, et al. A controlled trial of selegiline,
alpha-tocopherol, or both as treatment for Alzheimer’s disease.
N Engl J Med 1997;336:1216 –22.
[9] Lonn E, Bosch J, Yusuf S, et al. Effects of long-term vitamin E
supplementation on cardiovascular events and cancer: a randomized
controlled trial. JAMA 2005;293:1338 – 47.
[10] Miller ER III, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ,
Guallar E. Meta-analysis. high-dosage vitamin E supplementation
may increase all-cause mortality. Ann Intern Med. 2004.
449
Spring Research Roundtable
Spurs Wide-Ranging Discussion
Among Industry and Academic
Thought Leaders
The May 30 –31 meeting of the Alzheimer’s Association
Research Roundtable convened more than 90 members
and invited guests in Washington, D.C., to explore the
Academic/Industry Interface for Alzheimer’s Drug Dis-
covery. Historically, academic laboratories with deep ex-
pertise in basic science led the field during the first wave
of accelerating discovery in the 1970s and 1980s. Those
efforts focused on understanding the basic biology of the
brain and what goes wrong in Alzheimer’s disease. In-
sights gained identified a growing number of potential
points of intervention to treat symptoms or modify under-
lying disease processes. Work leading to development of
tacrine (Cognex), approved in 1993 as the first drug to
treat cognitive symptoms, took place chiefly in academic
labs.
Attracted by the steadily increasing number of potential
therapeutic targets, pharmaceutical companies during the
1990s began to develop major Alzheimer drug discovery
initiatives. Now, there is a rapidly expanding industry-
based clinical research enterprise, with new compounds
entering human clinical testing at an unprecedented rate.
The Research Roundtable Academic/Industry Interface
meeting explored how these two arenas can forge better
working relationships in the current drug discovery cli-
mate to speed movement of potentially valuable discover-
ies from basic academic science laboratories to further
development and clinical testing under the aegis of the
pharmaceutical companies. Although there are some aca-
demic labs that excel in collecting and packaging basic
science data in ways that dovetail well with industry pro-
grams, too often the two arenas operate as parallel uni-
verses filled with mismatched expectations and missed
opportunities.
Key issues raised by industry representatives included
the following:
● Academic researchers tend to overvalue molecules
that emerge very early in the target identification
process. One industry scientist emphasized that “tar-
get identification does not equal druggable target
identification.”
● Once drug discovery progresses to stages governed
by the regulatory framework of good laboratory
practices (GLPs), academic laboratories are seldom
Volume 3 / Issue 4 / October 2007
equipped to meet the code of federal regulations
(CFR) requirements for data acquisition.
● In general, academic researchers tend to be some-
what naïve about the drug development process. Ac-
ademics need education about Food and Drug Ad-
ministration (FDA) policies and procedures, but
industry should not be the sole source of this
education.
Despite the challenges complicating the current rela-
tionship, industry scientists emphasized their desire to
work successfully with academic colleagues. Areas pro-
posed as niches academia might most successfully fill to
complement rather than compete with industry efforts in-
clude the following:
● Develop better animal models. Current models are
good predictors of drug safety, with more than 70
percent of adverse events picked up in animal stud-
ies. But available models fall woefully short in doc-
umenting therapeutic impact and behavioral effects.
● Bring the needs and perspectives of patients and
caregivers to bear on clinical trial design, a function
academic investigators who work as both clinicians
and researchers are ideally positioned to fulfill.
● Continue to identify novel targets and elucidate
basic disease processes. Regardless of the success or
failure of current studies testing the amyloid hypoth-
esis, achieving a mature clinical environment will
require additional therapeutic strategies. Possible ap-
proaches include cytoskeletal stabilizers, novel anti-
inflammatories, metabolic enhancers, and neuropro-
tectants. In an academic environment, there is really
no “bad target.” Even if a molecule turns out to lack
potential as a drug, new data on its behavior may
produce publishable insights into key disease
processes.
Industry scientists cited the following as elements of an
ideal basic science “data package” with the potential to
pique the interest of a pharmaceutical company in further
drug development:
● Complete characterization of the compound. Various
speakers described the “ideal molecule” as smallest
in its class, non-lipophilic, low propensity to form
active intermediates or to induce or inhibit liver
enzymes, close to soluble, and biased toward CNS
penetrance.
450 Alzheimer’s Association Update / Alzheimer’s & Dementia 3 (2007) 449 – 452
● As much risk and safety data as possible, especially
information about structure-activity relationships
(SARs) and detail about the compound’s biology
beyond the immediate indication. One example
cited is the work, primarily from academic labs,
showing the need to modulate gamma-secretase ac-
tivity to avoid the Notch substrate.
● Preclinical studies that involve enough mice (at least
20 –30) and last long enough to provide meaningful
statistical power.
● Animal data should include such neuronal endpoints
as whether synaptic density or expected levels of cell
death changed.
● In addition to traveling with complete and compelling
data, an attractive compound fits well into a compa-
ny’s current portfolio and pipeline, and competes well
with the entire potential universe of other candidates
that might be pursued.
One recurring theme focused on the urgent need for
validated biomarkers. A speaker cited the case of natali-
zumab (Tysrabi), a multiple sclerosis drug whose clinical
data demonstrated the reliability of gadolinium-enhancing
lesions as an MS biomarker. With the advent of this biomar-
ker, which is capable of giving a good read on efficacy over
a 6-month trial, more than 30 companies are now looking at
MS drug discovery programs.
Several speakers urged colleagues to consider pooling
placebo group data from clinical studies in Alzheimer’s
disease and mild cognitive impairment (MCI) as one
strategy to provide an enhanced benchmark for docu-
menting the effects of potentially disease-modifying
drugs. An FDA spokesman cited an academic re-
searcher’s role in taking the initiative to collect all the
placebo data from every monotherapy epilepsy trial every
conducted. Despite its long-standing reluctance to accept
“historical” controls, the FDA decided to accept the
pooled epilepsy placebo data as the comparison group for
any investigational new epilepsy drug, and adopt an of-
ficial position that use of a placebo in an epilepsy trial is
now unethical.
Hosted by the Alzheimer’s Association, the Research
Roundtable currently includes 13 members from the phar-
maceutical, imaging, and diagnostic testing industries.
Topic-oriented meetings held each spring and fall have
become two of the most respected events in Alzheimer
research. Watch for a paper providing further details about
the Academic/Industry meeting in a forthcoming issue of
Alzheimer’s & Dementia: The Journal of the Alzheimer’s
Association.
New members are always welcomed to the roundtable.
For further information about participation, please contact
Jay Thompson, Alzheimer’s Association Corporate Initia-
tives, at jay.thomspon@alz.org.
Dementia Prevention Conference:
Emerging Treatments, Diagnostic
Tools and Risk Management
Strategies Could Transform the
Clinical Landscape
Current treatment options for Alzheimer’s disease are ex-
tremely limited, creating a huge unmet need for the 5.1 million
Americans living with the disease, their families, and the phy-
sicians who care for them. At the Alzheimer’s Association
International Conference on Prevention of Dementia, held June
9⫺12 in Washington, D.C., more than 1,000 attendees learned
how that bleak clinical landscape could change as a result of
accelerating progress toward effective treatments, earlier diag-
nosis, and risk management.
Consumer health surveys consistently show that avoid-
ing cognitive decline is a top concern of older adults. Ex-
tensive media coverage of the Prevention Conference sug-
gests general public interest in this issue is growing. In
2011, 78 million American baby boomers begin turning 65,
reaching the age when risk for Alzheimer’s disease rises
sharply. Expectations and values among this historically
activist group will drive a growing demand for health care
that encompasses cognitive well-being and provides access
to the best available information and interventions.
Based on unmet medical need and a potentially robust
return on investment, many pharmaceutical companies have
embarked on Alzheimer drug discovery programs. Several
drugs that may slow or stop progression of the disease have
reached Phase III clinical trials. Phase III is the final stage
in a three-part human testing sequence mandated by the
Food and Drug Administration (FDA) to demonstrate safety
and effectiveness of a new drug.
The first convincing clinical data showing that Alzhei-
mer’s can be slowed or stopped will spur further drug
development. This proof-of-concept data will also acceler-
ate the search for biological markers that can detect the
disease in its earliest stages and monitor effectiveness of
experimental drugs. The quest for biomarkers is already
under way, with current efforts focusing on brain imaging,
signature proteins in spinal fluid or blood, or characteristic
patterns in protein or gene expression profiles.
The Treatment Horizon
The next-generation treatments most advanced in clinical
testing target beta-amyloid, a protein fragment considered a
prime suspect in Alzheimer’s disease. According to a lead-
ing theory, all individuals produce beta-amyloid, but those
who develop Alzheimer’s generate abnormally large quan-
tities or have reduced ability to clear it from the brain. The
“chemically sticky” fragments build up into the plaques
 Alzheimer’s Association Update / Alzheimer’s & Dementia 3 (2007) 449 – 452
considered one pathological hallmark of the disease, and
trigger a cascade of other abnormalities contributing to
cognitive decline.
Prevention Conference sessions provided updates on the
following amyloid-targeting experimental drugs:
● Alzhemed (tramiprosate) blocks aggregation of beta-
amyloid into plaques. It is the first anti-amyloid drug
to complete a Phase III clinical trial. Based on an
earlier Phase II study suggesting Alzhemed reduced
beta-amyloid levels in cerebrospinal fluid and stabi-
lized patients with the mildest Alzheimer’s disease,
Phase III results were highly anticipated. Conference
attendees were disappointed to learn that no results
were yet available because unusually high data varia-
tions among trial sites invalidated the statistical
model. Alzhemed developer Neurochem is working
with the FDA to develop a new model reducing the
impact of site variations. Possible factors responsible
for the variation include differences in the number of
participants taking memantine, cholinesterase inhibi-
tors, antidepressants or vitamin E. According to Neu-
rochem, trial results could be announced soon.
● Flurizan (tarenflurbil) reduces beta-amyloid by
changing the activity of an enzyme involved in its
production. A conference session reported results of a
12-month follow-on study to a year-long Phase II trial.
After 24 months of treatment, patients in the mildest
Alzheimer stages who received the highest drug dose
showed significant stabilization or improvement in
cognition and overall function. Two large Phase III
studies sponsored by developer Myriad are now under
way testing that highest Phase II dose in early-stage
patients.
● LY450139 blocks activity of the same amyloid-
generating enzyme as Flurizan. The latest Phase II
results showed LY450139 decreased beta-amyloid
levels in both blood and spinal fluid. Developer Lilly
plans to launch a Phase III trial in 2008.
● AN-1792, the first drug in the anti-amyloid class ever
to reach clinical trials, was a “vaccine” that mobilized
the immune system to produce beta-amyloid antibod-
ies. A Phase II trial was halted in 2002 after 6 percent
of recipients developed encephalitis, but AN-1792 de-
velopers Elan and Wyeth continued to monitor partic-
ipants. In follow-up, vaccine recipients who devel-
oped the highest levels of anti-amyloid antibodies
have consistently shown encouraging hints of benefit.
The new four-year data shows trends toward less de-
cline in memory and overall function; greater likeli-
hood of living at home rather than in long-term care;
and no further cases of encephalitis or difference in
survival. Elan and Wyeth announced in May 2007 that
they would move forward in 2008 with a Phase III
trial of AAB-001, a monoclonal antibody against beta-
451
amyloid. Monoclonal antibodies, produced in the lab-
oratory, represent a “passive immunization” approach
that does not trigger an immune response or persist in
the body after drug administration stops.
Prevention Conference sessions also highlighted experi-
mental drugs targeting other mechanisms:
● Dimebon is an antihistamine developed in Russia,
selected for further development as an Alzheimer drug
by the Russian Academy of Sciences after it showed
several potential avenues of effectiveness. New 12-
month data from a Russian Phase II study directed by
two prominent American clinical trialists showed sig-
nificant benefit in cognition, global function, and be-
havior. Developer Medivation is considering a U.S.
Phase III trial.
● Ketasyn (AC-1202), taken in a modified nutritional
supplement “milkshake,” may correct metabolic defi-
ciencies in the Alzheimer brain by providing ketone
bodies, an alternate energy source for declining cells
that have lost the ability to use glucose. Phase II
studies suggest Ketasyn improves memory in people
with Alzheimer’s who do not have a common risk
gene for the disease. Developer Accera is considering
a Phase III study.
● Avandia (rosiglitazone), a PPAR-gamma receptor ag-
onist currently FDA-approved for Type 2 diabetes, is
under investigation in several Phase III trials for its
potential to improve brain glucose utilization. A con-
ference session summarized early-stage Alzheimer
trial results and reviewed recent data linking Avandia
to an increased risk of heart attack and stroke.
● Lipitor (atorvastatin) is being tested as a treatment for
mild-to-moderate Alzheimer’s in a large Phase III
study under way at 97 sites in 10 countries, with
results expected in 2008. Statins may impact Alzhei-
mer’s disease through several mechanisms, including
general benefit for brain blood vessels and anti-in-
flammatory effects. Some statins, including Lipitor,
may also influence beta-amyloid production.
Improved Diagnosis and Monitoring
As demonstrated by early-stage data on Alzhemed, Fluri-
zan, and other trials reported elsewhere, evidence suggests
that drugs aimed at fundamental Alzheimer pathology may
work best when taken as early as possible. The current
mainstay of Alzheimer diagnosis is cognitive evaluation,
often with the mini-mental status examination (MMSE).
The MMSE is imprecise and insensitive to the first signs of
cognitive change. Extensive neuropsychiatric test batteries
can detect more subtle deficits, but they are not a practical
alternative in general practice as they are time-consuming,
452 Alzheimer’s Association Update / Alzheimer’s & Dementia 3 (2007) 449 – 452
unpopular with patients, and often denied reimbursement by
third-party payers.
Better approaches to diagnosis and monitoring are
clearly needed. Two of the most promising avenues are
brain imaging and biomarkers in spinal fluid or blood.
A packed Prevention Conference session brought attend-
ees up to date on the Alzheimer’s Disease Neuroimaging
Initiative (ADNI). This nationwide study aims to standard-
ize brain imaging protocols for cognitive change and vali-
date their usefulness in early diagnosis and monitoring.
Both magnetic resonance imaging (MRI) and positron emis-
sion tomography (PET) are under investigation. ADNI is
also collecting blood and spinal fluid samples to explore the
accuracy of biomarkers. All ADNI data will be accessible to
any qualified investigator.
Although ADNI is funded chiefly by the National Insti-
tute on Aging (NIA), a number of commercial interests and
organizations, including the Alzheimer’s Association, are
also supporting the initiative. Alzheimer’s Association
funding includes support for evaluation of PET scans using
Pittsburgh compound B (PIB) as a strategy to measure brain
beta-amyloid levels. Initial development of PIB was funded
by the Alzheimer’s Association. The Association is also
funding an effort to bring brain imaging and biomarker
studies in progress in Europe and Asia into compliance with
ADNI protocols, broadening the power and scope of the
initiative.
It is unlikely any imaging technology or biomarker will
be widely accepted until ADNI results are available in
about five years. In the meantime, individual investigators
and companies continue to report promising results with
specific approaches. Conference sessions highlighted bi-
omarker strategies focusing on:
● Signature proteins in cerebrospinal fluid (CSF).
Results from research settings suggest these measures
may be the most sensitive and specific current options.
However, experts acknowledge that the lumbar punc-
ture (LP) needed to evaluate CSF status is unlikely to
achieve general acceptance in the United States.
● Signature proteins in blood. Poor chances of LP
acceptance have spurred efforts to correlate blood
levels with brain levels of key proteins. Although
these correlations are less well characterized than CSF
benchmarks, some strategies are generating promising
data.
● “Biomarker barcodes” in blood. These approaches
focus on multi-component protein or gene expression
profiles in which no single element is predictive, but a
characteristic pattern may distinguish people at in-
creased risk or in the earliest stages of Alzheimer’s.
Emerging Prevention Strategies
Early Alzheimer trial data and experience with other dis-
eases suggest primary prevention may be the most effective
intervention. A conference keynote address by Julie Gerb-
erding, M.D., M.P.H., director of the Centers for Disease
Control and Prevention (CDC), highlighted these leading
potentially modifiable risk factors for cognitive decline:
● Physical inactivity
● Cardiovascular risk factors, including elevated blood
pressure and cholesterol, insulin resistance and Type 2
diabetes, smoking, and excess weight
● Psychosocial issues (engagement, anxiety, sense of
control over life)
The CDC has launched The Healthy Brain Initiative, a
first-ever road map to make cognitive well-being part of our
overall national public health goals. CDC will collaborate
with the Alzheimer’s Association to develop a pilot com-
munity-level intervention program over the next few years.
Honoring Leon Thal: The Tomorrow’s
Leaders Prize
Dementia research lost one of its most widely admired
leaders when Leon J. Thal, M.D., died in a January 2007
private plane crash. Thal, a visionary in designing clinical
trials, led the Alzheimer’s Disease Cooperative Study
(ADCS), a federally funded consortium conducting clinical
studies, including ADNI. He was also a major force in
creation and scientific leadership of the Prevention Confer-
ence, so the event provided a fitting venue to announce a
new prize honoring the legacies of Thal and George Glen-
ner, who isolated and identified beta-amyloid as the chief
component of plaques. The Tomorrow’s Leaders in Alzhei-
mer’s Disease Research Prize, jointly sponsored by the
Alzheimer’s Association, the Cure Alzheimer’s Fund, and
the Lou Ruvo Brain Institute, will award annual prizes of
$100,000 to talented younger scientists with a record of
extraordinary current achievement and outstanding potential
for important future work. Modeled on other “genius”
grants, awardees can use the prize funds for any purpose
they wish. More details about the Tomorrow’s Leaders
Prize and application procedures will be available soon
from the sponsoring organizations.