xiii
SLEEP
Foreword
Teofilo Lee-Chiong, Jr., MD
Consulting Editor
A bidirectional relationship exists between the
sciences of psychiatry and sleep medicine. Indeed,
symptoms of psychiatric disorders are modified by
and, more importantly, can lead to sleep disrup-
tion. The association of insomnia and the risk of
a new psychiatric disorder, specifically major de-
pression, developing is well described. Further-
more, psychiatric disorders can give rise to
complaints of insomnia (eg, bipolar disorder,
depression, generalized anxiety disorder, obses-
sive–compulsive disorder, panic disorder, person-
ality disorders, posttraumatic stress disorder, and
schizophrenia) or excessive daytime sleepiness (eg,
atypical depression and seasonal affective disor-
der). Certain parasomnias, such as nightmares and
sleep terrors, may be more prevalent in patients
with psychiatric illnesses. Finally, the medications
used to manage psychiatric disorders, including
many antidepressant and antipsychotic agents, can
affect sleep quality, duration, and architecture.
The clinical course of schizophrenia may be
complicated by sleep disturbance, sleep-initiation
and sleep-maintenance, insomnia, reversal of day–
night sleep patterns, or alternating phases of sleep-
lessness and sleepiness. Since some antipsychotic
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) xiii–xiv
Teofilo Lee-Chiong, Jr., MD
Section of Sleep Medicine
National Jewish Medical and Research Center
University of Colorado Health Sciences Center
1400 Jackson Street
Room J232
Denver, CO 80206, USA
E-mail address:
lee-chiongt@njc.org
agents can cause sedation, insomnia may also
develop following discontinuation of these medi-
cations. During exacerbations of psychotic symp-
toms, prolonged periods of waking may be
maintained and terminated only by exhaustion.
Conversely, rebound sleepiness can occur during
the waning phase of schizophrenia or during
residual schizophrenia.
Insomnia is common among persons suffering
from mood disorders, and there is generally
a correlation between the severity of both con-
ditions. Sleep disturbances and changes in sleep
architecture (ie, increase in sleep onset latency or
reductions in sleep efficiency, N3 sleep, and REM
sleep latency) may both precede or persist after
remission of major depressive episodes. Insomnia
can be especially severe during a manic episode.
Excessive daytime sleepiness, along with an increase
in the requirement for sleep, may be seen during
the depressive phase of a bipolar disorder, seasonal
affective disorder, or atypical depression. In sea-
sonal affective disorder, major depressive episodes
occur during the fall and winter, when patients may
complain of daytime sleepiness, fatigue, and de-
creased energy levels; during spring and summer,
doi:10.1016/j.jsmc.2008.04.012
xiv Foreword
some patients’ moods improve, but they may
experience hypomanic symptoms. Patients with
atypical depression may present with lethargy,
increase in appetite, weight gain, sensation of
heaviness in the extremities, sensitivity to rejection,
and excessive sleepiness.
Anxiety disorders, including acute stress disorder,
generalized anxiety disorder, and posttraumatic
stress disorder, commonly manifest with insomnia,
frequent nighttime awakenings, recurring anxiety
dreams, or excessive daytime sleepiness. It is
important to distinguish generalized anxiety disor-
der from psychophysiologic insomnia in which
anxiety is chiefly restricted to the issues related to
sleep disturbance and insomnia. Patients with
posttraumatic stress disorder may describe
re-experiences of the original event in frequent
distressing dreams, nightmares, and sleep terrors.
In panic disorder, panic attacks can occur during
sleep, typically in the transition between light
NREM sleep and N3 sleep, but occasionally from
REM sleep; awakenings can be accompanied by
sympathetic activation and delayed return to sleep.
Sleep panic attacks can be triggered by sleep
deprivation.
Finally, other psychiatric disorders can also pro-
foundly affect sleep, including insomnia in somati-
zation disorders, obsessive–compulsive personality
disorder, and anorexia nervosa; excessive sleepiness
in bulimia; and greater rates of obstructive sleep
apnea and periodic limb movement disorder in
attention deficit hyperactivity disorder.

Preface
Karl Doghramji, MD
Guest Editor
Sleep has been of great interest to mankind
throughout the centuries. It was not, however,
a focus of scientific exploration until the 1900s
when Sigmund Freud advanced the notion, in The
Interpretation of Dreams, that dreams are a window
into the mysteries of the mind. Freud’s primary
interest was, no doubt, understanding the thoughts
and feelings that reside in the unconscious mind
and which motivate human behavior and lead to
psychic conflict. Sleep and dreams represented
a means towards that end. However, by necessity,
Freud’s interests led him to advance a number of
theoretical formulations regarding the psychologi-
cal processes that govern dream production and
sleep maintenance.
Whereas Freud and the field of psychoanalysis
that he championed remained focused on the
psychological aspects of sleep, later psychiatrists
delved into sleep’s physiological bases. Hans Berger,
a German psychiatrist, was the first to record and
describe human electroencephalographic wave
forms in 1924. This established the technological
foundation for the milestone discovery of rapid eye
movement (REM) sleep by Aesrinsky and Kleitman
in 1953. The subsequent surge in research into
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
xv
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) xv–xvi
Karl Doghramji, MD
Thomas Jefferson University
Sleep Disorders Center
211 South Ninth Street, Suite 500
Philadelphia, PA 19107, USA
E-mail address:
karl.doghramji@jefferson.edu
electrophysiological sleep formed the foundation
upon which much of the field of sleep medicine, as
we know it today, is based. It is fitting that we begin
this issue of the Sleep Medicine Clinics with a review
of the history of sleep medicine by a sleep re-
searcher who is widely regarded as the father of this
field, Dr. William Dement. Dr. Dement’s initial
interest in sleep research was engendered through
the exploration of dreams, dreaming, and psychi-
atric conditions.
Today, the clinical practice of sleep medicine
subsumes scientific knowledge and clinical con-
ditions that also belong to the mainstay of other
medical specialties. These include psychiatry, psy-
chology, neurology, pulmonary medicine, otolar-
yngology, and pediatrics. The goal of this issue of
Sleep Medicine Clinics is to explore those disorders
that are shared with the field of psychiatry for the
sleep medicine practitioner. We begin with a clari-
fication of the psychological underpinnings of sleep
and dreaming and demonstrate how these broaden
our understanding of the parasomnias and de-
pressive disorders. We then explore the many facets
of insomnia—clearly the most commonly ex-
pressed sleep-related complaint—through a review
doi:10.1016/j.jsmc.2008.04.011
xvi Preface
of its clinical presentation and evaluation and of its
management through pharmacological and non-
pharmacological means. We also explore excessive
daytime somnolence and fatigue, which affect the
majority of patients presenting to sleep medicine
clinics, and review the clinical characteristics and
management techniques of the emotional condi-
tions that feature these two complaints. We then
discuss the parasomnias, long considered to be
disorders of primarily emotional origin, and review
their neurophysiological underpinnings. Psychotic,
mood, and anxiety disorders feature complaints
surrounding sleep and wakefulness, and patients
with these disorders turn to sleep medicine practi-
tioners for assistance. The next few articles focus on
the clinical manifestations of these syndromes and
the nature, significance, and clinical management
of disturbed sleep in these disorders. We also
discuss special issues that confront children, se-
niors, and women. We conclude with a chapter on
seasonal affective disorder and review guidelines for
phototherapy, a therapeutic measure with applica-
tions in a myriad of sleep disorders.
I am indebted to the contributing authors of this
issue: luminaries in the field of sleep medicine and
highly respected for their research and clinical
work. Without them, a project of this scope would
have been impossible. I am also grateful to my
family—Laurel Jeanne, Mark, and Leah—for their
loving support.

History of Sleep Medicine
William C. Dement, MD, PhD
- Professional organizations
- Complications of obstructive sleep apnea
- Discipline identity and discipline overlap
- Minimal penetration of the mainstream
educational system
There have been a number of accounts of the his-
tory of sleep medicine. Perhaps the most notable
was developed by a group of sleep specialists for
the inaugural issue of the Journal of Clinical Sleep
Medicine [1]. Insofar as the history of anything is
known, it is what it is and cannot be changed,
although possibly it can be reinterpreted and ex-
panded. This article may have some novelty from
the perspective of the author’s participation in the
field from 1952 until present, spiced with some in-
formation not achieving publication in scientific
journals, or even the occasional autobiographic
material [2].
It is assumed that the first life forms evolved in
equatorial climates rather than the poles. This being
the case, early life forms were continuously exposed
to the consequences of the earth’s rotation. This
exposure almost certainly induced some sort of
24-hour rest-activity cycle. With diversity and eco-
logic specialization, it is likely that some organisms
became nocturnal primarily for safety and to avoid
daytime predators. Presumably, at the end of eons
of evolution primates and human beings arrived
whose interaction with the environment was pri-
marily visual. This interaction must have fostered
the evolution of the 24-hour sleep-wake cycle.
Interest in sleep almost certainly goes back to the
dawn of history. It is in marked contrast that sleep re-
search and particularly the diagnosis and treatment
Department of Psychiatry and Behavioral Sciences, Sleep Disorder/Sleep Center, 701 Welch Road, #2226,
Stanford, CA 94305–5744, USA
E-mail address: dement@stanford.edu
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
147
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 147–156
- Final thoughts
- References
of sleep disorders began so recently. It may be that
the major obstacle was the need for investigators
and clinicians to stay awake at night. Until the ad-
vent of the electric light, working at night had
none of the practicality that it currently possesses.
The recognition of the important specific sleep
disorders (eg, obstructive sleep apnea) may have
some partial roots in antiquity. Although human
beings have been concerned about their sleep for
centuries, there had not been a specific focus on
problems of sleep and particularly the development
of specific characterizations of individual sleep dis-
orders. It is difficult to select highlights over the last
hundred or so years; however, there is some consen-
sus. A very important early step was the discovery of
the electrical activity of the central nervous system
by Richard Caton in 1875. The German psychiatrist
Hans Berger was the first to describe human brain
waves in reports published in the late 1920s. In
the 1930s, a group at Harvard described different
patterns of sleeping brain waves, and in particular
described what are now known as ‘‘sleep spindles,’’
‘‘K-complexes,’’ and ‘‘high amplitude slow waves.’’
By this time, it had also been found that the resting
awake human adult had a very prominent 8 to
12 cycle per second nearly continuous oscillation
from the occipital surface of the skull, which disap-
peared at the onset of sleep. Further advances in
understanding the mechanisms of sleep and
doi:10.1016/j.jsmc.2008.01.003
148 Dement
wakefulness were delayed by the hostilities up to and
during World War II, and to some extent, the Korean
War and events associated with the Cold War.
Appropriately, the discovery that many say initi-
ated the modern era of sleep research and sleep
medicine was made in the laboratory of a man
whose focus was on the study of human sleep, Uni-
versity of Chicago Professor of Physiology
Nathaniel Kleitman. Slow rolling eye movements
during sleep had been described much earlier. Kleit-
man decided that this slow activity might be an ex-
cellent measure of the depth of sleep. He based this
hypothesis on the somewhat disproportionately
large areas of the brain devoted to the execution
of eye movements and the demanding nuances of
visual following and focusing. A graduate student
in Kleitman’s department, Eugene Aserinksy, was
put to work studying the aforementioned slow roll-
ing eye movements during sleep. Initially, Aserinsky
simply looked at the eyes of sleeping subjects (eg,
direct visual observation). He soon came across
the fact, however, of the existence of a corneoretinal
potential difference with the cornea consistently
about 70 mV positive to the retina. Aserinsky then
concluded that eye movements could be recorded
electrographically and proceeded to do so with
the use of an ink-writing oscillograph.
In the course of observing slow eye movements,
Aserinsky occasionally saw what appeared to be elec-
trical artifacts. In 1952, I was a sophomore medical
student at the University of Chicago. I approached
Kleitman and asked if I could work in his laboratory.
He immediately gave me the assignment of helping
Aserinsky. My somewhat tedious task was to sit by
the bed of the sleeping subject (only one subject at
a time was studied) and look at the eyes with a flash-
light when Aserinsky saw the ‘‘electrical artifacts.’’ Af-
ter a few observations, it became clear that the
electrical artifacts were actually the changes in elec-
tropolarity as the eyeballs moved rapidly (very dif-
ferent from the previously described slow
movements). Because the periods of eye movements
were associated with an elevated heart rate, Kleitman
wondered if they might represent dreaming.
To test this notion, 20 normal men were used in sev-
eral series of experiments. These observations [3] were
subsequently published in 1953 and marked what
some have said represents the true beginning of the
modern era of sleep research and sleep medicine. Sub-
jects were wakened when eye movement potentials
appeared in the electrooculogram. In the original ob-
servations, rapid eye movements (REM) were not seen
1
In 1983, I was driving from Chicago to southern Illinois on a route that took me past the Manteno State Hospital. Permission was given
for me to visit the buildings and grounds. They were eerily deserted, but quite familiar, including the building that housed the sleep
recordings.
in four of the subjects [3]. Nonetheless, when the oc-
currence and detail of dream recall reports from REM
awakenings were compared with dream recall reports
from awakenings when REM were not present, the dif-
ferences were highly significant.
In 1953, any interest in sleep by health profes-
sionals was engendered mainly by the theories of
Sigmund Freud, specifically about the meaning of
dreams. Accordingly, the interest was only in the
phenomenon of dreaming, with sleep as the neces-
sary background state. Freud’s work resulted in the
creation of a clinical and scientific discipline known
as ‘‘psychoanalysis,’’ and the technique of dream in-
terpretation was a very important part of its thera-
peutic and theoretic approach to psychiatric
problems. The discovery of an objective measure
of dreaming was extremely interesting. Aserinsky’s
initial observations described previously were per-
formed using a four-channel ink writing oscillo-
graph to record brain waves and eye movements,
and because electroencephalogram (EEG) paper at
the time was fairly expensive, sample recordings
were made about every 10 or 15 minutes. No con-
tinuous all-night recordings were performed in
the first wave of eye movement observations.
My first independent study was performed at the
Illinois Manteno State Hospital, in a special
research ward that housed approximately 5000
chronic schizophrenics1. No antipsychotic medica-
tions were available at the time, but electroconvul-
sive shock therapy was occasionally used. Because
Freud believed that dreams were the safety valve
of the mind, it was hypothesized that for some
reason this safety valve was not available in schizo-
phrenics and dreaming erupted into the waking
state giving rise to the symptomatology of psycho-
sis. During the summer and fall of 1953, I studied
17 schizophrenics and 13 medical students. All
had periods of REM during sleep [4].
Because of my intense interest in dreams and
dreaming at the time, I decided to carry out
a more complete description of all-night sleep by
recording brain waves and eye movements continu-
ously all through the night rather than sampling. To
do this, a relatively slow paper speed was used but
REM potentials were very easy to identify at any pa-
per speed. One purpose was further to expand and
quantify the description of these REM periods. Dur-
ing my junior and senior years in medical school, I
performed a total of 126 all-night recordings on 33
subjects and, by means of a unique categorization
of sleeping EEG patterns, scored the sleep

recordings in their entirety. When these 126 all-
night records were examined, I found that there
was a predictable sequence of patterns over the
course of the night that had been hinted at in Aser-
insky’s study but entirely overlooked in all previous
EEG studies of sleep. Furthermore, all subjects with-
out exception showed periods of REM [5].
Although this sequence of regular variations has
now been observed thousands of times in hundreds
of laboratories, the original description remains es-
sentially unchanged. The following is from my
chapter in the 4th edition of Principles and Practice
of Sleep Medicine [6].
The usual sequence was that after the onset of
sleep, the EEG progressed fairly rapidly to stage 4
(record dominated by high amplitude slow waves)
which persisted for varying amounts of time gener-
ally about 30 minutes, and then a lightening of
sleep indicated by EEG changes took place.
Whereas the progression from wakefulness to stage
4 at the beginning of the cycle was almost invari-
able through a continuum of change, the lighten-
ing was usually abrupt and often coincided with
a body movement or a series of body movements.
After the termination of stage 4, there was generally
a short period of stage 2 (low amplitude EEG with
sleep spindles) or stage 3 which gave way to stage 1
and rapid eye movements. When the first eye move-
ment period ended, the EEG again progressed
through a continuum of change through stage 3
or 4, which persisted for a time and then lightened,
often abruptly, with body movement to stage 2
which again gave way to stage 1 and the second
rapid eye movement period. This cyclic variation
of EEG patterns occurred repeatedly throughout
the night at intervals of 90 to 100 minutes from
the end of one eye movement period to the end
of the next. The regular occurrences of REM periods
and dreaming strongly suggested that dreams did
not occur in response to chance disturbances.
Despite these observations, sleep was still consid-
ered to be a single organismic state of being. In our
1957 paper, Kleitman and I characterized the EEG
during REM periods ‘‘as emergent stage 1 in contrast
to descending stage 1 at the onset of sleep’’ [5]. The
percentage of total sleep time occupied by REM pe-
riods was between 20% and 25%. The periods of
REM tended to be shorter in the early cycles of the
night. This typical pattern of all-night sleep has
been seen over and over in normal humans of
both genders in widely varying environments and
cultures and across the lifespan.
By the time I graduated from medical school, the
data allowed me to assume that all male adults in
the human population had a characteristic sleep ar-
chitecture that consisted of alternating periods of
REM sleep and non-REM sleep. The latter name
was consistent with the much greater interest in
History of Sleep Medicine 149
REM periods. Although it seemed unlikely, in the ab-
sence of data it remained a possibility that these REM
periods did not occur in females. Accordingly, it was
necessary to study at least one woman. Kleitman was
somewhat reluctant to have a woman sleeping all
night in his laboratory and insisted on having a chap-
erone. I recruited a chaperone, who came to Kleit-
man’s sleep laboratory on the same evening as the
first female subject and promptly went to sleep on
a cot in Kleitman’s office. The first female subject
did indeed have periods of REM during sleep that
were essentially the same as the male subjects. In
1955, Kleitman temporarily left the University of
Chicago to spend a sabbatical leave in Europe. Alone
in the sleep laboratory, I recorded several more fe-
males without a chaperone and satisfied myself
that the characteristic sleep architecture was present
in all human adults regardless of gender.
Because of the corneal bulge, it is very easy to see
the eyeballs rotate under the closed lids. It had been
assumed that dreaming might not occur until sev-
eral years after birth. It was also assumed that the
ability of the human brain to experience the subjec-
tive world of dreaming was not necessarily a require-
ment for the occurrence of the REM. In 1956, I went
to the newborn nursery at the University of Chicago
Hospital and observed sleeping 1-day old infants. I
saw the characteristic binocularly synchronous
REMs enough times to assume that newborn sleep
includes REM periods. In subsequent studies, Ho-
ward Roffwarg and I found that nearly 50% of the
sleep time in newborns consisted of REM periods
[7]. These observations have been subsequently
confirmed by many others. Assuming that all hu-
man sleep contained REM periods, the next ques-
tion was what about animals? In the Department
of Physiology at the University of Chicago, research
was being performed on cats. Although cats were
somewhat unwilling to sleep under laboratory cir-
cumstances, enough observations were eventually
accumulated to be certain that a state of sleep anal-
ogous to REM sleep in humans did occur with reg-
ularity in cats [8].
The technique of electrodes implanted directly
into the brain of animals had been developed in
1956; my studies of sleep in cats used pins inserted
into the scalp. When cats were awake, the record-
ings were dominated entirely by the electromyo-
graphic activity of the cats’ prominent temporal
muscles. During periods of REM sleep, this electro-
myographic activity completely disappeared and
the brain wave patterns were easy to observe. In ad-
dition, it was noteworthy that in these observations,
the brain wave patterns during eye movement pe-
riods in the cat were essentially identical with EEG
patterns recorded during wakefulness. These pat-
terns were of low amplitude with higher
150 Dement
frequencies. The assertion by me, however, that an
‘‘activated’’ EEG could be associated with unambig-
uous sleep was considered to be absurd.
I must admit that I did not fully appreciate the
significance of the absence of muscle potentials
during the REM periods. It remained for the French
scientist, Michel Jouvet, who began working in
Lyon, France, in 1959 to insist on the importance
of electromyographic suppression in his early
studies [9], the first of which was published in
1959. I left Chicago in 1957 to take a post at the
Mt. Sinai Medical Center in New York City. There
I collaborated with others to study muscle activity
during sleep. We monitored the electrically induced
H reflex during sleep and found it to be completely
suppressed in REM periods [10].
As the decade of the 1960s began, the concept of
sleep as a single organismic state with different
levels was giving way to the concept of the duality
of sleep. In all other mammals that were studied
in addition to humans, it was found that overall
sleep consisted of two entirely different alternating
organismic states: REM sleep and non-REM sleep.
An enormous amount of data was accumulated
supporting the duality concept; microelectrode
studies of single neurons in the brain during sleep,
more complex studies of motor atonia, and finally
the seminal work of Jouvet and Mounier [11] show-
ing that electrolytic lesions in the pontine reticular
formation of the cat brain eliminated REM sleep,
but left non-REM sleep intact.
Sleep research, emphasizing all-night sleep re-
cordings, burgeoned in the 1960s. This research
was the legitimate precursor of sleep medicine
and particularly of its core clinical test, polysom-
nography. Other than the early studies in schizo-
phrenic patients at the Manteno State Hospital,
very little interest in other clinical applications
had been manifested until investigators noted a sig-
nificantly shortened REM latency in association
with endogenous depression. In the ensuing years,
this phenomenon has been well investigated. Other
important harbingers of sleep medicine were the
following: (1) the observation of sleep-onset REM
periods in individuals afflicted with narcolepsy,
(2) an interest in epileptic seizures during sleep,
and (3) sleep studies to evaluate benzodiazepine
medications for the treatment of insomnia. The lat-
ter application represents the first time the sleep
laboratory and all-night polysomnography were
used as part of efficacy protocols to evaluate seda-
tive hypnotics.
In 1959, my colleague at the Mt. Sinai Hospital,
Dr. Charles Fisher, saw a patient who was referred
with the diagnosis of narcolepsy. At Fisher’s sugges-
tion, I arranged to record all-night sleep in this pa-
tient. Within several seconds after the patient fell
asleep, he was showing the dramatically characteris-
tic REM sleep and electromyographic suppression.
Simultaneously, observations were performed on
a single patient by Vogel [12] at the University of
Chicago, the results of which were published in
1960. In a subsequent collaborative study between
the University of Chicago and the Mt. Sinai Hospi-
tal, nine patients underwent all-night sleep record-
ings and the important sleep-onset REM periods
were described and reported in a 1963 publication
[13]. It is noteworthy that patients were diagnosed
with narcolepsy when they showed a characteristic
set of symptoms reported in 1880 by Gelineau
(sleep attacks, hypnagogic hallucinations, sleep pa-
ralysis, and cataplexy). The latter term was applied
to sudden attacks of muscular atonia experienced
by the narcoleptic patients usually precipitated by
strong emotion. Subsequent studies showed that
sleepy patients who did not report having attacks
of cataplexy also did not show sleep-onset REM pe-
riods. Conversely, sleepy individuals who reported
the occurrence of attacks of cataplexy always had
sleep-onset REM periods. It was clear that the best
explanation for the strange attacks of muscle atonia
was the normal motor inhibitory mechanism of
REM sleep occurring in a precocious or abnormal
manner [14].
Most of the aforementioned research was per-
formed after I moved to Stanford University in Jan-
uary 1963. My plans to carry out this research were
initially hampered by the fact that not a single nar-
coleptic patient could be identified at the Stanford
University Medical Center. Finally, in desperation,
I placed an advertisement in the local daily newspa-
per. More than 100 people responded! About 50 of
these respondents were bona fide narcoleptics hav-
ing daytime sleepiness and cataplexy and hypnago-
gic hallucinations and sleep paralysis. The response
to this advertisement was a seminal event in the
future development of sleep disorders medicine.
None of the narcoleptics had been previously diag-
nosed and I was forced to assume responsibility for
their clinical management for them to participate in
the research project. As the months passed, I and
a research colleague became responsible for manag-
ing and taking care of over 100 individuals with
narcolepsy. Mostly this involved visiting with the
patients at regular intervals and adjusting their
medications, which consisted mainly of methyl-
phenidate or dextroamphetamine. Providing this
patient care was the harbinger of the modern sleep
disorders clinic. I was managing sleep disorders
patients and in addition performing diagnostic
recordings to demonstrate the pathognomonic
sleep-onset REM periods. In 1964, I formally
launched the Stanford University Narcolepsy
Clinic. It was set up as a true fee-for-service

enterprise. Most of the patients were unable to pay
their bills, however, and insurance companies de-
nied payment because the diagnostic recordings of
narcolepsy patients were deemed ‘‘experimental.’’
Accordingly, this unique clinic was closed because
of financial bankruptcy.
One of the most important landmarks in the his-
tory of sleep disorders medicine occurred in Europe
around this time. Obstructive sleep apnea was dis-
covered independently by Gastaut, Tassinari, and
Duron in France and by Jung and Kuhlo in Ger-
many. Both groups reported their findings in
1965. Scholars have found hints in earlier works,
even biblical, that suggest obstructive sleep apnea
was being described. The aforementioned Euro-
pean reports, however, were the first clear-cut recog-
nition and description that had a direct continuity
to sleep disorders medicine as it is known today.
There is no evidence that the pulmonary medicine
community in the United States understood the im-
portance of these European reports. What should
have been an almost inevitable discovery by the
pulmonary medicine community or perhaps by
the ear-nose-throat surgery community did not oc-
cur because there was no interest in either specialty
for meticulously observing respiration during sleep.
The well-known and frequently cited study of Bur-
well and colleagues [15], although impressive in
a literary sense in its evoking of the somnolent
boy Joe from the Papers of the Pickwick Club by
Charles Dickens, erred badly in attributing the som-
nolence of their obese patients to hypercapnea. The
frequent citing of this paper further reduced the
likelihood that sleep apnea would be discovered
by the pulmonary medicine community. To this
day, evidence that hypercapnea causes true somno-
lence is completely lacking, although high levels of
PCO2 are certainly associated with impaired cerebral
function. Nonetheless, the term ‘‘pickwickian syn-
drome’’ became an instant neologistic success and
may have played a role in stimulating an interest
in this syndrome by the European neurologists
who were also interested in sleep.
In the 1960s, a small group of French neurolo-
gists began specializing in clinical neurophysiology
and electroencephalography. An individual trained
in France, Alberto Tassinari, joined the Italian neu-
rologist Elio Lugaresi, in Bologna. These scientists
then performed a crucial series of clinical sleep in-
vestigations that provided a complete description
of the sleep apnea syndrome including the first
observations of the occurrence of sleep apnea in
nonobese patients, a description of the cardiovascu-
lar correlates, and a clear identification of the im-
portance of snoring and hypersomnolence as
diagnostic indicators. These pioneering studies are
recounted in Lugaresi’s book, Hypersomnia with
History of Sleep Medicine 151
Periodic Apneas [16]. Henri Gastaut and Elio Lugar-
esi were prompted to organize an international
symposium in 1967, the proceedings of which
were published as Abnormalities of Sleep in Man.
These proceedings encompassed issues across
a full range of pathologic sleep in humans. The
meeting took place in Bologna and the papers pre-
sented covered many of what are now major topics
in the sleep medicine field: insomnia, sleep apnea,
narcolepsy, and periodic leg movements during
sleep. It was an epochal meeting from the point
of view of the clinical investigation of sleep. The
only major issues not represented were clear con-
cepts of clinical practice models and accurate esti-
mates of the high population prevalence of sleep
disorders. The event, however, finally triggered
a serious international interest in sleep apnea
syndromes. Another extremely important meeting
was organized by Lugaresi in 1972. An account of
this meeting is included in Lugaresi’s 1978 book
[16].
Despite all the laudatory clinical research, the
concept of the all-night sleep recording as a clinical
diagnostic test did not occur. An all-night diagnos-
tic test, particularly if it was conducted on out-pa-
tients, was completely unprecedented. In addition,
the cost of all-night polysomnographic recording
was already quite high without adding the cost of
hospitalization, although the latter would have
legitimized a patient spending the entire night in
a testing facility. It should also be clear that in
1970, only a tiny number of people understood
that the complaint of daytime sleepiness repre-
sented something that had clinical significance.
Even narcolepsy, which had been fully character-
ized as an interesting and disabling clinical syn-
drome, was not recognized as such in the larger
medical community. The extreme neglect of pa-
tients with narcolepsy (we documented a mean of
15 years from onset of the characteristic symptoms
of narcolepsy until recognition and accurate diag-
nosis by a clinician, and patients having seen
a mean of 5.5 different physicians during the
15 years), however, well justified creating a medical
specialty dedicated to such patients.
The evaluation of efficacy in the development of
sedative hypnotics had quite reasonably begun to
include all-night sleep recordings. When a pharma-
ceutical company wanted to evaluate a new hyp-
notic compound, it was necessary to recruit
volunteer subjects for the research who manifested
objective signs of insomnia. Recruiting insomniacs
typically involved giving clinical advice and staying
in touch. For these reasons, it became obvious to
me early in 1970 that there needed to be a new clin-
ical discipline that specialized in the diagnosis and
treatment of individuals afflicted with one or other
152 Dement
of the then known sleep disorders. In the summer
of 1970, with as much fanfare as we could muster,
we announced the opening of the Stanford Univer-
sity Sleep Disorders Clinic and organized a press
conference to publicize this event. Only one re-
porter attended the press conference, and he turned
out to be more interested in the issue of whether or
not children should sleep with their parents. In the
early months and probably the first year or so of the
existence of the Stanford University Sleep Disorders
Clinic, there were very few referrals from commu-
nity physicians. For quite some time, we were
unable to achieve more than five or six referrals
per week.
The all-night diagnostic sleep recordings were
done on the side by technologists who at the time
were funded by research projects. I and my sleep
colleague, Dr. Vincent Zarcone, were also paid
largely by research grants. One of the current giants
of sleep medicine, Dr. Christian Guilleminault,
a French neurologist and psychiatrist, had come
to the sleep center as a research fellow and was do-
ing basic research in the summer of 1970. He re-
turned to France later in the year. Because most of
our patients in the early days turned out to be af-
flicted with narcolepsy, we believed in the need to
have a neurologist on the clinical staff. In the sum-
mer of 1971, Dr. Zarcone and I recruited Dr. Guille-
minault to return to Stanford, which he did,
arriving in January of 1972. Until his arrival, the
Stanford group had not routinely used respiratory
and cardiac sensors in their all-night sleep studies.
Beginning in 1972, these measurements became
a routine part of the all-night diagnostic test, which
was finally dubbed ‘‘polysomnography’’ in 1974.
During 1972, the search for sleep abnormalities
in patients with sleep-related complaints contin-
ued. We also attempted to conceptualize the patho-
physiologic process as both an entity and as the
cause of the presenting symptom. With this ap-
proach, a number of phenomena seen during sleep
were rapidly linked to the fundamental sleep-
related presenting complaints. Toward the end of
1972, the basic concepts and protocols of sleep dis-
orders medicine were sculpted sufficiently to offer
a day-long course through Stanford University’s Di-
vision of Postgraduate Medicine. This course was
offered under the title ‘‘The Diagnosis and Treat-
ment of Sleep Disorders.’’ The topics covered were
normal sleep architecture; the diagnosis and treat-
ment of insomnia with drug-dependent insomnia,
pseudoinsomnia, central sleep apnea, and periodic
leg movement as diagnostic entities; and the diag-
nosis and treatment of excessive daytime sleepiness
or hypersomnia with narcolepsy, non-REM narco-
lepsy, and obstructive sleep apnea as diagnostic
entities.
Professional organizations
In 1964, the Stanford University Sleep Research
Program hosted the fourth annual meeting of per-
sons interested in sleep research. This was still
somewhat informal, but before the occasion a small
group met with the idea of creating some sort of
professional organization. The first suggestion for
the name of such an organization was the Associa-
tion for the Study of Sleep, until someone noticed
the acronym. Finally, because of the interest in
both the physiology of sleep and the phenomenon
of dreaming, it was agreed that the organization
should be named the Association for the Psycho-
physiological Study of Sleep (APSS). Every since,
there have been annual meetings of the APSS with-
out exception.
By 1975, there were five centers diagnosing and
treating patients with sleep-related complaints. Fol-
lowing the launching of the Stanford University
Sleep Disorders Clinic, the second such clinic was
initiated by the late Dr. Elliot Weitzman who spent
a sabbatical year, 1974 to 1975, at Stanford. In
a 1976 meeting at O’Hare Airport in Chicago, the
Association of Sleep Disorders Centers (ASDC)
was formally organized. The charter members
were from medical schools at Stanford, Montefiore,
Baylor, Cincinnati, and Pittsburgh.
The APSS continued to meet annually, and at
a certain point the issue of sponsoring a scientific
journal arose. The journal Sleep was sponsored
jointly by the APSS, the European Society for Sleep
Research, and the ASDC. The first issue appeared in
September 1978.
Because the sleep disorders discipline was clearly
established and growing, the first president of the
ASDC and secretary began to think about achieving
more notice from the National Institutes of Health
and some support from the US Congress. In a first
visit to Washington, Dr. Merrill Mitler encountered
a Washington lobbyist, Mr. Harley Dirks, who had
been chief of staff for the Chairman of the Senate
Appropriations Committee. After leaving his con-
gressional position, Mr. Dirks established the
Health and Medicine Council of Washington,
which is now headed by his son, Mr. Dale Dirks.
In 1986, to increase the lobbying clout of the sleep
organizations, it was decided to meld them into
one overall organization. This entailed the difficulty
that two of the organizations consisted of individ-
ual members, whereas the third consisted of mem-
ber centers. A fourth organization was temporarily
created called the Clinical Sleep Society to facilitate
this process and after several organizing meetings,
the Association of Professional Sleep Societies
came into being. There was, however, a problem
with leadership and finances such that the

organizations almost immediately separated into
a loose coalition renamed the Associated Profes-
sional Sleep Societies.
I was elected president of the ASDC for four con-
secutive 3-year terms, after which it was decided
that a 1-year presidency would suffice and more
managerial responsibility should devolve to the
Executive Director. The first formal national office
of the APSS and what is now the American Academy
of Sleep Medicine was located in Rochester, Minne-
sota; it was later moved to a less out of the way lo-
cation in Westbrook, Illinois, a suburb of Chicago.
A fourth professional society came into being in
1991 to serve the interests of dentists, called the
Academy of Dental Sleep Medicine. In addition, un-
der the aegis of the American Academy of Sleep
Medicine, several tax exempt foundations were
formed. The first examination by what is now the
American Board of Sleep Medicine was adminis-
tered in 1977. Today, there are more than 7000
individual members of the professional sleep socie-
ties, more than 1400 accredited sleep disorder cen-
ters, and more than 10 scientific journals around
the world devoted to sleep disorders, sleep research,
and biologic rhythms.
Complications of obstructive sleep apnea
The disability and cardiovascular complications of
severe obstructive sleep apnea are very serious. In
the early years of sleep medicine, treatment options
were limited to weight loss and chronic tracheos-
tomy. Chronic tracheotomy dramatically amelio-
rated the symptoms and complications of the
illness. Many patients with severe obstructive sleep
apnea, however, nonetheless strongly resisted being
treated by means of chronic tracheostomy. In
addition, the broader medical community was ex-
tremely skeptical about the use of chronic tracheos-
tomy mainly because of ignorance about the sleep
disorders field. Securing the proper management
was a major challenge, as illustrated by the case
history of one of the first patients referred to the
Stanford Clinic, a 10-year-old boy, Raymond M.
The overall difficulties are illustrated in a personal
account by Dr. Christian Guilleminault.
Raymond M. was a 10 1⁄2 -year-old boy referred to
the pediatrics clinic in 1971 for evaluation of unex-
plained hypertension which had developed pro-
gressively over the preceding 6 months. There was
a positive family history of high blood pressure,
but never so early in life. Raymond was hospital-
ized and had determination of renin, angiotensin,
and aldosterone, renal function studies including
contrast radiographs, and extensive cardiac evalua-
tion. All results had been normal except that his
blood pressure oscillated between 140-170/90-100.
History of Sleep Medicine 153
It was noticed that he was somnolent during the
daytime and Dr. S. suggested that I see him for
this ‘unrelated’ symptom.
I reviewed Raymond’s history with his mother.
Raymond had been abnormally sleepy ‘all his
life.’ However, during the past two to three years,
his schoolteachers were complaining that he would
fall asleep in class and was at times a ‘behavioral
problem’—not paying attention, hyperactive, and
aggressive. His mother confirmed that he had
been a very loud snorer since he was very young,
at least since age two, perhaps before.
Physical examination revealed an obese boy with
a short neck and a very narrow airway. I recommen-
ded a sleep evaluation which was accepted. An
esophageal balloon and measurement of end tidal
CO2 was added to the usual array. His esophageal
pressure reached 80 to 120 cm H2O, he had values
of 6% end tidal CO2, apneic events lasted between
25 and 65 sec, and the apnea index was 55. His
SaO2 [arterial oxygen saturation] was frequently
below 60%.
I called the pediatric resident and informed him
that the sleep problem was serious. I also suggested
that the sleep problem might be the cause of the as
yet unexplained hypertension. The resident could
not make sense of my information and passed it
to the attending physician. I was finally asked to
present my findings at the weekly pediatric case
conference which was led by Dr. S. I came with
the recordings, showed the results, and explained
why I believed that there was a relationship be-
tween the hypertension and the sleep problem.
There were a lot of questions. They simply could
not believe it. I was asked what treatment I would
recommend, and I suggested a tracheostomy. I
was asked how many patients had this treatment
in the United States, and how many children had
ever been treated with tracheostomy. When I had
to answer ‘zero’ to both questions, the audience
was somewhat shocked. It was decided that such
an approach was doubtful at best, and completely
unacceptable in a child. However, they did concede
that if no improvement was achieved by medical
management, Raymond would be reinvestigated,
including sleep studies.
This was Spring 1972. In the Fall, he was, if any-
thing, worse in spite of vigorous medical treatment.
At the end of 1972, Raymond finally had his trache-
ostomy. His blood pressure went down to 90/60
within 10 days, and he was no longer sleepy. Dur-
ing the five years we were able to follow Raymond,
he remained normotensive and alert, but I had to
fight continuously to prevent outside doctors
from closing his tracheostomy. I don’t know what
happened to him since.
Discipline identity and discipline overlap
In July of 1981 I wrote a note for the ASDC newslet-
ter. The title was ‘‘Does Somnology (a.k.a Sleep
154 Dement
Medicine) Exist?’’ It was patterned after an analogous
discussion titled ‘‘Does Gerontonology Exist?’’,
which appeared elsewhere. A dozen criteria defining
the existence of a clinical scientific discipline were
posed. They were paraphrased by me and answered.
In 1981, the title of someone who practiced sleep
medicine was ‘‘clinical polysomnographer,’’ a name
of which no one was particularly fond. The criteria
are presented below, slightly updated for this article.
1. Does clinical polysomnography and sleep research
have the broad academic organization to provide ca-
reer lines for teachers and researchers? No. Ten-
ured appointments are controlled by
departments. As far as I know, no medicinal
school has a department of sleep medicine.
2. Does society provide support for sleep research? Al-
though there are a fair number of federal grants
for sleep research, they are not easy to obtain.
Certainly, there is no official large government
mandate for such funding. There is, however,
a small Center for Sleep Disorders Research
(2 FTEs) in the National Heart, Lung, and
Blood Institute.
3. Is there a student demand for courses? Although
the need is great, the demand is small. It is in-
creasing, however, as are applications to sleep
centers for postgraduate training.
4. Is there a demand for graduates? Yes, definitely.
Many institutions interested in starting sleep
disorders centers need trained and experienced
sleep specialists to serve as medical directors;
the dearth of well-trained people actually ham-
pers development of the field.
Questions five through seven concern a would-be
discipline’s relation to its subject matter.
5. Does sleep research study some distinct part of na-
ture? Yes. Sleep research is concerned with the
sleeping organism, the determinants of sleep,
the mechanisms of sleep, the circadian rhythm
of sleep and wakefulness and its determinants,
the role of sleep in waking function, and the
pathology of all systems and mechanisms re-
sponsible for the sleep-wake cycle. Certainly,
sleep research overlaps with other disciplines.
Sleep-induced respiratory disturbances concern
pulmonary specialists. Many sleep disorders
have significant hemodynamic impact, of inter-
est to cardiologists. Hypersomnia is still con-
sidered to be a neurologic problem. Sleep
research provides the umbrella or the forum
in which professionals from different areas
can share their findings and forge new diagnos-
tic and therapeutic approaches.
6. Is there at least a social group to be studied? Not
applicable.
7. Is there at minimum a social problem to be studied?
Not applicable unless it is the impact of waking
sleepiness in all components of society.
The last set of questions concerns the internal cri-
teria or maturity of the discipline.
8. Does the work of sleep researchers share a unique
method? Certainly. Polysomnography, sleep
scoring, characterization of sleep patterns,
and sleep deprivation studies all are good ex-
amples of approaches unique to the field.
9. Does sleep research have a theory? Yes. Although
there are many questions as yet unanswered,
several models have been developed that gener-
ate testable hypotheses.
10. Can the knowledge of the field be linked in an
integrated whole? The answer is now ‘‘yes.’’ Al-
though there is minor disagreement about
what might be regarded as common knowledge
within the discipline, there are now several
textbooks.
11. Do scholars, teachers, and practitioners find it useful
to share interaction? As with any field investigating
widespread and serious medical disorders, the
answer to this question is an unqualified ‘‘yes.’’
The overall answer to these questions is ‘‘yes.’’
Clearly, sleep research qualifies as a full-fledged ac-
ademic discipline concerned with, for example,
physiology during sleep, sleep mechanisms, state
regulation and sequencing in the organism, chro-
nobiology as it relates to sleep and wakefulness
rhythms, sleepiness and alertness and related vari-
ables, and the diagnosis and treatment of patholo-
gies in all of these areas.
Unlike many newcomers to the medical main-
stream, sleep medicine is not the child of a single
parent discipline. Rather, clinical and basic sleep re-
search have been the foster children of many disci-
plines but have been the favored children of none.
In the early days, we were passed from hand to
hand as then current findings had practical signifi-
cance for other specialties. Early research on REM
deprivation provided models for psychosis and
depression. Sleep apnea syndromes, sleep-induced
gastroesophageal reflux with secondary laryngo-
spasm, and general sleep-related respiratory load
increase have intrigued those in pulmonary medi-
cine. REM sleep-related penile tumescence is of
great diagnostic value to urologists in the study of
impotence. Pediatricians work with sleep specialists
to unravel sudden infant death syndrome and the
impact of sleep disturbances on normal develop-
ment. REM sleep muscle inhibition and sleep-
related changes in the balance of the autonomic
and parasympathetic nervous system concern prac-
titioners caring for patients already compromised

(eg, those with poliomyelitis, familial dysautono-
mia, muscular dystrophy, diabetes, and so forth).
Those concerned with biologic rhythms share an
interest in the important circadian rhythms of sleep
and wakefulness.
Minimal penetration of the mainstream
educational system
The clinical scientific discipline of sleep medicine
has not yet been widely embraced by academic med-
icine. Although such surveys are difficult, it is likely
that there are fewer than five tenured professorships
in American medical schools designated specifically
for sleep research and sleep medicine. There does
not exist in any medical school a department of sleep
medicine. There are, however, at least a few divisions
of sleep medicine, eg, at Harvard University in the
Department of Medicine, at the University of Penn-
sylvania in the Department of Medicine, at Stanford
University in the Department of Psychiatry, and at
the University of Michigan in the Department of
Neurology. An unsuccessful proposal was made by
one medical school to establish a division in its de-
partment of ear, nose, and throat surgery.
At the undergraduate level, there is very little sys-
tematic teaching about sleep knowledge, circadian
rhythms, dreaming, and sleep disorders in the
nearly 4000 colleges and universities. In a very re-
cent survey of Stanford undergraduates, fewer
than 2% had received systematic teaching about
sleep before matriculating.
The first sleep disorders clinic did not exist before
1970 and until the invention of continuous positive
airway pressure for the treatment of sleep apnea, the
standard treatment using chronic tracheostomy
posed a major obstacle to the rapid expansion of
the diagnosis and treatment of sleep apnea. The dis-
cipline of sleep medicine did not become truly via-
ble until the 1980s.
Unfortunately, clinicians have now reached
a time when expansion and growth in many com-
ponents of society has flattened or slowed. The
number of medical schools in America today is es-
sentially the same as it was four decades ago, and no
undergraduate department of psychology or hu-
man biology offers a well-organized and complete
teaching program involving the sleep-related
disciplines.
Final thoughts
The need for an effective societal understanding of
sleep-related issues is great and urgent. It is possible
that the entire human race has not achieved its full
behavioral and intellectual capacity because of
chronic sleep deprivation and the accumulation of
History of Sleep Medicine 155
sleep debt. Extrapolating from the report by Young
and coworkers [17] in 1993, there are at least
30 million victims of obstructive sleep apnea in
the United States. It is not inconceivable that new
cases are developing at a rate that is higher than
the rate of cases that are identified and effectively
treated. The complaint of insomnia is extremely
widespread and the underlying causes are reason-
ably well characterized.
Sleep medicine in all its ramifications is a field
that should take its destiny into its own hands as
much as possible. We must penetrate the educa-
tional system. We must prevail on elected represen-
tatives to do more for the field both at the national
and the state levels. If both patients and sleep pro-
fessionals can be mobilized, the numbers are there
to achieve many things. There are still far too many
sleep-related accidents, far too many undiagnosed
and untreated victims of sleep disorders, and
many individuals continue to organize their lives
based more on mythologic beliefs than the true
facts of sleep and wakefulness.
References
[1] Shepard J, Buysse D, Chesson A, et al. History of
the development of sleep medicine in the United
States. J Clin Sleep Med 2005;1:61–82.
[2] Dement W, Vaughan C. Promise of sleep. New
York: Delacorte Press, Random House Inc.; 1999.
[3] Aserinsky E, Kleitman N. Regularly occurring pe-
riods of eye motility, and concomitant phenom-
ena, during sleep. Science 1953;118:273–4.
[4] Dement W. Dream recall and eye movements
during sleep in schizophrenics and normals.
J Nerv Ment Dis 1955;122:263–9.
[5] Dement W, Kleitman N. Cyclic variations in EEG
during sleep and their relation to eye move-
ments, body motility, and dreaming. Electroen-
cephalogr Clin Neurophysiol 1957;9:673–90.
[6] Kryger M, Roth T, Dement W. Principles and
practice of sleep medicine. Philadelphia: WB Sa-
unders Co.; 1994.
[7] Roffwarg H, Muzio J, Dement W. Ontogenetic
development of the human sleep-dream cycle.
Science 1966;152:604–19.
[8] Dement W. The occurrence of low voltage, fast,
electroencephalogram patterns during behav-
ioral sleep in the cat. Electroencephalogr Clin
Neurophysiol 1958;10:291–6.
[9] Jouvet M, Michel F, Courjon J. Sur un stade d’ac-
tivite electrique cerebrale rapide au cours du
sommeil physiologique. C R Seances Soc Biol
Fil 1959;153:1024–8 [in French].
[10] Hodes R, Dement W. Depression of electrically
induced reflexes (H-reflexes) in man during low
voltage EEG sleep. Electroencephalogr Clin Neu-
rophysiol 1964;17:617–29.
[11] Jouvet M, Mounier D. Effects des lesions de la
formation reticulaire pontique sur le sommeil
156 Dement
du chat. C R Seances Soc Biol Fil 1960;154:
2301–5 [in French].
[12] Vogel G. Studies in psychophysiology of dreams.
III: The dream of narcolepsy. Arch Gen Psychiatry
1960;3:421–8.
[13] Rechtschaffen A, Wolpert E, Dement W, et al.
Nocturnal sleep of narcoleptics. Electroencepha-
logr Clin Neurophysiol 1963;15:599–609.
[14] Dement W, Rechtschaffen A, Gulevich G. The na-
ture of the narcoleptic sleep attack. Neurology
1966;16:18–33.
[15] Burwell CS, Robin ED, Whaley RD, et al. Extreme
obesity associated with alveolar hypoventilation:
a pickwickian syndrome. Am J Med 1956;21:
811–8.
[16] Lugaresi E, Coccagna G, Mantovani M. Hyper-
somnia with periodic apneas. New York: Spec-
trum; 1978.
[17] Young T, Palta M, Dempsey J, et al. The occur-
rence of sleep-disordered breathing among
middle-aged adults. N Engl J Med 1993;328:
1230–5.

The Contribution of the Psychology
of Sleep and Dreaming
to Understanding
Sleep-Disordered Patients
Rosalind Cartwright, PhD, FAASM
- A brief history
- The first twenty years
- The focus on sleep disorders
- Current theories of dreaming and the
psychology of sleep circa 2007
- Non–rapid eye movement parasomnia:
sleepwalking
- What makes a sleepwalker walk?
- Sleepwalking in obstructive sleep apnea
This article reviews the major psychologic func-
tions occurring during sleep and the evidence that
patients who present with disorders of sleep have
specific psychologic dysfunctions related to the
type and degree of their sleep disturbance. Although
there are individual differences in the degree to
which people are aware of sleep thoughts and
dreams, mental activity is actually continuous
throughout the waking, sleeping, and dreaming
cycles. The neurologic mechanism responsible for
this is the cooperation recently identified between
the slow wave sleep (SWS) of the first hour and
the following rapid eye movement (REM) sleep.
As sleep begins, the neural networks that were
activated during the waking experience are reacti-
vated and remain active until REM sleep occurs.
At that point, the coded representation of that expe-
rience is matched to similar older memories, and
this combination is displayed as a dream. This is
the explanation of how normal sleep is basic to
Department of Behavioral Sciences, Rush University Medical Center, 1653 West Congress Parkway, Chicago,
IL 60612, USA
E-mail address: rcartwri@rush.edu
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
157
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 157–166
- Rapid eye movement parasomnia:
nightmares
- Dream enactment in rapid eye movement
behavior disorder
- Major depression
- Summary
- Future research
- References
‘‘learning’’ the overnight improvement in mood
and performance. The organization of new experi-
ence, and retention of it in long-term memory,
allows waking behavior to be more adaptive. The
selection of which experiences are to be saved
relates to their emotional relevance to the self-
system. The replay of these initiates a hippo-
campal-neocortical dialog during which negative
emotion, generated in response to experience that
challenges the sense of self, is downregulated in
an across-the-night, sequential processing. If there
is a persistent sleep disorder the learning process
and mood regulation process is less efficient or
stopped altogether. The study of sleep disorders
provides insights into this regular nocturnal updat-
ing of the self-program, and of the specific psycho-
logic effects of interrupted sleep or abnormalities
in the timing of the cycles. The application of these
findings to understanding and treating these
patients is discussed.
doi:10.1016/j.jsmc.2008.01.002
158 Cartwright
A brief history
Academic psychology did not recognize dreaming
as a legitimate area of study until the early 1950s,
when publications from Kleitman’s laboratory
announced there is a regularly occurring active
brain state in sleep, dubbed ‘‘rapid eye movement’’
sleep, which is a reliable indicator of when a distinc-
tive type of mental behavior, dreaming, is ongoing
[1,2]. A way was opened to access the psychology
of sleep on-line.
At that time psychology was still struggling to be
recognized as an independent empiric science
devoted to the understanding of human behavior
through the methods of hypothesis testing and con-
trols. As a new science, psychology staked out the
areas that defined its turf: learning and memory,
motivation, and emotion. The idea that these basic
aspects of behavior were also operating during sleep,
or that they were affected by sleep, was not then con-
sidered credible. After all, the sleeper was uncon-
scious and experimental psychologists could
neither observe their mental activity directly, nor
could they rely on the sleepers’ morning memory to
be a reliable data source. As a result, the formal study
of psychology, with few exceptions, ignored the 8
hours of sleep until the methods of sleep recording
made this investigation scientifically feasible.
Freud’s Interpretation of Dreams [3], originally
published in 1900, called attention to the impor-
tance of the unconscious as a strong influence on
the motives that drive behavior and the accompa-
nying emotions. Freud not only illustrated the
many ways the unconscious plays a role in waking
behavior, but he pointed to dreaming as the path
by which one could tap into this important infor-
mation [3]. Armed with the guidepost of the strong
correlation between REM sleep and dreaming, psy-
chologists set out to test the validity of Freud’s
premise, that adding information from dreaming
enhances one’s understanding of human behavior
and why it is that even some smart people do not
‘‘learn from experience’’ and continue to make
self-destructive behavior choices.
The first twenty years
Despite the drawbacks of the unnatural situation of
sleeping in a laboratory, early studies of dreaming
managed to address important questions: At what
age does the child begin to dream and how does
that relate to their waking cognitive development?
[4]; Are dreams the normal equivalent of the hallu-
cinations of psychosis? [5]; Do dreams have inher-
ent meaning or is this added to random images as
an awakening takes place [6]; Do dreams tell a con-
nected story from REM to REM across the night?
[7]; Why do some dreams repeat over and over?
[8]. Some of the more difficult questions addressing
the major topics of psychology (what is the role of
sleep in learning, memory, and motivation; and is
sleep a place where emotion is regulated [9])
remained controversial and only recently have
been addressed with more advanced methodology.
The focus on sleep disorders
Something else occurred to divert psychologists
from the basic studies. The years 1975 to 2000
saw the emphasis shift from research into the
nature and function of normal sleep to that of
pathology. That in turn led to the advent of special-
ized clinical facilities to diagnose and treat those
whose sleep was dysfunctional. This development
resulted in a tremendous increase in public aware-
ness of the role of sleep in both physical and mental
health and an expansion of the disciplines involved
in the applied area of clinical sleep medicine [10].
Studies of the mind were left behind; the very
existence of mind was denied. It was deemed a fuzzy
concept no longer necessary in the age of brain
imaging. Some clinical research on the psychologic
aspects of sleep and dreaming continued, however,
and is now ready to be integrated to better the
understanding and treatment of patients who pres-
ent with problems of sleep.
Current theories of dreaming and the
psychology of sleep circa 2007
Although early studies had noted that there was
some mental activity that could be retrieved from
sleep onset and in sleep stages other than REM
[11], most of the research of the first 50 years was
focused on REM-related dreaming. This article
begins with the legacy from that work, the consen-
sus on one major dream function, before moving to
the newer view of the influence of waking states of
mind on the mental activity of non-REM (NREM)
sleep and on the following dream production in
REM and the effect of sleep on tomorrow’s behav-
ior, the full 24-hour perspective on continuous
mental activity.
Many of the research-based investigators of
dreaming [12–17] all place emotion as the key to
understanding dreams and their relation to why
one generally feels better in the morning. This is
referred to as the ‘‘mood regulatory function of
dreaming.’’ Hartmann [13] states: ‘‘overall dreaming
makes connections more broadly than waking in
nets of the mind, and the connections are not
made randomly but guided by the dreamer’s emo-
tional concerns.’’ Breger [14] links presleep emotion
to postsleep waking adaptation: ‘‘(dreaming) serves
 The Contribution of the Psychology of Sleep and Dreaming
a unique function in the assimilation and mastery
of aroused material into the solutions embedded
in existing memory systems.’’ From my own research
[17] I have concluded that ‘‘dreaming has an active
self-regulatory role in emotional modulation.’’
How this self-regulating dream function adds to
an understanding of four sleep disorders is explored
next: sleepwalking (an NREM parasomnia); night-
mares and dream enactment (both are REM para-
somnias); and major depression (insomnia
associated with a mental disorder). These illustrate
how specific deficits in the integrity of sleep relate to
failures to regulate mood and to change behavior.
Fig. 1 is a model of mental activity throughout
night, drawn from reports from normal volunteers
awakened in various sleep stages at different times
of the night. Freud identified three parallel levels of
ongoing mental activity: (1) conscious, (2) precon-
scious, and (3) unconscious. If one superimposes
the results from awakening studies, the model pre-
dicts that aspects of the presleep waking thoughts
(level 1) are picked up and mingled with the lon-
ger-term preconscious emotional concerns of the
sleeper (level 2). The mental activity retrieved from
the first NREM episode depends on the strength of
the emotion evoked in waking and carried into sleep
by the reactivation mechanism. As REM sleep is
turned on, this activation is ‘‘mapped onto previously
stored memories’’ (level 3), matched to a neocortical
network with similar feelings. As the first REM period
ends and NREM sleep is resumed, some memory bits
continue to be carried forward into the next NREM
sleep. This process continues throughout the night.
The evidence supporting this continuity model has
recently been summarized by Ribeiro and Nicolelis
[18] and Giuditta and coworkers [19] and is next ex-
amined as it applies in several sleep disorders.
Non–rapid eye movement parasomnia:
sleepwalking
The continuation of motives and emotions from
waking into the first cycle of SWS can be seen
159
most clearly in the behavior of adult sleepwalkers.
This disorder has been difficult to study in the lab-
oratory because of its unpredictability and lack of
any distinctive marker of its presence using the stan-
dard sleep stage scoring of the polysomnogram
(PSG) [20]. What is well established is that the
sleepwalker is not acting out a dream. The walk
actually begins with an abrupt arousal out of SWS
in the first third of the night, not out of REM sleep
[21]. Nor can the patient clarify why they were
doing what they did, because they have little or
no memory of the event. At best they report a sense
of urgency. Although this disorder is relatively com-
mon in young children with an equal prevalence in
both genders, the frequency of these episodes
wanes or stops all together during adolescence as
SWS is reduced in amount and in the amplitude
of the delta waves. There can be a resurgence of
this disorder in young adulthood under certain
conditions, however, and when it does the preva-
lence is much higher in males.
What makes a sleepwalker walk?
The strongest predisposing factor is genetics. Those
with a family tree loaded with many relatives
affected with sleepwalking or sleep terrors are at
increased risk for this disorder [22]. Twin studies
in Sweden [23] find that monozygotic twins are
concordant for sleepwalking at six times the rate
of dyzygotic twins. Genetic testing by HLA typing
shows sleepwalkers more often display a DQB1
marker than do controls [24], although this marker
is also found in those affected by narcolepsy and
REM behavior disorder. All three of these disorders
have in common some motor control dysfunction.
More promising is new evidence that scoring the
PSG by spectral analysis for the presence of delta
frequency activity rather than by the usual sleep
stage criteria finds sleepwalkers differ from controls
in having less delta activity at the beginning of the
night [25–27]. Fig. 2 shows the reduced delta activ-
ity in the first sleep cycle of sleepwalkers compared
Fig. 1. R. Cartwright graph.
160 Cartwright
Fig. 2. The reduced delta activity in the first sleep cycle of sleepwalkers compared with controls. (From Guille-
minault C, Poyares D, Abat F, et al. Sleep and wakefulness in somnambulism: a spectral analysis study. J Psycho-
som Res 2001;51:411–16; with permission.)
with controls [25]. Another feature of the PSG is
that sleepwalkers have multiple arousals from delta
sleep in the first two sleep cycles, making this sleep
time very unstable.
An episode of adult sleepwalking typically begins
with an abrupt arousal during this unstable first
hour of sleep, when the brain is simultaneously
partially asleep and partially awake. One imaging
study caught a 16-year-old habitual sleepwalker
shortly after he had a behavioral arousal during
a PSG study [28]. The single-photon emission CT
showed a 25% increase in blood flow in the poste-
rior cingulate cortex and decreases in the frontopar-
ietal cortices. In this mixed state, a sleepwalker is
able to perform complex motor behaviors, includ-
ing driving a car sometimes for long distances, but
is not able to recognize faces, even those of loved
ones. This suggests that the area of the brain identi-
fied with face recognition, the fusiform gyrus, is not
active [29]. Once an episode is over, which may take
as long as an hour, the sleepwalker typically returns
to sleep spontaneously, although this may be in
a strange or even dangerous location.
If startled by a touch or sound while in this state
a sleepwalker may become swiftly aggressive, how-
ever, behaving as if they are in a fight or flight sur-
vival mode. In a large epidemiologic study the rate
of those adults reporting that they currently have
episodes of aggressive sleepwalking was 2.1%,
much higher than expected [30]. These vary in sever-
ity from benign to lethal, such as jumping out a win-
dow or thrusting an arm through a glass door.
There are many behaviors other than aggression
that have been recently reported performed without
conscious awareness, or any later memory, follow-
ing an arousal from SWS within the first 3 hours
of sleep. Some sleepwalkers eat, and even prepare
food; often these are strange combinations of
foods, such as raw bacon with chocolate bar sand-
wiches. Others attempt to rescue someone from
an imaginary danger, such as pulling their wife
out of bed because the mattress is thought to be
on fire. Recently, there have been patients who
have presented because of inappropriate sexual
behavior, such as lying on top of a sleeping child.
Exploring new territory is another sleepwalking
behavior, as it was in the 15 year old who climbed
a 130-foot crane. All of these are instances of basic
motives necessary for our survival: eating, procreat-
ing, fighting or fleeing in response to danger, pro-
tecting the family, and exploring new territory. All
show these remain active during SWS, the realm
of the unconscious. If not for the arousal that aborts
the transmission into REM, the activation of these
drives would most likely be expressed in dream im-
agery and down-regulated by morning.
In addition to a genetic predisposition, it takes
added factors for an overt episode of sleepwalking
to occur. The most powerful of these is sleep depri-
vation. When a genetically vulnerable person has
inadequate amounts of sleep over an extended
period of weeks, because of either an internal or
external arousing stimulus, sleepwalking is highly
likely to occur. Sleep deprivation increases the drive
for more SWS. Recently this has been used to stim-
ulate sleepwalking under laboratory conditions.
The study required both sleepwalkers and controls
to undergo 25 hours of sleep deprivation before
 The Contribution of the Psychology of Sleep and Dreaming
a PSG test. All the sleepwalkers had one or more
episodes, whereas none occurred in the controls
[31]. Under home sleep conditions, the degree of
sleep loss patients report having before a sleepwalk-
ing episode occurs is usually more chronic. This
protocol, a combination of acute sleep deprivation
beforehand, videotaping of any events during the
sleep study, and power density scoring of the PSG,
can now be used to confirm this diagnosis in the
laboratory. This will be important in future forensic
cases when sleepwalking is invoked by the defense
in the case of a serious assault.
In one such case, which received widespread me-
dia attention [32], PSG studies were ordered by the
court, but were conducted before the publication of
the work showing the usefulness of power density
scoring and the triggering effect of prior sleep dep-
rivation. Although the studies demonstrated that
K.P., a tall, well built, 23 year old, had very little
delta sleep and many abrupt arousals whenever
SWS was reached, he made no attempt to sleepwalk
in the laboratory and so these tests were inconclu-
sive. There was, however, other evidence that sup-
ported the argument that he may have been
sleepwalking. K.P. had a strong family history of
sleepwalking detailed by a court-appointed expert,
and was known to be a sleepwalker as a child. He
was also experiencing an ongoing major stress
that resulted in extended loss of sleep, and he no
memory for the event that brought him to trial.
The back story of this case reveals the psychology
of this event. K.P. a high school graduate, married,
with a new first baby, had begun to gamble on
horse races and quickly found himself in serious
debt. Without his wife’s knowledge he continued
to bet and lose many thousands of dollars. He emp-
tied their joint bank account, took out an addi-
tional home loan, and then embezzled from his
employer. When this was discovered he was fired.
Over the next 4 months he became increasingly im-
mobilized; he stopped seeking work and seeing his
friends; and did not visit his in-laws, with whom he
had a close relationship. His mother-in-law called
him her Gentle Giant. His sleep quality was so
poor he often did not go to bed at all but stayed
on the couch watching television.
The violent episode started late on a Saturday
night. That afternoon he and his wife had quarreled
about his behavior. She insisted he seek help and
that he accompany her the following day for a Sun-
day visit to her parents. This would mean revealing
his financial crisis to them, which he dreaded and
had avoided. His in-laws were modest people, not
able to help him financially, and had trusted him
to care for their daughter. Now that he was unable
to find work and was at risk of losing their home,
he feared he would be seriously diminished in their
161
eyes. He did promise his wife that he would go with
her next day. That night he fell asleep on the couch
while watching television and, as he reported, ‘‘I
woke up over the body of a woman.’’ He did not rec-
ognize this as his mother-in-law, whom he had
stabbed to death, nor did he remember that he
had driven 15 km to their house, attacked his sleep-
ing father-in-law, and killed the mother-in-law with
a knife from her kitchen.
When I interviewed K.P. while he awaited trial, he
asked me in a bewildered voice really wanting an
answer: ‘‘Why would I do that, when I had every-
thing to lose and nothing to gain?’’ He was right:
this episode made no rational sense. It was an emo-
tionally driven, complex set of responses to an iden-
tity crisis. He fell asleep with the visit on his mind.
He was determined to go, to carry through on his
promise to his wife, but once there was the threat
of exposing himself to her parents as a failure.
This prospect mobilized his drive to survive by
attacking those who would destroy a positive part
of his self-system. He would no longer be the
beloved Gentle Giant. The jury’s task was to decide
if the ensuing aggression was committed when he
was in a nonconscious state. Their verdict was to
acquit on the grounds that he was in a state of ‘‘non-
insane automatism.’’ Because his arousal from
NREM sleep aborted REM, he lost the opportunity
for dreaming to perform a dampening of his rage,
and no mood regulation took place.
Sleepwalking in obstructive sleep apnea
Cases of NREM parasomnia precipitated by the
sleep loss associated with obstructive sleep apnea
are not common but three cases have been pub-
lished; two of these involved serious violence
[33–35]. In all three the obstructive sleep apnea
was assessed to be severe by laboratory studies
and also showed the reduced delta and increased
number of arousals typical of obstructive sleep ap-
nea and of patients diagnosed with NREM para-
somnia. In the case published by Nofzinger and
Wettstein [33] a man shot his wife to death follow-
ing an arousal from sleep. In this case the sleep ex-
pert testified for the prosecution. He argued that
sleepwalking was unlikely because obstructive sleep
apnea is common but there had been no reports of
the disrupted sleep being associated with violence.
Also it did not help the defendant’s case that he
may have had conscious ‘‘motivation’’ for this act.
The police found a note written by the wife in her
handbag stating that she intended to leave him.
The jury found him guilty and he was convicted.
The second case [34] was one of a night terror dur-
ing a PSG study while continuous positive airway
pressure was being titrated. A return of SWS in which
162 Cartwright
a desaturation occurred prompted an abrupt
arousal. The patient jumped up and stood on the
bed in a state of fear but no aggression was involved.
In the third case [35], a 54-year-old woman with a
5-year history of weight gain and excessive daytime
sleepiness awoke one morning to find her hands
bloody. She then found more blood on a kitchen
cutting board. The remains of her cat were found
by the trash can. She was diagnosed with severe
obstructive sleep apnea and treated with continuous
positive airway pressure and had no further epi-
sodes. What her motivation was for dissecting the
cat could not be explored because she lived alone
and had complete amnesia for this act. A desatura-
tion episode may have triggered an arousal into
a sleepwalk to the kitchen where perhaps the cat in-
terrupted her by a touch or sound, but that is a guess.
Two other studies indicate that mild breathing
disorders are more common in those who have
a history of sleepwalking than they are in matched
controls [36,37]. In the Guilleminault study [37],
a surgical intervention to improve the respiration,
a tonsillectomy, or opening the nasal passages elim-
inated the sleepwalking in a group of children with
a history of sleepwalking.
Rapid eye movement parasomnia:
nightmares
Patients suffering from chronic nightmares present
another opportunity to examine the impact of
a sleep disorder on the psychologic function of
dreaming. In contrast to the NREM parasomnias,
the definition of nightmares includes an abrupt
arousal from REM sleep, usually toward the end
of the night, into a fully awake state with a clear
memory of a highly emotional dream. The dream
scenario usually involves a fear-invoking situation
in which the dreamer has a strong sense of power-
lessness. This may explain why they are more
common in young children and become less fre-
quent as youngsters acquire more coping skills.
Adult nightmares have been studied extensively
where they are found to be associated with psycho-
pathology [38–41]. Nightmare frequency increases
suicide potential in the depressed [42], and is
a defining symptom of posttraumatic stress disor-
der following major disasters and personal traumas,
such as physical abuse and rape [43–46].
Epidemiologic studies show a high prevalence of
nightmares in adults, particularly in inner city
women. Between 2% and 6% of the general popu-
lation report experiencing nightmares at least
once per week. Here too, the family and twin stud-
ies [47] implicate a genetic basis for this disorder. A
variety of explanatory theories link the
neurobiology of fear to a psychologic trait of poor
capacity to regulate stress [48,49].
In a recent review [48] the authors propose there
is a common link between emotion-evoking dream
imagery and the profound lack of muscle tone char-
acteristic of REM sleep. Being scared to death, while
literally unable to move the major muscles, mimics
the situation in the behavioral treatment called ‘‘de-
sensitization.’’ In this program, visually imagining
the fear-evoking memory while awake is coupled
with training in physical relaxation [50]. The paral-
lel in sleep is that the exposure from REM to REM
over 1 or more nights to the feared stimulus while
the atonia prevents a motor response ‘‘wears out’’
the arousal response. In chronic nightmares, how-
ever, the negative affect exceeds the capacity of
REM to sustain sleep. This may be because there is
no match in memory to the current stressful event
to help disperse the affect load and an arousal oc-
curs instead that prevents completion of the dream.
It is the failure to complete the dream that is
responsible for the repeated bad dream scenario
rather than its extinction. Evidence supporting
this model comes from three sources.
1. Nightmares are more common in those identi-
fied as having anxiety disorders, neuroticism,
schizophrenia-spectrum symptoms, posttrau-
matic stress disorder, and maladaptive coping.
2. The usual flattening of emotion indicators dur-
ing fearful dreams (low heart rate, respiratory
rate, and low eye movement counts during
REM) are even lower in the sleep just before
a nightmare arousal [41].
3. The usual reduction in negative mood following
the morning awakening fails to occur for the
nightmare sufferer [49].
Dream enactment in rapid eye movement
behavior disorder
REM behavior disorder is another sleep disorder in
which there is a failure to maintain REM sleep.
These patients not only have abrupt arousals from
REM with vivid recall of a fearful dream, they also
act it out [51]. Swinging at imaginary intruders
they punch holes in the bedroom walls, upset
lamps, hurt themselves, and damage property. Their
sleep recordings show bursts of muscle activity in
the chin leads during REM sleep and other indica-
tors of a failure of motor control: bruxism (tooth
grinding), hypnic jerks, and periodic leg move-
ments are all common throughout the sleep in
the PSGs of these patients. When this disorder
was originally described it was most often identified
in the elderly who also had some neurodegenera-
tive disease, primarily Parkinson’s disease. Then,
 The Contribution of the Psychology of Sleep and Dreaming
younger patients meeting the diagnostic criteria be-
gan to be seen in sleep centers, when no clinical
Parkinson’s disease symptoms were apparent. Lon-
gitudinal follow-up of these patients showed it
might be decades later before the neurologic symp-
toms were documented in waking [52].
Although rare, a milder form of one REM behav-
ior disorder symptom called ‘‘dream enactment’’
has been documented in healthy young adults. It
is possible that these may also be at risk for the later
appearance of Parkinson’s disease. Again, there is
a mixed picture of psychologic precipitating factors
on top of a genetic vulnerability caused by a flaw in
the REM motor inhibition system. One young
patient, a recently married medical student, pre-
sented with a self-diagnosis of dream enactment
with a 3-month history. He described his sleep
episodes as now occurring three to six times per
night on 5 or 6 nights a week. A typical recent event
began with an abrupt sit up in bed in a fearful state
imagining an anaconda in the bed. He described
himself as screaming to his wife ‘‘you grab the tail,
I’ll grab the head.’’ His PSG recording showed mul-
tiple bursts of activity in the submental muscle and
periodic leg movements during REM sleep. Clearly,
he needed some insight into the added stress his
marriage just 3 months ago imposed while he was
preparing for his second-year board examination.
He also needed some help with stress management.
The breakthrough of muscle activity during REM is
usually successfully treated with a small dose of
clonazepam at bedtime. In the case of this patient
where the psychologic function of sleep was so
clearly disturbed the treatment also needed to
include the rules of good sleep hygiene, especially
the importance of avoiding sleep deprivation and
training in relaxation techniques.
Major depression
In major depression the sleep marker in the PSG is
in the timing of the onset of the first REM period of
the night. The number of minutes of sleep before
the first REM episode, normally about 90 minutes,
is most often reduced to less than 65 minutes dis-
placing the SWS [53] or is skipped lengthening
the first NREM cycle. This robust finding suggests
dreams too may be abnormal. In a series of studies
volunteers, all of whom were undergoing the same
negative emotional experience (the breakup of
a first marriage), were followed longitudinally
tracking the early REM marker and their dreams.
The aim was to examine whether those who met
depression criteria would restore mood regulation
over time (wake in a better morning mood and
remit from depression) following a change in
dream scenarios. The focus was on two dream
163
variables: the expression of negative emotion and
the inclusion in the dream story aspects of both
the present life event and older memory images
[54]. Twenty men and women who met depression
criteria and 10 nondepressed divorcing controls all
had their dreams collected from each REM period
on three occasions over a 5-month period. At the
follow-up visit, 12 of the depressed were in remis-
sion without any intervening psychotherapy or
pharmacologic treatment and 8 remained de-
pressed. All controls remained free of depression
symptoms.
The major difference between the depressed who
improved and those who did not was the degree to
which the dreams integrated fragments of the recent
emotional experience with older memories relating
to the same emotion associated with the divorce
[54]. One example of this is a dream reported by a de-
pressed woman who when awakened from REM
gave this report: ‘‘My ex-spouse appeared at my par-
ent’s home where I was having my 16th birthday
party. He embarrassed me by exposing himself.’’
The dreams of those who were in remission by the
end of the study were longer than those who did
not change. They were also more complex and in-
cluded multiple characters and shifts of scene. This
heightened dream complexity was characteristic of
their REM reports from the start of the study [55].
Their remission could be predicted from the
dream-like quality of their reports on the first night
of REM awakenings. They seemed to be putting
things together in new ways,‘‘changing their minds’’
during sleep. Those who failed to remit without
treatment had short rather stark dreams or no recall
at all. These are the depressed that require some
treatment intervention, antidepressant medication,
psychotherapy, or a therapy directed to dream
change [56].
Summary
This article has looked into how one psychologic
function of sleep applies to expand the understand-
ing of a number of sleep disorders. Overall, this
function integrates new waking experience relevant
to the organized self-system, and modulates nega-
tive emotion invoked by experience that threatens
this program. The reactivation mechanism supports
the update of the ‘‘underlying strategies that guide
behavior’’ [57] and thereby prepares for more ap-
propriate responses to any challenges the next
day. In sleepwalking the arousal out of SWS aborts
REM and so stops this process at the beginning of
the night’s sleep. In nightmare disorder the reactiva-
tion matches some current experience to a reminder
of an earlier threat to the self-structure. The dream
constructed in the late night sleep, when REM is
164 Cartwright
most intense, has the additive effect of the old and
new threats leading to a spontaneous awakening
delaying any modification of the underlying self-
system. In dream enactment dreams are acted out
in response to an overload of new challenges to
a system with a genetic deficit in sustaining sleep
motor atonia. All these are examples of a failure
to sustain sleep because of a high level of some dis-
turbing affect in those with an inherited biologi-
cally weakened motor control system either in
early NREM sleep, late night REM sleep, or both.
In major depression the problem is not primarily
one of sustaining sleep but of its timing. The early
REM displaces the SWS responsible for the reactiva-
tion of relevant waking experience. The REM dream
content is either empty or at best simple and bland
and is not functional for change in morning mood.
Future research
There is still much work to be done to understand
with more precision all of the interactions of brain
and behavior in the different organizational states
humans cycle through each 24-hour period.
1. The evidence of a role for genetics in several
sleep disorders needs further work. As yet there
has been no application of the newer SNIPS
technique to differentiate between narcolepsy,
REM behavior disorder, and sleepwalking, all
of which have a common DBQ1 marker using
HLA typing.
2. There is likely a genetic differentiation within the
group of adult sleepwalkers between those who
are and those who are not violent [58]. This
should be addressed with better genetic testing.
3. The question of why the male gender is so dom-
inant in adult sleepwalkers also is a topic for fur-
ther study.
4. Besides the evidence that sleep deprivation and
zolpidem are precipitating factors in NREM par-
asomnias, there is controversy about the role of
substances, such as caffeine and alcohol
[59–61], which lighten the first NREM sleep
and so challenge the old model that it is in-
creased SWS with high thresholds for arousal
that are responsible for this disorder rather than
low thresholds and the reduced delta activity. Fur-
ther work exploring the effect of these and other
sleep medications is needed to test their potential
as triggers for sleepwalking in prone individuals.
5. There is also controversy about the continuity
of cognition, motivation, and emotion thro-
ughout the sleep-wake cycle based on the data
from sleep interruptions across the night. This
method may induce continuity because of the
memory of what was reported when awakened.
New protocols are needed to control for this
experimental artifact.
6. There is evidence that there are individual differ-
ences in both the speed and amount of over-
night improvement in behavior within normal
subjects related to sleep variables other than
number of arousals or timing of cycles. The eye
movement density is one predictor of learning
ability [62]. Further work on the role of more
specific sleep variables, such as sleep spindles
and K complexes, on learning and memory is
indicated [63].
These are a few of the topics for future research to
clarify the proposition put forth here: that when
sleep is intact, of adequate length, and undisturbed
by abnormal arousals, information from waking
continues to be actively carried forward through
neural circuits allowing it to be sorted, stored, and
tempers cooled. Sleep likely performs several psy-
chologic functions. This article focuses on the REM
and dream function, for which there is the most
data. This explores how sleep processes waking expe-
rience that has a negative emotional impact. Recent
work studying how more neutral experimental
learning tasks effect sleep and subsequent perfor-
mance is providing information about specific
changes in sleep characteristics that take place as
one learns. Correcting sleep disorders has the poten-
tial of restoring the neuropsychologic system to the
fine balance between stability and flexibility of
behavior characteristic of humans at their best.
References
[1] Aserinsky E, Kleitman K. Regularly occurring pe-
riods of eye motility and concomitant phenom-
enia during sleep. Sciences 1953;118:273–4.
[2] Dement W, Kleitman N. Cyclic variations in EEG
during sleep and their relation to eye move-
ments, body motility and dreaming. Electroence-
phalogr Clin Neurophysiol 1957;9:673–90.
[3] Freud S. The interpretation of dreams. (English
translation). New York: Basic Books; 1954. p. 608.
[4] Foulkes D. Children’s dreams: longitudinal stud-
ies. New York: John Wiley and Sons; 1982.
[5] West LJ. A general theory of hallucinations and
dreams. In: West LJ, editor. Hallucinations. New
York: Grune and Stratton; 1962. p. 275–91.
[6] Hobson JA, McCarley R. The brain as a dream
state generator: an activation-synthesis hypothe-
sis of the dream process. Am J Psychiatry 1977;
134:1335–48.
[7] Rechtschaffen A, Vogel G, Shaikun G. Interrelat-
edness of mental activity during sleep. Arch
Gen Psychiatry 1963;9:536–47.
[8] Cartwright R. The nature and function of repeti-
tive dreams: a survey and speculation. Psychiatry
1979;42:131–7.
 The Contribution of the Psychology of Sleep and Dreaming
[9] Kramer M. The selective mood regulatory func-
tion of dreaming: an update and revision. In:
Moffitt A, Kramer M, Hoffman R, editors. The
functions of dreaming. Albany (NY): State Uni-
versity of New York Press; 1993. p. 139–95.
[10] Cohen H, Altevogt B, editors. Sleep disorders and
sleep deprivation: an unmet public health prob-
lem. Washington, DC: National Academies Press;
2006.
[11] Foulkes D. Dream reports from different stages of
sleep. J Abnorm Soc Psychol 1962;65:14–25.
[12] Foulkes D. Dreaming: a cognitive-psychological
analysis. Hillsdale (NJ): Erlbaum; 1985.
[13] Hartmann E. Dreaming. In: Lee-Chiong T,
Sateia M, Carskadon M, editors. Sleep medicine.
Philadelphia: Hanley and Belfus; 2002. p. 223.
[14] Breger L. Function of dreams. J Abnorm Psychol
1967;72:1–28.
[15] Reiser M. The dream in contemporary psychiatry.
Am J Psychiatry 2001;158:351–9.
[16] Kramer M. The dream experience: a systematic ex-
ploration. New York: Routlege; 2007. p. 167–87.
[17] Cartwright R, Young M, Mercer P. The role of
REM sleep and dream variables in the prediction
of remission from depression. Psychiatry Res
1998;80:249–55.
[18] Ribeiro S, Nicolelis M. Reverberation, storage,
and post-synaptic propagation of memories dur-
ing sleep. Learn Mem 2004;11:686–96.
[19] Giuditta A, Mandile P, Montagnese P, et al. The
role of sleep in memory processing: the sequen-
tial hypothesis. In: Maquet P, Smith C,
Stickgold R, editors. Sleep and brain plasticity.
New York: Oxford Press; 2003. p. 157–78.
[20] Cartwright R. Sleep related violence: does the
polysomnogram help establish the diagnosis?
Sleep Med 2000;5:25–32.
[21] Broughton R. Sleep disorders: disorders of
arousal? Science 1968;159:1070–8.
[22] Kales A, Soldates C, Bixler E. Hereditary factors
in sleepwalking and night terrors. Br J Psychiatry
1980;137:111–8.
[23] Hublin C, Kaprio J, Partinen M. Prevalence and
genetics of sleepwalking: a population based
sleep study. Neurology 1997;48:177–81.
[24] Lecendreux M, Bassetti C, Dauvilliers Y. HLA and
genetic susceptibility to sleepwalking. Mol Psy-
chiatry 2003;8:114–7.
[25] Guilleminault C, Poyares D, Abat F, et al. Sleep
and wakefulness in somnambulism: a spectral
analysis study. J Psychosom Res 2001;51:411–6.
[26] Espa F, Ondze B, Deglise P, et al. Sleep architec-
ture, slow wave activity and sleep spindles in
adult patients with sleepwalking and sleep ter-
rors. Clin Neurophysiol 2000;11:929–39.
[27] Gaudreau H, Joncas S, Zadra A, et al. Dynamics
of slow wave activity during the NREM sleep of
sleep walkers and control subjects. Sleep 2000;
23:755–60.
[28] Bassetti C, Vella S, Donati S, et al. SPECT during
sleepwalking. Lancet 2000;356:484–5.
165
[29] Cartwright R. Sleepwalking violence: a sleep dis-
order, a legal dilemma and a psychological chal-
lenge. Am J Psychiatry 2004;161:1149–58.
[30] Ohayon M, Guilleninault C, Priest R. Night ter-
rors, sleep walking and confusional arousals in
the general population: their frequency and rela-
tionship to other sleep and mental disorders.
J Clin Psychiatry 1999;60:268–76.
[31] Pilon M, Zadra A, Adam B, et al. 25 hours of
sleep deprivation increases the frequency and
complexity of somnambulistic episodes in adult
sleepwalkers. Sleep 2005;28:A257.
[32] Broughton R, Billings R, Cartwright R, et al. Ho-
mocidal somnambulism: a case report. Sleep
1994;17:915–25.
[33] Nofzinger E, Wettstein R. Homicidal behavior
and sleep apnea: a case report and a medicolegal
discussion. Sleep 1995;18:776–82.
[34] Pressman M, Meyer T, Kendrick-Mohamed J.
Night terrors in an adult precipitated by sleep ap-
nea. Sleep 1995;8:773–5.
[35] Lateef O, Wyatt J, Cartwright R. A case of
non-REM parasomnia that resolved with treat-
ment of obstructive sleep apnea. Chest 2005;
28:461S.
[36] Espa F, Dauvillers Y, Ondze B. Arousal reactions
in sleep-walking in adults: the role of respiratory
events. Sleep 2002;25:871–5.
[37] Guilleminault C, Palombini L, Pelayo R, et al.
Sleepwalking and sleep terrors in prepubertal
children: what triggers them? Pediatrics 2003;
111:17–25.
[38] Spoormaker VI, Schredl M, Bout JV. Nightmares:
from anxiety symptom to sleep disorder. Sleep
Med Rev 2005;10:19–31.
[39] Levin R. Nightmares and schizotypy. Psychiatry
1998;61:206–16.
[40] Zadra A, Donderi DC. Nightmares and bad
dreams: their prevalence and relationship to
well-being. J Abnorm Psychol 2000;109:273–81.
[41] Nielsen T, Zadra A. Nightmares and other com-
mon dream disturbances. In: Kryger M, Roth T,
Dement W, editors. Principles and practice of
sleep medicine. 4th edition. Philadelphia: Elsev-
ier Saunders; 2005. p. 926–35.
[42] Agargun M, Cartwright R. REM sleep, dream var-
iables and suicidality in depressed patients. Psy-
chiatry Res 2003;119:33–9.
[43] Ohayon M, Morselli P, Guilleminault C. Preva-
lence of nightmares and their relationship to
psychopathology and daytime functioning in in-
somnia subjects. Sleep 1997;20:340–8.
[44] Mellman T, Pigeon W. Dreams and nightmares
in posttraumatic stress disorder. In: Kryger M,
Roth T, Dement W, editors. Principles and prac-
tice of sleep medicine. 4th edition. Philadelphia:
Elsevier Saunders; 2005. p. 573–8.
[45] Mellman T, David D, Bustemente V. Dreams in
the acute aftermath of trauma and their rela-
tionship to PTSD. J Trauma Stress 2001;14:
241–7.
166 Cartwright
[46] Wood J, Bootzin R, Rosenhan D. Effects of the
1989 San Francisco earthquake on frequency
and content of nightmares. J Abnorm Psychol
1992;101:219–24.
[47] Hublin C, Kaprio J, Partinen M. Nightmares,
familial aggregation and association with psychi-
atric disorders in a nation wide twin cohort. Am
J Med Genet 1999;88:329–36.
[48] Nielson T, Levin R. Nightmares: a new neurocog-
nitive model. Sleep Med Rev 2007;11:295–310.
[49] Kramer M. The nightmare: a failure of dream
function. Dreaming 1991;1:277–85.
[50] Perlis M, Nielsen T. Mood regulation, dream-
ing and nightmares: evaluation of a desensitiza-
tion function for REM sleep. Sleep 1993;3:
243–57.
[51] Schenck C, Bundlie S, Ettinger M, et al. Chronic
behavioral disorders of REM sleep: a new cate-
gory of parasomnia. Sleep 1986;9:293–308.
[52] Schenck C, Bundlie S, Mahowald M. Delayed
emergence of a parkinsonian disorder in 38%
of 29 older men initially diagnosed with idio-
pathic rapid eye movement sleep disorder. Neu-
rology 1996;46:388–93.
[53] Kupfer D, Foster F. Interval between the onset of
sleep and rapid eye movements as an indicator
of depression. Lancet 1972;11:684–6.
[54] Cartwright R, Agargun M, Kirkby J, et al. Relation
of dreams to waking concerns. Psychiatry Res
2006;141:261–70.
[55] Cartwright R, Baehr E, Kirkby J, et al. REM sleep
reduction, mood regulation and remission in
untreated depression. Psychiatry Res 2003;12:
159–67.
[56] Cartwright R, Lamberg L. Crisis dreaming. New
York: Harper Collins; 1992. [reprinted by iUni-
verse.com Inc. 2000, Lincoln, NB].
[57] Winson J. Brain and psyche: the biology of the
unconscious. Garden City (NJ): Anchor Press/
Doubleday; 1985.
[58] Moldofsky H, Gilbert R, Lue F, et al. Sleep-related
violence. Sleep 1995;18:731–9.
[59] Pressman M. Factors that pre-dispose, prime and
precipitate NREM parasomnias in adults: clinical
and forensic implications. Sleep Med Rev 2007;
11:5–30.
[60] Cartwright R, Pressman M. Factors that pre-
dispose, prime and precipitate NREM parasom-
nias in adults: clinical and forensic implications
[letter]. Sleep Med Rev 2007;11:327–9.
[61] Pressman M. Response to Rosalind Cartwright’s
letter to the editor. Sleep Med Rev 2007;11:329–33.
[62] Smith C, Nixon M, Nader R. Posttraining in-
creases in REM sleep intensity implicate REM
sleep in memory processing and provide a bio-
logical marker of learning potential. Learn
Mem 2004;11:714–9.
[63] Fogel S, Nader R, Cote K, et al. Sleep spindles
and learning potential. Behav Neurosci 2007;
121:1–10.

Insomnia: Prevalence, Impact,
Pathogenesis, Differential Diagnosis,
and Evaluation
Evelyn Mai, MD, Daniel J. Buysse,
- Insomnia prevalence
- Insomnia impact
Insomnia and psychiatric conditions
Insomnia and medical conditions
Socioeconomic impact of insomnia
- Insomnia pathogenesis
- Insomnia evaluation
Insomnia is the most common sleep disorder af-
fecting millions of people as either a primary or co-
morbid condition. Insomnia has been defined as
both a symptom and a disorder, and this distinction
may affect its conceptualization from both research
and clinical perspectives. Whether insomnia is
viewed as a symptom or a disorder, however, it nev-
ertheless has a profound effect on the individual
and society. The burden of medical, psychiatric,
interpersonal, and societal consequences that can
be attributed to insomnia underscores the impor-
tance of understanding, diagnosing, and treating
the disorder.
Insomnia prevalence
The prevalence of insomnia varies depending on the
specific case definition. Broadly speaking, insomnia
has been viewed as a symptom and as a disorder in
its own right. Insomnia has also been defined
by subtypes based on frequency; duration (acute
Supported by National Institutes of Health grants MH24652 and AG20677.
Sleep Medicine Institute, University of Pittsburgh, 3811 O’Hara Street, Pittsburgh, PA 15213, USA
* Corresponding author. Department of Psychiatry, E-1127 WPIC, 3811 O’Hara Street, Pittsburgh, PA
15213.
E-mail address: buyssedj@upmc.edu (D.J. Buysse).
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
167
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 167–174
MD*
Patient interview
Physical and mental status examination
Collateral sources interview
Objective data
- Summary
- References
versus chronic); and etiology. This picture is further
complicated by considerations of insomnia as
either a comorbid condition; as a symptom of
a larger sleep, medical, or psychiatric disorder;
or as a secondary disorder [1]. An illustration of
this idea is the overlap between insomnia and
depression. Do insomnia and depression coexist
in an individual as separate disorders? Is insomnia
only one symptom in the larger context of
depression? Did insomnia secondarily developed
as a distinct disorder from a primary depressive
disorder?
The three main diagnostic manuals, International
Classification of Sleep Disorders (ICSD-2) [2], Diag-
nostic and Statistic Manual (DSM IV-TR) [3], and
International Classification of Disease (ICD-10) [4],
vary in their approach to defining insomnia
(Box 1). ICSD-2 subdivides insomnia into descrip-
tive, etiologic categories. Examples include adjust-
ment insomnia (insomnia temporally related to
an identifiable stressor) and psychophysiologic
doi:10.1016/j.jsmc.2008.02.001
168 Mai & Buysse
Box 1: Insomnia diagnostic categories
ICSD-2 insomnia categories
Adjustment insomnia (acute insomnia)
Psychophysiologic insomnia
Paradoxical insomnia
Idiopathic insomnia
Insomnia caused by mental disorder
Inadequate sleep hygiene
Behavioral insomnia of childhood
Insomnia caused by drug or substance
Insomnia caused by medical condition
Insomnia not caused by substance or known
physiologic conditions, unspecified (nonor-
ganic insomnia)
Physiologic (organic) insomnia, unspecified
ICD-10 insomnia categories
Nonorganic insomnia
Nonorganic disorder of the sleep-wake
schedule
DSM-IV-TR insomnia categories
Primary insomnia
Insomnia related to axis I or II category
insomnia (increased arousal and conditioned sleep
difficulty) (Box 2) [2]. These categories also contain
insomnia caused by a mental disorder, substance,
or medical condition. The DSM IV-TR separates
out primary insomnia (insomnia symptoms associ-
ated with distress or daytime impairment) from
other ‘‘dyssomnias,’’ such as a breathing-related
sleep disorder [3]. ICD-10 uses the broadest ap-
proach, categorizing insomnia based on underlying
pathology: nonorganic insomnia and nonorganic
disorder of the sleep-wake schedule (see Box 2)
[4]. Duration of insomnia (at least 1 month of
symptoms) is noted in ICSD-2 and DSM IV-TR;
however, frequency of symptoms is broached only
in ICD-10.
As a result of these differences in insomnia case
definitions, estimates of insomnia prevalence have
varied widely, from 10% to 40% [5–12]. This prob-
lem is demonstrated by the findings of a prevalence
study from South Korea. When insomnia was
defined by frequency (symptoms occurring at least
3 nights per week), 17% of randomly selected sub-
jects from the population qualified for the diagno-
sis. If the symptom of difficulty maintaining sleep
was the defining factor, 11.5% of the sample was
affected. Using the more stringent criteria from
DSM-IV, however, 5% of the sample qualified for
the diagnosis [13]. Similar disparities were shown
in a prevalence study from France [14]. According
to a 2005 statement by the National Institutes of
Health, insomnia has a prevalence of 10% if the def-
inition necessitates daytime distress or impairment
Box 2: Insomnia definition
ICSD-2 general criteria for insomnia
1. A complaint of difficulty initiating sleep,
difficulty maintaining sleep, or waking
up too early or sleep that is chronically
unrestorative or poor in quality. In
children, the sleep difficulty is often
reported by the caretaker and may con-
sist of observed bedtime resistance or
inability to sleep independently.
2. The above sleep difficulty occurs despite
adequate opportunity and circumstances
for sleep.
3. At least one of the following forms of
daytime impairment related to the
nighttime sleep difficulty is reported by
the patient: fatigue or malaise; atten-
tion, concentration or memory impair-
ment; social or vocational dysfunction
or poor school performance; mood
disturbance or irritability; daytime sleep-
iness; motivation, energy, or initiative
reduction; proneness for errors or acci-
dents at work or while driving; tension,
headaches, or gastrointestinal symptoms
in response to sleep loss; concerns or
worries about sleep.
DSM-IV-TR criteria for primary insomnia
1. The predominant complaint is difficulty
initiating or maintaining sleep, or non-
restorative sleep, for at least 1 month.
2. The sleep disturbance (or associated day-
time fatigue) causes clinically significant
distress or impairment in social, occupa-
tional, or other important areas of
functioning.
3. The sleep disturbance does not occur ex-
clusively during the course of narcolepsy,
breathing-related sleep disorder, circadian
rhythm sleep disorder, or a parasomnia.
4. The disturbance does not occur exclu-
sively during the course of another
mental disorder (eg, major depressive
disorder, generalized anxiety disorder,
a delirium).
5. The disturbance is not caused by the
direct physiologic effects of a substance
(eg, a drug of abuse, a medication) or
a general medical condition.
ICD-10 criteria for nonorganic insomnia
A condition of unsatisfactory quantity or
quality of sleep, which persists for a consid-
erable period of time, including difficulty
falling asleep, difficulty staying asleep, or
early final wakening. Insomnia is a common
symptom of many mental and physical
disorders, and should be classified here in
addition to the basic disorder only if it dom-
inates the clinical picture.

[15]. Given all the information available, the prev-
alence of insomnia symptoms may be estimated
at 30% and specific insomnia disorders at 5% to
10% [16].
Several risk factors for insomnia have been iden-
tified. Female gender, advanced age, depressed
mood, snoring, low levels of physical activity,
comorbid medical conditions, nocturnal mictura-
tion, regular hypnotic use, onset of menses, previ-
ous insomnia complaints, and high level of
perceived stress have all been implicated as risk fac-
tors; the first three factors in particular (female
gender, advanced age, and depressed mood) are
consistent risk factors [7,17–22].
Precipitants of insomnia have also been studied.
Bastien and colleagues [23] examined precipitating
factors of insomnia and found that family, work or
school, and health events proved to be the most
common precipitants [23]. Another study of psy-
chosocial stressors in Japan demonstrated that em-
ployees with greater intragroup conflict and job
dissatisfaction had greater risk for insomnia [24].
Knowledge of both risk factors and possible pre-
cipitants of insomnia can help to guide the evalua-
tion and treatment of insomnia. Questions about
psychosocial stressors at home and at work in
high-risk individuals, such as those experiencing
depression or who are female or elderly, can help
to shape and direct patient care.
Insomnia impact
Insomnia and psychiatric conditions
An estimated 40% of individuals with insomnia
have a comorbid psychiatric condition [7,25]. In
a review of epidemiologic studies, Taylor and col-
leagues [26] found that insomnia predicted depres-
sion, anxiety, substance abuse or dependence, and
suicide [26]. The correlation between insomnia
and later development of depression within 1 to
3 years is particularly strong [27]. Johnson and col-
leagues [28] found that in a community sample of
adolescents, in 69% of cases insomnia preceded
comorbid depression, whereas an anxiety disorder
preceded insomnia 73% of the time [28]. In a large
group of subjects aged 15 to 100 years, insomnia
either appeared before (>40%) or at the same
time (>22%) as mood disorders. This study also
found that insomnia appeared at the same time as
(>38%) or after (34%) anxiety disorders [29].
As further evidence of morbidity, individuals
with insomnia complaints in the last year but with-
out any previous psychiatric history were shown to
have an increased risk of first-onset major depres-
sion, panic disorder, and alcohol abuse the follow-
ing year when compared with controls [30].
Furthermore, adolescents who completed suicide
Insomnia 169
were found to have higher rates of insomnia in
the week preceding death than community-control
adolescents [31,32].
Taken as a whole, these findings underscore the
impact of insomnia on the individual while sug-
gesting a possible relationship between insomnia
and psychiatric disorders. The nature of this rela-
tionship has yet to be established. Insomnia could
be an early symptom, part of a prodrome, of a de-
pressive or anxiety disorder. Similarly, insomnia
might also exist as a separate, comorbid disorder
that either gave rise to or developed from a psychiat-
ric condition. In either case the need to address
insomnia and psychiatric disorders together re-
mains important.
Insomnia and medical conditions
Associations between insomnia and a variety of
medical conditions have also been established. Tay-
lor and colleagues [33] found that in a community-
based sample chronic insomniacs reported more
heart disease, hypertension, chronic pain, and in-
creased gastrointestinal, neurologic, urinary, and
breathing difficulties. The converse was also shown
to be true, in which subjects with hypertension,
chronic pain, breathing, gastrointestinal, and
urinary problems complained of insomnia more
often than noninsomniacs [33]. Others have also
found increased odds ratios for insomnia in a vari-
ety of medical conditions, ranging from congestive
heart failure to hip impairment [34].
Ancoli-Israel [35] emphasized the different ways
that insomnia and chronic medical conditions
may relate to each other: sleep complaints may
function as a symptom of a disorder, such as con-
gestive heart failure and Cheyne-Stokes respiration
gastroesophageal reflux disease and increased
arousals. In other cases, insomnia may be a compo-
nent of the etiology of a disorder, such as diabetes
mellitus [35].
The connection between cardiovascular disease
and insomnia bears specific attention. After adjust-
ing for age and coronary risk factors, a risk ratio of
1.5 to 3.9 between difficulty falling asleep and cor-
onary heart disease has been demonstrated [36].
Men who experienced difficulty falling asleep were
also shown to have a threefold risk of death second-
ary to coronary heart disease [37].
The relationship between chronic pain and
insomnia is also of particular clinical relevance. In
one study, more than 40% of insomniacs reported
having at least one chronic painful physical condi-
tion. Moreover, chronic pain was in turn associated
with shorter sleep duration and decreased ability to
resume sleep following arousal [38]. Tang and
colleagues [39] found that 53% of chronic pain
patients had scores suggestive on the Insomnia
170 Mai & Buysse
Severity Index of clinical insomnia versus 3% of
subjects without pain [39].
Socioeconomic impact of insomnia
In addition to psychiatric and medical comorbid-
ities, insomnia is associated with substantial per-
sonal and societal consequences. One study that
examined the effect of insomnia on primary care
patients found insomniacs had double the number
of days with restricted activity because of illness
[11]. Another study showed that more insomniacs
rated their quality of life as poor (22%) compared
with subjects without any sleep complaints (3%)
[32]. Insomnia has also been shown to have a detri-
mental effect on health-related quality of life to the
same degree as chronic disorders, such as depres-
sion and congestive heart failure [40]. When the
economic costs that encompass health care use,
workplace effects of absenteeism, accidents, and
increased alcohol consumption secondary to
insomnia were considered, the annual cost was
estimated to be between $35 to $107 billion
a year [41,42]. Insomnia has not been found to be
associated with increased risk of death [43].
Health care use, as defined by increased office
visits and rates of hospitalization, is consistently
higher in insomniacs than in subjects without sleep
complaints [44,45]. The direct costs incurred
through inpatient, outpatient, pharmacy, and emer-
gency room usage are greater in insomniacs regard-
less of age [46]. An evaluation of the direct health
care costs of insomnia in 1995 placed estimates at
$13.9 billion in the United States and $2.1 billion
in France [47,48].
Function in the workplace is also negatively
affected. Insomniacs miss work twice as often as
good sleepers, with absenteeism particularly promi-
nent in men and blue-collar workers [49]. The extra
cost of work absenteeism secondary to insomnia,
through decreased productivity and salary replace-
ment, is then brought to bear on employers [50].
Insomnia pathogenesis
Insomnia is often believed to arise from a state of
hyperarousal. In the physiologic hyperarousal
model, an elevated level of alertness throughout
the day and night makes it difficult to sleep. In sup-
port of this theory, insomniacs have been found to
have an increased whole body metabolic rate when
compared with normal sleepers [51,52]. They also
score higher than normal sleepers on a Hyper-
arousal Scale, and even during the day when
complaining of fatigue, insomniacs still take a lon-
ger time to fall asleep [53,54].
On functional neuroimaging, insomniacs show
increased cerebral glucose metabolism during sleep
and wake states [55]. On electroencephalography, in-
somniacs demonstrate increased beta activity and
lower delta activity [56,57]. From an endocrine
perspective, insomniacs, like patients with major
depressive disorder, demonstrate corticotropin-
releasing factor hyperactivity, suggesting a role for hy-
pothalamic-pituitary-adrenal axis dysfunction [58].
Insomnia evaluation
The cornerstone of the insomnia evaluation is a de-
tailed history obtained during the patient interview.
Although the approach to the interview may vary
depending on the practitioner, key points should
be covered to ensure a thorough evaluation. Addi-
tional assessment tools, such as the sleep-wake
diary, actigraphy, and in specific cases polysomnog-
raphy, can supplement the information obtained in
the interview. A list of diagnoses and comorbid con-
ditions to consider during the insomnia evaluation
can be found in Box 3.
Patient interview
Detailed information about the nature of the com-
plaint is necessary, such as if insomnia is related to
sleep onset, sleep maintenance, early morning
awakening, nonrestorative sleep quality, or a combi-
nation of these problems. Information obtained
here may help to guide the diagnosis, such as
a sleep-onset complaint resulting from restless
legs syndrome as opposed to an early morning
awakening presenting as part of a depressive disor-
der. Additional information about the onset, course
and duration, current presentation, frequency,
severity, and precipitating or alleviating factors
also helps to define the problem. In particular, a life-
long course with an onset in the absence of medical
and psychiatric comorbidities may suggest a pri-
mary insomnia as opposed to a secondary insom-
nia that develops in late adulthood in the context
of chronic pain.
The sleep schedule, including bedtime, sleep la-
tency, number and length of nighttime awakenings,
sleep reinitiation time, wake time, time spent in
bed, and total sleep time, should be reviewed. A pa-
tient’s preferred bedtime may not coincide with
actual bedtime, as in a circadian rhythm disorder.
Similarly, nighttime awakenings caused by night-
mares from posttraumatic stress disorder as
opposed to awakenings from nocturia caused by
prostate enlargement suggest different disorders.
The daytime routine with a review of work sched-
ule, eating and exercise times, and duration and
timing of naps is also important. Eating and exer-
cise times that occur in close temporal relation to
bedtimes may inhibit the patient’s ability to fall
asleep. Moreover, naps of long duration that occur
 Insomnia 171

Box 3: Insomnia differential diagnosis and In this area, collateral report from family, teachers,
common comorbidities or coworkers may prove helpful if the patient is
unaware of the extent of his or her symptoms.
Medical conditions
Safety issues, such as the negative effect on driv-
Cardiovascular: congestive heart failure,
arrhythmia, coronary artery disease
ing and work performance in potentially hazardous
Pulmonary: chronic obstructive pulmonary areas, should be broached and may provide an
disease, asthma opportunity for patient education.
Neurologic: stroke, Parkinson’s disease, Sleep conditions and routines should be dis-
neuropathy traumatic brain injury cussed, such as the conditions of the room used
Gastrointestinal: gastroesophageal reflux for sleep (eg, effect of light, temperature, and
Renal: chronic renal failure noise); use of television, computer, or radio both
Endocrine: diabetes, hyperthyroidism in the prebedtime routine and during periods of
Rheumatologic: rheumatoid arthritis, nighttime awakenings; the effect of anxiety during
osteoarthritis, fibromyalgia, headaches
sleep latency and sleep reinitiation periods; and
Sleep disorders
the presence of clock-watching before and during
Restless legs syndrome
Periodic limb movement disorder sleep times. Too much noise or light exposure in
Sleep apnea the sleeping room may inhibit sleep initiation. Sim-
Circadian rhythm disorder ilarly, clock-watching with each nighttime awaken-
Parasomnias ing may only further heighten an already raised
Nocturnal panic attacks level of anxiety. Specific difficulty falling asleep at
Nightmares home but not while out of town may suggest
Rapid eye movement behavior disorder insomnia related to the bedroom environment.
Psychiatric conditions Previous treatments tried and their effects and
Depression
side effects should be discussed. Treatments may
Anxiety
include over-the-counter, homeopathic, herbal, or
Panic disorder
Posttraumatic stress disorder prescription medications and behavioral therapies.
Medications In addition to providing information on potential
Decongestants treatments that may not have yet been offered to
Antidepressants the patient, information obtained in this area may
Corticosteroids provide a sense of the kind of treatment for which
b-Agonists the patient is looking.
b-Antagonists Symptoms of other sleep disorders that could be
Stimulants affecting the complaint include such conditions as
Statins
restless legs syndrome, periodic limb movement
Substances
disorder, sleep apnea, and sleep phase syndromes.
Caffeine
Alcohol These should be considered as possible contribu-
Nicotine tors to insomnia.
Cocaine One should review comorbid medical conditions
that could play a role in the presentation. General
Data from Buysse DJ. Sleep disorders and psychiatry.
Arlington (VA): American Psychiatric Publishing, categories to consider include cardiovascular,
American Psychiatric Publishing Review of pulmonary, neurologic, gastrointestinal, renal,
Psychiatry; 2005; and Sateia MJ, Doghramji K, Hauri endocrine, and rheumatologic.
PJ, et al. Evaluation of chronic insomnia. An Review of underlying psychiatric conditions and
American Academy of Sleep Medicine review. Sleep
psychosocial stressors should be included. Eliciting
2000;23:243–308.
symptoms of depression, bipolar disorder, anxiety,
panic (including nocturnal panic attacks), and psy-
chosis can help to clarify the diagnostic picture
in the late afternoon or evening may have a similar while emphasizing the need to obtain or continue
negative effect on sleep latency and continuity. psychiatric care.
A discussion of daytime functioning and associ- A review of substance use, including nicotine,
ated symptoms includes daytime sleepiness; alcohol, and caffeine, should cover amount, fre-
fatigue; difficulty with memory and concentration; quency, and time of day the substance is used
depression; anxiety; irritability; impairment at because all of these substances may contribute to
work, school, or home; and overall quality of life. an insomnia complaint. Patient education about
A report of daytime impairment and patient distress the effects of nicotine, alcohol, and caffeine on
may underscore the severity of symptoms, and sleep should also be undertaken if it seems that sub-
highlight the need aggressively to treat insomnia. stance use has a negative effect on sleep quality.
172 Mai & Buysse
Fig. 1. Sleep diary.
Finally, one should undertake a review of family
history of sleep, medical, and psychiatric disorders.
Physical and mental status examination
The physical examination may reveal signs consis-
tent with sleep apnea (obesity, enlarged neck
circumference, crowded oropharynx) and thyroid,
cardiac, respiratory, and neurologic disorders.
The mental status examination may yield infor-
mation about the patient’s mood, affect, level of
alertness, and ability to attend.
Collateral sources interview
Interview the patient’s bed partner or family mem-
bers, if possible, to elicit symptoms of which the
patient may be unaware. This part of the evaluation
may also help to corroborate and expand on the
patient’s original description. Revelation about re-
spiratory symptoms (snoring, apneas, or gasping)
could suggest a sleep-disordered breathing etiology,
whereas report of repeated limb movements may
move the diagnosis toward restless legs syndrome
or periodic limb movement disorder.
Objective data
Actigraphy
Actigraphy helps to characterize rest-activity pat-
terns and may have some use as an objective mea-
sure when used in conjunction with a sleep-wake
dairy and formal interview. For insomniacs actigra-
phy can provide information about circadian
rhythms and sleep patterns [59]. Compared
with polysomnography, however, actigraphy in
insomniacs has had variable results: it has been
found to overestimate and underestimate total
sleep time [60–62]. Another study found that actig-
raphy was well validated by polysomnography with
respect to number of awakenings, wake time after
sleep onset, total sleep time, and sleep efficiency
[63]. When using actigraphy, increasing the dura-
tion of recording to more than 7 days may improve
the reliability of sleep time estimates [64].
Polysomnography
Polysomnography is not routinely used in the eval-
uation of insomnia; the onus of the diagnosis lies
instead on the patient interview. According to
2003 practice parameters established by the Ameri-
can Academy of Sleep Medicine, specific cases may
apply when polysomnography is warranted. These
cases include suspicion of sleep-related breathing
disorders or periodic limb movement disorders, un-
certain initial diagnosis, treatment failure, and
arousals leading to violent behavior [65].
Sleep diaries
Sleep diaries recorded over 1 to 2 weeks can help
track a patient’s sleep-wake patterns. Information
including actual sleep-wake times, duration of
time in bed, and day-to-day variability in sleep-
wake times can be gathered from the diaries
(Fig. 1).
Summary
Insomnia is thought to result from a state of hyper-
arousal. As a result of this elevated state of alertness,

sleep may prove difficult. Formulating a clinical
definition of insomnia has proved a challenge. Nev-
ertheless, some enduring characteristics of insomnia
include difficulty with sleep initiation or mainte-
nance, early morning awakening, and nonrestora-
tive sleep in the setting of daytime impairment or
distress in the setting of adequate sleep opportunity.
With these characteristics in mind the prevalence of
insomnia is thought to be approximately 10%.
The evaluation of insomnia emphasizes the inter-
view, during which information about the specific
complaint, comorbid sleep, medical or psychiatric
conditions, family histories, medication, and sub-
stance use may be gathered. Additional information
from collateral sources, sleep diaries, actigraphy,
and polysomnography may also prove useful.
Insomnia is a disorder that has far-reaching
effects: medical, psychiatric, personal, and societal
consequences have all been linked with insomnia.
The cost of insomnia can be measured not just in
dollars, but also in impaired quality of life from
comorbid conditions and impaired interpersonal
relationships.
References
[1] Harvey AG. Insomnia: symptom or diagnosis?
Clin Psychol Rev 2001;21:1037–59.
[2] American Academy of Sleep Medicine. The Inter-
national classification of sleep disorders, second
edition (ICSD-2): diagnostic and coding manual.
2nd edition. Westchester (IL): American Acad-
emy of Sleep Medicine; 2005.
[3] American Psychiatric Association. Diagnostic and
statistical manual of mental disorders (DSM-IV-
TR). 4th edition. Text Revision. Washington, DC:
American Psychiatric Association; 2000.
[4] World Health Organization. The ICD-10 classifi-
cation of mental and behavioural disorders:
clinical descriptions and diagnostic guidelines.
Geneva (Switzerland): World Health Organiza-
tion; 1992.
[5] Ancoli-Israel S, Roth T. Characteristics of insom-
nia in the United States: results of the 1991
National Sleep Foundation Survey. I. Sleep
1999;22(Suppl 2):S347–53.
[6] Bixler EO, Kales A, Soldatos CR, et al. Prevalence
of sleep disorders in the Los Angeles metropoli-
tan area. Am J Psychiatry 1979;136:1257–62.
[7] Ford DE, Kamerow DB. Epidemiologic study of
sleep disturbances and psychiatric disorders: an
opportunity for prevention? JAMA 1989;262:
1479–84.
[8] Kuppermann M, Lubeck DP, Mazonson PD, et al.
Sleep problems and their correlates in a working
population. J Gen Intern Med 1995;10:25–32.
[9] Mellinger GD, Balter MB, Uhlenhuth EH. Insom-
nia and its treatment: prevalence and correlates.
Arch Gen Psychiatry 1985;42:225–32.
Insomnia 173
[10] Ohayon MM. Epidemiology of insomnia: what
we know and what we still need to learn. Sleep
Med Rev 2002;6:97–111.
[11] Simon GE, Von Korff M. Prevalence, burden, and
treatment of insomnia in primary care. Am
J Psychiatry 1997;154:1417–23.
[12] Ustun TB, Privett M, Lecrubier Y, et al. Form,
frequency and burden of sleep problems in
general health care: a report from the WHO
collaborative study on psychological problems
in general health care. Eur Psychiatry 1996;11:
S5–10.
[13] Ohayon MM, Hong SC. Prevalence of insomnia
and associated factors in South Korea. J Psycho-
som Res 2002;53:593–600.
[14] Leger D, Guilleminault C, Dreyfus JP, et al. Prev-
alence of insomnia in a survey of 12,778 adults
in France. J Sleep Res 2000;9:35–42.
[15] National Institutes of Health. NIH State-of-the-
Science Conference Statement on Manifestations
and Management of Chronic Insomnia in
Adults. Bethesda (MD): 2005.
[16] Roth T. Insomnia: definition, prevalence, etiol-
ogy, and consequences. J Clin Sleep Med 2007;
3:S7–10.
[17] Johnson EO, Roth T, Schultz L, et al. Epidemiol-
ogy of DSM-IV insomnia in adolescence: lifetime
prevalence, chronicity, and an emergent gender
difference. Pediatrics 2006b;117:E247–56.
[18] Klink ME, Quan SF, Kaltenborn WT, et al. Risk
factors associated with complaints of insomnia
in a general adult population. Arch Intern Med
1992;152:1634–7.
[19] Morgan K. Daytime activity and risk factors for
late-life insomnia. J Sleep Res 2003;12:231–8.
[20] Morgan K, Clarke D. Risk factors for late-life
insomnia in a representative general practice
sample. Br J Gen Pract 1997;47:166–9.
[21] Murata C, Yatsuya H, Tamakoshi K, et al. Psycho-
logical factors and insomnia among male civil
servants in Japan. Sleep Med 2007;8:209–14.
[22] Su TP, Huang SR, Chou P. Prevalence and risk
factors of insomnia in community-dwelling
Chinese elderly: a Taiwanese urban area survey.
Aust N Z J Psychiatry 2004;38:706–13.
[23] Bastien CH, Vallieres A, Morin CM. Precipitating
factors of insomnia. Behav Sleep Med 2004;2:
50–62.
[24] Nakata A, Haratani T, Takahashi M, et al. Job
stress, social support, and prevalence of insom-
nia in a population of Japanese daytime workers.
Soc Sci Med 2004;59:1719–30.
[25] McCall WV. A psychiatric perspective on insom-
nia. J Clin Psychiatry 2001;62(Suppl 10):27–32.
[26] Taylor DJ, Lichstein KL, Durrence HH. Insomnia
as a health risk factor. Behav Sleep Med 2003;1:
227–47.
[27] Riemann D, Voderholzer U. Primary insomnia:
a risk factor to develop depression? J Affect
Disord 2003;76:255–9.
[28] Johnson EO, Roth T, Breslau N. The association
of insomnia with anxiety disorders and
174 Mai & Buysse
depression: exploration of the direction of risk.
J Psychiatr Res 2006a;40:700–8.
[29] Ohayon MM, Roth T. Place of chronic insomnia
in the course of depressive and anxiety disorders.
J Psychiatr Res 2003;37:9–15.
[30] Weissman MM, Greenwald S, Nino-Murcia G,
et al. The morbidity of insomnia uncomplicated
by psychiatric disorders. Gen Hosp Psychiatry
1997;19:245–50.
[31] Goldstein TR, Bridge JA, Brent DA. Sleep distur-
bance preceding completed suicide in adoles-
cents. J Consult Clin Psychol 2008;76:84–91.
[32] Hajak G, SINE Study Group, Study of Insomnia
in Europe. Epidemiology of severe insomnia and
its consequences in Germany. Eur Arch Psychia-
try Clin Neurosci 2001;251:49–56.
[33] Taylor DJ, Mallory LJ, Lichstein KL, et al. Comor-
bidity of chronic insomnia with medical prob-
lems. Sleep 2007;30:213–8.
[34] Katz DA, McHorney CA. Clinical correlates of
insomnia in patients with chronic illness. Arch
Intern Med 1998;158:1099–107.
[35] Ancoli-Israel S. The impact and prevalence of
chronic insomnia and other sleep disturbances
associated with chronic illness. Am J Manag
Care 2006;12:S221–9.
[36] Schwartz S, McDowell AW, Cole SR, et al. Insom-
nia and heart disease: a review of epidemiologic
studies. J Psychosom Res 1999;47:313–33.
[37] Mallon L, Broman JE, Hetta J. Sleep complaints
predict coronary artery disease mortality in
males: a 12-year follow-up study of middle-
aged Swedish population. J Intern Med 2002;
251:207–16.
[38] Ohayon MM. Relationship between chronic
painful physical condition and insomnia.
J Psychiatr Res 2005;39:151–9.
[39] Tang NK, Wright KJ, Salkovskis PM. Prevalence and
correlates of clinical insomnia co-occurring with
chronic back pain. J Sleep Res 2007;16:85–95.
[40] Katz DA, McHorney CA. The relationship be-
tween insomnia and health-related quality of
life in patients with chronic illness. J Fam Pract
2002;51:229–35.
[41] Chilcott LA, Shapiro CM. The socioeconomic
impact of insomnia. Pharmacoeconomics 1996;
10:1–14.
[42] Stoller MK. Economic effects of insomnia
[review]. Clin Ther 1994;16:873–97 [discussion:
854].
[43] Phillips B, Mannino DM. Does insomnia kill?
Sleep 2005;28:965–71.
[44] Leger D, Guilleminault C, Bader G, et al. Medical
and socio-professional impact of insomnia.
Sleep 2002;25:625–9.
[45] Novak M, Mucsi I, Shapiro CM, et al. Increased
utilization of health services by insomniacs: an
epidemiological perspective. J Psychosom Res
2004;56:527–36.
[46] Ozminkowski RJ, Wang S, Walsh JK. The direct
and indirect costs of untreated insomnia in
adults in the United States. Sleep 2007;30:
263–73.
[47] Leger D, Levy E, Paillard M. The direct costs of
insomnia in France. Sleep 1999;22:S394–401.
[48] Walsh JK, Engelhardt CL. The direct economic
costs of insomnia in the United States for
1995. Sleep 1999;22:S386–93.
[49] Leger D, Massuel MA, Metlaine A. Professional
correlates of insomnia. Sleep 2006;29:171–8.
[50] Godet-Cayre V, Pelletier-Fleury N, Le Vaillant M,
et al. Insomnia and absenteeism at work. Who
pays the cost? Sleep 2006;29:179–84.
[51] Bonnet MH, Arand DL. Hyperarousal and
insomnia. Sleep Med Rev 1997;1:97–108.
[52] Stepanski E, Zorick F, Roehrs T, et al. Daytime
alertness in patients with chronic insomnia com-
pared with asymptomatic control subjects. Sleep
1988;11:54–60.
[53] Bonnet MH, Arand DL. 24-hour metabolic rate
in insomniacs and matched normal sleepers.
Sleep 1995;18:581–8.
[54] Pavlova M, Berg O, Gleason R, et al. Self-
reported hyperarousal traits among insomnia
patients. J Psychosom Res 2001;51:435–41.
[55] Nofzinger EA, Buysse DJ, Germain A, et al.
Functional neuroimaging evidence for hyper-
arousal in insomnia. Am J Psychiatry 2004;161:
2126–31.
[56] Krystal AD, Edinger JD, Wohlgemuth WK, et al.
NREM sleep EEG frequency spectral correlates
of sleep complaints in primary insomnia
subtypes. Sleep 2002;25:630–40.
[57] Perlis ML, Smith MT, Andrews PJ, et al. Beta/
Gamma EEG activity in patients with primary
and secondary insomnia and good sleeper
controls. Sleep 2001;24:110–7.
[58] Roth T, Roehrs T, Pies R. Insomnia: pathophysi-
ology and implications for treatment. Sleep
Med Rev 2007;11:71–9.
[59] Morgenthaler T, Alessi C, Friedman L, et al.
Practice parameters for the use of actigraphy
in the assessment of sleep and sleep disor-
ders: an update for 2007. Sleep 2007;30:
519–29.
[60] Sadeh A, Acebo C. The role of actigraphy in sleep
medicine. Sleep Med Rev 2002;6:113–24.
[61] Sivertsen B, Omvik S, Havik OE, et al. A compar-
ison of actigraphy and polysomnography in
older adults treated for chronic primary insom-
nia. Sleep 2006;29:1353–8.
[62] Vallieres A, Morin CM. Actigraphy in the assess-
ment of insomnia. Sleep 2003;26:902–6.
[63] Lichstein KL, Stone KC, Donaldson J, et al. Actig-
raphy validation with insomnia. Sleep 2006;29:
232–9.
[64] Van Someren EJ. Improving actigraphic sleep
estimates in insomnia and dementia: how
many nights? J Sleep Res 2007;16:269–75.
[65] Littner M, Hirshkowitz M, Kramer M, et al. Prac-
tice parameters for using polysomnography to
evaluate insomnia. Sleep 2003;26:754–7.
 175

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 175–187

Efficacy and Safety

of Sleep-Promoting Agents
a,b,c, a,c
Thomas Roth, PhD *, Timothy Roehrs, PhD

- Efficacy of hypnotics - Drugs used for insomnia therapy

Defining hypnotic efficacy On-label use
Assays of hypnotic efficacy Off-label use
Therapeutic end points Self-treatment
Patient populations - Special populations
Duration of efficacy Elderly persons
- Safety of benzodiazepine receptor agonists Patients who have primary sleep disorders
Psychomotor impairment Persons who have hepatic/renal
Cognitive impairment impairments
Discontinuation effects Alcohol and substance abusers
Liability for abuse - References
Falls
Idiosyncratic side effects

The management of insomnia has been affected
dramatically by advances in the understanding of
the pathophysiology and morbidity of insomnia,
by new applications for behavioral treatments of
insomnia, and by the development of new thera-
peutic targets for the pharmacologic management
of insomnia [1].
Insomnia encompasses one or more of the
following symptoms: difficulty initiating sleep, dif-
ficulty maintaining sleep, waking up too early, or
sleep that is chronically nonrestorative or of poor
quality [2]. These difficulties with sleep occur de-
spite the individual’s having adequate opportunity
and circumstances for sleep. In addition to the re-
ported difficulties with sleep, the diagnosis of
insomnia requires a patient’s report of daytime im-
pairment or distress related to the nighttime sleep
difficulty. These impairments may include, but are
not limited to, problems such as fatigue, memory
impairment, mood disturbances, increased risk for
errors and accidents, tension headaches, and gastro-
intestinal symptoms in association with the diffi-
culty with sleep [3]. Most diagnostic systems
require these symptoms to be present three or
more times per week and to have been present for
at least a month.
Data presented at a recent state of the science
conference on insomnia [1] demonstrate clearly
that clinicians need to treat insomnia as a primary
disorder rather than as a symptom secondary to

a
Sleep Disorders and Research Center, Henry Ford Hospital, 2799 West Grand Boulevard, CFP-3, Detroit,
MI 48202, USA
b
Department of Psychiatry, University of Michigan School of Medicine, Ann Arbor, MI 48109, USA
c
Department of Psychiatry and Behavioral Neuroscience, Wayne State University, School of Medicine,
2751 East Jefferson, Suite 400, Detroit, MI 48207, USA
* Corresponding author. Sleep Disorders and Research Center, Henry Ford Hospital, 2799 West Grand
Boulevard, CFP-3, Detroit, MI 48202.
E-mail address: troth1@hfhs.org (T. Roth).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.03.001
sleep.theclinics.com
176 Roth & Roehrs
an underlying condition. Clinicians historically
have regarded insomnia as a consequence of an-
other disorder, because in most cases insomnia is
not present in isolation (primary insomnia) but
rather coexists with another medical, psychiatric,
or sleep disorder (comorbid insomnia). In this
view these other conditions do not cause insomnia;
rather, they precipitate it in vulnerable individuals.
Clinically it is important to recognize that many
patients who have these other conditions do not re-
port insomnia, that the insomnia often predates the
comorbid condition, that treating the comorbid
condition does not always reverse the insomnia,
and that treating the insomnia has benefits in the
management of comorbidities such as pain and
depression.
Efficacy of hypnotics
Defining hypnotic efficacy
The efficacy of any treatment is determined by its
ability to reverse the signs and symptom of a condi-
tion. Insomnia is a symptom-based diagnosis.
Specifically, the symptoms of insomnia include dif-
ficulty in falling asleep or in staying asleep and the
experience of nonrefreshing sleep. To meet the diag-
nostic criteria for insomnia, these symptoms
should be associated with daytime impairment or
daytime distress. Finally, these difficulties with sleep
and associated impairment in daytime function
should be present for at least 3 nights a week for
at least a month.
Assays of hypnotic efficacy
The efficacy of hypnotics is determined objectively
by polysomnography (PSG) and by patient reports
of nocturnal sleep (postsleep questionnaires or dia-
ries). In addition, a variety of measures are used to
evaluate daytime function. Finally, global ratings of
efficacy are determined by evaluations of overall
sleep by the patient and by clinicians.
Therapeutic end points
The therapeutic end points of hypnotics are im-
provements in the patient’s ability to fall asleep
and stay asleep and in the refreshing quality of
sleep. For sleep induction the primary end point
is the speed of falling asleep. In PSG studies the ac-
cepted measure is the time need to achieve 10 con-
secutive minutes of uninterrupted sleep (ie, latency
to persistent sleep). In patient reports, subjects sim-
ply are asked, ‘‘How long did it take you to fall
asleep last night?’’ This subjective assay usually is
collected in the morning, 1 to 2 hours after arising,
and is averaged over some period of time, typically
a week. It is preferable to collect this information by
using a method that time stamps the entries (eg, an
interactive voice response system or electronic
diary) to prevent patients from filling out all the
estimates at a single time point. For sleep mainte-
nance, the accepted PSG end point is the number
of minutes that the patient is awake after sleep on-
set (WASO) before getting out of bed. It includes
the times the patient woke in the middle of the
night well as early morning awakening. Another
measure of sleep maintenance is the number of
times the patient woke during the night before the
final awakening. Three measures are used for pa-
tient reports. The first is the question, ‘‘How long
were you awake during the night?’’ This question
is parallel to the WASO measure. Often, however,
the patient response to ‘‘How long did you sleep
last night?’’ is a better correlate of WASO and thus
is used more frequently. Finally, subjects are asked
how many times they awake during the night. In
PSG measurement, sleep duration (total sleep
time) is evaluated by the number of minutes that
the subject was asleep expressed as a percentage of
the time that the subject was in bed. This ratio of
total sleep time to total time in bed is termed ‘‘sleep
efficiency.’’ Although total sleep time and its deriva-
tive, sleep efficiency, are important, they do not
provide direct information about whether the med-
ication is facilitating sleep onset or sleep mainte-
nance. That is, a 30-minute reduction in sleep
latency and a 30-minute reduction in WASO have
the same effect on total sleep time. The final end
point in sleep efficacy is the refreshing quality of
sleep. All sleep diaries ask questions such as, ‘‘How
would you rate the quality of your sleep?’’ or
‘‘How refreshing was your sleep?’’ Although these
questions have great face validity, their validity in
demonstrating an improvement in the symptom
of nonrefreshing sleep has not been established to
date. This lack of validation is a challenge, because
there is no accepted PSG measure of sleep quality.
In evaluating hypnotics, measures of daytime
function have been used primarily to measure the
residual effects of hypnotics. Attempts to show im-
provement in a variety of tasks in association with
improved sleep have failed, in great part because
almost all studies evaluating daytime function in
insomniacs have failed to identify impairment.
Thus it is difficult to find a cognitive or psychomo-
tor task that shows improvement after an improved
nights’ sleep. In contrast, there are assays of daytime
function that show promise. It has been shown that
measures of fatigue, daytime sleepiness, work pro-
ductivity, quality of life, and disability are improved
with the pharmacologic management of insomnia
[4]. Such improvement is a critical element of
insomnia therapy and needs to be a primary thera-
peutic end point in future studies. These end points
probably have been neglected because historically

clinical trials in insomnia were of short duration,
and longer treatment periods were need to reverse
some of these morbidities.
The overall efficacy of treatment needs to be eval-
uated with a patient and/or a clinician global
impression. In these evaluations the rater typically
uses a five-point scale to determine the degree to
which the treatment has made the insomnia symp-
tom complex better or worse.
Patient populations
Most studies are conducted in patients who have
primary insomnia [1]. It is accepted that most cases
of insomnia are not primary insomnia but rather
insomnia that is comorbid with other conditions
[2]. Studying primary insomnia, however, allows
evaluation of the therapeutic effect of the medica-
tion without the confounding elements of concur-
rent disease and the medications used to treat it.
These studies are conducted in adults and in elderly
populations. There is a need to study insomnia in
elderly, because the therapeutic dose often is lower
and the incidence of side effects is greater in these
patients. Studies also have evaluated the effect of
drugs in normal volunteers undergoing an experi-
mental challenge that produces transient insomnia
[5]. These experimental manipulations, which in-
clude changing the sleep environment (sleeping
in a laboratory or with noise in the background),
changing the timing of sleep (as occurs with jet
travel across multiple time zones), and decreasing
homeostatic drive by requiring the subject to nap
in the afternoon or consuming caffeine or other
stimulants, produce a variety of sleep disturbances
that can be corrected with effective sleep agents.
More recently it has been become clear that most
individuals who have insomnia have a comorbid
condition and, more importantly, that the course
of the insomnia and the course of the comorbid
condition interact. Therefore efficacy studies also
are being conducted in comorbid insomnia [6,7].
The most common comorbid conditions are those
associated with a psychiatric disorder (eg, comorbid
depression and anxiety) and/or a medical disorder
(eg, comorbid disorders associated with pain and
with dyspnea). In menopausal women it also is
important to evaluate the effect of sleep agents on
sleep induction and continuity and on the occur-
rence of menopause-related hot flashes during the
night. In these studies a sample of patients meeting
the diagnostic criteria for insomnia and for a comor-
bid disorder are recruited. Typically the subjects are
allowed to take a medication for the comorbid
condition in combination with the hypnotic (or
placebo). To evaluate efficacy, the traditional sleep
end points are assayed, and the signs and symptoms
of the comorbid condition are evaluated also. For
Sleep-Promoting Agents 177
example, in studying patients who have insomnia
comorbid with depression, sleep and daytime func-
tion were measured, and the rating of depression
was evaluated [6]. In most of these studies the
results seem to indicate that treating the insomnia
has positive effects on the comorbid condition;
however, hypnotics are not indicated for the treat-
ment of these comorbid conditions.
Duration of efficacy
Until recently the clinical lore was that insomnia
was a symptom; therefore, the underlying condition
should be treated in the long term, and hypnotic
therapy should be undertaken only as a short-
term solution. As a result, hypnotic efficacy trials
were conducted only on a short-term basis; the du-
ration of most trials was 4 weeks or less. With the
realization that insomnia is a chronic disorder
and data showing that many insomniacs use hyp-
notics on a long-term basis, long-term trials of
hypnotic are being conducted routinely [1]. Almost
all sleep medications recently marketed or under
development undergo efficacy trials for nightly
use for 3 to 12 months [8]. These longer trials allow
the evaluation of daytime function, which requires
a longer therapeutic trial to demonstrate potential
benefit.
Safety of benzodiazepine receptor agonists
The drug class of choice for pharmacotherapy of
insomnia is the benzodiazepine receptor agonists
(BzRAs). This class includes all but one of the indi-
cated hypnotics. The major side effects associated
with BzRAs are psychomotor and cognitive (ie, an-
terograde amnesia) impairment, discontinuation
effects, and the risk of dependence [9]. Some of
these side effects are mediated by the primary phar-
macodynamic activity of BzRAs—sedation—and
thereby relate directly to the pharmacokinetic prop-
erties of specific BzRAs. Other side effects can be
attributed to both the drug’s pharmacokinetics
and the specificity of its receptor selectivity. Finally,
drug dosage and duration of use may determine
other of the side effects; drug dosage is the major
determinant of all these side effects.
Psychomotor impairment
Psychomotor impairment has been demonstrated
in laboratory performance tests and actual roadway
driving by slowed reaction times, response errors,
tracking errors, lapses of attention, and driving
deviations. At peak plasma concentrations, impair-
ment relates directly to drug concentration (dose
and time since ingestion). For example, the effects
of daytime administration of 0.125, 0.25, and
0.50 mg triazolam, of 5, 10, and 20 mg zolpidem,
178 Roth & Roehrs
and of 15, 30, and 60 mg temazepam were com-
pared [10]. The drugs were chosen for their differ-
ences in pharmacokinetics and receptor selectivity,
with temazepam being longer acting than zolpi-
dem, and triazolam and zolpidem being more
receptor selective than triazolam and temazepam.
At peak concentration, zolpidem, triazolam, and
temazepam each produced orderly dose-related im-
pairments of psychomotor performance, learning,
and recall. Their differential receptor selectivity
(ie, zolpidem’s greater selectivity for gamma-
aminobutyric acid type A [GABAA] alpha 1 recep-
tors) did not produce differing patterns of
impairment at peak concentrations.
The duration of impairment relates to both the
half-life and dose. The time-course of impairment
for the drugs in the study described previously
revealed a 6-hour duration of impairment relative
to placebo with temazepam (60 mg) and a 3-hour
duration impairment with zolpidem (20 mg),
although these drug doses had comparable impair-
ing effects at their peak [10].
When the BzRAs are administered before sleep,
and the impairment extends to the morning follow-
ing the nighttime administration, the impairment is
referred to as ‘‘residual effects’’ (ie, a prolongation of
the therapeutic effect of the drug). Residual effects
are not the same as a rebound effect; with residual
effects, plasma concentrations of drug are still pres-
ent, whereas rebound occurs after the plasma con-
centration has reached zero. Thus, the primary
determinant of residual effects is the duration of
drug action, which is determined by half-life of
the drug and secondarily by the dose of the drug
(eg, higher doses and longer half-lives extend the
duration of action). Studies using performance,
driving, and Multiple Sleep Latency Test (MSLT)
assessments show differences in residual effects
between short- and long-acting drugs and between
doses of the same drug. A classic early study in
healthy elderly persons compared the daytime
residual effects of triazolam (0.25 mg) and fluraze-
pam (15 mg) administered before sleep [11]. Both
drugs produced a comparable 1-hour increase in
total sleep time, but flurazepam, a long-acting
drug, resulted in greater daytime sleepiness as eval-
uated by the MSLT on the following day, whereas
triazolam, a short-acting drug, reduced sleepiness
as evaluated by the MSLT. Also next-day vigilance
performance was impaired with flurazepam, but
triazolam had no effect on vigilance.
The likelihood of residual effects is determined
by the time of drug administration relative to the
time of arising versus the pharmacokinetics of
the drug. This point is illustrated in a study that
compared the residual effects of zolpidem
(10 mg), which has a short half-life (2.5–4.5 hours)
and of zaleplon (10 mg), which has an ultra-short
half-life (1 hour) after middle-of-the-night admin-
istration (at 3:00, 4:00, 5:00, or 6:00 AM) before
an 8:00 AM awakening or 2 to 5 hours after admin-
istration [12]. Zolpidem showed residual effects on
digit symbol substitution and immediate and de-
layed memory recall after all the middle-of-the-
night administrations, but no effects were observed
with zaleplon, even when administered at 6 AM,
2 hours before awakening. Consequently, given
the distinct pharmacokinetics for various hyp-
notics, the Food and Drug Administration labels
may include the caution that 8 hours should be
devoted to sleep when using the medication.
Cognitive impairment
Cognitive impairment, typically anterograde amne-
sia, is another major side effect of BzRAs. Antero-
grade amnesia is memory failure for information
presented after consumption of the drug. It is deter-
mined by the pharmacokinetics and dose of the
drug: the plasma concentration at the time of infor-
mation input determines the degree of amnesia (ie,
memory consolidation failure). At peak plasma
concentrations, very orderly dose-dependent amne-
sic effects have been demonstrated for BzRAs [10].
The amnesia is related in part to the sedative effects
of the BzRAs, because the degree of the amnesic ef-
fects parallels the sedative effects as measured by the
MSLT [13]. That failed consolidation of the newly
acquired material is the cause of the amnesia was
supported by a study in which the drug-induced
rapid return to sleep was delayed for 15 minutes
(ie, wakefulness was maintained for 15 minutes),
and memory was preserved [14]. The extent to
which the sedative effects mediate the amnesic
effects has been disputed extensively, however. Sev-
eral studies have attempted to dissociate the two
effects by using drugs that have different effects on
sedation and memory or by using the antagonist,
flumazenil; results have been equivocal [15,16].
The problem with these studies is that sedation is
self-reported rather than assessed objectively.
Amnesia also is associated with the receptor se-
lectivity of the BzRAs. The BzRAs act as allosteric
modulators of GABAA receptors, and gene knock-
in studies have identified and characterized the
pharmacologic profiles of various GABAA receptor
subunits. Animal data indicate the alpha 1 receptor
subtype mediates both the sleep and amnesic
effects of the BzRAs [17]. When the first non-
benzodiazepine hypnotics were introduced, it was
hypothesized that amnesia could be avoided
because of the receptor selectivity of zolpidem. As
noted previously, however, zolpidem, which is se-
lective for the alpha 1 receptor, did not differ
from the nonselective BzRAs in its amnesic effects

[10]. The subsequent animal studies revealed that
the alpha 1 receptor subtype mediates both
sedation and memory. All BzRAs produce dose-
dependent anterograde amnesia, and no studies
have demonstrated differences among the various
drugs when sedative potency is controlled.
Long-term BzRA use purportedly is associated
with cognitive impairment, particularly in elderly
persons. The results of studies assessing cognitive
function in elderly chronic BzRA users are equivo-
cal, with some studies reporting impairment and
others finding minimal or no impairment [18–20].
It is difficult to make definitive conclusions,
because these reports are cross-sectional and retro-
spective in nature with a number of possible con-
founds, and determining the appropriate controls
for these studies is problematic. Furthermore,
most of the information is from patients who
have anxiety disorders who are using long-acting
BzRAs. The relevance of these data to current best
clinical practice for insomnia pharmacotherapy
(ie, short-acting non-benzodiazepine hypnotics) is
questionable.
Discontinuation effects
The most prominent discontinuation effect of the
BzRAs in clinical use is rebound insomnia [21].
Rebound insomnia is disturbed sleep for 1 to
2 nights relative to baseline after even 1 to 2 nights
of previous BzRA use. In the short term, rebound
insomnia does not seem to increase in severity
with the duration of nightly use. It was reported first
with the 0.5 mg dose, but not the 0.25 mg dose, of
the short-acting drug triazolam [21]. Although
proper multiple-dose studies exploring the thresh-
old dose for rebound in other hypnotics have not
been conducted, rebound is likely to occur after
high doses (ie, beyond minimally effective doses)
of all short- and intermediate-acting BzRAs. This
prediction is based on the multiple-dose studies
of daytime performance impairment that have
compared various drugs with triazolam and have
found comparable impairment at triazolam doses
that produce rebound [10]. Rebound is not likely
with any long-acting drugs because of the gradual
decline in plasma concentrations inherent in their
pharmacology. Clinically rebound can be mini-
mized with short- and intermediate-acting drugs
by tapering the dose gradually over a few nights.
Rebound insomnia is an exacerbation of the orig-
inal symptom (ie, insomnia) and thus differs from
recrudescence, which is the return of the original
symptom at its original severity. It is not a with-
drawal syndrome (ie, expression of new symp-
toms), at least in the available short-term studies
(ie, 2 weeks and less), which induced rebound
but in which no other new symptoms were
Sleep-Promoting Agents 179
observed [21]. The extent to which duration of use
and dosage might combine to increase the likeli-
hood of rebound, even at clinical doses, with
long-term use, is not known fully. A recent study as-
sessed rebound insomnia after 6 months of nightly
use of eszopiclone at its clinical dose (3 mg) [22].
No increase in self-reported sleep latency or
WASO relative to baseline was observed for
14 days after eszopiclone was discontinued. It has
been suggested that the experience of rebound
insomnia leads to continued chronic use of the
hypnotic. A study directly tested that notion and
showed that the experience of rebound insomnia
did not alter the subsequent likelihood of a patient’s
self-administering triazolam (0.25 mg) [23].
Liability for abuse
With long-term use there is concern about depen-
dence, because physical and behavioral dependence
have been reported with long-term daytime anxio-
lytic use of therapeutic doses of BzRAs [24]. System-
atic information regarding the risk of dependence
with long-term therapeutic use of BzRA hypnotics
at clinical doses is very limited, however. Epidemio-
logic studies indicate that most patients use hyp-
notics for 2 weeks or less [25,26]. Two recent
placebo-controlled, double-blind studies of eszopi-
clone (3 mg) reported no evidence of physical or
behavioral dependence after 6 months of nightly
use [22,27]. These studies, however, did not directly
test the risk of physical and behavioral dependence.
Short-term studies directly testing the risk of
behavioral dependence of BzRA hypnotics suggest
they carry a low risk of behavioral dependence
[28,29]. The risk of behavioral dependence was
tested directly by using color-coded capsules to
assess the self-administration of active drug versus
self-administration of placebo. After sampling
each color-coded capsule, patients chose a capsule
based on its color over 7 to 14 subsequent nights.
Self-administration of hypnotics by insomniacs
was not associated with dose escalation with re-
peated use when insomniacs were given the oppor-
tunity to self-administer multiple capsules [29], did
not increase with rebound insomnia [23], did not
generalize to daytime use [30], and varied as a func-
tion of the nature and severity of the patient’s sleep
disturbance [28]. This evidence indicates that the
insomnia patients’ self-administration of hypnotics
in these studies is therapy-seeking behavior and not
drug seeking or abuse. These conclusions hold true
for insomniacs and normal controls but not for
individuals who have a history of drug abuse.
One important question is the extent to which
receptor subtype selectivity may influence the risk
of abuse of the BzRAs. One assessment failed to
find differential receptor subtype selectivity as
180 Roth & Roehrs
a factor in the risk of abuse among drugs used as
hypnotics [31]. For example, the risk of abuse of
the alpha-1 receptor–selective drug zolpidem did
not differ from that of various nonselective BzRAs.
Few studies, however, have compared multiple
doses of multiple drugs, and thus the rating had
to be made across a variety of methodologies and
data sources. The rating also included drug toxic-
ities and thus was not specific to what is more nar-
rowly defined as drug-abuse liability.
Falls
The use of psychotropic medication has been
reported to be associated with an increased risk of
falls, particularly in elderly persons. A number of
studies done in hospitalized, nursing home, and
residential care patients have reported an associa-
tion between the use of psychotropic medication
and falls, but antidepressants seem to carry the
highest risks for falls [32–37]. The BzRAs typically
reported in these studies as being associated with
risk of falls are all long-acting drugs used as daytime
sedatives. The extent to which BzRA hypnotics used
to treat insomnia, particularly the short-acting
drugs, are associated with falls is not known fully.
These studies did not control for the presence of
insomnia, a known risk factor for falls in the elderly.
Several recent studies controlling for insomnia have
found the risk of falls associated with these medica-
tions is the same as, or even less than, the risk with
untreated insomnia [36,37].
Idiosyncratic side effects
Reports of idiosyncratic side effects associated with
BzRA hypnotics have appeared periodically in the
public press. These reports of ‘‘global amnesia,’’
somnambulism, and sleep-related eating disorders
are problematic because they raise unnecessary
concern among patients and their physicians. These
reports are not peer reviewed, generally are not
documented independently, are subject to confir-
mation bias, and overrepresent the real risk.
Peer-reviewed case reports of idiosyncratic side
effects associated with BzRA use also have appeared
in the scientific and medical literature. One must
view these reports with caution. Although case re-
ports do provide more accurate information that
includes contributing factors, they do not have the
evidence level of placebo-controlled information.
The real risk is unknown, because the rate of expo-
sure is not known: the number of prescriptions
written and the doses consumed at the time of the
event are unknown, and consequently the inci-
dence of the events cannot be determined.
Transient global amnesia has been reported in as-
sociation with the use of triazolam by otherwise
healthy individuals [38,39]. The memory loss was
for all autobiographical events transpiring over an
8- to 12-hour period. In some of these cases in
which clinical doses were used, prior stress, sleep
deprivation, and a virus may have contributed to
the amnesia. In other cases, supraclinical doses
and alcohol ingestion probably were contributory
factors. It is unlikely that this phenomenon is
unique to triazolam, because similar kinds of am-
nesia are produced by the intravenous administra-
tion of other BzRAs.
Somnambulism has been reported with zolpi-
dem and zaleplon [40,41]. These episodes of som-
nambulism have occurred in individuals taking
two to three times the clinical doses of the drug,
in individuals who have a prior history of somnam-
bulism, and in individuals who have experienced
prior traumatic head injury. Zolpidem-associated
somnambulism also has been reported in combina-
tion with antidepressant treatment [42]. Somnam-
bulism is believed to be associated with partial
arousals from sleep. Although BzRAs increase som-
nambulism, alcohol and sleep deprivation also pro-
duce partial arousals and increase somnambulism.
Finally, there have been recent case reports in-
volving sleep-related eating disorder and psychotro-
pic medications, including BzRAs [43–48]. It is
disputed whether sleep-related eating disorder is
a disorder of partial arousal from sleep with altered
levels of consciousness or is the psychiatric disorder
of nocturnal eating with awareness and recall
[46,48]. Sleep-related eating disorder is hypothe-
sized to share a common pathophysiology with
somnambulism. Zolpidem was reported to exacer-
bate sleep-related eating disorder and in several
cases to induce it de novo [46]. In some of these
cases doses of zolpidem greater than 10 mg were
being used, and in other cases there was use of
sedating antidepressants. Sleep-related eating disor-
der also has been reported with triazolam [47,49].
A common thread links much of this case-report
information: excessive hypnotic activity or sleep
drive. The excessive hypnotic activity can occur as
a result of high doses, clinical doses in vulnerable
individuals (ie, those who have a past history of
sleep disorders or brain injury), the combination
of clinical or high doses with prior sleep depriva-
tion caused by stress or illness, or the combination
of clinical or high doses with the prior consump-
tion of alcohol. The behaviors described in these
case reports also share a commonality. They all
are symptoms of excessive hypnotic activity or ex-
cessive sleepiness. Patients who have primary sleep
disorders report amnesia and memory difficulties
associated with excessive daytime sleepiness. Sleep
deprivation produces intense slow-wave sleep, and
abrupt arousal from slow-wave sleep after prior
sleep deprivation is known to be associated with
 Sleep-Promoting Agents 181

sleep inertia, behavior of which individuals have Table 1: Insomnia treatment medications
little consciousness or little memory. Patients who
have excessive sleepiness are known to engage in Half-life Available
automatic behavior. Sleep deprivation also is Drug (in hours) dose (in mg)
known to induce somnambulism in individuals BzRAs
who have a previous history of somnambulism. Estazolam 8–24 1, 2
Thus, this information from case reports is not Flurazepam 48–120 15, 30
quite as idiosyncratic as it first might seem. This in- Quazepam 48–120 7.5, 15
formation relates to the known effects of high doses Temazepam 8–20 7.5, 15, 22.5, 30
of BzRAs and to the effects of other manipulations Triazolam 2–4 0.125, 0.25
Non-BzRAs
of sleep, such as sleep-phase reversal, sleep depriva-
Imidazopyridine
tion, sleep fragmentation, and the consumption of Zolpidem 1.5–2.4 5, 10
alcohol and other sedating drugs. Zolpidem 2.8–2.9 6.25, 12.5
Clinically, two points should be emphasized. extended-release
First, excessive sleep drive and hypnotic activity pro- Pyrazolopyrimidine
duced by high doses (doses above the approved Zaleplon w1 5, 10
clinical doses), a combination of sedating drugs, Pyrrolopyrazine
or the combination of prior sleep deprivation and Eszopiclone 5–7 1, 2, 3
a sedating drug in vulnerable individuals should MT agonist
be avoided. Thus, the dose used, the concurrent Ramelteon 1–2.6 8
use of other sedating drugs, and the time in bed
after drug ingestion should be monitored carefully.
Second, by most indications these side effects are have a benzodiazepine chemical structure (ie, es-
rare when the medications are used appropriately. tazolam, flurazepam, quazepam, temazepam, tria-
One study of adverse reactions to sedative hyp- zolam); others (ie, zaleplon, zolpidem, zolpidem
notics over a 3-year period found the median CR, eszopiclone) do not. The one non-BzRA, re-
frequency of report adverse reactions was 0.01%, melteon, acts as an agonist at MT1 receptors in
or 1 in 10,000 doses [50]. In double-blind, pla- the super-chiasmatic nucleus (SCN). The SCN
cebo-controlled trials no reports of such adverse contains high concentrations of MT1 and MT2 re-
events associated with the BzRAs have been ceptors; the MT1 receptors are thought to attenu-
reported. ate the SCN’s alerting signal, and the MT2
receptors are thought to synchronize the circadian
Drugs used for insomnia therapy clock [52,53].
The binding affinity of the BzRAs for most GABAA
A variety of drugs from different drug classes are receptor subtypes is similar. In contrast, the affinity
prescribed for insomnia pharmacotherapy, and in- of the non-BzRAs for the receptor subtype with an
somniacs also report self-treating their insomnia alpha-1 subunit is much higher than for other sub-
with alcohol and with over-the-counter (OTC) types [54]. Because receptors with alpha-1 subunits
and herbal agents. The drugs currently approved mediate sedation, amnesia, and some of the anti-
for pharmacotherapy of insomnia are listed in convulsant properties, but not anxiolysis or myore-
Table 1. As noted earlier, with one exception, the laxation, it is possible that these more selective
BzRAs are the drug class of choice for insomnia drugs will have hypnotic effects with fewer side ef-
pharmacotherapy. The BzRAs share a common fects [54]. This conjecture remains to be demon-
mechanism of action; the one non-BzRA, ramel- strated definitely, however.
teon, has a unique mechanism, stimulation of the The other major difference among BzRAs is in
melatonin MT1 receptor. The BzRAs differ in their their pharmacokinetics: some have an ultra-short
receptor-binding specificity, time to maximum con- half-life, others have a short or an intermediate
centration, and half-life. half-life, and some have a long half-life. Drugs
that have an intermediate and long half-life are
On-label use likely to produce residual daytime sedation,
The term ‘‘benzodiazepine receptor agonists’’ is whereas zaleplon, which has an ultra-short half-
derived from the recognized mechanism of action life, is indicated only for sleep induction, not for
of these drugs, which involves occupation of ben- sleep maintenance. The non-BzRA, remelteon,
zodiazepine receptors on the GABAA receptor also is indicated only for sleep induction. All thein-
complex, resulting in the opening of chloride dicated hypnotics in short- and intermediate-term
ion channels and facilitation of GABA inhibition studies have been shown to reduce sleep latency
[51]. Some of these drugs, described in Table 1, and most, excepting zaleplon and remelteon,
182 Roth & Roehrs
increase total sleep time as assessed by patient re-
ports, by nocturnal PSGs, or both.
The development of tolerance to the hypnotic
effects of these drugs has been an area of dispute.
Tolerance classically is defined as the reduction of
a drug effect with repeated administration of a con-
stant dose or the need to increase the dosage to sus-
tain a specific level of effect. Despite speculations in
the medical literature, tolerance to the hypnotic
effects of BzRAs did not develop in most studies,
at least at the therapeutic doses for the periods of
time that have been studied. Investigations that
often are cited as evidence for the development of
tolerance (eg, the study by Mitler and colleagues
[55]) show gradual improvement over time in the
placebo group versus a constant effect in the drug
group, resulting in loss of statistical significance.
The sleep in the active drug groups does not worsen
with time with a stable dose of drug, as the defini-
tion of tolerance requires. Thus, one cannot con-
clude that tolerance has developed. It is likely that
unspecified changes have occurred over time with
placebo, such as spontaneous remission, regression
to the mean (if there are sleep disruption entry cri-
teria), sleep hygiene influences inherent in protocol
adherence, Hawthorne effects, and true placebo
effects. Several recent large-sample outpatient stud-
ies have shown continued hypnotic efficacy of eszo-
piclone (3 mg) for 6 months and more of nightly
administration [8,56].
Chronic primary insomnia is defined as difficulty
initiating or maintaining sleep or as nonrestorative
sleep that is associated with some type of daytime
impairment. One would expect that improved
nighttime sleep would result in improved daytime
function. Few studies, however, have documented
impaired daytime function in primary insomniacs
objectively, and consequently improved daytime
function associated with improved nighttime sleep
has been an elusive outcome. Among other things,
the problem relates to an incomplete understand-
ing of the pathophysiology of primary insomnia.
In the 6-month eszopiclone studies, however,
patient reports of daytime alertness, ability to func-
tion during the day, and physical sense of well being
were improved relative to placebo [8,56].
Off-label use
Sedating antidepressants
The most frequently prescribed medications for the
treatment of primary insomnia are sedating antide-
pressants, trazodone, amitriptyline, and mirtaze-
pine being the leading three [57]. Unfortunately,
little is known about their mechanism of action
for hypnotic effects or their efficacy and safety as
hypnotics. The transmitter systems altered by the
three leading sedating antidepressants differ.
Trazodone antagonizes serotonin 2a (5HT2a),
5HT2c, and alpha1-adrenergic receptors and also in-
hibits 5HT reuptake [58,59]. Amitriptyline blocks
acetylcholine and histamine binding and inhibits
reuptake of norepinephrine and 5HT [60–62]. Mir-
tazepine antagonizes alpha1-adrenergic, 5HT2a,
5HT2c, and 5HT3 receptors, and is a strong hista-
mine receptor type 1 (H1) antagonist [63,64].
Amitriptyline and mirtazepine share antihistami-
nergic activity, which may produce hypnotic effects.
The hypnotic activity of trazodone may occur
through 5HT2a, 5HT2c, and/or alpha1-adrenergic
mechanisms.
To the authors’ knowledge, the data concerning
the hypnotic efficacy of sedating antidepressant
agents in primary insomniacs are limited to two
studies of trazodone, two studies of trimipramine,
and two studies of doxepin [65–68]. Amitriptyline
and mirtazepine have not been studied in primary
insomnia. Trazodone (150 mg) in ‘‘poor sleepers’’
(participants were not further characterized) over
3 weeks reduced WASO and stage 1 sleep and
increased stage 3 to 4 sleep relative to a placebo
baseline, but it did not increase total sleep time
or reduce sleep latency [65]. In well-defined pri-
mary insomniacs, trazodone (50 mg), compared
with zolpidem (10 mg) and parallel placebo, re-
duced self-reported sleep latency and increased
sleep duration only during the first week of the
2-week study; zolpidem, however, continued to re-
duce sleep latency during the second week of the
study [66].
One of the trimipramine studies failed to find
a significant improvement of PSG-defined sleep
measures with a mean dose of 100 mg taken for 1
month, although self-rated sleep did show im-
provement relative to placebo [67]. The second
trimipramine study found an increase in total sleep
time and sleep quality relative to a baseline with
a average 166-mg dose taken for 1 month [68].
This study, however, did not include a parallel pla-
cebo group.
Doxepin (25–50 mg) taken nightly for 1 month
by primary insomniacs improved total sleep time
and WASO on both nights 1 and 28 relative to a par-
allel placebo group [69]. Lower doses of doxepin
(1, 3, and 6 mg), with less likelihood of anticholin-
ergic side effects, were studied recently in primary
insomniacs [70]. Taken for 2 nights in a crossover
design, 1, 3, and 6 mg of doxepin increased total
sleep time and reduced WASO relative to placebo.
The side-effect profile of the sedating antidepres-
sants may be more problematic, particularly com-
pared with the BzRAs,. The safety data for sedating
antidepressants have been compiled from studies
of patients who have primary depression using an-
tidepressant doses. As a case in point, in the 25- to

50-mg doxepin study cited previously, two of the
four insomniacs who discontinued doxepin treat-
ment did so because of significant adverse events
[69], but no serious adverse events were reported
in the study of low-dose doxepin [70]. With that
proviso, the margin of safety is much narrower
with antidepressants than with BzRAs and, because
the half-life of most of these drugs is 9 to 30 hours,
the likelihood of daytime residual effects is
enhanced.
Anticholinergic side effects are reported with
many of the tricyclic antidepressants, including am-
itriptyline and trimipramine. For example, in the
previously cited study of trimipramine in primary
insomniacs, dizziness, dry mouth, headache, and
nausea were more frequent than with the compara-
tor drug, lormetazepam [67]. With trazodone,
orthostatic hypotension, weakness, and lighthead-
edness are common; cardiac conduction abnormal-
ities have been reported in patients who have
pre-existing heart disease; and, although rare, priap-
ism is a potentially serious side effect [71–74].
Antipsychotics
The antipsychotics quetiapine and olanzapine also
are used frequently as hypnotics in people who
have primary insomnia. Unlike the sedating antide-
pressants, almost no information is available about
the efficacy and safety of their use as hypnotics in
primary insomnia. Any sedative effects with these
drugs may result from their antihistaminic activity.
These drugs also have adrenergic, muscarinic, dopa-
minergic, and serotonergic activity. Activity at these
multiple transmitter systems increases the likeli-
hood of side effects.
Anxiolytics
Anxiolytics, including clonazepam, alprazolam,
and lorazepam, also were among the 16 drugs
most frequently reported as being used to treat
insomnia [57]. Furthermore, the anxiolytics are pre-
scribed as hypnotics for a longer initial period of
treatment, and the prescriptions are refilled more
frequently [75]. The mechanism of action of these
drugs is the same as for the hypnotic BzRAs. To
the authors’ knowledge, however, there are no stud-
ies of their hypnotic efficacy in primary insomnia.
Clonazepam has been studied as a second-line
treatment for periodic limb movement disorder
and as a primary treatment for various parasomnia
disorders [76,77]. Having the same mechanism of
action, these drugs have side effect profiles similar
to those of the hypnotic BzRAs. The half-lives
of clonazepam and lorazepam are longer than
6 hours, and thus they are likely to produce residual
effects. Alprazolam is chemically similar to triazo-
lam and has a short half-life.
Sleep-Promoting Agents 183
Self-treatment
In the general population relatively few people who
have insomnia receive medical treatment; one study
reported that 5% received treatment [78]. Persons
who have insomnia do use other available sub-
stances to treat their condition. Population-based
studies have reported that 10% to 28% of respon-
dents report using alcohol as a sleep aid, and 10%
to 29% use OTC sleep aids [75,78–80].
Alcohol
Reported studies of the effects of alcohol on sleep
conducted in healthy normal persons have used
high alcohol doses, doses that raise alcohol breath
concentrations above 0.05% [81]. These doses dis-
rupt sleep, at least during the second half of the
night. Insomniacs, however, report using low doses,
one to two drinks before sleep [79]. The use of low-
dose alcohol as a sleep aid is potentially dangerous
for two reasons. Low-dose alcohol initially im-
proves the sleep of insomniacs, which is why they
self-administer it as a sleep aid [82]. Within 6 nights,
however, tolerance develops, sleep is worsened be-
yond baseline, and larger alcohol doses are self-
administered to achieve the sleep effect [83,84].
Also, in one population-based study, insomniacs
who reported using alcohol as a sleep aid reported
greater levels of daytime sleepiness than those who
used prescription or OTC drugs for sleep [26].
Over-the-counter sleep aids
The active component of most all OTC sleep aids is
an H1 antihistamine, typically diphenhydramine
(25 mg). Several studies of diphenhydramine
(25–50 mg) have shown hypnotic effects for several
nights of administration, but these were not paral-
lel, placebo-controlled studies [85,86]. There
is limited placebo-controlled evidence that diphen-
hydramine has hypnotic efficacy, and it has been
shown that tolerance to the sedative effects of di-
phenhydramine develops rapidly [87].
Herbals
According to the 2002 National Health Interview
Survey of the general population, 17% of respon-
dents reported insomnia within the past year, and
approximately 5% of persons who had experienced
insomnia reported using various alternative treat-
ments, mostly herbals, for their insomnia [88].
There is a paucity of rigorous scientific investigation
of the efficacy and safety of the use of herbals to
treat insomnia. The herbal that has received the
most investigation is valerian. Valerian is a root
occurring in several species (ie, Valeriana officinalis,
Valeriana wallichii, Valeriana edulis), which is
extracted and prepared by different methods, pro-
ducing differing chemical constituents in the final
184 Roth & Roehrs
product [89]. The mechanism of action for vale-
rian’s hypnotic effect is not certain, although one
study suggests that valerian has agonistic activity
at the adenosine A1 receptor [90]. A recent system-
atic review identified 29 controlled trials of the effi-
cacy of valerian in insomniacs and concluded that,
regardless of preparation, valerian did not improve
self-rated or PSG-rated sleep relative to placebo
[89]. The failure to find efficacy occurred although
the placebo condition in more than half the studies
did not control for the distinctive, unpleasant odor
of valerian.
Another herbal taken for a variety of conditions,
including depression, anxiety, and sleep distur-
bances, is St John’s Wort [91]. St John’s Wort is
a flowering herb that also is available in a number
of different preparations, although the active ingre-
dient is thought to be hyperforin [91]. Hyperforin
inhibits reuptake of serotonin, norepinephrine,
dopamine, L-glutamate, and GABA. Trials of its
efficacy in depression have been conducted, but
there are no trials in primary insomnia. The results
of the two studies of its hypnotic activity in healthy
normal persons are equivocal [91].
Special populations
Elderly persons
Treating insomnia in the elderly is complex for two
reasons: the normal change in drug pharmacokinet-
ics associated with aging, and the increased
frequency of primary sleep disorders in elderly
persons. A number of age-related changes in gastro-
intestinal structure and function have been docu-
mented that affect the absorption of drugs, and
age-related changes in body morphology alter
drug distribution [92]. Better known is the aging-
associated change in liver function that alters drug
metabolism.
Drugs that are metabolized primarily by conjuga-
tion are potentially safer for aged patients or
patients who have liver disease. The characteristic
pharmacokinetics of oxidated drugs are altered in
elderly people and in patients who have liver dis-
ease by increasing the area under the plasma-
concentration curve. This alteration occurs in
some drugs (eg, triazolam) by increasing the peak
plasma concentration and in others (eg, fluraze-
pam) by extending the duration of significant
blood levels. The reduced recommended dose for
most hypnotics when treating elderly patients is
related, in part, to these kinetic changes.
The second issue in treating insomnia in the
elderly relates to the increased frequency of primary
sleep disorders in elderly persons, a well-
documented phenomenon [92]. The issue of use
of hypnotics in persons who have primary sleep
disorders is discussed in the next section.
Patients who have primary sleep disorders
There is a potential negative effect of drugs used as
hypnotics in insomniacs who have the primary
sleep disorders of sleep-related breathing distur-
bances and periodic leg movements. Among the
drugs most frequently prescribed for treatment of
insomnia are the sedating antidepressants, one
of which is the tricyclic, amitriptyline. The tricyclics
are reported to exacerbate periodic leg movements,
although the risk factors, mechanisms, and dose-
relations for this effect are unclear [92,93]. On the
other hand, BzRA hypnotics improve periodic leg
movements by reducing the arousals associated
with the leg movements.
Drugs that have a sedative effect, including the
BzRAs, have the potential to exacerbate sleep-
related breathing disturbances. The early data were
scientifically weak and somewhat equivocal. The
evidence now suggests that hypnotics do not induce
sleep-related breathing disturbance in people with-
out such disturbance, do produce a small increase
in people who have occasional apnea and hypo-
pnea, and exacerbate sleep-related breathing in
patients who have clear obstructive sleep apnea syn-
drome [92].
Persons who have hepatic/renal impairments
Because most hypnotics undergo hepatic metabo-
lism, advanced liver disease requires the use of
a lower dose or avoidance of these medications.
The alteration of the pharmacokinetics of hypnotics
in cases of compromised liver function was dis-
cussed earlier in this article.
Alcohol and substance abusers
Although the risk of developing dependence on
BzRAs and other sedative drugs is low, most pa-
tients who have a history of alcoholism or drug
abuse should not receive BzRAs in outpatient set-
tings without close supervision. The BzRAs also
should be used cautiously by moderate users of
alcohol because the additive sedative effects with
hypnotics narrow the wide margin of safety.
Sleep-related eating disorder in association with
hypnotic use is often reported after prior alcohol
consumption, suggesting additive sedative effects
evoke the eating disorder.
Given the persisting sleep disturbance of alcohol-
ism and the proven risk of relapse associated with
the sleep disturbance, treating that disturbance is
clinically important. No alternative to the BzRAs
has emerged, however. GABA agonists, specifically
gabapentin, have shown some promise, but in stud-
ies to date their use either has improved sleep but

not drinking outcomes or has improved drinking
outcomes but not sleep [94]. Selective 5HT2A recep-
tor antagonists such as ritanserin, eplivanserin, and
several others are being investigated as hypnotics
because of their capacity of enhance slow-wave
sleep [94]. Like the BzRAs, these drugs have not
shown strong effects on sleep induction.
References
[1] National Institutes of Health. State of the science
conference statement on manifestations and
management of chronic insomnia in adults
June 13–15, 2005. Sleep 2005;28(9):1049–57.
[2] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
edition text revised. Washington, DC: American
Psychiatric Association; 2000.
[3] Edinger JD, Bonnett MH, Bootzin RR, et al. Der-
ivation of research diagnostic criteria for insom-
nia: report of an American Academy of Sleep
Medicine work group. Sleep 2004;27:1567–96.
[4] Walsh JK, Krystal AD, Amato DA, et al. Nightly
treatment of primary insomnia with eszopiclone
for six months: effect on sleep, quality of life and
work limitations. Sleep 2007;30(8):959–68.
[5] Roth T, Stubbs C, Walsh JK. Ramelteon (TAK-
375), a selective MT1/MT2-receptor against,
reduces latency to persistent sleep in a model
of transient insomnia related to a novel sleep
environment. Sleep 2005;28(3):303–8.
[6] Fava M, McCall WV, Krystal A, et al. Eszopiclone
co-administered coexisting with major depres-
sive disorder. Biol Psychiatry 2006;59:1052–60.
[7] Sores CN, Joffe H, Rubens R, et al. Eszopiclone in
patients with insomnia during perimenopause
and early postmenopause. Obstet Gynecol
2006;108:1402–10.
[8] Roth T, Walsh JK, Krystal A, et al. An evaluation
of the efficacy and safety of eszopiclone over
12 months in patients with chronic primary
insomnia. Sleep Med 2005;6(6):487–95.
[9] Roth T, Roehrs TA, Vogel GW, et al. Evaluation
of hypnotic medications. In: Prien RF,
Robinson DS, editors. Clinical evaluation of
psychotropic drugs: principles and guidelines.
New York: Raven Press; 1995. p. 579–92.
[10] Rush CR, Griffiths RR. Zolpidem, triazolam, and
temazepam: behavioral and subject-rated effects
in normal volunteers. J Clin Psychopharmacol
1996;16:146–57.
[11] Carskadon MA, Seidel WF, Greenblatt DJ, et al.
Daytime carryover of triazolam and flurazepam
in elderly insomniacs. Sleep 1982;5:361–71.
[12] Danjou P, Fruncillo PR, Worthington P, et al. A
comparison of the residual effects of zaleplon
and zolpidem following administration 5 to
2 h before awakening. Br J Clin Pharmacol
1999;48:367–74.
[13] Roehrs T, Merlotti L, Zorick F, et al.
Sedative, memory and performance effects
Sleep-Promoting Agents 185
of hypnotics. Psychopharmacology (Berl)
1994;116:130–4.
[14] Roehrs T, Zorick F, Sicklesteel J, et al. Effects of
hypnotics on memory. J Clin Psychopharmacol
1983;3:310–3.
[15] Curran HV, Birch B. Differentiating the
sedative, psychomotor and amnesic effects of
benzodiazepines: a study with midazolam
and the benzodiazepine antagonist, flumaze-
nil. Psychopharmacology (Berl) 1991;103:
519–23.
[16] Green JF, McElholm A, King DJ. A comparison of
the sedative and amnestic effects of chlorproma-
zine and lorazepam. Psychopharmacology (Berl)
1996;128:67–73.
[17] Mohler H, Crestani F, Rudolph U. GABA-A recep-
tor subtypes: a new pharmacology. Curr Opin
Pharmacol 2001;1:22–5.
[18] Paterniti S, Dufouil C, Alperovitch A. Long-term
benzodiazepine use and cognitive decline in the
elderly: the Epidemiology of Vascular Aging
Study. J Clin Psychopharmacol 2002;22:285–93.
[19] Allard J, Artero S, Ritchie K. Consumption of
psychotropic medication in the elderly: a re-
evaluation of its effect on cognitive performance.
Int J Geriatr Psychiatry 2003;18:874–8.
[20] McAndrews MP, Weiss RT, Sandor P, et al. Cognitive
effects of long-term benzodiazepine use in older
adults. Hum Psychopharmacol 2003;18:51–7.
[21] Roehrs TA, Vogel G, Roth T. Rebound insomnia:
its determinants and significance. Am J Med
1990;88:43S–6S.
[22] Krystal A, Walsh J, Roth T, et al. Evaluation of the
efficacy and safety of eszopiclone over six
months of treatment in patients with insomnia
[abstract]. Poster presented at Sleep 2006, Salt
Lake City UT, June 17–26, 2006.
[23] Roehrs T, Merlotti L, Zorick F, et al. Rebound
insomnia and hypnotic self administration. Psy-
chopharmacology (Berl) 1992;107:480–4.
[24] Woods JH, Winger G. Current benzodiazepine
issues. Psychopharmacology (Berl) 1995;118:
107–15.
[25] Mellinger GD, Balter MB, Uhlenhuth EH. Insom-
nia and its treatment. Arch Gen Psychiatry 1985;
42:225–32.
[26] Roehrs T, Hollebeek E, Drake C, et al. Substance
use for insomnia in Metropolitan Detroit.
J Psychosom Res 2002;53:571–6.
[27] Krystal AD, Walsh JK, Laska E, et al. Sustained
efficacy of eszopiclone over 6 months of nightly
treatment: results of a randomized, double-
blind, placebo-controlled study in adults with
chronic insomnia. Sleep 2003;26:793–9.
[28] Roehrs T, Bonahoom A, Pedrosi B, et al. Dis-
turbed sleep predicts hypnotic self administra-
tion. Sleep Med 2002;3:61–6.
[29] Roehrs T, Pedrosi B, Rosenthal L, et al. Hypnotic
self administration and dose escalation. Psycho-
pharmacology (Berl) 1996;127:150–4.
[30] Roehrs T, Bonahoom A, Pedrosi B, et al.
Nighttime versus daytime hypnotic
186 Roth & Roehrs
self-administration. Psychopharmacology (Berl)
2002;161:137–42.
[31] Griffiths RR, Johnson MW. Relative abuse liability
of hypnotic drugs: a conceptual framework and al-
gorithm for differentiating among compounds.
J Clin Psychiatry 2005;66(Suppl 9):31–41.
[32] Mendelson WB. The use of sedative/hypnotic
medication and its correlation with falling
down in the hospital. Sleep 1996;19:698–701.
[33] Ensrud KE, Blackwell TL, Mangione CM, et al.
Central nervous system-active medications and
risk for falls in older women. J Am Geriatr Soc
2002;50:1629–37.
[34] Kallin K, Lundin-Olsson L, Jensen J, et al. Predis-
posing and precipitating factors for falls among
older people in residential care. Public Health
2002;116:263–71.
[35] Thapa PB, Gideon P, Cost TW, et al. Antidepres-
sants and the risk of falls among nursing home
residents 263–271. N Engl J Med 1998;339:
875–82.
[36] Brassington GS, King AC, Bliwise DL. Sleep
problems as a risk factor for falls in a sample
of community-dwelling adults aged 64–99 years.
J Am Geriatr Soc 2000;48:1234–40.
[37] Avidan AY, Fries BE, James ML, et al. Insomnia
and hypnotic use recorded in the minimum
data set, as predictors of fall and hip fractures
in Michigan nursing homes. J Am Geriatr Soc
2005;53:955–62.
[38] Shader RI, Greenblatt DJ. Triazolam and antero-
grade amnesia: all is not well in the z-zone.
J Clin Psychopharmacol 1983;3:273.
[39] Morris HH, Estes ML. Traveler’s amnesia.
Transient global amnesia secondary to triazo-
lam. JAMA 1987;258:945–6.
[40] Yang W, Dollear M, Muthukrishnan SR. One rare
side effect of zolpidem—sleepwalking: a case
report. Arch Phys Med Rehabil 2005;86:1265–6.
[41] Liskow B, Pikalov A. Zaleplon overdose associ-
ated with sleepwalking and complex behavior.
J Am Acad Child Adolesc Psychiatry 2004;43:
927–8.
[42] Lange CL. Medication-associated somnambu-
lism. J Am Acad Child Adolesc Psychiatry 2005;
44:211–2.
[43] Paquet V, Strul J, Servais L, et al. Sleep-related
eating disorder induced by olanzapine. J Clin
Psychiatry 2002;63:597.
[44] Lu ML, Shen WW. Sleep-related eating disorder
induced by risperidone. J Clin Psychiatry 2004;
65:273–4.
[45] Morgenthaler TI, Silber MH. Amnestic sleep-
related eating disorder associated with zolpidem.
Sleep Med 2002;3:323–7.
[46] Vetrugno R, Manconi M, Strembi LF. Nocturnal
eating: sleep related eating disorder or nocturnal
eating syndrome? A videopolysomnographic
study. Sleep 2006;29:949–54.
[47] Menkes DB. Triazolam-induced nocturnal binge-
ing with amnesia. Aust N Z J Psychiatry 1992;26:
320–1.
[48] Schenk CH, Mahowald MW. Review of nocturnal
sleep-related eating disorders. Int J Eat Disord
1994;15:343–56.
[49] Lauerma H. Nocturnal wandering caused by
restless legs and short-acting benzodiazepines.
Acta Psychiatr Scand 1991;83:492–3.
[50] Mendelson WB, Thompson C, Franko T. Adverse
reactions to sedative hypnotics: three years’ expe-
rience. Sleep 1996;19:702–6.
[51] Mohler H, Fritschy JM, Rudolph U. A new benzo-
diazepine pharmacology. J Pharmacol Exp Ther
2002;300:2–8.
[52] Kato K, Hirai K, Nichiyama K, et al. Neurochem-
ical properties of ramelteon (TAK-375),
a selective MT1/MT2 receptor agonist. Neuro-
pharmacology 2005;48:301–10.
[53] Karin A, Tolbert D, Cao C. Disposition kinetics
and tolerance of escalating single doses of remel-
teon, a high-affinity MT1 and MT2 melatonin
receptor agonist indicated for treatment of
insomnia. J Clin Pharmacol 2006;46:140–8.
[54] Mohler H, Fritschy JM, Rudolph U. A new
benzodiazepine pharmacology. J Pharmacol
Exp Ther 2002;300:2–8.
[55] Mitler MM, Seidel WF, Van Den Hoed J, et al.
Comparative hypnotic effects of flurazepam,
triazolam, and placebo: a long-term simulta-
neous nighttime and daytime study. J Clin
Psychopharmacol 1984;4:2–13.
[56] Walsh JK, Krystal AD, Amato DA, et al. Nightly
treatment of primary insomnia with eszopiclone
for six months: effect on sleep, quality of life and
work limitations. Sleep 2007;30:959–68.
[57] Compton-McBride S, Schweitzer PK, Walsh JK.
Most commonly used drugs to treat insomnia
in 2002. Sleep, in press.
[58] Brogden RN, Heel RC, Speight TM, et al. Trazo-
done: a review of its pharmacologic properties
and therapeutic use in depression and anxiety.
Drugs 1981;21:401–29.
[59] Jenck F, Moreau JL, Mutel V, et al. Evidence for
a role of 5-HT1C receptors in the antiserotoner-
gic properties of some antidepressant drugs. Eur
J Pharmacol 1993;231:223–9.
[60] Frazer A. Pharmacology of antidepressants.
J Clin Psychopharmacol 1997;17(Suppl 1):
2S–18S.
[61] Preskorn SH. Pharmacokinetics of antidepres-
sants: why and how they are relevant to treat-
ment. J Clin Psychiatry 1993;54(Suppl):14–34.
[62] Richelson E. The pharmacology of antidepres-
sants at the synapse: focus on newer compounds.
J Clin Psychiatry 1994;55(Suppl A):34–9.
[63] De Boer T. The pharmacological profile of mirta-
zapine. J Clin Psychiatry 1996;57(Suppl 4):
19–25.
[64] Radhakishun FS, van den Bos J, van der
Heijden BC, et al. Mirtazapine effects on alert-
ness and sleep in patients as recorded by interac-
tive telecommunication during treatment with
different dosing regimens. J Clin Psychopharma-
col 2000;20(5):531–7.

[65] Montgomery I, Oswald I, Morgan K, et al. Trazo-
done enhances sleep in subjective quality but
not in objective duration. Br J Clin Pharmacol
1983;16:139–44.
[66] Walsh JK, Erman M, Erwin CW, et al. Subjective
hypnotic efficacy of trazodone and zolpidem in
DSM-III-R primary insomnia. Hum Psychophar-
macol 1998;13:191–8.
[67] Riemann D, Voderholzer U, Cohrs S, et al.
Trimipramine in primary insomnia: results of
a polysomnographic double-blind controlled
study. Pharmacopsychiatry 2002;35:165–74.
[68] Hohagen F, Montero RF, Weiss E, et al. Treatment
of primary insomnia with trimipramine: an
alternative to benzodiazepine hypnotics? Eur
Arch Psychiatry Clin Neurosci 1994;244(2):65–72.
[69] Hajak G, Rodenbeck A, Voderholzer U, et al.
Doxepin in the treatment of primary insomnia:
a placebo-controlled, double-blind, polysomno-
graphic study. J Clin Psychiatry 2001;62:453–63.
[70] Roth T, Rogowski R, Hull S. Efficacy and safety of
doxepin 1 mg, 3 mg, and 6 mg in adults with pri-
mary insomnia. Sleep 2007;30:1555–61.
[71] Golden RN, Dawkins K, Nicholas L. Trazadone
and nefazodone. In: Schatzberg A, Nemeroff C,
editors. The American psychiatric textbook of
psychopharmacology. Washington, DC: Ameri-
can Psychiatric Textbook, Inc.; 2004. p. 315–25.
[72] Levenson JL. Prolonged QT interval after trazo-
done overdose. Am J Psychiatry 1999;156:
969–70.
[73] Haria M, Fitton A, McTavish D. Trazodone. A re-
view of its pharmacology, therapeutic use in
depression and therapeutic potential in other
disorders. Drugs Aging 1994;4:331–55.
[74] Carson CC III, Mino RD. Priapism associated
with trazodone therapy. J Urol 1988;139:
369–70.
[75] Roehrs T, Roth T.‘Hypnotic’ prescription patterns
in a large managed-care population. Sleep Med
2004;5:463–6.
[76] Montplaisir J, Allen RP, Walters AS, et al. Restless
legs syndrome and periodic leg movements
during sleep. In: Kryger MH, Roth T,
Dement WC, editors. Principles and practice of
sleep medicine. 4th edition. Philadelphia: Elsev-
ier Saunders; 2005. p. 839–52.
[77] Mahowald MW, Schenck CH. REM sleep para-
somnias. In: Kryger MH, Roth T, Dement WC,
editors. Principles and practice of sleep medi-
cine. 4th edition. Philadelphia: Elsevier Saun-
ders; 2005. p. 897–916.
[78] Ancoli-Israel S, Roth T. Characteristics of insom-
nia in the United States: results of the 1991
National Sleep Foundation Survey. I. Sleep
1999;22:S347–53.
Sleep-Promoting Agents 187
[79] Gallup. Sleep in America. Princeton (NJ): Gallup
Organization; 1995. p. 1–78.
[80] Johnson EO, Roehrs T, Roth T, et al. Epidemiol-
ogy of alcohol and medication as aids to sleep in
early adulthood. Sleep 1998;21:178–86.
[81] Roehrs T, Roth T. Sleep, sleepiness, sleep disor-
ders and alcohol use and abuse. Sleep Med Rev
2001;5:287–97.
[82] Roehrs T, Papineau K, Rosenthal L, et al. Ethanol
as a hypnotic in insomniacs: self administration
and effects of sleep and mood. Neuropsycho-
pharmacology 1999;20:279–86.
[83] Roehrs T, Blaisdell B, Cruz N, et al. Tolerance to
hypnotic effects of ethanol in insomniacs. [ab-
stract]. Sleep 2004;27:A52.
[84] Roehrs TA, Blaisdell B, Richardson GS, et al.
Insomnia as a path to alcoholism: dose escala-
tion [abstract]. Sleep 2003;26:A307.
[85] Kudo Y, Kurihara M. Clinical evaluation of
diphenhydramine hydrochloride for the treat-
ment of insomnia in psychiatric patients:
a double-blind study. J Clin Pharmacol 1990;
30:1041–8.
[86] Rickels K, Morris RJ, Newman H, et al. Diphen-
hydramine in insomniac family practice pa-
tients: a double-blind study. J Clin Pharmacol
1983;23(5–6):234–42.
[87] Richardson GS, Roehrs TA, Rosenthal L, et al.
Tolerance to daytime sedative effects of H1 anti-
histamines. J Clin Psychopharmacol 2002;22:
511–5.
[88] Pearson NJ, Johnson LL, Nahin RL. Insomnia,
trouble sleeping, and complementary and alter-
native medicine. Arch Intern Med 2006;166:
1775–82.
[89] Taibi DM, Landis CA, Petry H, et al. A systematic
review of valerian as a sleep aid: safe but not
effective. Sleep Med Rev 2007;11:209–30.
[90] Muller CE, Schumacher B, Brattstrom A, et al.
Interactions of valerian extracts and a fixed
valerian-hop extract combination with adeno-
sine receptors. Life Sci 2002;71:1939–49.
[91] Clinical Practice Review Committee, American
Academy of Sleep Medicine. Oral nonprescrip-
tion treatment for insomnia: an evaluation of
products with limited evidence. J Clin Sleep
Med 2005;1:173–87.
[92] Roehrs T, Roth T. Drugs, sleep disorders, and
aging. Clin Geriatr Med 1989;5:395–404.
[93] Schweitzer PK. Drugs that disturb sleep and
wakefulness. In: Kryger MH, Roth T,
Dement WC, editors. Principles and practice of
sleep medicine. 4th edition. Philadelphia: Elsev-
ier Saunders; 2005. p. 499–518.
[94] Brower KJ. Insomnia, alcoholism and relapse.
Sleep Med Rev 2003;7:523–39.
 189

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 189–204

Nonpharmacologic Strategies
in the Management of Insomnia:
Rationale and Implementation
Paul B. Glovinsky, PhDa,b,*, Chien-Ming Yang, PhDc,
Boris Dubrovsky, PhDd, Arthur J. Spielman, PhDa,d

- Neurophysiologic models of insomnia Sleep hygiene education

- Psychologic and behavioral factors Stimulus control instructions
affecting sleep Sleep restriction therapy
Dysfunctional sleep cognitions Relaxation training
Behaviors adversely affecting sleep Cognitive therapy
Emotional arousal Chronotherapy
- The 3P model of insomnia Light therapy
- Evaluation of insomnia - Selecting and delivering cognitive
- Cognitive behavioral interventions behavioral treatments
for insomnia - References

As a physician, you probably offer a fair amount of
support and encouragement along with more spe-
cific recommendations. Whether you are exhorting
your patients to diet, exercise, or more assiduously
monitor blood sugar, the underlying message they
are likely to hear is that they should be making
more of an effort. This advice is not necessarily
regrettable, because trying harder is generally benefi-
cial when coping with chronic illness; however, the
situation grows trickier when dealing with insom-
nia. Trying harder to sleep often proves counterpro-
ductive. Natural sleep can occasionally be coerced
into making an appearance, for example, by
prolonged sleep deprivation, but typically the pros-
pects for sleep dissipate when too much force is
applied.
Patients who plant themselves in bed hoping
to net an adequate amount of sleep even if it takes
all night and half the morning usually end up with
sleep that is unsatisfying and broken. The same is
true of those who obsess over the ingredients of
a good night, searching for a fail-safe recipe. The
best way to fall asleep and stay asleep is not to think
so much about it. This prescription is easy enough
for good sleepers to follow but not so for those
who have been plagued by months or years of sleep

a
Department of Psychology, The City College of New York NAC 7/120, 138th Street and Convent Avenue,
New York, NY 10031, USA
b
St. Peter’s Sleep Center, Pine West Plaza # 1, Washington Avenue Ext., Albany, NY 12205, USA
c
Department of Psychology, National Chengchi University, 64, Sec. 2, Chih-Nan Road, Taipei, Taiwan 116
d
Center for Sleep Disorders Medicine and Research, New York Methodist Hospital, 519 Sixth Street, Brooklyn,
NY 11215, USA
* Corresponding author. St. Peter’s Sleep Center, Pine West Plaza # 1, Washington Avenue Ext., Albany, NY
12205.
E-mail address: glov@earthlink.net (P.B. Glovinsky).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.007
sleep.theclinics.com
190 Glovinsky et al
difficulties. These poor sleepers have learned to
their chagrin not only that their sleep is unreliable
but also that they can conjure up a night of sleep-
lessness ‘‘out of thin air’’ merely by imagining it
might happen.
This article surveys cognitive behavioral treat-
ments for insomnia (CBT-I), providing a basic
review of their theoretic rationale and practical
strategies for their implementation. Substantial evi-
dence demonstrates that CBT-I offers effective and
sustained benefits for patients contending with
insomnia, regardless of whether the sleep distur-
bance is of a primary nature [1–4] or secondary to
a psychiatric or medical disorder [5–10]. The cur-
rent evidence-based view is that non-drug treat-
ments are at least as effective as pharmacologic
remedies [11–14].
Demonstrated efficacy notwithstanding, CBT-I
can present as a motley group of interventions. In
the text to follow, we discuss therapies that involve
altering bedtime schedules, initiating worry jour-
nals, making rounds between the bed and an easy
chair, muscle relaxation, bright light exposure,
mental imagery, and more. Is the only attribute
that binds such an array the fact that they are non-
pharmacologic? We submit that the therapies have
something more substantial in common, that is,
CBT-I represents efforts to counter the unreliable
and evanescent nature of sleep, especially as it is
experienced by veteran insomniacs, by recruiting
from among a wide range of processes known to
affect sleep—be these physiologic, cognitive, behav-
ioral, environmental, or social—and aligning them
in the service of sleeping well.
In their roundabout approach, CBT-I presents
a contrast with pharmacologic treatments, which
act directly on neurotransmitters subserving sleep
to increase the propensity of falling and staying
asleep. Of course sleeping pills do require modest
cooperation from patients (in terms of maintaining
behavioral and cognitive quiescence) to work effec-
tively. Typically, poor sleepers who have just taken
a hypnotic are able to do their part in this regard.
Instead of becoming more keyed up as night falls,
they can relax as they perceive the balance tipping
in favor of sleep. They may even relish the sense
that the impending slide is a fait accompli, that is,
this time out they will not be able to sabotage their
own prospects.
An unheralded but important psychologic effect
of sleeping pills, therefore, is that they allow poor
sleepers to cede control of the mission to sleep. Al-
though beneficial in the short run, this consequence
of resorting to sleeping pills is a core component of
the psychologic dependence that can accrue with
long-term use. Poor sleepers tend to attribute what-
ever sleep they have accumulated to the pill itself,
shortchanging their own contributions. Anticipa-
tory anxiety and pharmacologically based with-
drawal symptoms often collude to produce
‘‘rebound insomnia’’ when such patients refrain
from sleeping pills, confirming the need for contin-
ued use. CBT-I is useful in transitioning patients
away from sleeping pill dependency, because it pro-
vides means by which these patients can improve
their sleep by dint of their own efforts.
Given that CBT-I draws upon wide-ranging
models of sleep/wake functioning, we feel it appro-
priate to review these explanatory models. A deeper
understanding is not only of academic interest but
also clinically useful, providing a sturdy theoretic
framework with which to buttress treatment recom-
mendations. In contrast to sleeping pills, CBT-I is
not coated with face validity. These treatments
may be met with initial skepticism on the part of
patients because they typically involve interven-
tions that run counter to current practices. Success
often hinges on fostering patience in the weeks
before the beneficial effects of treatment become
apparent. This goal is best accomplished by accom-
panying your seemingly outlandish advice with
a solid theoretic rationale.
The following sections provide a brief overview
of strategies for evaluating insomnia, with the aim
of helping one elicit clinical material than can guide
the choice of treatment. Readers interested in more
in-depth treatment of this topic are referred to sev-
eral comprehensive reviews [15–17]. We then
describe the practical application of CBT-I, high-
lighting critical issues and potential pitfalls.
Neurophysiologic models of insomnia
Insomnia can be construed as a disruption of the
balance between three major neural systems regu-
lating sleep: (1) a homeostatic system that with
each passing hour of wakefulness increases the pro-
pensity to sleep; (2) a circadian process that gener-
ates a biologic rhythm of sleep and wake tendency
irrespective of recent sleep history; and (3) an
arousal system that promotes wakefulness, counter-
ing the homeostatic sleep drive [18–20].
The homeostatic system works to maintain an
adequate amount of total sleep over successive
nights. The level of sleep drive present on the basis
of this homeostatic mechanism is at any given time
determined by prior durations of sleep and wake-
fulness. If an individual’s sleep is curtailed, this
leads to an augmented sleep drive and an increased
likelihood of accumulating extra recovery sleep dur-
ing subsequent time in bed, thereby restoring the
balance. Oversleeping and napping by contrast
reduce the homeostatic sleep drive, leading to
shorter or lighter stints of sleep.

The circadian system is based on an internal clock
in the hypothalamus that generates a rhythm of
sleepiness and alertness independent of prior sleep
history. Studies of animals and humans have iden-
tified the genetic basis of this cycle [21,22]. The typ-
ical endogenous circadian cycle in human beings
has a period of slightly over 24 hours [23,24]; there
is an innate tendency for our bedtimes and rising
times to slowly drift later around the clock. This
propensity is often revealed during vacation pe-
riods, when we are not as constrained by externally
imposed schedules. Exposure to environmental
time cues, especially daylight or other bright light,
can stabilize an endogenous circadian rhythm that
would otherwise ‘‘free run.’’
The arousal system counteracts the sleep drive
through the promotion of alertness. Activated by
internal thoughts and emotions as well as by exter-
nal stimulation, it can be viewed as a mobilizing
system intended to arouse the organism when it is
at risk. In contrast to the homeostatic mechanism
that gradually strengthens the sleep drive as our
waking hours pass, alertness can soar in a moment,
as required in an emergency. Although this arrange-
ment may be adaptive from an evolutionary stand-
point, it works against the prospects for sleep in
insomnia patients, whose careful preparations for
sleep can be overturned in an instant by an errant
thought.
Several models highlighting the role of arousal as
a cause of insomnia have been proposed [25,26].
Individuals with insomnia have been shown to
have elevated autonomic activity, indicated by
a higher metabolic rate, body temperature, heart
rate, urinary cortisol and adrenaline excretion,
skin conduction, and muscle tension [27,28], as
well as increased cognitive processing around sleep
onset or during sleep as reflected by faster electroen-
cephalographic frequencies [29–31]. Recent studies
using event-related potentials [32] and PET imaging
[33] further demonstrate a relative inability to
lower general attentional or arousal processes as
well as impairment in the sleep-specific inhibitory
process associated with sleep initiation in patients
with insomnia.
Difficulty sleeping can arise from inherited
anomalies pertaining to each of the three systems
subserving sleep and wakefulness described previ-
ously. Some individuals possess an inherently
weak homeostatic sleep drive. Others are under
the sway of atypical circadian clocks. For example,
persons who have clocks that are strongly biased
toward sleep phase delay may comfortably follow
a ‘‘night owl’’ pattern under typical circumstances
but be at greater risk for sleep initiation difficulties
when stressed. Although persons who have hyper-
aroused ‘‘type A’’ constitutions may benefit from
Nonpharmacologic Management of Insomnia 191
that makeup during waking hours in terms of pro-
ductivity, their sleep may be especially vulnerable to
disruption.
In addition to inherited predispositions, insom-
nia may be learned. Formerly good sleepers can en-
gage in maladaptive behavioral practices that lead
to a weakened sleep drive, an attenuated circadian
sleep/wake cycle, or hyperarousal at bedtime. For
example, the freedom from work-imposed sleep
schedules that comes with retirement can lead to
irregular sleep patterns and eventually to insomnia
even as the stakes regarding performance the next
day have been reduced.
Psychologic and behavioral factors
affecting sleep
Although the three physiologically based systems
described previously interact to create a stable
sleep/wake cycle under ideal conditions, this out-
come is by no means assured because sleep is so
susceptible to interruption by psychologic and
behavioral factors. Sleep is readily deferred (at least
in the short run) in the aftermath of emotionally
charged experiences, to complete tasks deemed suf-
ficiently important, or to maintain vigilance in the
face of real or perceived threats. This tilt toward
wakefulness renders sleep, in the minds of chroni-
cally poor sleepers, vulnerable and capricious.
Dysfunctional sleep cognitions
Several studies have provided evidence that anxious
or dysphoric thoughts are likely to inhibit sleep
[34–39]. Insomnia is clearly associated with worry-
ing that has a ‘‘real world’’ basis or with exposure to
traumatic events; however, even neutral cognitions
such as planning for the week ahead can interfere
with sleep. Sometimes it is not so much the content
as the form of thinking that proves deleterious.
Many patients insist they are not particularly wor-
ried about anything. Rather, their minds simply
‘‘will not shut off’’; they may race instead from
thought to thought or incessantly replay an adver-
tising jingle.
A special case obtains when concern centers on
sleep itself. A string of poor nights provides fodder
for lamenting the past and worrying about the
future. Recalling long hours of nocturnal wakeful-
ness and consequent daytime listlessness and antic-
ipation of more of the same is a surefire way to avert
sleep. A similar pattern may be seen if patients
harbor unreasonable expectations about sleep.
Poor sleepers who feel that their insomnia is ‘‘inev-
itable’’ and will necessarily lead to dire health con-
sequences may become frenzied by bedtime. The
sleeplessness that predictably follows only cements
their original premonitions. Directly challenging
192 Glovinsky et al
dysfunctional thinking about sleep through edu-
cation and cognitive therapy has been shown to
improve sleep [40,41].
Behaviors adversely affecting sleep
The impact of evening activity on subsequent sleep
quality is underappreciated. Poor sleepers are often
surprised to learn that they need at least 4 hours in
which to wind down before sleep. They cannot
expect to work an evening shift, finish up an assign-
ment brought home from the office, or play in an
evening basketball league and still fall asleep at a rel-
atively early hour. Even seemingly relaxing activities
such as calling friends or surfing the Web may prove
sufficiently stimulating to preclude sleep. Poor
sleepers should strive instead to disengage from
real world concerns, be they work related or social,
in the hours before bedtime.
A particularly pernicious set of behaviors origi-
nates in the desire to compensate for poor sleep
or its daytime effects [42]. It is common to hear
that patients will sleep into the morning following
a disrupted night when their schedules allow or
take an afternoon nap. These responses may bring
short-term relief from the effects of sleep loss but
at the cost of reducing the amount of sleep drive
available to induce sleep at bedtime, or of disrupt-
ing the circadian processes that regulate sleep and
wakefulness. Box 1 lists common daily life practices
that may interfere with sleep in these ways.
Emotional arousal
Transient insomnia in the face of acute stress is
a nearly universal experience resulting directly
from autonomic nervous system activation and
hormone release subserving the ‘‘fight or flight’’
reaction [27]; however, some individuals are more
vulnerable to chronic sleep disruption than others
[43]. Persons who internalize conflicts through
self-inhibition, denial, or suppression seem to be
more susceptible to sleeplessness [44]. The need
for perfection and the need to maintain control
are associated with insomnia, just as are more pre-
dictable psychologic traits such as a predisposition
to anxiety and depression [44–47]. Arousal can
also be a learned response, appearing in specific
contexts such as the approach of bedtime or entry
into the bedroom after repeated pairings of these
contexts with the experience of sleeplessness.
Once such associative links are established, bed-
time with its attendant rituals begins to offer con-
textual cues for arousal rather than sleep [48].
Fig. 1 offers a schematic representation of how
behavioral and psychologic factors influence sleep
through the mediation of the three neural systems
described above.
The 3P model of insomnia
The waxing and waning of the various physiologic,
psychologic, and behavioral factors contributing to
insomnia and their interaction can complicate clin-
ical assessment of the disorder. We have introduced
a model that has proven useful for understanding
the genesis of a particular case of insomnia and
focusing treatment efforts. Termed the 3P model, it
groups etiologic factors temporally into predispos-
ing characteristics, precipitating events, and perpet-
uating attitudes and practices [49].
Predisposing characteristics are often present for
years before chronic insomnia takes hold. Many
are thought to be congenital, such as tendencies
toward physiologic or cognitive hyperarousal, or
innate preferences for activity in the evening versus
the morning. The 3P model allows for acquired pre-
disposing factors as well. For example, residual pain
following an injury may not in itself be accompa-
nied by chronic insomnia, but it can lower the
threshold for the disorder’s appearance.
Precipitating events within the 3P model cor-
respond to what patients are wont to label as the
‘‘cause’’ of insomnia. Appearing just before or concur-
rently with the sleep disturbance, typical precipitating
events lead to a few nights or even weeks of poor
sleep in just about everyone. They may be as dire
as divorce or serious illness or elating as a newborn.
Precipitating events may also be fairly innocuous,
such as the acquisition of a new mattress or the
introduction of flexible starting times on the job.
The disruption associated with a precipitating event
usually subsides with the passage of time, and sleep
generally regroups in turn.
By the time people have labeled themselves ‘‘poor
sleepers’’ and presented this complaint to their phy-
sician, the precipitating events identified as triggers
of their sleeplessness are often long resolved. This
can be a source of consternation. A patient may ap-
pear years after a divorce and demonstrate convinc-
ingly that she has moved on with her life yet still be
unable to count on a good night’s sleep. In this case,
perpetuating attitudes and practices, the third com-
ponent of the 3P model, have likely become pre-
dominant. As we have seen, the experience of sleep
disturbance on a chronic basis becomes self-sustain-
ing. Poor sleepers begin to associate bedtime and
their bedrooms with an anxious hyperaroused state,
and they settle for short-term relief from the effects
of sleep loss through ultimately maladaptive mea-
sures such reliance on caffeine or frequent napping
(Fig. 2).
Perpetuating factors are what telescope acute sleep
disruption into chronic insomnia; as such, they
often present the most opportune targets for behav-
ioral treatment. Their presence is actually grounds
 Nonpharmacologic Management of Insomnia 193

Box 1: Daily life behaviors and sleep-related habits that may interfere with sleep
Practices that reduce homeostatic drive at bedtime
Daily life behaviors
 Insufficient activity during the day
 Lying down to rest during the day
Sleep-related habits
 Napping, nodding, and dozing off during the day or evening
 In a trance, semi-awake in the evening
 Spending too much time in bed
 Extra sleep on weekends
Practices that disrupt circadian regularity
Daily life behaviors
 Insufficient morning light exposure, leading to a phase delay in circadian rhythm
 Early morning light exposure, producing early morning awakening due to a phase advance in cir-
cadian rhythm
Sleep-related habits
 Irregular sleep-wake schedule
 Sleeping in in the morning during weekends
Practices that enhance the level of arousal
Daily life behaviors
 Consuming caffeine excessively or too late in the day
 Smoking in the evening
 Alcohol consumption in the evening
 Exercising in the late evening
 Late evening meal or fluid intake (may cause nocturnal acid reflux or frequent urination)
 Getting home late, leaving insufficient time to wind down
Sleep-related habits
 Evening apprehension regarding sleep
 Preparations for bed are arousing
 No regular pre-sleep ritual
 Distressing pillow talk
 Watching TV, reading, or engaging in other sleep-incompatible behaviors in bed before lights
out, or falling asleep with TV or radio left on
 Trying too hard to sleep
 Clock watching during the night
 Staying in bed during prolonged awakenings, or lingering in bed awake in the morning
 Non-conducive sleep environment, such as bed partner snoring, noises, direct morning sunlight,
or pets in the bedroom
Adapted from Yang CM, Spielman AJ, Glovinsky PB. Nonpharmacologic strategies in the management of Insomnia.
Psychiatr Clin North Am 2006;29(4):900; with permission.

for optimism. When patients have become disheart-
ened by the entanglement of their sleep with seem-
ingly intractable problems such as chronic illness
or the loss of financial security, addressing perpetu-
ating factors can yield moderate improvement rela-
tively quickly.
The other factors in the 3P model should not be
overlooked. Because predisposing characteristics
increase the risk of developing insomnia, any
mitigation of their contribution would be helpful.
A similar notion holds true for precipitating events.
These stressors pile onto the patient’s preexisting
propensity for sleep disturbance, eventually breach-
ing the threshold for insomnia. Addressing triggers
of sleep loss such as marital strife or performance
anxiety directly with targeted treatments can roll
back the level of sleep disturbance to a subclinical
state.
Box 2 lists factors contributing to insomnia cate-
gorized according to the 3P model, with each cate-
gory further subdivided by the three processes
governing sleep and wakefulness.
194 Glovinsky et al

Psychological/Behavioral Factors Neurophysiological Systems

Homeostatic
System

Behavioral
Practices

Sleep Circadian
Cognition Sleep
System

Emotional
Arousal

Arousal
System

Fig. 1. A conceptual model illustrating how psychologic/behavioral factors influence sleep through the media-
tion of neurophysiologic mechanisms. (Adapted from Yang CM, Spielman AJ, Glovinsky PB. Nonpharmacologic
strategies in the management of insomnia. Psychiatr Clin North Am 2006;29(4):898; with permission.

Evaluation of insomnia
The complex nature of insomnia presents a diagnos-
tic challenge for physicians practicing under rigor-
ous time constraints. There is a lot of territory to
cover, and it can be difficult to elicit a balanced
and well-articulated history. Patients may readily
provide a summary such as ‘‘I can’t fall asleep’’ or
‘‘I keep waking up through the night,’’ or review
a blow-by-blow account of last night’s fiasco, but
they often have a harder time making sense of their
sleeplessness in terms that facilitate diagnosis and
treatment. The physician’s task will be considerably
eased by the adoption of a structured interview and
the use of sleep logs.
Although a semi-structured interview for insom-
nia has been published [50] and others are in devel-
opment, one may have to use an abridged version
in daily practice. A comprehensive evaluation will
elicit the chief nocturnal sleep complaints, daytime

Fig. 2. The 3P model of insomnia, in this case illustrating the major contributions of precipitating and perpetu-
ating factors and the minor contribution of predisposing factors. (Adapted from Spielman AJ, Yang CM,
Glovinsky PB. Assessment techniques for insomnia. In: Kryger MH, Roth T, Dement WC, editors. Principles
and practice of sleep medicine. 4th edition. Philadelphia: Elsevier Saunders; 2005. p. 1405; with permission.)
 Nonpharmacologic Management of Insomnia 195

Box 2: Common contributing factors associated with the development of insomnia

Predisposing factors
Homeostatic process
 Abnormality or weakness of the neurophysiologic system that generates sleep
Circadian process
 Extreme circadian type as a trait (eg, ‘‘owls’’ predisposed to activity in the late evening or ‘‘larks’’
inclined to the early morning)
 Less flexible circadian system
Arousal system
 Anxiety-prone and depressive personality traits as well as tendencies toward neuroticism and
somatization lead to a higher level of emotional and physiologic arousal
 Personality traits associated with sustained level of arousal, such as perfectionism and excessive
need for control
 Heightened or more sensitive physiologic arousal system
Precipitating factors
Homeostatic process
 Lack of, or decrease of, daytime activities, such as retirement
Circadian process
 Change of sleep-wake schedule, such as jet lag or starting a night shift job
Arousal system
 Life stressors or events leading to emotional and physiologic distress
Perpetuating factors
Homeostatic process
 Increased resting in bed
 Discharge of the sleep drive by sleeping outside of the nocturnal sleep period through planned
daytime naps or inadvertent dozing
 Reduced daytime activities
Circadian process
 Sleeping in during weekend to catch up on sleep
Arousal system
 Dysfunctional beliefs and attitudes about sleep that lead to increased emotional arousal and
worries over sleep loss
 Conditioning between bedtime cues and arousal
Adapted from Yang CM, Spielman AJ, Glovinsky PB. Nonpharmacologic strategies in the management of insomnia.
Psychiatr Clin North Am 2006;29(4):900; with permission.

consequences, life circumstances at the onset of the
sleep disorder and during its subsequent course,
typical weekday and weekend sleep patterns, beliefs
about sleep and the likely effects of sleeplessness,
behavioral changes that have been made to com-
pensate for poor sleep, any concurrent sleep disor-
ders, prior treatments, general medical status,
medication and substance use, family history, and
assessments of psychologic and social functioning.
Both underlying sleep patterns and the extent of
variability masking those patterns can best be ap-
preciated by the use of a nightly sleep log kept for
1 or 2 weeks. We favor a graphic log for its ability
to quickly convey copious amounts of temporal
information as well as any changes occurring
from week to week. Patients clock retiring and rising
times (at night and for daytime naps), estimate how
long it took to first fall asleep, and indicate the du-
ration and distribution of subsequent sleep and
waking episodes. The night’s pattern is supplied in
the morning as a holistic impression rather than
by fastidiously watching the clock. Patients rate
sleep quality and, when getting into bed the next
night, their level of fatigue during the day just
passed. Other variables of interest such as caffeine
and alcohol intake, exercise, the phase of the men-
strual cycle, light exposure, and medication use may
also be logged. Besides aiding clinicians in assessing
196 Glovinsky et al

a complex and protean problem, sleep logs offer pa-
tients a wider perspective. They learn that their
sleep’s appearance is not totally random, that their
worst nights are, in fact, not representative, and,
hopefully, that over time sleep does respond to rec-
ommended interventions, even if that improve-
ment is not apparent every night (Fig. 3).
If the physician’s assessment suggests that a phys-
iologic sleep disorder such as obstructive sleep
apnea or periodic leg movements may be contribut-
ing to sleep disruption, referral for nocturnal poly-
somnographic recording would be indicated. One
may also decide that other medical or psychiatric
co-morbidities warrant additional evaluation.
Even so, CBT-I may proceed apace, because concur-
rent treatment of predisposing, precipitating, and
perpetuating factors often proves most effective.
Cognitive behavioral interventions
for insomnia
Most patients will accept and tolerate CBT-I [51].
Similar to the prevailing view among clinicians,
they typically conceive of these treatments in terms
of what they are not, that is, that they are nonph-
armacologic. Not having a pill to swallow may be
considered a plus in that there is no danger of
drug dependency; however, the treatments may
also be viewed as counterintuitive, requiring too
much effort, and of dubious efficacy. ‘‘Why should
I leave my bed in the middle of the night,’’ your
patient might protest, ‘‘then I’ll be losing sleep for
sure.’’ A common objection is that the treatments
do not provide short-term relief.
It is useful to anticipate negative attitudes and to
counter them with clear rationales for treatment
while at the same time lowering expectations. De-
clare at the outset that improvement will likely
take a few weeks rather than a few nights, and con-
trast this against the months or years that your
patient’s insomnia has resisted the strategies de-
ployed thus far. It is also helpful to aim for an early
victory. For example, total sleep time is not so read-
ily extended through CBT-I. Even sleeping pills
yield only about 30 minutes more sleep on average
over placebo in clinical trials. Nevertheless, pro-
longed sleep latencies can be reduced fairly quickly
with several of the treatments described in the next
sections. This outcome should not go unheralded.
Sleep hygiene education
Instruction regarding proper sleep hygiene is a front-
line approach to correcting maladaptive behaviors
that may have been identified from perusal of Box 1.
Sleep hygiene is premised on the notion that practi-
cally every decision we make in our waking lives affects
sleep. Sometimes the issues are straightforward, such
as when a double espresso is habitual after dinner.
Others are more open to debate, such as whether
one can really ‘‘wind down’’ with a video game

EXAMPLE:
Into bed Out of bed
Fatigue
Morning’s Time to Amount of Rating
6 7 8 9 10 11 Mid 1 2 3 4 5 6 7 8 9 10 11 Noon 1 2 3 4 5 6 Date Fall asleep Sleep 1 Lo—10 Hi
1 C Mo12/10 100 min 5 hours 3
C
1 = Benadryl, 50 mg Caffeine

Medication Asleep

Fatigue
Morning’s Time to Amount of Rating
6 7 8 9 10 11 Mid 1 2 3 4 5 6 7 8 9 10 11 Noon 1 2 3 4 5 6 Date Fall asleep Sleep 1 Lo—10 Hi

Medication 1 ______________ Dosage _________ Medication 2 ______________ Dosage _________

Alcohol Day 1 ___________2 __________3 ___________4 __________5 __________6 __________ 7__________

Fig. 3. A sample sleep log.

before bed, or whether carving out a home office in
the bedroom is such a good idea.
Clinicians and patients often discount sleep hy-
giene education, albeit, for different reasons. Clini-
cians tend to regard the guidelines as self-evident,
assuming that someone who has had trouble sleep-
ing for years would know to silence the TV and ban-
ish the golden retriever from bed. Patients may
acknowledge that sleep hygiene directives are gener-
ally beneficial but just not applicable in their case.
They have tried following the rules to no avail. Years
of trial and error have honed their tactics, and even
if their sleep is still disrupted, it only gets worse
when they forego a nightcap or sleep in a room
that is too dark. It should not surprise the physician
to discover that simply handing out a list of ‘‘Good
Sleep Hygiene Practices’’ does not trigger an epiph-
any. Sleep hygiene really does require education,
a process of building upon general principles to
help patients arrive at an understanding of how
their thoughts and actions affect their prospects
for sleep. They must also learn not to expect a payoff
within a night or two after instituting the recom-
mendations. Improved sleep hygiene by itself is
often insufficient to uproot chronic insomnia,
although it should render sleeplessness more ame-
nable to other CBT-I.
Stimulus control instructions
Stimulus control instructions were one of the first
behavioral interventions specifically developed to
treat insomnia [48]. They target the maladaptive
association between bedtime cues and conditioned
arousal that strengthens each time the act of going
to bed leads to a sleepless night. Stimulus control
instructions treatment accomplishes this by banish-
ing from the bed behaviors that are incompatible
with sleep, such as eating, watching TV, reading,
or just worrying about being unable to sleep. An
exception is made for sex. Patients are instructed
to get out of bed if not asleep after about 20 min-
utes and to sit in a chair reading, listening to music,
or engaged in some other only mildly stimulating
activity, returning to bed when they feel sleepy.
See Box 3 for detailed instructions.
Initially, patients may accrue considerable sleep
loss following the 20-minute rule, which will
enhance the sleep drive and eventually foster more
rapid sleep onset. Patients should be forewarned
that this sleep loss may also lead to daytime deficits.
Repeated association of bedroom cues with rapid
sleep onset rather than with tossing and turning,
such as before treatment, is said to bring sleep under
the ‘‘stimulus control’’ of the bedroom environment.
Stimulus control instructions treatment may raise
several objections from patients. As noted previ-
ously, they may argue that the treatment guarantees
Nonpharmacologic Management of Insomnia 197
Box 3: Stimulus control instructions
1. Go to sleep only when you feel sleepy.
2. Do not use your bed or bedroom for any-
thing except sleep (sexual activity is the
only exception).
3. If you have not fallen asleep within approx-
imately 20 minutes, get up and go into an-
other room. Engage in relaxing activities,
such as non–work-related light reading,
and go back to bed when you feel sleepy
or ready for sleep.
4. If you cannot fall back to sleep, repeat step
3.
5. Set the alarm for the same time each
morning.
Adapted from Bootzin RR. Stimulus control
treatment for insomnia. Proc Am Psychol Assoc
1972;7:395.
sleep loss, and that there would be at least a chance
of their returning to sleep if they stayed in bed.
Some patients become wide awake when they
read. Others may grow despondent sitting up alone
while everyone else is fast asleep. It may help to re-
mind patients that their experiences with stimulus
control instructions are properly compared with
a night spent tossing and turning in bed rather
than to some idealized sleep experience. One might
also point out that their current pattern already en-
tails sleep loss, along with frustration and helpless-
ness. Stimulus control instructions, by contrast,
puts more control in their hands. Although the
treatment cannot make sleep appear on command,
it does allow patients to put their own stamp on the
wakefulness they experience at night (see Box 3).
Sleep restriction therapy
Sleep restriction therapy addresses both the weak-
ened homeostatic sleep drive and attenuated circa-
dian sleep/wake cycle characteristic of chronic
insomnia. Restricting time in bed to prescribed
hours leads to the gradual accumulation of signifi-
cant sleep debt and, in so doing, replenishes the
sleep drive [52]. It also ensures that sleep consis-
tently appears in the same relatively narrow time
slot rather than being strewn in snippets across
a wide span of nighttime and perhaps daytime
hours. The regular timing of sleep begins to reest-
ablish the circadian sleep/wake cycle. It also allays
the anticipatory anxiety that perpetuates chronic
sleeplessness.
Sleep restriction therapy is applied following a 1-
or 2-week log of baseline sleep. The estimated total
sleep time averaged across the log is initially used to
set time in bed. For example, a patient who retires
for nearly 9 hours but who reports sleeping for
198 Glovinsky et al
only 6 of these would be assigned 6 hours in bed
and asked to refrain from napping and oversleeping
on weekends. The exact times of retiring and rising
should factor in work and social obligations while
at the same time taking note of when patients are
most likely to be awake. Patients who generally
have difficulty falling asleep should be assigned
a relatively late bedtime, whereas those who log
broken sleep toward morning should rise early.
We set a lower limit of about 5 hours to avoid severe
sleep loss, even when patients claim they hardly
sleep at all.
In our original formulation of sleep restriction
therapy, the time in bed was adjusted weekly based
on subjective sleep efficiency (total sleep time/time
in bed  100%) according to the following rules:
(1) for a mean sleep efficiency greater than 90%
(85% in seniors), increase the time in bed by 15
minutes; (2) for a mean sleep efficiency less than
85% (80% in seniors), decrease the time in bed
by 15 minutes; and (3) for a mean sleep efficiency
greater than 85% and less than 90% (80% to 85%
in seniors), keep the time in bed the same [53].
More recently, we have employed an alternative
approach that, following acute restriction, progres-
sively increases the time in bed each week as long
as the patient’s subjectively estimated wake time
in bed is 45 minutes or less [54]. Fifteen-minute
increments in the time in bed are preferred to give
the sleep drive adequate time to build, but 30 min-
utes might be added if daytime sleepiness becomes
too pronounced. This variation of sleep restriction
therapy is often better tolerated, avoiding the dispir-
iting effect of reducing bedtime yet again after the
initial curtailment (Box 4).
Sleep restriction therapy can be a trying experi-
ence for patients, often on account of daytime
sleepiness. While acknowledging that this can be
a significant problem and issuing precautions
regarding driving or operating dangerous machin-
ery, one should also underscore that the sleepiness
is direct evidence of a strengthened sleep drive
which, when harnessed properly, will be at the
service of nocturnal sleep. For patients whose in-
somnia reflects chronic hyperarousal, the mere
appearance of sleepiness, even if initially ill timed,
may be welcomed as a sign that the homeostatic
mechanism is still viable. Patients may also object
that using an alarm clock to curtail precious sleep
is inane, that they are at a loss for things to do dur-
ing the extra hours of wakefulness, or that they
experience a ‘‘second wind’’ by the time their
assigned bedtimes roll around.
Such objections might be countered by drawing
patients’ attention to the shorter sleep latencies
and more consistent sleep that begin to appear on
their logs, improvements arising in part because
Box 4: Sleep restriction therapy
From the information provided on a sleep log
completed for at least 1 week, set the initial
time in bed equal to the reported average total
sleep time. To avoid severe sleep deprivation,
the minimum time in bed is 5 hours.
Version 1
A. Increase the time in bed by 15 to 30 minutes
when the average reported sleep efficiency
(sleep efficiency 5 average sleep time/time
in bed  100%) for 5 days is R90% (85%
in older individuals).
B. When the sleep efficiency from 5 days docu-
mented on a sleep log is <85% (80% in older
individuals), decrease the time in bed by 15
minutes.
C. When the sleep efficiency from 5 days docu-
mented on a sleep log is R85% and <90%
(R80 to <85% in older individuals), keep
the time in bed the same.
Version 2
After the original restriction, increase the time
in bed progressively by 15 or 30 minutes each
week, contingent upon a subjective total
wake time of 45 minutes or less, until the pa-
tient is spending 7 hours in bed. Further
changes are based on daytime functioning, fa-
tigue, and sleepiness.
Adapted from Yang CM, Spielman AJ, Glovinsky PB.
Nonpharmacologic strategies in the management of
insomnia. Psychiatr Clin North Am 2006;29(4):903;
with permission.
sleep is not allowed to run its course in the morning.
With sleep restriction therapy, there may not be any
fully satisfying nights of sleep, but there are also
fewer awful nights. Night-to-night variability in
sleep length and quality is replaced by predictable
strings of relatively short but consolidated sleep.
This change, in turn, lessens anticipatory anxiety
about what the coming night will bring. Patients
who experience a second wind are probably allow-
ing themselves to dip too close to sleep in the early
evening out of boredom, perhaps ‘‘zoning out’’ in
front of the TV, and should be counseled to become
more intellectually engaged or, if need be, physically
active after dinner. Because they will be likely stay-
ing up quite late while undergoing sleep restriction
therapy, there will still be plenty of time in which to
wind down before getting into bed.
Relaxation training
We have discussed how insomnia can be brought
about by maladaptive behaviors and thoughts.
Mental and physical repose, on the other hand,
facilitate sleep. Several classic cognitive behavioral

treatments for tension and anxiety have proved
useful in the management of insomnia. These treat-
ments include progressive relaxation, which re-
duces tension by sequentially tensing and relaxing
the main muscle groups [55,56]; autogenic train-
ing, which promotes relaxation by inducing sensa-
tions of warmth and heaviness [56]; guided
imagery, which provides a safe path for the mind
to follow, keeping it from straying into potentially
dangerous territory [57]; and biofeedback, helpful
in teaching patients to recognize and maintain
calmer behavioral and mental states [58,59].
It takes time and practice to learn how to relax.
Patients can quickly grow frustrated at their inabil-
ity to attain such a seemingly fundamental state.
Initial practice should not take place at bedtime
when the stakes are too high. Relaxation scripts
and inductions recorded on tapes or CDs may be
helpful. In any case, do not let the declaration
‘‘I just can’t relax’’ go unchallenged. Emphasize
that the ability to achieve physical and mental equi-
librium is an acquired skill much like riding a bicy-
cle. Once mastered, relaxation is useful precisely
because the state can be induced on cue. If sleep
is not quickly forthcoming, a restful state can be
summoned before agitation sets in and really ruins
the night. Restfulness is also pleasant in its own
right, besides serving as an effective staging ground
for sleep (Box 5).
Cognitive therapy
The dysfunctional thinking about sleep so often
found in cases of entrenched insomnia can be clas-
sified into five categories: (1) misconceptions con-
cerning the causes of insomnia, (2) misplaced
concerns regarding its consequences, (3) unrealistic
sleep expectations, (4) diminished perceptions of
control, and (5) mistaken beliefs about the predict-
ability of sleep [60]. Directly challenging such mis-
guided beliefs and attitudes through cognitive
therapy has been demonstrated to bring about
improvements in sleep [37,40].
One such therapy, cognitive restructuring, in-
volves identifying specific dysfunctional thoughts
and replacing them with more realistic assessments
and positive ideas [60]. A spirit of objectivity is fos-
tered, with clinician and patient as co-investigators.
For example, a patient may assume that he requires
at least 8 hours of sleep to stay awake the next day.
This claim might be evaluated using sleep logs and
repeated assessments on a simple sleepiness scale.
The opportunity might be taken to provide educa-
tion about how our circadian rhythms and arousal
system counter the effects of moderate sleep loss.
Another patient may arrive at your office frantically
clutching an Internet printout headlining research
on mortality and sleep. It may be useful to retrieve
Nonpharmacologic Management of Insomnia 199
Box 5: Relaxation training (progressive
muscle relaxation, autogenic training, slow
deep abdominal breathing, guided imagery)
1. Explain the rationale for the specific
technique.
2. The clinician demonstrates the technique
for the patient during an office visit.
3. The patient then practices the technique at
home once or twice a day (for about 10 min-
utes) in between office visits.
4. The clinician’s instructions and demonstra-
tion can be recorded, or commercial tapes
can be used to facilitate the practice at
home.
5. It may take a few weeks of practice before
the patient develops the skill required.
6. The patient is told not to use the technique
for sleep until a moderate level of skill in
performing the technique is achieved.
Adapted from Yang CM, Spielman AJ, Glovinsky PB.
Nonpharmacologic strategies in the management of
insomnia. Psychiatr Clin North Am 2006;29(4):910;
with permission.
the original journal article and place the findings in
context, noting similarities and differences with the
patient’s own circumstances.
Keeping a ‘‘worry journal’’ is another way to ad-
dress distressing thoughts that often preoccupy the
minds of would-be sleepers. In the evening before
the buffer period begins, the authors have patients
sit down and predict which issues and problems
are most likely to snag their descent into sleep during
the coming night, listing these on the left side of
their journal. On the right side, they are free to for-
mulate a credible solution, come up with a ‘‘quick
fix,’’ or just find a good way to procrastinate. The
goal here is not one of problem solving but rather
of getting off the hook. The journal is then closed
for the night. Committing a problem to paper,
affixed permanently on page 5, and paired with
even a flimsy solution has the effect of inoculating
the mind against this same problem racing around
it later on (Box 6).
Chronotherapy
Because the endogenous circadian pacemaker typi-
cally has a period slightly longer than 24 hours, sleep
has a propensity to drift progressively later around
the clock. To be able to fall asleep around the same
time each night, we must advance our sleep phase
a few minutes every cycle. Persons who have internal
clocks that run particularly slow or that are especially
hard to advance can reach a point where it habitually
takes hours to fall asleep. This problem, in turn,
leads to difficulty arising in the morning and
200 Glovinsky et al
Box 6: Cognitive therapy
1. Discuss with patients their general beliefs
regarding sleep and their sleep problems.
Subjective rating scales such as the Dysfunc-
tional Beliefs and Attitudes Scale [60] can be
used to help with the evaluation of patients’
sleep cognitions.
2. Identify counterproductive beliefs.
3. Enlist the patient as a co-investigator to
help gather data that will test and refute
the dysfunctional beliefs.
4. Provide correct information to address the
specific dysfunctional beliefs.
Adapted from Yang CM, Spielman AJ, Glovinsky PB.
Nonpharmacologic strategies in the management of
insomnia. Psychiatr Clin North Am 2006;29(4):910;
with permission.
tardiness for school or work. Sleep latencies are ap-
preciably shorter on weekends or during vacations,
when bedtimes are very late and oversleeping well
past noon is not uncommon. When this sleep pat-
tern is associated with adverse psychosocial conse-
quences, a circadian rhythm sleep disorder known
as delayed sleep phase disorder may be diagnosed
[61]. This condition is relatively common in adoles-
cents and young adults [62].
A specialized treatment for delayed sleep phase
disorder known as chronotherapy takes advantage
of the slightly slow circadian clock, which makes
sleep amenable to shifting later [63]. Chronother-
apy involves a progressive delay of bedtime, gener-
ally 3 hours per night, bringing sleep around the
clock to the desired earlier bedtime. The process is
analogous to traveling nearly all the way around
the world in a westward direction, only to end up
slightly east of one’s point of departure.
Before beginning chronotherapy, patients should
adhere for 1 week to a slightly restricted bedtime
schedule, starting when they are typically able to
fall asleep, for example, from 3 AM to 10 AM. Their
times of retiring and rising are then delayed by 3
hours until they approach the desired bedtime.
For example, after the first phase delay, patients
would be following a 6 AM to 1 PM schedule (per-
haps a typical weekend pattern) and after the sec-
ond, a 9 AM to 4 PM schedule. For the first time in
years, patients may wish to go to bed early. They
should be discouraged from doing so, with others
in the household enlisted to help keep them on
track. We have found it useful to ‘‘apply the brakes’’
by limiting the last few phase shifts to 1 hour, be-
cause there is a risk of overshooting the mark and
resuming a late pattern. Continuing with our exam-
ple, after bedtimes of 12 noon, 3 PM, 6 PM, and 9 PM,
with 7 hours of allotted time in bed at each step,
patients might then follow a slightly extended
10 PM to 6 AM schedule for a full week before finally
settling into an 11 PM to 7 AM routine (Box 7).
Chronotherapy does not counter the patient’s
innate phase delay tendencies, and it provides no
bulwark against further slippage once the target is
reached; however, it can provide relief in what had
been deemed an intractable situation. This improve-
ment can endure if patients reschedule their days as
well as nights, that is, when they eat, work, and play.
Whereas 10 PM was previously ‘‘prime time’’ for activ-
ities of all sorts, it will have to signal the start of the
wind-down period if sleep is to follow shortly.
For several days under chronotherapy, patients
will have no chance of meeting school or work
obligations. If the treatment cannot coincide with
a vacation, teachers or supervisors will have to
sign off on the absence. The middle phase is most
problematic in terms of compliance, when patients
are asked to retire in the afternoon and early even-
ing. There is also resistance against leading such
a regimented life. Patients ask if they will ever again
be able to enjoy a late night out with friends. You
might assure them that if they can come to see 2 AM
rather than 6 AM as ‘‘late’’ and limit oversleeping to
an hour or so rather than sleeping into the afternoon,
such outings can be accommodated.
Light therapy
Light is the most important influence on circadian
rhythms. Typically, bright outdoor light stabilizes
or ‘‘entrains’’ rhythms that might otherwise free
run. Exposure to sunlight can also shift the phase
of these rhythms, a phenomenon that lets us adapt
Box 7: Chronotherapy
1. The patient chooses a retiring time and aris-
ing time that can be adhered to for 1 to 2
weeks.
2. The patient completes a sleep log daily.
3. The retiring time and arising time are
delayed by 2 or 3 hours each day until they
are within 1 hour of the desired sleep
schedule.
4. For 1 week, the retiring and arising time are
maintained within 1 hour of the desired
schedule.
5. The retiring time and arising time are
delayed by 1 hour to the desired schedule.
6. The patient should not sleep later than the
scheduled arising time to help avoid
relapse.
Adapted from Yang CM, Spielman AJ, Glovinsky PB.
Nonpharmacologic strategies in the management of
insomnia. Psychiatr Clin North Am 2006;29(4):912;
with permission.

to new time zones following jet travel. For the past
20 years, artificial bright light has been used to reset
circadian rhythms in individuals who never leave
the ground. The size and direction of the phase shift
depend on the intensity and timing of the exposure
[64,65]. Morning light exposure in individuals fol-
lowing a nocturnal sleep schedule will tend to ad-
vance the circadian sleep phase. It counters our
inherent drift toward a later phase, helping main-
tain a 24-hour period. Evening light exposure, by
contrast, suppresses the expected melatonin secre-
tion and tends to delay the circadian sleep phase.
Light exposure upon awakening is now the treat-
ment of choice for most patients with a delayed
sleep phase [66]. Because the magnitude of the
phase shift is greatest when exposure occurs just af-
ter the circadian temperature nadir (approximately
2 hours before habitual awakening in most individ-
uals), there should be as little delay as possible be-
tween arising and light exposure. Initially, patients
should keep typical bedtimes, even if that means
arising at midday. As treatment proceeds, bedtimes
and rising times are progressively shifted, perhaps
an hour earlier each week, with light exposure also
shifting to immediately follow awakening (Box 8).
As is true for chronotherapy, bright light treat-
ment for delayed sleep phase disorder is abetted
by the rescheduling of waking activities to com-
plement earlier sleep times. Light therapy has the
advantage that patients adhere to progressively ear-
lier rising times over the course of treatment; there-
fore, they are more likely to comply with a fixed
schedule at its conclusion. It should be cautioned
that bright light treatment has been noted to precip-
itate manic episodes in bipolar patients [67].
Bright light therapy can be administered with
artificial light boxes or natural sunlight. Traditional
commercial devices use low UV fluorescent bulbs
and special reflectors to provide illuminances of
2500 to 10,000 lux at the eye when obliquely placed
about 1.5 to 3 ft away. Exposure times are typically
about 60 minutes, during which patients can have
breakfast, read, or watch TV. Patients should not
stare directly into the light. If obtaining much
brighter outdoor light (75,000 lux or more on clear
summer days) is feasible, good results can be had
with exposure times as short as 30 minutes. Cloudy
days are sufficiently bright to maintain progress;
patients should be counseled to obtain light expo-
sure on a consistent basis.
Recently, it has been discovered that the circadian
system responds specifically to light in the shorter,
deep blue wavelengths of the visible spectrum. It re-
lies on dedicated receptors in the retina with a peak
sensitivity at about 470 nanometers [68–70]. Some
manufacturers have devised light boxes emitting less
intense blue light, allowing shorter durations of
Nonpharmacologic Management of Insomnia 201
Box 8: Light therapy
Sleep schedule
Patients can shift their circadian phase by a half
hour to an hour each week with little difficulty
with the help of light exposure.
Each week of treatment, the wake-up time is
progressively shifted to an earlier time for pa-
tients with a delayed circadian sleep phase; in
contrast, the retiring time is gradually shifted
to a later time for patients with an advanced
circadian sleep phase.
For patients with a delayed sleep-wake pat-
tern, bright light exposure is administered in
the morning as close to the patient’s scheduled
arising time as possible. In contrast, for pa-
tients with an advanced sleep-wake pattern,
bright light should be administered in the early
evening, a couple of hours before their sched-
uled bedtime.
The source of bright light can be an artificial
light therapy device or natural outdoor sun-
light. Illuminance of approximately 2500 lux
or more at eye level usually is required to ob-
tain successful results. A 1- to 2-hour period
of treatment each day is optimal. Scheduling
constraints may necessitate a shorter exposure
duration or less frequent exposure that will re-
sult in slower treatment response.
Adapted from Yang CM, Spielman AJ, Glovinsky PB.
Nonpharmacologic strategies in the management of
insomnia. Psychiatr Clin North Am 2006;29(4):913;
with permission.
exposure. It is important to follow their instructions
carefully, because overexposure to blue light can be
especially hazardous to the eye.
Because light exposure before bedtime delays the
sleep phase, blocking exposure to the blue wave-
lengths of evening light with special sunglasses
can be helpful for patients contending with a de-
layed sleep phase, especially during summer
months when the photoperiod is prolonged [71].
Blue-blocking glasses (with lenses that render the
world in amber shades) may be worn from about
6 PM to sunset (except when driving at dusk for rea-
sons of safety). The glasses might also be worn by
persons who spend hours in front of computer
screens during the evening.
Evening bright light exposure has been used to
treat patients (more often the elderly) whose circa-
dian sleep phases are advanced [72,73]. These
patients typically complain of evening sleepiness
with inadvertent dozing, as well as early morning
awakenings. Light boxes are generally employed
because outdoor evening light is not available dur-
ing much of the year. Because evening light is del-
ivered further from the temperature nadir, it is
generally less potent in effecting phase shifts;
202 Glovinsky et al
therefore, exposure times toward the longer end of
the recommended range may be required. An oph-
thalmic consultation may be advisable in the el-
derly population before using bright light
treatment. The use of blue-blocking sunglasses in
the morning hours may also be helpful in promot-
ing a phase delay. As light treatment progresses,
bedtime schedules are shifted later to accommodate
the gradual shift in sleep phase.
Selecting and delivering cognitive
behavioral treatments
CBT-I are effective, producing moderate-to-large
effect sizes in controlled studies [1–4]. Currently,
most behavioral medicine practitioners tend to em-
ploy a combination of treatments. Multi-component
CBT and CBT in combination with hypnotics have
been proved to be superior to hypnotic use alone
in long-term follow-up [3,11,12].
The authors advocate an initial choice of treat-
ment based on the predominant insomnia com-
plaint and an analysis based on the 3P model,
with special attention given to perpetuating factors
[54]. If sleep has a broken pattern, with short pe-
riods of sleep and wakefulness in succession, or if
the time in bed is excessive relative to the amount
of sleep actually obtained, sleep restriction therapy
may be the initial treatment of choice. Stimulus
control instructions are especially useful when pro-
longed periods of wakefulness are seen, whether
in the form of difficulty falling asleep or staying
asleep, or when patients report getting a second
wind upon climbing into bed, evidence that the
bedroom has become associated with arousal
rather than sleep. Dysfunctional thinking that inter-
feres with sleep is best addressed directly with cog-
nitive treatments, whereas relaxation training is of
particular benefit when hyperarousal is deemed
a critical predisposing factor. Circadian rhythm dis-
orders are most effectively addressed with light ther-
apy, with chronotherapy reserved for severe cases of
delayed sleep phase disorder in which patients are
not falling asleep until near sunrise or later.
CBT-I can be delivered individually or in groups;
in either situation, treatment usually takes 4 to 8
weeks. Recently, an abbreviated two-session pro-
gram was reported to be effective in a primary care
setting [74]. Treatments for insomnia have also
been delivered through structured bibliotherapy
[75], self-help books [54,76,77], telephone consul-
tations [78,79], and the Internet [80]. Although
not all formats have been demonstrated to be
effective, the ongoing effort to disseminate CBT-I
reflects an acknowledged need to bring these well-
established treatments to the millions who suffer
from chronic insomnia, despite a dearth of experts
in the field. There are also initiatives underway to
train nurses, mental health counselors, and others
to deliver CBT-I in primary care settings, efforts
that have already shown promise [81].
References
[1] Lacks P, Morin CM. Recent advances in the
assessment and treatment of insomnia. J Consult
Clin Psychol 1992;60:586–94.
[2] Morin CM, Culbert JP, Schwartz SM. Nonphar-
macological interventions for insomnia:
a meta-analysis of treatment efficacy. Am J Psy-
chiatry 1994;151:1172–80.
[3] Morin CM, Hauri PJ, Espie CA, et al. Nonphar-
macologic treatment of chronic insomnia. Sleep
1999;22:1–23.
[4] Murtagh DRR, Greenwood KM. Identifying effec-
tive psychological treatment for insomnia:
a meta-analysis. J Consult Clin Psychol 1995;
63:79–89.
[5] Krakow B, Johnston L, Melendrez D, et al. An
open-label trial of evidence-based cognitive be-
havior therapy for nightmares and insomnia in
crime victims with PTSD. Am J Psychiatry 2001;
158:2043–7.
[6] Krakow B, Hollifield M, Johnston L, et al. Imag-
ery rehearsal therapy for chronic nightmares in
sexual assault survivors with posttraumatic stress
disorder. JAMA 2001;286:537–45.
[7] Lichstein KL, Wilson NM, Johnson CT. Psycho-
logical treatment of secondary insomnia. Psychol
Aging 2000;15:232–40.
[8] Perlis ML, Sharpe MC, Smith MT, et al. Behav-
ioral treatment of insomnia: treatment outcome
and the relevance of medical and psychiatric
morbidity. J Behav Med 2001;24:281–96.
[9] Rybarczyk B, Lopez M, Benson R, et al. Efficacy of
two behavioral treatment programs for comor-
bid geriatric insomnia. Psychol Aging 2002;17:
288–98.
[10] Rybarczyk B, Stepanski E, Fogg L, et al. A pla-
cebo-controlled test of cognitive-behavioral ther-
apy for comorbid insomnia in older adults. J
Consult Clin Psychol 2005;73:1164–74.
[11] Smith MT, Perlis ML, Park A, et al. Comparative
meta-analysis of pharmacotherapy and behavior
therapy for persistent insomnia. Am J Psychiatry
2002;159:5–11.
[12] Jacobs GD, Pace-Schott EF, Stickgold R, et al.
Cognitive behavior therapy and pharmacother-
apy for insomnia. Arch Intern Med 2004;164:
1888–96.
[13] Morin CM, Colecchi C, Stone J, et al. Behavioral
and pharmacological therapies for late-life in-
somnia: a randomized controlled trial. JAMA
1999;281:991–9.
[14] National Institutes of Health. National Institutes
of Health State of the Science conference state-
ment on manifestations and management of
chronic insomnia in adults, June 13–15, 2005.
Sleep 2005;28:1049–57.

[15] Spielman AJ, Glovinsky P. The diagnostic inter-
view and differential diagnosis for complaints
of insomnia. In: Pressman MR, Orr WC, editors.
Understanding sleep: the evaluation and treat-
ment of sleep disorders. Washington, DC: Amer-
ican Psychological Association; 1997. p. 125–60.
[16] Sateia MJ, Doghramji K, Hauri PJ, et al. Evaluation
of chronic insomnia: an American Academy of
Sleep Medicine review. Sleep 2000;23:243–308.
[17] Spielman AJ, Yang CM, Glovinsky PB. Assess-
ment techniques for insomnia. In: Kryger M,
Roth T, Dement W, editors. Principles and prac-
tices of sleep medicine. 4th edition. Toronto: WB
Saunders; 2005. p. 1403–16.
[18] Borbely AA. A two process model of sleep regula-
tion. Hum Neurobiol 1982;1:195–204.
[19] Mignot E, Taheri S, Nishino S. Sleeping with the
hypothalamus: emerging therapeutic targets for
sleep disorders. Nat Neurosci 2002;5:1071–5.
[20] Saper CB, Scammell TE, Lu J. Hypothalamic reg-
ulation of sleep and circadian rhythms. Nature
2005;437:1257–63.
[21] Wager-Smith K, Kay SA. Circadian rhythm ge-
nerics: from flies to mice to humans. Nat Genet
2000;26:23–7.
[22] Klei L, Reitz P, Miller M, et al. Heritability of
morningness-eveningness and self-report sleep
measures in a family-based sample of 521 Hut-
terites. Chronobiol Int 2005;22:1041–54.
[23] Czeisler CA, Duffy JF, Shanabar TL, et al. Stabil-
ity, precision, and near-24-hour period of the
human circadian pacemaker. Science 1999;284:
2177–81.
[24] Dijk DJ, Czeisler CA. Contribution of the circa-
dian pacemaker and the sleep homeostat to
sleep propensity, sleep structure, electroencepha-
lographic slow waves, and sleep spindle activity
in humans. J Neurosci 1995;15:3526–38.
[25] Bonnet MH, Arand DL. Hyperarousal and in-
somnia. Sleep Med Rev 1997;2:97–108.
[26] Perlis ML, Giles DE, Mendelson WB, et al. Psy-
chophysiological insomnia: the behavioral
model and a neurocognitive perspective. J Sleep
Res 1997;6:179–88.
[27] Vgontzas AN, Tsigos C, Bixler EO, et al. Chronic
insomnia and activity of the stress system: a pre-
liminary study. J Psychosom Res 1998;45:21–31.
[28] Bonnet MH, Arand DL. 24-Hour metabolic rate
in insomniacs and matched normal sleepers.
Sleep 1995;18:581–8.
[29] Freedman RR. EEG power spectra in sleep-onset
insomnia. Electroencephalogr Clin Neurophy-
siol 1986;63:408–13.
[30] Merica H, Blois R, Gaillard JM. Spectral charac-
teristics of sleep EEG in chronic insomnia. Eur
J Neurosci 1998;10:1826–34.
[31] Perlis ML, Smith MT, Andrews PJ, et al. Beta/
gamma EEG activity in patients with primary
and secondary insomnia and good sleeper con-
trols. Sleep 2001;24:110–7.
[32] Yang CM, Lo HS. ERP evidence of enhanced
excitatory and reduced inhibitory processes of
Nonpharmacologic Management of Insomnia 203
auditory stimuli during sleep in patients with
primary insomnia. Sleep 2007;30:585–92.
[33] Nofzinger EA, Nissen C, Germain A, et al.
Regional cerebral metabolic correlates of WASO
during NREM sleep in insomnia. J Clin Sleep
Med 2006;2:316–22.
[34] Harvey AG. Pre-sleep cognitive activity: a com-
parison of sleep-onset insomniacs and good
sleepers. Br J Clin Psychol 2000;39:275–86.
[35] Nelson J, Harvey AG. An exploration of pre-sleep
cognitive activity in insomnia: imagery and ver-
bal thought. Br J Clin Psychol 2003;42:271–88.
[36] Fichten CS, Creti L, Amsel R, et al. Poor sleepers
who do not complain of insomnia: myths and
realities about psychological and lifestyle charac-
teristics of older good and poor sleepers. J Behav
Med 1995;18:189–223.
[37] Edinger JD, Fins AI, Glenn DM, et al. Insomnia
and the eye of the beholder: are there clinical
markers of objective sleep disturbances among
adults with and without insomnia complaints?
J Consult Clin Psychol 2000;68:586–93.
[38] Wicklow A, Espie CA. Intrusive thoughts and
their relationship to actigraphic measurement
of sleep: towards a cognitive model of insomnia.
Behav Res Ther 2000;38:679–93.
[39] Morin CM, Stone J, Trinkle D, et al. Dysfunc-
tional beliefs and attitudes about sleep among
older adults with and without insomnia com-
plaints. Psychol Aging 1993;8:463–7.
[40] Harvey L, Inglis SJ, Espie CA. Insomniac’s
reported use of CBT components and relation-
ship to long-term clinical outcome. Behav Res
Ther 2002;40:75–83.
[41] Morin CM, Blais F, Savard J. Are changes in
beliefs and attitudes about sleep related to sleep
improvements in the treatment of insomnia?
Behav Res Ther 2002;40:741–52.
[42] Harvey AG. Identifying safety behaviors in
insomnia. J Nerv Ment Dis 2002;190:16–21.
[43] Drake C, Richardson G, Roehrs T, et al. Vulnera-
bility to stress-related sleep disturbance and
hyperarousal. Sleep 2004;27:285–91.
[44] Kales A, Caldwell AB, Preston TA, et al. Personal-
ity patterns in insomnia: theoretical implica-
tions. Arch Gen Psychiatry 1976;33:1128–34.
[45] Dorsey CM, Bootzin RR. Subjective and psycho-
physiologic insomnia: an examination of sleep
tendency and personality. Biol Psychiatry 1997;
15(41):209–16.
[46] Kales A, Caldwell AB, Soldatos CR, et al. Biopsy-
chobehavioral correlates of insomnia. II. Pattern
specificity and consistency with the Minnesota
Multiphasic Personality Inventory. Psychosom
Med 1983;45:341–56.
[47] Vincent NK, Walker JR. Perfectionism and chronic
insomnia. J Psychosom Res 2000;49:349–54.
[48] Bootzin RR. Stimulus control treatment for in-
somnia. Proceedings of the American Psycholog-
ical Association 1972;7:395–6.
[49] Spielman AJ. Assessment of insomnia. Clin Psy-
chol Rev 1986;6:11–25.
204 Glovinsky et al
[50] Spielman AJ, Anderson MA. The clinical evalua-
tion as a guide to understanding the nature of in-
somnia: the CCNY semi-structured interview for
insomnia. In: Chokroverty S, editor. Sleep disor-
ders medicine. 2nd edition. Boston: Butterworth
and Heinemann; 1999. p. 385–426.
[51] Morin CM, Gaulier B, Barry T, et al. Patients’
acceptance of psychological and pharmacologi-
cal therapies for insomnia. Sleep 1992;15:302–5.
[52] Spielman AJ, Saskin P, Thorpy MJ. Treatment of
chronic insomnia by restriction of time in bed.
Sleep 1987;10:45–56.
[53] Spielman AJ, Caruso LS, Glovinsky PB. A behav-
ioral perspective on insomnia treatment. Psy-
chiatr Clin North Am 1987;10(4):541–53.
[54] Glovinsky P, Spielman A. The insomnia answer:
breakthrough solutions for getting to sleep,
staying asleep, broken sleep. New York: Perigee;
2006.
[55] Borkovec TD, Grayson JB, O’Brien GT, et al.
Relaxation treatment of pseudoinsomnia and id-
iopathic insomnia: an electroencephalographic
evaluation. J Appl Behav Anal 1979;12:37–54.
[56] Nicassio P, Bootzin R. A comparison of progres-
sive relaxation and autogenic training as treatment
for insomnia. J Abnorm Psychol 1974;83:253–60.
[57] Woolfolk RL, Carr-Kaffashan L, McNulty TF.
Meditation training as a treatment for insomnia.
Behav Ther 1976;7:359–65.
[58] Haynes SN, Sides H, Lockwood G. Relaxation
instructions and frontalis electromyographic
feedback intervention with sleep-onset insom-
nia. Behav Ther 1977;8:644–52.
[59] Hauri PJ. Treating psychophysiologic insomnia
with biofeedback. Arch Gen Psychiatry 1981;
38:752–8.
[60] Morin CM. Insomnia: psychological assessment
and management. New York: Guilford; 1993.
[61] Weitzman ED, Czeisler CA, Coleman RM, et al.
Delayed sleep-phase syndrome: a chronobiologi-
cal disorder with sleep-onset insomnia. Arch Gen
Psychiatry 1981;38:737–46.
[62] Thorpy MJ, Korman E, Spielman AJ, et al.
Delayed sleep-phase syndrome in adolescents.
J Adolesc Health Care 1988;9:22–7.
[63] Czeisler CA, Richardson GS, Coleman RM, et al.
Chronotherapy: resetting the circadian clocks of
patients with delayed sleep phase insomnia.
Sleep 1981;4:1–21.
[64] Czeisler CA, Allan JS, Strogatz SH, et al. Bright
light resets the human circadian pacemaker in-
dependent of the timing of the sleep-wake cycle.
Science 1986;233:667–71.
[65] Minors DS, Waterhouse JM, Wirz-Justice A. A hu-
man phase-response curve to light. Neurosci Lett
1991;133:354–61.
[66] Rosenthal NE, Joseph-Vanderpool JR,
Levendosky AA, et al. Phase-shifting effects of
bright morning light as treatment for delayed
sleep phase syndrome. Sleep 1990;13:354–61.
[67] Pande AC. Light-induced hypomania. Am J Psy-
chiatry 1985;142:1126.
[68] Thapan K, Arendt J, Skene DJ. An action spec-
trum for melatonin suppression: evidence for
a novel non-rod, non-cone photoreceptor sys-
tem in humans. J Physiol 2001;535:261–7.
[69] Brainard GC, Hanifin JP, Greeson JM, et al. Action
spectrum for melatonin regulation in humans:
evidence for a novel circadian photoreceptor.
J Neurosci 2001;21:6405–12.
[70] Lockley SW, Brainard GC, Czeisler CA. High sen-
sitivity of the human circadian melatonin
rhythm to resetting by short wavelength light.
J Clin Endocrinol Metab 2003;88:4502–5.
[71] Sasseville A, Paquet N, Sevigny J, et al. Blue
blocker glasses impede the capacity of bright
light to suppress melatonin production. J Pineal
Res 2006;41:73–8.
[72] Campbell SS, Dawson D, Anderson MW. Allevi-
ation of sleep maintenance insomnia with timed
exposure to bright light. J Am Geriatr Soc 1993;
41:829–36.
[73] Singer CM, Lewy AJ. Case report: use of the dim
light melatonin onset in the treatment of ASPS
with bright light. Sleep Research 1989;18:445.
[74] Edinger JD, Sampson WS. A primary care
‘‘friendly’’ cognitive behavioral insomnia ther-
apy. Sleep 2003;26:177–82.
[75] Mimeault V, Morin CM. Self-help treatment for
insomnia: bibliotherapy with and without pro-
fessional guidance. J Consult Clin Psychol
1999;67:511–9.
[76] Hauri P, Linde S. No more sleepless nights. New
York: John Wiley & Sons; 1990.
[77] Jacobs G, Benson H. Say good night to insom-
nia: the six-week, drug-free program developed
at Harvard Medical School. New York: Henry
Holt & Company; 1998.
[78] Bastien CH, Morin CM, Ouellet MC, et al. Cog-
nitive-behavioral therapy for insomnia: compar-
ison of individual therapy, group therapy, and
telephone consultation. J Consult Clin Psychol
2004;72:653–9.
[79] Verbeek I, Declerck G, Knuistingh NA, et al.
Sleep information by telephone: callers indicate
positive effects on sleep problems. Sleep and
Hypnosis 2002;4:47–51.
[80] Strom L, Pettersson R, Andersson G. Internet-
based treatment for insomnia: a controlled
evaluation. J Consult Clin Psychol 2004;72:
113–20.
[81] Espie CA, Inglis SJ, Tessier S, et al. The clinical ef-
fectiveness of cognitive behaviour therapy for
chronic insomnia: implementation and evalua-
tion of a sleep clinic in general medical practice.
Behav Res Ther 2001;39:45–60.

Sleepiness and Fatigue in Patients
with Psychiatric Disorders
a,
Chad C. Hagen, MD *, Jed E. Black,
- Sleepiness and fatigue
- Psychiatric diagnosis
- Cognitive disorders
Attention Deficit-Hyperactivity Disorder
(ADHD)
Dementia
- Mood disorders
Fatigue and daytime sleepiness are common
complaints in the patient with a psychiatric diagno-
sis. Although both psychiatric illness and psycho-
tropic medications may cause sleepiness and
fatigue, when complaints of sleepiness or fatigue
arise in these patients other potential etiologies
must be considered. Patients with a psychiatric dis-
order and comorbid sleepiness or fatigue warrant
appropriate evaluation to determine whether a pri-
mary sleep disorder is accounting for, contributing
to, or complicating the treatment of a presumed
psychiatric disorder. A review of cognitive and
mood effects secondary to sleep fragmentation
and deprivation are beyond the scope of this article,
but it is well-demonstrated that fatigue and sleepi-
ness arise from processes that compromise the effi-
ciency of sleep or restrict a person from an adequate
sleep duration. These include conditions such as
periodic limb movement disorder, sleep-related
breathing disorders, chronobiological or circadian
rhythm sleep disorders, insomnia, and neurologic
or medical conditions adversely effecting sleep
a
Sleep Disorders Program, CR-139, Department of Psychiatry, Oregon Health and Science University, 3181 SW
Sam Jackson Park Rd., Portland, OR 97239-3098, USA
b
Stanford Sleep Disorders Center, Department of Psychiatry and Behavioral Sciences, 401 Quarry Rd., #3301,
Stanford, CA 94305, USA
* Corresponding author.
E-mail address: hagench@ohsu.edu (C.C. Hagen).
1556-407X/08/$ – see front matter. ª Chad C. Hagen and Jed E. Black.
sleep.theclinics.com
205
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 205–215
b
MD
Depression
Bipolar disorder
- Anxiety disorders
Post traumatic stress disorder
Generalized anxiety disorder
- Emerging circadian findings
- Psychiatric medications
- References
quantity or quality. If sleepiness, decreased vigi-
lance, or fatigue is secondary to a sleep disorder,
then they should not be used to support a psychiat-
ric diagnosis such as depression, attention-deficit
hyperactivity disorder, or generalized anxiety disor-
der. Indeed such sleep disorders commonly exacer-
bate comorbid psychiatric conditions and often
render primary psychiatric syndromes more treat-
ment refractory.
Insufficient sleep is the most common explana-
tion for sleepiness and fatigue in the general popu-
lation. Sleep-disordered breathing accounts for
most cases of sleepiness and fatigue that present
to sleep disorders clinics. Other causes of sleepiness
such as circadian misalignment, insufficient or
disrupted sleep, head trauma, narcolepsy, and idio-
pathic hypersomnia should be considered in all
patients complaining of excessive daytime sleepi-
ness (EDS) or fatigue. Lack of an understanding of
these conditions not only leads to misdiagnosis or
lack of recognition, but often also results in gross
mismanagement. Examples include the use of
doi:10.1016/j.jsmc.2008.01.013
206 Hagen & Black
antipsychotic medications in a patient with narco-
lepsy-related hypnagogic hallucinations or antide-
pressant therapy for patients with a complaint of
hypersomnolence related to a primary sleep disor-
der that is misdiagnosed as depression. This article
summarizes psychiatric disorders and treatments
that are associated with EDS or fatigue, and the
role of the sleep physician in addressing these com-
plaints in the patient with comorbid sleep and psy-
chiatric disorders.
Sleepiness and fatigue
A patient with EDS often struggles to maintain
wakefulness in monotonous situations, whereas
a patient with the complaint of fatigue may experi-
ence listlessness and lethargy with or without EDS
and the tendency to fall asleep. There is consider-
able overlap between these two presentations, and
both complaints may indicate a significant prob-
lem. EDS has been easier to define as a specific
physiologic state by measuring a patient’s capacity
to fall asleep or capacity to remain awake in seden-
tary situations throughout the day. Fatigue, however
has remained more difficult to isolate and quantify
and may represent numerous chronic or acute phys-
iologic or psychologic processes. The American
Academy of Sleep Medicine defines EDS in the
International Classification of Sleep Disorders
(ICSD) as ‘‘a complaint of difficulty in maintaining
desired wakefulness or a complaint of excessive
amount of sleep’’ [1]. It also notes that while sleep-
iness may be more objectively defined, it is still
essentially a subjective report of difficulty maintain-
ing the alert awake state, usually accompanied by an
increased propensity for rapid entrance into sleep
when the individual is sedentary [1]. The second
edition of the ICSD describes strategies for quanti-
fying the capacity to stay awake or the tendency to
fall asleep using the Maintenance of Wakefulness
Test (MWT) or the Multiple Sleep Latency Test
(MSLT), respectively. The second edition also de-
fines fatigue as ‘‘Lack of energy, listlessness, or wea-
riness as from labor. Not necessarily associated with
an increased tendency to fall asleep’’ [2]. While a va-
riety of scales have been developed for assessing fa-
tigue and sleepiness, a recent study by Bailes and
colleagues [3] indicates that six items from the
Epworth sleepiness scale may be more specific for
sleepiness, while three items selected from two
fatigue scales (Chalder fatigue scale [4] and Fatigue
severity scale [5]) were more specific for weakness
and tiredness arising from exertion. Their analysis
indicated that the three most specific items for
this type of fatigue are the following: ‘‘Exercise
brings on my fatigue,’’ ‘‘I start things without diffi-
culty but get weak as I go on,’’ and ‘‘I lack energy.’’
The Fatigue severity scale in its entirety has been val-
idated and found to correlate well with ‘‘disabling
fatigue’’ independent from depressive symptom-
atology in patients with disorders such as lupus,
multiple sclerosis, and insomnia [3,5,6].
Psychiatric diagnosis
The Diagnostic and Statistical Manual (DSM-IV-TR)
[7], much like ICSD-2, outlines the diagnostic crite-
ria, epidemiologic data, and associated features for
each disorder summarized. Within the diagnostic
criteria for each psychiatric disorder, the DSM stip-
ulates that the symptoms in question are not better
explained by a medical condition (eg, a sleep disor-
der). Although sleep clinicians may feel unqualified
or not obligated to assess psychiatric disorders in
the sleep clinic, some familiarity with the diagnostic
criteria and how to use them is essential. Awareness
of psychiatric disorders and their means of diagno-
sis is so critical that 10% of the ICSD-2 is dedicated
to a review of the information found within DSM-
IV-TR (Appendix B, ICSD-2) [2]. This knowledge
facilitates the recognition of possible erroneous
psychiatric diagnoses in a patient suffering primar-
ily from a sleep disorder, as well as for the diagnosis
of comorbid sleep disorders in those with primary
psychiatric conditions. These distinctions are more
difficult after the burden of social consequence
and adjustment to sleepiness or fatigue takes its
toll. For example, a sleepy patient with undiag-
nosed sleep apnea may feel ‘‘depressed’’ because
of difficulty maintaining his or her performance at
work. When liberated from sleepiness this patient
may perform better and subsequently feel relatively
well; hence, a diagnosis of depression would be in-
appropriate. We briefly outline the diagnostic crite-
ria for these disorders at the beginning of each
section, but recommend Appendix B of the ICSD-
2, or the DSM-IV-TR, for sleep physicians wishing
to review psychiatric diagnostic criteria in more
detail.
Cognitive disorders
Attention Deficit-Hyperactivity
Disorder (ADHD)
DSM-IV-TR characterizes attention deficit disorder/
inattentive type as a problem with attention and
concentration that is accompanied by six symptoms
of inattentiveness from a list of nine potential
symptoms. These symptoms must arise in a variety
of circumstances (eg, work, school, home), begin
before age 7, and persist for greater than 6 months.
An additional six symptoms of hyperactivity or
impulsivity from a list of nine symptoms are re-
quired for the diagnosis of attention deficit

disorder/hyperactive type. Given evolving diagnos-
tic criteria and changes in identification of the dis-
order in clinical practice, estimates of prevalence
have varied, but prevalence in schoolchildren is
roughly estimated to be approximately 3% to 8%.
Characterized as a cognitive disorder and treated
for decades with stimulant medications [8], ADHD
is essentially a disorder of cognitive function that
has significant overlap with the deficits in attention,
concentration, vigilance, and impulse regulation
seen in conditions of sleep fragmentation and/or
sleep deprivation. Until recent versions of the
DSM, presence of a sleep disturbance was included
in the diagnostic criteria. The DSM-IV-TR has
excluded sleep disturbance from the diagnostic cri-
teria for ADHD and instead qualifies the diagnosis
stating that the symptoms of poor attention, con-
centration, or hyperactivity cannot be attributed
to a general medical condition. Therefore, sleep
deprivation or sleep fragmentation from a sleep dis-
order must be reasonably excluded before a patient
receiving a diagnosis of, and treatment for ADHD.
This process can be complicated in the patient
who presents for a sleep evaluation while receiving
treatment for ADHD. Sleepiness may be missed in
these children if masked by stimulant medication,
which can obscure the common symptoms of
childhood sleepiness such as irritability; reduced
frustration threshold; impulsivity; and decreased
vigilance, alertness, and concentration.
Cohen-Zion and Anconi-Israel distilled
47 ADHD-research studies, published between
1980 and 2004, including 13 involving stimulant
intervention and another 34 using medication
free subjects [9]. Their review clarifies that many
studies to date are limited due to reliance on subjec-
tive reports of sleep; none-the-less, these reports
documented a high rate of sleep disturbance, in-
creased nocturnal activity, decreased REM sleep,
and excessive daytime somnolence in stimulant-
free children with ADHD when compared with
controls. Another recent review by Cortese and col-
leagues [10] summarizes that apnea-hypopnea
index (AHI, a measure of sleep-related breathing
disturbance severity), EDS, and movements during
sleep were all higher in children with ADHD com-
pared with controls. Much research evidences that
children, adolescents, and adults carrying a diagno-
sis of ADHD often have an associated sleep disorder
that may better account for their symptoms of
inattention, poor concentration, or hyperactivity
[11–14]. Beyond the high rates of association for
these syndromes, recent work has demonstrated
successful management of ADHD with surgical or
positive air-pressure treatments [15–20] in some
patients with comorbid sleep-related breathing dis-
orders. Stimulant use carries some risk including
Sleepiness and Fatigue 207
dose-dependent increases in insomnia, systolic
and diastolic blood pressure, and heart rate [21].
While a recent actigraphy study in patients without
sleep disordered breathing (SDB), suggested meth-
ylphenidate reduces movement during nocturnal
sleep [22], these other side effects may be particu-
larly disadvantageous in a patient with an unrecog-
nized sleep-related breathing disorder.
Other groups have demonstrated high associa-
tions between periodic limb movements, restless
leg symptoms, and ADHD symptoms in children
[12,23] and suggest that the strength of these asso-
ciations warrants increased investigation. Oosterloo
and colleagues [24] compared symptoms of sleepi-
ness and deficiencies in attention and concentra-
tion between patients with narcolepsy and those
with ADHD. They found a high incidence of sleep-
iness in their ADHD patients, with 38% having an
Epworth greater than or equal to 12. Nineteen per-
cent of their narcoleptic patients met criteria for
ADHD based on self-report scales. From this, they
concluded that given the high degree of overlap
between these two populations, the validity of
self-report scales in these populations is unclear
and there should be attentiveness to the possibility
of misdiagnosis in sleep-disorder patients with
psychiatric disorders.
In view of the mounting evidence demonstrating
ADHD-like symptoms are common among individ-
uals with sleep disorders, patients with the diagno-
sis of ADHD, or in whom ADHD is suspected,
should undergo a sleep evaluation that includes,
when indicated, full polysomnography.
While there are reports of circadian misalignment
in children with ADHD, published studies thus far
indicate that interventions such as sleep hygiene
and melatonin use timed for phase-advancing
ADHD children with delayed sleep phase or initia-
tion insomnia is beneficial for decreasing sleep
latency, but does not appear to have a significant
impact on overall ADHD symptomatology
[25,26]. However, adults with ADHD treated with
phase-advancing light therapy during winter
months had improvements in ADHD based on
improved symptom scales and neuropsychological
testing [27].
Given the similarities between ADHD symptoms
and sleepiness and the associations between ADHD
and sleep disorders in the literature, it is clearly
important to detect and treat any comorbid sleep
disorder [28].
Dementia
Dementia arises from a variety of conditions includ-
ing Alzheimer’s, Huntington’s, and Parkinson’s
diseases, as well as from conditions such as cerebro-
vascular disease, human immunodeficiency virus
208 Hagen & Black
infection, and head injury. Dementia is character-
ized by loss of memory and is often associated
with a variety of other cognitive impairments. Prev-
alence increases with age, with an estimated 1%
prevalence from age 60 to 64, but increases up to
30% to 50% in those older than 85.
Patients with Alzheimer’s disease often struggle
with excessive daytime sleepiness and circadian
abnormalities. In contrast to the normal decline
in sleep period for aging adults, dementia patients
have been shown to have a longer sleep period,
which is positively correlated with dementia sever-
ity [29] Objective measures of daytime sleepiness
are also higher in patients with Alzheimer’s disease.
Furthermore, the magnitude of sleepiness in this
population, measured by MSLT, appears to be asso-
ciated with the level of cognitive impairment on
a variety of neuropsychologic measures [30]. Sun-
downing is a common phenomenon in demented
patients, referring to increased disorientation and
agitation that tends to occur in evening hours.
There is evidence that sundowning is related to an
inappropriate decrease in vigilance and alertness
during the waking hours [31]. This may be in part
related to a genetic predisposition for some people
to phase advance into a circadian phase advanced
or ‘‘morningness’’ pattern with age [32]. Reports
of treatment with morning light, and evening mel-
atonin have shown improvement in the consistency
of nighttime sleep and consolidation of the sleep-
wake cycle [29,33–36]. Two studies have reported
some degree of behavioral improvement with
such treatment as well [31,37]. Overall, demented
patients may show inappropriate sleepiness or
inappropriately timed sleepiness, both of which
will often benefit from good sleep hygiene, daytime
light and activity, and maintaining a quiet and
appropriately dark environment conducive to
sleeping at night [34].
Mood disorders
Depression
Major depressive disorder can be a devastating
illness causing significant social or occupational
disruption. Death by suicide occurs at an alarming
rate of up to 15% in patients with severe major
depression [38]. The lifetime risk of depression
varies from about 5% to 25% of the population,
depending on the study, but occurs more com-
monly in women than men with an average age
of onset in the mid 20s [38]. The disorder is charac-
terized by one or more incidents of a major depres-
sive episode. Diagnostically, these episodes must be
longer than 2 weeks in duration, represent a change
from baseline, and include at least five symptoms,
at least one of which must be either depressed
mood or anhedonia (loss of interest or pleasure
in activities). The remaining symptoms may
include unintentional weight loss, weight gain,
insomnia, hypersomnia, changes in activity level,
fatigue or loss of energy, feelings of worthlessness
or excessive guilt, poor concentration or indecisive-
ness, or recurrent thoughts of death or suicide. Like
nearly all psychiatric diagnoses, the critical stipula-
tion within the DSM-IV-TR is: ‘‘Do not include
symptoms that are clearly due to a general medical
condition’’ [38]. It is easy for sleep clinicians to rec-
ognize patients with five of these symptoms. How-
ever, it is difficult for us to tell whether the
symptoms are due to a general medical condition
or sleep disorder, such as restless legs syndrome or
obstructive sleep apnea (OSA) (Boxes 1–3). Pa-
tients and their psychotropic prescribers should be
informed about which of their previously attributed
depressive symptoms, such as hypersomnia, fa-
tigue, sleepiness, performance changes, changes in
activity level, reductions in interest or pleasure in
activities, sleep alterations, poor concentration,
and weight gain might have a sleep disorder etiol-
ogy. Indeed all of these symptoms are common
complaints in patients with OSA.
Some authors have suggested that, fatigue, un-
controlled sleep deprivation, and the inability to
initiate or maintain sleep may carry increased risk
for suicide. Polysomnographic findings consistent
with both depression and sleepiness, such as
decreased sleep latency, shortened rapid-eye move-
ment (REM) latency, and increased REM percent-
ages are seen in both sleep-deprived patients and
suicidal patients [39,40]. Several studies have indi-
cated that insomnia, but not sleepiness or fatigue,
is correlated with thoughts of suicide, plans for sui-
cide, or attempted suicide [41–43]. Patients with de-
pression-related symptoms that present initially to
the sleep disorders specialist should be screened
in the sleep disorders clinic for depression and sui-
cide and referred to appropriate mental health
services.
Box 1: DSM-IV diagnostic criteria for
depression
Symptoms associated with sleep disturbance
Depressed mood
Weight change
Insomnia
Decreased activity
Fatigue/poor energy
Poor concentration
Symptoms specific to depression
Anhedonia
Worthlessness
Suicidal thoughts

Box 2: DSM-IV Diagnostic criteria for
dysthymia
Symptoms specific Symptoms associated
to dysthymia with sleep loss
Poor appetite Overeating
Low self esteem Insomnia
Feelings of Hypersomnia
hopelessness Low energy or fatigue
Decreased energy
Poor concentration
Research has been conducted in an effort to dis-
criminate sources of fatigue and sleepiness in
depressed patients with sleep disorders such as
obstructive sleep apnea. Given the circular nature
of this line of inquiry, the research is difficult and
inherently confounded. Some findings from cross-
sectional data suggest that fatigue may be more sub-
stantively explained by self-reported depression
scales than by severity of obstructive sleep apnea
[44]. The authors of this study also make the critical
point that management of depression in patients
with comorbid OSA may yield better improvement
in their fatigue than pursuing OSA treatment solely.
Other authors have placed more emphasis on the
need for optimization of treatment for OSA, and
the necessity of adequately treating sleepiness and
fatigue for successful treatment of the patient with
comorbid depression [40,42,45–47]. Depression
symptoms often improve or are resolved in patients
successfully treated for obstructive sleep apnea [48].
Optimal treatment of the patient with depression
who is newly diagnosed with a sleep disorder in-
cludes treatment continuation for depression while
the sleep disorder is formally diagnosed and
treated. Mood and fatigue have been shown to cor-
relate with sleep fragmentation associated with
sleep-related breathing disorders more than with
the frequency of oxygen desaturations [6,42].
Therefore, optimizing the treatment of fatigue and
sleepiness in these patients requires attention to
Box 3: DSM-IV Diagnostic criteria for atypical
depression
Symptoms specific Symptoms
to atypical associated with
depression sleep loss
Long standing Mood reactivity
interpersonal Appetite and
rejection weight increase
sensitivity Decreased energy
‘‘leaden paralysis’’
(fatigue)
Sleepiness and Fatigue 209
sensitive measures of airflow limitation, such as
snoring, hypopneas, and flow signal alteration
and not simply dependence on a slowly responding
and insensitive indicator of disturbance such as
oxygen desaturation.
If depression treatment is enhanced by treatment
for a sleep disorder, a cautious re-consideration of
the psychiatric diagnosis by a psychiatrist (or appro-
priate professional) able to provide follow-up and
observation of the patient over time should take
place. Observation over time is critical given the re-
lapsing-remitting cycle of mood disorders. In some
cases, it may be appropriate for the psychiatrist to
reconsider the dosage or need for medications
depending on the patient’s history of response to
antidepressants and sleep-related treatments. Al-
though not Food and Drug Administration (FDA)
indicated, several studies support the use of moda-
finil as an augmentation strategy in depressed
patients presenting with a significant component
of fatigue or sleepiness.
Dysthymia is a condition similar to depression
that presents with a more mild but chronic symp-
tomatology characterized by depression occurring
more days than not and persisting for more than 2
years. In some cases, dysthymia may occur with ma-
jor depression. Dysthymia is typically associated
with complaints such as fatigue and poor energy
and can be difficult to discriminate from chronic fa-
tigue syndrome [49] or the fatigue of a sleep disor-
der. Like depression, dysthymia patients benefit
from diagnosis and appropriate treatment for any
comorbid sleep disorder compounding their fatigue.
Seasonal affective disorder (SAD) is not a term
found in the DSM-IV-TR, rather the term ‘‘seasonal
pattern’’ is used as a qualifier to describe a patient
with a mood disorder that tends to occur with a sea-
sonal association more often than not. The qualifier
can be applied to major depressive disorder or
Bipolar types I or II. SAD of the depressed type is
considered a psychiatric disorder, but interestingly
appears to arise from changes in sleepiness, fatigue,
and cognition secondary to mismatch of life and
social obligations with circadian phase, much like
other circadian disorders such as jet lag, shift
work, or delayed sleep phase syndrome. SAD is
arguably the only psychiatric disorder with a biolog-
ical marker correlating with depression severity and
its response to treatment [50]. Degree of circadian
phase misalignment has been correlated with sever-
ity of depression symptoms in this population
throughout illness, treatment, and posttreatment
[50]. Impressive associations with variants of bio-
logical clock genes (Per2, Arntl, Npas2) have been
demonstrated in humans at higher risk for seasonal
depression [51]. While antidepressant treatment
and cognitive behavioral therapies may be
210 Hagen & Black
beneficial, treatment with appropriately timed light
(usually early morning light for patients with a de-
layed phase presentation) is currently the treatment
of choice [50,52–54]. Appropriately timed low-
dose melatonin (eg, 0.5 mg in the early evening
for patients with a delayed phase presentation)
has also demonstrated a robust effect [50].
Bipolar disorder
Bipolar disorder may be one of the most fascinating
human models of disordered sleep homeostasis,
but research has lagged given the complexities of
studying this population during active manic epi-
sodes. Bipolar disorder is divided into two major
categories including Bipolar I and Bipolar II. The
essential distinguishing feature between the two is
that Bipolar I patients have experienced at least
one manic episode (manic symptoms lasting
7 days or longer). Bipolar II patients have a similar
presentation that tends to be slightly less severe
with manic symptoms lasting less than 4 days per
episode. Diagnostically, three manic symptoms
are required in addition to an ‘‘abnormally and per-
sistently elevated, expansive, or irritable mood’’
lasting the appropriate duration. The other three
symptoms may include increased self-esteem or
grandiosity that is nondelusional, decreased need
for sleep, increased rate of speech or thoughts,
and increased activity in risky or pleasurable activi-
ties with disregard for the consequences or associ-
ated risks. Regarding the nature of the sleep
disturbance, ‘‘almost invariably, there is a decreased
need for sleep. The person usually awakens several
hours earlier than usual, feeling full of energy.
When the sleep disturbance is severe, the person
may go for days without sleep and yet not feel tired.’’
This fascinating clinical presentation has been ne-
glected from a sleep and fatigue research standpoint
and more work is certainly needed in this area given
the limited efficacy of available treatment options,
high morbidity of the bipolar population, and the
opportunity for furthering our understanding of
sleep-wake homeostasis and the generation and
experience of sleepiness and fatigue.
While it is rare to find an actively manic patient
presenting to the sleep clinic, it is common to see
bipolar spectrum patients on maintenance medica-
tions for their mood disorder. Essentially all of the
common mood-stabilizing agents have a high inci-
dence of fatigue or sedation as a side effect. These
medications include typical and ‘‘atypical’’ antipsy-
chotic medications, anticonvulsants, and lithium.
An effort to clarify whether the fatigue or sleepiness
changed with initiation or dosage adjustment in
one of these medications is warranted.
Anxiety disorders
Post traumatic stress disorder
Posttraumatic stress disorder (PTSD) affects 8% of
the general population and is characterized by the
development of a variety of symptoms occurring
after a traumatic or life-threatening event [38]. It
can arise after a broad range of emotionally laden
exposures, but occurs more commonly in a third
to half of patients that have been exposed to rape,
combat, war-related captivity, or genocide [38].
Traumatic exposures are disturbingly common in
children, but only 0.5% of children will manifest
the number and severity of symptoms required for
the diagnosis of PTSD in an adult [55]. While
DSM-IV PTSD criteria have been derived more
from work in combat exposure, the next edition
may tailor PTSD criteria more specifically to other
populations such as rape, abuse, and childhood
trauma.
Difficulty initiating sleep (considered a symptom
of hyperarousal) and nightmares (considered much
like a reexperiencing of the traumatic event) are two
sleep-specific symptoms that are included in the
DSM-IV-TR diagnostic criteria for PTSD. Although
not required for the diagnosis, recurring nightmares
have been reported to occur in approximately 50%
of combat veterans [56]. These miserable reliving
experiences often compromise sleep quality, quan-
tity, and daytime mood and function. Polysomno-
graphic evaluation evidences increased sleep
latency, decreased sleep efficiency, and the presence
of dream recall and increased motor activity in both
REM and non-REM (NREM) sleep in these subjects.
Many patients have fatigue or frank sleepiness that
they attribute to this obvious sleep disruption, but
as of yet these symptoms have not been examined
with more objective measures in the literature.
Like many PTSD symptoms vivid, frightening,
reexperiencing nightmares and sleep disruption
have proved difficult to treat. A variety of pharmaco-
logic strategies have been attempted and reported,
including antidepressant medications, anticonvul-
sant medications, hypnotics, and anti-adrenergic
medications [57]. The most effective and well-
demonstrated treatment to date is the alpha-1
adrenergic antagonist prazosin titrated up to 2 to
10 mg as tolerated at bedtime [58–62]. Some
patients report lethargy when dosed during waking
hours and titration requires appropriate vigilance
for orthostatic hypotension, but studies in a variety
of populations have found prazosin to be well
tolerated. Initially studied in a Vietnam-era com-
bat-veteran population, benefits are now being
reported in civilian and ethnically and culturally
diverse PTSD populations that struggle with

nightmares, difficulty maintaining sleep, and auto-
nomic arousal [63–65]. Polysomnographic data,
from a placebo-controlled study, has demonstrated
that while decreasing nightmares and sleep com-
plaints, prazosin increases total sleep time, REM
sleep time, and average REM episode duration with-
out changing sleep latency [64]. While there is evi-
dence prazosin improves patient global level of
functioning and sleep, there is currently no research
available on the subsequent impact on fatigue or
sleepiness in patients with PTSD and prazosin is
not FDA indicated for use in PTSD.
Generalized anxiety disorder
Generalized anxiety disorder (GAD) is character-
ized by excessive anxiety or worry about multiple
concerns that is difficult to control and occurs
more days than not for 6 months. To receive the
diagnosis, three of the following six associated
symptoms should be present: restlessness, fatigue,
irritability, muscle tension, poor concentration, or
sleep disturbance. The sleep disturbances can be
either difficulty initiating or maintaining sleep or
nonrestorative sleep. Interestingly, fatigue, irritabil-
ity, poor concentration, and disturbed sleep often
occur in the setting of a primary sleep disorder. Pol-
ysomnographic studies have indicated lower sleep
efficiency, slow wave sleep, and sleep time [2]. Re-
search findings demonstrate increased sleep
complaints in GAD [66], but it is difficult to tell
whether these subjects have an increased incidence
of sleep disorders, or just the increased subjective
report of sleep complaints. Disturbance of sleep
worsens anxiety symptoms [67], and although anx-
iety can make the treatment of sleep disorders more
complicated, these patients should be diagnosed
and treated aggressively for any comorbid sleep
conditions.
Panic disorder is characterized by the sudden
onset of intense anxiety that is unexpected and
followed by at least 1 month of persistent worrying
about recurrence of the attacks, significance of the
symptoms, or a sustained change in behavior due
to the attacks. Panic disorder can occur with or
without agoraphobia. One third of patients with
panic disorder have attacks that occur during sleep
and it is important to differentiate these from sleep-
disordered breathing given the often associated
symptoms of sweating, racing heart, and shortness
of breath. Excessive daytime sleepiness, fatigue, or
consistent changes in polysomnographic measures
have not been reported.
First-line pharmacologic treatment for GAD,
panic disorder, and PTSD often includes selective
serotonergic reuptake inhibitors (SSRIs). Tricyclic
antidepressants (TCAs) and serotonergic-noradren-
ergic reuptake inhibitors (SNRIs) may also be used,
Sleepiness and Fatigue 211
although tricyclic antidepressants are not FDA ap-
proved for panic disorder, generalized anxiety dis-
order, or PTSD. SSRIs and SNRIs are generally
well tolerated, but sedation from those with higher
antihistaminic activity (ie, paroxetine, sertraline,
mirtazapine, trazodone) can exacerbate sleepiness
or fatigue. TCAs commonly cause sedation and
are therefore typically administered at bedtime.
Emerging circadian findings
The rapidly growing field of chronobiology sits at
the intersection of the expanding frontiers of sleep
disorders research and psychiatric research. Circa-
dian abnormalities may be an occult problem caus-
ing inappropriately timed sleepiness in a wide
range of patient presentations. In addition to find-
ings mentioned previously in the respective sec-
tions on dementia, ADHD, and SAD, there are
recently emerging associations between circadian
abnormalities and other psychiatric disorders. A re-
cent study demonstrated that Clock gene mutant
mice demonstrate maniclike behavior (high-risk
behavior, lowered anxiety and depression, hyperac-
tivity, and decreased sleep), and that many of these
symptoms returned to normal with lithium admin-
istration—a first-line treatment for bipolar disorder
[68]. Furthermore, these abnormal behaviors were
reversed via virally mediated gene transfer to
reinstate functional Clock protein in the ventral teg-
mental area [54,68]. These animal findings are of
note given recent reports of abnormal circadian
rhythms in some bipolar and schizophrenic popu-
lations [52,54]. There has been some speculation
of a possible circadian misalignment in some cases
of nonseasonal depression but data are limited at
this time. Sleep specialists are uniquely suited to
help patients understand how inappropriately
timed sleepiness may complicate the management
of the patient’s psychiatric condition.
Psychiatric medications
The impact of medications and nonprescription
substances on daytime fatigue and sleepiness is
important to consider in the evaluation of a patient
complaining of EDS. Special consideration should
be given to sedating antidepressants, antipsy-
chotics, sedating anti-epileptics (often used in psy-
chiatry for mood stabilization), hypnotic or sleep
aids, and stimulant use. Many psychotropic medica-
tions involve antidopaminergic, anti-adrenergic,
anticholinergic, or antihistaminic activity all of
which may contribute to sleepiness –or fatigue
(Tables 1–3). While side effects are typically worse
at drug initiation or during dose escalations, pa-
tients on stable dosing may continue to have
212 Hagen & Black

Table 1: Relative sedation of antipsychotic Table 3: Relative sedation of mood stabilizers

medications (antimanic agents)a
Typical antipsychotics Carbamazepine (Tegretol) 11
Lithium (Eskalith, lithobid) 11
Chlorpromazine (Thorazine) 111 Valproic acid (Depakote) 11
Thioridazine (Mellaril) 111 Topiramate (Topomax) 1
Loxapine (Loxitane) 11 Lamotrigene (Lamictal) 1
Molindone (Moban) 11
Perphenazine (Trilafon) 1 a
Relative sedation to other medications in Table 1; Seda-
Thiothixene (Navane) 1 tion and fatigue can occur with all mood stabilizing
Droperidol (Inapsine) 1 agents.
Haloperidol (Haldol) 1 1, Low; 11, Moderate; 111, High.
Atypical antipsychotics
Clozapine (Clozaril) 111
Olanzapine (Zyprexa) 11 these symptoms that are due to the underlying dis-
Quetiapine (Seroquel) 11 order versus a medication side effect; however, re-
Risperidone (Risperdal) 1 ductions in fatigue or sleepiness on medication
Ziprasidone (Geodon) 1 withdrawal suggest a significant component is di-
Fluoxetine (Prozac) 1 rectly attributable to medications for many patients.
Citalopram (Celexa) 1 Stimulant medications for ADHD or antidepres-
Escitalopram (Lexapro) 1 sant augmentation carry risks of insomnia. Addi-
1, Low; 11, Moderate; 111, High.
tionally, a careful review of supplements, alcohol,
tobacco, and illicit substances may yield insight
sedation or fatigue related to their medication regi- into complaints of disrupted nocturnal sleep,
sleepiness, and fatigue. When side effects of antide-
men. A study evaluating side effects of antidepres-
pressant medication contribute to a patient’s sleep-
sants found that 40% of 117 patients deemed to
have a successful response to their antidepressant iness or fatigue, sleep physicians play an important
role in collaborating with the prescribing physician.
had symptoms of persistent fatigue or sleepiness
General recommendations include the use of re-
[69]. It is difficult to determine the proportion of
duced medication dosage to the extent adequate
efficacy is retained, use of antidepressant medica-
Table 2: Relative sedation of antidepressant tions that are less likely to contribute to fatigue or
medications sedation, use of antidepressants with more activat-
Tricyclic antidepressants ing side effects (fluoxetine, duloxetine, venlafax-
ine), and/or adjunct stimulant medications, if
Amitriptyline (Elavil) 111 appropriate [40,45,70,71]. While there are case
Clomipramine (Anafranil) 111
reports of combating medication-related fatigue
Doxepin (Sinequan) 111
Trimipramine (Surmontil) 111 and sedation with modafinil, this is exceedingly
Maprotiline (Ludiomil) 11 rare and does not have an FDA indication. The
Amoxapine (Asendin) 1 management of medication-induced fatigue or
Desipramine (Norpramin) 1 sleepiness is more complicated in disorders such
Nortriptyline (Pamelor) 1 as bipolar disorder or schizophrenia. A sleep physi-
Protriptyline (Vivactil) 1 cian respectfully noting potential sedating side
Selective serotonin reuptake inhibitorsa effects and suggesting the use of less-sedating alter-
Paroxetine (Paxil) 1 natives, if clinically appropriate, may be helpful for
Sertraline (Zoloft) 1 physicians working with these populations.
Fluvoxamine (Luvox) 1
Communicating the observation of a sleep disor-
Mono amine oxidase inhibitors
Phenelzine (Nardil) 1 der that interacts with the patient’s psychiatric dis-
Tranylcypromine (Parnate) 1 order, coordinating evaluation and treatment with
Other antidepressants the prescribing primary care physician or psychia-
Trazadone (Desyrel) 111 trist, and alerting him or her to the potential need
Nefazadone (Serzone) 11 for altered medication requirements over time are
Mirtazapine (Remeron) 11 all critical roles for the sleep physician consulting
Venlafaxine (Effexor) 1 on a patient with a psychiatric disorder. As fre-
Bupropion (Wellbutrin) 1 quently occurs in the management of various med-
a
Relative sedation to other medications in Table 1;
ical conditions, medication regimens may need to
SSRI sedation is most common with paroxetine. be changed in some patients after treatment for
1, Low; 11, Moderate; 111, High. OSA, PLMD, and circadian or other sleep disorders.

According to the DSM-IV diagnostic qualifier that
psychiatric disorders or their symptoms are not bet-
ter accounted for by a medical condition, in some
cases psychiatrists must reconsider the original psy-
chiatric diagnosis once a confounding diagnosis,
such as a sleep disorder, is identified and treated.
In summary, there is significant symptom overlap
and uncertainty when fatigue and sleepiness are
found at the intersection of sleep and psychiatric dis-
orders. Recognition of comorbid sleep disorders in
this population will only occur with appropriate vig-
ilance and evaluation. Sleep disorder specialists are
obligated to attend to sleepiness and fatigue in the
psychiatric patient as thoroughly as they would in
the medical or surgical patient. Detection of psychi-
atric symptoms and appropriate mental health refer-
ral is important in the management of the sleepy or
fatigued patient, particularly when a risk of suicide
exists. While it is difficult to determine the origin
of sleepiness and fatigue in the patient with comor-
bid sleep and psychiatric disorders, careful evalua-
tion and treatment for both provides themost
comprehensive and appropriate care. Depressed
and fatigued patients may require more sensitive
techniques for detection of comorbid sleep disor-
ders as mood and fatigue are correlated more with
respiratory related arousal frequency than oxygen
desaturation frequency in some studies. Clarifying
the potential benefits and limitations of sleep disor-
der treatment with both the patient and the mental
health provider is essential in the co-management
of the sleepy or fatigued patient.
References
[1] American Academy of Sleep Medicine. Interna-
tional classification of sleep disorders. Revised.
Diagnostic and coding manual. Chicago: Ameri-
can Academy of Sleep Medicine; 2001.
[2] Sateia MJ, editor. International classification of
sleep disorders: diagnostic and coding manual.
2nd edition. Westchester (IL): American Acad-
emy of Sleep Medicine; 2005.
[3] Bailes S, Libman E, Baltzan M, et al. Brief and
distinct empirical sleepiness and fatigue scales.
J Psychosom Res 2006;60:605–13.
[4] Chalder T, Berelowitz G, Pawlikowska J, et al. De-
velopment of a fatigue scale. Journal of Psycho-
somatic Research 1993;37:147–53.
[5] Krupp LB, LaRocca NG, Muir-Nash J, et al. The
fatigue severity scale. Application to patients
with multiple sclerosis and systemic lupus eryth-
ematosus. Arch Neurol 1989;46:1121–3.
[6] Guilleminault C, Lopes MC, Hagen CC, et al.
The cyclic alternating pattern demonstrates
increased sleep instability and correlates with
fatigue and sleepiness in adults with upper air-
way resistance syndrome. Sleep 2007;30:641–7.
Sleepiness and Fatigue 213
[7] American Psychiatric Association. (DSM-IV-TR)
Diagnostic and statistical manual of mental dis-
orders. 4th edition. Washington, DC: American
Psychiatric Press, Inc.; 2000.
[8] Amphetamine-type drugs for hyperactive chil-
dren. Med Lett Drugs Ther 1972;14:21–3.
[9] Cohen-Zion M, Ancoli-Israel S. Sleep Medicine
Reviews 2004;8(5):379–402.
[10] Cortese S, Konofal E, Yateman N, et al. Sleep and
alertness in children with attention-deficit/hyper-
activity disorder: a systematic review of the liter-
ature. Sleep 2006;29:504–11.
[11] Huang YS, Chen NH, Li HY, et al. Sleep disorders
in Taiwanese children with attention deficit/
hyperactivity disorder. J Sleep Res 2004;13:269–77.
[12] Chervin RD, Archbold KH, Dillon JE, et al. Inat-
tention, hyperactivity, and symptoms of sleep-
disordered breathing. Pediatrics 2002;109:
449–56.
[13] Chervin RD, Archbold KH, Dillon JE, et al. Asso-
ciations between symptoms of inattention,
hyperactivity, restless legs, and periodic leg
movements. Sleep 2002;25:213–8.
[14] Gau SS, Kessler RC, Tseng WL, et al. Association
between sleep problems and symptoms of atten-
tion-deficit/hyperactivity disorder in young
adults. Sleep 2007;30:195–201.
[15] Huang YS, Guilleminault C, Li HY, et al. Atten-
tion-deficit/hyperactivity disorder with obstruc-
tive sleep apnea: a treatment outcome study.
Sleep Med 2007;8:18–30.
[16] Li HY, Huang YS, Chen NH, et al. Impact of
adenotonsillectomy on behavior in children
with sleep-disordered breathing. Laryngoscope
2006;116:1142–7.
[17] Weber SA, Lima Neto AC, Ternes FJ, et al. Hyper-
activity and attention deficit syndrome in
obstructive sleep apnea syndrome: is there
improvement with surgical management? Rev
Bras Otorinolaringol 2006;72:124–9 [in
Portuguese].
[18] Naseem S, Chaudhary B, Collop N. Attention
deficit hyperactivity disorder in adults and
obstructive sleep apnea. Chest 2001;119:294–6.
[19] Wei JL, Mayo MS, Smith HJ, et al. Improved be-
havior and sleep after adenotonsillectomy in
children with sleep disorderd breathing. Arch
Otolaryngol Head Neck Surg 2007;133(10):
974–9.
[20] Mitchell RB, Kelly J. Behavioral changes in chil-
dren with mild sleep-disordered breathing or ob-
structive sleep apnea after adenotonsillectomy.
Larnygoscope 2007;117(9):1685–8.
[21] Stowe CD, Gardner SF, Gist CC, et al. 24-hour
ambulatory blood pressure monitoring in male
children receiving stimulant therapy. Ann Phar-
macother 2002;36:1142–9.
[22] Boonstra AM, Kooij JJ, Oosterlaan J, et al. Hy-
peractive night and day? Actigraphy studies in
adult ADHD: a baseline comparison and the
effect of methylphenidate. Sleep 2007;30:
433–42.
214 Hagen & Black
[23] Picchietti DL, Underwood DJ, Farris WA, et al.
Further studies on periodic limb movement
disorder and restless legs syndrome in children
with attention-deficit hyperactivity disorder.
Mov Disord 1999;14:1000–7.
[24] Oosterloo M, Lammers GJ, Overeem S, et al. Pos-
sible confusion between primary hypersomnia
and adult attention-deficit/hyperactivity disor-
der. Psychiatry Res 2006;143:293–7.
[25] Van der Heijden KB, Smits MG, Van Someren EJ,
et al. Effect of melatonin on sleep, behavior, and
cognition in ADHD and chronic sleep-onset
insomnia. J Am Acad Child Adolesc Psychiatry
2007;46:233–41.
[26] Weiss MD, Wasdell MB, Bomben MM, et al.
Sleep hygiene and melatonin treatment for chil-
dren and adolescents with ADHD and initial in-
somnia. J Am Acad Child Adolesc Psychiatry
2006;45:512–9.
[27] Rybak YE, McNeely HE, Mackenzie BE, et al. An
open trial of light therapy in adult attention-
deficit/hyperactivity disorder. J Clin Psychiatry
2006;67:1527–35.
[28] Schredl M, Alm B, Sobanski E. Sleep quality in
adult patients with attention deficit hyperactivity
disorder (ADHD). Eur Arch Psychiatry Clin
Neurosci 2007;257:164–8.
[29] Fetveit A, Bjorvatn B. Sleep duration during the
24-hour day is associated with the severity of
dementia in nursing home patients. Int J Geriatr
Psychiatry 2006;21:945–50.
[30] Bonanni E, Maestri M, Tognoni G, et al. Daytime
sleepiness in mild and moderate Alzheimer’s
disease and its relationship with cognitive im-
pairment. J Sleep Res 2005;14:311–7.
[31] Doljansky JT, Dagan Y. A chronobiological
approach in treatment of sleep disturbances in
Alzheimer’s dementia patients. Harefuah 2006;
145:437–40 [in Hebrew].
[32] Wu YH, Fischer DF, Kalsbeek A, et al. Pineal
clock gene oscillation is disturbed in Alz-
heimer’s disease, due to functional disconnec-
tion from the ‘‘master clock.’’ FASEB J 2006;20:
1874–6.
[33] Ancoli-Israel S, Gehrman P, Martin JL, et al.
Increased light exposure consolidates sleep and
strengthens circadian rhythms in severe Alz-
heimer’s disease patients. Behav Sleep Med
2003;1:22–36.
[34] Koch S, Haesler E, Tiziani A, et al. Effectiveness
of sleep management strategies for residents of
aged care facilities: findings of a systematic
review. J Clin Nurs 2006;15:1267–75.
[35] Staedt J, Stoppe G. Treatment of rest-activity
disorders in dementia and special focus on
sundowning. Int J Geriatr Psychiatry 2005;20:
507–11.
[36] Ancoli-Israel S, Alessi C. Sleep and aging. Am J
Geriatr Psychiatry 2005;13:341–3.
[37] Skjerve A, Holsten F, Aarsland D, et al. Improve-
ment in behavioral symptoms and advance of
activity acrophase after short-term bright light
treatment in severe dementia. Psychiatry Clin
Neurosci 2004;58:343–7.
[38] First MB, editor. Diagnostic and statistical
manual—text revision (DSM-IV-TR, 2000).
Washington, DC: American Psychiatric Associa-
tion; 2000.
[39] Liu X, Buysse DJ. Sleep and youth suicidal
behavior: a neglected field. Curr Opin Psychiatry
2006;19:288–93.
[40] Fava M. Daytime sleepiness and insomnia as
correlates of depression. J Clin Psychiatry 2004;
65:27–32.
[41] Chellappa SL, Araujo JF. Sleep disorders and
suicidal ideation in patients with depressive
disorder. Psychiatry Res 2007;153(2):131–6.
[42] McCall WV, Harding D, O’Donovan C. Corre-
lates of depressive symptoms in patients with
obstructive sleep apnea. J Clin Sleep Med 2006;
2:424–6.
[43] Sjostrom N, Waern M, Hetta J. Nightmares and
sleep disturbances in relation to suicidality in
suicide attempters. Sleep 2007;30:91–5.
[44] Bardwell WA, Ancoli-Israel S, Dimsdale JE. Com-
parison of the effects of depressive symptoms
and apnea severity on fatigue in patients with
obstructive sleep apnea: a replication study. J Af-
fect Disord 2007;97:181–6.
[45] Baldwin DS, Papakostas GI. Symptoms of fatigue
and sleepiness in major depressive disorder.
J Clin Psychiatry 2006;67:9–15.
[46] Kawahara S, Akashiba T, Akahoshi T, et al. Nasal
CPAP improves the quality of life and lessens the
depressive symptoms in patients with obstruc-
tive sleep apnea syndrome. Intern Med 2005;
44:422–7.
[47] Theorell-Haglow J, Lindberg E, Janson C. What
are the important risk factors for daytime sleepi-
ness and fatigue in women? [see comment].
Sleep 2006;29:751–7.
[48] McMahon JP, Foresman BH, Chisholm RC. The
influence of CPAP on the neurobehavioral
performance of patients with obstructive sleep
apnea hypopnea syndrome: a systematic review.
WMJ 2003;102:36–43.
[49] Brunello N, Akiskal H, Boyer P, et al. Dysthymia:
clinical picture, extent of overlap with chronic
fatigue syndrome, neuropharmacological con-
siderations, and new therapeutic vistas. J Affect
Disord 1999;52:275–90.
[50] Lewy AJ, Lefler BJ, Emens JS, et al. The circadian
basis of winter depression. Proc Natl Acad Sci U
S A 2006;103:7414–9.
[51] Partonen T, Treutlein J, Alpman A, et al. Three
circadian clock genes Per2, arntl, and Npas2
contribute to winter depression. Ann Med
2007;39:229–38.
[52] Bhattacharjee Y. Psychiatric research. Is internal
timing key to mental health? Science 2007;317:
1488–90.
[53] Desan PH, Weinstein AJ, Michalak EE, et al. A
controlled trial of the litebook light-emitting
diode (LED) light therapy device for treatment

of seasonal affective disorder (SAD). BMC Psy-
chiatry 2007;7:38.
[54] McClung CA. Circadian genes, rhythms and the
biology of mood disorders. Pharmacol Ther
2007;114:222–32.
[55] Copeland W, Keeler G, Angold A, et al. Trau-
matic events and posttraumatic stress in child-
hood. Arch Gen Psychiatry 2007;64:577–84.
[56] Neylan TC, Marmar CR, Metzler TJ, et al. Sleep
disturbances in the Vietnam generation: findings
from a nationally representative sample of male
Vietnam veterans. Am J Psychiatry 1998;155:
929–33.
[57] van Liempt S, Vermetten E, Geuze E, et al. Phar-
macotherapy for disordered sleep in post-
traumatic stress disorder: a systematic review.
Int Clin Psychopharmacol 2006;21:193–202.
[58] Dierks MR, Jordan JK, Sheehan AH. Prazosin
treatment of nightmares related to posttraumatic
stress disorder. Ann Pharmacother 2007;41:
1013–7.
[59] Peskind ER, Bonner LT, Hoff DJ, et al. Prazo-
sin reduces trauma-related nightmares in older
men with chronic posttraumatic stress disor-
der. J Geriatr Psychiatry Neurol 2003;16:
165–71.
[60] Raskind MA, Peskind ER, Hoff DJ, et al. A par-
allel group placebo controlled study of
prazosin for trauma nightmares and sleep dis-
turbance in combat veterans with post-trau-
matic stress disorder. Biol Psychiatry 2007;61:
928–34.
[61] Raskind MA, Thompson C, Petrie EC, et al. Pra-
zosin reduces nightmares in combat veterans
with posttraumatic stress disorder. J Clin Psychi-
atry 2002;63:565–8.
[62] Raskind MA, Dobie DJ, Kanter ED, et al. The
alpha1-adrenergic antagonist prazosin amelio-
rates combat trauma nightmares in veterans
Sleepiness and Fatigue 215
with posttraumatic stress disorder: a report of 4
cases. J Clin Psychiatry 2000;61:129–33.
[63] Boehnlein JK, Kinzie JD. Pharmacologic reduc-
tion of CNS noradrenergic activity in PTSD: the
case for clonidine and prazosin. J Psychiatr Pract
2007;13:72–8.
[64] Taylor FB, Martin P, Thompson C, et al. Prazosin
effects on objective sleep measures and clinical
symptoms in civilian trauma posttraumatic
stress disorder: a placebo-controlled study. Biol
Psychiatry 2008;63(6):629–32.
[65] Taylor F, Raskind MA. The alpha1-adrenergic
antagonist prazosin improves sleep and night-
mares in civilian trauma posttraumatic stress
disorder. J Clin Psychopharmacol 2002;22:82–5.
[66] Alfano CA, Beidel DC, Turner SM, et al. Prelim-
inary evidence for sleep complaints among
children referred for anxiety. Sleep Med 2006;7:
467–73.
[67] Mellman TA. Sleep and anxiety disorders. Psy-
chiatr Clin North Am 2006;29:1047–58.
[68] Roybal K, Theobold D, Graham A, et al. Mania-
like behavior induced by disruption of CLOCK.
Proc Natl Acad Sci U S A 2007;104:6406–11.
[69] Fava M, Graves LM, Benazzi F, et al. A cross-
sectional study of the prevalence of cognitive and
physical symptoms during long-term antidepres-
sant treatment. J Clin Psychiatry 2006;67:1754–9.
[70] Fava M, Thase ME, DeBattista C. A multicenter,
placebo-controlled study of modafinil augmen-
tation in partial responders to selective serotonin
reuptake inhibitors with persistent fatigue and
sleepiness. J Clin Psychiatry 2005;66:85–93.
[71] Stahl SM, Zhang L, Damatarca C, et al. Brain
circuits determine destiny in depression: a novel
approach to the psychopharmacology of wake-
fulness, fatigue, and executive dysfunction in ma-
jor depressive disorder. J Clin Psychiatry 2003;
64:6–17.
 217

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 217–229

Parasomnias: Psychiatric
Considerations
a,b c,d,
David T. Plante, MD , John W. Winkelman, MD, PhD *

- Non–rapid eye movement parasomnias: Rapid eye movement behavior disorder

disorders of arousal Sleep paralysis
Confusional arousals Nightmare disorder
Sleepwalking - Other parasomnias
Sleep terrors Sleep-related eating disorder
- Sleep-related sexual behavior and Sleep-related hallucinations
sleep-related violence Sleep-related dissociative disorders
- Evaluation and treatment of non–rapid - Summary
eye movement parasomnias - References
- Rapid eye movement–related parasomnias

Parasomnias are a group of sleep disorders
broadly defined as undesirable physical or experi-
ential events that occur within entry into sleep,
during sleep, or during arousals from sleep [1]. In
centuries past, parasomnias were often thought re-
lated to mystical or supernatural forces. In the early
twentieth century, with the development of psycho-
analytic theory, which introduced for the first time
the notion that unconscious drives affect human
behavior, psychoanalysts debated the nature of
parasomnias, often under the assumption that
such behavior represented the expression of latent
desires [2]. As a result, parasomnias were believed
for decades to be caused by underlying mental ill-
ness. With the more recent rise and development
of sleep medicine as a discipline, most parasomnias
are not believed to be directly related to psychiatric
illness [3]. However, since these disorders are be-
havioral in nature, often co-occur with psychiatric
illness, may present as psychiatric complaints, and
can be induced by psychotropic medications, para-
somnias are best conceptualized from an interdisci-
plinary perspective, and their optimal management
requires that psychiatric considerations be under-
stood by all clinicians who treat them, regardless
of discipline of origin.
Within psychiatry, the standard of diagnostic cat-
egorization is the Diagnostic and Statistical Manual
(DSM), which in its most current edition minimally
delineates the parasomnias into one of four diagno-
ses (Table 1). Unfortunately, this diagnostic
schema fails to capture the heterogeneity of the

a
Department of Psychiatry, Massachusetts General Hospital and McLean Hospital, Harvard Medical School,
Boston, MA, USA
b
McLean Hospital, Outpatient Clinic, 115 Mill Street, Belmont, MA 02478, USA
c
Divisions of Sleep Medicine and Psychiatry, Brigham & Women’s Hospital, Harvard Medical School, Boston,
MA, USA
d
Sleep Health Center Affiliated with Brigham & Women’s Hospital, 1505 Commonwealth Avenue, Brighton,
MA 02135, USA
* Corresponding author. Sleep Health Centers, Affiliated with Brigham & Women’s Hospital, 1505 Common-
wealth Avenue, Brighton, MA 02135.
E-mail address: jwwinkelman@partners.org (J.W. Winkelman).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.005
sleep.theclinics.com
218 Plante & Winkelman

Table 1: Parasomnia classification of DSM-IV and ICSD-2

Disorders of arousal
(from NREM sleep)
DSM-IV-TR  Sleepwalking
 Sleep terrors
 Confusional arousals
ICSD-2
Parasomnias usually
associated with REM
sleep Other parasomnias
 Nightmare disorder  Parasomnia not otherwise
specified/unspecifieda
 REM sleep behavior  Sleep-related dissociative
disorder disorders
 Recurrent isolated  Sleep enuresis
sleep paralysis  Sleep-related groaning
(catathrenia)
 Exploding head syndrome
 Sleep-related
hallucinations
 Sleep-related eating
disorder

Abbreviations: DSM-IV, Diagnostic and Statistical Manual, 4th edition; ICSD, International classification of sleep disor-
ders; NREM, non–rapid eye movement; REM, rapid eye movement.
a
‘‘Parasomnia not otherwise specified’’ according to DSM-IV-TR includes those parasomnias not boxed; whereas ‘‘par-
asomnia unspecified’’ according to ICSD-2 includes parasomnias not listed. Of note, ICSD-2 notes that ‘‘parasomnia
unspecified’’ is intended for parasomnias that cannot be classified elsewhere or for cases in which the physician
has a clinical suspicion that an underlying psychiatric condition may cause the parasomnia. Omitted are ‘‘parasomnias
due to medical condition’’ and ‘‘parasomnia due to drug or substance’’ (ICSD-2), which are classified as sleep disorder
caused by a general medical condition, parasomnia type, and substance-induced sleep disorder, parasomnia type,
respectively under DSM-IV-TR [1,72].

parasomnias, and is not based on current knowl-

edge of the underlying pathology of these disorders.
As the brain cycles through three primary states of
being (wakefulness, non–rapid eye movement
[non-REM] sleep, and rapid eye movement [REM]
sleep) separation between the states may become
blurred or oscillate rapidly, giving rise to parasom-
nias [3,4]. The International Classification of Sleep
Disorders (ICSD), used primarily by sleep medicine
clinicians, is structured around this pathologic par-
adigm, and is further differentiated than the DSM.
Fortunately, most practicing psychiatrists realize
the limitations of the DSM, and it is the authors’
hope that the upcoming fifth edition of the DSM
(expected to be published in 2012) will reflect
more current understanding of these disorders.
This article focuses on parasomnias that are most
frequently seen in psychiatric settings or have con-
nections to neuropsychiatric illness including
non-REM parasomnias of arousal, REM-related
parasomnias, and other parasomnias including
sleep-related eating disorder (SRED), sleep-related
hallucinations, and sleep-related dissociative disor-
ders (SRDD) (Table 2). It should be noted that
there are no pharmacologic agents approved by
the Food and Drug Administration for the treat-
ment of parasomnias, and recommendations are
based on the available scientific literature or the
clinical experience of the authors.
Non–rapid eye movement parasomnias:
disorders of arousal
The non-REM parasomnias (sleepwalking, sleep ter-
rors, and so forth) are a group of related disorders of
arousal that share common features and likely un-
derlying pathology. During a partial arousal from
sleep, behaviors (eg, ambulation, eating, and so
forth) or mood states (eg, anger, fear) are expressed
that are not fully (or at all) under conscious control;
nor do they occur with ordered judgment or full in-
tegration of environmental feedback [5]. Episodes
tend to arise from slow wave sleep ([SWS] stages 3
and 4 of non-REM sleep); typically occur during
the first 1 to 2 hours of the sleep period; and are sel-
dom remembered by the individual on final awak-
ening. Non-REM parasomnias are common in
childhood, but usually diminish with increasing
age. There is frequently a family history of disorders
of arousal in individuals who present for evaluation,
and these disorders are likely expressed when envi-
ronmental factors affect a genetically predisposed
individual [6]. In susceptible individuals, precipitat-
ing events can be either endogenous (eg, breathing
or movement abnormalities, pain) or exogenous
(eg, noise, light, sleep deprivation, medications) fac-
tors that disturb sleep [4].
As a result of similarities between these disorders,
the categorization of non-REM parasomnias is
 Table 2: Characteristics of parasomnias
Non-REM parasomnias REM parasomnias Other parasomnias
Confusional Sleep Nightmare
Parasomnia arousals Sleepwalking terrors RBD Sleep paralysis disorder SRED SRHa SRDD
Stage of arousal II, III, IV III, IV III, IV REM REM REM II, III, IV REM NREM or REM
Typical time Anytime First 2 h First 2 h Anytime Anytime Anytime Anytime Onset/ Anytime
of night (first 2 h) offset
of sleep
EEG during event NA Mixed Mixed REM Wake NA Mixed Wake Wake pattern
pattern pattern pattern
EMG activity [ [ [ [ Y NA [ NA [
during event
Decreased 1 1 1 1 1 1 1
responsiveness
during event
Autonomic 1 1  1  

Parasomnias: Psychiatric Considerations

hyperactivity
Amnesia 1 1 1 (dream (experience (dream  (partial) 1
recall) recall) recall)
Confusion 1 1 1 1 1
postepisode
Family history 1 1 1  1  Unknown

Abbreviations: EEG, electroencephalogram; EMG, electromyogram; NREM, non–rapid eye movement; RBD, REM behavior disorder; REM, rapid eye movement; SRED, sleep-related
eating disorder; SRH, sleep-related hallucinations; SRDD, sleep-related dissociative disorders.
a
SRH for this table includes hypnagogic/pompic hallucinations. The few EEG studies of complex nocturnal hallucinations suggest they can occur from NREM.

219
220 Plante & Winkelman
typically based on the types of behaviors and mood
states exhibited. In the following sections, these
disorders are presented along a continuum of
behavioral and affective arousal, acknowledging
that clinical overlap often exists among these
disorders.
Confusional arousals
Confusional arousals are typically brief, simple
motor behaviors that occur with little emotional
expression, lack of responsiveness to the environ-
ment, and are associated with mental confusion
on arousal or awakening [7]. The motoric behaviors
are simple and may be accompanied by indistinct
vocalization. Episodes are brief, and because of
dense amnesia for the episode, without collateral
information from a bedpartner or parent, they of-
ten go unnoticed. Cross-sectional prevalence is esti-
mated at 4.2%, with comparable rates among men
and women, and prevalence that decreases with age
[7].
A variant of confusional arousals has been de-
scribed as ‘‘sleep drunkenness’’ or excessive sleep in-
ertia [8,9]. Confusional arousals are distinguished
from excessive sleep inertia because the latter phe-
nomenon occurs from final awakening; however,
both are similar in regards to their immediate devel-
opment from sleep, impaired mentation, automatic
behavior, and relative unresponsiveness to the envi-
ronment. The duration during which excessive
sleep inertia can affect an individual is typically
minutes, but can be up to 4 hours [9]. Sleep inertia
occurs from both naps and full sleep periods, and
its severity and duration is likely related to the
depth of prior sleep. Studies using self-report to ex-
amine the prevalence of confusional arousals may
focus on sleep inertia, because patients are more
apt to recall these episodes as they achieve full
wakefulness. One such study found bipolar disor-
der, a major mood disorder characterized by epi-
sodes of depression and mania, was strongly
associated with such self-reported confusional
arousals, although the significance of this finding
is unclear [7].
Sleepwalking
Sleepwalking, or somnambulism, is distinguished
by greater complexity of behavior than confusional
arousals. Individuals who sleepwalk have limited
conscious control during an episode, but may recall
simple motivations (eg, desire to urinate) if awak-
ened during an episode. Sleepwalkers typically
have eyes open during an event; may be clumsy in
their behavior; and if undisturbed, typically return
to sleep, although they may do so in atypical places
[10,11]. Although sleepwalking behaviors are not
usually dangerous, individuals may injure
themselves (eg, climbing out a window) or others,
and interruption may evoke a range of responses,
including agitation or violence. Similar to other
non-REM parasomnias, episodes typically arise
from SWS during the first part of the sleep episode,
and amnesia for the episode is usually present.
Somnambulism is common in childhood, occur-
ring in 10% to 20% of all children, with the greatest
prevalence occurring between 3 and 10 years of age
[12]. Because its prevalence decreases with increas-
ing age, sleepwalking is considered in most cases to
be a transiently disruptive, rather than pathologic,
process that typically resolves by adolescence. Som-
nambulism occurs in 1% to 4% of adults, however,
with roughly 80% of cases a continuation of child-
hood behavior [7]. Evaluation in adults is typically
prompted when an individual’s bedpartner is con-
cerned by the frequency or dangerousness of the be-
havior. Prevalence of sleepwalking does not seem to
be associated with gender, race, or socioeconomic
conditions [12]. Individuals with sleepwalking
likely have a genetic predisposition, as evidenced
by epidemiologic, twin, and HLA mapping studies
[13,14]. The risk of somnambulism is roughly dou-
bled if one parent, and tripled if both parents, have
a history of sleepwalking.
The relationship between mental illness and
sleepwalking has been a long debated issue [15]. Al-
though childhood sleepwalking does not seem to
be directly related to psychiatric pathology, it has
been suggested that psychopathology may be asso-
ciated with sleepwalking in adolescence and adult-
hood [7,16]. There is insufficient evidence to
suggest sleepwalking behaviors represent uncon-
scious motivations acted out during sleep; however,
there are no controlled studies of sleepwalkers
simultaneously in psychotherapy during which
unconscious desires might be explored in detail
[17]. It is critical to realize, however, that psychotro-
pic medications may raise the risk of adult som-
nambulism because of their effects on sleep and
wakefulness [18]. Also, because nearly every major
psychiatric illness (eg, depression, bipolar disorder,
schizophrenia) is associated with disturbed sleep,
this may increase the risk of sleepwalking in these
individuals, particularly if they had sleepwalking
behaviors as a child [19].
Sleep terrors
Sleep terrors are similar to sleepwalking; however,
they are distinguished by more intense motor, emo-
tional, and autonomic activity. Rather than constru-
ing sleep terrors and somnambulism as distinct
disorders, it is more useful to consider them related
entities that can evolve from one another. Similar to
somnambulism, sleep terrors usually occur in the
first third of the sleep period and are believed to

be caused by a confluence of genetic susceptibility
with precipitating factors [20]. Sleep terrors occur
in roughly 5% of children and are typically dra-
matic: a piercing scream, followed by fear, crying,
and inconsolability [7,21]. In adults, the preva-
lence of sleep terrors is 1% to 2%, and presentation
is less stereotypical, usually involving agitation, of-
ten with injury to self, others, or property during
an episode [7,21]. Similar to somnambulism, sim-
ple thoughts may be recalled (eg, ‘‘I am in danger’’),
which can be difficult to dispel even once awak-
ened, but patients typically do not report dreaming
and are amnesic for the episode. Confrontation of
an individual in the midst of an episode can be
dangerous, because there is the real danger of be-
ing incorporated into the sleep terror leading to
violence.
Sleep-related sexual behavior
and sleep-related violence
Sleep-related sexual behavior and sleep-related vio-
lence are non-REM parasomnias that consist of
more complex behaviors and emotional states and
are poorly understood compared with those previ-
ously discussed. Because these disorders cause sig-
nificant harm and may have forensic implications,
however, there is an urgency to develop a more
comprehensive understanding of these disorders.
Sleep-related sexual behavior, or ‘‘sexsomnia,’’ is
a parasomnia in which sexual behavior occurs
with limited awareness during the act, relative unre-
sponsiveness to the external environment, and am-
nesia for the event [22]. The sexual behavior may
range from sexual vocalizations to intercourse,
and may be atypical for the patient in terms of part-
ner or type of sexual act (eg, anal intercourse) [23].
It has been proposed that distinguishing features of
sexsomnia that differentiate it from sleepwalking
include more widespread autonomic activation,
sexual arousal, and duration of behavior that
can occasionally exceed 30 minutes [22]. Of note,
parasomnias are not the only potential cause of
sleep-related sexual behavior, and other disorders
including nocturnal seizures and Kleine-Levin
syndrome, a rare disorder typically presenting in
adolescence with episodic hypersomnolence, hy-
perphagia, and hypersexuality, should be included
in the differential diagnosis [24].
Sleep-related violence is best conceptualized as
an overlap disorder of sleep terrors following sleep-
walking [25]. Violent behavior occurs in a state con-
sistent with night terrors, with anger or fear as the
primary emotion, and agitation directed toward
individuals who may be in close proximity or con-
front the individual [26]. Individuals slowly return
to normal levels of alertness, or may go back to
Parasomnias: Psychiatric Considerations 221
sleep, and typically have amnesia for the episode.
Like sexsomnia, the violent behavior is often atypi-
cal for the patient. Most cases have been young to
middle-aged men with a previous history of sleep-
walking [27].
The legal concept of mens rea, criminal intent, is
of great import in the prosecution of individuals
with sleep-related violence or sleep-related sexual
behavior. Because the pathophysiology of disorders
of arousal may involve relative deactivation of the
frontal lobe, which is largely responsible for logic
and reasoned judgment, and the inappropriate acti-
vation of limbic areas, regions in the brain respon-
sible for the expression of emotion, it may be
argued that criminal intent does not exist in these
cases [26]. Often defendants may argue that alcohol
induced the violent or sexual act; however, there is
currently no direct experimental evidence that alco-
hol triggers sleepwalking or related disorders [28].
Typically, these cases are quite complex and convic-
tion or acquittal by a jury in cases where a non-REM
parasomnia is used as a defense is relatively
unpredictable.
Evaluation and treatment of non–rapid
eye movement parasomnias
Polysomnography (PSG) is often not necessary for
the evaluation of non-REM parasomnias, and
attempts to document somnambulism and sleep
terrors by PSG are often unsuccessful [29]. Poly-
somnographic markers of susceptibility have been
studied, however, for both their diagnostic use
and potential insight into pathogenesis. Most poly-
somnographic studies of sleepwalkers demonstrate
increased brief arousals from SWS with a preserved
sleeping electroencephalogram, and autonomic ac-
tivation following the arousal [30,31]. Similarly,
multiple brief arousals with autonomic hyperactiv-
ity may be observed as a marker of sleep terrors
[32]. PSG may be indicated when there is a new
onset of parasomnias without a prior history of
childhood parasomnias, because there may be
a treatable cause of arousal from SWS present,
such as sleep-related breathing disorder, periodic
limb movements of sleep, or nocturnal seizures.
The differential diagnosis of unwanted nocturnal
behaviors also includes nocturnal panic attacks,
frontal lobe seizures, delirium associated with med-
ical or neurologic disorders, and nocturnal dissocia-
tive disorders and REM sleep behavior disorder
(discussed later). Of note, nocturnal panic attacks
refer to episodes where individuals awaken from
sleep, typically without dream recall, and experi-
ence panic attacks with full awareness that may in-
clude tachycardia, diaphoresis, shaking, shortness
of breath, chest discomfort, nausea, paresthesias,
222 Plante & Winkelman
or intense fear. These episodes are common among
individuals with panic disorder, and a daytime his-
tory of similar events should be elicited, although
nocturnal panic attacks may occasionally occur in
isolation [33].
The frequency of the parasomnia event, the risk
of injury to self or others, and the distress the be-
havior is causing the patient or family members,
should all be carefully considered when managing
non-REM parasomnias [5]. For most children, para-
somnias do not require pharmacologic treatment
because the behavior may be intermittent, self-lim-
ited, poses little risk of harm to the child, and does
not negatively affect daytime functioning. Keeping
regular sleep-wake times, avoidance of sleep depri-
vation, voiding before bed, and treatment of under-
lying sleep disorders (eg, sleep apnea, restless legs
syndrome) may decrease the frequency of episodes.
Also, improving the safety of the sleeping environ-
ment (eg, locking doors and windows, keeping
hallways and stairs well lit) is important in reducing
potential harm from such episodes.
When treatment of non-REM parasomnias in
adults is indicated, it is done following a three-
step model: (1) modifying predisposing and
precipitating factors, (2) improving the safety of
the sleeping environment, and (3) pharmacother-
apy if necessary. In the decision to pursue pharma-
cologic therapy, patients and physicians should
collaboratively decide whether continuous treat-
ment for behaviors that are usually episodic in
nature is warranted. The agents most frequently
used to treat non-REM parasomnias are benzodiaz-
epines; however, there are no controlled trials of
these agents, and their use is generally guided by
clinical experience. Clonazepam (0.5–2 mg at bed-
time), a long-acting benzodiazepine, has been used
successfully for extended periods without the devel-
opment of tolerance in most cases [34]. The long
half-life of clonazepam increases the likelihood of
daytime side effects, however, and in such instances
a shorter-acting benzodiazepine (eg, lorazepam,
triazolam) may be efficacious. Although clinically
useful, it is unclear whether these medications
work by suppressing arousals during sleep or de-
creasing SWS.
Rapid eye movement–related parasomnias
Specific physiologic and experiential changes that
occur during REM sleep include atonia of the
voluntary muscles (except extraocular); elevated
autonomic activity; and dreaming. REM-related
parasomnias involve either the incoordination of
these processes or the inappropriate admixture of
REM sleep and wakefulness.
Rapid eye movement behavior disorder
REM behavior disorder (RBD) is characterized by
the loss of coordination of dreaming and paralysis
of the skeletal muscles during sleep, causing indi-
viduals with this disorder to act out their dreams,
with a mixture of simple and complex motor be-
haviors, which can include screaming, punching,
kicking, and so forth. Behavior is enacted with
eyes closed and unresponsiveness to the surround-
ing environment. When awakened while acting
out a dream, the individual rapidly achieves full
alertness and often reports a dream that corre-
sponds to their behavior. Agitated or violent behav-
ior leading to self-injury or injury of a bedpartner is
often the impetus that leads to evaluation by a phy-
sician. RBD may also present coincidentally with
primarily psychiatric symptoms (eg, as part of
a depressive episode), with symptoms of RBD
symptoms unearthed when taking a sleep history
[35]. Similarly, given the propensity of serotonergic
antidepressants to provoke RBD, psychiatrists may
be the first physicians consulted for this problem.
The prevalence of RBD is estimated to be between
0.04% and 0.5% of the population [36]. RBD is typ-
ically separated into acute versus chronic forms,
which are thought to have different underlying
causes. Acute RBD is often associated with medica-
tions; drugs of abuse; or withdrawal (particularly
alcohol) [37]. Chronic RBD is most commonly
seen in men over 50 years of age, and is further sub-
divided into two types: idiopathic and secondary to
a neurologic process. The diseases most commonly
associated with RBD are the a-synucleinopathies in-
cluding Parkinson’s disease, dementia with Lewy
bodies, and multiple system atrophy, all of which
are characterized by pathologic accumulation of
the protein a-synuclein [38]. Although the three
largest cohorts of chronic RBD suggest that roughly
60% of chronic RBD is idiopathic, ongoing follow-
up studies suggest that idiopathic RBD may be
a prodromal symptom of an underlying neurologic
illness that may or may not manifest during the life-
span of patients with the disorder [36,39].
Although the pathophysiology of RBD is not en-
tirely elucidated, the extrapyramidal and REM sleep
systems in the brainstem share specific neuronal
connections, which may be central to RBD patho-
genesis. Animal models of RBD in which brainstem
lesions near the locus coeruleus produced REM
sleep without atonia existed well in advance of clin-
ical description of the disorder, and more recent re-
search suggests neurons may be reduced in similar
regions in RBD [40,41]. Furthermore, reduced do-
pamine transporters and reduced dopaminergic in-
nervation in the striatum have been demonstrated
in RBD [42,43]. One intriguing hypothesis is that

the pedunculopontine nucleus may play a role in
the REM-atonia circuitry and its disruption in
RBD, connecting clinical observations regarding
the a-synucleinopathies and extrapyramidal system
and RBD with observations in pontine-lesioned an-
imal models [44].
PSG along with clinical history is necessary to
confirm the diagnosis of RBD. PSG demonstrates
elevated muscle tone or increased phasic muscle
activity in the chin (submental) or limb (anterior
tibialis) electromyogram during REM sleep [1].
Periodic limb movements of sleep are common in
RBD, although otherwise the PSG is typically nor-
mal. Further nonspecific findings in RBD may in-
clude general slowing of the waking EEG, subtle
neuropsychologic dysfunction, autonomic dysfunc-
tion, subtle abnormalities of motor and gait speed,
impairment in color discrimination, and olfactory
dysfunction [45–48].
The management of RBD, like other parasom-
nias, is typically behavioral and pharmacologic. En-
suring safety of the bedpartner is important, and it
may be recommended that the bedpartner sleep in
another room until symptoms are controlled, par-
ticularly if there is a history of injurious behavior.
Patients may devise their own home remedies,
such as tying themselves to the bed, but these
may be dangerous as evidenced by the fact that re-
straint of agitated psychiatric patients in inpatient
settings is a significant cause of morbidity and mor-
tality [37,49]. Prudent behavioral management
includes locking doors and windows; limiting ob-
jects (eg, furniture) on which a patient may injure
themselves in the bedroom; sleeping on a mattress
on the floor; and sleeping on the ground floor if
possible. Before starting a medication to treat
RBD, it is important to limit any medications that
may be contributing to or causing the disorder. Se-
rotonergic antidepressants, monamine oxidase in-
hibitors, and tricyclic antidepressants have been
associated with subclinical RBD (the disinhibition
of motor tone during REM sleep without obvious
clinical symptoms), and may lead to or exacerbate
RBD and should be discontinued if clinically feasi-
ble [50,51]. Any discontinuation of antidepressant
medication should be discussed with the prescrib-
ing physician (eg, psychiatrist) before discontinua-
tion, however, because 15% of those with
depressive disorders commit suicide, necessitating
coordination of care after antidepressant discontin-
uation [52].
Benzodiazepines are typically the first-line phar-
macologic agents in the treatment of RBD. Clonaze-
pam (0.5–2 mg) is most commonly used and has
been shown to decrease the frequency and extent
of problematic dream-enacting behavior [51]. In
general, clonzaepam is well-tolerated; however,
Parasomnias: Psychiatric Considerations 223
cognitive impairment, motor disturbance, and day-
time sedation may limit its use, particularly in older
individuals. Alternative treatment strategies include
using a benzodiazepine with a shorter half-life
(eg, lorazepam, 1–2 mg), or using a different class
of medications. There are small case series suggest-
ing efficacy of melatonin (3–15 mg qhs) and prami-
pexole (0.5–1 mg qhs) in RBD, which may be
useful in patients who cannot tolerate a benzodiaz-
epine or who may have a history of substance abuse
or dependence [53,54].
Sleep paralysis
Sleep paralysis (SP) refers to paralysis of the volun-
tary musculature associated with a conscious state
at the onset or offset of sleep. Episodes may last sec-
onds to minutes, and often resolve spontaneously
on reentry into sleep or when touched. SP is
thought to result from inappropriate REM intrusion
into wakefulness, or conversely the failure to main-
tain sleep during REM periods [55]. SP can occur at
any time of night, but tends to be clustered in the
first 2 hours of the sleep period or at the final awak-
ening, and may be worsened by sleep deprivation
and supine positioning [55–57]. Estimates of the
lifetime prevalence of SP vary widely between
2.3% and 40%, with multiple episodes occurring
in 1% to 10% of the population [55]. Although
most patients with SP likely do not have associated
psychiatric illness, SP occurs at higher rates among
individuals with bipolar disorder, depression, anxi-
ety disorders, and posttraumatic stress disorder
(PTSD) compared with the general population
[58,59]. Cultural factors may influence a patient’s
subjective experience and report of symptoms,
with many cultures relating the phenomenon to
beliefs about spirits or supernatural forces [60].
Although many patients may have symptoms in
isolation, SP with or without hypnogogic-popmic
hallucinations (discussed later) should prompt in-
quiry about other symptoms of narcolepsy, because
SP is a classic, although nonspecific, symptom of
this disorder [61]. Unless narcolepsy is suspected
by history, PSG is typically not indicated for SP.
Treatment of underlying psychiatric illness (eg, de-
pression) may be helpful in the management of
SP; however, in the absence of comorbid illness, re-
assurance and education is most useful in isolated
cases. If the frequency of SP is bothersome to pa-
tients, there is a suggestion that antidepressants
may be of benefit, likely because of their REM-
suppressing properties [62].
Nightmare disorder
The study of dreams has long been of interest in
psychiatry, likely stemming from early psychoana-
lytic theory. Freud, who himself developed myriad
224 Plante & Winkelman
revolutionary theories on the workings of the mind
(although many have been criticized in modern
times), considered his theory on the meaning of
dreams to have been his greatest accomplishment
[63]. Although the complexities of his theory are
beyond the scope of this discussion, Freud believed
dreams to be ‘‘the royal road to a knowledge of the
unconscious,’’ because they were disguised fulfill-
ments of wishes and desires that could not be
expressed consciously [64]. With advances in sleep
medicine, it is known from a physiologic perspec-
tive that dreams occur predominantly (although
not exclusively) during REM sleep.
Nightmare disorder is characterized by recurrent,
affect-laden dreams, followed by awakening, typi-
cally with detailed recall of the dream. Fear is the
most common emotion; however, anger, embar-
rassment, and sorrow might also be experienced.
Nightmares typically occur in the latter third of
the night, which coincides with the increased pro-
portion of REM sleep within that portion of the
sleep period [65]. Unlike RBD, nightmares are not
typically associated with motoric dream enactment.
After awakening, nightmare recall and heightened
autonomic activity may make it difficult for an indi-
vidual to return to sleep. In addition, fear of night-
mares may lead some individuals to be afraid to
initiate sleep at the beginning of the night. Night-
mares can lead to either sleep onset or sleep main-
tenance insomnia.
Estimates of the prevalence of nightmares and
their association with psychiatric illness are hin-
dered by inconsistent definitions among re-
searchers. Still, roughly 5% to 8% of adults in the
general population experience current nightmares,
and nightmares are more common in women
than men [66,67]. Most studies have found night-
mares to be associated with many psychiatric diag-
noses including depression, substance abuse
disorders, and personality disorders [68]. Interest-
ingly, there is growing evidence of a connection be-
tween nightmares and suicidality, although a causal
relationship cannot be inferred at present [69,70].
There is some evidence suggesting distress from
nightmares, rather than frequency of occurrence,
may be more closely tied to mental illness [71].
The psychiatric illness most commonly associ-
ated with nightmares is PTSD. PTSD is currently
classified as an anxiety disorder in the DSM-IV,
and its technical diagnostic criteria are somewhat
complex [72]. In brief, diagnosis of the disorder re-
quires that an individual has been exposed to a trau-
matic event and experiences symptoms from three
clusters: (1) re-experiencing of the trauma (eg,
nightmares about the event); (2) avoidance of
stimuli associated with the trauma; and (3) persis-
tent symptoms of arousal [72]. Within current
DSM-based nosology, nightmare disorder should
not technically be diagnosed if the nightmares are
thought secondary to PTSD. Conversely, the
ICSD-2 considers PTSD a predisposing factor for
nightmare disorder [1]. In clinical practice, many
psychiatrists treat nightmares as a separate, treatable
component of PTSD, directing specific interven-
tions toward this symptom. Nightmares associated
with PTSD tend to be recurrent, and have similar
thematic content to the experienced trauma history
[73]. The prevalence of nightmares in those with
PTSD varies depending on the study population
and nature of the trauma. Surveys of the general
population reveal individuals with PTSD report
nightmares at rates five times that of the general
population [74]. Interestingly, a tendency to experi-
ence ‘‘bad dreams’’ and disrupted sleep preceding
Hurricane Andrew was found to be a risk factor
for developing PTSD after the event, suggesting pre-
morbid sleep disruption may be a susceptibility
marker for those who develop PTSD [75].
PSG is not routinely indicated for evaluation of
nightmares unless necessary to rule out another
sleep disorder. Numerous small studies have sug-
gested effectiveness of prazosin, topiramate, and
atypical antipsychotics in the treatment of PTSD-
related nightmares, although no uniform consen-
sus strategy exists and choice of medication is best
made on a case-by-case basis [76–78]. Psychothera-
peutic strategies including imagery rehearsal ther-
apy and guided hypnosis, in which alternative
versions of nightmares with better outcomes are re-
hearsed while awake or in a state of deep relaxation,
have also shown potential benefit for the treatment
of nightmares, both trauma and nontrauma related
[79,80].
Other parasomnias
The following section details three parasomnias
characterized in the ICSD-2 as ‘‘other’’ parasomnias:
SRED, sleep-related hallucinations, and SRDD [1].
Although in some instances they may be more
closely related to non-REM parasomnias (eg,
SRED) or REM parasomnias (eg, sleep-related hal-
lucinations), their categorization as ‘‘other’’ reflects
divergence in presentation and management from
more classic non-REM and REM parasomnias
already discussed.
Sleep-related eating disorder
SRED is a parasomnia that has only recently been
described in the medical literature, but has become
a popular topic in the media [81]. SRED is best con-
ceptualized as a combination of the binge eating
behaviors of bulimia nervosa or binge eating disor-
der (eating disorders in which patients eat excessive

quantities with or without purging) with the disor-
dered arousal, confusional behavior, and amnesia
of a non-REM parasomnia [82,83]. Episodes typi-
cally occur with repetitive partial arousals within
the first 2 to 3 hours of sleep, and ingestion of
food occurs in a hurried or ‘‘out of control’’ manner,
despite frequent lack of hunger at the time of the
episode. Foods consumed are often high-carbohy-
drate, but may also be unusually combined foods,
frozen foods, or nonnutritive substances. Patients
often feel ashamed of their behaviors, gain weight
because of the binges, and may forcibly attempt
to control their overall caloric intake by daytime an-
orexia. Awareness during episodes can be variable,
with patients often reporting that they were mostly
asleep or half-awake, half-asleep. In some instances,
patients report complete amnesia for the episodes,
and specifics may be elucidated by witnesses to
the episodes, or reconstructed from evidence on
awakening (eg, food missing, messy eating place,
food in bed). Unlike other classic non-REM para-
somnias, the level of awareness may vary between
episodes within the same night, and from night to
night over the longitudinal course of the disorder.
Level of awareness during the eating episode is
what distinguishes SRED from night eating syn-
drome (NES), a disorder characterized by eating
excessive amounts of food either before bed or dur-
ing nocturnal awakenings, while maintaining full
consciousness [84]. It is likely most useful to con-
sider NES and SRED as related disorders that exist
at opposite ends of a continuum of awareness
during nocturnal eating [85]. The prevalence of
SRED is estimated to be 1% to 5% in the general
population, two to four times more common in
females, and tends to have onset in late adolescence
or early adulthood, although patients often do not
present until many years after they first develop
symptoms [86]. Patients with SRED may have a his-
tory of sleepwalking; however, once eating behav-
iors become established, they tend to replace any
other distinct sleepwalking behaviors. The patho-
physiology of SRED and NES is unclear; however,
it has been hypothesized that nocturnal eating
may be caused by abnormal coordination of hor-
mones regulating appetite and the sleep-wake cycle
[87,88].
SRED seems to be more common in patients with
a daytime eating disorder (eg, anorexia nervosa, bu-
limia nervosa); however, most patients with SRED
do not have an eating disorder that manifests while
awake [86,89]. It is noteworthy, however, that the
prevalence estimates of NES are as high as 12.3%
in psychiatric outpatients [90]. Approximately one
third of patients with SRED have a first-degree
family member with nocturnal eating, similar to
familial patterns observed in sleepwalking and
Parasomnias: Psychiatric Considerations 225
certain eating disorders [91,92]. Polysomnographic
studies have demonstrated frequent arousals from
SWS, similar to other non-REM parasomnias, but
eating episodes can emerge at any time of night
and from all states of non-REM sleep [82,83].
Occasionally, SRED may be secondary to condi-
tions that cause arousal from sleep (eg, obstructive
sleep apnea, periodic limb movements of sleep), in
which case PSG may be helpful to rule out such dis-
orders [93]. Restless legs syndrome is often comor-
bid with SRED, and treatment of RLS with
dopamine agonists may diminish the nocturnal eat-
ing behaviors [89]. Also, many psychotropic medi-
cations may induce the disorder in susceptible
individuals, including benzodiazepine-receptor ag-
onists (eg, zolpidem) and atypical antipsychotics
(eg, risperidone, olanzapine) [94–97].
Treatment of SRED is similar to other non-REM
parasomnias, with some notable exceptions. Be-
sides avoidance of sleep deprivation and maintain-
ing the safety of the sleeping environment,
normalization of a daytime eating schedule is im-
portant. Pharmacotherapy is typically tailored to
the individual patient. Those with a history of sleep-
walking can be given short- to intermediate-acting
benzodiazepines or other non–appetite-stimulat-
ing sedatives (eg, trazodone); however, this may
have the paradoxical effect of worsening eating be-
haviors and amnesia. Antidepressants including
selective serotonin reuptake inhibitors and buprop-
rion may also be of benefit in some patients
[98,99]. There is growing evidence that topiramate,
an antiepileptic medication, may be the treatment
of choice in SRED because numerous open-label
case series have shown it to be efficacious in de-
creasing frequency of nocturnal eating, often with
decrease in weight [100–102].
Sleep-related hallucinations
Sleep-related hallucinations include hypnagogic
and hypnopompic hallucinations, and complex
nocturnal visual hallucinations [1]. The latter is
a less common syndrome in which the individual
experiences visual hallucinations, often of a person
or animal, after full awakening from sleep, which
may remain present for several minutes, but usually
disappears if illumination is increased. Although
these may be related to neuropsychiatric disease
(eg, dementia with Lewy bodies), there are individ-
uals who experience complex nocturnal visual hal-
lucinations in whom anxiety is the only associated
feature [103].
Hypnagogic and hypnopompic hallucinations
are classically visual, but may include auditory, tac-
tile, or kinetic hallucinations that occur at the onset
of sleep or on awakening. Although the precise
pathophysiology is unknown, it is thought that
226 Plante & Winkelman
they may represent REM dreaming intruding into
wakefulness. This may account for the observation
that these hallucinations often co-occur with SP,
the combination of which can be quite distressing
to patients. Hypnagogic and hypnopompic halluci-
nations are common in the general population,
with prevalence estimates of 37% and 12.5%, re-
spectively [104]. Individuals with mood, anxiety,
and psychotic disorders experience hypnagogic-
pompic hallucinations at greater rates than the gen-
eral population, but they are neither sensitive nor
specific for mental illness [104]. Because hypnago-
gic-pompic hallucinations are more frequently
seen in psychotic disorders (eg, schizophrenia)
than other psychiatric disorders, when encountered
in a clinical setting, practitioners should inquire
about other signs of psychosis (eg, daytime halluci-
nations, delusions, and so forth) and refer to psy-
chiatrists when appropriate.
Sleep-related dissociative disorders
SRDDs have been included as a parasomnia in the
ICSD-2, reflecting a shift from previous nosology,
and merit discussion [1]. SRDD occurs when an
individual with a dissociative disorder (eg, dissocia-
tive amnesia, dissociative fugue, dissociative iden-
tity disorder [formerly multiple personality
disorder], and depersonalization disorder) has noc-
turnal episodes of dissociation. Although the path-
ophysiology of dissociative disorders is not known,
they are thought to arise when normally integrated
functions of consciousness, memory, identity, or
perception are separated (ie, dissociated) from
one another [72]. Dissociative disorders are more
predominant in women, and often (although not
always) there is a history of past trauma. Nocturnal
dissociative episodes occur during electroencepha-
logram established wakefulness, and can occur at
varied intervals from sleep, ranging from the transi-
tion from sleep to wakefulness to several minutes
after achieving wakefulness. Myriad behaviors
may manifest during these episodes that may be
complex, violent, self-mutilating, abuse re-enact-
ments, or fugue [105]. Prevalence is not known;
however, 7 out of 100 consecutive patients referred
to a sleep disorders center for sleep-related injury
were found to have SRDD [106]. Nocturnal
episodes of dissociation may occur in roughly one
quarter of patients with dissociative disorders
[107]. Although nocturnal dissociative episodes
tend to arise from more clearly established wakeful-
ness than other disorders previously discussed,
their undesirable behavior and relatedness to sleep
challenge the understanding of parasomnias as
a group of disorders [108]. Treatment of nocturnal
dissociation is typically directed at the underlying
dissociative illness. PSG is not routinely indicated,
but may be indicated to rule out frontal lobe sei-
zures, RBD, or more clearly delineate a nocturnal
dissociative episode from a non-REM parasomnia.
Summary
Parasomnias represent undesirable experiences or
behavior that arise from sleep but are not fully un-
der voluntary control. As a group of disorders, they
challenge the arbitrary separation between disci-
plines, and reflect the interdisciplinary nature of
sleep medicine. Appreciation of the psychiatric
aspects of parasomnias is necessary by those who
manage these disorders because they are often
comorbid with mental illness, may present with
psychiatric complaints, may be induced by psy-
chotropic medications, and are often effectively
managed with psychopharmocologic and/or psy-
chotherapeutic interventions.
References
[1] American Academy of Sleep Medicine. Interna-
tional classification of sleep disorders: diagnostic
and coding manual. 2nd Edition. Westchester
(IL): American Academy of Sleep Medicine;
2005.
[2] Nunberg H, Federn E. Minutes of the Vienna
Psychoanalytic Society volume I: 1906–1908.
New York: International Universities Press,
Inc.,; 1962.
[3] Mahowald MW, Schenck CH. Non-rapid eye
movement sleep parasomnias. Neurol Clin
2005;23:1077–106.
[4] Mahowald MW, Bornemann MA. NREM sleep-
arousal parasomnias. In: Kryger MH, Roth T,
Dement WC, editors. Principles and practice
of sleep medicine. 4th edition. Philadelphia
(PA): Elsevier Saunders; 2005. p. 889–96.
[5] Schenck CH, Mahowald MW. Parasomnias:
managing bizarre sleep-related behavior disor-
ders. Postgrad Med 2000;107:145–56.
[6] Hublin C, Kaprio J, Partinen M, et al. Parasom-
nias: co-occurrence and genetics. Psychiatr
Genet 2001;11:65–70.
[7] Ohayon MM, Guilleminault C, Priest RG. Night
terrors, sleepwalking, and confusional arousals
in the general population: their frequency and
relationship to other sleep and mental disor-
ders. J Clin Psychiatry 1999;60:268–76 [quiz:
277].
[8] Roth B, Nevsimalova S, Rechtschaffen A. Hyper-
somnia with sleep drunkenness. Arch Gen
Psychiatry 1972;26:456–62.
[9] Tassi P, Muzet A. Sleep inertia. Sleep Med Rev
2000;4:341–53.
[10] Crisp AH. The sleepwalking/night terrors syn-
drome in adults. Postgrad Med J 1996;72:
599–604.
[11] Kavey NB, Whyte J, Resor SR Jr, et al. Somnam-
bulism in adults. Neurology 1990;40:749–52.

[12] Laberge L, Tremblay RE, Vitaro F, et al. Develop-
ment of parasomnias from childhood to early
adolescence. Pediatrics 2000;106:67–74.
[13] Hublin C, Kaprio J. Genetic aspects and genetic
epidemiology of parasomnias. Sleep Med Rev
2003;7:413–21.
[14] Lecendreux M, Bassetti C, Dauvilliers Y, et al.
HLA and genetic susceptibility to sleepwalking.
Mol Psychiatry 2003;8:114–7.
[15] Schenck CH, Mahowald MW. On the reported
association of psychopathology with sleep
terrors in adults. Sleep 2000;23:448–9.
[16] Gau SF, Soong WT. Psychiatric comorbidity of
adolescents with sleep terrors or sleepwalking:
a case-control study. Aust N Z J Psychiatry
1999;33:734–9.
[17] Hartman D, Crisp AH, Sedgwick P, et al. Is there
a dissociative process in sleepwalking and night
terrors? Postgrad Med J 2001;77:244–9.
[18] Landry P, Warnes H, Nielsen T, et al. Somnam-
bulistic-like behaviour in patients attending
a lithium clinic. Int Clin Psychopharmacol
1999;14:173–5.
[19] Joncas S, Zadra A, Paquet J, et al. The value of
sleep deprivation as a diagnostic tool in adult
sleepwalkers. Neurology 2002;58:936–40.
[20] Kales A, Soldatos CR, Bixler EO, et al. Heredi-
tary factors in sleepwalking and night terrors.
Br J Psychiatry 1980;137:111–8.
[21] Mehlenbeck R, Spirito A, Owens J, et al. The clin-
ical presentation of childhood partial arousal
parasomnias. Sleep Med 2000;1:307–12.
[22] Shapiro CM, Trajanovic NN, Fedoroff JP. Sex-
somnia: a new parasomnia? Can J Psychiatry
2003;48:311–7.
[23] Ebrahim IO. Somnambulistic sexual behaviour
(sexsomnia). J Clin Forensic Med 2006;13:
219–24.
[24] Schenck CH, Arnulf I, Mahowald MW. Sleep
and sex: what can go wrong? A review of the lit-
erature on sleep related disorders and abnormal
sexual behaviors and experiences. Sleep 2007;
30:683–702.
[25] Cartwright R. Sleepwalking violence: a sleep
disorder, a legal dilemma, and a psychological
challenge. Am J Psychiatry 2004;161:1149–58.
[26] Pressman MR. Disorders of arousal from sleep
and violent behavior: the role of physical con-
tact and proximity. Sleep 2007;30:1039–47.
[27] Bonkalo A. Impulsive acts and confusional
states during incomplete arousal from sleep:
criminological and forensic implications.
Psychiatr Q 1974;48:400–9.
[28] Pressman MR, Mahowald MW, Schenck CH, et al.
Alcohol-induced sleepwalking or confusional
arousal as a defense to criminal behavior: a review
of scientific evidence, methods and forensic
considerations. J Sleep Res 2007;16:198–212.
[29] Broughton RJ. Sleep disorders: disorders of
arousal? Enuresis, somnambulism, and night-
mares occur in confusional states of arousal,
Parasomnias: Psychiatric Considerations 227
not in dreaming sleep. Science 1968;159:
1070–8.
[30] Espa F, Ondze B, Deglise P, et al. Sleep architec-
ture, slow wave activity, and sleep spindles in
adult patients with sleepwalking and sleep ter-
rors. Clin Neurophysiol 2000;111:929–39.
[31] Gaudreau H, Joncas S, Zadra A, et al. Dynamics
of slow-wave activity during the NREM sleep of
sleepwalkers and control subjects. Sleep 2000;
23:755–60.
[32] Llorente MD, Currier MB, Norman SE, et al.
Night terrors in adults: phenomenology and
relationship to psychopathology. J Clin Psychi-
atry 1992;53:392–4.
[33] Craske MG, Tsao JC. Assessment and treatment
of nocturnal panic attacks. Sleep Med Rev 2005;
9:173–84.
[34] Schenck CH, Mahowald MW. Long-term,
nightly benzodiazepine treatment of injurious
parasomnias and other disorders of disrupted
nocturnal sleep in 170 adults. Am J Med
1996;100:333–7.
[35] Ebrahim IO, Peacock KW. REM sleep behavior
disorder–psychiatric presentations: a case series
from the United Kingdom. J Clin Sleep Med
2005;1:43–7.
[36] Fantini ML, Ferini-Strambi L, Montplaisir J. Id-
iopathic REM sleep behavior disorder: toward
a better nosologic definition. Neurology 2005;
64:780–6.
[37] Schenck CH, Mahowald MW. REM sleep behav-
ior disorder: clinical, developmental, and
neuroscience perspectives 16 years after its
formal identification in SLEEP. Sleep 2002;25:
120–38.
[38] Boeve BF, Silber MH, Parisi JE, et al. Synuclein-
opathy pathology and REM sleep behavior
disorder plus dementia or parkinsonism.
Neurology 2003;61:40–5.
[39] Schenck CH, Bundlie SR, Mahowald MW. REM
behavior disorder (RBD): delayed emergence of
parkinsonism and/or dementia in 65% of older
men initially diagnosed with idiopathic RBD,
and an analysis of the minimum and maximum
tonic and/or phasic electromyographic abnor-
malities found during REM sleep. Sleep 2003;
26:A316.
[40] Jouvet M, Delorme F. Locus coeruleus et som-
meil paradoxal. Comptes Rendus de la Société
de Biologie (Paris) 1965;159:895–9.
[41] Turner RS, D’Amato CJ, Chervin RD, et al. The
pathology of REM sleep behavior disorder
with comorbid Lewy body dementia. Neurol-
ogy 2000;55:1730–2.
[42] Eisensehr I, Linke R, Noachtar S, et al. Reduced
striatal dopamine transporters in idiopathic
rapid eye movement sleep behaviour disorder:
comparison with Parkinson’s disease and
controls. Brain 2000;123(Pt 6):1155–60.
[43] Albin RL, Koeppe RA, Chervin RD, et al. De-
creased striatal dopaminergic innervation in
228 Plante & Winkelman
REM sleep behavior disorder. Neurology 2000;
55:1410–2.
[44] Rye DB. Contributions of the pedunculopon-
tine region to normal and altered REM sleep.
Sleep 1997;20:757–88.
[45] Gagnon JF, Fantini ML, Bedard MA, et al. Asso-
ciation between waking EEG slowing and REM
sleep behavior disorder in PD without demen-
tia. Neurology 2004;62:401–6.
[46] Postuma RB, Lang AE, Massicotte-Marquez J,
et al. Potential early markers of Parkinson dis-
ease in idiopathic REM sleep behavior disorder.
Neurology 2006;66:845–51.
[47] Ferini-Strambi L, Di Gioia MR, Castronovo V,
et al. Neuropsychological assessment in idio-
pathic REM sleep behavior disorder (RBD):
does the idiopathic form of RBD really exist?
Neurology 2004;62:41–5.
[48] Ferini-Strambi L, Oldani A, Zucconi M, et al.
Cardiac autonomic activity during wakefulness
and sleep in REM sleep behavior disorder. Sleep
1996;19:367–9.
[49] Pinninti NR, Rissmiller D. Incidence of re-
straint-related deaths. Psychiatr Serv 2001;52:
975.
[50] Winkelman JW, James L. Serotonergic antide-
pressants are associated with REM sleep without
atonia. Sleep 2004;27:317–21.
[51] Schenck CH, Mahowald MW. Rapid eye move-
ment sleep parasomnias. Neurol Clin 2005;23:
1107–26.
[52] Sadock BJ, Sadock VA. Emergency psychiatric
medicine. In: Cancro R, editor. Kaplan and
Sadock’s synopsis of psychiatry: behavioral
sciences, clinical psychiatry. 9th edition. Phila-
delphia (PA): Lippincott, Williams & Wilkins;
2003. p. 901–22.
[53] Boeve BF, Silber MH, Ferman TJ. Melatonin for
treatment of REM sleep behavior disorder in
neurologic disorders: results in 14 patients.
Sleep Med 2003;4:281–4.
[54] Fantini ML, Gagnon JF, Filipini D, et al. The ef-
fects of pramipexole in REM sleep behavior
disorder. Neurology 2003;61:1418–20.
[55] Girard TA, Cheyne JA. Timing of spontaneous
sleep-paralysis episodes. J Sleep Res 2006;15:
222–9.
[56] Takeuchi T, Fukuda K, Sasaki Y, et al. Factors
related to the occurrence of isolated sleep paral-
ysis elicited during a multi-phasic sleep-wake
schedule. Sleep 2002;25:89–96.
[57] Cheyne JA. Situational factors affecting sleep
paralysis and associated hallucinations: posi-
tion and timing effects. J Sleep Res 2002;11:
169–77.
[58] Ohayon MM, Zulley J, Guilleminault C, et al.
Prevalence and pathologic associations of sleep
paralysis in the general population. Neurology
1999;52:1194–200.
[59] Hinton DE, Pich V, Chhean D, et al. The ghost
pushes you down: sleep paralysis-type panic
attacks in a Khmer refugee population. Trans-
cult Psychiatry 2005;42:46–77.
[60] Hinton DE, Hufford DJ, Kirmayer LJ. Culture
and sleep paralysis. Transcult Psychiatry 2005;
42:5–10.
[61] Guilleminault C, Fromherz S. Narcolepsy: diag-
nosis and management. In: Kryger MH, Roth T,
Dement WC, editors. Principles and practice of
sleep medicine. 4th edition. Philadelphia (PA):
Elsevier Saunders; 2005. p. 780–90.
[62] Koran LM, Raghavan S. Fluoxetine for isolated
sleep paralysis. Psychosomatics 1993;34:184–7.
[63] Mitchell SA, Black MJ. Freud and beyond:
a history of modern psychoanalytic thought.
New York: Basic Books; 1995.
[64] Freud SF. Third lecture. In: Strachey J, editor.
Five lectures on psycho-analysis. New York:
W.W. Norton & Company; 1961. p. 28–41.
[65] Pagel JF. Nightmares and disorders of dreaming.
Am Fam Physician 2000;61:2037–42, 2044.
[66] Zadra A, Donderi DC. Nightmares and bad
dreams: their prevalence and relationship to
well-being. J Abnorm Psychol 2000;109:
273–81.
[67] Ohayon MM, Morselli PL, Guilleminault C.
Prevalence of nightmares and their relation-
ship to psychopathology and daytime func-
tioning in insomnia subjects. Sleep 1997;20:
340–8.
[68] Nielsen TA, Laberge L, Paquet J, et al. Develop-
ment of disturbing dreams during adolescence
and their relation to anxiety symptoms. Sleep
2000;23:727–36.
[69] Agargun MY, Cilli AS, Kara H, et al. Repetitive
and frightening dreams and suicidal behavior
in patients with major depression. Compr
Psychiatry 1998;39:198–202.
[70] Tanskanen A, Tuomilehto J, Viinamaki H, et al.
Nightmares as predictors of suicide. Sleep 2001;
24:844–7.
[71] Levin R, Fireman G. Nightmare prevalence, night-
mare distress, and self-reported psychological
disturbance. Sleep 2002;25:205–12.
[72] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
Edition. Washington, D.C: American Psychiat-
ric Association; 2000. Text Revision.
[73] Harvey AG, Jones C, Schmidt DA. Sleep and
posttraumatic stress disorder: a review. Clin
Psychol Rev 2003;23:377–407.
[74] Ohayon MM, Shapiro CM. Sleep disturbances
and psychiatric disorders associated with post-
traumatic stress disorder in the general popula-
tion. Compr Psychiatry 2000;41:469–78.
[75] Mellman TA, David D, Kulick-Bell R, et al. Sleep
disturbance and its relationship to psychiatric
morbidity after Hurricane Andrew. Am J Psychi-
atry 1995;152:1659–63.
[76] Raskind MA, Peskind ER, Kanter ED, et al. Re-
duction of nightmares and other PTSD symp-
toms in combat veterans by prazosin:

a placebo-controlled study. Am J Psychiatry
2003;160:371–3.
[77] Berlant JL. Prospective open-label study of
add-on and monotherapy topiramate in civil-
ians with chronic nonhallucinatory posttrau-
matic stress disorder. BMC Psychiatry 2004;4:
24.
[78] Ahearn EP, Krohn A, Connor KM, et al. Pharma-
cologic treatment of posttraumatic stress disor-
der: a focus on antipsychotic use. Ann Clin
Psychiatry 2003;15:193–201.
[79] Krakow B, Hollifield M, Johnston L, et al. Imag-
ery rehearsal therapy for chronic nightmares in
sexual assault survivors with posttraumatic
stress disorder: a randomized controlled trial.
JAMA 2001;286:537–45.
[80] Hauri PJ, Silber MH, Boeve BF. The treatment of
parasomnias with hypnosis: a 5-year follow-up
study. J Clin Sleep Med 2007;3:369–73.
[81] Barrett J, Underwood A. Perchance to eat?
Newsweek 2006;147:54.
[82] Winkelman JW. Clinical and polysomno-
graphic features of sleep-related eating disorder.
J Clin Psychiatry 1998;59:14–9.
[83] Schenck CH, Hurwitz TD, Bundlie SR, et al.
Sleep-related eating disorders: polysomno-
graphic correlates of a heterogeneous syndrome
distinct from daytime eating disorders. Sleep
1991;14:419–31.
[84] Marshall HM, Allison KC, O’Reardon JP, et al.
Night eating syndrome among nonobese
persons. Int J Eat Disord 2004;35:217–22.
[85] Winkelman JW. Sleep-related eating disorder
and night eating syndrome: sleep disorders,
eating disorders, or both? Sleep 2006;29:876–7.
[86] Winkelman JW, Herzog DB, Fava M. The preva-
lence of sleep-related eating disorder in psychi-
atric and non-psychiatric populations. Psychol
Med 1999;29:1461–6.
[87] Birketvedt GS, Florholmen J, Sundsfjord J, et al.
Behavioral and neuroendocrine characteristics
of the night-eating syndrome. JAMA 1999;282:
657–63.
[88] Allison KC, Ahima RS, O’Reardon JP, et al.
Neuroendocrine profiles associated with energy
intake, sleep, and stress in the night eating
syndrome. J Clin Endocrinol Metab 2005;90:
6214–7.
[89] Schenck CH, Hurwitz TD, O’Connor KA, et al.
Additional categories of sleep-related eating
disorders and the current status of treatment.
Sleep 1993;16:457–66.
[90] Lundgren JD, Allison KC, Crow S, et al.
Prevalence of the night eating syndrome in a psy-
chiatric population. Am J Psychiatry 2006;163:
156–8.
[91] Hublin C, Kaprio J, Partinen M, et al. Prevalence
and genetics of sleepwalking: a population-
based twin study. Neurology 1997;48:177–81.
Parasomnias: Psychiatric Considerations 229
[92] Klump KL, Kaye WH, Strober M. The evolving
genetic foundations of eating disorders.
Psychiatr Clin North Am 2001;24:215–25.
[93] Howell MJ, Schenck C, Crow SJ. Curbing noc-
turnal binges in sleep-related eating disorder.
Current Psychaitry 2007;6:19–24.
[94] Schenck CH, Connoy DA, Casetllanos M, et al.
Zolpidem-induced amnestic sleep-related
eating disorder (SRED) in 19 patients. Sleep
2005;28:A259.
[95] Morgenthaler TI, Silber MH. Amnestic sleep-
related eating disorder associated with zolpi-
dem. Sleep Med 2002;3:323–7.
[96] Lu ML, Shen WW. Sleep-related eating disorder
induced by risperidone. J Clin Psychiatry 2004;
65:273–4.
[97] Paquet V, Strul J, Servais L, et al. Sleep-related
eating disorder induced by olanzapine. J Clin
Psychiatry 2002;63:597.
[98] O’Reardon JP, Allison KC, Martino NS, et al. A
randomized, placebo-controlled trial of sertra-
line in the treatment of night eating syndrome.
Am J Psychiatry 2006;163:893–8.
[99] Schenck CH, Mahowald MW. Combined bupro-
pion-levodopa-trazodone therapy of sleep-re-
lated eating and sleep disruption in two adults
with chemical dependency. Sleep 2000;23:587–8.
[100] Winkelman JW. Efficacy and tolerability of
open-label topiramate in the treatment of
sleep-related eating disorder: a retrospective
case series. J Clin Psychiatry 2006;67:1729–34.
[101] Winkelman JW. Treatment of nocturnal eating
syndrome and sleep-related eating disorder
with topiramate. Sleep Med 2003;4:243–6.
[102] Schenck C, Mahowald M. Topiramate therapy
of sleep related eating disorder (SRED). Sleep
2006;29:A268.
[103] Silber MH, Hansen MR, Girish M. Complex
nocturnal visual hallucinations. Sleep Med
2005;6:363–6.
[104] Ohayon MM, Priest RG, Caulet M, et al. Hypna-
gogic and hypnopompic hallucinations: patho-
logical phenomena? Br J Psychiatry 1996;169:
459–67.
[105] Schenck CH, Milner DM, Hurwitz TD, et al.
Dissociative disorders presenting as somnam-
bulism: Polysomnographic video and clinical
documentation (8 cases). Dissociation 1989;2:
194–204.
[106] Schenck CH, Milner DM, Hurwitz TD, et al. A
polysomnographic and clinical report on
sleep-related injury in 100 adult patients. Am J
Psychiatry 1989;146:1166–73.
[107] Agargun M, Kara H, Ozer O, et al. Characteris-
tics of patients with nocturnal dissociative dis-
orders. Sleep and Hypnosis 2001;3:131–4.
[108] Mahowald MW, Schenck CH. Nocturnal disso-
ciation: awake? asleep? both? or neither? Sleep
and Hypnosis 2001;3:129–30.
 231

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 231–249

Sleep in Mood Disorders

a a,b,
Michael J. Peterson, MD, PhD , Ruth M. Benca, MD, PhD *

- Epidemiology Advanced sleep electroencephalogram

- Classification and diagnosis of mood analysis
disorders Power spectral analysis
Prevalence and types of sleep problems Automated slow wave analysis
in mood disorders Coherence of sleep electroencephalogram
Epidemiology of sleep disturbance and rhythms
mood disorders Topography of sleep
Predictive value and specificity of sleep electroencephalogram activity
disturbance for mood disorders - Biologic mechanisms of sleep changes
Interepisodic persistence of subjective in mood disorders
sleep disruption Neurotransmitters
Psychiatric comorbidities associated with Neuroimaging
sleep and mood disorders Endocrine changes
Comorbidity of primary sleep disorders Genetic polymorphisms
and mood disorders - Treatment of sleep disturbance and mood
- Sleep findings in mood disorders disorders
Subjective sleep complaints Sleep-deprivation therapy
Polysomnographic and architectural sleep Sleep loss and bipolar disorder
changes Clinical use of polysomnography
- Summary
- References

Sleep disturbances are among the most com-
mon symptoms in patients with acute episodes
of mood disorders, and patients with mood disor-
ders exhibit higher rates of sleep disturbances than
the general population, even during periods of re-
mission. Insomnia and hypersomnia are associ-
ated with an increased risk for the development
or recurrence of mood disorders, and increased se-
verity of psychiatric symptoms. Similarly, primary
sleep disorders, such as obstructive sleep apnea
(OSA) and restless legs syndrome (RLS), are also
associated with an increased incidence of depres-
sion. Sleep electroencephalogram (EEG) record-
ings have identified objective abnormalities
associated with mood disorders, providing insight
into the neurobiologic relationships between
mood and sleep. Future studies will continue
to investigate this association, and potentially

This work was supported by NIH Roadmap Interdisciplinary Award T32 MH75880 (MJP), and a NARSAD Young
Investigator Award (MJP).
a
Department of Psychiatry, University of Wisconsin-Madison, 6001 Research Park Boulevard, Madison, WI
53719, USA
b
University of Wisconsin Sleep Program, 6001 Research Park Boulevard, Madison, WI, USA
* Corresponding author. Department of Psychiatry, University of Wisconsin-Madison, 6001 Research Park
Boulevard, Madison, WI 53719.
E-mail address: rmbenca@wisc.edu (R.M. Benca).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.009
sleep.theclinics.com
232 Peterson & Benca

improve the treatment of both sleep and mood
disorders.
Epidemiology
Classification and diagnosis of mood
disorders
Sleep complaints are one of the most consistent
symptoms associated with mood disorders. Disrup-
tions of typical sleep patterns are a core diagnostic
criterion of mood episodes in the Diagnostic and sta-
tistical manual of mental disorders-IV-TR (Tables 1
and 2) [1], reflecting their importance and preva-
lence in the presentation of these disorders. Mood
disorders are among the most common categories
of psychiatric diagnoses, second only to anxiety
disorders. They are responsible for tremendous
socioeconomic costs worldwide, including eventual
suicide in 15% of persons with major depression, in-
creased morbidity and mortality from other illnesses,
and economic impacts from associated disability
(http://www.who.int/topics/depression/en/).
Mood disorders are diagnosed based on the
pattern of depressive and/or manic episodes (see
Table 1). Major depressive disorder is characterized
by the occurrence of one or more episodes of major
depression. Patients with a bipolar disorder diagno-
sis must have at least one manic, hypomanic, or
mixed episode, and frequently have episodes of
major depression. Major depressive or manic epi-
sodes are categorized as psychotic or nonpsychotic,
based on the presence or absence of delusions and
hallucinations. Depressive episodes, in both major
depression and bipolar disorder, can have

Table 1: Diagnostic and Statistical Manual of Mental Disorders-IV-TR mood disorders and subtypesa
Mood disorder Clinical features
Major depression Either single or recurrent MDEs. No periods of mania or hypomania.
Bipolar I At least one manic episode. Recurrent MDEs are typical, but not
formally required for DSM diagnosis.
Dysthymia Persistent depressed mood and symptoms that are not severe enough
for diagnosing an MDE. Symptoms must be present most of the time
for at least 2 years.
Bipolar II At least one hypomanic episode and one MDE. Typical course consists
of recurrent MDEs and some hypomanias.
Cyclothymia Recurrent hypomanic episodes and recurrent or chronic depressive
symptoms. Depressive symptoms not severe enough to meet MDE criteria.
Subtype Clinical features
Atypical Clinical picture of depression characterized by mood reactivity
(mood brightens in response to positive events), and at least
two of the following features:
Significant weight gain or increase in appetite
Hypersomnia
Heavy, leaden feeling in arms or legs
Long-standing pattern of interpersonal rejection sensitivity
(not limited to depressive episodes) that results in significant
social or occupational impairment
Melancholic Clinical picture of depression characterized by a lack of pleasure in almost all
activities or does not feel better, even temporarily, when something good
happens. Also includes at least three of the following:
Depression distinct from feeling after significant loss
(eg, death of a loved one)
Depression usually worse in the morning (diurnal pattern of symptoms)
Early morning awakening (at least 2 h before usual time of awakening)
Marked psychomotor retardation or agitation
Significant anorexia or weight loss
Excessive or inappropriate guilt
Seasonal Periods of depression consistently recur at the same time of year (most often
fall and winter), unrelated to nonseasonal factors (eg, being unemployed
each winter). Full remissions of depression (or switch to (mania or
hypomania) also occur at a characteristic time of year (eg, spring). This
pattern must be consistently present for at least the last 2 years.

Abbreviations: DSM, Diagnostic and Statistical Manual of Mental Disorders; MDE, major depressive episodes.
a
Sleep-related criteria are in bold.
 Sleep in Mood Disorders 233

Table 2: Diagnostic and Statistical Manual of Mental Disorders-IV-TR diagnostic criteria for mood
episodesa
Type Criteria
Major depressive episode Depressed mood for most of the day or decreased interest
or pleasure in almost all activities, present nearly every day
for at least 2 weeks. Must include at least three or more of
the following symptoms:
1. Significant weight loss (without dieting) or weight gain
(>5% in a month), or decrease or increase in appetite
2. Insomnia or hypersomnia
3. Psychomotor agitation or retardation
4. Fatigue or loss of energy
5. Feelings of worthlessness or excessive or inappropriate guilt
6. Diminished ability to think or concentrate, or indecisiveness
7. Recurrent thoughts of death (not just fear of dying), suicidal
thoughts (with or without a plan), or a suicide attempt
Manic episode A distinct period of abnormally and persistently elevated,
expansive, or irritable mood, lasting at least 1 week
(or shorter if hospitalization necessary). Must include
at least three or more of the following symptoms:
1. Inflated self-esteem or grandiosity
2. Decreased need for sleep (eg, feels rested after only 3 h sleep)
3. More talkative than usual or pressure to keep talking
4. Flight of ideas or subjective experience that thoughts are racing
5. Distractibility
6. Increased activity (either socially, at work or school, or sexually)
or psychomotor agitation
7. Excessive involvement in pleasurable activities without considering
risks (eg, unrestrained buying sprees, sexual indiscretions, driving
too fast, and so forth)
a
Sleep-related criteria are in bold.

a seasonal pattern, with a typical onset in the fall or
winter and remission (or even hypomania) in the
spring, suggesting a correlation with diurnal pat-
terns of light exposure (see Table 1). Major depres-
sion has been further subcategorized based on
common symptom clusters (see Table 1). Current
diagnostic subtypes include melancholic and atypi-
cal depression, and both involve distinct features
related to sleep patterns. Interestingly, patients
with melancholic depression are more likely to
have an improvement of depression with total sleep
deprivation, further suggesting that sleep patterns
may define biologic subtypes of depression.
amounts of sleep. During manic periods, patients
usually report reductions in total sleep, often with
a sense of decreased sleep need. Similar to major
depression, patients with bipolar disorder may
also report insomnia during a depressed episode
[4]. Hypersomnia, generally defined as daily sleep
in excess of 9 to 10 hours and reported in 3% to
8% of the general population [5,6], is also com-
monly reported in major depression with atypical
features, a seasonal pattern, or in bipolar disorders.
A subset of patients may have complaints of both
insomnia and hypersomnia, perhaps suggesting
a more severe pathophysiology [2,6].

Prevalence and types of sleep problems
in mood disorders
Complaints of sleep disruptions are common both
preceding and during major depressive or manic
episodes, and may even persist during remission.
From 65% to 75% of both adults and children
and adolescents with depression reported insom-
nia, hypersomnia, or both compared with healthy
controls [2,3]. Other common complaints include
more frequent nocturnal awakenings, nonrestora-
tive sleep, disturbing dreams, or decreased total
Epidemiology of sleep disturbance and mood
disorders
The relationship between sleep and mood is bidi-
rectional: sleep complaints are more common in
people with mood disorders than in the general
population, and mood disorders are more common
in those with sleep complaints. Population surveys
indicated that adults with insomnia were up to nine
times more likely to have major depression at the
time of the interview than those without insomnia
[7,8]. Additionally, the degree and duration of
234 Peterson & Benca
insomnia were positively correlated with more se-
vere or recurrent major depression [7]. In a survey
of the general population, those with insomnia at
baseline and at a 1-year follow-up had a relative
risk of 39.8 for major depression compared with
those without insomnia [5].
The correlation between sleep and mood seems
to be even higher in medical populations. In outpa-
tients at general medical clinics, sleep and fatigue
symptoms had the highest positive predictive value
for major depression (61% and 69%, respectively
[9]). Over three quarters of patients with insomnia
evaluated in a multicenter study also met criteria for
psychiatric disorders, including mood disorders
[10]. The prevalence of major depression in patients
with insomnia in general medical (31%) [11] or
sleep center clinics (32.3%) [12] is much higher
than patients without insomnia (4%).
Predictive value and specificity of sleep
disturbance for mood disorders
Remarkably, sleep problems even without concur-
rent psychiatric diagnoses are predictive of future
mood disorders. Young adults with a history of
insomnia, hypersomnia, or both showed a 10- to
20-fold increase in the lifetime prevalence of major
depression compared with those with no sleep
problems [6]. In a survey of adolescents (13–
16 years old) with both insomnia and a psychiatric
diagnosis, onset of insomnia more commonly pre-
ceded depression, whereas anxiety disorders more
often preceded insomnia, suggesting distinct, direc-
tional associations [13]. Insomnia or hypersomnia
are also associated with an increased risk of a new
onset of major depressive or manic episodes
[2,14]. Subjects reporting insomnia at an initial as-
sessment were 2 to 5.4 times more likely than those
without insomnia to develop major depression dur-
ing follow-up periods of 3.5 to 34 years [6,15]. Sim-
ilarly, hypersomnia was associated with a relative
risk of 2.9 for developing major depression [6]. No-
tably, two of these studies were longitudinal studies
of young adults [6,15], demonstrating that insom-
nia at a young age conferred a lifetime risk for devel-
oping mood disorders.
The presence of disruptive or stressful life events
in the preceding 4 months was also predictive of in-
creased time awake after sleep onset, but only for
subjects with a previous history of depression and
not healthy controls [16]. Women with more dis-
rupted sleep also had more depression both ante-
partum and postpartum, and their newborns
perhaps also had more disrupted sleep and less
deep sleep [17,18]. The presence of initial insomnia
(delayed sleep-onset latency) seemed to be the
most relevant screening question for identifying
women at risk for postpartum depression [18],
which corresponds to findings in the aforemen-
tioned studies.
Insomnia and fatigue were the most frequently
reported symptoms preceding a recurrent depres-
sive episode [2,19]. In about half of new-onset or
recurrent depressive episodes, and in about three
quarters of manic episodes, insomnia preceded
the appearance of mood changes [14]. More omi-
nously, an increase or decrease of 3 hours or more
of sleep suggested imminent onset of a recurrent
mood episode in patients diagnosed with bipolar
disorder [20].
Interepisodic persistence of subjective sleep
disruption
Subjective reports of insomnia may improve, but
not necessarily normalize, with remission from ma-
jor depression. Insomnia is the most commonly re-
ported residual symptom during remission of
mood episodes in patients with major depression
[21] or bipolar disorder [22]. Persistence of sleep
disturbances has been shown to be predictive of in-
creased severity and recurrence of major depression
[23].
Psychiatric comorbidities associated
with sleep and mood disorders
Suicide and thoughts about suicide (suicidal idea-
tion) are common symptoms of depression with se-
rious implications. Independent of the severity of
depression, decreased sleep time, insomnia, and
particularly nightmares were predictive of suicide
attempts and suicidal ideation. Among patients
with a recent suicide attempt, almost 90% reported
some sleep disturbance, further suggesting a link
between insomnia and suicidality in depression
[24]. Although suicide and suicidal ideation are
closely linked to major depression, individuals
who had suicide attempts were more likely than de-
pressed nonattempters to report insomnia or night-
mares. Moreover, major depression combined with
insomnia confers an increased risk of suicide in
both adolescents [25] and adults [24]. This associa-
tion again emphasizes the importance of monitor-
ing insomnia (and nightmares) in patients with
or without depression at the time.
Comorbidity of primary sleep disorders
and mood disorders
Such conditions as OSA and RLS, generally consid-
ered primary sleep disorders, also seem to be more
prevalent in patients with mood disorders than the
general population [26–28]. This relationship is bi-
directional for OSA, because approximately one in
five patients with OSA also were diagnosed with
major depression, and nearly one in five patients

with major depression were diagnosed with OSA
[26]. In a recent longitudinal study, individuals
with sleep-related breathing disorders showed a se-
verity-related risk of developing depression: odds
ratios of 1.6 (minimal sleep-related breathing disor-
ders); 2 (mild sleep-related breathing disorders); to
2.6 (moderate or worse sleep-related breathing dis-
orders) [29]. Depression is also a modifier of clini-
cal course: in patients with OSA and depression,
depression is directly associated with daytime
fatigue severity, independent of OSA severity [30].
Furthermore, recent studies have shown that treat-
ing OSA results in a sustained improvement in de-
pression [31], and the degree of improvement in
OSA with continuous positive airway pressure also
correlated with the improvement in depression
[32].
The relationship between major depression and
RLS is less well defined. Several studies have, but
other studies have not, shown a correlation be-
tween the two disorders [28,33,34]. This relation-
ship is complicated by the fact that serotonergic
antidepressant medications can exacerbate RLS. At
least some studies have accounted for this, however,
such as a recent population survey that indicated
RLS is associated with depressed mood, but antide-
pressant medication use was not higher in the RLS
group than the general population [33].
Sleep findings in mood disorders
Subjective sleep complaints
Some reports suggest that there is a discrepancy be-
tween subjective and objective sleep measurements
in patients with mood disorders. For example, in
many of the depressed patients who reported sleep
complaints, no abnormalities were identified by
polysomnographic recordings [35]. Similarly, a re-
cent study of adolescents with major depression
found subjective, but not objective, sleep distur-
bances [36]. A number of studies have investigated
the underestimation or overestimation of sleep pa-
rameters, including sleep-onset latency, number of
awakenings, sleep depth, and total sleep time
[37–39]. The study designs and results have varied,
however, and there is no clear consensus on the
topic. Depressed subjects did not differ from con-
trols in their self-report accuracy in some studies,
but substantially overestimated or underestimated
sleep parameters compared to controls in others
studies. Similarly, although less studied, objective
measures of hypersomnia and daytime sleepiness
often correlate poorly with subjective reports [38].
It is important to keep in mind that insomnia,
like major depression, is a clinical diagnosis based
on the patient’s reported symptoms. Nevertheless,
Sleep in Mood Disorders 235
general trends suggest that in patients with major
depression, increased stage 1 sleep correlates with
subjective reports of poor sleep quality, and with
treatment, increased amounts of slow wave sleep
([SWS] stage 3–4) correlate with reports of ‘‘deeper’’
or more satisfying sleep [37,40]. This subjective
decrement in sleep quality is consistent with the
objective decreases in SWS seen in the polysomnog-
raphy of patients with major depression [41]. Cor-
respondingly, in an 8-week treatment response
study of patients with major depression, increases
in SWS correlated with subjective scores of sleep
depth and satisfaction [37].
Objective measures from polysomnographic re-
cordings may be better biologic markers, but based
on epidemiologic studies described previously, sub-
jective measures do seem to have a value in the clin-
ical assessment and treatment of major depression,
and may be a better indicator of treatment onset
and measure of response. This is particularly rele-
vant because reports of insomnia improve, but do
not normalize, with remission of depression. Ide-
ally, however, both subjective and objective mea-
sures are used because they relate to different
clinical aspects of sleep and mood.
Polysomnographic and architectural sleep
changes
Since the late 1960s, sleep EEG changes have been
studied extensively as biologic markers of major de-
pression. Most patients with major depression have
some objective findings of sleep disturbance [41],
but these findings are not always specific for depres-
sion and can vary with age and other factors.
Sleep architecture abnormalities in major depres-
sion can be grouped into three general categories
(Table 3) [42].
1. Sleep continuity disturbances. Patients with ma-
jor depression often show prolonged sleep-onset
latency (time from lights out to first appearance
of sleep); increased number and duration of
waking periods during sleep; and early morning
awakening. These disturbances are reflected by
increased sleep fragmentation and decreased
sleep efficiency.
2. SWS deficits. Stages 3–4 of non–rapid eye move-
ment sleep are considered SWS, colloquially re-
ferred to as ‘‘deep’’ sleep, because the arousal
threshold is high in this state. Depressed patients
often have a reduction of SWS, both as the per-
cent of total sleep and as actual minutes of
SWS. Furthermore, an abnormal temporal distri-
bution of sleep has been identified in some stud-
ies of this population. Healthy subjects have
a peak in SWS time during the first sleep cycle,
with a gradual decline through subsequent
236 Peterson & Benca
Table 3: Sleep abnormalities in depression (and mania)
Subjective complaints
Insomnia
Difficulty falling asleep (initial insomnia)
Increased awakening at night/restless sleep
(middle insomnia)
Early morning awakening (terminal insomnia)
Decreased amounts of sleep
Sleep less deep or less refreshing
Disturbing dreams
Abbreviations: REM, rapid eye movement; SWS, slow wave sleep.
cycles. Depressed subjects often have less SWS
during the first cycle, and a relative peak during
the second cycle [43]. Recently, computerized
analysis has led to further descriptions of these
patterns (discussed in the next section).
3. Rapid eye movement sleep abnormalities.
Changes in rapid eye movement sleep parame-
ters have long been thought to be the most con-
sistent and relatively specific sleep abnormalities
in major depression [41,42]. Decreased time to
the onset of rapid eye movement sleep (reduced
rapid eye movement sleep latency) is the most
commonly reported and studied sleep finding
in major depression. Other abnormalities in-
clude a prolonged first rapid eye movement
sleep period; increased rapid eye movements
during rapid eye movement sleep periods (in-
creased rapid eye movement density); and in-
creased percentage of rapid eye movement sleep.
Similar findings have also been documented in
patients with bipolar disorder with depression or
mania, although these populations have been less
well studied. During manic episodes, disrupted
sleep continuity, shortened rapid eye movement
(REM) sleep latency, and increased REM density
have been reported in polysomnographic studies
[44]. Interestingly, patients with bipolar depression
and hypersomnia did not consistently have de-
creased REM sleep latency, and also did not show
decreased sleep latency despite complaints of day-
time sleepiness; however, these results are from
a single study [45].
As with bipolar disorder, relatively few studies
have investigated sleep patterns in dysthymia
Polysomnographic findings
Sleep continuity disturbances
Prolonged sleep-onset latency
Increased wake time during sleep
Increased early morning wake time
Decreased total sleep time
SWS deficits
Decreased SWS amount
Decreased SWS percentage of total sleep
REM sleep abnormalities
Reduced REM sleep latency
Prolonged first REM sleep period
Increased REM activity (total number of eye
movements during the night)
Increased REM density (REM activity/total
REM sleep time)
Increased REM sleep percentage of total sleep
(chronic low-grade depressive symptoms [see
Table 1]), and the results have been variable.
Some studies lacked control or comparison groups
with major depression, limiting the comparability
of this small pool of studies. In general, it seems
that patients with dysthymia may have some of
the sleep findings characteristic of major depression,
but not to the same extent [46]. Although the EEG
changes described previously occur frequently in de-
pression, they are not specific to depression. Some
studies have found similar changes in REM sleep
or SWS in other psychiatric disorders, such as
schizophrenia and anxiety disorders [41].
Attempts have been made to correlate polysom-
nographic findings in major depression with spe-
cific symptoms, symptom severity, or course of
illness. Giles and colleagues [47] showed that fea-
tures of melancholic depression (see Table 1)
were more strongly associated with reduced REM
sleep latency. Another multivariate analysis identi-
fied 15 core depressive symptoms that correlated
with nine sleep variables, suggesting a biologic rela-
tionship between core mood and sleep findings [2].
Some sleep changes, including increased REM den-
sity and reduced sleep efficiency, seem more corre-
lated with illness severity and seem to normalize
with remission from depression [48]. Reduced
REM sleep latency and decreased SWS can persist
for long periods of time, even while patients are
asymptomatic [48,49]. These results lend support
to the hypothesis that sleep findings may reflect
both the ‘‘state’’ (current level of symptoms) and
‘‘trait’’ (biologic disposition) of major depression.
The severity of these persistent findings may reflect
a more pronounced biologic subtype of illness and

an increased risk of recurrence [50]. The hypothesis
that REM sleep latency is a trait marker for major
depression is also supported by evidence from fam-
ily studies; first-degree relatives of patients with ma-
jor depression tend to have shortened REM sleep
latency, whether or not they have a personal history
of depression [51].
EEG sleep variables in depression are also af-
fected by gender and age. Studies have shown that
loss of SWS in major depression was more promi-
nent in men than women, for both adults [52]
and adolescents [53]. This may suggest differing bi-
ologic effects of depression between women and
men, perhaps in part related to endocrine or devel-
opmental differences.
Age effects may also increase the depression-
related changes in sleep. Some studies have shown
clear differences between older adults (up to the
seventh decade of life) with major depression and
age-matched controls [54]. In contrast to adults,
younger patients with major depression are often
indistinguishable from age-matched controls
[55,56]. Another study suggested that no differ-
ences in SWS were identified between depressed
and control subjects until the middle of the fourth
decade of life, and these differences remained until
the seventh decade [57].
Amount of SWS is greatest in children and ado-
lescents, and it decreases to adult amounts by the
end of puberty. This shift is likely related to synaptic
pruning and maturation of the prefrontal cortex.
The maturation of endocrine systems (including
the hypothalamic-pituitary-adrenal axis) is in-
volved in this process and is also dynamically
changing during this time [42,55,56]. Overall, sleep
changes associated with mood disorders are best
characterized and most easily identified in adult
subjects, particularly compared with children and
adolescents. The relationships between gender,
age, major depression, and sleep, however, are not
yet fully elucidated.
Advanced sleep electroencephalogram
analysis
Recent advances in EEG recording, including digital
recording and increased availability of powerful
computational analyses, have provided new tools
to study sleep in relation to mood disorders. Tech-
niques include quantitative analysis of EEG activity
across a broad range of frequencies (power spectral
analysis); the automated detection and investiga-
tion of specific EEG waveforms, such as slow waves,
spindles, and REMs; and measures of EEG synchro-
nization across brain regions (synchrony and co-
herence). The use of high-density EEG, with the
application of 60 to 256 recording electrodes to
the scalp, now allows topographic analysis of
Sleep in Mood Disorders 237
sleep-related waveforms, allowing localization of
EEG activity over specific cortical regions, and the
possibility of accurate source localization to specific
brain regions involved in their generation.
Power spectral analysis
Analysis of the EEG power spectra (also known as
‘‘power density’’) allows quantification of EEG activ-
ity changes during sleep in different frequency
ranges. Power spectra for individual subjects are ex-
tremely consistent between nights [58], and also
show characteristic changes when comparing pa-
tient and control groups. An advantage of power
spectral analysis over traditional sleep scoring is
that activity can be quantified over sleep cycles or
whole nights independent of staging. Because sleep
EEG waveforms, such as slow waves, are not re-
stricted to single stages (eg, slow waves are present
in stage 2 and in SWS [stages 3 and 4]), power spec-
tral analysis within this frequency band (the ‘‘delta’’
frequency band, 0.5–4.5 Hz) allows a more com-
plete assessment of slow-wave activity ([SWA] delta
band EEG power). Total minutes of SWS for the
night are decreased in most studies of depressed pa-
tients [41]. Borbely and colleagues [59] used power
spectral analysis to characterize SWA in control and
depressed subjects, both confirming that SWS re-
ductions in major depression were correlated with
decrements in SWA and validating the use of EEG
power spectra to quantify SWS. A similar study of
bipolar depression did not identify SWA differ-
ences, however, and speculated that this result
could be caused by differences between the biologic
basis of unipolar and bipolar depression [60]. Addi-
tionally, subjects with major depression who had
an acute increase of SWA over the whole night fol-
lowing antidepressant administration were more
likely to respond to subsequent treatment [61,62].
Kupfer and colleagues [63] extended their findings
to demonstrate that delta band power was a stable
finding between acute depression and remission,
and more likely is a trait marker of major depres-
sion. Buysse and colleagues [64] did not identify
significant differences in SWA between women in
remission from depression who responded to
psychotherapy alone and either psychotherapy
nonremitters or psychotherapy and medication
(a serotonin reuptake inhibitor). They speculated
that the severity of depression, gender, and re-
sponse to psychotherapy in their subjects could ac-
count for the differences. Also, these differences
across studies highlight a difficulty inherent in psy-
chiatric research: illnesses based on patient and cli-
nician descriptions may reflect a diverse spectrum
of biologic illness. In contrast, they also reflect the
importance of objective measures, such as sleep
EEG findings, that might identify biologic subtypes
238 Peterson & Benca
of illness, and could be important in selecting study
populations.
Most research on sleep and major depression us-
ing power spectral analysis has focused on differ-
ences in the delta frequency range, because of its
relationship to slow waves. A few studies have iden-
tified differences in other frequency ranges, how-
ever, including an increase in 12 to 20 Hz power
[63] for subjects in remission as compared with
those same subjects when acutely depressed; and
increases from 20 to 35 Hz (beta2) and 35 to
45 Hz (gamma) ranges in depressed subjects com-
pared with controls [59,61,65]. The significance of
these findings is not as clear, because they do not
correlate with predominant sleep waveforms; how-
ever, these investigators have speculated that higher-
frequency EEG activity could reflect changes in
overall levels of arousal or integrative processing
associated with a depressive state.
Automated slow wave analysis
The technique of computerized slow wave counts
using automated algorithms has also advanced
slow wave analysis. This technology has allowed
for more sensitive detection of slow waves, and
a more quantifiable measure of slow waves than
from sleep staging alone [42].
In addition to a reduction of total minutes of
SWS, an altered distribution of SWS during the
night has been shown in subjects with major de-
pression. Minutes of SWS generally are greatest in
the first non-REM (NREM) sleep period, and de-
crease linearly through subsequent periods. Many
depressed subjects show increased or equal minutes
of SWS in the second compared with the first pe-
riod. This difference has been more clearly defined
by automated slow wave counts. Kupfer and
coworkers [66] defined the delta sleep ratio as the
ratio of the average slow (delta) wave counts per
minute in the first NREM sleep period to the aver-
age counts per minute in the second NREM sleep
period. In control subjects, this ratio (typically
>1.6) reflects the higher density of slow waves in
the first NREM sleep period. Depressed subjects typ-
ically have a lower ratio (%1.10) reflecting this ab-
normal distribution. The delta sleep ratio has been
thought to reflect an abnormal homeostatic regula-
tion of sleep in depression (deficient process ‘‘S’’
[67]).
With prolonged wakefulness, there is an increase
in SWS at the beginning of the night and delay in
REM sleep onset, reflecting a homeostatic increase
in SWS during the first NREM sleep period. In ma-
jor depression, REM sleep onset is earlier, and there
is less SWS during the first NREM sleep period, the
opposite of the normal homeostatic pattern. It has
been suggested that a higher ratio of SWA in the first
to second NREM sleep periods (delta sleep EEG
power, not slow wave counts) predicted a more ro-
bust antidepressant response effect of sleep depriva-
tion [68]. The delta sleep ratio, as with other SWA
markers, has been shown to predict treatment
response and likelihood of recurrence [48,49,66].
Although reduced SWA in the first NREM sleep pe-
riod has also been found in subjects with schizo-
phrenia, only subjects with depression showed
a reduced delta sleep ratio, suggesting this measure
may be more specific for major depression [69,70].
Coherence of sleep electroencephalogram
rhythms
Coherence is a measure of the similarity of EEG
rhythms at different cortical locations. Measures
of EEG coherence are thought to reflect the func-
tional relationships between different cortical re-
gions. Functional cortical connections may be
impaired in psychiatric disorders including major
depression. Both intrahemispheric and interhemi-
spheric coherence have been shown to be lower in
major depression compared with controls [71]. De-
creased coherence has been identified primarily in
the delta (0.5–4 Hz) and beta (16–32 Hz) fre-
quency ranges in both adolescents and adults
with major depression [71–73]. Reduced coherence
also seems to be present in offspring at risk of devel-
oping major depression [74,75], and correlates
with risk of recurrence in children and adolescents
with major depression [71], suggesting it may be
a biologic marker of depression. Most reports sug-
gest that coherence changes are more prominent
in girls and women with major depression, in con-
trast with reduced numbers of slow waves that tend
to be more prominent in men [76]. Fewer studies
have investigated coherence changes in adults
with major depression, and the relationship of co-
herence to age and other sleep measures, such as
SWA, should be further delineated.
Topography of sleep electroencephalogram
activity
Sleep EEG patterns have classically been recorded
from one or two central electrodes. EEG recordings
from different scalp locations can vary significantly,
however, and analyses of EEG patterns across the
scalp can yield substantial additional information.
The use of high-density EEG (generally >20–256
electrodes) has facilitated the description of normal
sleep EEG topography and how it varies over the
course of the night. Topographic patterns seem to
be stable between nights, despite differences in
sleep architecture [77–79]. Additionally, different
frequency ranges have characteristic distributions
of activity, likely related to different cortical sources
of rhythm generation [78,79]. In particular, SWA

shows a characteristic frontal distribution, with an
increased power density but stable topography after
sleep deprivation, which is consistent with the ho-
meostatic increase in SWS [67,77,79,80]. Despite
the potential advantages of high-density EEG, few
studies have applied this technology to psychiatric
populations. Given the convergence of imaging
data demonstrating regional differences in brain ac-
tivity (see previous and following sections summa-
rizing this research) and the clear associations of
altered SWS-SWA during depression, high-density
EEG studies of sleep are likely to yield important in-
formation that furthers the understanding of the
common biologic bases of sleep and mood
disorders.
Biologic mechanisms of sleep changes
in mood disorders
The high coincidence and overlapping symptoms of
major depression and insomnia suggest common
neurobiology. Reflecting their common clinical pre-
sentations, many of the criteria in the recently pub-
lished American Academy of Sleep Medicine
research diagnostic criteria for insomnia [81] are
shared with the Diagnostic and Statistical Manual of
Mental Disorders-IV-TR criteria for major depressive
episodes. This raises the question as to which is
the primary or secondary disorder, or if they are
manifestations of the same underlying process rep-
resenting a spectrum disorder. Despite the preva-
lence and impact of mood disorders, the exact
etiologies are still not fully understood. Similarly,
there have been many speculations about the mech-
anisms for sleep changes in mood disorders and
correlations with other biologic abnormalities iden-
tified in depression. At an even more fundamental
level, the regulation of and biologic need for sleep
are still incompletely defined (see the article by K.
Doghramji elsewhere in this issue) [80]. The close
association of mood and sleep suggest that the
neurobiology is closely intertwined; it is likely that
advances in the understanding of either component
will lead to a more complete explanation of the
other. The following sections discuss some of the
hypotheses explaining this association.
Neurotransmitters
The classic neurotransmitter hypothesis of mood
regulation was based on the discovery that increases
or decreases in monoaminergic neurotransmitters
(eg, serotonin, norepinephrine, and dopamine)
correlated with improved or worsened depression,
respectively. Most pharmaceutical agents (including
tricyclics, monoamine oxidase inhibitors, and sero-
tonin-reuptake inhibitors) used to treat depression
primarily increase synaptic levels of these
Sleep in Mood Disorders 239
neurotransmitters. Conversely, the same medica-
tions can trigger manic episodes in susceptible indi-
viduals, suggesting that the other pole of the mood
spectrum relates to excessive monoamine transmis-
sion. Recent evidence continues to support this
hypothesis, and has identified alterations in neuro-
transmitter levels, activity of brain areas primarily
associated with monoaminergic activity, and of can-
didate genes associated with serotonin levels and
function [82].
Normal regulation of sleep is closely tied to these
systems; the onset of REM sleep requires a decrease
in monoaminergic tone (including serotonin and
norepinephrine) and increased cholinergic tone
[83]. Most antidepressant medications increase se-
rotonin, and correspondingly increase REM sleep
latency, decrease REM sleep amount, and increase
SWS, reversing the typical architectural abnormali-
ties of sleep in depression [84,85]. Although this
has been proposed to be the primary mechanism
for antidepressant effect, some antidepressants do
not alter either REM sleep or serotonin levels [85].
More recently, investigations have suggested roles
for additional neurotransmitter systems in mood
disorders. Amino acid neurotransmitters, such as
glutamate acting by a-amino-3-hydroxy-5- methyl-
4-isoxazolepropionic acid (AMPA) receptors signal-
ing pathways, have increasingly been found to play
a role in depression [86]. Particularly relevant are
the associations of glutamate signaling with plastic-
ity (by increased brain-derived neurotrophic factor
levels) and learning [87–89]. A decrease in neuro-
trophic factors, such as brain-derived neurotrophic
factor, related to depression could result in de-
creased neurogenesis, or even neural cell loss, in
brain regions critical to mood regulation and
responsiveness.
The glutamate system is also intimately tied to
both REM and NREM sleep regulation. Glutamate
interacts with cholinergic neurons to increase activ-
ity of the reticular system associated with REM sleep
onset. During NREM sleep, excitatory glutamate
neurotransmission has a prominent role in the tha-
lamocortical generation of sleep EEG oscillations
[83]. Additionally, sleep has increasingly been
shown to be necessary for plastic processes, such
as learning and memory, and it affects the expres-
sion of plasticity-related genes [90]. The inter-
twined processes of mood, sleep, and plasticity,
and their modulation through such factors as gluta-
mate and brain-derived neurotrophic factor, make
them appealing targets for future therapies [89].
Evidence indicates that conventional serotoner-
gic antidepressants may indirectly potentiate
AMPA receptors, possibly relating to their efficacy
[86]. The evidence for involvement of neuroplastic-
ity and other signaling cascades perhaps explains
240 Peterson & Benca
the therapeutic lag between drug initiation and
clinical effect. Similarly, overlapping signaling path-
ways that regulate cell death and survival may be
long-term targets for both mood stabilizers and an-
tidepressants [89]. As one of such pathways, the
glutamate system and downstream signaling cas-
cades may also provide a therapeutic target for fu-
ture generations of antidepressants [86,89].
Neuroimaging
A growing body of literature has started to identify
the brain regions involved in regulation of sleep
and how their activity is altered in mood disorders
with sleep disturbances. Imaging studies are identi-
fying brain areas involved in the sleep disturbances
exhibited by depressed patients, such as reduced
SWS and increased REM sleep. During normal
NREM sleep, metabolic activity is broadly decreased
in the frontal, temporal, and parietal cortexes com-
pared with waking levels. Nofzinger and colleagues
[91] demonstrated that subjects with current major
depression have a smaller decrease in these cortical
regions from waking to sleep, and a relative
hypoactivity in waking. It is possible that the wak-
ing hypofrontality could reflect a deficit in a sleep-
wake–related process present in major depression,
such as decreased synaptic potentiation (during
waking) or decreased downscaling (during sleep)
[80]. Other brain areas involved in emotional regu-
lation (anterior cingulate cortex, amygdala, para-
hippocampal cortex, thalamus [92]) also had
a smaller decline in metabolic level from waking
to NREM sleep. Relative to control subjects, how-
ever, these areas have elevated metabolic levels dur-
ing sleep. Altered function in these regions could
relate to a failure of arousal mechanisms to decline
from waking to sleep, and changes in cognition, at-
tention, and emotional regulation in depression
[93]. Similarly, investigations of increased REM
sleep (which correlates with depression severity
and clinical outcomes) in subjects with major de-
pression demonstrated increased metabolic activity
during REM sleep compared with controls in dif-
fuse cortical and subcortical structures [94]. Because
there is a shift from predominantly monoaminergic
activity during waking to cholinergic activity during
REM sleep, these alterations could also reflect an
imbalance of monoaminergic-cholinergic systems
in altered mood states. A state of relative arousal
with lower monoaminergic activity in depression
could explain the increased REM sleep, decreased
REM sleep latency, and decreased SWS.
Other imaging studies have focused on changes
in brain activity in depressed patients after total or
partial sleep deprivation, which results in an antide-
pressant response in about 50% of patients with
major depression [95]. These studies suggest that
there is a biologic subtype of depression with defi-
cits that can be corrected by sleep deprivation, lend-
ing further support to the hypothesis that sleep and
mood regulation are controlled by overlapping
brain regions. Several studies show consistency
with this hypothesis: patients who responded to
sleep deprivation initially had increased metabolic
activity in the amygdala, orbital prefrontal cortex,
inferior temporal, and anterior cingulate, which
normalized after sleep deprivation [96–99]. Volk
and colleagues [96] demonstrated predeprivation
perfusion levels correlating with the reduction of
depressive symptoms. Functional imaging studies
with single-photon emission CT suggest that sleep
deprivation responders may have a particular deficit
in monoaminergic systems involved with attention,
arousal, and mood, particularly in dopaminergic
and serotonergic systems [97,100].
Endocrine changes
Neuroendocrine dysregulation, particularly over-
activation of the hypothalamic-pituitary-adrenal
axis, has also been long recognized as playing
a key role in the genesis of mood disorders, and
could potentially lead to sleep disturbance [101]. El-
evations of both corticotrophin-releasing hormone
and cortisol have been associated with major de-
pression and could contribute to atrophy of hippo-
campal neurons, in turn reducing their inhibition of
adrenocorticotropic hormone secretion, further
exacerbating the elevation of hypothalamic-pitui-
tary-adrenal axis activity.
Abnormalities of the hypothalamic-pituitary-
adrenal axis are found in almost half of patients
with major depression. The most common abnor-
mality is hypercortisolemia, which has classically
been assessed by the dexamethasone suppression
test. Elevated levels of cortisol are also associated
with stress and can lead to more fragmented sleep
and hippocampal damage. Coincident with cortisol
elevations, corticotrophin-releasing hormone is
also secreted based on circadian rhythms and is el-
evated in depressed patients. Increased nocturnal
corticotrophin-releasing hormone may actually be
responsible for increased awakenings with hypo-
thalamic-pituitary-adrenal axis hyperactivity [102].
Supporting this hypothesis, administration of a cor-
ticotrophin-releasing hormone receptor antagonist
was reported to improve sleep EEG patterns of de-
pressed patients [103]. Growth hormone–releasing
hormone has a reciprocal relationship with cortico-
trophin-releasing hormone, and promotes sleep.
Growth hormone–releasing hormone and growth
hormone may also be decreased in some patients
with depression, further contributing to SWS decre-
ments [101].

Genetic polymorphisms
Genetic factors likely account for at least 33% of the
risk for major depression, and over 85% of the risk
for bipolar disorder [104,105]. Both of these disor-
ders are likely polygenic and heterogeneous, and
the result of a combined genetic and environmental
factors. Based on the understanding of the neurobi-
ology of these disorders, there are a number of
candidate genes and possible associations of poly-
morphisms that could at least partially account
for some cases of mood disorders. A handful of
genes have been implicated in both mood disorders
and sleep regulation.
Genes that regulate monoamine levels (particu-
larly serotonin and norepinephrine) have been
particularly intriguing candidates, given the impor-
tance of these neurotransmitters in the response to
most antidepressant medications. Both the gene for
monoamine oxidase–A and the serotonin trans-
porter gene promoter (‘‘linked polymorphic region’’
5-HTTLPR) have been implicated in depression,
and may correlate with insomnia scores (mono-
amine oxidase–A) or treatment response to sleep
deprivation (5-HTTLPR) [106].
There have recently been reports that the angio-
tensin I-converting enzyme gene and mineralocorti-
coid receptor gene expression are altered in major
depression and bipolar disorder, respectively
[107,108]. Both genes are candidates to explain, at
least partially, abnormalities of the hypothalamic-
pituitary-adrenal axis in both mood and sleep
disturbances.
Recent years have seen rapid advances in identify-
ing ‘‘clock’’ genes involved in regulating circadian
rhythms. No specific circadian genes have been
clearly linked to depression or bipolar disorder,
but a number of them have been implicated and
may help explain treatment responses and some as-
pects of these disorders. A weak association has
been found between susceptibility for major de-
pression with a seasonal pattern (see Table 1) and
an NPAS2 gene polymorphism [109]. Similarly,
some reports suggest that the period-3 (per3) gene
variants may be associated with certain features of
mood disorders [110]. Multiple reports have now
linked a CLOCK gene polymorphism to the pres-
ence of insomnia and decreased sleep time in
depressed and bipolar patients [111], and the geno-
type may interact with lithium treatment [111]. Lith-
ium, the prototypic pharmacologic mood stabilizer,
has been shown to inhibit glycogen synthase kinase
3 (gsk3), which is also a circadian regulator. The
gsk3 gene has been under intense scrutiny as
a possible candidate gene for bipolar disorder, but
several studies have revealed only a moderate link-
age in relatively small populations [112]. More
Sleep in Mood Disorders 241
importantly, the polymorphism in question has
not been shown to have an effect on gene expres-
sion or activity. Yin and colleagues [113] recently
demonstrated that lithium also affects stability of
the Rev-erb a protein, which in turn regulates the ac-
tivation of other clock genes. This is an intriguing
possible link between bipolar disorder and circa-
dian genes, but an involvement in patients has
not yet been demonstrated.
Plasticity-related cascades are a developing area
of investigation for identifying both candidate
genes and novel molecular targets for therapeutics
[89]. These include genes involved in regulation
of DNA replication, such as histone deacetylase,
and others that are members of signal transduction
cascades. Glutamate-AMPA receptor cascades are
particularly interesting targets, and a number of ex-
perimental therapeutic agents affect this system
[89]. Plasticity is closely linked to learning, sleep,
and hormonal (cortisol) regulation. Supporting
this connection, molecular investigations of the
genes regulated by sleep and sleep deprivation iden-
tified a number of plasticity-related gene targets
[90,114]; genes related to plasticity and synaptic po-
tentiation tend to be expressed during wakefulness,
and genes related to synaptic downscaling tend to
be expressed during sleep [80]. It is feasible that
sleep is required for the downscaling of synapses
on a daily basis, and that alteration in sleep or
mood disorders could affect this normal process.
Conversely, sleep deprivation could strengthen syn-
apses in brain regions involved in affect regulation,
potentially explaining some of the acute effects of
sleep deprivation therapy.
Although anatomic, neurochemical, neuroendo-
crine, and genetic evidence seems to be converging,
it remains unknown which abnormalities are re-
sponsible for initiating mood disorders and sleep
disturbance. Nevertheless, approaches to identify
both gene linkages and molecular targets poten-
tially involved in illness continue to be critical for
nderstanding and treating the overlapping distur-
bances of mood and sleep.
Treatment of sleep disturbance and mood
disorders
Because of the high comorbidity of mood disorders
and insomnia (or hypersomnia), patients present-
ing with complaints of one must be assessed for
the other. Both pharmacologic and nonpharmaco-
logic treatment modalities for insomnia have al-
ready been discussed in this issue; however, a few
specific topics regarding mood disorders deserve
attention. Although pharmacologic treatments for
depression are more commonly prescribed, empir-
ically validated psychotherapies (eg, cognitive
242 Peterson & Benca
behavioral therapy and interpersonal therapy) are
also efficacious treatments. Similarly, increasing ev-
idence supports the efficacy of cognitive behavioral
therapy for insomnia for improvement of sleep dis-
turbances associated with mood disorders [115].
Psychotherapy may be a valuable treatment option
that can address both mood and sleep symptoms of
depression, particularly for patients who do not tol-
erate the sleep-related side effects of antidepressant
medications.
Of note, almost all of the available antidepressant
medications, including tricyclic antidepressants,
monoamine oxidase inhibitors, trazodone and
nefazodone serotonin reuptake inhibitors (fluoxe-
tine, citalopram, escitalopram, sertraline, pa-
roxetine), or serotonin-norepinephrine reuptake
inhibitors (venlafaxine or duloxetine), bupropion,
and mirtazapine can have effects on sleep (Table 4)
[116]. Despite their typical profiles, any of them can
exacerbate insomnia or hypersomnia, impair or im-
prove sleep quality, and affect EEG measures of
sleep architecture. Unfortunately, it is not always
possible to predict specific medication effects in
particular patients. This can have significant clinical
consequences, because exacerbation of insomnia
could contribute to medication noncompliance.
Daytime sedation, similarly, could further impair
the ability of patients to carry out daily activities. Al-
though not approved by the Food and Drug Admin-
istration (FDA) specifically for this purpose,
sedating antidepressants, such as tricyclic antide-
pressants, trazodone, and mirtazapine, are some-
times used to treat insomnia associated with
depression, with varying levels of empiric data
[116]. The use of the antidepressant trazodone de-
serves particular mention, because it is one of the
most prescribed agents to improve sleep, even in
those without depression. Trazodone is not FDA-
approved for treating insomnia, but it is used far
more frequently as a sleep aid than as an antide-
pressant. Despite its prevalent use, relatively little
objective data on its effects on sleep are available,
although some studies have suggested it may im-
prove insomnia in depression [117]. It is important
to remember that in general, doses of antidepres-
sant medications used specifically for insomnia
are far below the therapeutic dose for depression
and likely provide little benefit for depressed mood.
It is critical to monitor both insomnia and hyper-
somnia when treating depression. Even in ‘‘ade-
quately treated’’ patients in remission from mood
symptoms, sleep disturbance is the most common
residual symptom [118]. Insomnia is a strong pre-
dictor for recurrent episodes of depression.
Whether treatment of residual insomnia prevents
recurrence is not clear. Just as the standard of care
for treating a mood disorder with psychotic features
requires specific interventions for mood and psy-
chosis, treatment of mood disorder with sleep dis-
turbance should address both aspects of illness. A
few recent studies have addressed this issue by coad-
ministering antidepressants (serotonin reuptake in-
hibitors or tricyclic antidepressants) with hypnotic
agents, such as eszolpiclone [119,120], zolpidem,
or lorazapam. Although the number of studies
and number of subjects within the studies is lim-
ited, the results suggested that addition of a hyp-
notic was generally well tolerated and improved
insomnia. The addition of zolpidem [121] or lora-
zepam [122] to antidepressant therapy in depres-
sives did not slow antidepressant response or
cause clinically significant adverse drug interac-
tions, and decreased symptoms of insomnia
[121,122]. One more recent study suggested that
coadministration of fluoxetine with eszolpiclone
led to greater decreases in depression ratings (mea-
sured at 4 and 8 weeks) and improved sleep mea-
sures compared with fluoxetine with placebo
alone [119]; after an 8-week course, discontinuation
did not result in significant withdrawal or rebound
insomnia, and sleep and depression improvements
persisted [120]. Hypnotics alone are unlikely to
have a direct antidepressant response, however,
and are not a substitute for approved antidepres-
sant treatments.
Second-generation antipsychotics, including ris-
peridone, olanzapine, quetiapine, and clozapine,
have also been used as primary or adjunctive treat-
ments for mood disorders. Atypical antipsychotics
are approved treatments for psychotic mood epi-
sodes, and for mania and mood stabilization in
bipolar disorder, but are not FDA-approved for
treatment of nonpsychotic depression. Addition-
ally, they have also been used off-label to treat
mood-related sleep disruption. Despite their in-
creasingly common use for sleep, there are few stud-
ies investigating sleep changes related to the use of
these medications in depression. Most studies re-
port only limited subjective reports of sleep
changes, whereas the few studies including poly-
somnography data were not blinded and were per-
formed with only small numbers (N 5 8–15) of
subjects [123–125]. Although the newer antipsy-
chotics could be useful adjunctive treatments to se-
rotonin reuptake inhibitors in treatment-resistant
depression, the limited data regarding their efficacy
on sleep and their significant potential side effects
(eg, weight gain and metabolic syndrome) do not
warrant more widespread use.
At the other end of the spectrum, stimulants (in-
cluding both amphetamine derivatives and modafi-
nil) are sometimes used as an adjunctive, but not
FDA-approved, treatment for (unipolar or bipolar)
depression. Although these medications certainly
 Sleep in Mood Disorders 243

Table 4: Antidepressant medications and sleep

Medication class Dosage: depression Pharmacologic
examples (insomniaa) mechanism Effects on sleep
Tricyclic antidepressants
Amitryptiline 75–150 mg Inhibit serotonin and Sedation
(25–50 mg) norepinephrine reuptake; REM sleep suppression
anticholinergic and Increased stage 2 sleep
Nortriptyline 50–150 mg antihistaminergic effects
(25–50 mg)
Doxepin 75–300 mg
(6–50 mg)
Clomipramine 100–250 mg
Monoamine oxidase
inhibitors
Phenelzine 45–90 mg Inhibit monoamine Insomnia
Tranylcypromine 30–60 mg oxidase, thus increasing Potent REM suppression
norepinephrine,
serotonin, and dopamine
Serotonin reuptake
inhibitors
Fluoxetine 20–80 mg Inhibit serotonin reuptake Insomnia
Sertraline 50–200 mg REM suppression
Paroxetine 15–60 mg Increased eye
Citalopam 20–60 mg movements in NREM
Escitalopram 10–30 mg sleep
Serotonin-norepinephrine
reuptake inhibitors
Venlafaxine 150–450 mg Inhibit serotonin and Insomnia
Duloxetine 20–60 mg norepinephrine reuptake REM suppression
Increased eye
movements in NREM
sleep
Other antidepressants
Trazodone 150–600 mg Inhibit serotonin Sedation
(25–75 mg) reuptake; blocks a1
adrenoreceptors;
serotonin-2A receptor
antagonist
Bupropion 100–450 mg Inhibits norepinephrine Insomnia/activation
and dopamine reuptake
Mirtazapine 15–45 mg a2 receptor antagonist; Sedation; REM sleep
(7.5–15 mg) serotonin-2 and -3 suppression
receptor antagonist;
antihistaminergic

Abbreviation: REM, rapid eye movement.

a
Use of these medications for treatment of insomnia is an off-label usage, not approved by the Food and Drug
Administration.

have an effect on sleepiness, there are no data that
they directly affect mood. Modafinil has received in-
creasing attention, given its novel but not well-char-
acterized mechanism, and apparently low abuse
potential, in contrast to amphetamine derivatives.
Fatigue or hypersomnia associated with (seasonal
or atypical) depression were reportedly improved
with the addition of modafinil, but the effect on
mood was not clear [126–128]. Two of these
studies were not placebo controlled, however
[126,128], and in the controlled study, differences
in mood and sleepiness ratings did not reach statis-
tical significance between the modafinil and pla-
cebo groups [127]. Recently, a trial of adjunctive
modafinil combined with a mood stabilizer was
shown to improve both response and remission
244 Peterson & Benca
rates in bipolar depression resistant to treatment
with mood stabilizer alone [129]. Although this
study did not observe an increased incidence of
treatment-emergent mania or hypomania with
modafinil, this is an off-label use, and evidence is
too limited to suggest that modafinil is safe in
bipolar disorder; it should be considered to be as-
sociated with a potential risk of inducing mania,
similar to other stimulants. Furthermore, in all pa-
tients treated with modafinil, as with all stimulants,
clinicians need to be vigilant for treatment-emer-
gent insomnia and decreased sleep time.
Sleep-deprivation therapy
One of the most rapid, but perhaps least frequently
used, treatments for depression is sleep deprivation.
Total sleep deprivation (preventing any sleep dur-
ing the night) can reduce depressive symptoms (a
50% reduction of Hamilton Depression Rating
Scale scores) within hours in 30% to 60% of pa-
tients with major depression [130]. Partial sleep
deprivation (particularly during the latter part of
the night) has been shown in some studies to
provide a similar improvement and is easier to im-
plement. The effect of total sleep deprivation or par-
tial sleep deprivation is often short-lived, however,
and a relapse of symptoms may occur after even
short periods of sleep in at least 50% to 80% of re-
sponders. Sleep deprivation has been combined
with other treatment modalities (including medica-
tions, sleep phase advance, light therapy, and trans-
cranial magnetic stimulation) to combine the rapid
response with sustained improvements from other
modalities.
The response to sleep deprivation provides an-
other clear link between the neurobiology of
mood disorders and sleep regulation. Although nei-
ther total sleep deprivation nor partial sleep depri-
vation has become a widely used therapy, both
provide opportunities simultaneously to investigate
rapidly occurring changes in mood and other bio-
logic variables [95,131,132]. The antidepressant re-
sponse to sleep deprivation has been correlated
with other biologic markers in sleep EEG [68], im-
aging [96,97,99,100], and genetic [133] studies. Ad-
ditionally, response to sleep deprivation may serve
as a biologic marker of a major depression subtype,
and the basis for designing novel antidepressants.
Sleep loss and bipolar disorder
Sleep loss has long been recognized as a trigger for
manic episodes in patients with bipolar disorder
[20,134,135]. Although it can be difficult to isolate
insomnia’s possible role as a prodromal symptom
of mania, studies have used various designs to dem-
onstrate that sleep loss may be a risk factor for ma-
nia, independent of prodromal mood symptoms.
Additionally, laboratory-based experiments suggest
that sleep loss induced by forced wakefulness, med-
ication, or other factors can trigger mania in the
absence of other changes [135]. Interestingly, sleep
loss has also been associated, although not as
strongly, with bipolar depression [19,20,134]. Sleep
loss in bipolar patients is often followed by the on-
set of mood episodes in the following 24-hour pe-
riod, and the magnitude of sleep change seems to
correlate with the likelihood of a subsequent
mood change [20,134]. These findings stress the
importance of closely monitoring sleep patterns
in patients with bipolar disorder, and aggressively
treating the first signs of sleep loss and insomnia.
Clinical use of polysomnography
Despite the numerous EEG abnormalities of sleep
associated with mood disorders, none is currently
considered specific or sensitive enough to warrant
the use of routine polysomnography in the diagno-
sis of mood disorders. As EEG technologies and
analysis tools become more advanced, and routine
evaluation of sleep recordings moves beyond sleep
architecture, use of sleep EEG will be an invaluable
tool in evaluating insomnia, fatigue, and psychiatric
disorders.
A careful sleep history and evaluation is valuable
and necessary in all patients with mood disorders. A
polysomnogram may be indicated when a primary
sleep disorder is suspected (eg, OSA, sleep-related
movement disorders, or narcolepsy), given their
high comorbidity. For example, recent reports sug-
gest that the incidence of OSA is higher in a cohort
of subjects with major depression, and mood disor-
ders are more common in patients with OSA than
in controls [26,27]. Furthermore, increasing evi-
dence shows that appropriate treatment of the sleep
disorder can positively affect mood [31,32]. Com-
bined with the increasing prevalence of obesity
and other risk factors in the general population, as-
sessing and treating OSA is becoming increasingly
important in the psychiatric setting. Additionally,
psychiatrists and other physicians need to be vigi-
lant for iatrogenic sleep disorders caused or exacer-
bated by psychopharmacologic agents. Weight gain
associated with antipsychotics, mood stabilizers,
and antidepressants can contribute to OSA. Addi-
tionally, serotonergic antidepressants can induce
or worsen several primary sleep disorders, such as
RLS, periodic leg movements, and REM sleep be-
havior disorder [136,137].
Summary
Because of the close association between mood and
sleep disorders, it is critical to assess all patients

with sleep complaints for mood disorders and vice
versa. The presence of sleep disturbance alone is
also strongly predictive of onset of mood problems
in the future. Simultaneous treatment of both in-
somnia and major depression is often helpful, be-
cause insomnia is the most frequent residual
symptom, and its persistence is an important pre-
dictor of future illness.
Ongoing studies of the relationship between
sleep disturbances and mood disorders should pro-
vide a better understanding of their neurobiologic
underpinnings, and more importantly for those
suffering from these conditions, safer and more
effective treatments. Just as understanding other
medical disorders can help clinicians understand
normal biologic processes, it is hoped that an un-
derstanding of the sleep disorders will lead to an
understanding of normal sleep and mood.
References
[1] American Psychiatric Association. Task force on
DSM-IV. Diagnostic and statistical manual of
mental disorders: DSM-IV-TR. 4th edition.
Washington, DC: American Psychiatric Associa-
tion; 2000.
[2] Perlis ML, Giles DE, Buysse DJ, et al. Self-
reported sleep disturbance as a prodromal
symptom in recurrent depression. J Affect
Disord 1997;42(2–3):209–12.
[3] Liu X, Buysse DJ, Gentzler AL, et al. Insomnia
and hypersomnia associated with depressive
phenomenology and comorbidity in childhood
depression. Sleep 2007;30(1):83–90.
[4] Detre T, Himmelhoch J, Swartzburg M, et al.
Hypersomnia and manic-depressive disease.
Am J Psychiatry 1972;128(10):1303–5.
[5] Ford DE, Kamerow DB. Epidemiologic study of
sleep disturbances and psychiatric disorders: an
opportunity for prevention? JAMA 1989;
262(11):1479–84.
[6] Breslau N, Roth T, Rosenthal L, et al. Sleep dis-
turbance and psychiatric disorders: a longitudi-
nal epidemiological study of young adults. Biol
Psychiatry 1996;39(6):411–8.
[7] Taylor DJ, Lichstein KL, Durrence HH, et al.
Epidemiology of insomnia, depression, and
anxiety. Sleep 2005;28(11):1457–64.
[8] Roth T, Jaeger S, Jin R, et al. Sleep problems, co-
morbid mental disorders, and role functioning
in the national comorbidity survey replication.
Biol Psychiatry 2006;60(12):1364–71.
[9] Gerber PD, Barrett JE, Barrett JA, et al. The rela-
tionship of presenting physical complaints to
depressive symptoms in primary care patients.
J Gen Intern Med 1992;7(2):170–3.
[10] Buysse DJ, Reynolds CF III, Kupfer DJ, et al.
Clinical diagnoses in 216 insomnia patients
using the International Classification of Sleep
Sleep in Mood Disorders 245
Disorders (ICSD), DSM-IV and ICD-10 cate-
gories: a report from the APA/NIMH DSM-IV
Field Trial. Sleep 1994;17(7):630–7.
[11] Simon GE, VonKorff M. Prevalence, burden, and
treatment of insomnia in primary care. Am
J Psychiatry 1997;154(10):1417–23.
[12] Buysse DJ, Reynolds CF III, Hauri PJ, et al. Diag-
nostic concordance for DSM-IV sleep disorders:
a report from the APA/NIMH DSM-IV field trial.
Am J Psychiatry 1994;151(9):1351–60.
[13] Johnson EO, Roth T, Breslau N. The association
of insomnia with anxiety disorders and depres-
sion: exploration of the direction of risk. J Psy-
chiatr Res 2006;40(8):700–8.
[14] Jackson A, Cavanagh J, Scott J. A systematic
review of manic and depressive prodromes.
J Affect Disord 2003;74(3):209–17.
[15] Chang PP, Ford DE, Mead LA, et al. Insomnia in
young men and subsequent depression. The
Johns Hopkins Precursors Study. Am J Epidemiol
1997;146(2):105–14.
[16] Haynes PL, McQuaid JR, Ancoli-Israel S, et al.
Disrupting life events and the sleep-wake cycle
in depression. Psychol Med 2006;36(10):
1363–73.
[17] Field T, Diego M, Hernandez-Reif M, et al. Sleep
disturbances in depressed pregnant women and
their newborns. Infant Behav Dev 2007;30(1):
127–33.
[18] Goyal D, Gay CL, Lee KA. Patterns of sleep dis-
ruption and depressive symptoms in new
mothers. J Perinat Neonatal Nurs 2007;21(2):
123–9.
[19] Perlman CA, Johnson SL, Mellman TA. The pro-
spective impact of sleep duration on depression
and mania. Bipolar Disord 2006;8(3):271–4.
[20] Bauer M, Grof P, Rasgon N, et al. Temporal re-
lation between sleep and mood in patients
with bipolar disorder. Bipolar Disord 2006;
8(2):160–7.
[21] Reynolds CF III, Hoch CC, Buysse DJ, et al.
Sleep in late-life recurrent depression: changes
during early continuation therapy with nortrip-
tyline. Neuropsychopharmacology 1991;5(2):
85–96.
[22] Harvey AG, Schmidt DA, Scarna A, et al. Sleep-
related functioning in euthymic patients with
bipolar disorder, patients with insomnia, and
subjects without sleep problems. Am J Psychiatry
2005;162(1):50–7.
[23] Dew MA, Reynolds CF III, Houck PR, et al. Tem-
poral profiles of the course of depression during
treatment.: predictors of pathways toward
recovery in the elderly. Arch Gen Psychiatry
1997;54(11):1016–24.
[24] Sjostrom N, Waern M, Hetta J. Nightmares and
sleep disturbances in relation to suicidality in
suicide attempters. Sleep 2007;30(1):91–5.
[25] Barbe RP, Williamson DE, Bridge JA, et al. Clin-
ical differences between suicidal and nonsuici-
dal depressed children and adolescents. J Clin
Psychiatry 2005;66(4):492–8.
246 Peterson & Benca
[26] Ohayon MM. The effects of breathing-related
sleep disorders on mood disturbances in the
general population. J Clin Psychiatry 2003;
64(10):1195–200 [quiz: 1274–6].
[27] Sharafkhaneh A, Giray N, Richardson P, et al.
Association of psychiatric disorders and sleep
apnea in a large cohort. Sleep 2005;28(11):
1405–11.
[28] Picchietti D, Winkelman JW. Restless legs syn-
drome, periodic limb movements in sleep,
and depression. Sleep 2005;28(7):891–8.
[29] Peppard PE, Szklo-Coxe M, Hla KM, et al. Lon-
gitudinal association of sleep-related breathing
disorder and depression. Arch Intern Med
2006;166(16):1709–15.
[30] Bardwell WA, Ancoli-Israel S, Dimsdale JE.
Comparison of the effects of depressive symp-
toms and apnea severity on fatigue in patients
with obstructive sleep apnea: a replication
study. J Affect Disord 2007;97(1–3):181–6.
[31] Schwartz DJ, Karatinos G. For individuals with
obstructive sleep apnea, institution of CPAP
therapy is associated with an amelioration of
symptoms of depression which is sustained
long term. J Clin Sleep Med 2007;3(6):631–5.
[32] Wells RD, Freedland KE, Carney RM, et al. Ad-
herence, reports of benefits, and depression
among patients treated with continuous posi-
tive airway pressure. Psychosom Med 2007;
69(5):449–54.
[33] Ulfberg J, Bjorvatn B, Leissner L, et al. Comor-
bidity in restless legs syndrome among a sample
of Swedish adults. Sleep Med 2007;8(7–8):
768–72.
[34] Cuellar NG, Strumpf NE, Ratcliffe SJ. Symp-
toms of restless legs syndrome in older adults:
outcomes on sleep quality, sleepiness, fatigue,
depression, and quality of life. J Am Geriatr
Soc 2007;55(9):1387–92.
[35] Feinberg M, Carroll BJ, Greden JF, et al. Sleep
EEG, depression rating scales, and diagnosis.
Biol Psychiatry 1982;17(12):1453–8.
[36] Bertocci MA, Dahl RE, Williamson DE, et al.
Subjective sleep complaints in pediatric depres-
sion: a controlled study and comparison with
EEG measures of sleep and waking. J Am Acad
Child Adolesc Psychiatry 2005;44(11):1158–66.
[37] Argyropoulos SV, Hicks JA, Nash JR, et al. Cor-
relation of subjective and objective sleep mea-
surements at different stages of the treatment
of depression. Psychiatry Res 2003;120(2):
179–90.
[38] Matousek M, Cervena K, Zavesicka L, et al. Sub-
jective and objective evaluation of alertness and
sleep quality in depressed patients. BMC Psychi-
atry 2004;4:14.
[39] Armitage R, Trivedi M, Hoffmann R, et al. Rela-
tionship between objective and subjective sleep
measures in depressed patients and healthy
controls. Depress Anxiety 1997;5(2):97–102.
[40] Tsuchiyama K, Nagayama H, Kudo K, et al. Dis-
crepancy between subjective and objective sleep
in patients with depression. Psychiatry Clin
Neurosci 2003;57(3):259–64.
[41] Benca RM, Obermeyer WH, Thisted RA, et al.
Sleep and psychiatric disorders: a meta-analysis.
Arch Gen Psychiatry 1992;49((8):651–68
[discussion: 669–70].
[42] Kupfer DJ. Sleep research in depressive illness:
clinical implications—a tasting menu. Biol Psy-
chiatry 1995;38(6):391–403.
[43] Kupfer DJ, Reynolds CF III, Ulrich RF, et al.
Comparison of automated REM and slow-
wave sleep analysis in young and middle-aged
depressed subjects. Biol Psychiatry 1986;21(2):
189–200.
[44] Hudson JI, Lipinski JF, Keck PE Jr, et al. Poly-
somnographic characteristics of young manic
patients: comparison with unipolar depressed
patients and normal control subjects. Arch
Gen Psychiatry 1992;49(5):378–83.
[45] Nofzinger EA, Thase ME, Reynolds CF III, et al.
Hypersomnia in bipolar depression: a comparison
with narcolepsy using the multiple sleep latency
test. Am J Psychiatry 1991;148(9):1177–81.
[46] Howland RH, Thase ME. Biological studies of
dysthymia. Biol Psychiatry 1991;30(3):283–304.
[47] Giles DE, Roffwarg HP, Schlesser MA, et al.
Which endogenous depressive symptoms relate
to REM latency reduction? Biol Psychiatry
1986;21(5–6):473–82.
[48] Thase ME, Fasiczka AL, Berman SR, et al. Elec-
troencephalographic sleep profiles before and
after cognitive behavior therapy of depression.
Arch Gen Psychiatry 1998;55(2):138–44.
[49] Jindal RD, Thase ME, Fasiczka AL, et al. Electro-
encephalographic sleep profiles in single-epi-
sode and recurrent unipolar forms of major
depression: II. Comparison during remission.
Biol Psychiatry 2002;51(3):230–6.
[50] Thase ME, Buysse DJ, Frank E, et al. Which de-
pressed patients will respond to interpersonal
psychotherapy? The role of abnormal EEG sleep
profiles. Am J Psychiatry 1997;154(4):502–9.
[51] Giles DE, Kupfer DJ, Rush AJ, et al. Controlled
comparison of electrophysiological sleep in
families of probands with unipolar depression.
Am J Psychiatry 1998;155(2):192–9.
[52] Reynolds CF III, Kupfer DJ, Thase ME, et al.
Sleep, gender, and depression: an analysis of
gender effects on the electroencephalographic
sleep of 302 depressed outpatients. Biol Psychi-
atry 1990;28(8):673–84.
[53] Robert JJ, Hoffmann RF, Emslie GJ, et al. Sex
and age differences in sleep macroarchitecture
in childhood and adolescent depression. Sleep
2006;29(3):351–8.
[54] Reynolds CF III, Buysse DJ, Kupfer DJ, et al.
Rapid eye movement sleep deprivation as
a probe in elderly subjects. Arch Gen Psychiatry
1990;47(12):1128–36.
[55] Ivanenko A, Crabtree VM, Gozal D. Sleep and
depression in children and adolescents. Sleep
Med Rev 2005;9(2):115–29.

[56] Dahl RE. The development and disorders of
sleep. Adv Pediatr 1998;45:73–90.
[57] Lauer CJ, Riemann D, Wiegand M, et al. From
early to late adulthood: changes in EEG sleep
of depressed patients and healthy volunteers.
Biol Psychiatry 1991;29(10):979–93.
[58] Buckelmuller J, Landolt HP, Stassen HH, et al.
Trait-like individual differences in the human
sleep electroencephalogram. Neuroscience
2006;138(1):351–6.
[59] Borbely AA, Tobler I, Loepfe M, et al. All-night
spectral analysis of the sleep EEG in untreated
depressives and normal controls. Psychiatry
Res 1984;12(1):27–33.
[60] Mendelson WB, Sack DA, James SP, et al. Fre-
quency analysis of the sleep EEG in depression.
Psychiatry Res 1987;21(2):89–94.
[61] Kupfer DJ, Ehlers CL, Pollock BG, et al.
Clomipramine and EEG sleep in depression.
Psychiatry Res 1989;30(2):165–80.
[62] Luthringer R, Minot R, Toussaint M, et al. All-
night EEG spectral analysis as a tool for the
prediction of clinical response to antidepres-
sant treatment. Biol Psychiatry 1995;38(2):
98–104.
[63] Kupfer DJ, Ehlers CL, Frank E, et al. Electroen-
cephalographic sleep studies in depressed
patients during long-term recovery. Psychiatry
Res 1993;49(2):121–38.
[64] Buysse DJ, Hall M, Begley A, et al. Sleep and
treatment response in depression: new findings
using power spectral analysis. Psychiatry Res
2001;103(1):51–67.
[65] Tekell JL, Hoffmann R, Hendrickse W, et al.
High frequency EEG activity during sleep: char-
acteristics in schizophrenia and depression.
Clin EEG Neurosci 2005;36(1):25–35.
[66] Kupfer DJ, Frank E, McEachran AB, et al. Delta
sleep ratio: a biological correlate of early recur-
rence in unipolar affective disorder. Arch Gen
Psychiatry 1990;47(12):1100–5.
[67] Borbely AA. The S-deficiency hypothesis of de-
pression and the two-process model of sleep reg-
ulation. Pharmacopsychiatry 1987;20(1):23–9.
[68] Nissen C, Feige B, Konig A, et al. Delta sleep ra-
tio as a predictor of sleep deprivation response
in major depression. J Psychiatr Res 2001;35(3):
155–63.
[69] Hoffmann R, Hendrickse W, Rush AJ, et al.
Slow-wave activity during non-REM sleep in
men with schizophrenia and major depressive
disorders. Psychiatry Res 2000;95(3):215–25.
[70] Ganguli R, Reynolds CF III, Kupfer DJ. Electro-
encephalographic sleep in young, never-medi-
cated schizophrenics: a comparison with
delusional and nondelusional depressives and
with healthy controls. Arch Gen Psychiatry
1987;44(1):36–44.
[71] Armitage R, Hoffmann R, Emslie G, et al. Sleep
microarchitecture in childhood and adolescent
depression: temporal coherence. Clin EEG Neu-
rosci 2006;37(1):1–9.
Sleep in Mood Disorders 247
[72] Armitage R, Hoffmann RF, Rush AJ. Biological
rhythm disturbance in depression: temporal
coherence of ultradian sleep EEG rhythms.
Psychol Med 1999;29(6):1435–48.
[73] Armitage R, Roffwarg HP, Rush AJ. Digital pe-
riod analysis of EEG in depression: periodicity,
coherence, and interhemispheric relationships
during sleep. Prog Neuropsychopharmacol
Biol Psychiatry 1993;17(3):363–72.
[74] Morehouse RL, Kusumakar V, Kutcher SP, et al.
Temporal coherence in ultradian sleep EEG
rhythms in a never-depressed, high-risk cohort
of female adolescents. Biol Psychiatry 2002;
51(6):446–56.
[75] Fulton MK, Armitage R, Rush AJ. Sleep electro-
encephalographic coherence abnormalities in
individuals at high risk for depression: a pilot
study. Biol Psychiatry 2000;47(7):618–25.
[76] Armitage R, Hoffmann RF. Sleep EEG, depres-
sion and gender. Sleep Med Rev 2001;5(3):
237–46.
[77] De Gennaro L, Ferrara M, Vecchio F, et al. An
electroencephalographic fingerprint of human
sleep. Neuroimage 2005;26(1):114–22.
[78] Finelli LA, Achermann P, Borbely AA. Individual
fingerprints in human sleep EEG topography.
Neuropsychopharmacology 2001;25(5 Suppl):
S57–62.
[79] Tinguely G, Finelli LA, Landolt HP, et al. Func-
tional EEG topography in sleep and waking:
state-dependent and state-independent fea-
tures. Neuroimage 2006;32(1):283–92.
[80] Tononi G, Cirelli C. Sleep function and synaptic
homeostasis. Sleep Med Rev 2006;10(1):49–62.
[81] Edinger JD, Bonnet MH, Bootzin RR, et al. Deri-
vation of research diagnostic criteria for insom-
nia: report of an American Academy of Sleep
Medicine Work Group. Sleep 2004;27(8):
1567–96.
[82] Adrien J. Neurobiological bases for the relation
between sleep and depression. Sleep Med Rev
2002;6(5):341–51.
[83] Pace-Schott EF, Hobson JA. The neurobiology
of sleep: genetics, cellular physiology and sub-
cortical networks. Nat Rev Neurosci 2002;
3(8):591–605.
[84] Thase ME. Depression, sleep, and antidepres-
sants. J Clin Psychiatry 1998;59(Suppl 4):55–65.
[85] Argyropoulos SV, Wilson SJ. Sleep disturbances
in depression and the effects of antidepressants.
Int Rev Psychiatry 2005;17(4):237–45.
[86] Alt A, Nisenbaum ES, Bleakman D, et al. A role
for AMPA receptors in mood disorders.
Biochem Pharmacol 2006;71(9):1273–88.
[87] Manji HK, Moore GJ, Rajkowska G, et al. Neu-
roplasticity and cellular resilience in mood
disorders. Mol Psychiatry 2000;5(6):578–93.
[88] Manji HK, Quiroz JA, Sporn J, et al. Enhancing
neuronal plasticity and cellular resilience to de-
velop novel, improved therapeutics for difficult-
to-treat depression. Biol Psychiatry 2003;53(8):
707–42.
248 Peterson & Benca
[89] Zarate CA Jr, Singh J, Manji HK. Cellular plastic-
ity cascades: targets for the development of
novel therapeutics for bipolar disorder. Biol
Psychiatry 2006;59(11):1006–20.
[90] Cirelli C. A molecular window on sleep:
changes in gene expression between sleep
and wakefulness. Neuroscientist 2005;11(1):
63–74.
[91] Nofzinger EA, Buysse DJ, Germain A, et al.
Alterations in regional cerebral glucose metabo-
lism across waking and non-rapid eye move-
ment sleep in depression. Arch Gen Psychiatry
2005;62(4):387–96.
[92] Davidson RJ, Pizzagalli D, Nitschke JB, et al.
Depression: perspectives from affective neuro-
science. Annu Rev Psychol 2002;53(1):545–74.
[93] Nofzinger EA, Buysse DJ, Germain A, et al.
Functional neuroimaging evidence for hyper-
arousal in insomnia. Am J Psychiatry 2004;
161(11):2126–8.
[94] Nofzinger EA, Buysse DJ, Germain A, et al. In-
creased activation of anterior paralimbic and
executive cortex from waking to rapid eye
movement sleep in depression. Arch Gen Psy-
chiatry 2004;61(7):695–702.
[95] Berger M, van Calker D, Riemann D. Sleep and
manipulations of the sleep-wake rhythm in
depression. Acta Psychiatr Scand Suppl 2003;
108(s418):83–91.
[96] Volk SA, Kaendler SH, Hertel A, et al. Can re-
sponse to partial sleep deprivation in depressed
patients be predicted by regional changes of ce-
rebral blood flow? Psychiatry Res 1997;75(2):
67–74.
[97] Wu JC, Buchsbaum M, Bunney WE Jr. Clinical
neurochemical implications of sleep depriva-
tion’s effects on the anterior cingulate of de-
pressed responders. Neuropsychopharmacology
2001;25(5 Suppl):S74–8.
[98] Clark CP, Brown GG, Archibald SL, et al. Does
amygdalar perfusion correlate with antidepres-
sant response to partial sleep deprivation in
major depression? Psychiatry Res 2006;146(1):
43–51.
[99] Clark CP, Brown GG, Frank L, et al. Improved
anatomic delineation of the antidepressant re-
sponse to partial sleep deprivation in medial
frontal cortex using perfusion-weighted
functional MRI. Psychiatry Res 2006;146(1):
43–51.
[100] Ebert D, Berger M. Neurobiological similarities
in antidepressant sleep deprivation and psy-
chostimulant use: a psychostimulant theory of
antidepressant sleep deprivation. Psychophar-
macology (Berl) 1998;140(1):1–10.
[101] Steiger A. Neurochemical regulation of sleep.
J Psychiatr Res 2007;41(7):537–52.
[102] Buckley TM, Schatzberg AF. On the interactions
of the hypothalamic-pituitary-adrenal (HPA)
axis and sleep: normal HPA axis activity and
circadian rhythm, exemplary sleep disorders.
J Clin Endocrinol Metab 2005;90(5):3106–14.
[103] Held K, Kunzel H, Ising M, et al. Treatment with
the CRH1-receptor-antagonist R121919 im-
proves sleep-EEG in patients with depression.
J Psychiatr Res 2004;38(2):129–36.
[104] Fava M, Kendler KS. Major depressive disorder.
Neuron 2000;28(2):335–41.
[105] McGuffin P, Rijsdijk F, Andrew M, et al. The her-
itability of bipolar affective disorder and the ge-
netic relationship to unipolar depression. Arch
Gen Psychiatry 2003;60(5):497–502.
[106] Benedetti F, Colombo C, Serretti A, et al. Anti-
depressant effects of light therapy combined
with sleep deprivation are influenced by a func-
tional polymorphism within the promoter of
the serotonin transporter gene. Biol Psychiatry
2003;54(7):687–92.
[107] Baghai TC, Schule C, Zwanzger P, et al. Influ-
ence of a functional polymorphism within the
angiotensin I-converting enzyme gene on
partial sleep deprivation in patients with major
depression. Neurosci Lett 2003;339(3):223–6.
[108] Xing GQ, Russell S, Webster MJ, et al. Decreased
expression of mineralocorticoid receptor
mRNA in the prefrontal cortex in schizophrenia
and bipolar disorder. Int J Neuropsychophar-
macol 2004;7(2):143–53.
[109] Johansson C, Willeit M, Smedh C, et al. Circa-
dian clock-related polymorphisms in seasonal
affective disorder and their relevance to diurnal
preference. Neuropsychopharmacology 2003;
28(4):734–9.
[110] Artioli P, Lorenzi C, Pirovano A, et al. How do
genes exert their role? Period 3 gene variants
and possible influences on mood disorder phe-
notypes. Eur Neuropsychopharmacol 2007;
17(9):587–94.
[111] Benedetti F, Dallaspezia S, Fulgosi MC, et al. Ac-
timetric evidence that CLOCK 3111 T/C SNP in-
fluences sleep and activity patterns in patients
affected by bipolar depression. Am J Med Genet
B Neuropsychiatr Genet 2007;144(5):631–5.
[112] Benedetti F, Serretti A, Pontiggia A, et al. Long-
term response to lithium salts in bipolar illness
is influenced by the glycogen synthase kinase
3-beta -50 T/C SNP. Neurosci Lett 2005;376(1):
51–5.
[113] Yin L, Wang J, Klein PS, et al. Nuclear receptor
Rev-erbalpha is a critical lithium-sensitive com-
ponent of the circadian clock. Science 2006;
311(5763):1002–5.
[114] Cirelli C, Gutierrez CM, Tononi G. Extensive
and divergent effects of sleep and wakefulness
on brain gene expression. Neuron 2004;41(1):
35–43.
[115] Smith MT, Huang MI, Manber R. Cognitive be-
havior therapy for chronic insomnia occurring
within the context of medical and psychiatric
disorders. Clin Psychol Rev 2005;25(5):559–92.
[116] DeMartinis NA, Winokur A. Effects of psychiat-
ric medications on sleep and sleep disorders.
CNS Neurol Disord Drug Targets 2007;6(1):
17–29.

[117] Mendelson WB. A review of the evidence for the
efficacy and safety of trazodone in insomnia.
J Clin Psychiatry 2005;66(4):469–76.
[118] Nierenberg AA, Keefe BR, Leslie VC, et al. Resid-
ual symptoms in depressed patients who re-
spond acutely to fluoxetine. J Clin Psychiatry
1999;60(4):221–5.
[119] Fava M, McCall WV, Krystal A, et al. Eszopiclone
co-administered with fluoxetine in patients
with insomnia coexisting with major depressive
disorder. Biol Psychiatry 2006;59(11):1052–60.
[120] Krystal A, Fava M, Rubens R, et al. Evaluation of
eszopiclone discontinuation after cotherapy
with fluoxetine for insomnia with coexisting
depression. J Clin Sleep Med 2007;3(1):48–55.
[121] Asnis GM, Chakraburtty A, DuBoff EA, et al.
Zolpidem for persistent insomnia in SSRI-
treated depressed patients. J Clin Psychiatry
1999;60(10):668–76.
[122] Buysse DJ, Reynolds CF III, Houck PR, et al.
Does lorazepam impair the antidepressant
response to nortriptyline and psychotherapy?
J Clin Psychiatry 1997;58(10):426–32.
[123] Sharpley AL, Bhagwagar Z, Hafizi S, et al. Ris-
peridone augmentation decreases rapid eye
movement sleep and decreases wake in treat-
ment-resistant depressed patients. J Clin Psychi-
atry 2003;64(2):192–6.
[124] Sharpley AL, Attenburrow ME, Hafizi S, et al.
Olanzapine increases slow wave sleep and sleep
continuity in SSRI-resistant depressed patients.
J Clin Psychiatry 2005;66(4):450–4.
[125] Armitage R, Cole D, Suppes T, et al. Effects of
clozapine on sleep in bipolar and schizoaffec-
tive disorders. Prog Neuropsychopharmacol
Biol Psychiatry 2004;28(7):1065–70.
[126] Ninan PT, Hassman HA, Glass SJ, et al. Adjunc-
tive modafinil at initiation of treatment with
a selective serotonin reuptake inhibitor en-
hances the degree and onset of therapeutic
effects in patients with major depressive disor-
der and fatigue. J Clin Psychiatry 2004;65(3):
414–20.
Sleep in Mood Disorders 249
[127] Fava M, Thase ME, DeBattista C. A multicenter,
placebo-controlled study of modafinil augmen-
tation in partial responders to selective serotonin
reuptake inhibitors with persistent fatigue and
sleepiness. J Clin Psychiatry 2005;66(1):85–93.
[128] Lundt L. Modafinil treatment in patients with
seasonal affective disorder/winter depression:
an open-label pilot study. J Affect Disord
2004;81(2):173–8.
[129] Frye MA, Grunze H, Suppes T, et al. A placebo-
controlled evaluation of adjunctive modafinil
in the treatment of bipolar depression. Am
J Psychiatry 2007;164(8):1242–9.
[130] Wu JC, Bunney WE. The biological basis of an
antidepressant response to sleep deprivation
and relapse: review and hypothesis. Am J Psy-
chiatry 1990;147(1):14–21.
[131] Gillin JC, Buchsbaum M, Wu J, et al. Sleep dep-
rivation as a model experimental antidepressant
treatment: findings from functional brain imag-
ing. Depress Anxiety 2001;14(1):37–49.
[132] Wirz-Justice A, Van den Hoofdakker RH. Sleep
deprivation in depression: what do we know,
where do we go? Biol Psychiatry 1999;46(4):
445–53.
[133] Benedetti F, Serretti A, Colombo C, et al. Dopa-
mine receptor D2 and D3 gene variants are not
associated with the antidepressant effect of
total sleep deprivation in bipolar depression.
Psychiatry Res 2003;118(3):241–7.
[134] Leibenluft E, Albert PS, Rosenthal NE, et al. Re-
lationship between sleep and mood in patients
with rapid-cycling bipolar disorder. Psychiatry
Res 1996;63(2–3):161–8.
[135] Wehr TA, Sack DA, Rosenthal NE. Sleep reduc-
tion as a final common pathway in the genesis
of mania. Am J Psychiatry 1987;144(2):201–4.
[136] Yang C, White DP, Winkelman JW. Antidepres-
sants and periodic leg movements of sleep. Biol
Psychiatry 2005;58(6):510–4.
[137] Schenck CH, Mahowald MW. Rapid eye move-
ment sleep parasomnias. Neurol Clin 2005;
23(4):1107–26.

Sleep in Schizophrenia
Kathleen L. Benson, PhD
- Schizophrenia: a brief overview
- The abnormalities of sleep
in schizophrenia
Subjective assessment
Objective assessment
- Clinical and biologic correlations
Clinical correlates
Biologic correlates
- Treatment issues
Aserinsky and Kleitman’s [1] historic discovery of
rapid eye movement (REM) sleep and its association
with dream reports suggested to many that one of
the defining features of the schizophrenic psychosis
(ie, hallucinations) might constitute an intrusion of
the dream state into waking. Although the ensuing
years witnessed many attempts to validate this hy-
pothesis, polysomnographic (PSG) studies of pa-
tients with schizophrenia failed to identify any
consistent REM sleep abnormalities or any intru-
sions of REM sleep into wakefulness [2–4]. In con-
trast, many of these studies revealed other sleep
abnormalities or dyssomnias that are more consis-
tently characteristic of patients with schizophrenia.
This article describes many of these dyssomnias
and discusses their significance. It also discusses
the relationship of these dyssomnias to some of
the clinical and neuropathologic features of schizo-
phrenia. Finally, it presents an overview of antipsy-
chotic (AP) treatments; their effects on sleep; and
their potential to facilitate or augment clinical sleep
disorders, such as sleep disordered breathing (SDB)
and restless legs syndrome (RLS). A brief overview
of the clinical features and neuropathology of
schizophrenia is first presented.
PO Box 335, Barnstable, MA 02630–0335, USA
E-mail address: be97@stanford.edu
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
251
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 251–260
First- and second-generation
antipsychotics: an overview
Antipsychotic medications: their effects on
sleep patterns
Antipsychotics: side effects of insomnia
and somnolence
Adjunct medications
Antipsychotics: associated sleep disorders
- Summary
- References
Schizophrenia: a brief overview
Schizophrenia has been variously described as psy-
choticism, a gross impairment of reality testing, or
a fundamental cognitive dysfunction known as
‘‘formal thought disorder.’’ Currently, the defining
features and diagnostic criteria are best defined in
the American Psychiatric Association’s [5] Diagnos-
tic and Statistical Manual of Mental Disorders, Fourth
Edition. The defining features include a mixture of
both positive and negative symptoms. Positive
symptoms reflect ‘‘an excess or distortion of normal
functions’’ and include delusions, hallucinations,
disorganized speech, and disorganized or catatonic
behavior. Negative symptoms reflect a ‘‘diminution
or loss of normal functions’’ and include ‘‘restric-
tions in the range and intensity of emotional ex-
pression (affective flattening), in the fluency and
productivity of thought and speech (alogia), and
in the initiation of goal-directed behavior (avoli-
tion).’’ The diagnostic criteria also include signifi-
cant social or occupational impairment.
Estimates of the prevalence rate of schizophrenia
range from 0.5% to 1%. The age of onset typically
occurs in the late teenage years upward to the
doi:10.1016/j.jsmc.2008.01.001
252 Benson
early 30s. Although there is not a significant lifetime
preponderance of males over females, males typi-
cally express an earlier age of onset. An early age
of onset is also associated with poorer clinical out-
come, greater negative symptoms, and increased
cognitive impairment. For some, the onset of the
disease may be abrupt; for others, a constellation
of symptoms is expressed gradually over time. The
course of illness is also a variable feature. Some
schizophrenics remain chronically impaired and
some are permanently institutionalized. For others,
the illness is punctuated by episodic flare-ups of
positive symptoms. Negative symptoms may persist
during intervening periods of partial remission.
These episodic flare-ups are often preceded by a pro-
dromal phase, a gradual development of signs and
symptoms. Schizophrenic patients also experience
a higher mortality rate from suicide and poor
health. Finally, it is not uncommon for a patient
with schizophrenia to be diagnosed with a concom-
itant substance-related disorder.
There is no disease-specific laboratory abnormal-
ity diagnostic of schizophrenia; rather, the diagno-
sis is based entirely on a comprehensive clinical
assessment. Although the etiology and pathophysi-
ology of the illness continue to be the focus of vig-
orous study, there is neither a cure nor definitive
preventive measures. The consensus view is that
schizophrenia is a neurodevelopmental disorder in-
volving the interaction of multiple susceptibility
genes with one another and with environmental
factors, some of which may be prenatal. The broad
array of presenting signs and symptoms of schizo-
phrenia is consistent not only with the clinical het-
erogeneity of schizophrenia but also with the
diversity of potential etiologic factors. The early
age of onset of schizophrenia, coupled with its
poor prognosis, is consonant with the devastating
human costs associated with this illness. One of
these human costs is the marked dyssomnia subjec-
tively reported by patients with schizophrenia and
objectively validated by overnight sleep studies.
The abnormalities of sleep in schizophrenia
Subjective assessment
Although it is a common clinical experience that
major depression and primary insomnia are associ-
ated with disturbed sleep, patients with schizophre-
nia describe the subjective quality of their sleep in
very similar terms [6]. Their subjective assessment
of poor sleep quality is predictive of self-assessed
poor quality of life and impaired coping skills
[7,8]. Self-assessed poor sleep quality includes sub-
jective reports bearing on measures of sleep mainte-
nance (ie, loss of total sleep time [TST], degraded
sleep efficiency [SE]) defined as the percent of TST
relative to time in bed, early insomnia as measured
by sleep onset latency [SL], middle insomnia, and
early morning awakenings). Early and middle in-
somnia are among the most common complaints
[9,10]. Patients may also report a degraded quality
of sleep including restlessness and agitation. Epi-
sodic flare-ups of psychotic symptoms may be ac-
companied by nights of significant insomnia or
total sleeplessness. There is also a greater likelihood
of sleep-wake reversals (ie, sleeping during the day
and wakefulness at night). This sleep-wake reversal
is also correlated with subjective complaints of
poor sleep quality [11]. Finally, it is important for
clinicians to take note of two observations. First, se-
vere insomnia is one of the prodromal signs associ-
ated with impending psychotic exacerbation or
with relapse following the discontinuation of AP
treatment [12–16]. Second, these studies of subjec-
tive sleep quality have, for the most part, sampled
schizophrenics on standard doses of APs; this sug-
gests that some APs may have limited efficacy in
treating schizophrenia-associated dyssomnias.
Objective assessment
The impetus for many of the earliest studies of sleep
patterns in schizophrenics rested with the potential
role of REM sleep intrusions in the pathogenesis of
schizophrenia. Subsequent to their failure to iden-
tify any consistent REM sleep abnormality, investi-
gations turned to an examination of other aspects
of sleep architecture: measures of sleep mainte-
nance (SL, TST, SE); measures of non-REM sleep,
in particular sleep stages 3 and 4; REM latency (ie,
the interval between sleep onset and the first REM
period); REM sleep eye movement activity; and
the quantification of sleep-related brain wave pat-
terns, such as electroencephalogram (EEG) delta
(0–3 Hz) activity, EEG beta and gamma (20–45
Hz) activity, and sleep spindles (12–15 Hz events).
Many of these investigations hoped to find
a unique sleep pattern or abnormality that might
serve as a biologic marker to identify those with
schizophrenia. Others hoped that sleep abnormali-
ties in schizophrenia might be predictive of pro-
gnosis or treatment outcome or, better yet, might
provide some insight into the underlying patho-
physiology of schizophrenia. Although many of
these goals went unmet, these studies did provide
a comprehensive description of the range of sleep
abnormalities seen in patients with schizophrenia.
In the 50 plus years subsequent to the discovery
of REM sleep, scores of published studies have
added to this description. These studies have dif-
fered in many ways including sample size; protocol
design; control groups; algorithms quantifying
sleep parameters; the age of the subjects; their med-
ication status and history; and their clinical features

(eg, hallucinating versus nonhallucinating) and
clinical history (eg, early onset versus late onset).
In recent years, two meta-analyses [17,18] and
two reviews [19,20] have brought some coherence
to the diversity of these studies. The next section
summarizes their observations regarding the more
salient dyssomnias associated with schizophrenia.
Measures of sleep maintenance
Subjective complaints of insomnia have been exten-
sively validated by objective PSG study. Broadly
speaking, unmedicated schizophrenics who are ac-
tively symptomatic evidence poor SE; reductions
in TST; and early, middle, and late insomnia.
Schizophrenic patients typically take a protracted
amount of time to reach a state of persistent sleep.
This early insomnia (or long SL) is the most consis-
tently reported abnormality shown in empiric stud-
ies of sleep patterns in schizophrenia. In healthy
controls, SL rarely exceeds 30 minutes, but in em-
piric studies of symptomatic schizophrenics SL fre-
quently exceeds 30 minutes, and often exceeds 1
hour. This characteristic insomnia may be associ-
ated with the pathophysiology of schizophrenia
or may reflect hyperarousal secondary to ongoing
emotional and psychotic turmoil.
Measures of non–rapid eye movement sleep
Attention now turns to those components of sleep
that have been extensively studied and scrutinized:
notable non-REM stages 3 and 4, known collec-
tively as slow wave sleep (SWS); and REM latency
(REML). The reader is again referred the meta-anal-
yses and reviews previously cited [17–20] for
a more detailed enumeration and exposition of
the many investigative efforts. In contrast to the
consistent demonstration of sleep maintenance
failures, SWS deficits and abnormally short REMLs
have been found in some studies and not in others.
Because of this variability, abnormalities of SWS
and REML have not been confirmed by the meta-
analytic approach. The lack of consistent findings
is frequently attributed to between-study differ-
ences in medication status or clinical features,
such as concomitant depression; however, it may
also reflect the heterogeneity of the disease itself.
Although SWS deficits and short REMLs are not
disease specific, these dyssomnias have been
observed in most studies including first episode,
AP-naive patients with schizophrenia [21].
The motivation to document SWS deficits and
short REMLs in patients with schizophrenia re-
flected not only an interest in identifying, within
the framework of those sleep abnormalities, bio-
logic markers for the disease, but also the hope
that SWS deficits and short REMLs might shed
some light on underlying pathophysiologic
Sleep in Schizophrenia 253
mechanisms. The homeostatic model of SWS was
first advanced in 1974 [22]. According to this
model, the homeostatic drive builds up during wak-
ing hours and dissipates in SWS across successive
non-REM sleep cycles. The strength of the drive
reflects both the amount of prior waking and the in-
tensity of prior waking brain activity. In healthy
subjects, this homeostatic drive is clearly demon-
strated by an augmentation of SWS following a pe-
riod of sleep deprivation. This dynamic overshoot
suggests that SWS serves a restorative role in brain
function. By implication, SWS restoration in
patients with schizophrenia may be an important
contributor to their clinical and neurocognitive
outcome.
SWS deficits are potentially related to another
non-REM sleep abnormality, short REML. There is
lack of agreement regarding the mechanism(s) un-
derlying short REMLs in patients with schizophre-
nia. First, SWS deficits, particularly in the first
non-REM cycle, might permit the passive advance
or early onset of the first REM period. Alternatively,
short REML might represent a primary abnormality
of REM sleep (ie, an augmented drive for REM
sleep).
Measures of rapid eye movement time
and rapid eye movement sleep eye movement
activity
The meta-analyses and reviews previously cited
[17–20] revealed no systematic augmentations or
reductions in the amount of REM sleep time
when comparing schizophrenics with healthy or
psychiatric controls. In addition to empiric studies
of the tonic amounts of REM sleep time, other re-
search has quantified one of the phasic events of
REM sleep, notably REM sleep eye movements.
These studies have demonstrated no consistent in-
crease or decrease in REM sleep eye movements
activity in schizophrenic patients (both AP-naive or
currently unmedicated) relative to nonpsychiatric
and psychiatric controls [23–26].
Measures of sleep-related brain wave activity
In contrast to visual scoring of SWS, computer
algorithms have been developed to quantify the in-
cidence and amplitude of underlying EEG delta
(0–3 Hz) frequencies of non-REM sleep. Studies us-
ing computer quantifications have confirmed the
loss of power in the delta range in schizophrenics
relative to nonpsychiatric controls [24,27,28]. This
loss of non-REM delta power can occur despite
comparable amounts of visually scored SWS [25].
In 1983, Feinberg [29] proposed a neurodevelop-
mental model of schizophrenia that accounts for
both SWS deficits and the loss of underlying delta
EEG. According to this model, schizophrenia
254 Benson
develops during the second decade of life because
of a fault in the normal maturational process of syn-
aptic elimination scheduled for this stage of devel-
opment. Less synchronous slow or delta EEG
activity (and associated SWS deficits) would reflect
excess synaptic pruning.
In addition to sleep-related abnormalities in the
slow or delta range of EEG activity, it has been re-
ported that unmedicated schizophrenics exhibit
greater power in the high-frequency beta (20–35
Hz) and gamma (35–45 Hz) EEG ranges than
healthy controls during all stages of sleep [30]. Fi-
nally, a reduction in both the number and ampli-
tude of non-REM sleep spindles (12–15 Hz) has
been observed in medicated schizophrenics; this
finding hints at some abnormality in thalamic-retic-
ular and thalamocortical function in schizophrenia
[31].
Clinical and biologic correlations
Clinical correlates
The relationship of the dyssomnias of schizophre-
nia to clinical symptoms, neurocognitive impair-
ment, and prognosis has been extensively studied.
Although some studies investigated global assess-
ments of symptom severity, others examined the
components of symptom severity, such as positive
or negative symptoms and cognitive dysfunction.
Studies have documented a positive correlation
between global symptom severity and increased
waking, reduced REM sleep time, SWS deficits,
and short REML [32–34]. Positive symptoms,
such as hallucinations and delusions, have been di-
rectly associated with long SLs [35], impaired SE
[36], short REML [21,23,33,37], increased REM
sleep eye movements density [26,38], and high-fre-
quency EEG brain wave activity [30]. Negative
symptoms have also been directly associated with
short REML [33,39], SWS deficits [27,40–44], and
underlying high-amplitude delta wave activity
[45]. Formal thought disorder, or cognitive dys-
function, also correlates with SWS deficits, which
is perhaps indicative of frontal lobe dysfunction
[46]. In addition, poor clinical and psychosocial
outcome have been associated with SWS deficits
[47] and short REMLs [39,48]. Finally, in compari-
sons of schizophrenics with healthy controls,
a small collection of studies demonstrated sleep-
related impairments in neuropsychologic perfor-
mance tasks: procedural memory [49], reaction
time in a selective attention task [50], visuospatial
memory [51], and tasks of attention and cognitive
flexibility [52].
These correlational studies of sleep abnormalities
reflect a diversity of clinical assessment instruments,
different algorithms to quantify sleep parameters,
small sample sizes, major differences in medication
status and history, and clinical heterogeneity of the
subjects. Consequently, an overarching synthesis is
premature. Finally, most of these studies were
cross-sectional in design; a longitudinal, within-pa-
tient, assessment of sleep patterns across changing
clinical states may prove to be a more productive
methodology.
Biologic correlates
Research has also documented significant associa-
tions between brain structures and the dyssomnias
of schizophrenia. The four studies reporting these
linkages used CT brain imaging technology. Two
studies reported that SWS deficits or reductions in
its stage 4 component were associated with en-
larged ventricular system volume [40,53]. This find-
ing was not confirmed in a later study of AP-naive
schizophrenics [23]. Poor sleep maintenance has
also been associated with brain dysmorphologies.
Longer sleep latencies [40] and the number of
awakenings [23] have been linked to size increases
in the proportion of ventricular system volume to
whole brain volume, known as the ‘‘ventricular
brain ratio’’ [40]. A positive correlation has also
been reported between the number of awakenings
and both global and prefrontal cortical atrophy
[40]. Finally, negative correlations have been re-
ported between TST and third ventricle width [40]
and between sleep maintenance and third ventri-
cle/brain ratio, caudate/brain ratio, and anterior
horn/brain ratio [54].
Interpretation of these diverse findings requires
a note of caution. The four studies summarized
here represent a mere snapshot in time. All were
cross-sectional designs. Only longitudinal studies
could justify the use of such terms as ‘‘enlargement’’
and ‘‘atrophy,’’ with the former suggesting illness-re-
lated growth and the latter suggesting shrinkage.
Changes in the volumes of these brain structures
before the point of study are speculative. A second
note of caution concerns the interpretation of brain
structure correlates. Neuroanatomic correlates of
sleep abnormalities might suggest a stable or trait-
like impairment. As is seen in subsequent sections,
however, patterns of sleep maintenance and staging
can and do undergo change, particularly in re-
sponse to AP treatment. Furthermore, anatomic
structures in the human brain are subject to adult
neurogenesis [55], and longitudinal brain imaging
studies of schizophrenics have demonstrated brain
structure changes associated with AP treatment
[56–58]. The observation that adult neurogenesis
can be inhibited by sleep loss [59] suggests that
the chronic dyssomnias of schizophrenia might
themselves contribute to brain structure
abnormalities.

In addition to neuroanatomic correlates, neuro-
chemical associations have also been explored,
but these investigations have been few and not sys-
tematically related. There are no studies of the dys-
somnias of schizophrenia in relationship to
dopamine (DA), the neurotransmitter long thought
to be a major factor in the pathophysiology of
schizophrenia. Four other neurotransmitter systems
(ie, acetylcholine, serotonin [5-HT], norepineph-
rine, and hypocretin), however, have been exam-
ined. With regard to acetylcholine, short REMLs
have been associated with cholinergic supersensi-
tivity [60]. SWS deficits have been linked to a seroto-
nergic (5-HT) dysfunction in a study finding
a positive correlation between SWS time and
cerebrospinal fluid levels of the principal 5-HT
metabolite [61]. Increased cerebrospinal fluid levels
of norepinephrine and its primary metabolite,
3-methoxy–4-hydroxyphenylglycol (MHPG), have
accompanied psychotic decompensation and re-
lapse-related insomnia [14]. Finally, a positive
correlation has been reported between sleep latency
and levels of cerebrospinal fluid hypocretin, a wake-
promoting neurotransmitter; the authors suggest
a possible relationship between the neurotransmit-
ter hypocretin and hyperarousal in schizophrenia
[62].
Treatment issues
First- and second-generation antipsychotics:
an overview
Most patients diagnosed with schizophrenia are ex-
posed to AP medications. APs have signature effects
on neurotransmitter receptors (eg, DA, 5-HT, a-ad-
renergic, cholinergic, and histaminic receptors) and
their numerous subtypes; these unique receptor-
binding profiles are associated with their therapeu-
tic efficacy and a wide range of potentially adverse
effects.
The first-generation AP medications are known as
‘‘traditional’’ or ‘‘typical’’ APs. Generally speaking,
the typical APs have a strong affinity for the DA
D2 postsynaptic receptor. Although this affinity
has been credited with their therapeutic efficacy,
this same affinity, particularly to D2 receptors in
the nigrostriatal pathways, has been linked to an ar-
ray of extrapyramidal side effects, such as akathisia,
dystonia, and parkinsonism. Furthermore, D2 re-
ceptor-binding is associated with a more adverse ef-
fect, namely tardive dyskinesia. The adverse effects
of extrapyramidal side effects and tardive dyskinesia
have contributed to the poor compliance associated
with first-generation APs.
The second-generation AP medications are
known collectively as ‘‘novel’’ or ‘‘atypical’’ APs. In
current clinical practice, atypical APs are first-line
Sleep in Schizophrenia 255
therapy for schizophrenia. The atypical APs include
clozapine, risperidone, olanzapine, quetiapine, zi-
prasidone, aripiprazole, and paliperidone. Although
each has a unique receptor-binding profile, all have
high 5-HT to DA binding ratios [63]. As a conse-
quence, the incidence of extrapyramidal side effects
and tardive dyskinesia is lower in second-generation
APs. Among the atypicals, risperidone, olanzapine,
and paliperidone show some dose-related increase
in extrapyramidal side effects [63]. Second-genera-
tion APs have also been linked to other morbidities,
such as weight gain, glucose dysregulation, type 2 di-
abetes, and hyperlipidemia [63].
The amount of weight gain associated with the
atypical APs is greatest for those taking clozapine
and olanzapine; moderate for those taking risperi-
done and quetiapine; and lower for patients on
ziprasidone, aripiprazole, and paliperidone [63].
Weight gain is a risk factor for metabolic distur-
bances, such as glucose dysregulation, but it is
also a risk factor for sleep-disordered breathing,
a dyssomnia further taxing sleep-related restorative
processes.
Antipsychotic medications: their effects
on sleep patterns
Several studies have evaluated the effect of first-gen-
eration APs on the sleep of patients with schizo-
phrenia using PSG methodology. Haloperidol was
most commonly studied, but other studies have in-
cluded such agents as chlorpromazine, thiothixene,
and fluphenazine. Traditional APs have been
shown to improve measures of sleep maintenance,
increasing both TST and SE and reducing both SL
and wake after sleep onset; their effects on REML,
stage REM minutes, and REM sleep eye movements
density were less consistent [64–69]. Chlorproma-
zine was found to increase SWS minutes and
REML [64].
Empiric studies of the sleep of schizophrenic
patients treated with second-generation APs have
examined the effects of clozapine, risperidone,
olanzapine, and paliperidone. Clozapine, being
the first of the atypical APs, has been the most ex-
tensively studied [69–71]. These three studies have
been in broad agreement in finding clozapine-re-
lated increases in TST, SE, and stage 2 minutes;
they have also reported clozapine-related decreases
in SL, waking, and SWS. PSG studies of schizo-
phrenic patients treated with olanzapine have re-
ported increased SE and SWS [72,73]. One of
these two studies also noted significantly increased
TST, stage 2 minutes, REM sleep eye movements,
density, and significantly decreased amounts of
waking and stage 1 minutes [72]. The second olan-
zapine study also reported significantly increased
stage REM minutes [73]. An enhancement of SWS
256 Benson
has also been observed in risperidone-treated
schizophrenics [74]. The effect of paliperidone on
sleep patterns in patients with schizophrenia re-
ported the following: increased SE, TST, stage 2
minutes, and stage REM minutes; decreased SL,
waking, and stage 1 minutes [75]. There are no em-
piric PSG studies of schizophrenics treated with
quetiapine, ziprasidone, and aripiprazole; however,
in a PSG study of healthy controls, quetiapine was
associated with significant improvements in sleep
induction and sleep continuity [76].
Broadly speaking, these acute and short-term
dosing studies of the effects of APs on the sleep of
patients with schizophrenia demonstrate a positive
improvement in sleep maintenance and architec-
ture. Given that clinical improvement typically
lags behind the initiation of AP treatment, improve-
ments in sleep maintenance and architecture are
not mere by-products of clinical improvement.
Rather, the restoration or normalization of sleep
processes may contribute to a positive clinical
outcome.
Antipsychotics: side effects of insomnia
and somnolence
The National Institute of Mental Health–sponsored
Clinical Antipsychotic Trials of Intervention Effec-
tiveness study documents the prevalence of adverse
effects in AP-treated schizophrenics [77]. These
adverse events include reported rates of daytime
sedation and somnolence, and rates of residual in-
somnia. AP-treated schizophrenics reported som-
nolence rates ranging from 24% to 31%. Rates of
residual insomnia ranged from 16% to 30%. Al-
though improvements in sleep maintenance and ar-
chitecture have been amply documented, these
figures suggest that a large number of AP-treated
schizophrenics suffer from daytime somnolence
or from residual insomnia. These sleep-related
adverse effects may contribute to noncompliance
and potentially to a poorer outcome.
Adjunct medications
Benzodiazepine tranquilizers and hypnotics may be
prescribed to address complaints of residual insom-
nia; however, they should be prescribed cautiously,
particularly for those schizophrenics with a sleep-
related breathing disorder or a history of alcohol
or drug abuse. Mood stabilizers and antidepressants
also have positive effects on insomnia; these agents
are frequently prescribed for patients diagnosed
with concomitant affect disorders or impulse con-
trol problems.
Melatonin
Melatonin, the chief hormonal product of the pi-
neal gland, has been used to treat insomnias
associated with disturbed patterns of melatonin se-
cretion. Studies have shown that the nighttime peak
in melatonin secretion is blunted in medication-
free schizophrenics and is not normalized even af-
ter clinical improvement with AP treatment
[78,79]. Two studies have tested the effect of exoge-
nous melatonin on residual insomnia in AP-treated
schizophrenics. First, melatonin replacement (2 mg
controlled release) significantly improved SE as
measured by actigraphy [80]. Second, exogenous
melatonin (3–12 mg/night, modal dose of 3 mg)
increased self-reported TST and reduced self-
reported nighttime awakenings [81].
Modafinil
Modafinil is a wakefulness-promoting agent cur-
rently approved by the Food and Drug Administra-
tion for the treatment of excessive daytime
sleepiness associated with narcolepsy. Modafinil
has been considered an adjunct medication to
offset the somnolence associated with many AP
agents. Case studies [82] and an open-label pilot
study [83] show that modafinil, as an adjunct to
AP treatment, can increase wake time, reduce
fatigue and TST, and improve quality of life. Be-
cause stimulant drugs that promote wakefulness
may increase the risk of relapse or exacerbation of
psychosis in patients with schizophrenia [84], off-
label use of modafinil to control AP-related seda-
tion in schizophrenia requires more extensive
investigation.
Antipsychotics: associated sleep disorders
It is not uncommon for patients with schizophrenia
to present with symptoms of other sleep disorders
regularly seen in sleep disorders clinics. These may
include inadequate sleep hygiene, irregular sleep-
wake patterns, parasomnias, sleep-related move-
ment disorders, and sleep-related breathing
disorders. Because there are no large-scale preva-
lence studies of sleep disorders in AP-naive schizo-
phrenics, their baseline rates for comorbid sleep
disorders are unknown. Clearly, some of these dys-
somnias, such as irregular sleep-wake patterns, may
be associated with their illness. Unfortunately,
others may be enhanced by, or induced by, AP
treatment.
Somnambulism and sleep-related eating
disorder: two parasomnias
Somnambulism or ‘‘sleep-walking’’ is typically initi-
ated during an arousal from SWS. It has been de-
scribed as a potential side effect of treatment with
first-generation APs, particularly in combination
with lithium [85]. Sleepwalking also has been ob-
served in patients treated with the atypical AP olan-
zapine [86]. Both lithium and olanzapine have

been credited with enhancing SWS, and thus may
be associated with an increased risk of impaired
arousal. Clonazepam, a treatment option in pri-
mary arousal disorders, might be considered for
the treatment of AP-induced somnambulism [87].
Sleep-related eating disorder is another parasomnia
that can be induced by AP medication. This syn-
drome has been associated with haloperidol [88],
olanzapine [89], and risperidone treatment [90].
Sleep-disordered breathing
High prevalence rates for SDB in schizophrenic pa-
tients are suggested by research protocols. These
studies reported the following estimates of SDB:
17% with a respiratory disturbance index greater
than five events per hour of sleep [91], 48% with
a respiratory disturbance index greater than 10
events per hour of sleep [92], and 19% with a desa-
turation index greater than five events per hour
[93]. In contrast, a study of schizophrenic patients
referred to a sleep clinic for a suspected sleep disor-
der determined that more than 46% had a respira-
tory disturbance index greater than 10 events per
hour; the mean respiratory disturbance index was
64.8 events per hour; the best predictor of SDB
was obesity [94]. Weight gain secondary to AP treat-
ment, particularly second-generation APs, carries
a serious morbidity risk including the development
of moderate to severe SDB [95]. Clinicians must
consider the differential diagnosis of comorbid
SDB for schizophrenics who present with daytime
somnolence and who are obese by history or who
have undergone weight gain secondary to AP treat-
ment. These patients may be poor historians and
frequently have no bed partner to provide informa-
tion regarding snoring and breathing pauses. Im-
portantly, daytime somnolence in patients with
schizophrenia may signal more than AP-related se-
dation. Patients who are comorbid for schizophre-
nia and SDB can be treated effectively with nasal
continuous positive airway pressure; they also can
demonstrate relatively good compliance and signif-
icant clinical improvement [96,97].
Sleep-related movement disorders
DA deficiency has been linked to the pathophysiol-
ogy of sleep-related movement disorders, such as
RLS and periodic limb movement disorder
(PLMD); consequently, RLS and PLMD may be
more prevalent in schizophrenics because of antag-
onism of the DA receptor by APs. The diagnosis of
RLS is based on the self-report of symptoms evalu-
ated by a trained clinician using well-defined diag-
nostic criteria. In the case of schizophrenics, RLS
must be distinguished from the restlessness of aka-
thisia. In contrast, overnight PSG is the objective
methodology used to make the diagnosis of
Sleep in Schizophrenia 257
PLMD. Most patients diagnosed with RLS are often
comorbid for PLMD.
In a study of the prevalence of RLS in AP-treated
schizophrenics, 21.4% of the schizophrenics met
diagnostic criteria for RLS in contrast to 9.3% of
healthy controls [98]. This group also found that
the severity of psychiatric symptoms (as measured
by the Brief Psychiatric Rating Scale) was greater
in those schizophrenic with RLS than in the
schizophrenics without RLS. The prevalence of
PLMD in patients with schizophrenia has been
less rigorously examined. Prevalence rates for pa-
tients treated with first-generation APs are in the
13% to 14% range [91,92]. Two case reports
link second-generation APs to the development
of RLS and a clinically significant PLMD index
in patients with schizophrenia; the first case was
associated with olanzapine treatment [99], the
second case with risperidone treatment [100].
Both cases resolved on switching to a different
atypical AP. Although DA agonists are first-line
therapeutic options to treat RLS or PLMD, they
are not treatment options for patients diagnosed
with schizophrenia. Rather, a reduction in AP
dose or a change in medication must be consid-
ered whenever RLS or PLMD develop secondary
to AP treatment.
Summary
Patients diagnosed with schizophrenia may be co-
morbid for dyssomnias either induced by or exacer-
bated by their treatment with AP agents. These
dyssomnias include somnambulism, sleep-related
eating disorders, sleep-related breathing disorders,
and sleep-related movement disorders. Every effort
should be made to treat comorbid sleep disorders
vigorously in patients with schizophrenia. A favor-
able prognosis or positive clinical outcome may re-
quire some normalization of sleep and its
restorative processes.
References
[1] Aserinsky E, Kleitman N. Regularly occurring
periods of eye motility, and concomitant phe-
nomena, during sleep. Science 1953;118:273–4.
[2] Dement W. Dream recall and eye movements
during sleep in schizophrenics and normals.
J Nerv Ment Dis 1955;122:263–9.
[3] Rechtschaffen A, Schulsinger F, Mednick S.
Schizophrenia and physiological indices of
dreaming. Arch Gen Psychiatry 1964;10:89–93.
[4] Benson KL, Zarcone VP. Testing the REM phasic
event intrusion hypothesis of schizophrenia.
Psychiatry Res 1985;15:163–73.
[5] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
258 Benson
edition. Washington (DC): American Psychiat-
ric Press; 1994. p. 273–90.
[6] Doi Y, Minowa M, Uchiyama M, et al. Psycho-
metric assessment of subjective sleep quality us-
ing the Japanese version of the Pittsburgh Sleep
Quality Index in psychiatric disordered and
control subjects. Psychiatry Res 2000;97(2–3):
165–72.
[7] Ritsner M, Kurs R, Ponizovsky A, et al. Perceived
quality of life in schizophrenia: relationships to
sleep quality. Qual Life Res 2004;13:783–91.
[8] Hofstetter JR, Lysaker PH, Mayeda AR. Quality
of sleep in patients with schizophrenia is associ-
ated with quality of life and coping. BMC Psy-
chiatry 2005;5(1):13.
[9] Haffmans PM, Hoencamp E, Knegtering HJ,
et al. Sleep disturbance in schizophrenia. Br
J Psychiatry 1994;165:697–8.
[10] Serretti A, Mandelli L, Lattuada E, et al. Depres-
sive syndrome in major psychoses: a study on
1351 subjects. Psychiatry Res 2004;127(1–2):
85–99.
[11] Hofstetter JR, Mayeda AR, Happel CG, et al. Sleep
and daily activity preferences in schizophrenia:
associations with neurocognition and symp-
toms. J Nerv Ment Dis 2003;191(6):408–10.
[12] Herz MI, Melville C. Relapse in schizophrenia.
Am J Psychiatry 1980;137(7):801–5.
[13] Kumar S, Thara R, Rajkumar S. Coping with
symptoms of relapse in schizophrenia. Eur
Arch Psychiatry Neurol Sci 1989;239(3):213–5.
[14] Van Kammen DP, van Kammen WB, Peters JL,
et al. CSF MHPG, sleep and psychosis in schizo-
phrenia. Clin Neuropharmacol 1986;9(Suppl 4):
575–7.
[15] Dencker SJ, Malm U, Lepp M. Schizophrenic re-
lapse after drug withdrawal is predictable. Acta
Psychiatr Scand 1986;73:181–5.
[16] Chemerinski E, Ho B, Flaum M, et al. Insomnia
as a predictor for symptom worsening follow-
ing antipsychotic withdrawal in schizophrenia.
Compr Psychiatry 2002;43(5):393–6.
[17] Benca RM, Obermeyer WH, Thisted RA, et al.
Sleep and psychiatric disorders: a meta-analysis.
Arch Gen Psychiatry 1992;49:651–68.
[18] Chouinard S, Poulin J, Stip E, et al. Sleep in un-
treated patients with schizophrenia: a meta-
analysis. Schizophr Bull 2004;30(4):957–67.
[19] Benson KL, Zarcone VP. Schizophrenia. In:
Kryger MH, Roth T, Dement WC, editors. Prin-
ciples and practice of sleep medicine. 4th edi-
tion. Philadelphia: Elsevier Saunders; 2005. p.
1327–36.
[20] Benson KL. Sleep in schizophrenia: impair-
ments, correlates, and treatment. Psychiatr
Clin North Am 2006;29:1033–45.
[21] Poulin J, Daoust A, Forest G, et al. Sleep archi-
tecture and its clinical correlates in first episode
and neuroleptic-naive patients with schizophre-
nia. Schizophr Res 2003;62:147–53.
[22] Feinberg I. Changes in sleep cycle patterns with
age. J Psychiatr Res 1974;10:283–306.
[23] Lauer CJ, Schreiber W, Pollmächer T, et al. Sleep
in schizophrenia: a polysomnographic study on
drug-naive patients. Neuropsychopharmacol-
ogy 1997;16:51–60.
[24] Keshavan MS, Reynolds CF III, Miewald JM,
et al. Delta sleep deficits in schizophrenia.
Arch Gen Psychiatry 1998;55:443–8.
[25] Hoffman R, Hendrickse W, Rush AJ, et al. Slow-
wave activity during non-REM sleep in men
with schizophrenia and major depressive disor-
ders. Psychiatry Res 2000;95:215–25.
[26] Benson KL, Zarcone VP. REM sleep eye move-
ment activity in schizophrenia and depression.
Arch Gen Psychiatry 1993;50:474–82.
[27] Ganguli R, Reynolds CF III, Kupfer DJ. EEG
sleep in young, never medicated, schizophrenic
patients: a comparison with delusional and
nondelusional depressives and with healthy
controls. Arch Gen Psychiatry 1987;44:36–45.
[28] Hiatt JF, Floyd TC, Katz PH, et al. Further evi-
dence of abnormal NREM sleep in schizophre-
nia. Arch Gen Psychiatry 1985;42:797–802.
[29] Feinberg I. Schizophrenia: caused by a fault in
programmed synaptic elimination during ado-
lescence? J Psychiatr Res 1983;17:319–34.
[30] Tekell JL, Hoffman R, Hendrickse W, et al. High
frequency EEG activity during sleep: characteris-
tics in schizophrenia and depression. Clin EEG
Neurosci 2005;36(1):25–35.
[31] Ferrarelli F, Huber R, Peterson MJ, et al. Reduced
sleep spindle activity in schizophrenia patients.
Am J Psychiatry 2007;164(3):483–92.
[32] Kempenaers C, Kerkhofs M, Linkowski P, et al.
Sleep EEG variables in young schizophrenic
and depressive patients. Biol Psychiatry 1988;
24:833–8.
[33] Tandon R, Shipley JE, Taylor S, et al. Electroen-
cephalographic sleep abnormalities in schizo-
phrenia: relationship to positive/negative
symptoms and prior neuroleptic treatment.
Arch Gen Psychiatry 1992;49:185–94.
[34] Thaker GK, Wagman AMI, Tamminga CA. Sleep
polygraphy in schizophrenia: methodological
issues. Biol Psychiatry 1990;28:240–6.
[35] Zarcone VP, Benson KL. BPRS symptom factors
and sleep variables in schizophrenia. Psychiatry
Res 1997;66:111–20.
[36] Neylan TC, van Kammen DP, Kelley ME, et al.
Sleep in schizophrenic patients on and off hal-
operidol therapy. Arch Gen Psychiatry 1992;49:
643–9.
[37] Howland RH. Sleep-onset rapid eye movement
periods in neuropsychiatric disorders: implica-
tions for the pathophysiology of psychosis.
J Nerv Ment Dis 1997;185:730–8.
[38] Feinberg I, Koresko RL, Gottlieb F. Further
observations on electrophysiological sleep
patterns in schizophrenia. Compr Psychiatry
1965;6:21–4.
[39] Taylor SF, Tandon R, Shipley JE, et al. Sleep on-
set REM periods in schizophrenic patients. Biol
Psychiatry 1991;30:205–9.

[40] Van Kammen DP, van Kammen WM, Peters J,
et al. Decreased slow-wave sleep and enlarged
lateral ventricles in schizophrenia. Neuropsy-
chopharmacology 1988;1:265–71.
[41] Keshavan MS, Pettegrew JW, Reynolds CF III,
et al. Biological correlates of slow wave sleep
deficits in functional psychoses: 31P-magnetic
resonance spectroscopy. Psychiatry Res 1995;
57:91–100.
[42] Keshavan MS, Miewald J, Haas G, et al. Slow-
wave sleep and symptomatology in schizophre-
nia and related psychotic disorders. J Psychiatr
Res 1995;29:303–14.
[43] Kato M, Kajimura N, Okuma T, et al. Associa-
tion between delta waves during sleep and
negative symptoms in schizophrenia. Neuro-
psychobiology 1999;39:165–72.
[44] Tandon R, DeQuardo JR, Taylor SF, et al. Phasic
and enduring negative symptoms in schizophre-
nia: biological markers and relationship to
outcome. Schizophr Res 2000;45(3):191–201.
[45] Kajimura N, Kato M, Okuma T, et al. Relation-
ship between delta activity during all-night
sleep and negative symptoms in schizophrenia:
a preliminary study. Biol Psychiatry 1996;39(6):
451–4.
[46] Yang C, Winkelman J. Clinical significance of
sleep EEG abnormalities in chronic schizophre-
nia. Schizophr Res 2006;82:251–60.
[47] Keshavan MS, Reynolds CF, Miewald J, et al.
Slow-wave sleep deficits and outcome in schizo-
phrenia and schizoaffective disorder. Acta Psy-
chiatr Scand 1995;91(5):289–92.
[48] Goldman M, Tandon R, DeQuardo JR, et al. Bi-
ological predictors of 1-year outcome in schizo-
phrenia in males and females. Schizophr Res
1996;21:65–73.
[49] Manoach DA, Cain MS, Vangel MG, et al. A fail-
ure of sleep-dependent procedural learning in
chronic, medicated schizophrenia. Biol Psychi-
atry 2004;56:951–6.
[50] Forest G, Poulin J, Daoust AM, et al. Attention
and non-REM sleep in neuroleptic-naive per-
sons with schizophrenia and control partici-
pants. Psychiatry Res 2007;149:33–40.
[51] Göder R, Boigs M, Braun S, et al. Impairment of
visuospatial memory is associated with de-
creased slow wave sleep in schizophrenia. J Psy-
chiatr Res 2004;38:591–9.
[52] Göder R, Aldenhoff JB, Boigs M, et al. Delta
power in sleep in relation to neuropsychologi-
cal performance in healthy subjects and schizo-
phrenia patients. J Neuropsychiatry Clin
Neurosci 2006;18(4):529–35.
[53] Benson KL, Sullivan EV, Lim KO, et al. Slow wave
sleep and CT measures of brain morphology in
schizophrenia. Psychiatry Res 1996;60:125–34.
[54] Keshavan MS, Reynolds CF III, Ganguli R, et al.
Electroencephalographic sleep and cerebral
morphology in functional psychosis: a prelimi-
nary study with computed tomography. Psychi-
atry Res 1991;39:293–301.
Sleep in Schizophrenia 259
[55] Gross CG. Neurogenesis in the adult brain:
death of a dogma. Nat Rev Neurosci 2000;1:
67–73.
[56] Corson PW, Nopoulos P, Miller DD, et al.
Change in basal ganglia volume over 2 years
in patients with schizophrenia: typical versus
atypical neuroleptics. Am J Psychiatry 1999;
156:1200–4.
[57] Dazzan P, Morgan KD, Orr K, et al. Different
effects of typical and atypical antipsychotics
on grey matter in first episode psychosis: the
ǼSOP study. Neuropsychopharmacology
2005;30:765–74.
[58] Newton SS, Duman RS. Neurogenic actions of
atypical antipsychotic drugs and therapeutic
implications. CNS Drugs 2007;21(9):715–25.
[59] Guzman-Marin R, Suntsova N, Methippara M,
et al. Sleep deprivation suppresses neurogenesis
in the adult hippocampus of rats. Eur J Neuro-
sci 2005;22:2111–6.
[60] Riemann D, Hohagen F, Krieger S, et al. Cholin-
ergic REM induction test: muscarinic supersen-
sitivity underlies polysomnographic findings
in both depression and schizophrenia. J Psy-
chiatr Res 1994;28(3):195–210.
[61] Benson KL, Faull KF, Zarcone VP. Evidence for the
role of serotonin in the regulation of slow wave
sleep in schizophrenia. Sleep 1991;14:133–9.
[62] Nishino S, Ripley B, Mignot E, et al. CSF hypo-
cretin-1 levels in schizophrenia and controls:
relationship to sleep architecture. Psychiatry
Res 2002;110:1–7.
[63] Conley RR, Kelly DL. Clinical pharmacology
and medication-associated side effects: a review
of second-generation antipsychotic for schizo-
phrenia. Clinical Schizophrenia and Related
Psychoses 2007;1(2):135–46.
[64] Kaplan J, Dawson S, Vaughan T, et al. Effect of
prolonged chlorpromazine administration on
the sleep of chronic schizophrenics. Arch Gen
Psychiatry 1974;31:62–6.
[65] Taylor SF, Tandon R, Shipley JE, et al. Effect of
neuroleptic treatment on polysomnographic
measures in schizophrenia. Biol Psychiatry
1991;30:904–12.
[66] Nofzinger EA, van Kammen DP,
Gilbertson MW, et al. Electroencephalographic
sleep in clinically stable schizophrenic patients:
two-weeks versus six-weeks neuroleptic free.
Biol Psychiatry 1993;33:829–35.
[67] Keshavan MS, Reynolds CF III, Miewald JM,
et al. A longitudinal study of EEG sleep in
schizophrenia. Psychiatry Res 1996;59:203–11.
[68] Maixner S, Tandon R, Eiser A, et al. Effects of
antipsychotic treatment on polysomnographic
measures in schizophrenia: a replication and ex-
tension. Am J Psychiatry 1998;155(11):1600–2.
[69] Wetter TC, Lauer CJ, Gillich G, et al. The electro-
encephalographic sleep pattern in schizo-
phrenic patients treated with clozapine or
classical antipsychotic drugs. J Psychiatr Res
1996;30(6):411–9.
260 Benson
[70] Hinze-Selch D, Mullington J, Orth A, et al. Ef-
fects of clozapine on sleep: a longitudinal study.
Biol Psychiatry 1997;42:260–6.
[71] Lee JH, Woo JI, Meltzer HY. Effects of clozapine
on sleep measures and sleep-associated changes
in growth hormone and cortisol in patients
with schizophrenia. Psychiatry Res 2001;103:
157–66.
[72] Salin-Pascual RJ, Herrera-Estrella M, Galicia-
Polo L, et al. Olanzapine acute administration
in schizophrenic patients increases delta sleep
and sleep efficiency. Biol Psychiatry 1999;46:
141–3.
[73] Müller MJ, Rossbach W, Mann K, et al. Sub
chronic effects of olanzapine on sleep EEG in
schizophrenic patients with predominantly
negative symptoms. Pharmacopsychiatry 2004;
37(4):157–62.
[74] Yamashita H, Morinobu S, Yamawaki S, et al.
Effect of risperidone on sleep in schizophrenia:
a comparison with haloperidol. Psychiatry Res
2002;109(2):137–42.
[75] Luthringer R, Staner L, Noel N, et al. A double-
blind, placebo controlled, randomized study
evaluating the effect of paliperidone extended-
release tablets on sleep architecture in patients
with schizophrenia. Int Clin Psychopharmacol
2007;22:299–308.
[76] Cohrs S, Rodenbeck A, Guan Z, et al. Sleep-pro-
moting properties of quetiapine in healthy sub-
jects. Psychopharmacology (Berl) 2004;174(3):
421–9.
[77] Lieberman JA, Stroup TS, McEvoy JP, et al. Effec-
tiveness of antipsychotic drugs in patients with
chronic schizophrenia. N Engl J Med 2005;
353(12):1209–23.
[78] Robinson S, Rosca P, Durst R, et al. Serum mel-
atonin levels in schizophrenic and schizoaffec-
tive hospitalized patients. Acta Psychiatr Scand
1991;84(3):221–4.
[79] Monteleone P, Natale M, LaRocca A, et al. De-
creased nocturnal secretion of melatonin in
drug-free schizophrenics: no change after sub-
chronic treatment with antipsychotics. Neuro-
psychobiology 1997;36:159–63.
[80] Shamir E, Laudon M, Barak Y, et al. Melatonin
improves sleep quality of patients with chronic
schizophrenia. J Clin Psychiatry 2000;61:
373–7.
[81] Kumar PNS, Andrade C, Bhakta SG, et al. Mela-
tonin in schizophrenic outpatients with insom-
nia: a double-blind, placebo-controlled study.
J Clin Psychiatry 2007;68(2):237–41.
[82] Makela EH, Miller K, Cutlip WD. Three case
reports of modafinil use in treating sedation
induced by antipsychotic medications. J Clin
Psychiatry 2003;64(4):485–6.
[83] Rosenthal MH, Bryant SL. Benefits of adjunct
modafinil in an open-label, pilot study in
patients with schizophrenia. Clin Neurophar-
macol 2004;27(1):38–43.
[84] Narendran R, Young CM, Valenti AM, et al. Is
psychosis exacerbated by modafinil? Arch Gen
Psychiatry 2002;59(3):292–3.
[85] Charney DS, Kales A, Soldatos CR, et al. Som-
nambulistic-like episodes secondary to com-
bined lithium-neuroloeptic treatment. Br J
Psychiatry 1979;135:418–24.
[86] Kolivakis TT, Margolese HC, Beauclair L, et al.
Olanzapine-induced somnambulism [letter].
Am J Psychiatry 2001;158(7):1158.
[87] Goldbloom D, Chouinard G. Clonazepam in
the treatment of neuroleptic-induced somnam-
bulism [letter]. Am J Psychiatry 1984;141:1486.
[88] Horiguchi J, Yamashita H, Mizuno S, et al. Noc-
turnal eating/drinking syndrome and neurolep-
tic-induced restless legs syndrome. Int Clin
Psychopharmacol 1999;14(1):33–6.
[89] Paquet V, Strul J, Servais L, et al. Sleep-related
eating disorder induced by olanzapine [letter].
J Clin Psychiatry 2002;63(7):597.
[90] Lu ML, Shen WW. Sleep-related eating disorder
induced by risperidone [letter]. J Clin Psychia-
try 2004;65(2):273.
[91] Benson KL, Zarcone VP. Sleep abnormalities in
schizophrenia and other psychotic disorders.
Review of Psychiatry 1994;13:677–705.
[92] Ancoli-Israel S, Martin J, Jones DW, et al. Sleep-dis-
ordered breathing and periodic limb movements
in sleep in older patients with schizophrenia. Biol
Psychiatry 1999;45(11):1426–32.
[93] Takahashi KI, Shimizu T, Sugita T, et al. Preva-
lence of sleep-related respiratory disorders in
101 schizophrenic patients. Psychiatry Clin
Neurosci 1998;52:229–31.
[94] Winkelman JW. Schizophrenia, obesity, and
obstructive sleep apnea. J Clin Psychiatry
2001;62(1):8–11.
[95] Wirshing DA, Pierre JM, Wirshing WC. Sleep
apnea associated with antipsychotic-induced
obesity. J Clin Psychiatry 2002;63:369–70.
[96] Boufidis S, Kosmidis MH, Bozikas VP, et al.
Treatment outcome of obstructive sleep apnea
syndrome in a patient with schizophrenia:
case report. Int J Psychiatry Med 2003;33(3):
305–10.
[97] Karanti A, Landén M. Treatment refractory psy-
chosis remitted upon treatment with continu-
ous positive airway pressure: a case report.
Psychopharmacol Bull 2007;40(1):113–7.
[98] Kang SG, Lee HJ, Jung SW, et al. Characteristics
and clinical correlated of restless legs syndrome
in schizophrenia. Prog Neuropsychopharmacol
Biol Psychiatry 2007;31:1078–83.
[99] Kraus T, Schuld A, Pollmächer T. Periodic leg
movements in sleep and restless legs syndrome
probably caused by olanzapine. J Clin Psycho-
pharmacol 1999;19:478–9.
[100] Wetter TC, Brunner J, Bronisch T. Restless legs
syndrome probably induced by risperidone
treatment. Pharmacopsychiatry 2002;35(3):
109–11.

Sleep and Anxiety Disorders
Thomas A. Mellman, MD
- Sleep in specific and social phobias
- Sleep in obsessive-compulsive disorder
- Sleep in generalized anxiety disorder
- Sleep in panic disorder
Sleep panic attacks
Sleep disturbances and anxiety symptoms are in-
extricably intertwined. With insomnia, anxious
arousal interferes with sleep onset. Insufficient
sleep sustains and predisposes to persisting anxiety
states. Anxiety disorders are psychiatric conditions
whose primary features are anxiety that is persis-
tent, maladaptively triggered, and of sufficient in-
tensity to disrupt function. The anxiety disorders
in the Diagnostic and Statistical Manual, 4th Edition
(DSM-IV) are generalized anxiety, panic, posttrau-
matic stress, obsessive-compulsive, and phobic dis-
orders. Sleep disturbances are frequently associated
with, and can comprise core features of, anxiety
disorders.
Posttraumatic stress disorder (PTSD) and gener-
alized anxiety disorder (GAD) feature sleep distur-
bances among their DSM-IV diagnostic criteria.
PTSD develops in some individuals after exposure
to severely threatening stress and manifests with
symptoms of re-experiencing the trauma, emo-
tional numbing and avoidance behaviors, and
heightened arousal. Specific criteria for the disorder
related to sleep include nightmares with trauma-
related content and difficulty initiating and main-
taining sleep, which is the common definition of
insomnia. The principal feature of GAD is chronic
worry and tension. Impaired sleep initiation and
maintenance is also a symptom criterion for GAD.
Panic disorder features recurring severe and
Department of Psychiatry, Howard University Hospital, 2041 Georgia Avenue, NW, Washington, DC 20060,
USA
E-mail address: tmellman@howard.edu
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
261
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 261–268
- Sleep in posttraumatic stress disorder
- Sleep anxiety symptoms and sleep apnea
- Treatment and prevention
- Acknowledgments
- References
unpredictable episodes of anxiety with crescendo-
like onsets called ‘‘panic attacks,’’ which are often
complicated by anticipatory anxiety and phobic
avoidance. Although not among the specific diag-
nostic criteria, panic disorder has also been associ-
ated with complaints of difficulty initiating and
maintaining sleep in many studies. In addition,
panic attacks can arise from sleep in many patients
diagnosed with the disorder. Sleep disturbances can
occur with, but seem to be less salient features of,
obsessive-compulsive disorder (OCD) and specific
and social phobic disorders.
In addition frequently to being a part of their pre-
senting symptoms, insomnia is a risk factor for the
subsequent onset of anxiety disorders [1–3]. There
is overlap between interventions that target insom-
nia and other sleep disturbances and those that are
used in treating anxiety disorders. Overlapping
approaches include medications, and cognitive
behavioral strategies that target worry, tension,
and maladaptive cognitions. Optimal sequencing
or integration of treatments targeting insomnia
and sleep disturbance, however, are not well
investigated.
Much of the emphasis regarding overlap of anxi-
ety and sleep disturbance is appropriately focused
on insomnia. There can be overlapping features
of the sleep manifestations of anxiety disorders
with other primary sleep disorders. For example,
doi:10.1016/j.jsmc.2008.01.010
262 Mellman
symptomatic episodes in panic disorder sometimes
need to be differentiated from clinical manifesta-
tions of sleep apnea, and GAD from restless legs
syndrome.
In the following sections clinical issues and labo-
ratory information where available regarding sleep
aspects of specific anxiety disorders are reviewed.
This review is followed by discussion of interfaces
of anxiety and sleep apnea and treatment issues
that overlap sleep and anxiety disorders.
Sleep in specific and social phobias
Fear and avoidance of situations are the key features
of phobic disorders. Because these situations occur
during interactions with the environment during
wakefulness, sleep disturbances are not typically re-
garded as central to or commonly associated with
these conditions. Nonetheless, persons with phobic
disorders may experience anticipatory anxiety that
affects their sleep and dreams. Investigations relat-
ing sleep to phobias are limited. In one study, per-
sons with social phobia subjectively reported had
poorer sleep quality, longer sleep latency, more
frequent sleep disturbance, and increased daytime
dysfunction compared with controls [4]. The one
pilot study of social phobia identified that used pol-
ysomnography (PSG), however, reported normal
findings [5]. Clark and colleagues [6] noted that
sleep architecture was similar in depressed persons
with and without simple phobias (the term that
predated the DSM-IV). A study on parasomnias,
including sleep terrors and sleepwalking, among
adolescents found increased comorbidity with
simple phobias and other anxiety disorders [7].
Sleep in obsessive-compulsive disorder
Sleep disturbances are also not included among the
symptom criteria, and are not commonly associated
with OCD. PSG has been applied to OCD and other
disorders to evaluate overlap with depression where
reduced latency to rapid eye movement (REM) sleep
is a well-established biologic marker. An early poly-
somnographic study noted impaired sleep mainte-
nance and a reduced latency to REM sleep in
a group with persons with OCD, which is consistent
with a linkage between OCD and affective illness
[8]. Two more recent polysomnographic studies of
persons with OCD failed to replicate these results,
however, reporting instead that the sleep patterns
of persons with OCD were essentially normal [9,10].
Sleep in generalized anxiety disorder
There is a high degree of overlap between GAD and
insomnia. DSM-IV criteria for GAD are chronic
worry and three of six additional criteria that in-
clude difficulty initiating or maintaining sleep, or
restless and unsatisfying sleep. Two of the other
symptom criteria, fatigue and irritability, can be
consequences of sleep loss. In addition, the princi-
pal attribute of GAD, excessive worry or apprehen-
sive expectation, is commonly implicated in the
genesis and maintenance of insomnia problems.
Ohayon and colleagues [3] found that the co-
morbidity of GAD and insomnia was greater than
for all of the other psychiatric disorders surveyed.
Studies using objective sleep recordings corroborate
the reported associations of GAD and insomnia by
demonstrating impaired sleep initiation and main-
tenance in persons with GAD [11–13]. High comor-
bidity with major depression has generated interest
in comparing biologic markers of the disorders.
Latency to REM sleep was normal in these studies,
in contrast to findings from major depression
where REM sleep latency is reduced [11–13].
Consistent with their high degree of overlap and
comorbidity, there is also substantial overlap of
treatment approaches for GAD and insomnia.
Overlapping approaches include the use of medica-
tions that target benzodiazepine receptors and psy-
chotherapeutic interventions that target excessive
worry. Application of these approaches in treating
co-occurring generalized anxiety and sleep distur-
bances is discussed later.
Sleep in panic disorder
Panic attacks are distinguished from other anxiety
episodes by their sudden, crescendo-like onset, in-
tensity and number of symptoms, and at times
unpredictable pattern of occurrence. Panic attacks
can emerge from sleep. Panic disorder also typically
features chronic anxiety related to anticipating
subsequent attacks and phobic avoidance (agora-
phobia). Panic attacks arising from sleep (sleep
panic attacks) have been suggested to condition
fear and apprehension of sleep resulting in second-
ary insomnia [14]. Surveys document that insom-
nia is more frequent in patients with panic
disorder than in control populations [15,16].
Most [17–20] but not all [21,22] published studies
of panic disorder that used objective methods of
sleep recording (PSG) have found evidence of
impaired sleep initiation and maintenance.
Survey data have noted associations between
sleep complaints and comorbid depression in per-
sons with panic disorder [21]. There are several pos-
sible explanations for the relationship between
sleep disturbance and the presence of depression
in persons with panic disorder. First, much of the
associated sleep disturbance may be caused by de-
pressive illness that commonly coexists with panic

disorder. The two studies that failed to identify any
impairment in sleep duration and maintenance
specifically excluded depression. One of the studies
documenting sleep disturbance in panic disorder
excluded depressive illness, however, and it is un-
certain whether depression accounted for the entire
sleep disturbance documented in the remaining
positive studies. A second consideration is that co-
morbid depression may be more common with
a more severe variant of panic disorder that also fea-
tures sleep disturbance. Because insomnia is a risk
factor for the subsequent onset of depression
[1,2], a third possibility is that depression is more
likely to evolve as a comorbid condition when
panic disorder features disturbed sleep.
Sleep panic attacks
Sleep panic attacks are not uncommon among per-
sons with panic disorder. In a study that prospec-
tively monitored panic attacks, 18% occurred
during sleep hours [23]. In surveys and clinical eval-
uations 33% to 71% of panic disorder patients re-
ported having experienced sleep panic attacks
[15,24–26]. As many as a third of panic disorder
patients experience sleep panic as often as or
more frequently than wake panic attacks [14,15].
It is not known how common it is for patients
only to have sleep panic; however, in this author’s
experience patients who exclusively panic from
sleep are rare. Sleep panic attacks have been de-
scribed as being awakened with a jolt. They also fea-
ture apprehension and somatic symptoms, similar
to panic attacks that are triggered during wake
states. Studies that have captured sleep panic attacks
during polysomnographic recordings find that the
episodes were preceded by either stage 2 or 3 of
non-REM sleep [17,27]. Mellman and Uhde [17]
more specifically noted that the sleep panic attacks
originated during the transition from stage 2 into
early slow wave sleep, which is a period of dimin-
ishing arousal. Slow wave sleep is also a state where
cognitive activity is at a relative nadir [28]. Panic be-
ing triggered during sleep might seem counterintu-
itive in view of the more intuitive circumstance of
panic attacks evolving from states of heightening
arousal where apprehension is building. The phe-
nomenon of sleep panic indicates that panic attacks
can also be precipitated during states of diminish-
ing arousal. This phenomenon adds to evidence
from pharmacologic challenge and treatment stud-
ies and twin and familial genetic data that endoge-
nous neurobiologic mechanisms can underlie
anxiety. Specific mechanisms postulated to underlie
sleep panic include increased sensitivity to in-
creased carbon dioxide blood levels [29], irregular
breathing during slow wave sleep [30], and re-
bound noradrenergic surges [17,18]. A cognitive
Sleep and Anxiety Disorders 263
mechanism of sensitivity to and catastrophic inter-
pretation of interoceptive stimuli has also been
suggested to underlie sleep panic [25].
The greater sensitivity of panic disorder patients
to pharmacologic challenges that induce panic
has provided an important research paradigm for
investigating the psychobiology of panic attacks.
Sodium lactate and pentagastrin challenges, which
trigger panic attacks from wake states, have been
demonstrated also to trigger panic attacks from
sleep [31,32]. Greater cardiac and respiratory re-
sponses to lactate infusion during sleep absent
panic awakenings have also been noted [18,33].
Findings that panicogenic triggers can elicit attacks
from sleep states indicate that elevated basal arousal
is not required for experimentally inducing panic.
Several investigations have explored the signifi-
cance of sleep panic by comparing patients who ex-
perience sleep panic attacks with patients who only
experience panic attacks from wake states. These
studies have indicated that patients with sleep panic
have early illness onset, higher symptom load, de-
pression, and suicidal ideation. Sleep panic may
be associated with a more severe variant of the
illness [24,34]. Patients with sleep panic have also
been noted to experience anxiety from relaxation
and hypnosis, and to have less agoraphobic avoid-
ance and fewer catastrophic cognitions compared
with panic patients who do not experience sleep
panic [15,25,35,36]. Having sleep panic seems
also to mark a propensity to have panic triggered
by reductions in arousal and for attacks to occur rel-
atively independently of situational and cognitive
stimuli that are associated with nonsleep panic.
Sleep in posttraumatic stress disorder
The trauma-related nightmares and difficulty initi-
ating and maintaining sleep denoted in the DSM-
IV criteria for PTSD are often prominent among
the symptom complaints of patients with the disor-
der [37,38]. Nightmares and insomnia are also
common in the early aftermath of trauma, espe-
cially among those who are developing PTSD
[39–42]. Furthermore, sleep disruption leads to fa-
tigue and irritability, which are daytime symptoms
of PTSD. Sleep disruption may also interfere with
healthy emotional adaptation and regulation and
thereby contribute to the development of PTSD.
Findings from sleep laboratory studies have not
yielded a consensus regarding the fundamental na-
ture of sleep disturbances in PTSD. All but a few of
the studies are focused on the chronic phase of the
disorder and many include only male war veterans.
These studies have been mixed in terms of finding
objective indices of impaired sleep initiation and
maintenance [43,44].
264 Mellman
Evidence for abnormalities related to REM sleep
in PTSD has been more consistent. Increased phasic
motor activity and eye movement density during
REM sleep have been reported in combat veterans
with PTSD [45–47]. Nightmares and other symp-
tomatic awakenings disproportionately arise from
REM sleep [48,49]. Breslau and colleagues [50] re-
cently reported more frequent transitions from
REM sleep to stage 1 or wake in a community
sample with either lifetime only (ie, remitted) or
current PTSD compared with trauma-exposed and
trauma-unexposed controls. There is converging
evidence for disruptions of REM sleep continuity
(symptomatic awakenings, increased awakening
and arousals, and motor activity) and increased
REM activation (eye movement density) with
chronic PTSD. The limited number of studies that
used objective recordings of sleep following trauma
also suggest the relevance of REM sleep disruption.
An early report of three cases with ‘‘acute combat
fatigue’’ described ‘‘markedly disrupted sleep’’ and
‘‘rare or absent REM episodes’’ [51]. Mellman and
colleagues [52] reported PSG findings from PSG re-
cordings conducted within a month of trauma in 21
recently injured patients and 10 healthy controls.
REM density was elevated in the recently trauma-
exposed, injured patients compared with healthy
controls but was similar among those who did
and did not develop PTSD. The patients who were
developing PTSD, however, had shorter continuous
periods of REM sleep before stage shifts or arousals.
Findings suggesting a relationship between frag-
mented patterns of REM sleep and PTSD were
also provided by a recent study of PTSD patients
with limited chronicity and comorbidity [53].
Studies have demonstrated that insomnia is very
common among people who have been recently ex-
posed to trauma. Green [39] found insomnia to be
the most frequent symptom endorsed by survivors
in the aftermath of a natural disaster. Koren and col-
leagues [41] found that complaints of insomnia
and excessive daytime sleepiness 1 month after mo-
tor vehicle accidents predicted being diagnosed
with PTSD at 3 months. In contrast, although Koren
and colleagues [41] found an association of early
subjective reports of sleep disturbance with the de-
velopment of PTSD, these investigators did not find
differences in early actigraphic measurements of
sleep initiation or maintenance in their prospective
study of traffic accident victims, nor in PSG mea-
sures in a subgroup recorded 1 year later [54].
Trauma-related nightmares are among the
DSM-IV criteria for PTSD. Mellman and colleagues
[42] evaluated relationships of recalled dream con-
tent elicited within a month of traumatic injuries
with the development of PTSD. Reports of dreams
rated as ‘‘highly similar’’ to the traumatic experience
and distressing were associated with concurrent and
subsequent PTSD severity. The trauma-exposed
group who did not subsequently develop PTSD ei-
ther did not recall dreaming or reported dreams
that did not depict actual memories, although
some represented threatening scenarios. The au-
thors theorized that dreams with highly replicative
content represent a failure of adaptive emotional
memory processing that is a normal function of
REM sleep and dreaming.
A role for noradrenergic functioning in sleep dis-
turbances during the early development of PTSD is
suggested by previously established relationships of
noradrenergic activity with PTSD [55] and PTSD
sleep disturbances [56], and the noradrenergic
signal terminating REM sleep [57]. Mellman and
colleagues [58] also evaluated heart rate variability
during sleep following trauma, which indexes
autonomic regulation of heart rate including sym-
pathetic nervous system activity, which is a periph-
eral manifestation of noradrenergic function. The
index of sympathetic nervous system activity, the
low-frequency/high-frequency ratio, was greater in
the subgroup that developed PTSD during their
initial REM sleep periods.
Subjective sleep complaints are common in the
aftermath of trauma and with PTSD. Nightmares
that are similar to trauma memories seem to be rel-
atively specific to the disorder. Overall, studies do
not indicate that sleep initiation and maintenance
is markedly more impaired among those develop-
ing or who have been diagnosed with PTSD. Several
studies now converge in suggesting that disruptions
of REM sleep may have a role in PTSD and its
development.
Sleep anxiety symptoms and sleep apnea
When a patient reports waking up in the middle of
the night with his or her heart pounding and feeling
short of breath, can this be confidently attributed to
anxiety, or should the physician first evaluate
cardiac and respiratory parameters before even con-
sidering psychiatric disorders as the likely primary
diagnosis? Clinical experience and review of rele-
vant literature indicate that clinical acumen and
common sense can often guide the direction of
the evaluation and that there is a place for physio-
logic monitoring and treatment of comorbid
medical and psychiatric symptoms. With sleep ap-
nea, arousals are usually below the threshold of
awareness and memory, although awakenings
with gasps, snorts, or symptoms of gastroesopha-
geal reflux are sometimes reported. Excessive day-
time sleepiness and reports of loud snoring are
reliable signs of the diagnosis [59]. Sleep anxiety-
related episodes have characteristic features that

have been previously described. The differential
diagnosis between primary sleep disorders and
sleep-related features of anxiety disorders, which
characteristically manifest with abrupt awakenings
to high levels of arousal, can often be made with
reasonable confidence on clinical grounds based
on their features and associated clinical findings.
It is also the case that anxiety disorders with sleep
manifestations not infrequently co-occur with pri-
mary sleep disorders. Some co-occurrence is inevi-
table because anxiety disorders and primary sleep
disorders are both common. To the best of this au-
thor’s knowledge there is not currently evidence
available from large community samples that indi-
cate whether the relationship between sleep apnea
and anxiety disorders is greater than expected by
random association. An analysis from the large
Veteran’s Administration Healthcare database,
however, did indicate an increased association of
diagnoses of ‘‘anxiety’’ and PTSD (and other psychi-
atric disorders) with sleep apnea [60].
There is some evidence that detection and treat-
ment of anxiety disorder, sleep breathing disorder
comorbidity is relevant to patient outcomes. One
study reported very high rates of sleep breathing
disorders (apnea and upper airway resistance syn-
drome) in a group of female research participants
seeking treatment for PTSD related to sexual assault
[61]. A study that recruited PTSD cases from a com-
munity sample, however, did not find elevated rates
of sleep apnea or other primary sleep disorders
[50]. Krakow and colleagues [62] have described
clinically significant improvement of PTSD symp-
toms with treatment of sleep breathing disorders.
Edlund and colleagues [63] have similarly reported
frequent associations of sleep panic attacks and
sleep apnea and response of nocturnal anxiety
symptoms to continuous positive airway pressure
treatment. That panic attacks are exacerbated by
interruptions of respiration is consistent with
observations of increased sensitivity to anxiogenic
effects of carbon dioxide in panic disorder [29]
and the ‘‘suffocation alarm’’ hypothesis of the etiol-
ogy of the disorder [29]. Indirect support for the
hypothesis that increased carbon dioxide receptor
sensitivity underlies forms of pathologic anxiety
includes the observation that children with congen-
ital central hypoventilation syndrome (Ondine’s
curse) have lower rates of anxiety symptoms than
age-matched children [64].
Treatment and prevention
Sleep disturbances are commonly associated with
anxiety disorders, particularly GAD, panic, and
PTSD. In contrast to melancholic subtypes of de-
pression where mood can paradoxically improve,
Sleep and Anxiety Disorders 265
anxiety disorders do not benefit, and can worsen
from sleep deprivation [65–67]. Insomnia has
also been found to be a prospective risk factor for
psychiatric disorders including anxiety disorders
[1,2]. In addition to alleviating distress from insom-
nia, amelioration of sleep disturbances could possi-
bly have therapeutic impact on other symptoms
and serve to prevent relapse and exacerbation.
Therapies for anxiety disorders and sleep distur-
bances overlap. Cognitive behavioral treatments
that were developed for insomnia have well-estab-
lished efficacy [68]. In addition to recommenda-
tions for maintaining consistent bedtimes and
wake times, avoidance of maladaptive use of sub-
stances, and not spending excessive time awake in
bed, effective cognitive behavioral interventions
for insomnia often include components that are
also used in the treatment of anxiety [69]. These
include relaxation techniques and identifying
and challenging dysfunctional beliefs that per-
petuate symptoms. Given the overlapping use of
anxiety management, exposure, and cognitive re-
structuring it seems that behavioral interventions
designed for insomnia and anxiety disorders can
be synergistically applied. One study documents
improvement in insomnia symptoms in associa-
tion with cognitive behavioral treatment of GAD
[70]. In contrast, DeViva and colleagues [71] iden-
tified a group of patients with significant residual
insomnia who had otherwise benefited from cog-
nitive behavioral treatment for PTSD. They further
describe a series of these cases where the residual
insomnia was reduced by a subsequently adminis-
tered cognitive behavioral intervention focused
specifically on the insomnia. A technique where
recurrent distressing dream content is the target
of exposure and cognitive restructuring (nightmare
imagery rehearsal) has been found to ameliorate
nightmares and sleep disruption with PTSD [72].
These observations notwithstanding, development
and evaluation of sequential or integrated treat-
ments for insomnia and anxiety disorders has
been limited.
Various benzodiazepine receptor agonist medica-
tions are approved and marketed for hypnotic indi-
cations or treatment of anxiety disorders,
particularly for GAD. One study found agents that
are marketed and approved as hypnotics had bene-
fits toward daytime anxiety in treating insomnia
associated with GAD (zopiclone and triazolam are
not approved by the Food and Drug Administration
[FDA] for GAD and zopiclone is not marketed in
the United States) [73]. There is preliminary evi-
dence that the novel agents pregabalin and tiage-
bine, which also target benzodiazepine receptors
or related GABAergic neurotransmission, benefit in-
somnia symptoms associated with GAD, although
266 Mellman
neither have FDA approval for GAD or insomnia
[74,75].
The newer antidepressant medications, particu-
larly those in the selective serotonin reuptake inhib-
itor and selective serotonin and norepinephrine
inhibitor categories, have become established as
effective for a range of anxiety disorders. They also
have advantages with respect to tolerance and de-
pendence concerns relative to benzodiazepines
and are now considered to be first-line treatments
for panic disorder, social phobia, GAD, OCD, and
PTSD. The effects on sleep of these agents vary be-
tween agents and to a greater degree between indi-
viduals and some have been noted to stimulate
insomnia [76]. Among the novel antidepressants
mirtazipine, which is neither a selective serotonin
reuptake inhibitor nor a selective serotonin and
norepinephrine inhibitor, tends to have sedating
and sleep effects. Mirtazapine was recently reported
to have been beneficial to GAD patients in a prelim-
inary open label trial [77], although it is not FDA
approved for GAD or insomnia. A study using the
selective serotonin reuptake inhibitor citalopram
for late-life anxiety disorders indicated improve-
ment in subjective sleep quality with treatment in
this subpopulation [78].
Presently, selective serotonin reuptake inhibitors,
specifically sertraline and paroxetine, are the only
agents approved by the FDA for the treatment of
PTSD. Benefits of these treatments tend to be mod-
est and do not typically include reductions in sleep
disturbance. Adjunctive interventions are often
used, often with the intent of targeting nightmare
and insomnia symptoms [79]. Among these, there
is support from controlled trials for adjunctive pre-
scription of olanzapine [80] and prazosin [81],
which are not FDA approved for PTSD. The emerg-
ing use of prazosin for this indication is consistent
with the role of noradrenergic stimulation in dis-
rupting REM sleep suggested by previously reviewed
research.
Acknowledgments
The author acknowledges Denver Brown for her
assistance in preparing this article.
References
[1] Ford D, Kamerow D. Epidemiologic study of
sleep disturbances and psychiatric disorders: an
opportunity for prevention? JAMA 1989;262:
1479–84.
[2] Breslau N, Roth T, Rosenthal L, et al. Sleep distur-
bance and psychiatric disorders: a longitudinal
epidemiological study of young adults. Biol Psy-
chiatry 1996;39:411–8.
[3] Ohayon M, Caulet M, Lemoine P. Comorbidity
of mental and insomnia disorders in the general
population. Compr Psychiatry 1998;39:185–97.
[4] Stein M, Kroft C, Walker J. Sleep impairment in
patients with social phobia. Psychiatry Res
1993;49:251–6.
[5] Brown T, Black B, Uhde T. The sleep architecture
of social phobia. Biol Psychiatry 1994;35:420–1.
[6] Clark C, Gillin J, Golshan S. Do differences in
sleep architecture exist between depressives
with comorbid simple phobia as compared
with pure depressives? J Affect Disord 1995;33:
251–5.
[7] Gau S, Soong W. Psychiatric comorbidity of ado-
lescents with sleep terrors or sleepwalking:
a case-control study. Aust N Z J Psychiatry
1999;33:734–9.
[8] Insel T, Gillin J, Moore A, et al. The sleep of
patients with obsessive-compulsive disorder.
Arch Gen Psychiatry 1982;39:1372–7.
[9] Robinson D, Walsleben J, Pollack S, et al.
Nocturnal polysomnography in obsessive-com-
pulsive disorder. Psychiatry Res 1998;80:257–63.
[10] Hohagan F, Lis S, Krieger S, et al. Sleep EEG of
patients with obsessive-compulsive disorder.
Eur Arch Psychiatry Clin Neurosci 1994;243:
273–8.
[11] Saletu-Zyhlarz G, Saletu B, Anderer P, et al. Non-
organic insomnia in generalized anxiety disor-
der: controlled studies on sleep, awakening and
daytime vigilance utilizing polysomnography
and EEG mapping. Neuropsychobiology 1997;
36:117–29.
[12] Arriaga F, Paiva T. Clinical and EEG sleep
changes in primary dysthymia and generalized
anxiety: a comparison with normal controls.
Neuropsychobiology 1990–1991;24:109–14.
[13] Papdimitriou G, Kerkhofs M, Kempenaers C,
et al. EEG sleep studies in patients with general-
ized anxiety disorder. Psychiatry Res 1988;26:
183–90.
[14] Mellman T, Uhde T. Patients with frequent sleep
panic: clinical findings and response to medica-
tion treatment. J Clin Psychiatry 1990;51:513–6.
[15] Mellman T, Uhde T. Sleep panic attacks: new
clinical findings and theoretical implications.
Am J Psychiatry 1989;146:1204–7.
[16] Stein M, Chartier M, Walker J. Sleep in nonde-
pressed patients with panic disorder: I. System-
atic assessment of subjective sleep quality and
sleep disturbance. Sleep 1993;16:724–6.
[17] Mellman T, Uhde T. Electroencephalographic
sleep in panic disorder: a focus on sleep-related
panic attacks. Arch Gen Psychiatry 1989;46:
178–84.
[18] Sloan E, Natarajan M, Baker B, et al. Nocturnal
and daytime panic attacks: comparison of sleep
architecture, heart rate variability, and response
to sodium lactate challenge. Biol Psychiatry
1999;45:1313–20.
[19] Lydiard R, Zealberg J, Laraia M, et al. Electroen-
cephalography during sleep of patients with

panic disorder. J Neuropsychiatry Clin Neurosci
1989;1:372–6.
[20] Arriaga F, Paiva T, Matos-Pires A, et al. The sleep
of non-depressed patients with panic disorder:
a comparison with normal controls. Acta Psy-
chiatr Scand 1996;93:191–4.
[21] Stein M, Enns M, Kryger M. Sleep in nonde-
pressed patients with panic disorder. II. Polysom-
nographic assessment of sleep architecture and
sleep continuity. J Affect Disord 1993;28:1–6.
[22] Uhde T, Roy-Byrne P, Gillin J, et al. The sleep of
patients with panic disorder. Psychiatry Res
1984;12:251–9.
[23] Taylor C, Skeikh J, Agras S, et al. Ambulatory
heart rate changes in patients with panic attacks.
Am J Psychiatry 1986;143:478–82.
[24] Krystal J, Woods S, Hill C, et al. Characteristics of
panic attack subtypes: assessment of spontane-
ous panic, situational panic, sleep panic, and
limited symptom attacks. Compr Psychiatry
1991;32:474–80.
[25] Craske M, Lang A, Rowe M, et al. Presleep attri-
butions about arousal during sleep: nocturnal
panic. J Abnorm Psychol 2002;111:53–62.
[26] Shapiro C, Sloan E. Nocturnal panic: an underre-
cognized entity. J Psychosom Res 1998;44:21–3.
[27] Hauri P, Freidman M, Ravaris C, et al. Sleep in
patients with spontaneous panic attacks. Sleep
1989;12:323–37. Paper presented at 139th An-
nual Meeting, American Psychiatric Association,
Washington, DC, 1986.
[28] Hobson JA, Pace-Schott E, Stickgold R. Dream-
ing and the brain: towards a cognitive neurosci-
ence of conscious states. Behav Brain Sci 2000;
23:793–842.
[29] Klein D. False suffocation alarms, spontaneous
panics, and related conditions: an integrative
hypothesis. Arch Gen Psychiatry 1993;50:
306–17.
[30] Stein M, Millar T, Larsen D, et al. Irregular
breathing during sleep in patients with panic
disorder. Am J Psychiatry 1995;152:1168–73.
[31] Koenigsberg H, Pollack C, Ferro D. Can panic be
induced in deep sleep? Examining the necessity
of cognitive processing for panic. Depress Anxi-
ety 1998;8:126–30.
[32] Geraci M, Anderson T, Slate-Cothren S, et al.
Pentagastrin-induced sleep panic attacks: panic
in the absence of elevated baseline arousal.
Biol Psychiatry 2002;52:1183–9.
[33] Koenigsberg H, Pollack C, Fine J, et al. Cardiac
and respiratory activity in panic disorder: effects
of sleep and sleep lactate infusions. Am J Psychi-
atry 1994;151:1148–52.
[34] Labbate L, Pollack M, Otto M, et al. Sleep panic
attacks: an association with childhood anxiety
and adult psychopathology. Biol Psychiatry
1994;36:57–60.
[35] Tsao J, Craske M. Fear of loss of vigilance: devel-
opment and preliminary validation of a self-
report instrument. Depress Anxiety 2003;18:
177–86.
Sleep and Anxiety Disorders 267
[36] Craske M, Lang A, Mystkowski J, et al. Does
nocturnal panic represent a more severe form
of panic disorder? J Nerv Ment Dis 2002;190:
611–8.
[37] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
edition. Washington, DC: American Psychiatric
Press; 1994.
[38] Neylan T, Marmar C, Metzler T, et al. Sleep dis-
turbances in the Vietnam generation: findings
from a nationally representative sample of
male Vietnam veterans. Am J Psychiatry 1998;
155:929–33.
[39] Green B. Disasters and posttraumatic stress dis-
order. In: Davidson JRT, Foa EB, editors. Post-
traumatic stress disorder DSM-IV and beyond.
Washington, DC: American Psychiatric Press;
1993. p. 75–97.
[40] Mellman T, David D, Kulick-Bell R, et al. Sleep
disturbance and its relationship to psychiatric
morbidity following Hurricane Andrew. Am
J Psychiatry 1995;152:1659–63.
[41] Koren D, Arnon I, Lavie P, et al. Sleep complaints
as early predictors of posttraumatic stress
disorder: a 1-year prospective study of injured
survivors of motor vehicle accidents. Am
J Psychiatry 2002;159:855–7.
[42] Mellman T, David D, Bustamante V, et al.
Dreams in the acute aftermath of trauma and
their relationship to PTSD. J Trauma Stress
2001;14:241–7.
[43] Mellman TA. Sleep and the pathogenesis of
PTSD. In: Shalev A, Yehuda R, McFarlane AC,
editors. International handbook of human
response to trauma. New York: Plenum Publish-
ing Company; 2000. p. 299–306.
[44] Lavie P. Current concepts: sleep disturbances in
the wake of traumatic events. N Engl J Med
2001;345:1825–32.
[45] Ross R, Ball W, Dinges D, et al. Rapid eye move-
ment sleep disturbance in posttraumatic stress
disorder. Biol Psychiatry 1994;35:195–202.
[46] Ross R, Ball W, Dinges D, et al. Motor dysfunction
during sleep in posttraumatic stress disorder.
Sleep 1994;17:723–32.
[47] Mellman T, Nolan B, Hebding J, et al. A poly-
somnographic comparison of veterans with
combat-related PTSD, depressed men, and non-
ill controls. Sleep 1996;20:46–51.
[48] Mellman T, Kulick-Bell R, Ashlock L, et al. Sleep
events among veterans with combat-related post-
traumatic stress disorder. Am J Psychiatry 1995;
152:110–5.
[49] Woodward S, Arsenault N, Santerre C, et al. Pol-
ysomnographic characteristics of trauma-related
nightmares. Presented at the Annual Meeting of
Association of Professional Sleep Societies. Las
Vegas (Nevada), June, 10–15 2000.
[50] Breslau N, Roth T, Burduvali E, et al. Sleep in life-
time posttraumatic stress disorder: a commu-
nity-based polysomnographic study. Arch Gen
Psychiatry 2004;61:508–16.
268 Mellman
[51] Schlosberg A, Benjamin M. Sleep patterns in
three acute combat fatigue cases. J Clin Psychia-
try 1978;39:546–9.
[52] Mellman T, Bustamante V, Fins A, et al. REM
sleep and the early development of posttrau-
matic stress disorder. Am J Psychiatry 2002;
159:1696–701.
[53] Habukawa M, Uchimura N, Maeda M, et al.
Sleep findings in young adult patients with post-
traumatic stress disorder. Biol Psychiatry 2007;
62:1179–82.
[54] Klein E, Koren D, Arnon I, et al. Sleep com-
plaints are not corroborated by objective sleep
measures in post-traumatic stress disorder:
a 1-year prospective study in survivors of motor
vehicle crashes. J Sleep Res 2003;12:35–41.
[55] Southwick S, Bremner J, Rasmusson A, et al. Role
of norepinephrine in the pathophysiology and
treatment of posttraumatic stress disorder. Biol
Psychiatry 1999;46:1192–204.
[56] Mellman T, Kumar A, Kulick-Bell R, et al. Norad-
renergic and sleep measures in combat-related
PTSD. Biol Psychiatry 1995;38:174–9.
[57] Hobson J, McCarley R, Wyzinski P. Sleep cycle
oscillation: reciprocal discharge by two brain-
stem neuronal groups. Science 1975;189:55–8.
[58] Mellman T, Knorr B, Pigeon W, et al. Heart rate
variability during sleep and the early develop-
ment of PTSD. Biol Psychiatry 2004;55:953–6.
[59] Vaughn B, D’Cruz O. Cardinal manifestations of
sleep disorders. In: Kryger M, Roth T, Dement W,
editors. Principles and practice of sleep medi-
cine. Pennsylvania (PA): Elsevier Saunders;
2005. p. 594–601.
[60] Sharafkhaneh A, Giray N, Richardson P, et al.
Association of psychiatric disorders and sleep
apnea in a large cohort. Sleep 2005;28:1405–11.
[61] Krakow B, Melendrez D, Johnston L, et al. Sleep-
disordered breathing, psychiatric distress, and
quality of life impairment in sexual assault survi-
vors. J Nerv Ment Dis 2002;190:442–52.
[62] Krakow B, Melendrez D, Pederson B, et al. Com-
plex insomnia: insomnia and sleep-disordered
breathing in a consecutive series of crime victims
with nightmares and PTSD. Biol Psychiatry 2001;
49:948–53.
[63] Edlund M, McNamara M, Millman R. Sleep
apnea and panic attacks. Compr Psychiatry
1991;32:130–2.
[64] Pine D, Weese-Mayer D, Silvestri J, et al. Anxiety
and congenital central hypoventilation syn-
drome. Am J Psychiatry 1994;151:864–70.
[65] Labbate L, Johnson M, Lydiard R, et al. Sleep
deprivation in panic disorder and obsessive-com-
pulsive disorder. Can J Psychiatry 1997;42:982–3.
[66] Labbate L, Johnson M, Lydiard R, et al. Sleep
deprivation in social phobia and generalized
anxiety disorder. Biol Psychiatry 1998;43:840–2.
[67] Roy-Byrne P, Uhde T, Post R. Effects of one
night’s sleep deprivation on mood and behavior
in panic disorder: patients with panic disorder
compared with depressed patients and normal
controls. Arch Gen Psychiatry 1986;43:895–9.
[68] Morin C, Culbert J, Schwartz S. Nonpharmaco-
logical interventions for insomnia: a meta-analy-
sis of treatment efficacy. Am J Psychiatry 1994;
151:1172–80.
[69] Falsetti SA, Combs-Lane A, Davis JL. Cognitive
behavioral treatment of anxiety disorders. In:
Nutt DJ, Ballenger JC, editors. Anxiety disorders.
Oxford UK: Blackwell Science; 2003. p. 425–44.
[70] Belanger L, Morin CM, Langlois F, et al. Insom-
nia and generalized anxiety disorder: effects of
cognitive behavior therapy for gad on insomnia
symptoms. J Anxiety Disord 2004;18:561–71.
[71] DeViva J, Zayfert C, Pigeon W, et al. Treatment of
residual insomnia after CBT for PTSD: case stud-
ies. J Trauma Stress 2005;18:155–9.
[72] Krakow B, Hollifield M, Johnston L, et al. Imag-
ery rehearsal therapy for chronic nightmares in
sexual assault survivors with posttraumatic stress
disorder: a randomized controlled trial. JAMA
2001;286:537–45.
[73] Fontaine R, Beaudry P, Le Morvan P, et al. Zopi-
clone and triazolam in insomnia associated with
generalized anxiety disorder: a placebo-con-
trolled evaluation of efficacy and daytime anxi-
ety. Int Clin Psychopharmacol 1990;5:173–83.
[74] Rickels K, Pollack M, Feltner D, et al. Pregabalin
for treatment of generalized anxiety disorder:
a 4-week, multicenter, double-blind, placebo-
controlled trial of pregabalin and alprazolam.
Arch Gen Psychiatry 2005;62:1022–30.
[75] Rosenthal M. Tiagabine for the treatment of
generalized anxiety disorder: a randomized,
open-label, clinical trial with paroxetine as
a positive control. J Clin Psychiatry 2003;64:
1245–9.
[76] Winokur A, Gary KA, Rodner S, et al. Depression,
sleep physiology, and antidepressant drugs.
Depress Anxiety 2001;14:19–28.
[77] Gambi F, De Berardis D, Campanella D, et al.
Mirtazapine treatment of generalized anxiety
disorder: a fixed dose, open label study. J Psycho-
pharmacol 2005;19:483–7.
[78] Blank S, Lenze EJ, Mulsant B, et al. Outcomes of
late-life anxiety disorders during 32 weeks of
citalopram treatment. J Clin Psychiatry 2006;
67:468–72.
[79] Freidman M, Davidson J, Mellman T, et al.
Guidelines for treatment of PTSD: pharmaco-
therapy. J Trauma Stress 2000;13:563–6.
[80] Stein M, Kline N, Matloff J. Adjunctive olanza-
pine for SSRI-resistant combat-related PTSD:
a double-blind, placebo-controlled study. Am
J Psychiatry 2003;160:1189–90.
[81] Raskind M, Peskind E, Kanter E, et al. Reduction
of nightmares and other PTSD symptoms in
combat veterans by prazosin: a placebo-con-
trolled study. Am J Psychiatry 2003;160:371–3.
 269

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 269–279

Behavioral Sleep Disorders

in Children and Adolescents
a, b,c
Lisa J. Meltzer, PhD *, Jodi A. Mindell, PhD

- Assessment of sleep disorders Nighttime fears

Sleep history - Sleep and psychiatric disorders in children
Further assessment and adolescents
- Behavioral sleep disorders Attention-Deficit–Hyperactivity Disorder
Behavioral insomnia of childhood Autism
Insufficient sleep and inadequate sleep Depression
hygiene Anxiety
Insomnia - Summary
Circadian rhythm disorder, delayed sleep - References
phase type

Sleep is essential, accounting for approximately Assessment of sleep disorders

40% of a child’s typical day. When children and ad-
olescents do not get enough sleep, aspects of their
physical, emotional, cognitive, and social develop-
ment are negatively affected. Furthermore, the clin-
ical symptoms of medical and psychiatric disorders
are likely to be worsened if a child has a sleep prob-
lem. Sleep problems in children and adolescents
are quite common (estimated prevalence of 25%–
40% [1]) and can be chronic; however; sleep prob-
lems are also highly treatable.
This article reviews the assessment, features,
prevalence, and treatment of behavioral sleep dis-
orders in children and adolescents. In addition,
behavioral sleep issues that are frequently experi-
enced by children and adolescents with common
psychiatric disorders are discussed.
In children and adolescents, not all sleep problems
meet criteria for a disorder based on diagnostic cri-
teria of the International Classification of Sleep Disor-
ders, 2nd Edition (ICSD-2) [2] or the Diagnostic and
Statistical Manual, 4th Edition [3]; however, they are
of significance and should always be considered
as part of any medical or psychiatric evaluation. It
is important for all health care specialists who reg-
ularly interact with children and adolescents (eg,
pediatricians, psychiatrists, psychologists, nurses)
to assess for sleep concerns. Most often, informa-
tion about sleep problems is provided by the parent
or caregiver, rather than the child or adolescent;
however, older children and adolescents can
provide important additional information.

a
Division of Pulmonary Medicine, The Children’s Hospital of Philadelphia and University of Pennsylvania
School of Medicine, 3535 Market Street, 14th Floor, Philadelphia, PA 19104, USA
b
Department of Psychology, Saint Joseph’s University, Philadelphia, PA 19131, USA
c
The Sleep Center at The Children’s Hospital of Philadelphia, 34th & Civic Center Boulevard, Wood Building,
5th Floor, Philadelphia, PA 19104, USA
* Corresponding author.
E-mail address: meltzerl@email.chop.edu (L.J. Meltzer).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.004
sleep.theclinics.com
270 Meltzer & Mindell
Furthermore, given that a child’s day-to-day experi-
ence is dependent on the systems around him or
her (parents, teachers, peers), input from all in-
volved is important to provide a complete assess-
ment of the sleep issues. Such an assessment that
integrates information from multiple sources also
supports treatment success, given the need for open
communication with all family members. Thus
a comprehensive assessment allows practitioners to
understand better the nature of the child’s sleep
problems, the impact of the sleep problems on day-
time functioning, and to set feasible treatment goals.
Sleep history
A thorough sleep history covers all aspects of
a child’s sleep patterns and behaviors related to
sleep. Questions should focus on (1) the child’s
or adolescent’s sleep schedule, including bedtime,
sleep-onset latency, wake time, and variations in
the sleep schedule on nonschool nights (eg, week-
ends, summer, holidays); (2) the sleep routine
and sleep environment, including the consistency
of the bedtime routine, if the bedroom is shared,
the presence of cosleeping, and any technology
(eg, television, computer) in the bedroom; (3) bed-
time behaviors, including bedtime stalling or re-
fusal, inability to fall asleep independently, and
anxiety or fears at bedtime; (4) nocturnal behaviors,
such as snoring, pauses in breathing, the frequency
and duration of night wakings, nightmares, sleep
terrors, and enuresis; and (5) daytime behaviors, in-
cluding daytime functioning, naps, caffeine intake,
and medications. For this last factor, it is important
to recognize that daytime sleepiness often presents
differently across developmental age groups. For ex-
ample, young children may seem more energetic
when sleepy, whereas older children and adoles-
cents may be more likely to be moody, fatigued,
and withdrawn.
Along with sleep and daytime behaviors, a psy-
chosocial history provides information about po-
tential aspects of the child’s social, academic, and
family life that may affect sleep. For example,
changes or stressors in a child’s life can have a signif-
icant impact on both their sleep quality and quan-
tity. For young children, common disruptive events
are the birth of a new sibling or a sudden change to
their environment or routine (eg, a parent returning
to work). School-age children and adolescents may
experience sleep problems related to a death in the
family, concerns about being bullied or getting
poor grades, or marital discord between the child’s
parents. Finally, for children of all ages who ‘‘worry’’
more than their peers, sleep onset and maintenance
may be affected by thoughts of peers, school, family
relations, or current events (eg, September 11th,
Iraq war).
Further assessment
Although parents and often children or adolescents
provide a significant amount of information about
sleep difficulties, further information can be gath-
ered with the use of sleep diaries, actigraphy, or pol-
ysomnography (PSG). Sleep diaries track bedtime,
wake times, sleep-onset latency, night wakings,
and daytime naps over a period of 1 to 2 weeks.
One advantage of a sleep diary over the sleep his-
tory is that parents or patients complete them daily,
allowing for an assessment of night-to-night vari-
ability and other inconsistencies in sleep patterns
over an extended period of time. This variability
can be a significant contributor to some behavioral
sleep disorders, such as insufficient sleep, insomnia,
and delayed sleep phase syndrome.
Actigraphy also provides an objective assessment
of sleep patterns over an extended period (typically
1–2 weeks) that can be used to obtain additional
information about sleep patterns, especially night-
time awakenings, or it can be used as an alternative
for families who are poor historians. This small de-
vice is the size of a wristwatch, and is worn on the
child’s nondominant wrist (or ankle in very young
children). Using an internal motion detector, actig-
raphy has been shown reliably to distinguish
between sleep and wake in children and adoles-
cents [4–6]. The sleep-wake patterns data gathered
by actigraphy can assist with differential diagnoses
for behavioral sleep disorders, guide treatment deci-
sions, and measure treatment effectiveness. Draw-
backs of actigraphy include the cost of the units,
and the requirement of a valid daily sleep diary to
assist with the interpretation of the results.
PSG is considered the gold standard for the
assessment of sleep stages and physiologic sleep dis-
orders (eg, obstructive sleep apnea [OSA], periodic
limb movement disorder [PLMD]). PSG is less ef-
fective, however, in the diagnosis of behavioral
sleep disorders. PSG most commonly involves a sin-
gle night assessment in a sleep laboratory, which
may or may not represent the typical patterns or
problems for children or adolescents with behav-
ioral sleep concerns. Other limitations of PSG
include the cost and limited availability of sleep
laboratories, especially those that are pediatric-
based. However, since behavioral sleep disorders of-
ten have comorbid presentations with other sleep
disorders (eg, OSA, PLMD), PSG can be an impor-
tant adjunct to rule out underlying sleep disruptors
that manifest as daytime sleepiness or irritability.
Behavioral sleep disorders
In general, behavioral sleep disorders present with
at least one of the following complaints: (1) bed-
time problems, including bedtime stalling or

resistance; (2) difficulties falling asleep; (3) fre-
quent or prolonged night wakings; (4) early morn-
ing wakings; or (5) excessive daytime sleepiness.
Despite these common symptoms, the causes, diag-
noses, and treatments for behavioral sleep disorders
vary depending on the nature of the disorder and
the child’s age. The next section reviews the preva-
lence, causes, and recommended treatment
approaches for behavioral sleep disorders in chil-
dren and adolescents.
Behavioral insomnia of childhood
Behavioral insomnia of childhood (BIC) presents
with complaints of bedtime problems or night wak-
ings [2]. Across cultures, the prevalence of these
sleep problems is 20% to 30% [7–10]. There are
three subtypes of BIC, with the sleep difficulties
linked to an identified behavior in the parent or
child.
Sleep-onset association type
A sleep-onset association is an environmental con-
dition required for the child to fall asleep at bed-
time and return to sleep following normal
nighttime arousals. Without the sleep-onset associ-
ation, the child may have a prolonged sleep-onset
latency and frequent night wakings. There are two
types of sleep associations: positive and negative.
A positive sleep association is a condition that the
child can create independently (eg, thumb sucking,
cuddle object). A negative sleep association is a con-
dition that can require another individual (eg, the
parent who nurses or rocks an infant to sleep or
lays next to a toddler until the child is asleep) or ex-
ternal stimuli (eg, riding in the car, television on).
Because all children typically arouse two to six
times per night [11], any condition present at bed-
time is required again following a naturally occur-
ring arousal. When the association is present, the
child returns to sleep quickly during the night. Be-
cause negative sleep associations involve parental
assistance, these associations result in frequent
night wakings and sometimes prolonged periods
of wakefulness.
Self-soothing without the need for parental assis-
tance to fall asleep is a developmental skill that typ-
ically occurs between 3 and 6 months [9]; therefore
it is not appropriate to diagnose BIC sleep-onset as-
sociation type before 6 months of age. BIC sleep-
onset association type is seen most commonly in
infants and toddlers (6 months–3 years). Some
negative associations naturally cease with time as
conditions that facilitate sleep are faded out (eg,
weaning from nursing). If negative associations
continue (eg, rocking instead of nursing), however,
frequent and persistent night wakings also con-
tinue. Behavioral treatments (described in the next
Behavioral Sleep Disorders 271
section), including extinction, graduated extinc-
tion, and positive routines with faded bedtime,
have been found to be highly efficacious in the
treatment of BIC sleep-onset association type [12].
Limit-setting type
Bedtime refusal or bedtime stalling (at an age-ap-
propriate bedtime) is the defining feature of BIC
limit-setting type [2]. Bedtime refusal is when
a child refuses to get ready for bed, go to bed, or
stay in bed, often involving temper tantrums and
resulting in a delayed bedtime. Bedtime stalling is
an attempt to delay bedtime, and manifests as re-
peated requests for additional activities (eg, another
television show, one more book or trip to the bath-
room) or attention (eg, another hug). Once the
child falls asleep, he or she has normal sleep qual-
ity, although the delayed sleep onset often results
in decreased sleep quantity. Bedtime problems
have been reported in 10% to 30% of toddlers,
and up to 15% of school-aged children (4–10 years)
[1].
Just as sleeping through the night is a develop-
mental skill, the behaviors associated with BIC
limit-setting type are also related to normal child
development. Toddlers and preschoolers (the
most common populations diagnosed with BIC
limit-setting type) are learning to navigate the
world by testing limits and exerting their newfound
independence during the day and at bedtime. As
long as parents set no limits (eg, allow child to
chose the bedtime or fall asleep in front of the tele-
vision) or inconsistent limits (eg, some nights the
child is allowed to fall asleep in own bed, other
nights allowed to fall asleep in parents’ bed), bed-
time problems persist. If limits are consistently
set, the child usually falls asleep quickly and easily.
As children get older, parental involvement with the
bedtime routine decreases, limiting the opportunity
for problematic behaviors at bedtime. Behavioral
interventions have been shown to be highly effica-
cious for the treatment of bedtime problems in
young children, as discussed later [12].
Combined type
Some children may experience both BIC sleep-on-
set association type and BIC limit-setting type, thus
a separate diagnosis of BIC-combined type is in-
cluded in the revised ICSD [2]. A typical example
of BIC-combined type is a child who stalls, makes
multiple requests, and has tantrums at bedtime re-
sulting in a prolonged sleep-onset latency (limit-
setting type), with a parent finally lying with the
child in his or her bed at which time the child falls
asleep quickly (sleep-onset association type). Dur-
ing the night, the child has frequent night wakings
when he or she needs the parent to lay with him or
272 Meltzer & Mindell
her until the child returns to sleep (sleep-onset
association type).
Treatment of behavioral insomnia
of childhood
There are a number of efficacious treatments for
BIC [12], yet basic sleep hygiene is required for all
children with this sleep disorder. First, children
should have an age-appropriate bedtime (typically
7:00–8:30 PM). If the child’s bedtime is too late,
they often become overtired, resulting in hyperac-
tivity and emotion disregulation [13]. Naps are
also an essential aspect of sleep for young children,
with most children napping until age 3, and 26%
until age 5 [14]. Skipping or withholding naps in
a younger child does not facilitate an easier sleep
onset at bedtime, but rather can result in the child
becoming overtired and having more difficulty
falling asleep. Second, children should have a
consistent bedtime routine that is short (20–30
minutes) and involves the same three to four activ-
ities every night [15]. Third, parents must be consis-
tent every night in terms of their management of
the child’s bedtime behavior.
Recently, the American Academy of Sleep Medi-
cine published standards of practice documents
on behavioral treatments of bedtime problems
and night wakings in young children [16]. The cor-
nerstone of these empirically supported behavioral
interventions is having children fall asleep indepen-
dently [12]. Infants should be placed in the crib
drowsy but awake, and children should fall asleep
in their own crib or bed at bedtime without a parent
present. To achieve this, the most common
approach is an intervention based on extinction
(a behavioral technique based on operant condi-
tioning). Although standard extinction (or ‘‘cry it
out’’) is the fastest treatment approach, it is not tol-
erated well by most parents [17,18]. Modified ver-
sions of this approach (graduated extinction,
faded parental presence) have been developed and
are found to be effective [19–21]. Graduated extinc-
tion is the approach most commonly recommen-
ded in popular parenting publications [22,23].
Other treatment approaches for BIC include (1)
scheduled awakenings, where the child is woken
approximately 15 minutes before a typical night-
time waking for approximately 10 days; (2) parent
preventive education, through written materials or
classes that can prevent the onset of sleep problems
in infants; and (3) positive routines with faded bed-
time and response cost, where the child’s bedtime is
delayed to coincide with their regular sleep-onset
time. Bedtime is preceded by an enjoyable routine
that ends immediately if the child begins to tan-
trum. The bedtime is then advanced gradually in
15-minute increments over a period of several
weeks.
Insufficient sleep and inadequate sleep
hygiene
Insufficient sleep, or getting inadequate sleep rela-
tive to the child’s sleep need, is a significant prob-
lem for youth of all ages. A national survey of
sleep in children ages 0 to 10 years found that
45% to 59% of children are sleeping less than
what is typically recommended for their age [14].
Preadolescents and adolescents reported averaging
7.6 hours of sleep on school nights, with less than
20% of respondents obtaining the recommended
9 hours per night in this age group [24].
Over time, insufficient sleep results in chronic
partial sleep deprivation, which leads to a number
of negative daytime consequences, including exces-
sive daytime sleepiness, mood disturbances, behav-
ior problems, cognitive impairment, and increased
risk-taking behaviors [13,25,26]. In adolescents, in-
sufficient sleep can result in drowsy driving, which
can be fatal to these new drivers if they fall asleep
at the wheel. Over 60% of juniors and seniors
have reported driving drowsy at least once in the
previous year, with 15% reporting drowsy driving
at least once in the prior week. Further, 3% of ju-
niors and 9% of seniors reported nodding off or
falling asleep behind the wheel in the past year
[24].
Multiple factors contribute to insufficient sleep.
For children and adolescents, the most common
are academic and extracurricular demands, social
activities, part-time employment; electronics (eg,
television, computer, video games), and early
school start times. For youth of all ages, inadequate
sleep hygiene can also contribute to insufficient
sleep.
There are two types of behaviors that contribute
to inadequate sleep hygiene: practices that increase
arousal (eg, caffeine, rough play, or watching televi-
sion at bedtime), and inconsistent sleep organiza-
tion (eg, naps late in the day, irregular sleep-wake
schedule, excessive time in bed relative to actual
time asleep). Prolonged inadequate sleep hygiene
may result in insomnia (see next section on insom-
nia). Causes of inadequate sleep hygiene include in-
adequate parental supervision of bedtime and sleep
behaviors, and insufficient education about sleep
needs and appropriate sleep behaviors.
Interventions for insufficient sleep and inade-
quate sleep hygiene involve changes to a child’s
daily routine and sleep-related behaviors. To in-
crease sleep time, parents may need to weigh the
benefits of multiple extracurricular activities with
the costs of insufficient sleep. In addition, a number
of schools have begun to change school start times,

resulting in increased sleep time for junior high stu-
dents [27] and high school students [28]. Children
of all ages should have a consistent and age-appro-
priate sleep-wake schedule. This schedule includes
no more than 1- to 2-hour differences between
weekday and weekend bedtimes and wake times.
Bedrooms should be sleep conducive, including be-
ing comfortable, cool, dark, and quiet. In addition,
all technology should be removed from the bed-
room. Youth of all ages should avoid eating or
drinking products with caffeine (eg, soda, iced tea,
coffee, energy drinks, chocolate), especially in the
late afternoon and evening. Finally, naps should
be encouraged as appropriate for age and develop-
mental stage. Regular naps are important for infants
and toddlers, whereas children and adolescents
who are unable to obtain sufficient sleep at night
may benefit from a 30- to 45-minute nap in the
early afternoon.
Insomnia
The hallmark features of insomnia are difficulties
initiating and maintaining sleep [2]. Because of de-
velopmental considerations, the ICSD-2 definition
of pediatric insomnia is the ‘‘repeated difficulty
with sleep initiation, duration, consolidation, or
quality that occurs despite age-appropriate time and
opportunity for sleep and results in functional im-
pairment for the child and/or family.’’ This differs
from insomnia in adults in that the complaints of
insomnia may not come from the child, but can
be reported by the parent. Further, the age-appro-
priate bedtime takes into consideration factors for
BIC, and circadian factors seen in adolescents (see
next section on delayed-sleep phase). Estimates of
insomnia in children and adolescents range from
6% to 39% depending on the definition used
[29–32]. Insomnia is more prevalent in girls than
boys postpuberty, but few racial differences have
been found [30,32].
Insomnia is a symptom of many psychiatric dis-
orders and is associated with a number of medical
conditions. A thorough history is needed to deter-
mine the cause of the insomnia before deciding
on a treatment approach. When insomnia is not re-
lated to a psychiatric or medical disorder, the two
primary factors that contribute to the insomnia
are maladaptive sleep behaviors and negative cogni-
tions (beliefs and attitudes) about sleep. Insomnia
typically results from a combination of predispos-
ing factors (eg, genetic vulnerability, underlying
medical or psychiatric conditions) and perpetuat-
ing factors (poor sleep habits, caffeine use, malad-
aptive cognitions) [33]. Treatment focuses on
changing the maladaptive sleep behaviors and
negative sleep cognitions. Cognitive-behavioral
Behavioral Sleep Disorders 273
treatment for insomnia typically includes a combi-
nation of the following interventions [34]:
 Sleep hygiene: Good sleep hygiene includes
having a consistent and appropriate bedtime,
avoiding caffeine, maintaining an appropriate
sleep environment, and having a consistent
wake time regardless of the amount of sleep
achieved the previous night.
 Stimulus control: Individuals are instructed
that if they are unable to fall asleep at bedtime
or return to sleep following night wakings
within 20 to 30 minutes, they should get out
of bed and engage in a quiet activity (eg, read-
ing), only returning to bed when they feel
sleepy. This may need to be repeated multiple
times before sleep occurs.
 Sleep restriction: After determining an individ-
ual’s current sleep quantity, he or she is in-
structed to limit the amount of time in bed to
the number of hours they are currently sleep-
ing. The goal of both stimulus control and
sleep hygiene is to improve sleep efficiency,
consolidate sleep, and disrupt the negative as-
sociation between not sleeping and being in
bed.
 Cognitive restructuring: Through challenging
and reframing negative cognitions that inter-
fere with sleep, this intervention strives to
shorten sleep-onset latency and wake after
sleep onset. The three steps of cognitive restruc-
turing are (1) identifying inappropriate sleep
cognitions (eg, I’ll never fall asleep); (2) chal-
lenging the validity of the cognitions (eg, I
did fall asleep last night eventually, and there
has never been a night where I haven’t fallen
asleep at all); and (3) replacing the thoughts
with more realistic and productive cognitions
(eg, I may not fall asleep right away, but even-
tually I will).
 Relaxation techniques: As with standard behav-
ior therapy, these include progressive muscle
relaxation, visual imagery, medication, and
diaphragmatic breathing.
Circadian rhythm disorder, delayed sleep
phase type
Circadian rhythm disorder, delayed sleep phase
type (also known as ‘‘delayed sleep phase syn-
drome’’ [DSPS]) is most commonly seen in adoles-
cents, although occasionally is experienced by
children. The defining feature of DSPS is a sleep-
wake schedule that is significantly and persistently
delayed by 2 or more hours beyond the desired bed-
time, and conflicts with an individual’s activities of
daily living (eg, school, work, scheduled activities).
Once asleep, there are no problems with sleep
274 Meltzer & Mindell
quality [2]. The feature that distinguishes DSPS
from insomnia is that if sleep is attempted at the
time the adolescent typically falls asleep (eg, 3:00
AM), he or she falls asleep quickly. Individuals
with insomnia continue to have difficulties initiat-
ing sleep regardless of bedtime. It has been esti-
mated that approximately 5% to 10% of
adolescents have DSPS [35,36]. DSPS is a multicom-
ponent disorder, caused by genetic, biologic, and
psychosocial factors [37–39]. The most common
clinical presentation of DSPS is a complaint of the
adolescent being awake until the early morning
hours (eg, 3:00 or 4:00 AM), and then being ex-
tremely difficult to wake in the morning.
Treatment for DSPS involves shifting an individ-
ual’s sleep timing, and requires strict maintenance
of a consistent sleep-wake schedule. Adolescents
need to be highly motivated for treatment to be suc-
cessful. There are two approaches that can be used.
In phase advancement, adolescents do not go to
bed until their usual sleep-onset time (eg, 3:00 AM),
and once they are falling asleep quickly each night,
the bedtime is advanced by 15 minutes every few
nights. For phase delay (or chronotherapy), both
bedtime and wake time are delayed for 2 to 3 hours
each day until the desired sleep-wake scheduled is
reached (eg, starting with a 3:00 AM bedtime, day
1:3:00 AM–11:00 AM, day 2:6:00 AM–2:00 PM, day
3:9:00 AM–5:00 PM, and so forth) [40]. For both ap-
proaches, once the desired sleep-wake schedule is
reached, it must be adhered to every night of the
week, including weekends.
Melatonin has also been recommended for the
treatment of DSPS [41–43]. Although dose recom-
mendations have ranged from 0.3 to 5 mg, there
are a number of shortcomings with melatonin.
First, the studies on the effectiveness and potential
side effects for children and adolescents are sparse
and inconclusive. Second, melatonin is only sold
as an over-the-counter supplement, with no regula-
tion for the actual concentration of melatonin in
each dose. Third, there is no clear consensus on
the timing and dosing of melatonin, with recom-
mendations ranging from 30 minutes to 4 hours be-
fore the desired bedtime.
Nighttime fears
Nighttime fears are a normal feature of develop-
ment, with 73% of children ages 4 to 12 experienc-
ing fears at some point [44,45]. The development
of nighttime fears parallels cognitive development
in young children, with imagination, creativity,
and an awareness of ‘‘bad things’’ contributing to
these fears. Nightmares may also contribute to
nighttime fears [45,46]. Along with normal develop-
mental fears, nighttime fears and nightmares may
occur following a traumatic event or be a symptom
of a more severe anxiety disorder. In these cases,
symptoms are also seen during the day. If nighttime
fears persist or cause significant distress for the child
or family, a further evaluation of psychiatric issues is
recommended.
If not addressed, nighttime fears may interfere
with a child’s sleep by delaying sleep onset or pro-
longing nighttime wakings. The presence of a parent
may alleviate these fears at bedtime, but may also
create a negative sleep association. A recent litera-
ture review found that cognitive-behavioral
interventions, such as positive self-talk, positive im-
agery, relaxation, and desensitization, have been
successfully used to address nighttime fears [46].
A key component to treatment may be behavioral
reinforcement [47], with children being rewarded
for making steps toward confronting their fears
and sleeping independently.
Sleep and psychiatric disorders in children
and adolescents
There is a complex and bidirectional relationship
between sleep disturbances and psychiatric disor-
ders in children and adolescents. For example,
insomnia and hypersomnia can be signs of depres-
sion, and sleep disturbances can be a sign of anxiety
[3]. Conversely, sleep disturbances can cause or ex-
acerbate negative mood and psychiatric problems.
Studies of patients referred to pediatric sleep clinics
have found that 31% to 50% of children and ado-
lescents have a diagnosed psychiatric disorder
[48,49]. In one of these centers, 40% of children
and adolescents without a psychiatric diagnosis
had significant psychiatric symptoms based on
both a validated questionnaire and clinical inter-
view [48]. Furthermore, children and adolescents
seen in a mental health clinic have significantly
more sleep complaints (25%–68%) compared
with nonpsychiatric controls (1%–24%) [50]. Sleep
problems are most common in children and ado-
lescents with ADHD, autism, and mood-anxiety
disorders. The following sections review the preva-
lence of sleep problems in these populations, and
identify potential causes and interventions for these
sleep disturbances.
Attention-Deficit–Hyperactivity Disorder
Sleep problems have been reported in 25% to 82%
of children and adolescents with ADHD [51,52].
Prevalence estimates have been found to vary, how-
ever, based on the type of assessment. For example,
significant differences have been found in bedtime
resistance, sleep-onset latency, night wakings, and
total sleep time when comparing parent reports of
children with and without ADHD [51]. Actigraphy
data, however, have shown that when averaged,

sleep continuity variables may not significantly dif-
fer between children and adolescents with and
without ADHD. Rather, there is significantly more
night-to-night variability in children with ADHD
[53]. This variability may contribute to the in-
creased parental reports of sleep problems. Simi-
larly, there are discrepancies in findings across
studies that have examined sleep in children with
ADHD using PSG with some studies noting differ-
ences, and other studies not reporting differences
in sleep between children with and without
ADHD. Overall, the conflicting results of studies
that use different assessment methodologies sug-
gest the need for a comprehensive, multimodal ap-
proach to the assessment of sleep problems in
children and adolescents with ADHD.
There are a number of factors that can contribute
to sleep problems in children and adolescents with
ADHD, including intrinsic sleep disorders, behav-
ior problems at bedtime, medications, and comor-
bid psychiatric disorders [51]. First, children and
adolescents with ADHD have been found to have
more intrinsic sleep disorders (ie, OSA, PLMD)
than children and adolescents without ADHD
[54–57]. Both OSA and PLMD disrupt sleep quality
and total sleep time, resulting in exacerbation of the
daytime behavior problems seen in children and
adolescents with ADHD. When sleep-disordered
breathing is treated, studies have found that symp-
toms of ADHD (eg, hyperactivity, inattention, poor
emotion regulation) improve [58,59]. When PLMD
is treated by dopamine agonists, one study found
both sleep quantity and quality improved [60]. Fur-
ther, signs of ADHD that had previously been resis-
tant to psychostimulants also improved. In terms of
behavioral problems, children and adolescents
with ADHD have more bedtime struggles than chil-
dren without ADHD [50]. It is unclear, however, if
sleep-onset latency differs between these groups of
children [55,61].
Psychostimulants and other medications that are
used to treat ADHD can also contribute to pro-
longed sleep-onset latency and poor sleep quality.
The timing and dosage of these medications need
to be considered when determining if a child’s sleep
problems are intrinsic, behavioral, a result of med-
ications, or a combination of these factors. Further-
more, some medications for ADHD have been
found to have less impact on sleep than others
[62]. Children and adolescents with ADHD com-
monly have comorbid psychiatric diagnoses that
need to be evaluated carefully before the diagnosis
and treatment of a sleep problem. For example,
children with ADHD and comorbid anxiety may
benefit from relaxation strategies, whereas a child
with ADHD and oppositional defiant disorder
may benefit from a specific reinforcement of desired
Behavioral Sleep Disorders 275
behaviors (eg, positive reinforcement for desired
behaviors, ignoring negative behaviors).
Finally, if a behaviorally based sleep problem is
diagnosed (eg, BIC, DSPS), behavioral interven-
tions should be tailored to the child’s individual
needs. For example, consistent bedtime routines
and limit-setting may be required for a child with
ADHD and bedtime resistance.
Autism
It is clear that children with autism have a significant
number of sleep problems. Prevalence estimates
range from 44% to 83%, again with differences
caused by assessment method. Compared with
both typically developing children and children
with intellectual disabilities, children with autism
have a significant number of sleep problems by par-
ent report (difficulty falling asleep, frequent or pro-
longed night wakings, or early morning wakings)
[63–70]. When sleep patterns in children with au-
tism have been assessed with actigraphy, differences
in sleep have not been as pronounced as with par-
ent report [65,71]. Sleep problems in children
with autism have been found to be related to
more energetic, excited, and problematic daytime
behaviors [69], and stereotypic behaviors [72].
The etiology of sleep problems in children with
autism remains to be determined. A number of po-
tential causes, however, can help guide treatment
decisions for behavioral sleep problems in this pop-
ulation. For example, it has been suggested that the
timing of melatonin secretion is altered in children
with autism; thus, exogenous melatonin may help
facilitate sleep onset [73]. More research is needed,
however, to determine the efficacy and safety of
melatonin use in children with autism. An alterna-
tive to melatonin for treatment of circadian-related
sleep problems is light therapy, with bright light
therapy in the morning advancing the secretion of
melatonin, and light therapy late in the day delay-
ing melatonin onset. Other potential etiologies
that may benefit from some type of physiologic or
pharmacologic intervention are abnormal electro-
encephalograms or brain pathology.
Children with autism also experience behavioral
sleep problems including bedtime struggles and the
inability to self-soothe at bedtime, all of which can
prolong sleep onset. As with typically developing
children, bedtime problems and night wakings
that are associated with inconsistent limits or nega-
tive sleep-onset association can be treated with
behavioral interventions, including a consistent
sleep schedule, consistent bedtime routines, and
graduated extinction. Behavioral interventions
need to be tailored to the specific needs of the child,
and take into consideration the age and intellectual
functioning of the child [73]. More research is
276 Meltzer & Mindell
needed on the efficacy of behavioral interventions
for children with autism and sleep problems, espe-
cially in light of findings suggesting that parents
find behavioral interventions more helpful than
medications, even though medication is a far
more common treatment approach [74].
Depression
The prevalence of sleep problems in children and
adolescents with depression ranges from 66% to
90% [75–77]. Sleep disturbances (ie, insomnia, hy-
persomnia) can be a symptom of depression for
children and adolescents [3]. Furthermore, disrup-
ted or insufficient sleep can contribute to, or exacer-
bate, signs of depression [76,78].
A study of sleep in depressed youth that used
actigraphy found poor sleep quality and abnormal
circadian rhythms [79]. There have been mixed
results when sleep in depressed youth has been
measured by PSG [80,81].
Because sleep problems and depression are so
highly related, a multimodal approach to treatment
is often necessary [77]. Treatment may need to
include pharmacology in combination with behav-
ioral techniques described previously for other
sleep disorders, especially insomnia (eg, consistent
sleep routine and schedule, relaxation, cognitive
restructuring). For pharmacologic treatments, the
impact of the medication on sleep should be con-
sidered, because some antidepressants can exacer-
bate sleep problems.
Anxiety
As with the other psychiatric disorders previously
discussed, there is a strong relationship between
sleep and anxiety. In particular, children who are
anxious during the day may have difficulties initiat-
ing sleep because of worries or fears, resulting in
shortened sleep that can heighten signs of anxiety.
Other underlying causes of anxiety can also contrib-
ute to sleep problems. For example, children who
have experienced a traumatic event may experience
excessive arousal, hypervigilance, and fears when
expected to fall asleep alone in a dark room [82].
The treatment of sleep problems in children and
adolescents with anxiety should also be multi-
modal, not only addressing the underlying cause
of the anxiety, but having good sleep hygiene, pos-
itive reinforcement for desired behaviors, and grad-
uated extinction to help children learn to fall asleep
independently. In addition, at bedtime, children
should be made to feel safe and secure.
Summary
Behavioral sleep disorders are common, and if left
untreated can have a significant impact on the
cognitive, social, and emotional functioning of chil-
dren and adolescents. A complete assessment of
sleep patterns, sleep disruptions, psychosocial
factors, and psychiatric disorders is essential to
disentangle the complex and often comorbid
presentation of behavioral sleep disorders. In addi-
tion, age and developmental stage need to be con-
sidered when weighing different diagnoses, and
when selecting an appropriate intervention. Non-
pharmacologic treatments for behavioral sleep dis-
orders have been found to be efficacious and often
preferred, especially by parents of children with
psychiatric disorders. It is suggested that all pediat-
ric health care practitioners consider sleep issues as
part of their comprehensive assessment of all
children.
References
[1] Owens JA. Epidemiology of sleep disorders dur-
ing childhood. In: Sheldon SH, Ferber R,
Kryger MH, editors. Principles and practices of
pediatric sleep medicine. Philadelphia: Elsevier
Saunders; 2005. p. 27–33.
[2] American Academy of Sleep Medicine. Interna-
tional classification of sleep disorders. In: Diag-
nostic and coding manual. 2nd edition.
Westchester (IL): American Academy of Sleep
Medicine; 2005.
[3] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
edition. Washington, DC: American Psychiatric
Association; 1994.
[4] Acebo C, Sadeh A, Seifer R, et al. Estimating sleep
patterns with activity monitoring in children and
adolescents: how many nights are necessary for
reliable measures? Sleep 1999;22:95–103.
[5] Acebo C, Sadeh A, Seifer R, et al. Sleep/wake pat-
terns derived from activity monitoring and
maternal report for healthy 1- to 5-year-old chil-
dren. Sleep 2005;28:1568–77.
[6] Sadeh A, Acebo C. The role of actigraphy in sleep
medicine. Sleep Med Rev 2002;6:113–24.
[7] Armstrong KL, Quinn RA, Dadds MR. The sleep
patterns of normal children. Med J Aust 1994;
161:202–6.
[8] Burnham MM, Goodlin-Jones BL, Gaylor EE,
et al. Nighttime sleep-wake patterns and self-
soothing from birth to one year of age: a longitu-
dinal intervention study. J Child Psychol
Psychiatry 2002;43:713–25.
[9] Goodlin-Jones BL, Burnham MM, Gaylor EE,
et al. Night waking, sleep-wake organization,
and self-soothing in the first year of life. J Dev
Behav Pediatr 2001;22:226–33.
[10] Lozoff B, Wolf AW, Davis NS. Sleep problems seen
in pediatric practice. Pediatrics 1985;75:477–83.
[11] Sadeh A. Assessment of intervention for infant
night waking: parental reports and activity-based
home monitoring. J Consult Clin Psychol 1994;
62:63–8.

[12] Mindell JA, Kuhn BR, Lewin DS, et al. Behavioral
treatment of bedtime problems and night wak-
ings in infants and young children. Sleep 2006;
29:1263–76.
[13] Dahl RE. The impact of inadequate sleep on chil-
dren’s daytime cognitive function. Semin Pediatr
Neurol 1996;3:44–50.
[14] National Sleep Foundation. Sleep in America Poll.
2004. Available at: http://www.sleepfoundation.
org/. Accessed May 11, 2007.
[15] Mindell JA, Luedtke K, Wiegand B, et al. Effects
of a consistent bedtime routine on infant sleep.
Sleep 2007;30:A91.
[16] Morgenthaler TI, Owens J, Alessi C, et al. Practice
parameters for behavioral treatment of bedtime
problems and night wakings in infants and
young children: an American Academy of Sleep
Medicine report. Sleep 2006;29:1415–9.
[17] Rickert VI, Johnson CM. Reducing nocturnal
awakening and crying episodes in infants and
young children: a comparison between sched-
uled awakenings and systematic ignoring. Pedi-
atrics 1988;81:203–12.
[18] Richman N. A double-blind drug trial of
treatment in young children with waking
problems. J Child Psychol Psychiatry 1985;
26:591–8.
[19] Adams LA, Rickert VI. Reducing bedtime tan-
trums: comparison between positive routines
and graduated extinction. Pediatrics 1989;84:
756–61.
[20] Hiscock H, Wake M. Randomised controlled trial
of behavioural infant sleep intervention to
improve infant sleep and maternal mood. BMJ
2002;324:1062–5.
[21] Reid MJ, Walter AL, O’Leary SG. Treatment of
young children’s bedtime refusal and nighttime
wakings: a comparison of standard and gradu-
ated ignoring procedures. J Abnorm Child Psy-
chol 1999;27:5–16.
[22] Mindell JA. Sleeping through the night: how
infants, toddlers, and their parents can get
a good night’s sleep. Revised edition. New
York: Harper Collins; 2005.
[23] Ferber R. Solve your child’s sleep problems. New
York: Fireside; 2006.
[24] National Sleep Foundation. Sleep in America Poll.
2006. Available at: http://www.sleepfoundation.
org/. Accessed March 31, 2006.
[25] O’Brien EM, Mindell JA. Sleep and risk-taking
behavior in adolescents. Behav Sleep Med
2005;3:113–33.
[26] Fallone G, Owens JA, Deane J. Sleepiness in chil-
dren and adolescents: clinical implications.
Sleep Med Rev 2002;6:287–306.
[27] Wolfson AR, Spaulding NL, Dandrow C, et al.
Middle school start times: the importance of
a good night’s sleep for young adolescents. Be-
hav Sleep Med 2007;5:194–209.
[28] Wahlstrom K. Changing times: findings from the
first longitudinal study of later high school start
times. NAASP Bulletin 2002;86:3–21.
Behavioral Sleep Disorders 277
[29] Sadeh A, Raviv A, Gruber R. Sleep patterns and
sleep disruptions in school-age children. Dev
Psychol 2000;36:291–301.
[30] Ohayon MM, Roth T. Place of chronic insomnia
in the course of depressive and anxiety disorders.
J Psychiatr Res 2003;37:9–15.
[31] Blader JC, Koplewicz HS, Abikoff H, et al. Sleep
problems of elementary school children: a com-
munity survey. Arch Pediatr Adolesc Med 1997;
151:473–80.
[32] Johnson EO, Roth T, Schultz L, et al. Epidemi-
ology of DSM-IV insomnia in adolescence:
lifetime prevalence, chronicity, and an emer-
gent gender difference. Pediatrics 2006;117:
e247–56.
[33] Spielman AJ, Caruso LS, Glovinsky PB. A behav-
ioral perspective on insomnia treatment. Psy-
chiatr Clin North Am 1987;10:541–53.
[34] Edinger JD, Means MK. Cognitive-behavioral
therapy for primary insomnia. Clin Psychol Rev
2005;25:539–58.
[35] Thorpy MJ, Korman E, Spielman AJ, et al. De-
layed sleep phase syndrome in adolescents.
J Adolesc Health Care 1988;9:22–7.
[36] Tomoda A, Miike T, Uezono K, et al. A school
refusal case with biological rhythm disturbance
and melatonin therapy. Brain Dev 1994;16:
71–6.
[37] Takahashi Y, Hohjoh H, Matsuura K. Predispos-
ing factors in delayed sleep phase syndrome. Psy-
chiatry Clin Neurosci 2000;54:356–8.
[38] Carskadon MA, Vieira C, Acebo C. Association
between puberty and delayed phase preference.
Sleep 1993;16:258–62.
[39] Crowley SJ, Acebo C, Carskadon MA. Sleep, cir-
cadian rhythms, and delayed phase in adoles-
cence. Sleep Med 2007;8:602–12.
[40] Czeisler CA, Richardson GS, Coleman RM, et al.
Chronotherapy: resetting the circadian clocks of
patients with delayed sleep phase insomnia.
Sleep 1981;4:1–21.
[41] Mundey K, Benloucif S, Harsanyi K, et al.
Phase-dependent treatment of delayed sleep
phase syndrome with melatonin. Sleep 2005;
28:1271–8.
[42] Smits MG, Nagtegaal EE, van der HJ, et al. Mela-
tonin for chronic sleep onset insomnia in chil-
dren: a randomized placebo-controlled trial.
J Child Neurol 2001;16:86–92.
[43] Scheer FA, Cajochen C, Turek FW, et al. Melato-
nin in the regulation of sleep and circadian
rhythms. In: Kryger MH, Roth C, Dement WC,
editors. Principles and practices of sleep medi-
cine. 4th edition. Philadelphia: Elsevier Saun-
ders; 2005. p. 395–404.
[44] Gordon J, King N, Gullone E, et al. Nighttime
fears of children and adolescents: frequency,
content, severity, harm expectations, disclosure,
and coping behaviours. Behav Res Ther 2007;
45:2462–72.
[45] Muris P, Merckelbach H, Gadet B, et al. Fears,
worries, and scary dreams in 4- to 12-year-old
278 Meltzer & Mindell
children: their content, developmental pattern,
and origins. J Clin Child Psychol 2000;29:43–52.
[46] Muris P, Merckelbach H, Ollendick TH, et al.
Children’s nighttime fears: parent-child ratings
of frequency, content, origins, coping behav-
iors and severity. Behav Res Ther 2001;39:
13–28.
[47] Ollendick TH, Hagopian LP, Huntzinger RM.
Cognitive-behavior therapy with nighttime fear-
ful children. J Behav Ther Exp Psychiatry 1991;
22:113–21.
[48] Ivanenko A, Barnes ME, Crabtree VM, et al. Psy-
chiatric symptoms in children with insomnia
referred to a pediatric sleep medicine center.
Sleep Med 2004;5:253–9.
[49] Meltzer LJ, Moore M, Mindell JA. The need for
interdisciplinary pediatric sleep clinics. Behav-
ioral Sleep Medicine, in press.
[50] Ivanenko A, Crabtree VM, Obrien LM, et al.
Sleep complaints and psychiatric symptoms in
children evaluated at a pediatric mental health
clinic. J Clin Sleep Med 2006;2:42–8.
[51] Owens JA. The ADHD and sleep conundrum:
a review. J Dev Behav Pediatr 2005;26:312–22.
[52] Crabtree VM, Ivanenko A, Gozal D. Clinical and
parental assessment of sleep in children with at-
tention-deficit/hyperactivity disorder referred to
a pediatric sleep medicine center. Clin Pediatr
(Phila) 2003;42:807–13.
[53] Gruber R, Sadeh A. Sleep and neurobehavioral
functioning in boys with attention-deficit/
hyperactivity disorder and no reported breath-
ing problems. Sleep 2004;27:267–73.
[54] Melendres MC, Lutz JM, Rubin ED, et al. Day-
time sleepiness and hyperactivity in children
with suspected sleep-disordered breathing. Pedi-
atrics 2004;114:768–75.
[55] O’Brien LM, Holbrook CR, Mervis CB, et al.
Sleep and neurobehavioral characteristics of 5-
to 7-year-old children with parentally reported
symptoms of attention-deficit/hyperactivity dis-
order. Pediatrics 2003;111:554–63.
[56] Picchietti DL, England SJ, Walters AS, et al. Peri-
odic limb movement disorder and restless legs
syndrome in children with attention-deficit
hyperactivity disorder. J Child Neurol 1998;13:
588–94.
[57] Picchietti DL, Underwood DJ, Farris WA, et al.
Further studies on periodic limb movement dis-
order and restless legs syndrome in children with
attention-deficit hyperactivity disorder. Mov Dis-
ord 1999;14:1000–7.
[58] Gozal D. Sleep-disordered breathing and school
performance in children. Pediatrics 1998;102:
616–20.
[59] Ali NJ, Pitson D, Stradlin JR. Sleep disordered
breathing: effects of adenotonsillectomy on
behavior and psychological function. Eur J Pe-
diatr 1996;155:55–62.
[60] Walters AS, Mandelbaum DE, Lewin DS, et al.
Dopaminergic therapy in children with restless
legs/periodic limb movements in sleep and
ADHD. Dopaminergic Therapy Study Group. Pe-
diatr Neurol 2000;22:182–6.
[61] Corkum P, Tannock R, Moldofsky H. Sleep
disturbances in children with attention-deficit/
hyperactivity disorder. J Am Acad Child Psychia-
try 1998;37:637–46.
[62] Sangal RB, Owens J, Allen AJ, et al. Effects of
atomoxetine and methylphenidate on sleep in
children with ADHD. Sleep 2006;29:1573–85.
[63] Allik H, Larsson JO, Smedje H. Insomnia in
school-age children with Asperger syndrome or
high-functioning autism. BMC Psychiatry 2006;
6:18.
[64] Couturier JL, Speechly KN, Steele M, et al. Paren-
tal perception of sleep problems in children of
normal intelligence with pervasive developmen-
tal disorders: prevalence, severity, and pattern.
J Am Acad Child Psychiatry 2005;44:815–22.
[65] Hering E, Epstein R, Elroy S, et al. Sleep patterns
in autistic children. J Autism Dev Disord 1999;
29:143–7.
[66] Patzold LM, Richdale AL, Tonge BJ. An investiga-
tion into sleep characteristics of children with
autism and Asperger’s disorder. J Paediatr Child
Health 1998;34:528–33.
[67] Polimeni MA, Richdale AL, Francis AJP. A survey
of sleep problems in autism, Asperger’s disorder
and typically developing children. J Intellect Dis-
abil Res 2005;49:260–8.
[68] Cotton S, Richdale A. Brief report: parental
descriptions of sleep problems in children with
autism, Down syndrome, and Prader-Willi syn-
drome. Res Dev Disabil 2006;27:151–61.
[69] Richdale AL, Prior MR. The sleep/wake rhythm
in children with autism. Eur Child Adolesc Psy-
chiatry 1995;4:175–86.
[70] Schreck KA, Mulick JA. Parental report of sleep
problems in children with autism. J Autism
Dev Disord 2000;30:127–35.
[71] Wiggs L, Stores G. Sleep patterns and sleep disor-
ders in children with autistic spectrum disorders:
insights using parent report and actigraphy. Dev
Med Child Neurol 2004;46:372–80.
[72] Schreck KA, Mulick JA, Smith AF. Sleep problems
as possible predictors of intensified symptoms of
autism. Res Dev Disabil 2004;25:57–66.
[73] Richdale AL. Sleep problems in autism: preva-
lence, cause, and intervention. Dev Med Child
Neurol 1999;41:60–6.
[74] Wiggs L, Stores G. Sleep problems in children
with severe intellectual disabilities: what help is
being provided? J Appl Res Intelect 1996;9:
159–64.
[75] Ryan ND, Puig-Antich J, Ambrosini P, et al. The
clinical picture of major depression in children
and adolescents. Arch Gen Psychiatry 1987;44:
854–61.
[76] Roberts RE, Lewinsohn PM, Seeley JR. Symp-
toms of DSM-III-R major depression in adoles-
cence: evidence from an epidemiological
survey. J Am Acad Child Psychiatry 1995;34:
1608–17.

[77] Ivanenko A, Crabtree VM, Gozal D. Sleep and
depression in children and adolescents. Sleep
Med Rev 2005;9:115–29.
[78] Breslau N, Roth T, Rosenthal L, et al. Sleep dis-
turbance and psychiatric disorders: a longitudinal
epidemiological study of young adults. Biol Psy-
chiatry 1996;39:411–8.
[79] Sadeh A, McGuire JP, Sachs H, et al. Sleep and
psychological characteristics of children on a psy-
chiatric inpatient unit. J Am Acad Child Adolesc
Psychiatry 1995;34:813–9.
Behavioral Sleep Disorders 279
[80] Puig-Antich J, Goetz R, Hanlon C, et al. Sleep ar-
chitecture and REM sleep measures in prepuber-
tal children with major depression: a controlled
study. Arch Gen Psychiatry 1982;39:932–9.
[81] Young W, Knowles JB, MacLean AW, et al. The
sleep of childhood depressives: comparison
with age-matched controls. Biol Psychiatry
1982;17:1163–8.
[82] Sadeh A. Stress, trauma and sleep in children.
Child Adolesc Psychiatr Clin N Am 1996;5:
685–700.
 281

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 281–293

Sleep and Its Disorders in Seniors

a b,
Carl J. Stepnowsky, Jr., PhD , Sonia Ancoli-Israel, PhD *

- Sleep and aging Sleep-related breathing disorder

- Circadian rhythm disturbances Periodic limb movements in sleep
- Insomnia and restless legs syndrome
Depression Rapid eye movement sleep behavior
Sleep and medical illness disorder
Sleep and medications - Sleep in dementia
Insomnia treatment - Sleep in institutionalized elderly
- Primary sleep disorders - Summary
- References

Over the past decade, knowledge about age-re-
lated changes in sleep has significantly increased.
It is now known that there are both normal, age-
related changes in sleep architecture and sleep pat-
terns, and a variety of sleep complaints and sleep
disorders that increase with age. This article reviews
both normal and abnormal sleep in the elderly.
Sleep and aging
Survey data show that half of elderly individuals re-
port some form of sleep difficulty, including longer
sleep-onset times, lower rates of sleep efficiency,
more time in bed, more awakenings during the
night, earlier wake-up times, and more daytime
naps. Elderly individuals complain primarily about
insomnia, which is often comorbid with other dis-
orders. The symptoms in the elderly are more likely
to be comorbid with an underlying physiologic
problem, rather than with stress as seen in younger
adults.
A number of subjective changes in sleep are expe-
rienced in the elderly:
Increase in time to fall asleep
Spend less time asleep
Increase in number of awakenings
Spend too much time in bed
Less satisfied with nighttime sleep
Significant increase in daytime sleepiness
Napping more often and longer
Objective evidence of these subjective changes in
sleep is corroborated by polysomnography. With
age, sleep becomes more fragmented and lighter
with an increase in the number of arousals and
awakenings. There is a reduction in the amount of
slow wave sleep (stages 3 and 4), beginning in mid-
dle age, with some evidence suggesting that slow
wave sleep is completely absent after the age of 90
[1,2]. There is a compensatory increase in stages
1 and 2, and there is a decrease in rapid eye move-
ment (REM) sleep, which is proportional to the

This work was supported by NIA AG08415, NIA AG15301, NCI CA112035, NIH M01 RR00827, VA IIR 02-275,
and the Research Service of the Veterans Affairs San Diego Healthcare System, and HS17246-01 (CS).
a
Department of Medicine, University of California, San Diego, 3350 La Jolla Village Drive, San Diego,
CA 92161, USA
b
Department of Psychiatry (116A), University of California, 9500 Gilman Drive, La Jolla, CA 92093, USA
* Corresponding author.
E-mail address: sancoliisrael@ucsd.edu (S. Ancoli-Israel).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.011
sleep.theclinics.com
282 Stepnowsky & Ancoli-Israel
decrease in total sleep time. Sleep efficiency and to-
tal sleep time are reduced with age and there are an
increased number of sleep stage shifts. Van Cauter
and colleagues [3] found that total sleep time de-
creased on average by 27 minutes per decade from
mid-life until the eighth decade in a sample of
men aged 16 to 83. Older adults spend more time
in bed but have deterioration in both the quality
and quantity of sleep.
All of these sleep changes can lead to excessive
daytime sleepiness, which in turn can lead to nap-
ping (both intentional and unintentional). Objec-
tive tests (eg, the Multiple Sleep Latency Test) of
daytime sleepiness in the elderly show that they
are sleepier than younger adults [4], suggesting
that the elderly may not be able to obtain an ade-
quate amount of nighttime sleep [5].
It is still not clear whether older adults need less
sleep; however, it is clear that there is a reduced abil-
ity to obtain adequate sleep in this population
[1,6]. This reduced ability can be linked to several
potential causes:
Circadian rhythm changes
Primary sleep disturbances (eg, sleep-related
breathing disorder, periodic limb movements
in sleep, REM sleep-behavior disorder)
Medical illness (eg, hyperthyroidism, arthritis)
Psychiatric illness (eg, depression, anxiety
disorders)
Multiple medications
Dementia
Poor sleep hygiene habits
Effective treatments exist for many of these sleep
difficulties. Given the high prevalence of sleep com-
plaints and sleep disorders in this population and
the link between insufficient sleep and heightened
levels of morbidity and mortality, there is a clear
need for increased awareness, assessment, and treat-
ment of sleep disturbances in the elderly.
Circadian rhythm disturbances
Circadian (24-hour) rhythms are biologic rhythms
or changes that control many physiologic func-
tions, including core body temperature, endoge-
nous hormone secretions, and the sleep-wake
cycle. These rhythms originate in the suprachias-
matic nucleus in the anterior hypothalamus, which
houses the internal circadian pacemaker. The
rhythms are also under the control of external
cues, such as light, time of day, social activities,
and meals. Circadian rhythm sleep disturbances
typically develop when there is a dysynchrony be-
tween the internal circadian pacemaker and exter-
nal environment demands.
Several factors likely contribute to circadian
rhythm desynchronization in the elderly. First, the
suprachiasmatic nucleus deteriorates with age,
which may result in weaker or more disrupted
rhythms [7]. Second, other circadian rhythm distur-
bances known to be involved in the entrainment of
the circadian rhythm of sleep may develop, such as
the gradual reduction of nocturnal secretion of
melatonin with age [8]. The decline in melatonin
secretion may result in reduced sleep efficiency
and an increased incidence of circadian rhythm
sleep disturbances. Third, elderly patients may
have exogenous cues that are too weak to entrain
the circadian rhythm of sleep-wake. For example,
light is one of the most powerful zeitgebers (liter-
ally ‘‘time-giver’’ or cues), yet studies have shown
that elderly patients, especially those who are insti-
tutionalized, spend too little time in daylight. Expo-
sure to daily bright light averages about 1 hour for
healthy elderly, 30 minutes for Alzheimer’s disease
patients living at home, and less than 10 minutes
for nursing home patients [9–12]. This reduced
level of bright light is associated with nighttime
sleep fragmentation and circadian rhythm sleep
disorders [12].
Another common circadian rhythm change in
older age is a shift in the timing of the sleep-wake
cycle. Many older patients experience a phase ad-
vance in their sleep-wake cycle, causing them to
feel sleepy early in the evening. Individuals with ad-
vanced sleep phase syndrome typically fall asleep
between 7 PM and 9 PM and wake up between 3 AM
and 5 AM. Not uncommonly, many older individ-
uals may stay up late in spite of their sleepiness,
yet still awaken early in the morning because of
their advanced sleep-wake cycle. This cycle can
cause sleep deprivation, excessive daytime sleepi-
ness, and subsequent daytime napping. Finally,
the amplitude (ie, the difference in the level be-
tween the peak and trough values) of the circadian
rhythm may also decrease with age, which can in-
crease the frequency of nighttime awakenings and
the severity of excessive daytime sleepiness [13].
Circadian rhythm changes are considered to be
common with age, and presenting symptoms may
mimic those of primary insomnia (discussed later).
Making a distinction between the two disturbances
is important, however, because each warrants differ-
ent treatment approaches. In addition to a careful
and detailed sleep history, sleep diaries and activity
monitoring with wrist actigraphy can be useful in
distinguishing between the two conditions.
The most appropriate therapies for shifts in the
circadian rhythm are those known to strengthen
and entrain the sleep-wake cycle. Because light is
the strongest cue for circadian entrainment, one
of the most effective and common treatments for

circadian rhythm shifts is bright light therapy. Even-
ing light exposure has been found to delay circadian
rhythms and strengthen the sleep-wake cycle in
both healthy community living older subjects and
in nursing home patients [14,15]. Patients with ad-
vanced rhythms should spend more time outdoors
during the late afternoon or early evening and avoid
bright light in the morning hours. If patients are un-
able to spend enough time outdoors, studies have
shown that exposure to artificial light by a bright
light box in the early evening can improve sleep
continuity in both healthy and institutionalized el-
derly patients [14,16]. In addition, a regular sleep
schedule helps to promote a stronger sleep-wake
cycle.
Endogenous secretion of melatonin is known to
promote sleep and is reduced in older adults.
Some studies suggest that melatonin-replacement
therapy may improve sleep efficiency in this popu-
lation [17,18]. There is little consensus, however,
on the recommended dose or timing of adminis-
tration. In addition, melatonin is not regulated
by the Food and Drug Administration (FDA),
and there is no control over the purity and exact
drug composition of the various brands currently
available. Little is known about the possible drug
interactions or side effects related to the long-
term administration of melatonin. Clinicians
should exercise caution when considering a trial
of melatonin-replacement therapy in elderly pa-
tients. The National Institutes of Health (NIH)
State-of-the-Science Insomnia Conference con-
cluded that although melatonin seems to be effec-
tive for the treatment of circadian rhythm
disorders, there is little evidence for efficacy in
the treatment of insomnia [19]. It was also con-
cluded that there is no definition of an effective
dose. Although melatonin is thought to be safe
in the short term, there is no information about
the safety of long-term use [19]. It should be
noted, however, that the FDA recently approved
the first melatonin agonist, ramelteon, for the
treatment of sleep-onset insomnia [20].
Insomnia
Insomnia is defined as the inability to initiate or
maintain sleep that results in daytime conse-
quences. Studies have found insomnia to be the
most common sleep disturbance in older adults,
with up to 40% to 50% of those over the age of
60 reporting difficulty sleeping [21] and an annual
incidence rate of 5% in those over the age of 65
[22]. Insomnia complaints include difficulty falling
asleep, difficulty staying asleep, and early morning
awakenings. Women tend to have higher rates of in-
somnia than men [23].
Sleep and Its Disorders in Seniors 283
There are a variety of factors associated with or
comorbid with insomnia in the elderly including
depression and other psychiatric conditions, medi-
cal conditions, medications, and circadian rhythm
disturbances [24]. Foley and colleagues [22] re-
ported that only 7% of the incident cases of insom-
nia in the elderly occur in the absence of one of
these risk factors.
Depression
Patients with insomnia often have comorbid psy-
chiatric conditions. In the classic study by Ford
and Kamerow [25], 40% of insomnia patients had
a psychiatric diagnosis, with anxiety being the
most common, followed closely by depression.
The same study also showed that persistent insom-
nia was associated with an increased risk of a future
psychiatric disorder. Studies have suggested, how-
ever, that insomnia also puts an individual at
greater risk for a new, future episode of depression
[26–30]. In a study by Weissman and colleagues
[31], over 7000 adults with insomnia were followed
for 1 year. The results confirmed that the odds ratio
of a new-onset psychiatric disorder in the baseline
insomnia group was 5.4 for major depression,
20.3 for panic disorder, and 2.3 for alcohol abuse.
The Breslau and colleagues [32] and Chang and col-
leagues [33] studies showed similar results; how-
ever, these studies have primarily been conducted
in younger adults. The only study to include older
adults was a study by Roberts and colleagues [34]
of 2370 community residents with a mean age of
64.9 years. Survey data were collected at baseline
and again 1 year later. The prevalence of insomnia
at baseline was 23%. At follow-up, for those who ei-
ther had insomnia 1 year previously, or still had in-
somnia, there was an 8.08 odds ratio for new-onset
major depression. More data examining the comor-
bid relationship between psychiatric disorders and
insomnia in the elderly are needed.
The overall conclusions from research studies are
that about 20% of patients with insomnia have de-
pression, and about 90% of patients with depres-
sion report a sleep disturbance. Insomnia can be
a symptom of depression, can contribute to the on-
set of depression or depressive episodes, can predict
a prognosis and response to antidepressant therapy,
can be linked to recurrence or relapse of depression,
and can be linked to anxiety and other psychiatric
disorders [35].
Because insomnia and depression are considered
comorbid conditions [19], the treatment implica-
tions are that the two conditions should be treated
concurrently. In a study by Fava and colleagues
[36], depressed patients were randomized to either
a fluoxetine-placebo arm or a fluoxetine-eszopi-
clone arm. For those on both an antidepressant
284 Stepnowsky & Ancoli-Israel
and a sedative hypnotic, sleep was significantly lon-
ger and less disrupted than for those on fluoxetine
and placebo. Of even greater interest, however, is
that the response to depression was also greater in
the group treated concurrently. Although these
data were in younger adults, the results suggest
that treating the psychiatric condition at the same
time as treating the insomnia might result in a better
overall response. It is important to remember, how-
ever, that hypnotic agents are not FDA indicated for
the treatment of depression.
Sleep and medical illness
Older individuals often suffer from medical comor-
bidity. In a National Sleep Foundation survey of
adults aged 65 years and over, those with more
medical conditions, including cardiac and pulmo-
nary disease and depression, reported significantly
more sleep complaints [37]. Osteoarthritic pain,
shortness of breath caused by chronic obstructive
pulmonary disease or congestive heart failure, noc-
turia caused by enlarged prostate, and neurologic
deficits related to cerebrovascular accidents or Par-
kinson’s disease all can lead to difficulty with sleep
initiation and maintenance. Studies examining the
prevalence of sleep disturbances in patients with
chronic medical diseases have reported that 31%
of arthritis and 66% of chronic pain patients report
difficulty falling asleep, whereas 81% of arthritis,
85% of chronic pain, and 33% of diabetes patients
report difficulty staying asleep [38,39].
Sleep and medications
The issue of polypharmacy is of significant concern
in the elderly. The medications used to treat the un-
derlying geriatric medical problems can also cause
disruptions in sleep. Bronchodilators, b-blockers,
corticosteroids, decongestants, and diuretics are all
well-known to cause sleep disturbances, as are other
cardiovascular, neurologic, psychiatric, and gastro-
intestinal medications. Whenever feasible, the
offending medications should be stopped, or at
minimum the dose and timing adjusted. Sedating
medications should be administered before bed-
time, whereas stimulating medications and di-
uretics should be taken during the day.
Insomnia treatment
Treatments for insomnia are comprised of behav-
ioral, pharmacologic, and combined treatment
approaches.
Nonpharmacologic interventions
Nonpharmacologic interventions are effective in
the treatment of insomnia [19,40]. Good sleep
hygiene, or the practice of appropriate sleep behav-
iors, provides the basis for the behavioral approach
to insomnia. Sleep hygiene rule for older adults in-
clude the following:
Check effect of medication on sleep and
wakefulness
Keep a regular bedtime-waketime schedule
Avoid naps or limit to one nap a day, no longer
than 30 minutes
Restrict naps to late morning or early afternoon
Avoid caffeine, alcohol, and tobacco after lunch
Increase overall daytime light exposure (eg,
spend more time outside, especially late in
the day)
Exercise regularly
Eat a light snack (ie, milk, bread) before bed
Limit liquids in the evening
Do not spend too much time in bed
Get out of bed if unable to fall asleep
Poor sleep hygiene practices can be associated
with behavioral patterns that contribute to sleep
disturbances. Patients should be educated on how
to identify specific factors that affect their own
sleep. The use of alcohol, which is widely used as
a sleep aid because of its ability to shorten sleep la-
tency, should be discouraged, because it has been
shown to contribute to sleep fragmentation and
early morning awakenings [41].
Two commonly prescribed behavioral therapies
are stimulus control therapy and sleep-restriction
therapy. Stimulus control is based on the belief
that insomnia may be the result of maladaptive
classical conditioning [42]. Patients are instructed
to eliminate all in-bed activities other than sleep,
such as reading and television watching. If they
are not able to fall asleep within 20 minutes, they
are instructed to get out of bed until they feel suffi-
ciently sleepy, when they can return to bed and
again attempt to fall asleep. If they are not able to
fall asleep within 20 minutes, the pattern of getting
out of bed until sleepy repeats itself. This therapy
tries to break the association between the bed and
wakefulness.
Sleep-restriction therapy limits the time spent in
bed to about 15 minutes beyond the duration of
time spent asleep at night [43]. As sleep efficiency
improves (ie, the amount of sleep relative to the
amount of time in bed), the time in bed gradually
increases.
Cognitive behavioral therapy
Cognitive behavioral therapy (CBT) for insomnia
involves educational, behavioral, and cognitive
components. The educational component involves
encouraging the patient to determine which factors
might be predisposing, precipitating, or perpetuat-
ing the insomnia. The therapist explains that CBT
is effective by eliminating the perpetuating factors

with behavioral and cognitive strategies. The behav-
ioral component involves the behavioral tech-
niques (ie, stimulus control, sleep-restriction
therapy) described previously. The cognitive com-
ponent deals with the maladaptive thoughts or dys-
functional beliefs that the patient has about the
insomnia.
CBT has been shown to be as effective as medica-
tions in the short run and to have better long-term
outcomes in the treatment of insomnia, in both
younger and older adults [44]. In an 8-week dou-
ble-blind treatment longitudinal outcome study,
CBT, an intermediate-acting benzodiazepine (tema-
zepam), a combined CBT-temazepam condition,
and a placebo condition were compared in a sample
of older adults [45]. Compared with baseline, all
three active treatments reduced night wakings at
posttreatment; however, only CBT alone and
CBT-temazepam were associated with continued
improvement at 3-, 12-, and 24-month follow-up
interviews. In addition, one study found even two
25-miunte CBT sessions for insomnia are effective
in reductive nocturnal awakenings, which may be
a more practical approach in the primary care set-
ting. The NIH 2005 State-of-the-Science conference
on insomnia concluded that CBT is the most effec-
tive treatment for insomnia; that CBT has de-
monstrated efficacy; that CBT is as effective as
prescription medications for brief treatment of
chronic insomnia; that the beneficial effects of
CBT (in contrast to those produced by medications)
may last well beyond the termination of treatment;
and that there is no evidence that CBT produces
adverse effects [19]. Although pharmacologic treat-
ments may be of more immediate help in the acute
treatment phase, nonpharmacologic or combined
approaches may be more effective for long-term
clinical gains.
Pharmacologic interventions
Historically, a number of different classes of medi-
cations have been used to treat insomnia in the el-
derly including sedative-hypnotics, antihistamines,
antidepressants, antipsychotics, and anticonvul-
sants. The 2005 NIH State-of-the-Science Confer-
ence on Insomnia concluded with several
recommendations regarding medications for in-
somnia [19]. All antidepressants have potentially
significant adverse effects, raising concerns about
their risk-benefit ratio. Barbiturates and antipsy-
chotic medications have significant risks, and their
use in the treatment of chronic insomnia cannot
be recommended at this time.
There is particular concern with the use of anti-
histamines for insomnia in the elderly, although
these drugs are easy to obtain and are cheap. In
a study of 426 older hospitalized medical patients,
Sleep and Its Disorders in Seniors 285
all 70 years and older, 27% received 25 to 50 mg di-
phenhydramine during hospitalization. Compared
with patients who were not given diphenhydra-
mine, these patients were shown to be at increased
risk for any delirium symptoms, inattention, disor-
ganized speech, altered consciousness, urinary cath-
eter placement, and longer median length of stay. A
dose-response relationship was demonstrated for
most adverse outcomes [46]. The NIH concluded
that there is no systematic evidence for the efficacy
of antihistamines, yet there are significant concerns
about the widespread use and risks with these
agents, particularly in the older patient.
Sedative-hypnotic medications are at times ap-
propriate for the management of insomnia, and
choosing the sedative-hypnotic that best fits the
specific complaint related to insomnia is the key
to using this class of medications successfully. Po-
tentially harmful effects must be taken into account
when prescribing sedative-hypnotics, particularly
benzodiazepines, in the elderly. The administration
of long-acting hypnotics can cause adverse daytime
effects, such as excessive daytime sleepiness and
poor motor coordination, which can lead to in-
juries [47]. In the elderly, the risk of falls, cognitive
impairment, and respiratory depression are of par-
ticular concern, although some recent studies have
suggested that insomnia is a risk for falls and hyp-
notic use is not [48]. Chronic use of long-acting
benzodiazepines can lead to tolerance and with-
drawal symptoms if abruptly discontinued, and
the benefits of these agents for long-term use have
not been studied with randomized clinical trials.
Additionally, the potential for exacerbating coexist-
ing medical conditions, such as hepatic or renal dis-
orders, exists when these medications are used.
The newer selective short-acting type-1 g-amino-
butyric acid benzodiazepines receptor agonists (ie,
zolpidem [49,50], zolpidem [51], zaleplon
[52,53], eszopiclone [54,55]) have been shown to
be effective in older adults, with a low propensity
for causing withdrawal, dependence, tolerance, or
clinical residual effects. All were shown either to de-
crease the time it takes to fall asleep or to increase
total sleep time. In younger adults, eszopiclone
has been found to be safe and effective in the
long-term treatment of chronic insomnia [56].
These long-term studies have not yet been pub-
lished, however, in older adults. Ramelteon, a mela-
tonin agonist, has also been shown to be safe and
effective in the treatment of insomnia in older
adults [57]. The NIH concluded that although the
older benzodiazepines are safe in the short-term
treatment of insomnia, the frequency and severity
of adverse effects are much lower in the newer non-
benzodiazepines [19]. The NIH panel also ex-
pressed significant concerns, however, about the
286 Stepnowsky & Ancoli-Israel
risks associated with the use of these medications in
older adults.
Primary sleep disorders
Three primary sleep disorders are commonly found
in the elderly: (1) sleep-related breathing disorder
(SRDB), (2) restless legs syndrome–periodic limb
movements in sleep (RLS-PLMS), and (3) REM
sleep-behavior disorder (RBD).
Sleep-related breathing disorder
SRBD has been shown to be quite common in the
elderly. In the largest series of randomly selected
community-dwelling elderly (65–95 years of age),
Ancoli-Israel and colleagues [58] reported that
81% of the study subjects had an apnea-hypopnea
index (AHI) greater than or equal to 5, with preva-
lence rates of 62% for an AHI greater than or equal
to 10, 44% for an AHI greater than or equal to 20,
and 24% for an AHI greater than or equal to 40.
The Sleep Heart Health Study studied a large cohort
of 6400 patients with a mean age of 63.5 (range:
40–98 years) and reported on prevalence rates of
SRBD by 10-year age groups: for those aged 60 to
69, 32% had an AHI 5 to 14 and 19% had an
AHI greater than or equal to 15; for those aged 70
to 79, 33% had an AHI 5 to 14 and 21% had an
AHI greater than or equal to 15; for those aged 80
to 98, 36% had an AHI 5 to 14 and 20% had an
AHI greater than or equal to 15 [59]. In contrast,
middle-aged adults 30 to 60 years of age have an es-
timated prevalence of 4% in men and 2% in
women (with SRBD defined as AHI R5 and the
presence of excessive daytime somnolence [60]).
Longitudinal and cross-sectional studies have
shown that the prevalence of SRBD increases or sta-
bilizes with increasing age [58,59]. The Sleep Heart
Health Study found a small increase in SRBD prev-
alence with increasing 10-year age groups for those
subjects with an AHI greater than or equal to 15
[59]. In a longitudinal study where older adults
were followed for 18 years, Ancoli-Israel and col-
leagues [61] found that AHI remained stable and
only changed with associated changes in body
mass index.
Elderly nursing home patients, in particular those
with dementia, have been shown to have higher
prevalence rates of SRBD than those who live inde-
pendently, with prevalence rates ranging from 33%
to 70% [62,63]. Several studies have also found that
the severity of the dementia was positively corre-
lated with the severity of the SRBD [62,64]. Despite
these findings, several other studies have failed to
show a significant difference in the amount of
SRBD in demented elderly compared with age-
matched controls [65,66].
SRBD risk factors in the elderly include increas-
ing age, gender, obesity, and symptomatic status
[67]. Other factors that increase the risk of develop-
ing SRBD include the use of sedating medications,
alcohol consumption, family history, race, smok-
ing, and upper airway configuration [67].
Snoring and excessive daytime sleepiness are two
principal symptoms of SRBD in the elderly. Other
less common presentations in the elderly include
insomnia, nocturnal confusion, and daytime cogni-
tive impairment including difficulties with concen-
tration and attention, and short-term memory loss.
It is also not uncommon for the symptoms and
clinical presentations of SRBD to be similar to
that of younger adults.
Approximately 50% of patients with habitual
snoring have some degree of SRBD, and snoring
has been identified as an early predictor of SRBD
[68]. In subjects 65 years and older, Enright and col-
leagues [69] found that loud snoring was indepen-
dently associated with BMI, diabetes, and arthritis
in elderly women and alcohol use in elderly men,
but that self-reported snoring decreased with age.
It should be noted, however, that not all patients
who snore have SRBD and not all patients with
SRBD snore. Because many elderly live alone, this
symptom may be difficult to identify.
Excessive daytime sleepiness results from recur-
rent nighttime arousals and sleep fragmentation
and is a major feature of SRBD in the elderly. The
presence of excessive daytime sleepiness may be
manifested as unintentional napping because indi-
viduals may fall asleep at inappropriate times dur-
ing the day, such as while watching television or
movies, while reading, during conversations, while
working, and while driving. Excessive daytime
sleepiness can cause reduced vigilance and is associ-
ated with cognitive deficits, which may be particu-
larly serious in older adults who may already have
some cognitive impairment [70].
There is a rapidly evolving body of literature on
cardiovascular consequences related to SRBD, in-
cluding hypertension, cardiac arrhythmias, conges-
tive heart failure, myocardial infarction, and stroke.
Most of the research to date has focused on younger
or middle-aged adults, however, and the exact rela-
tionship between SRBD and these comorbidities in
the elderly remains unknown.
Earlier studies have reported a positive associa-
tion between SRBD and hypertension in the elderly
[71]. The Sleep Heart Health Study found no associ-
ation between SRBD and systolic-diastolic hyper-
tension in those aged greater than or equal to
60 years [72]. The study did find a positive associa-
tion between the SRBD severity and the risk of
developing cardiovascular disease, however, includ-
ing coronary artery disease and stroke and the

development of congestive heart failure [73]. Im-
portantly, even mild to moderate SRBD was associ-
ated with its development.
The negative effect of severe SRBD (AHI R30) on
cognitive dysfunction in the healthy elderly is well
established, with consistent reports of impairment
on attentional tasks, immediate and delayed recall
of verbal and visual material, executive tasks, plan-
ning and sequential thinking, and manual dexterity
[74]. Studies examining the relationship between
milder SRBD and cognition are less clear-cut, be-
cause some studies have found that milder SRBD
(AHI 10–20) does not cause cognitive dysfunction
in the absence of sleepiness [75].
In addition to the cognitive deficits that may oc-
cur as a result of SRBD, there is evidence that many
of the progressive dementias including Alzheimer’s
disease and Parkinson’s disease involve degenera-
tion in areas of the brainstem that are responsible
for regulating respiration and other autonomic
functions relevant to sleep maintenance. This de-
generation may place the patient at an increased
risk of developing SRBD. For example, Ancoli-Israel
and colleagues [62] found that those institutional-
ized elderly with severe dementia had more severe
SRBD compared with those with mild-moderate
or no dementia. Furthermore, those with more se-
vere SRBD performed worse on the dementia rating
scales, suggesting that more severe SRBD was asso-
ciated with more severe dementia.
Higher rates of mortality are seen with SRBD. In
general, rates from all causes increase 30% during
the night, and for those aged 65 and over, the excess
deaths typically occur between the hours of 2 AM
and 8 AM [76]. The presence of unrecognized or
untreated SRBD may partially account for these
findings because several studies have found an asso-
ciation between SRBD in the elderly and increased
mortality rates [77,78], although some studies of
community-dwelling, nondemented elderly sub-
jects have not found AHI to be an independent pre-
dictor of mortality [79,80]. Rather than directly
causing an increased mortality, these studies have
found that SRBD may be one of several predispos-
ing factors for cardiopulmonary disease, which in
combination leads to increased mortality. This hy-
pothesis is strengthened by a study by Ancoli-Israel
and colleagues [81], which reported that elderly
men with congestive heart failure had more severe
SRBD than those with no heart disease, and men
with both conditions (congestive heart failure and
SRBD) had shortened life spans compared with
those with just congestive heart failure, just SRBD,
or neither. More studies need to be undertaken to
understand better the exact nature of the relation-
ship of SRBD and mortality in the elderly, including
studies specific to older women because most of the
Sleep and Its Disorders in Seniors 287
studies completed in this age category have in-
volved predominantly men.
To assess accurately the SRBD in the elderly,
a multiple step process should be used. A complete
sleep history should be obtained, focusing on
symptoms of SRBD including excessive daytime
sleepiness, unintentional napping, snoring, symp-
toms of other sleep disorders (ie, RLS), and sleep-re-
lated habits and routines, if possible in the presence
of a bed partner, roommate, or caregiver. The pa-
tient’s medical history, including psychiatric and
medical records, should be thoroughly reviewed,
paying particular attention to associated medical
conditions and medications, the use of alcohol,
and evidence of cognitive impairment. Lastly,
when there is a suggestion of SRBD, an overnight
sleep recording should be obtained.
Treatment of SRBD in the elderly should be
guided by the significance of the patient’s symp-
toms and the severity of the SRBD [82]. Patients
with more severe SRBD (AHI >20) deserve a trial
of treatment. For those with milder SRBD (AHI
<20), treatment should be considered if comorbid
conditions are present, such as hypertension, cogni-
tive dysfunction, or excessive daytime sleepiness.
Age alone should never be a reason to withhold
treatment, nor should assumed noncompliance
[83].
There are a number of effective treatments for
SRBD. Continuous positive airway pressure
(CPAP) is the gold standard treatment for SRBD
and provides positive pressure by the nasal passages
or oral airway, creating a pneumatic splint to keep
the airway open during inspiration. When used ap-
propriately, CPAP has been shown safely and effec-
tively to manage SRBD at night with minimal side
effects. Three months of compliant CPAP use in
older adults has been reported to improve cognition,
particularly in the areas of attention, psychomotor
speed, executive functioning, and nonverbal delayed
recall [74].
CPAP compliance can be an issue for any adult
with SRBD, and clinicians should not assume that
elderly patients would be noncompliant simply
because of age. The authors’ laboratory has deter-
mined that patients with mild-moderate Alz-
heimer’s disease and SRBD can tolerate treatment
with CPAP devices [84], and the only factor associ-
ated with poor CPAP compliance was the presence
of depression; age, severity of dementia, or severity
of SRBD were not associated with poor compliance
[84].
Alternatives to CPAP include oral appliances and
surgery; however, neither has been shown to be as
effective as CPAP. All patients should be counseled
on weight loss and smoking cessation, if indicated.
Longer-acting benzodiazepines should generally be
288 Stepnowsky & Ancoli-Israel
avoided in the elderly with SRBD because most of
these medications are respiratory depressants and
may actually increase the number and duration of
apneas. Elderly patients with SRBD should be en-
couraged to abstain completely from alcohol con-
sumption, because even small amounts can make
SRBD worse.
Although there is a growing body of literature ex-
ploring SRBD in the elderly, there is also an ongo-
ing debate in the field as to what the presence of
SRBD in the elderly means. Some propose that a dis-
tinction should be made between age-dependent
conditions, in which aging causes the pathology,
and age-related conditions, in which the disease
only occurs during a particular age period [85].
Whether SRBD is an age-dependent or an age-re-
lated condition remains unknown. Ancoli-Israel
and colleagues [58] and others [86–88] have found
that the prevalence of SRBD increases with age, and
SRBD may be thought of as an age-dependent con-
dition. If this is the case and SRBD in older adults is
the same condition with the same outcomes, then
the presence of SRBD in the elderly warrants the
same aggressive treatment. Although the prevalence
of SRBD in the elderly may be age-dependent, the
severity of the SRBD and its clinical significance in
the elderly may be age-related. Ancoli-Israel and
colleagues [61] showed in an 18-year follow-up
study of elderly patients with SRBD that AHI did
not continue to increase with age, and that if the pa-
tient’s body mass index remained stable, so did the
AHI. Bixler and colleagues [88] reported in a sample
of older men that the prevalence of SRBD increased
but that after controlling for body mass index, the
severity based on number of events and oxygen sat-
uration actually decreased with age. Studies aimed
at answering these questions and those related to
mortality and SRBD in the elderly are ongoing [83].
In terms of treatment, it does not matter if sleep
apnea in older adults is the same as sleep apnea
in younger adults. The driving force behind the de-
cision whether to treat or not to treat should be the
clinical presentation of the patient and conse-
quences of sleep apnea that patient is experiencing.
The bottom line is that if the sleep apnea is associ-
ated with clinical symptoms, then it should be
treated regardless of the age of the patient [83].
Periodic limb movements in sleep and restless
legs syndrome
Insomnia complaints may be associated with RLS
or with PLMS. Although the prevalence of RLS is
about 10% in younger adults, several studies have
shown that the prevalence almost doubles with
age [89,90]. The prevalence of a related disorder,
PLMS, also increases with age [91].
Pharmacologic intervention is required to man-
age RLS-PLMS. Dopamine agonists are effective in
both reducing the number of kicks and associated
arousals and are the FDA-approved treatments for
RLS, specifically ropinirole and pramipexole. Thera-
peutic studies, however, have only been conducted
in younger adults.
Rapid eye movement sleep behavior disorder
RBD is characterized by the intermittent absence of
normal skeletal muscle atonia during REM sleep,
associated with excessive motor activity while
dreaming and typically occurring during the second
half of the night when REM is more common. Pa-
tients may walk, talk, eat, or seem to be acting out
their dreams, which can result in violent move-
ments that are potentially harmful to themselves
and their bed partner. Vivid dreams, consistent
with the patient’s aggressive or violent behavior,
may be recalled on waking.
The exact prevalence of RBD is unknown but
studies have found it to be most common in elderly
men [92,93]. Although the etiology of RBD remains
unknown, there seems to be a strong association
between idiopathic RBD and degenerative neuro-
logic diseases, including Parkinson’s disease, multi-
ple system atrophy, and Lewy body dementia
[92,94,95]. In addition, in many cases of neurode-
generative disease, RBD may precede other symp-
toms of the neurodegenerative disorder by years
[92,95,96]. Olson and colleagues [92] reported
that 50% of patients diagnosed with idiopathic
RBD developed Parkinson’s disease or multiple sys-
tem atrophy within 3 to 4 years. Schenck and col-
leagues [96] found that Parkinsonism developed
in 38% of men a mean of 3.7 years after an initial
diagnosis of idiopathic RBD. Withdrawal of REM-
suppressing agents, such as alcohol, tricyclic antide-
pressants, amphetamines, and cocaine, has been
strongly linked to the onset of acute RBD [92,97].
Other medications and conditions reported to
induce acute RBD include monoamine oxidase in-
hibitors, fluoxetine, and stress disorders [92,97].
As with other primary sleep disorders, the diag-
nosis of RBD requires a thorough sleep history
along with bed partner report, if possible. The
RBD diagnosis requires a simultaneous overnight
polysomnogram and video recording of the night-
time behavior, to confirm a relationship between
REM sleep and the complex motor behaviors ex-
hibited by the patient. On reading the polysomnog-
raphy, clinicians should pay close attention to
intermittent elevations in muscle tone or limb
movements during REM sleep, which should be
relatively rare.
The treatment of choice for RBD, although used off-
label, is clonazepam, a long-acting benzodiazepine.

Although not in randomized clinical trials, clonaze-
pam has been shown to result in partial or complete
cessation of abnormal nocturnal motor movements
in 90% of patients [98]. Patients may complain of
residual sleepiness, however, because of its long
half-life, and it is contraindicated in patients with co-
existing SRBD. Several alternative medications have
shown some positive effects in RBD including carba-
mazepine [99], melatonin [100], and dopaminergics
agents [101], although none has been shown to be as
effective as clonazepam.
In addition to pharmacologic treatment, sleep
hygiene education of the patient and the bed part-
ner are important aspects of RBD treatment, includ-
ing (1) efforts to make the bedroom safer (eg,
removing heavy or breakable or potentially injuri-
ous objects from the bed’s vicinity); (2) placing
heavy curtains on bedroom windows and locking
doors and windows at night; and (3) having pa-
tients sleep on a mattress on the floor to avoid fall-
ing out of bed.
Sleep in dementia
There is considerable evidence that dementia affects
sleep differently from the normal aging process [1].
This is not surprising considering that dementing
illnesses, such as Alzheimer’s disease, Parkinson’s
disease, multi-infarct dementia, or Lewy body de-
mentia, may involve irreversible damage to the
brain in areas responsible for regulating sleep. In
general, patients with dementia have disturbed
sleep at night, and laboratory sleep studies of de-
mented patients have found increased sleep frag-
mentation and sleep-onset latency, and decreased
sleep efficiency, total sleep time, and slow wave
sleep [13]. Furthermore, the severity of dementia
seems to be associated with the severity of the sleep
disruption [102].
Because of these changes in sleep architecture, pa-
tients with dementia may have excessive daytime
sleepiness; nighttime wandering; confusion; and
agitation (sundowning). Such nighttime behavior
and disruptions may eventually lead to institution-
alization [103]. Addressing the issues related to
sleep disturbances in the community-dwelling de-
mented elderly is especially important, because it
may potentially enable caregivers to postpone
institutionalization.
It may be difficult to determine the exact nature
of the sleep disturbance in patients with dementia,
although caregivers can be a valuable source of
information. The same causes of sleep disruption
in the nondemented older adult are also found in
the patient with dementia. Pain from medical ill-
nesses, medications, circadian rhythm changes,
and depression are all potential causes of sleep
Sleep and Its Disorders in Seniors 289
disturbances in this population. It is also important
to inquire about treatable primary sleep disorders,
such as SRBD, RLS, or PLMS. Depending on the se-
verity of the dementia, overnight sleep studies may
not be an option and actigraphy may serve as a use-
ful method to assess sleep and circadian rhythms in
these patients [104].
Treatment of specific sleep disturbances in the el-
derly with dementia should be guided by that spe-
cific sleep disorder. SRBD should be treated with
continuous CPAP, RLS-PLMS should be treated
with a dopamine agonist, and circadian rhythm dis-
turbances should be treated with bright light ther-
apy. Maintenance of regular physical activity and
social interaction can also promote a more robust
sleep-wake cycle. Sedative-hypnotics including ben-
zodiazepines, tricyclic antidepressants, antihista-
mines, anticonvulsants, and antipsychotics, are
frequently prescribed in an off-label fashion for
the nighttime restlessness associated with demen-
tia. Attempting to enhance sleep continuity with
these medications, however, may paradoxically re-
sult in increased sleep disturbance and daytime
sleepiness. Because of the many side effects associ-
ated with these medications, including drug inter-
actions and residual daytime sleepiness (hangover
effect) resulting in impaired motor and cognitive
function, nonpharmacologic interventions are
preferred. In addition, the FDA recently put
a black-box warning on the use of the atypical anti-
psychotics agents in patients with dementia.
Sleep in institutionalized elderly
Institutionalized elderly experience extremely frag-
mented sleep [105]. Middelkoop and colleagues
[106] reported that patients living in nursing homes
had poorer sleep quality, more disturbed sleep on-
set, more phase-advanced sleep periods, and higher
use of sedative-hypnotics when compared with
those elderly living in the community or in assisted
living environments. Studies have found that for
older adults living in nursing homes, not a single
hour in a 24-hour day was spent fully awake or fully
asleep [102,105,107].
There are a variety of environmental factors that
contribute to the reduction in sleep quality. Noise
and light exposure occur intermittently throughout
the 24-hour day. Schnelle and colleagues [108]
demonstrated that both ambient light and night-
time noise contributed significantly to sleep disrup-
tion in nursing home patients. This study also
found that those patients living in nursing homes
where nighttime noise and light were kept to a min-
imum had better sleep. Ancoli-Israel and Kripke
[105] reported that the nursing home patients
were exposed to less than 10 minutes of bright light
290 Stepnowsky & Ancoli-Israel
per day and those with more light exposure had
fewer sleep disruptions [12]. Chronic bed rest is
known to disrupt circadian rhythms, yet institution-
alized patients typically spend large amounts of the
24-hour day in bed [105]. Changes in sleep hygiene
and the sleep environment may greatly improve the
sleep quality of nursing home inhabitants. Strate-
gies to reduce nighttime disturbances and to pro-
mote stronger sleep-wake cycles are consistent
with those listed previously.
Summary
Significant changes in sleep accompany aging for
most adults. There are a variety of potential causes
including medical illnesses; medications; circadian
rhythm disturbances; depression and other psychi-
atric disorders; and primary sleep disorders
(SRBD, RLS-PLMS, and RBD). The diagnosis re-
quires a good sleep history and, when indicated,
a sleep study. Treatment should address the primary
problem rather than the complaint itself and can re-
sult in significant improvement in quality of life
and daytime functioning in the elderly.
References
[1] Bliwise DL. Review: sleep in normal aging and
dementia. Sleep 1993;16:40–81.
[2] Ohayon MM, Carskadon MA, Guilleminault C,
et al. Meta-analysis of quantitative sleep param-
eters from childhood to old age in healthy indi-
viduals: developing normative sleep values
across the human lifespan. Sleep 2004;27(7):
1255–73.
[3] Van Cauter EV, Leproult R, Plat L. Age-related
changes in slow wave sleep and REM sleep and
relationship with growth hormone and cortisol
levels in healthy men. JAMA 2000;284(7):
861–8.
[4] Carskadon MA, van den Hoed J, Dement WC.
Sleep and daytime sleepiness in the elderly.
J Geriatr Psychiatry 1980;13:135–51.
[5] Dement WC, Seidel W, Carskadon MA. Daytime
alertness, insomnia and benzodiazepines. Sleep
1982;5:S28–45.
[6] Ancoli-Israel S. Sleep problems in older adults:
putting myths to bed. Geriatrics 1997;52(1):
20–30.
[7] Swaab DF, Fliers E, Partiman TS. The supra-
chiasmatic nucleus of the human brain in rela-
tion to sex, age and senile dementia. Brain Res
1985;342:37–44.
[8] Touitou Y. Human aging and melatonin: clinical
relevance. Exp Gerontol 2001;36(7):1083–100.
[9] Campbell SS, Kripke DF, Gillin JC, et al. Expo-
sure to light in healthy elderly subjects and
Alzheimer’s patients. Physiol Behav 1988;42:
141–4.
[10] Espiritu RC, Kripke DF, Ancoli-Israel S, et al.
Low illumination by San Diego adults: associa-
tion with atypical depressive symptoms. Biol
Psychiatry 1994;35:403–7.
[11] Ancoli-Israel S, Klauber MR, Jones DW, et al.
Variations in circadian rhythms of activity, sleep
and light exposure related to dementia in
nursing home patients. Sleep 1997;20(1):
18–23.
[12] Shochat T, Martin J, Marler M, et al. Illumina-
tion levels in nursing home patients: effects
on sleep and activity rhythms. J Sleep Res
2000;9(4):373–80.
[13] Vitiello MV. Sleep disorders and aging. Curr
Opin Psychiatry 1996;9(4):284–9.
[14] Campbell SS, Terman M, Lewy AJ, et al. Light
treatment for sleep disorders: consensus report.
V. Age-related disturbances. J Biol Rhythms
1995;10(2):151–4.
[15] Lack L, Wright H, Kemp K, et al. The treatment
of early-morning awakening insomnia with 2
evenings of bright light. Sleep 2005;28(5):
616–23.
[16] Ancoli-Israel S, Martin JL, Kripke DF, et al. Effect
of light treatment on sleep and circadian
rhythms in demented nursing home patients.
J Am Geriatr Soc 2002;50(2):282–9.
[17] Garfinkel D, Laudon M, Nof D, et al. Improve-
ment of sleep quality in elderly people by
controlled-release melatonin. Lancet 1995;
346:541–4.
[18] Haimov I, Lavie P. Potential of melatonin re-
placement therapy in older patients with sleep
disorders. Drugs Aging 1995;7(2):75–8.
[19] National Institutes of Health. State of the Sci-
ence Conference statement on manifestations
and management of chronic insomnia in adults,
June 13–15, 2005. Sleep 2005;28(9):1049–57.
[20] Roth T, Stubbs CD, Walsh JK. Ramelteon
(TAK-375): a selective MT1/MT2 receptor ago-
nist reduces latency to persistent sleep in
a model of transient insomnia related to a novel
sleep environment. Sleep 2005;28(3):303–7.
[21] Foley DJ, Monjan AA, Brown SL, et al. Sleep
complaints among elderly persons: an epidemi-
ologic study of three communities. Sleep 1995;
18(6):425–32.
[22] Foley DJ, Monjan A, Simonsick EM, et al. Inci-
dence and remission of insomnia among elderly
adults: an epidemiologic study of 6,800 persons
over three years. Sleep 1999;22(Suppl 2):
S366–72.
[23] Rediehs MH, Reis JS, Creason NS. Sleep in old
age: focus on gender differences. Sleep 1990;
13(5):410–24.
[24] Ancoli-Israel S. Insomnia in the elderly: a review
for the primary care practitioner. Sleep 2000;
23(Suppl 1):S23–30.
[25] Ford DE, Kamerow DB. Epidemiologic study of
sleep disturbances and psychiatric disorders: an
opportunity for prevention? JAMA 1989;
262(11):1479–84.

[26] Dryman A, Eaton WW. Affective symptoms as-
sociated with the onset of major depression in
the community: findings from the US National
Institute of Mental Health Epidemiologic
Catchment Area Program. Acta Psychiatr Scand
1991;84(1):1–5.
[27] Livingston G, Blizard B, Mann A. Does sleep
disturbance predict depression in elderly peo-
ple? A study in inner London. Br J Gen Pract
1993;43:445–8.
[28] Cole MG, Dendukuri N. Risk factors for depres-
sion among elderly community subjects: a system-
atic review and meta-analysis. Am J Psychiatry
2003;160(6):1147–56.
[29] Nowell PD, Buysse DJ. Treatment of insomnia in
patients with mood disorders. Depress Anxiety
2001;14(1):7–18.
[30] Fava M. Daytime sleepiness and insomnia as
correlates of depression. J Clin Psychiatry
2004;65(Suppl 16):27–32.
[31] Weissman MM, Greenwald S, Nino-Murcia G,
et al. The morbidity of insomnia uncomplicated
by psychiatric disorders. Gen Hosp Psychiatry
1997;19(4):245–50.
[32] Breslau N, Roth T, Rosenthal L, et al. Sleep
disturbance and psychiatric disorders: a longitu-
dinal epidemiological study of young adults.
Biol Psychiatry 1996;39:411–8.
[33] Chang PP, Ford DE, Mead LA, et al. Insomnia in
young men and subsequent depression. The
Johns Hopkins Precursors Study. Am J Epidemiol
1997;146(2):105–14.
[34] Roberts RE, Shema SJ, Kaplan GA, et al. Sleep
complaints and depression in an aging cohort:
a prospective perspective. Am J Psychiatry
2000;157(1):81–8.
[35] Kupfer DJ. Pathophysiology and management
of insomnia during depression. Ann Clin Psy-
chiatry 1999;11(4):267–76.
[36] Fava M, McCall WV, Krystal A, et al. Eszopi-
clone co-administered with fluoxetine in pa-
tients with insomnia coexisting with major
depressive disorder. Biol Psychiatry 2006;
59(11):1052–60.
[37] Foley D, Ancoli-Israel S, Britz P, et al. Sleep dis-
turbances and chronic disease in older adults:
results of the 2003 National Sleep Foundation
Sleep in America Survey. J Psychosom Res
2004;56(5):497–502.
[38] Wilcox S, Brenes GA, Levine D, et al. Factors
related to sleep disturbance in older adults ex-
periencing knee pain or knee pain with radio-
graphic evidence of knee osteoarthritis. J Am
Geriatr Soc 2000;48(10):1241–51.
[39] Sridhar GR, Madhu K. Prevalence of sleep dis-
turbance in diabetes mellitus. Diabetes Res
Clin Pract 1994;23(3):183–6.
[40] Morin CM, Hauri PJ, Espie CA, et al. Nonphar-
macologic treatment of chronic insomnia: an
American Academy of Sleep Medicine review.
Sleep 1999;22(8):1134–56.
Sleep and Its Disorders in Seniors 291
[41] Roehrs T, Roth T. Sleep, sleepiness, sleep disor-
ders and alcohol use and abuse. Sleep Med
Rev 2001;4:287–97.
[42] Bootzin RR, Nicassio PM. Behavioral treat-
ments for insomnia. In: Hersen M, Eisler RM,
Miller PM, editors, Progress in behavior modifi-
cation, vol. 6. New York: Academic Press, Inc.;
1978. p. 1–45.
[43] Spielman AJ, Saskin P, Thorpy MJ. Treatment of
chronic insomnia by restriction of time in bed.
Sleep 1987;10:45–56.
[44] Morin CM, Colecchi C, Stone J, et al. Behavioral
and pharmacological therapies for late life in-
somnia. JAMA 1999;281(11):991–9.
[45] Morin CM, Bastien CH, Brink D, et al. Adverse
effects of temazepam in older adults with
chronic insomnia. Hum Psychopharmacol
2003;18(1):75–82.
[46] Agostini JV, Leo-Summers LS, Inouye SK. Cog-
nitive and other adverse effects of diphenhydra-
mine use in hospitalized older patients. Arch
Intern Med 2001;161(17):2091–7.
[47] Roth T, Roehrs T, Zorick F. Pharmacological
treatment of sleep disorders. In: Williams RL,
Karacan I, Moore CA, editors. Sleep disorders: di-
agnosis and treatment. New York: John Wiley &
Sons; 1988. p. 373–95.
[48] Avidan AY, Fries BE, James MC, et al. Insomnia
and hypnotic use, recorded in the minimum
data set, as predictors of falls and hip fractures
in Michigan nursing homes. J Am Geriatr Soc
2005;53(6):955–62.
[49] Scharf MB, Mayleben DW, Kaffeman M, et al.
Dose response effects of zolpidem in normal
geriatric subjects. J Clin Psychiatry 1991;52(2):
77–83.
[50] Roger M, Attali P, Coquelin JP. Multicenter,
double-blind, controlled comparison of zolpi-
dem and triazolam in elderly patients with
insomnia. Clin Ther 1993;15(1):127–36.
[51] Hindmarch I, Legangneux E, Stanley N, et al. A
double-blind, placebo-controlled investigation
of the residual psychomotor and cognitive ef-
fects of zolpidem-MR in healthy elderly volun-
teers. Br J Clin Pharmacol 2006;62(5):538–45.
[52] Ancoli-Israel S, Walsh JK, Mangano RM, et al. Za-
leplon Clinical Study Group. Zaleplon, a novel
nonbenzodiazepine hypnotic, effectively treats
insomnia in elderly patients without causing
rebound effects. Primary Care Companion
J Clin Psychiatry 1999;1(4):114–20.
[53] Ancoli-Israel S, Richardson GS, Mangano R,
et al. Long-term use of sedative hypnotics in
older patients with insomnia. Sleep Med
2005;6:107–13.
[54] McCall WV, Erman M, Krystal AD, et al. A poly-
somnography study of eszopiclone in elderly
patients with insomnia. Curr Med Res Opin
2006;22(9):1633–42.
[55] Scharf M, Erman M, Rosenberg R, et al. A
2-week efficacy and safety study of eszopiclone
292 Stepnowsky & Ancoli-Israel
in elderly patients with primary insomnia.
Sleep 2005;28(6):720–7.
[56] Krystal AD, Walsh JK, Laska E, et al. Sustained
efficacy of eszopiclone over 6 months of nightly
treatment: results of a randomized, double-
blind, placebo-controlled study in adults with
chronic insomnia. Sleep 2003;26(7):793–9.
[57] Roth T, Seiden D, Wang-Weigand S, et al. A
2-night, 3-period, crossover study of ramel-
teon’s efficacy and safety in older adults with
chronic insomnia. Curr Med Res Opin 2007;
23(5):1005–14.
[58] Ancoli-Israel S, Kripke DF, Klauber MR, et al.
Sleep disordered breathing in community-
dwelling elderly. Sleep 1991;14(6):486–95.
[59] Young T, Shahar E, Nieto FJ, et al. Predictors of
sleep-disordered breathing in community-
dwelling adults: the Sleep Heart Health Study.
Arch Intern Med 2002;162(8):893–900.
[60] Young T, Palta M, Dempsey J, et al. The occur-
rence of sleep disordered breathing among
middle-aged adults. N Engl J Med 1993;328:
1230–5.
[61] Ancoli-Israel S, Gehrman P, Kripke DF, et al.
Long-term follow-up of sleep disordered
breathing in older adults. Sleep Med 2001;
2(6):511–6.
[62] Ancoli-Israel S, Klauber MR, Butters N, et al.
Dementia in institutionalized elderly: relation
to sleep apnea. J Am Geriatr Soc 1991;39(3):
258–63.
[63] Gehrman PR, Martin JL, Shochat T, et al. Sleep
disordered breathing and agitation in institu-
tionalized adults with Alzheimer’s disease.
Am J Geriatr Psychiatry 2003;11(4):426–33.
[64] Hoch CC, Reynolds CFI. Cognitive function
and sleep disordered breathing in dementia:
the Pittsburg experience. In: Kuna ST,
Suratt PM, Remmers JE, editors. Sleep and res-
piration in aging adults. New York: Elsevier;
1991. p. 245–50.
[65] Smallwood RG, Vitiello MV, Giblin EC, et al.
Sleep apnea: relationship to age, sex, and
Alzheimer’s dementia. Sleep 1983;6:16–22.
[66] Bliwise DL, Yesavage JA, Tinklenberg JR, et al.
Sleep apnea in Alzheimer’s disease. Neurobiol
Aging 1989;10:343–6.
[67] Phillips B, Ancoli-Israel S. Sleep disorders in the
elderly. Sleep Med 2001;2(2):99–114.
[68] Collop NA, Cassell DK. Snoring and sleep-disor-
dered breathing. In: Lee-Chiong TL, Sateia MJ,
Carskadon MA, editors. Sleep medicine. Phila-
delphia: Hanley & Belfus; 2002. p. 349–55.
[69] Enright PL, Newman AB, Wahl PW, et al. Prev-
alence and correlates of snoring and observed
apneas in 5,201 older adults. Sleep 1996;
19(7):531–8.
[70] Martin J, Stepnowsky C, Ancoli-Israel S. Sleep
apnea in the elderly. In: McNicholas WT,
Phillipson EA, editors. Breathing disorders dur-
ing sleep. London: W.B. Saunders Company,
Lmt.; 2002. p. 278–87.
[71] Stoohs RA, Gingold J, Cohrs S, et al. Sleep-dis-
ordered breathing and systemic hypertension
in the older male. J Am Geriatr Soc 1996;
44(11):1295–300.
[72] Haas DC, Foster GL, Nieto FJ, et al. Age-depen-
dent associations between sleep-disordered
breathing and hypertension: importance of dis-
criminating between systolic/diastolic hyper-
tension and isolated systolic hypertension in
the Sleep Heart Health Study. Circulation
2005;111:614–21.
[73] Shahar E, Whitney CW, Redline S, et al. Sleep-
disordered breathing and cardiovascular dis-
ease: cross sectional results of the Sleep Heart
Health Study. Am J Respir Crit Care Med
2001;163(1):19–25.
[74] Aloia MS, Ilniczky N, Di Dio P, et al. Neuropsy-
chological changes and treatment compliance
in older adults with sleep apnea. J Psychosom
Res 2003;54:71–6.
[75] Redline S, Strauss ME, Adams N, et al. Neuro-
psychological function in mild sleep-disordered
breathing. Sleep 1997;20(2):160–7.
[76] Mitler MM, Hajdukovic RM, Shafor R, et al.
When people die: cause of death versus time
of death. Am J Med 1987;822:66–74.
[77] He J, Kryger MH, Zorick FJ, et al. Mortality and
apnea index in obstructive sleep apnea: experi-
ence in 385 male patients. Chest 1988;94:9–14.
[78] Bliwise DL, Bliwise NG, Partinen M, et al. Sleep
apnea and mortality in an aged cohort. Am J
Public Health 1988;78:544–7.
[79] Ancoli-Israel S, Kripke DF, Klauber MR, et al.
Morbidity, mortality and sleep disordered
breathing in community dwelling elderly. Sleep
1996;19(4):277–82.
[80] Mant A, King M, Saunders NA, et al. Four-year
follow-up of mortality and sleep-related respi-
ratory disturbance in non-demented seniors.
Sleep 1995;18(6):433–8.
[81] Ancoli-Israel S, DuHamel ER, Stepnowsky C,
et al. The relationship between congestive heart
failure, sleep disordered breathing and mortal-
ity in older men. Chest 2003;124(4):1400–5.
[82] Ancoli-Israel S, Coy TV. Are breathing distur-
bances in elderly equivalent to sleep apnea
syndrome? Sleep 1994;17:77–83.
[83] Ancoli-Israel S. Sleep apnea in older adults: is it
real and should age be the determining factor
in the treatment decision matrix? Sleep Med
Rev 2007;11(2):83–5.
[84] Ayalon L, Ancoli-Israel S, Stepnowsky C, et al.
Treatment adherence in patients with Alz-
heimer’s disease and obstructive sleep apnea.
Am J Geriatr Psychiatry 2006;14(2):176–80.
[85] Young T. Sleep-disordered breathing in older
adults: is it a condition distinct from that of
middle-aged adults? Sleep 1996;19(7):529–30.
[86] Bliwise DL, Carskadon MA, Carey E, et al. Lon-
gitudinal development of sleep-related respira-
tory disturbance in adult humans. J Gerontol
1984;39:290–3.

[87] Hoch CC, Reynolds CFI, Monk TH, et al. Com-
parison of sleep-disordered breathing among
healthy elderly in the seventh, eighth, and
ninth decades of life. Sleep 1990;13(6):502–11.
[88] Bixler EO, Vgontzas AN, Ten Have T, et al. Ef-
fects of age on sleep apnea in men. Am J Respir
Crit Care Med 1998;157:144–8.
[89] Ohayon MM, Roth T. Prevalence of restless legs
syndrome and periodic limb movement disor-
der in the general population. J Psychosom
Res 2002;53(1):547–54.
[90] Phillips B, Young T, Finn L, et al. Epidemiology
of restless legs symptoms in adults. Arch Intern
Med 2000;160(14):2137–41.
[91] Ancoli-Israel S, Kripke DF, Klauber MR, et al. Pe-
riodic limb movements in sleep in community-
dwelling elderly. Sleep 1991;14(6):496–500.
[92] Olson EJ, Boeve BF, Silber MH. Rapid eye move-
ment sleep behaviour disorder: demographic,
clinical and laboratory findings in 93 cases.
Brain 2000;123:331–9.
[93] Schenck CH, Hurwitz TD, Mahowald MW. Sym-
posium: normal and abnormal REM sleep regula-
tion: REM sleep behaviour disorder: an update on
a series of 96 patients and a review of the world
literature. J Sleep Res 1993;2(4):224–31.
[94] Montplaisir J. Abnormal motor behavior during
sleep. Sleep Med 2004;5(Suppl 1):S31–4.
[95] Boeve BF, Silber MH, Ferman TJ, et al. REM
sleep behavior disorder and degenerative de-
mentia: an association likely reflecting Lewy
body disease. Neurology 1998;51(2):363–70.
[96] Schenck CH, Bundlie SR, Mahowald MW.
Delayed emergence of a parkinsonian disorder
in 38% of 29 older men initially diagnosed
with idiopathic rapid eye movement sleep
behavior disorder. Neurology 1996;46:388–93.
[97] Sforza E, Krieger J, Petiau C. REM sleep behav-
ior: clinical and physiopathological findings.
Sleep Med Rev 1997;1(1):57–69.
[98] Schenck CH, Mahowald MW. Polysomno-
graphic, neurologic, psychiatric, and clinical
Sleep and Its Disorders in Seniors 293
outcome report on 70 consecutive cases with
the REM sleep behavior disorder (RBD): sus-
tained clonazepam efficacy in 89.5% of 57
treated patients. Cleve Clin J Med 1990;57:
S10–24.
[99] Schenck CH, Bundlie SR, Patterson AL, et al.
Rapid eye movement sleep behavior disorder:
a treatable parasomnia affecting older adults.
JAMA 1987;257:1786–9.
[100] Boeve BF, Silber MH, Ferman TJ. Melatonin for
treatment of REM sleep behavior disorder in
neurologic disorders: results in 14 patients.
Sleep Med 2003;4(4):281–4.
[101] Bamford CR. Carbamazepine in REM sleep be-
havior disorder. Sleep 1993;16:33–4.
[102] Pat-Horenczyk R, Klauber MR, Shochat T, et al.
Hourly profiles of sleep and wakefulness in se-
verely versus mild-moderately demented nurs-
ing home patients. Aging Clin Exp Res 1998;
10:308–15.
[103] Pollak CP, Perlick D. Sleep problems and insti-
tutionalization of the elderly. J Geriatr Psychia-
try Neurol 1991;4:204–10.
[104] Ancoli-Israel S, Cole R, Alessi CA, et al. The role
of actigraphy in the study of sleep and circadian
rhythms. Sleep 2003;26(3):342–92.
[105] Ancoli-Israel S, Kripke DF. Now I lay me down
to sleep: the problem of sleep fragmentation in
elderly and demented residents of nursing
homes. Bull Clin Neurosci 1989;54:127–32.
[106] Middelkoop HA, Kerkhof GA, Smilde-van den
Doel DA, et al. Sleep and ageing: the effect of
institutionalization on subjective and objective
characteristics of sleep. Age Ageing 1994;23(5):
411–7.
[107] Jacobs D, Ancoli-Israel S, Parker L, et al.
Twenty-four hour sleep-wake patterns in a nurs-
ing home population. Psychol Aging 1989;4(3):
352–6.
[108] Schnelle JF, Cruise PA, Alessi CA, et al. Sleep hy-
giene in physically dependent nursing home
residents. Sleep 1998;21(5):515–23.
 295

SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 295–306

Sleep Disturbances in Women:

Psychiatric Considerations
Claudio N. Soares, MD, PhD, FRCPCa,b,*,
Brian J. Murray, MD, FRCPC, D, ABSMc

- Insomnia in women: prevalence, risk Perinatal changes

factors, and clinical characteristics Menopausal transition, mood symptoms,
- Reproductive-related mood and sleep and sleep disorders
problems in women - Diagnosis and management of sleep
Perimenstrual changes, premenstrual disturbances in women: clinical
syndrome, and premenstrual dysphoric considerations
disorder - Summary
- References

Insomnia has been more frequently described in
women than in men [1–3]. The risk for developing
insomnia among women is 1.3- to 1.8-fold greater
than that observed in men [4]. Moreover, insomnia
and other sleep disturbances have been reported in
association with female reproductive cycle events,
such as premenstrual periods [5], pregnancy [6,7],
and menopause [8]. The association between
changes and fluctuations in female sex hormones
and the occurrence of sleep disturbances has be-
come a subject of increasing research and some
controversy [4]. Reports suggest, for example, that
women are more likely to experience insomnia
during perimenopausal and early postmenopausal
years [9]. Insomnia is a common complaint during
the menopausal transition, affecting up to 60%
of women [10–12]. Putative causes of insomnia
associated with the menopausal transition and
postmenopausal years include the occurrence and
severity of nocturnal hot flashes, mood disorders,
and sleep-disordered breathing [8,13–15].
This article reviews what is currently known
about sleep problems in women, with emphasis
on the association between common sleep distur-
bances and the presence and co-occurrence of
psychiatric conditions across the female lifespan.
Existing evidence on the clinical characteristics,
risk factors, and treatment options for insomnia
in women and the potential implications of con-
comitant psychiatric problems for diagnosis and
treatment are reviewed. Lastly, some of the hor-
monal and nonhormonal strategies for the clinical
management of sleep disturbances in women are
discussed.

a
Department of Psychiatry and Behavioural Neurosciences, McMaster University, Ontario, Canada
b
Women’s Health Concerns Clinic, St. Joseph’s Healthcare Hamilton, 301 James Street South, FB #638,
Hamilton, Ontario L8P 3B6, Canada
c
Sunnybrook Health Sciences Centre, 76 Grenville Street, Burton Hall 418, Toronto, Ontario M5S 1B2,
Canada
* Corresponding author. Women’s Health Concerns Clinic, McMaster University, St. Joseph’s Healthcare
Hamilton, 301 James Street South, FB #638, Hamilton, Ontario L8P 3B6, Canada.
E-mail address: csoares@mcmaster.ca (C.N. Soares).

1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jsmc.2008.01.006
sleep.theclinics.com
296 Soares & Murray
Insomnia in women: prevalence, risk factors,
and clinical characteristics
The National Sleep Foundation’s Women and Sleep
Poll, involving 1012 women aged 30 to 60 years,
found that 53% of women experienced one or
more symptoms of insomnia during the month
before the assessment; the most common symptom
reported, more associated with sleep quality, was
‘‘feeling tired upon awakening’’ (35% of the
women). Other characteristics included difficulty
falling asleep; several awakenings during the night
followed by difficulties in getting back to sleep;
and waking too early in the morning (reported by
approximately 20% of the women surveyed) [16].
Female gender has been consistently identified as
a risk factor for insomnia, particularly in commu-
nity-based studies [3,17–20]. When compared
with men, women 65 years of age or older have
more trouble falling asleep (36% versus 29%);
wake up too early (31% versus 21%); or are unable
to fall asleep again (25% versus 20%) [18]. Gender
differences also exist in younger populations; in
a survey of 529 participants aged 20 to 45 years,
there were a significantly higher percentage of
women reporting difficulties maintaining sleep
(20% versus 10%, female versus male) [21]. The
gender difference remained significant after adjust-
ments for age, smoking, and psychologic status.
These findings are consistent with other reports of
difficulty initiating or maintaining sleep in a signif-
icant percentage of women, particularly among
certain subgroups, such as the elderly [22]; subjects
with depression [23]; or those that are medically ill
[24,25], with pain or dyspnea [26,27].
Various factors contribute to a heightened risk for
insomnia in women compared with men [28]. The
presence of a psychiatric disorder is a strong risk fac-
tor for insomnia in both men and women [29];
however, women are at higher risk (lifetime) for
the development of depression and anxiety disor-
ders [30,31], which may ultimately contribute to
their heightened prevalence of insomnia. Lindberg
and colleagues [21] examined the presence of
psychologic symptoms as a possible explanation
for gender differences in sleep disturbances in
529 young adults (aged 20–45 years). There was
a higher percentage of women suffering from anxi-
ety symptoms (33% versus 19%, females compared
with males, respectively); in addition, women with
anxiety showed an increased prevalence of difficulty
initiating sleep (9% versus 3%), difficulty maintain-
ing sleep (32% versus 15%), and early morning
awakening (10% versus 3%), compared with those
without anxiety. In contrast, men with or without
anxiety did not show significant differences respect
to sleep disturbances.
These findings suggest that women might be
more susceptible to develop sleep disturbances
concomitant to depression or anxiety. It is impor-
tant to note, however, that the possible relation-
ships between sleep difficulties and psychiatric
disorders are numerous. Insomnia and other sleep
disturbances can be precursors to the onset of
major depressive disorder, so they may act as risk
factors for or predictors of depression [20,32–34].
Conversely, persistent altered sleep is one of the
perpetuating factors for difficult-to-treat depressive
episodes; recent studies even suggest that treatment
of depression with concomitant insomnia could be
optimized with the use of sleep-inducing agents in
association with antidepressants [35].
Other female-specific risk factors have been iden-
tified in studies involving subpopulations of differ-
ent ethnic or racial backgrounds. In a study of
gender differences in insomnia in a Chinese popu-
lation (N 5 9851, aged 18–65 years), women who
were divorced or widowed were more susceptible to
insomnia than men; exposure to environmental
noise at night and frequent alcohol intake also con-
tributed more significantly to the occurrence of
insomnia in women compared with men [1]. In
Finland, in a community survey of 1600 individ-
uals, the presence of worries, human relationship
problems, and regrets were the top-ranked risk
factors affecting sleep, more significantly among
women (19%) than men (13%) [36].
The relative contribution of sex steroids and
reproductive-related events for the occurrence of
gender differences in insomnia is complex and still
understudied. Existing data on this topic are exam-
ined in more detail in the next few sections.
Reproductive-related mood and sleep
problems in women
Perimenstrual changes, premenstrual
syndrome, and premenstrual dysphoric
disorder
For some women, sleep problems may emerge sec-
ondary to menstrual symptoms (eg, cramping,
bloating, headaches, and tender breasts) or dys-
menorrhea [37,38]. Polysomnographic (PSG) data
indicate that women with dysmenorrhea experience
less efficient sleep and more wakefulness than
women without painful menstrual cycles [39].
Polycystic ovarian syndrome also increases the
occurrence of sleep-disordered breathing, which
can contribute to sleep fragmentation and daytime
sleepiness [40].
There have been reports of hypersomnia [41] and
insomnia [42] temporally linked to the premen-
strual phase of the menstrual cycle. Despite fre-
quent subjective reports of sleep being disrupted

during this period of time, consistent evidences of
changes in sleep architecture (eg, PSG measures)
associated with symptomatic premenstrual periods
are lacking. The inconsistency in the literature is
likely related to small sample sizes studied, and
significant heterogeneity of underlying sleep or en-
docrine characteristics assessed [9].
Although clinically intuitive, some [42] but not
all [43] studies corroborated the existence of signif-
icant sleep disruption associated with premenstrual
complaints, premenstrual syndrome, or with the
diagnosis of premenstrual dysphoric disorder [44–
46]. Premenstrual dysphoric disorder constitutes
a more severe form of premenstrual syndrome
and affects 3% to 9% of women during reproduc-
tive years; women with premenstrual dysphoric
disorder report significant dysphoria (irritability,
mood swings, and aggressive behavior) and func-
tional impairment during the late luteal phase of
the menstrual cycle (3–10 symptomatic days) with
full remission of symptoms on onset of menses
[47]. More definitive data on altered sleep in pa-
tients with premenstrual dysphoric disorder are
lacking; in addition, most studies that explored
the efficacy of hormonal and nonhormonal treat-
ments for premenstrual dysphoric disorder have
focused primarily on mood and anxiety symptoms
rather than insomnia [48,49]. Selective serotonin
reuptake inhibitors are the treatment of choice for
premenstrual dysphoric disorder and it is reason-
able to expect that sleep disturbances secondary to
mood swings and increased anxiety would respond
to such treatment [48]. The intermittent use of
benzodiazepines may also constitute an interesting
option for women with perimenstrual, circum-
scribed symptoms of anxiety and altered sleep and
low risk for potential abuse [50].
Perinatal changes
Sleep is substantially disrupted during pregnancy
and the postpartum period; prevalence rates for al-
tered sleep during pregnancy range from 15% to
80%, depending on the population studied and
the time of assessment (eg, higher rates during the
third trimester of pregnancy compared with first
trimester) [51]. Sleep in the postpartum period is
markedly disrupted by care of the infant [52].
Nonetheless, even in the absence of infant care,
the sleep that follows pregnancy is disrupted and
quite often does not return to the prepregnancy
state [53]. This may reflect the outcome of some
of the physiologic changes that had occurred over
the course of pregnancy or even the emergence of
anxiety and hypervigilance behaviors that some
mothers develop toward the newborn.
Survey studies indicate that women attribute
many different reasons to justify their altered sleep
Sleep Disturbances in Women 297
during pregnancy; during the first trimester, nausea
and vomiting are frequently associated with sleep
problems [54]; psychosocial stressors are also
reported particularly in cases involving first-time
pregnancies, unplanned pregnancies, or in the
absence of a solid psychosocial support or network
[55]. As pregnancy progresses into the second and
third trimesters, there are more reports of increased
number of awakenings, fatigue, leg cramps, and
shortness of breath. Overall, women are more likely
to go to bed earlier and frequently nap during the
day as they try to compensate for the disruption
of the nighttime sleep pattern [56].
Objective assessments of sleep during pregnancy
have produced inconsistent results, again likely
because of limited sample sizes and differing
research methodology. One study revealed changes
in sleep during the first trimester, with a significant
increase in total sleep time and decrease in slow
wave sleep (restorative sleep) compared with pre-
pregnancy assessments [7]. In this study, total sleep
time tended to return to normal levels at the end of
the third trimester, as women reported more fre-
quent awakenings and a decline in sleep efficiency.
Pregnant women who report sleep disturbances
during pregnancy should also be screened for
depressive symptoms. Pregnancy does not confer
any protective effect against depression and the
presence of disrupted sleep in this population
may be indicative of mood changes. In a recent pro-
spective study of 201 nondepressed pregnant
women (all taking antidepressants because of prior
history of depression but asymptomatic for at least
3 months before the conception) the decision of
discontinuing treatment early in pregnancy resulted
in a fivefold greater risk for relapse over time com-
pared with treatment maintenance. These results re-
inforce the importance of carefully assessing mood
symptoms in women at risk during pregnancy [57].
Hormonal changes in pregnant women may con-
tribute to the occurrence of sleep-disordered breath-
ing and altered sleep. Snoring, for example, tends to
increase during pregnancy, possibly caused by
changes in upper airway resistance, which is
sensitive to hormonal factors, particularly proges-
terone levels [58]. In addition, pregnant women,
particularly during the final trimester, seem to
have a heightened risk for developing sleep apnea
and restless legs syndrome [59]. Studies suggest
that a restless legs syndrome might be more associ-
ated with lower ferritin and folate levels. Nutritional
supplements should be considered for symptomatic
individuals or subpopulations at risk [60].
Depletion of iron can be noted in basal ganglia
structures in the brain of patients with restless legs
syndrome [61], which is even more interesting,
given the roles of these structures for mood
298 Soares & Murray
regulation [62]. Furthermore, iron is the rate-limit-
ing step in the synthesis of dopamine in the brain.
The relationship between iron loss, restless legs syn-
drome, and peripartum mood disorders warrants
further investigation, because iron deficiency re-
mains common in women of childbearing age [63].
Other factors contribute to sleep-disordered
breathing over the course of pregnancy including
weight gain, mucosal edema, and changes in respi-
ratory mechanics [64]. More commonly, upper air-
way resistance increases throughout pregnancy
[65], and may contribute to frequent arousals and
sleep maintenance insomnia with subsequent day-
time sleepiness.
Interestingly, disrupted sleep during pregnancy
has been associated with poor obstetric outcomes,
particularly length of labor and type of delivery.
In a prospective, longitudinal follow-up of 131 preg-
nant women, Lee and Gay [66] demonstrated that
women who slept less than 6 hours at night had
longer labors and were 4.5 times more likely to
have cesarean deliveries. The researchers high-
lighted the importance of monitoring sleep quan-
tity and quality during prenatal assessments as
potential predictors of labor duration and delivery
type.
The assessment of changes in sleep architecture
and sleep efficiency during the postpartum period
constitutes a challenging task given the hormonal
changes inherent to the postpartum period, sleep
deprivation caused by frequent awakenings associ-
ated with breastfeeding, and the presence and sever-
ity of postnatal mood and behavior changes.
Objective assessments (PSG) of women immedi-
ately following delivery indicate an increase in
wake time after sleep onset, and awakenings even
in the absence of nocturnal activities involving the
infant’s care and nursing [53]. Women in the post-
partum period develop lower sleep efficiency, with
shorter latency to rapid eye movement sleep (char-
acteristic of depression) and reduction in total sleep
time. Although most studies suggest that most
awakenings and changes in sleep/wake ratios are
directly related to the newborn’s direct nursing
care, it has been hypothesized that other factors
could contribute to disrupted sleep patterns; these
include an abrupt decline in progesterone immedi-
ately after delivery, a hormone with known sedative
properties, and changes in melatonin levels, the lat-
ter possibly affecting the normal circadian rhythm
[9]. Moreover, the occurrence of increased irritabil-
ity and mood swings postpartum (postnatal blues)
could lead to a disruption of new mothers’ sleep
efficiency and increase the risk for subsequent de-
velopment of postpartum depression [67]. Another
possibility is that the sleep disruption itself directly
contributes to many of these mood changes.
Preliminary, but intriguing, evidence suggests
that some strategies to promote protected sleep
time immediately following delivery (eg, prolonged
hospital stay, ‘‘rooming out’’ infants under the part-
ner or nurse’s care, and use of sedatives as needed)
may reduce the risk of clinically significant depres-
sion or anxiety in the first few months of the post-
partum period (Steiner, personal communication,
2006).
Menopausal transition, mood symptoms,
and sleep disorders
For some women, the menopausal transition is
characterized by the occurrence of vasomotor
symptoms (eg, hot flashes, night sweats), sleep dis-
turbances, and changes in sexual function that can
adversely affect quality of life [68]. More recently,
several studies revealed that women entering peri-
menopause are at higher risk for developing depres-
sion, even in the absence of prior episodes of
depression [69–71].
About 45% to 75% of women in the menopausal
transition experience hot flashes [72]. Hot flashes
are transient sensations of heat dissipation, and
may be accompanied by palpitations, nausea, dizzi-
ness, headache, and ultimately insomnia [73]. The
presence of nocturnal hot flashes is commonly
associated with sleep disturbances [9]. The patho-
physiology of hot flashes is not fully understood,
but it is thought to be mediated through the ante-
rior hypothalamus, an area that regulates tempera-
ture and sleep [13,74]. The existence of a more
direct association between hot flashes and insom-
nia has been considered controversial, because
most studies using objective sleep measures (ie,
PSG) or subjective measures (ie, sleep question-
naires) have shown inconsistent findings. Early
PSG studies in menopausal women suggested a sig-
nificant correlation between hot flashes and sleep
disturbances [75], with estrogen therapy resulting
in a significant reduction of sleep latency, frequency
of awakenings, and improvement of rapid eye
movement sleep [76]. More recent studies, however,
failed to show positive effects of hormone therapy
on sleep architecture in menopausal women [77].
Many women who develop depressive symptoms
during the menopausal years also experience hot
flashes and insomnia [71], suggesting a potential
association between sleep disruption and the occur-
rence or severity of vasomotor symptoms, leading
to an adverse impact on mood and well-being
[78]. Depression has been found in some meno-
pausal women, however, even in the absence of
clinically significant hot flashes [79].
It is possible that fluctuating estrogen (E2) levels
during the perimenopause exacerbate the hypotha-
lamic-pituitary-adrenal axis activity in response

to stressful events, resulting in mood and sleep
disorders [80]. Estrogen modulates the synthesis,
release, and metabolism of monoamines that affect
mood, behavior, and sleep [81]. More specifically,
estrogen exerts an agonist effect on serotonergic
activity by increasing the number of serotonin
receptors, increasing transport and uptake of the
neurotransmitter, increasing synthesis of serotonin,
up-regulating 5-HT1 receptors, down-regulating
5-HT2 receptors, and decreasing monoamine oxi-
dase activity [82]. It is intuitive that intense estrogen
fluctuations and depletion could affect different
monoaminergic systems leading to disrupted sleep,
mood changes, and the emergence of vasomotor
symptoms [83].
Fluctuations in estrogen and progesterone may
also alter g-aminobutyric acid (GABA) function
and play a role in menopause-related insomnia
[84]. GABA acid plays a central role in sleep initia-
tion and maintenance. Preclinical studies have
shown that gonadal hormones have a barbiturate-
like action on the GABA receptor complex. Estrogen
reduces the number of GABA-A receptors, but
progesterone has been shown to counter that effect
[85].
Several factors may contribute to the occurrence
of obstructive sleep apnea syndrome in meno-
pausal women; diminishing progesterone levels
may be a factor, because progesterone is a known
respiratory stimulant and upper airway dilator
[86]. Increased body weight associated with meno-
pause may also be an important factor, although an
increased risk of obstructive sleep apnea syndrome
has been observed independently of body weight
[87].
Diagnosis and management of sleep
disturbances in women: clinical
considerations
The first step in management is to establish a clear
diagnosis. A careful interview of the patient (and
more importantly the bed partner, who is an objec-
tive observer for behaviors of which the patient may
be unaware) is essential. To address women’s sleep
problems fully, a routine history and physical exam-
ination is required. When necessary, hormonal
status (ie, serum follicle-stimulating hormone, E2,
progesterone concentrations) to confirm meno-
pausal staging should be obtained. A typical day’s
sleep behavior can be obtained chronologically
and should include details of sleep onset, sleep
maintenance, and daytime alertness.
Often, a sleep problem has both medical and
behavioral components. Both factors need to be
addressed to ensure clinical improvement. To that
end, it is important to ensure that medical causes
Sleep Disturbances in Women 299
of insomnia (restless legs syndrome in particular
for sleep initiation difficulties, and sleep-disordered
breathing for sleep maintenance problems) have
been adequately treated before embarking on
behavioral management. Some intrinsic sleep dis-
orders, such as poor sleep hygiene or restless legs
syndrome, can be established purely based on
history, whereas others, such as obstructive sleep
apnea, require formal PSG assessment.
Studies on nonpharmacologic and pharmaco-
logic therapies for female-specific sleep disorders
are scarce. In pregnancy and breast-feeding scenar-
ios, special attention must be made to the safety of
medications that may be considered for treatment.
Unfortunately, many medications are not formally
tested in these populations, leaving significant
knowledge gaps. Nonpharmacologic therapies
become critically important for these patients. In
addition, some patients are reluctant to use medica-
tions for insomnia, particularly because of the pos-
sibility of habituation, withdrawal, and rebound
insomnia with the use of pharmacologic therapies
[88]. A greater acceptability of the use of nonphar-
macologic interventions, however, has other obsta-
cles including limited number of controlled
studies for specific interventions (particularly for
herbal preparations); cost and lack of insurance
reimbursement; and slower speed of onset [89,90].
Behavioral approaches provide similar short-
term clinical benefits for the management of in-
somnia compared with pharmacologic therapies,
but have better long-term benefits, especially in
older adults [90]. Overall, stimulus control and
sleep-restriction therapies are the most effective
behavioral approaches [91]. Other behavioral tech-
niques include meditation, biofeedback, and
hypnotherapy. Hypnotherapy has shown some effi-
cacy in small trials [92]. In one of these studies,
hypnotherapy was used in women treated for breast
cancer, and had a positive impact on hot flashes
and improved quality of sleep [93].
Data on the efficacy of behavioral therapies for
the treatment of insomnia in menopausal women
are limited. Given many unique factors contribut-
ing to insomnia in this population (ie, the higher
incidence of hot flashes and sleep-disordered
breathing) it is difficult to predict the extent to
which these treatments would be efficacious when
used alone.
For symptomatic menopausal women, hormone
therapy has been the treatment of choice for most
menopausal complaints, including hot flashes and
insomnia [77,94]. There is also evidence for its effi-
cacy for the treatment of depression during the
menopausal transition and, to a lesser extent, for
obstructive sleep apnea syndrome [95–98]. The
long-term safety of hormone therapy has been

Table 1: Clinical scenarios and treatment strategies for the management of sleep and mood disorders in women across the reproductive life cycle
Clinical scenario
1. A young woman (mid-20s) reports disrupted sleep associated with
premenstrual discomfort (back pain, menstrual migraines, bloating,
increased irritability) and menstrual irregularity. Symptoms seem to
remit with the onset of menses. No significant history of sleep
disorders or mood symptoms are experienced except for the ones
noted during the premenstrual periods.
2. A woman in her mid-30s is pregnant (second trimester) and reports
increasing sleep difficulties (sleep onset and sleep maintenance)
associated with low motivation, increasing marital conflicts caused
by her irritability, and ruminative thoughts about her baby’s health
and her financial conditions. Prior history of depression (several
depressive episodes), but stable for many years while using
antidepressants (SSRIs); recent treatment discontinuation (at 6 weeks
of gestation) because of pregnancy.
300
Soares & Murray
Treatment strategies and points to consider
1. Rule out an underlying mood or sleep disorder and confirm possible
diagnosis of severe PMS or PMDD through careful history and prospective
tracking of symptoms.
2. Oral contraceptives may help to regulate cycles and promote alleviation
of PMS symptoms; additional benefits for sleep symptoms are unknown.
3. Antidepressant use, particularly SSRIs, is the treatment of choice for the
management of both depressive and somatic symptoms. Its use may have
a positive impact on disrupted sleep.
4. Sleep hygiene measures and CBT for depression and for sleep symptoms
may be helpful; efficacy data focused on these specific subpopulations
are scarce.
1. Possible recurrence of major depressive disorder, with concomitant
symptoms of anxiety and disrupted sleep. Careful psychiatric assessment
(including the evaluation of potential risk for suicidal thoughts and
behaviors) is recommended.
2. Risks/benefits of reintroducing treatment with SSRIs should be reviewed
with the patient and her partner or family members.
3. Benzodiazepines could be a viable treatment strategy for the alleviation
of these symptoms with careful monitoring of use throughout the
remaining time of pregnancy.
4. Sleep-breathing disorder should be ruled out by history and partner’s
information.
5. Sleep hygiene measures and CBT for depression and for sleep symptoms
may be helpful; efficacy data focused on these specific subpopulations are
scarce.
 3. A patient (early-50s) reports amenorrhea for 2 years, accompanied
by the occurrence of moderate vasomotor symptoms (4–6 hot flashes
per day, night sweats). Significant sexual dysfunction, mainly low
libido and dryness. Decreased energy, cognitive decline (difficulty
concentrating), and lack of motivation have become more evident
over the past year along with sleep problems (constant awakenings,
poor sleep efficiency).
Abbreviations: CBT, cognitive-behavioral therapy; PMDD, premenstrual dysphoric disorder; PMS, premenstrual syndrome; SSRIs, selective serotonin reuptake inhibitors.
1. Possible occurrence of menopause-related symptoms (vasomotor
symptoms, sexual dysfunction).
2. Further assessment is needed to characterize the severity of mood and
sleep problems and whether these symptoms constitute a new complaint
(in the context of the menopause) or a re-emergence of prior mood and
sleep disorders.
3. If there are no contraindications to hormone therapies, consider a trial
with estrogens, preferably transdermal E2 (50–100 mg/day). If efficacious,
response (mood, sleep, sexual function, and vasomotor symptoms)
should be observed with 4–8 weeks of treatment.
4. The use of progestins is fundamental in patients with intact uterus;
patient might benefit from using micronized progesterone (100–200
mg/day, either continuously or cyclically) because of sedating effects.
5. Patients with partial response may benefit from the use of
antidepressants in combination with hormone therapy. Consider the use
of antidepressants with a more favorable profile with respect to sexual
adverse events.
6. Consider sleep hygiene measures and behavioral techniques to address
sleep problems.
7. Consider the use of nonbenzodiazepines (eg, zolpidem, eszopiclone)
to improve sleep problems.
8. CBT for depression may be helpful, despite scarce efficacy data available
for these specific subpopulations.
Sleep Disturbances in Women
301
302 Soares & Murray
questioned, however, particularly in light of the
results from the Women’s Health Initiative study
[99]. Consequently, many physicians and patients
are seeking nonhormonal therapies for the relief
of menopausal symptoms; as a result, herbal prep-
arations have gained considerable attention.
Most herbal preparations have showed inconsis-
tent results in well-controlled trials. Valerian (Vale-
rian officinalis) has sedative and muscle-relaxant
effects and its use has become increasingly popular
among women [100]. The evidence for its efficacy as
a treatment for insomnia is inconclusive, based on
a systematic review of nine clinical trials [101]. In
addition, sleep benefit may be delayed for 2 weeks
and some patients experience residual daytime
effects [102]. Cimicifuga racemosa (black cohosh)
has shown some positive results for the treatment
of menopause-related physical and psychologic
complaints in some [103–105] but not all [106]
studies. No specific studies with black cohosh for
the treatment of menopause-related insomnia
have been published to date.
The use of antidepressants in perimenopausal
and postmenopausal women (eg, paroxetine, cita-
lopram, mirtazapine) may have an indirect effect
on sleep and quality of life by improving other
menopause-related symptoms (eg, vasomotor
symptoms, pain, and mood swings); these agents
might constitute a treatment option for symptom-
atic women who are unable or unwilling to receive
hormone therapies [107–109].
Other potential pharmacologic therapies include
the use of benzodiazepines and non–benzodiaze-
pine receptor agonists. Benzodiazepines primarily
modulate GABA-A receptors and exert their effect
on the benzodiazepine site of the receptors, thereby
potentiating the inhibitory effect of GABA on neu-
rotransmission [110]. Non–benzodiazepine recep-
tor agonists, also known as ‘‘Z-compounds’’ (eg,
zolpidem, zaleplon, and eszopiclone), are recom-
mended for short-term treatment of insomnia;
some have demonstrated sustained efficacy and
safety without the development of tolerance or re-
bound insomnia over 6 months to 1 year [111,112].
A 4-week, placebo-controlled trial (N 5 141)
with zolpidem showed statistically significant
improvement in some parameters of insomnia
(increased total sleep time, decreased wake time
after sleep onset, and number of awakenings) in
perimenopausal and postmenopausal women;
however, the study did not assess whether zolpidem
affected positively or negatively other menopause-
related complaints [113].
In a larger study (N 5 410), eszopiclone showed
superior efficacy compared with placebo for the
treatment of insomnia in perimenopause and early
menopause subjects who had developed insomnia
in the context of transitioning to menopause, and
reported other menopause-related symptoms (eg,
hot flashes, night sweats) without comorbid depres-
sion or anxiety [114]. Subjects receiving eszopiclone
reported significantly greater improvement in sleep
induction; sleep maintenance (total awakenings,
awakenings caused by hot flashes, and time awake
after sleep onset); sleep duration; sleep quality;
and next-day functioning. The use of eszopiclone
resulted in improvement of quality of life and other
menopause-specific symptoms, assessed by the
Greene Climacteric Scale, the Menopause-Specific
Quality of Life questionnaire, and by changes in
the family life and home disability domains of
the Sheehan Disability Scale.
Summary
Women commonly report sleep disorders that may
predispose or contribute to the occurrence of other
psychiatric problems, particularly mood and anxi-
ety disorders. In addition, many reproductive life
events in women are associated with unique fea-
tures of sleep disruption that are at least partially
modulated by hormonal factors. Pregnancy, peri-
partum, and perimenopausal states are particularly
associated with sleep disruption, and have the clear-
est published evidence base. Many factors contrib-
ute to a heightened incidence of insomnia in
women, including the occurrence and severity of
somatic symptoms (eg, hot flashes, dysmenorrhea);
psychiatric symptoms, such as depression or anxi-
ety; and intrinsic sleep disorders. The presence of
insomnia has been associated with poorer quality
of life and impaired daytime functioning; its man-
agement is imperative. Table 1 explores some of
the common clinical scenarios in which sleep disor-
ders may occur in women associated with reproduc-
tive life events.
Given the strong possibility that sleep disorders
are contributing to mood changes during preg-
nancy and postpartum, careful attention must be
taken for the development of treatment strategies
that address both medical and psychologic factors
to assist fully this vulnerable population. Hormone
therapy is still the treatment of choice for short-
term management of menopause-related symp-
toms, particularly hot flashes; however, its safety
as a long-term treatment option has been ques-
tioned and many physicians and their patients are
seeking nonhormonal alternatives. An accurate
diagnosis must be made in this population, and
treatment of underlying mood disorders is crucial.
Most nonpharmacologic symptomatic treatments
for insomnia (eg, stimulus control, sleep restric-
tion) have not been systematically studied in men-
opausal patients, despite their proved efficacy for

insomnia in general. The use of antidepressants
seems to be more helpful when sleep complaints re-
sult from the emergence of menopause-related de-
pression or anxiety.
Women who suffer insomnia during the meno-
pausal transition or postmenopause may benefit
from the therapeutic effects of newer agents, such
as the non–benzodiazepine receptor agonists (zol-
pidem, eszopiclone), and melatonin receptor an-
tagonists; more studies are needed to determine
the extent to which these agents can offer a safe
and efficacious long-term management of insom-
nia, with a positive impact on quality of life and
overall functioning.
References
[1] Li RH, Wing YK, Ho SC, et al. Gender differ-
ences in insomnia: a study in the Hong Kong
Chinese population. J Psychosom Res 2002;
53(1):601–9.
[2] Klink ME, Quan SF, Kaltenborn WT, et al. Risk
factors associated with complaints of insomnia
in a general adult population: influence of
previous complaints of insomnia. Arch Intern
Med 1992;152(8):1634–7.
[3] Rocha FL, Guerra HL, Lima-Costa MF. Preva-
lence of insomnia and associated socio-
demographic factors in a Brazilian community:
the Bambui study. Sleep Med 2002;3(2):121–6.
[4] Krystal AD. Insomnia in women. Clin Corner-
stone 2003;5(3):41–50.
[5] Schenck CH, Mahowald MW. Two cases of
premenstrual sleep terrors and injurious sleep-
walking. J Psychosom Obstet Gynaecol 1995;
16(2):79–84.
[6] Sahota PK, Jain SS, Dhand R. Sleep disorders in
pregnancy. Curr Opin Pulm Med 2003;9(6):
477–83.
[7] Lee KA, Zaffke ME. Longitudinal changes in
fatigue and energy during pregnancy and the
postpartum period. J Obstet Gynecol Neonatal
Nurs 1999;28(2):183–91.
[8] Dennerstein L, Dudley EC, Hopper JL, et al. A
prospective population-based study of meno-
pausal symptoms. Obstet Gynecol 2000;96(3):
351–8.
[9] Moline ML, Broch L, Zak R, et al. Sleep in women
across the life cycle from adulthood through
menopause. Sleep Med Rev 2003;7(2):155–77.
[10] Owens JF, Matthews KA. Sleep disturbance in
healthy middle-aged women. Maturitas 1998;
30(1):41–50.
[11] Moline M, Broch L, Zak R. Sleep problems
across the life cycle in women. Curr Treat
Options Neurol 2004;6(4):319–30.
[12] Shin C, Lee S, Lee T, et al. Prevalence of insom-
nia and its relationship to menopausal status in
middle-aged Korean women. Psychiatry Clin
Neurosci 2005;59(4):395–402.
Sleep Disturbances in Women 303
[13] Ohayon MM. Severe hot flashes are associated
with chronic insomnia. Arch Intern Med
2006;166(12):1262–8.
[14] Shahar E, Redline S, Young T, et al. Hormone
replacement therapy and sleep-disordered
breathing. Am J Respir Crit Care Med 2003;
167(9):1186–92.
[15] Krystal AD, Edinger J, Wohlgemuth W, et al.
Sleep in peri-menopausal and post-menopausal
women. Sleep Med Rev 1998;2(4):243–53.
[16] National-Sleep-Foundation. Sleep in America
Poll 2005. vol. 2005, 2005. Available at: http://
www.sleepfoundation.org. Accessed July 2005.
[17] Radecki SE, Brunton SA. Management of
insomnia in office-based practice: national
prevalence and therapeutic patterns. Arch Fam
Med 1993;2(11):1129–34.
[18] Foley DJ, Monjan AA, Brown SL, et al. Sleep
complaints among elderly persons: an epidemi-
ologic study of three communities. Sleep 1995;
18(6):425–32.
[19] Bixler EO, Vgontzas AN, Lin HM, et al. Preva-
lence of sleep-disordered breathing in women:
effects of gender. Am J Respir Crit Care Med
2001;163(3 Pt 1):608–13.
[20] Neckelmann D, Mykletun A, Dahl AA. Chronic
insomnia as a risk factor for developing anxiety
and depression. Sleep 2007;30(7):873–80.
[21] Lindberg E, Janson C, Gislason T, et al. Sleep dis-
turbances in a young adult population: can gen-
der differences be explained by differences in
psychological status? Sleep 1997;20(6):381–7.
[22] Liu X, Liu L. Sleep habits and insomnia in
a sample of elderly persons in China. Sleep
2005;28(12):1579–87.
[23] McCall WV, Reboussin BA, Cohen W. Subjective
measurement of insomnia and quality of life in
depressed inpatients. J Sleep Res 2000;9(1):
43–8.
[24] Gislason T, Almqvist M. Somatic diseases and
sleep complaints: an epidemiological study of
3,201 Swedish men. Acta Med Scand 1987;
221(5):475–81.
[25] Davison SN, Jhangri GS. The impact of chronic
pain on depression, sleep, and the desire to with-
draw from dialysis in hemodialysis patients.
J Pain Symptom Manage 2005;30(5):465–73.
[26] Ohayon MM. Relationship between chronic
painful physical condition and insomnia. J Psy-
chiatr Res 2005;39(2):151–9.
[27] Meuser T, Pietruck C, Radbruch L, et al. Symp-
toms during cancer pain treatment following
WHO-guidelines: a longitudinal follow-up
study of symptom prevalence, severity and eti-
ology. Pain 2001;93(3):247–57.
[28] Soares CN. Insomnia in women: an overlooked
epidemic? Arch Womens Ment Health 2005;
8(4):205–13.
[29] Mallon L, Broman JE, Hetta J. Relationship be-
tween insomnia, depression, and mortality:
a 12-year follow-up of older adults in the com-
munity. Int Psychogeriatr 2000;12(3):295–306.
304 Soares & Murray
[30] Kessler RC, McGonagle KA, Swartz M, et al. Sex
and depression in the National Comorbidity
Survey. I: lifetime prevalence, chronicity and re-
currence. J Affect Disord 1993;29(2–3):85–96.
[31] Piccinelli M, Wilkinson G. Gender differences
in depression: critical review. Br J Psychiatry
2000;177:486–92.
[32] Fava M. Daytime sleepiness and insomnia as
correlates of depression. J Clin Psychiatry
2004;65(Suppl 16):27–32.
[33] Morphy H, Dunn KM, Lewis M, et al. Epidemi-
ology of insomnia: a longitudinal study in a UK
population. Sleep 2007;30(3):274–80.
[34] Perlis ML, Smith LJ, Lyness JM, et al. Insomnia
as a risk factor for onset of depression in the
elderly. Behav Sleep Med 2006;4(2):104–13.
[35] Fava M, McCall WV, Krystal A, et al. Eszopiclone
co-administered with fluoxetine in patients
with insomnia coexisting with major depressive
disorder. Biol Psychiatry 2006;59(11):1052–60.
[36] Urponen H, Vuori I, Hasan J, et al. Self-
evaluations of factors promoting and disturb-
ing sleep: an epidemiological survey in Finland.
Soc Sci Med 1988;26(4):443–50.
[37] Baker FC, Driver HS. Self-reported sleep across
the menstrual cycle in young, healthy women.
J Psychosom Res 2004;56(2):239–43.
[38] Cluydts R, Visser P. Mood and sleep. I. Effects of
the menstrual cycle. Waking Sleeping 1980;
4(3):193–7.
[39] Baker FC, Driver HS, Rogers GG, et al. High
nocturnal body temperatures and disturbed
sleep in women with primary dysmenorrhea.
Am J Physiol 1999;277(6 Pt 1):E1013–21.
[40] Vgontzas AN, Legro RS, Bixler EO, et al.
Polycystic ovary syndrome is associated with
obstructive sleep apnea and daytime sleepiness:
role of insulin resistance. J Clin Endocrinol
Metab 2001;86(2):517–20.
[41] Sachs C, Persson HE, Hagenfeldt K. Menstrua-
tion-related periodic hypersomnia: a case study
with successful treatment. Neurology 1982;
32(12):1376–9.
[42] Manber R, Bootzin RR. Sleep and the menstrual
cycle. Health Psychol 1997;16(3):209–14.
[43] Lee KA, McEnany G, Zaffke ME. REM sleep and
mood state in childbearing women: sleepy or
weepy? Sleep 2000;23(7):877–85.
[44] Mauri M, Reid RL, MacLean AW. Sleep in the
premenstrual phase: a self-report study of PMS
patients and normal controls. Acta Psychiatr
Scand 1988;78(1):82–6.
[45] Sheldrake P, Cormack M. Variations in men-
strual cycle symptom reporting. J Psychosom
Res 1976;20(3):169–77.
[46] Parry BL, Mostofi N, LeVeau B, et al. Sleep EEG
studies during early and late partial sleep
deprivation in premenstrual dysphoric disorder
and normal control subjects. Psychiatry Res
1999;85(2):127–43.
[47] Freeman EW, Sondheimer SJ. Premenstrual
dysphoric disorder: recognition and treatment.
Prim Care Companion J Clin Psychiatry 2003;
5(1):30–9.
[48] Steiner M, Pearlstein T, Cohen LS, et al. Expert
guidelines for the treatment of severe PMS,
PMDD, and comorbidities: the role of SSRIs.
J Womens Health (Larchmt) 2006;15(1):
57–69.
[49] Pearlstein TB, Bachmann GA, Zacur HA, et al.
Treatment of premenstrual dysphoric disorder
with a new drospirenone-containing oral
contraceptive formulation. Contraception
2005;72(6):414–21.
[50] Freeman EW. Luteal phase administration of
agents for the treatment of premenstrual
dysphoric disorder. CNS Drugs 2004;18(7):
453–68.
[51] Hedman C, Pohjasvaara T, Tolonen U, et al.
Effects of pregnancy on mothers’ sleep. Sleep
Med 2002;3(1):37–42.
[52] Shinkoda H, Matsumoto K, Park YM. Changes
in sleep-wake cycle during the period from
late pregnancy to puerperium identified
through the wrist actigraph and sleep logs.
Psychiatry Clin Neurosci 1999;53(2):133–5.
[53] Karacan I, Williams RL, Hursch CJ, et al. Some
implications of the sleep patterns of pregnancy
for postpartum emotional disturbances. Br
J Psychiatry 1969;115(525):929–35.
[54] Schweiger MS. Sleep disturbance in pregnancy:
a subjective survey. Am J Obstet Gynecol
1972;114(7):879–82.
[55] Walker LO, Cooney AT, Riggs MW. Psychosocial
and demographic factors related to health
behaviors in the 1st trimester. J Obstet Gynecol
Neonatal Nurs 1999;28(6):606–14.
[56] Mindell JA, Jacobson BJ. Sleep disturbances
during pregnancy. J Obstet Gynecol Neonatal
Nurs 2000;29(6):590–7.
[57] Cohen LS, Altshuler LL, Harlow BL, et al.
Relapse of major depression during pregnancy
in women who maintain or discontinue
antidepressant treatment. JAMA 2006;295(5):
499–507.
[58] Edwards N, Middleton PG, Blyton DM, et al.
Sleep disordered breathing and pregnancy.
Thorax 2002;57(6):555–8.
[59] Manconi M, Govoni V, De Vito A, et al. Restless
legs syndrome and pregnancy. Neurology 2004;
63(6):1065–9.
[60] Lee KA, Zaffke ME, Baratte-Beebe K. Restless
legs syndrome and sleep disturbance during
pregnancy: the role of folate and iron.
J Womens Health Gend Based Med 2001;
10(4):335–41.
[61] Earley CJ, Barker P, Horska A, et al. MRI-
determined regional brain iron concentrations
in early- and late-onset restless legs syndrome.
Sleep Med 2006;7(5):458–61.
[62] Bejjani BP, Damier P, Arnulf I, et al. Transient
acute depression induced by high-frequency
deep-brain stimulation. N Engl J Med 1999;
340(19):1476–80.

[63] Looker AC, Dallman PR, Carroll MD, et al.
Prevalence of iron deficiency in the United
States. JAMA 1997;277(12):973–6.
[64] Franklin KA, Holmgren PA, Jonsson F, et al.
Snoring, pregnancy-induced hypertension, and
growth retardation of the fetus. Chest 2000;
117(1):137–41.
[65] Edwards E, Whitaker-Azmitia PM, Harkins K.
5-HT1A and 5-HT1B agonists play a differential
role on the respiratory frequency in rats. Neuro-
psychopharmacology 1990;3:129–36.
[66] Lee KA, Gay CL. Sleep in late pregnancy predicts
length of labor and type of delivery. Am
J Obstet Gynecol 2004;191(6):2041–6.
[67] Ross LE, Murray BJ, Steiner M. Sleep and
perinatal mood disorders: a critical review.
J Psychiatry Neurosci 2005;30(4):247–56.
[68] Soares CN, Joffe H, Steiner M. Menopause and
mood. Clin Obstet Gynecol 2004;47(3):
576–91.
[69] Maartens LW, Knottnerus JA, Pop VJ. Meno-
pausal transition and increased depressive
symptomatology: a community based prospec-
tive study. Maturitas 2002;42(3):195–200.
[70] Freeman EW, Sammel MD, Lin H, et al. Associ-
ations of hormones and menopausal status
with depressed mood in women with no
history of depression. Arch Gen Psychiatry
2006;63(4):375–82.
[71] Cohen LS, Soares CN, Vitonis AF, et al. Risk for
new onset of depression during the meno-
pausal transition: the Harvard study of moods
and cycles. Arch Gen Psychiatry 2006;63(4):
385–90.
[72] Freeman EW, Sammel MD, Grisso JA, et al. Hot
flashes in the late reproductive years: risk
factors for African American and Caucasian
women. J Womens Health Gend Based Med
2001;10(1):67–76.
[73] Freedman RR. Physiology of hot flashes. Am
J Hum Biol 2001;13(4):453–64.
[74] Freedman RR. Biochemical, metabolic, and
vascular mechanisms in menopausal hot
flashes. Fertil Steril 1998;70(2):332–7.
[75] Erlik Y, Tataryn IV, Meldrum DR, et al. Associa-
tion of waking episodes with menopausal hot
flushes. JAMA 1981;245(17):1741–4.
[76] Schiff I, Regestein Q, Tulchinsky D, et al. Effects
of estrogens on sleep and psychological state of
hypogonadal women. JAMA 1979;242(22):
2405–4.
[77] Polo-Kantola P, Erkkola R, Irjala K, et al. Effect
of short-term transdermal estrogen replacement
therapy on sleep: a randomized, double-blind
crossover trial in postmenopausal women.
Fertil Steril 1999;71(5):873–80.
[78] Joffe H, Hennen J, Soares CN, et al. Hot flushes
associated with depression in perimenopausal
women seeking primary care. Menopause
2002;9(6):392–8.
[79] Freeman EW, Sammel MD, Liu L, et al. Hor-
mones and menopausal status as predictors
Sleep Disturbances in Women 305
of depression in women in transition to men-
opause. Arch Gen Psychiatry 2004;61(1):
62–70.
[80] Puder JJ, Freda PU, Goland RS, et al. Estrogen
modulates the hypothalamic-pituitary-adrenal
and inflammatory cytokine responses to endo-
toxin in women. J Clin Endocrinol Metab
2001;86(6):2403–8.
[81] McEwen BS, Alves SE. Estrogen actions in the
central nervous system. Endocr Rev 1999;
20(3):279–307.
[82] Halbreich U, Kahn LS. Role of estrogen in the
aetiology and treatment of mood disorders.
CNS Drugs 2001;15(10):797–817.
[83] Soares CN. Menopausal transition and depres-
sion: who is at risk and how to treat it? Expert
Rev Neurother 2007;7(10):1285–93.
[84] Krystal AD. Depression and insomnia in
women. Clin Cornerstone 2004;6(1):S19–28.
[85] Harrison NL, Majewska MD, Harrington JW,
et al. Structure-activity relationships for steroid
interaction with the gamma-aminobutyric acid
A receptor complex. J Pharmacol Exp Ther
1987;241(1):346–53.
[86] Young T, Palta M, Dempsey J, et al. The occur-
rence of sleep-disordered breathing among
middle-aged adults. N Engl J Med 1993;
328(17):1230–5.
[87] Young T, Rabago D, Zgierska A, et al. Objective
and subjective sleep quality in premenopausal,
perimenopausal, and postmenopausal women
in the Wisconsin Sleep Cohort Study. Sleep
2003;26(6):667–72.
[88] Erman MK. Therapeutic options in the treat-
ment of insomnia. J Clin Psychiatry 2005;
66(Suppl 9):18–23 [quiz: 42–3].
[89] Wang MY, Wang SY, Tsai PS. Cognitive behav-
ioural therapy for primary insomnia: a system-
atic review. J Adv Nurs 2005;50(5):553–64.
[90] Montgomery P, Dennis J. A systematic review of
non-pharmacological therapies for sleep prob-
lems in later life. Sleep Med Rev 2004;8(1):
47–62.
[91] Morin CM. Cognitive-behavioral approaches to
the treatment of insomnia. J Clin Psychiatry
2004;65(Suppl 16):33–40.
[92] Stanton HE. Hypnotic relaxation and the reduc-
tion of sleep onset insomnia. Int J Psychosom
1989;36(1–4):64–8.
[93] Younus J, Simpson I, Collins A, et al. Mind
control of menopause. Womens Health Issues
2003;13(2):74–8.
[94] Boyle GJ, Murrihy R. A preliminary study of
hormone replacement therapy and psychologi-
cal mood states in perimenopausal women.
Psychol Rep 2001;88(1):160–70.
[95] Schmidt PJ, Nieman L, Danaceau MA, et al.
Estrogen replacement in perimenopause-related
depression: a preliminary report. Am J Obstet
Gynecol 2000;183(2):414–20.
[96] Soares CN, Almeida OP, Joffe H, et al. Efficacy
of estradiol for the treatment of depressive
306 Soares & Murray
disorders in perimenopausal women: a double-
blind, randomized, placebo-controlled trial.
Arch Gen Psychiatry 2001;58(6):529–34.
[97] Cistulli PA, Barnes DJ, Grunstein RR, et al.
Effect of short-term hormone replacement in
the treatment of obstructive sleep apnoea in
postmenopausal women. Thorax 1994;49(7):
699–702.
[98] Keefe DL, Watson R, Naftolin F. Hormone
replacement therapy may alleviate sleep apnea
in menopausal women: a pilot study. Meno-
pause 1999;6(3):196–200.
[99] Rossouw JE, Anderson GL, Prentice RL, et al.
Risks and benefits of estrogen plus progestin
in healthy postmenopausal women: principal
results from the women’s health initiative
randomized controlled trial. JAMA 2002;
288(3):321–33.
[100] Wheatley D. Kava and valerian in the treatment
of stress-induced insomnia. Phytother Res
2001;15(6):549–51.
[101] Stevinson C, Ernst E. Valerian for insomnia:
a systematic review of randomized clinical
trials. Sleep Med 2000;1(2):91–9.
[102] Larzelere MM, Wiseman P. Anxiety, depression,
and insomnia. Prim Care 2002;29(2):339–60.
[103] Vermes G, Banhidy F, Acs N. The effects of
remifemin on subjective symptoms of meno-
pause. Adv Ther 2005;22(2):148–54.
[104] Pockaj BA, Loprinzi CL, Sloan JA, et al. Pilot
evaluation of black cohosh for the treatment
of hot flashes in women. Cancer Invest 2004;
22(4):515–21.
[105] Uebelhack R, Blohmer JU, Graubaum HJ, et al.
Black cohosh and St. John’s wort for climacteric
complaints: a randomized trial. Obstet Gynecol
2006;107(2 Pt 1):247–55.
[106] Newton KM, Reed SD, LaCroix AZ, et al.
Treatment of vasomotor symptoms of meno-
pause with black cohosh, multibotanicals, soy,
hormone therapy, or placebo: a randomized
trial. Ann Intern Med 2006;145(12):869–79.
[107] Soares CN, Poitras JR, Prouty J, et al. Efficacy of
citalopram as a monotherapy or as an adjunc-
tive treatment to estrogen therapy for perime-
nopausal and postmenopausal women with
depression and vasomotor symptoms. J Clin
Psychiatry 2003;64(4):473–9.
[108] Stearns V, Beebe KL, Iyengar M, et al. Paroxetine
controlled release in the treatment of meno-
pausal hot flashes: a randomized controlled
trial. JAMA 2003;289(21):2827–34.
[109] Soares CN, Arsenio H, Joffe H, et al. Escitalo-
pram versus ethinyl estradiol and norethin-
drone acetate for symptomatic peri- and
postmenopausal women: impact on depres-
sion, vasomotor symptoms, sleep, and quality
of life. Menopause 2006;13(5):780–6.
[110] Bateson AN. The benzodiazepine site of the
GABAA receptor: an old target with new poten-
tial? Sleep Med 2004;5(Suppl 1):S9–15.
[111] Krystal AD, Walsh JK, Laska E, et al. Sustained
efficacy of eszopiclone over 6 months of nightly
treatment: results of a randomized, double-
blind, placebo-controlled study in adults with
chronic insomnia. Sleep 2003;26(7):793–9.
[112] Roth T, Walsh JK, Krystal A, et al. An evaluation
of the efficacy and safety of eszopiclone over
12 months in patients with chronic primary
insomnia. Sleep Med 2005;6(6):487–95.
[113] Dorsey CM, Lee KA, Scharf MB. Effect of zolpi-
dem on sleep in women with perimenopausal
and postmenopausal insomnia: a 4-week, ran-
domized, multicenter, double-blind, placebo-
controlled study. Clin Ther 2004;26(10):1578–86.
[114] Soares CN, Joffe H, Rubens R, et al. Eszopiclone
in patients with insomnia during perimeno-
pause and early postmenopause: a randomized
controlled trial. Obstet Gynecol 2006;108(6):
1402–10.

Seasonal Affective Disorder
and Light Therapy
Brenda Byrne, PhD*, George C. Brainard,
- Definition
- SAD symptoms
- Epidemiology of SAD
- Pathogenesis of SAD
- Physiologic aspects of SAD
- Light therapy
The syndrome of fall and winter depression
known as Seasonal Affective Disorder (SAD) has
had a relatively short history—less than 30 years—
in psychiatric research and practice. Observations
of the effects of environment on human health
were made over 2000 years ago by Hippocrates
[1], and reports of seasonal depression and of the
benefits of sunlight are found through accounts of
Greek and Roman medicine and in occasional
case studies during the past two centuries [2]. In
1980, Lewy and colleagues [3] reported that bright
artificial light, but not ordinary room light, sup-
pressed melatonin secretion in humans. This find-
ing established that the response of humans to
light is comparable to that of other mammals and
was followed by a case report of the effective light
treatment of seasonal manic-depressive disorder
[4]. Light also was found to phase shift circadian
rhythms [5,6], and SAD was early suspected of being
a circadian rhythm disorder. In 1984, Rosenthal and
colleagues [7] offered both diagnostic criteria for
SAD and the first clinical trial of the effectiveness
of bright light therapy for major depression marked
Both authors are supported by the National Space Biomedical Research Institute under NASA Cooperative
Agreement NCC 9-58.
Department of Neurology, Thomas Jefferson University, 1025 Walnut Street, Suite 507, Philadelphia,
PA 19107, USA
* Corresponding author. Margolis Berman Byrne Health Psychology, 1015 Chestnut Street, Suite 901,
Philadelphia, PA 19107.
E-mail address: brenbyr@aol.com (B. Byrne).
1556-407X/08/$ – see front matter ª 2008 Elsevier Inc. All rights reserved.
sleep.theclinics.com
307
SLEEP
MEDICINE
CLINICS
Sleep Med Clin 3 (2008) 307–315
PhD
- Risks and side effects
- Assessment of SAD
- Other strategies of SAD treatment
- Summary
- Acknowledgments
- References
by fall/winter onset and spring/summer remission.
In the intervening years, numerous controlled stud-
ies, case reports, review papers, and meta-analyses
have contributed to a widening base of science
and medical practice regarding the phenomenon
of SAD.
Definition
SAD is a blend of physiologic and mood distur-
bances with a clear seasonal pattern. Diagnostic
criteria in DSM-IV-TR [8] permit a Seasonal Pattern
Specifier to be applied to a pattern of major depres-
sive episodes in bipolar disorders (Bipolar I or
Bipolar II) or in recurrent major depressions. In
the case of winter depressions, a seasonal pattern
of onset in the fall and winter and full remissions
(or a switch to mania or hypomania) in the spring
must be established for the 2 years preceding clini-
cal assessment. Over a patient’s lifetime, seasonal
depressions must outnumber nonseasonal epi-
sodes, and there must be no obvious seasonally
recurring stressors (eg, unemployment or
doi:10.1016/j.jsmc.2008.01.012
308 Byrne & Brainard
anniversaries of a loved one’s death) that might
account for recurrent depression. SAD criteria
from Rosenthal and colleagues [7] are similar, add-
ing the requirement than no other major psychiat-
ric disorder is present.
Seasonality refers to changes in mood and behav-
ior that alter with changes of season. Some forms of
seasonality are mild and common; some, as in SAD,
are more severe.
SAD symptoms
SAD sufferers typically describe a cluster of com-
plaints and symptoms including decreased activity,
sadness, anxiety, social withdrawal, increased appe-
tite (especially for carbohydrates), weight gain, de-
creased libido, and hypersomnia. Depressed SAD
sufferers sometimes report instead the more typical
vegetative symptoms of decreased appetite, weight
loss, and insomnia. SAD sufferers seem to fre-
quently report cold intolerance and to complain
of never feeling warm enough in winter. Women
with SAD often report intensified premenstrual
mood difficulties in the fall and winter.
The most common spontaneous complaints of
SAD sufferers are an overwhelming lack of motiva-
tion or energy and a disinclination to ‘‘be bothered’’
by anyone. SAD sufferers generally have to push
themselves through the fall and winter after the on-
set of symptoms, and—despite presumably suffer-
ing from light deprivation—prefer being indoors,
eating and sleeping. Cold intolerance, when pres-
ent, naturally increases the tendency to stay in-
doors; notions of ‘‘hibernating’’ through the
winter may have strong appeal. Sleep may be longer
and daytime sleepiness greater, but sleep itself is
likely to be of poorer quality [7].
Persons predisposed to SAD commonly become
symptomatic in response to prolonged dark or
cloudy weather in any season. They may report in-
creased symptoms when exposed for prolonged pe-
riods to dark indoor settings. Relief frequently
comes from periods spent in warm bright environ-
ments, with mood and energy lifting within a few
days. In cases of classic SAD, complete remission
of symptoms occurs in the longer daylight and
warmer weather of the springtime.
Seasonal depressive symptoms may be combined
with mood disorders other than major depression,
for instance in a milder subsyndromal form
(termed by researchers ‘‘S-SAD’’). Compared with
symptoms of the full SAD syndrome, symptoms
of S-SAD affect more people but are less disruptive
of mood, activity, or productivity. Light therapy,
which is effective in reversing SAD symptoms, has
been reported to be beneficial as well in treating
S-SAD [2].
In early reports, SAD was presented as a form of
manic-depressive, or bipolar, disorder. Of the first
group of 29 SAD patients treated with bright light
[7], 22 (76%) had been diagnosed with bipolar II
affective disorder; that is, a history of recurrent ma-
jor depressive episodes and of recurrent hypomanic
episodes (elevated mood not meeting criteria for
mania). As evidence from multiple studies has ac-
cumulated, however, the reported proportion of
SAD sufferers with diagnosed bipolar disorders is
between 11% and 50% [9]. Most SAD sufferers al-
ternate between a winter symptom cluster and nor-
mal good mood and energy.
Epidemiology of SAD
Most studies of the prevalence of seasonal symp-
toms have been completed in North America and
have varied widely in method. The Seasonal Pattern
Assessment Questionnaire (SPAQ) [10], the most-
used assessment instrument in such studies, obtains
self-report of how a variety of experiences and be-
haviors change with seasons and with environmen-
tal conditions. A total score, the Global Seasonality
Score (GSS), is derived from indications of seasonal
changes. Specifically, item 12 of the SPAQ asks to
what degree seasonal changes are experienced in
sleep length, social activity, mood (overall feeling
of well-being), weight, appetite, and energy level.
Each is rated on a 5-point scale from ‘‘no change’’
to ‘‘extremely marked change’’ and the item re-
sponses yield a GSS score from 0 to 24. Designed
as a screening tool for use with clinical samples,
the SPAQ does not reliably detect major depression
and so does not differentiate SAD from subsyndro-
mal SAD conditions or clarify how completely win-
ter seasonal symptoms remit in the spring/summer.
Community-based surveys using the SPAQ have
produced estimates of prevalence in North America
as high as 9.7% in northern latitudes. Studies using
more stringent DSM diagnostic criteria have re-
ported lower estimates of prevalence, between
0.8% and 2.2% in North America and between
1.7% and 2.2% in Canada. Prevalence in Europe
has been reported at between 1% and 3% and in
Asia as less than 1% [2]. A positive correlation be-
tween SAD/S-SAD prevalence and latitude has
been reported more for North America than for Eu-
rope. Overall, latitude may be less important than
are other factors such as genetic vulnerability, cli-
mate, and cultural factors [11]. Although a summer
version of seasonal depression has been reported
[12], winter SAD appears much more common
and has been almost exclusively the focus of sea-
sonal mood studies [13].
SAD symptoms are reported in children but are
more common following puberty. Women are

more affected than men of the same age group, espe-
cially during childbearing years, with typical age of
onset in the early to mid-twenties. Rates of SAD de-
cline among elderly populations, and men and
women are more equally affected. Overall, commu-
nity studies have reported the female-to-male ratio
to be about 1.6 to 1 [2]. Although by definition a re-
current disorder, SAD is not necessarily a life-long
disorder. Follow-up studies of SAD subjects reveal
variable courses, with a core group of SAD patients
continuing in a winter depression, summer remis-
sion pattern. A follow-up study of the first 59 pa-
tients of the National Institutes of Health (NIH)
Seasonal Studies Program was performed after
a mean of 8.8 years following their treatment. Re-
sults showed that 42% of this sample continued to
experience symptoms of SAD, and 41% of the sam-
ple continued to use light treatment. Other
patients, 14% of the sample, had fully remitted,
and 44% were experiencing forms of nonseasonal
depression [14]. In two other studies, 18% to 20%
of patients followed up after several years had recov-
ered and others had milder symptoms or recurrent
but nonseasonal major depressions or some other
altered forms of mood disorder [15,16].
Pathogenesis of SAD
Early in the study of SAD diagnosis and treatment,
Rosenthal and colleagues [7] suggested that poten-
tial causal factors deserving study were photoperiod
(number of hours between sunrise and sunset),
hours of daily sunshine, and mean daily tempera-
ture. Light treatment initially was applied in the
morning and evening, based on the ‘‘photoperiod
hypothesis’’ that extending the duration of daylight
would be corrective. Animal models of seasonal
rhythms such as hibernation identify length of the
dark period as the potent cue for behavioral shifts
[17,18], Indeed, careful studies done at the Na-
tional Institute of Mental Health showed that
healthy humans retain photoperiodic responses to
light [19,20].
Because daily hours of sunshine, daily tempera-
ture, and other weather factors are variable and
highly correlated, these have not been promising
explanatory variables for predicting SAD onset or
duration. Photoperiod, however, is predictable,
varying with latitude and day of the year, and is
highly correlated with other environmental factors
[21]. In two studies of SAD symptom onset in the
Chicago area, Young and colleagues [22] found
that photoperiod correlated 0.97 with SAD onset
risk and accounted for 26% of the total variance in
the weekly risk of SAD symptom onset in the fall,
with no other environmental variables accounting
for much additional variance.
Seasonal Affective Disorder and Light Therapy 309
Alternative hypotheses to the photoperiod hy-
pothesis have been proposed, however: a photon-
counting hypothesis based on the total amount of
light received within a day regardless of day length,
a phase-shift hypothesis based on the assumption
of a mismatch between the sleep-wake cycle and
other circadian rhythms, and a melatonin-based
hypothesis based on presumed differences in mela-
tonin levels in affected versus normal subjects. Mel-
atonin has been of great interest based on the
finding that bright light presented within nighttime
darkness suppresses melatonin secretion by the
pineal gland in humans [3,23,24] and that morn-
ing light advances the onset of melatonin secretion
in the evening [6]. Periods of elevated melatonin
secretion are extended during the longer nights of
winter and have been produced in laboratory stud-
ies of the effects of ‘‘summer’’ versus ‘‘winter’’ sched-
ules of environment lighting [19]. Both bright light
and melatonin can be administered in schedules
that will advance or delay circadian rhythms and
have been used for circadian disorders such as jet
lag, shift-work disorders, and advanced or delayed
sleep-wake patterns [25]. Since SAD patients are
more typically phase delayed in their circadian
and sleep-wake patterns, the effects of light in ad-
vancing circadian rhythms is of interest.
Of the alternative explanations offered for SAD,
the phase-shift hypothesis has received the most
support, based on findings in multiple studies
that morning light therapy is superior to evening
light therapy for most subjects [26–29]. Although
some studies have failed to support the superiority
of morning to evening light, no study has found
evening light to be superior. Further, a grouped
analysis of studies failed to find that morning plus
evening light was superior to morning light alone
[30]. The conclusion of a rigorous meta-analysis
of 14 studies of light-box therapy for SAD con-
cluded that bright light was superior to dimmer
light and that morning treatment was superior to
treatment at any other time of day [31].
Morning light treatment has been proposed as
effective because it produces phase advances of
the internal circadian clock, revealed in measure-
ments of the DLMO, or dim light melatonin onset;
that is, the onset of melatonin production mea-
sured in the evening during a dim-light condition
[29,32]. Terman and colleagues [33] report that cir-
cadian rhythm phase advances in melatonin onset
often accompany antidepressant responses and
that the magnitude of the phase advance correlates
with degree of clinical improvement. In other SAD
studies, the degree of phase advance has not been
found to reliably correlate with clinical
improvement [32–35]. Further, since some SAD pa-
tients improve on evening-light treatment, the best
310 Byrne & Brainard
outcomes may result from matching a phase-
shifting treatment to the phase-delayed or advanced
state of the individual patient [25,36].
Despite support for the phase-shift hypothesis,
several studies have reported contrary evidence;
eg, that timing of light treatment may not be crucial
[34,37] (this finding would support the photon-
counting hypothesis) and that circadian profiles
of plasma hormones (cortisol, prolactin, and thyro-
tropin) did not distinguish SAD patients from
healthy controls [38]. Numbers of SAD patients
may not show altered circadian phases and may
respond to light for other reasons or may respond
to alternate treatment approaches.
Exogenous administration of melatonin, which,
when properly timed, can advance or delay circa-
dian rhythms, showed promise in reversing SAD
symptoms in one pilot study [39]. A larger study
found no overall improvement of SAD subjects
treated with melatonin, but did identify improve-
ment in the subjects who were most phase-delayed
and responded to the phase-advancing effect of
melatonin [40].
One effort to account for the variability in types
and degrees of SAD symptoms has been offered in
the ‘‘dual-vulnerability hypothesis’’ [41,42]. This hy-
pothesis assumes that separate factors of seasonality
and depression may occur separately or together.
When only the seasonality factor is expressed, the
diagnosis of S-SAD would apply (energy, sleep, or
appetite/weight changes, for instance). When de-
pression but not seasonality is expressed, a nonsea-
sonal depression would be diagnosed. When
seasonality and depression occur together, SAD is
a likely diagnosis, although some milder forms of
SAD occur in which depression is present but not
severe enough to meet criteria for major depression.
In one study, 552 patients with seasonal complaints
were sorted into SAD, S-SAD, and incomplete sum-
mer remission (ISR) groups and then treated with
light for 2 weeks. Rates of improvement were high-
est in the S-SAD group (78%), intermediate in the
SAD group (66%), and lowest in the ISR group
(51%) [36]. This study provides support for the pro-
vision of light treatment to S-SAD sufferers, since
improvement with light does not depend on sever-
ity of depressive symptoms [43].
Physiologic aspects of SAD
Numerous investigations of neurotransmitter func-
tion in SAD have focused on serotonin (5-HT) and
on the catecholamines dopamine (DA) and nor-
adrenaline (NA), given the importance of these in
studies of major depression. Levels of 5-HT were
found to be lower in winter than in summer in
healthy men, with production of 5-HT increasing
with increased bright sunlight [44], lending support
to a role for 5-HT in the development of the sea-
sonal symptoms of SAD. A technique of reducing
5-HT by depleting tryptophan, a dietary precursor
of 5-HT, has been used in studies of nonseasonal
depression and of antidepressants; this technique
applied to SAD patients in remission after light
therapy led to their relapse [45]. A similar relapsing
effect on SAD patients following successful light
treatment was found in one study both with trypto-
phan depletion and catecholamine depletion [46].
A variety of studies have converged toward the con-
clusion that 5-HT plays a role in the pathophysiol-
ogy of SAD, and light therapy may contribute to
correcting this underlying factor [47]. Evidence
has been less conclusive for the roles of DA and
NA in SAD [2].
Family studies have found that first-degree rela-
tives of SAD patients are more vulnerable to sea-
sonal and nonseasonal depression than would be
expected in the general population. Thirteen per-
cent to 17% of these relatives have been reported
to be affected by SAD and 25% to 67% by nonsea-
sonal affective disorders [48,49]. Lifetime expec-
tancy of developing a unipolar depression in the
general population is estimated to be from 10%
in men to 20% in women [50].
A twin study of seasonality (seasonal changes in
functioning, not necessarily meeting criteria for
SAD) in Australia assessed sleep, social activity,
weight, appetite, and energy level among a cohort
of 4639 adult twins. Genetic effects were estimated
to account for 29% of the variance in seasonality
[51]. A higher degree of heritability of SAD was
found in a twin study in Canada, in which genetics
accounted for 45% to 69% of the total variance
[52].
Genetics may play a role in the association be-
tween SAD and alcoholism. In one family study,
41% of SAD patients reported alcoholism among
their first-degree relatives, with similar elevated
rates found by other researchers [53]. Studies of
molecular genetics of SAD and seasonality have
expanded in recent years, with special interest in
the 5-HT transporter promoter repeat length poly-
morphism (5-HTTLPR) and 5-HT-related genes
[32,49,53].
Light therapy
Bright white light was the original treatment tested
for SAD [4] and numerous studies have shown it to
be effective as a therapeutic intervention for this
condition [2,31]. Three studies published in 1998
provided strong evidence of its superiority to pla-
cebo [26–28]. Light therapy was originally adminis-
tered via large light boxes holding fluorescent light

tubes behind a diffusing screen; these emitted white
light at 2500 lux, a measure of illuminance, and
have generally filtered out ultraviolet light to reduce
health risks to the skin and retina [54,55]. In early
studies, patients were exposed to such lighting for
2 to 6 hours per day, usually in morning and even-
ing doses. Since light must enter the eye to reverse
symptoms [32,56], patients were advised to look
directly at the light for a few seconds every minute
or so. Compliance in study protocols was improved
when it was demonstrated that comparable results
could be obtained with 30-minute exposures to
light measured at 10,000 lux, and by the develop-
ment of units that produced sufficient intensity of
light to be effective when reflected from a desk or
table surface. Clinical response to light therapy is
often reported within 1 week, but full response
may take up to 3 to 4 weeks to develop.
Fluorescent light sources have typically produced
polychromatic white light, with great variability in
the balance of wavelengths [57]. From early years
of SAD study, interest in the potential effects of
varying wavelengths of light across the visible
spectrum has been the focus of a series of studies
[58–60]. To date, the ideal blend of wavelengths
for reversing SAD symptoms is not known, and
broadband white light remains the standard ther-
apy. Recent interest, however, has focused on the
shorter wavelengths in the blue portion of the spec-
trum. Studies of melatonin suppression by light in
healthy humans identified 446 to 477 nm as the
region of the visible spectrum providing the most
potent input for regulating melatonin [23,61]. In-
deed, a series of eight published action spectra
with rodents, monkeys, and humans have con-
firmed that the blue part of the spectrum is most
potent for circadian, neuroendocrine, and neurobe-
havioral responses to light [62]. Further, equal pho-
ton densities of monochromatic blue light at a peak
of 460 nm compared with green light with a peak of
555 nm induced a twofold greater circadian phase
delay and was twice as potent in suppressing mela-
tonin [24]. The development of light-emitting
diodes (LEDs), which can restrict wavelengths emis-
sions to narrow bandwidths, makes possible more
precise studies of the effects of wavelength in
SAD. In one study, narrow-band blue light at
468 nm significantly reduced SAD symptoms com-
pared with a placebo condition of dimmer red light
[63]. Conclusions about the relative efficacy of
various wavelengths in SAD compared with broad-
band white light await further clinical trials.
Light units for use in SAD therapy are now pro-
duced in a variety of sizes and configurations [57].
Fluorescent light boxes have been the most studied
and are still the most widely used clinically, al-
though much smaller units have been produced,
Seasonal Affective Disorder and Light Therapy 311
including head-mounted visors. LED light sources
are now available in addition to fluorescent tubes.
Light therapy has been presented in the form of
‘‘dawn simulation,’’ or light that turns on in one’s
bedroom during morning sleep and gradually in-
creases until wake-up time [64]. A recently com-
pleted 6-year study found that dawn simulation
compared favorably to standard bright-light treat-
ment (30 minutes at 10,000 lux) [29].
Many SAD sufferers obtain information and or-
der treatment devices from the Internet. Web sites
such as those offered by the Society of Light Treat-
ment and Biological Rhythms [65] and the Center
for Environmental Therapeutics [66] provide reli-
able information about the disorder and forms of
treatment.
As most treatment studies have shown morning
light to be more effective than light at other times
of day [31], a starting regimen of 30 minutes at
10,000 lux upon awakening is a common recom-
mendation. The authors of one study [33] argue
for individualizing the timing of morning light treat-
ment to begin no later than 8.5 hours after evening
melatonin onset, or 2.5 hours from the midpoint
of sleep, and they offer the Morningness-Evening-
ness Questionnaire as a tool to estimate the ideal
starting time for light treatment [67,68].
Risks and side effects
A number of photobiological hazards have been
identified in the biomedical literature relative to
overexposure of the skin and eyes to the ultraviolet,
visible, and infrared portions of the electromagnetic
spectrum. Different parts of the spectrum can in-
duce problems such as infrared cataract, photoker-
atitis, photoretinitis, retinal thermal injury,
ultraviolet cataract, and ultraviolet erythema [see
Refs. 69–71 for reviews]. If light therapy equipment
blocks the transmission of ultraviolet radiation to
the eyes, this hazard is not a problem for patients
using the equipment. Most commercial light units
for treatment of SAD or of circadian disorders
generally filter out ultraviolet radiation, which
subtracts from risk potential but not from therapeu-
tic effect [54]. The Blue Light Hazard Function is
potentially relevant to the light exposures from
equipment that emits broad-spectrum white light,
broad-spectrum blue light, or narrow bandwidth
blue light. This function relates to photochemically
induced retinal injury resulting from optical radia-
tion exposure at wavelengths primarily between
400 nm and 500 nm. The action spectrum for this
adverse effect has a peak activity in the 435 to
440 nm wavelength range with short and long
wavelength limbs that drop rapidly with increasing
and decreasing wavelengths [70–74]. Problems
312 Byrne & Brainard
related to the blue light hazard can be avoided if the
manufacturer verifies the safety of their product by
having an independent hazard analysis performed
on the equipment according to national and inter-
national safety guidelines [72–74]. Requesting
ultraviolet radiation and blue light safety informa-
tion from the manufacturer is reasonable when
one purchases or recommends specific light-
therapy equipment.
Side effects to light therapy are infrequent and
generally mild, especially when compared with
the side effects of pharmacologic agents. They gen-
erally subside within the first few days of light treat-
ment or after light exposure is shortened or
intensity of light reduced. The most common com-
plaints reported have been headache, eyestrain,
nausea, irritation, anxiety, or overactivity [1,2,75].
Patients with a history of bipolar disorder are at
increased risk for manic or hypomanic episodes in
response to light therapy, especially if not taking
mood-stabilizing medication.
Assessment of SAD
Basic assessment of seasonality relies on the pa-
tient’s retrospective account of symptoms and their
timing. Although a diagnosis of major depression
combined with a seasonal pattern is required by
the formal DSM criteria for SAD, inquiry should
cover seasonal changes that may be milder or less
definitive. As stated above, subsyndromal SAD (S-
SAD) has been shown to improve with light
treatment, sometimes more fully than will classic
SAD [43,76]. Further, light may also reduce the
severity of winter symptoms in patients who report
combined seasonal and nonseasonal mood
disturbances.
In most but not all cases of SAD, the atypical de-
pressive symptoms of hyperphagia, weight gain,
and hypersomnia will be present, and light therapy
has been predicted to be most effective for patients
reporting these atypical depressive symptoms [77].
Among the standard elements of clinical assess-
ment, inquiry of possible SAD sufferers should in-
clude seasonal patterns over previous years,
seasonality and mood symptoms, whether unipolar
or bipolar, anticipatory anxiety about the fall and
winter, evidence of ‘‘skipped’’ winters or year-round
depression, effect of travel to warm sunny locales,
cold tolerance, ophthalmologic history, current
medications and photosensitizing medications,
and family history of affective disorders and of
alcoholism.
Severity of depression must be assessed, as de-
pression whether seasonal or nonseasonal can pres-
ent risks of suicide. The SPAQ is a brief and useful
screening tool for SAD, which can be supplemented
with the Beck Depression Inventory II and other
self-report measures. Research studies have com-
monly used as an outcome measure responses on
the Structured Interview Guide for the Hamilton
Depression Scale—Seasonal Affective Disorder
Version (SIGH-SAD) [78]. The SIGH-SAD is useful
clinically as well and is available in a self-report
form [79]. Several other assessment instruments
have been developed and are commercially avail-
able [66].
Other strategies of SAD treatment
Most studies of medication management of SAD
have focused on selective serotonin reuptake inhib-
itors (SSRIs), several of which have been shown to
be superior to placebo. Effective doses are similar
to those applied to nonseasonal depression [2,9].
The few comparisons of light treatment to medica-
tions in treating SAD have not resulted in clear
differences in efficacy [9]. For some patients, a com-
bination of light therapy and medication manage-
ment may have additive benefits.
Cognitive behavioral therapy (CBT) has been
tested in SAD treatment alone and in combination
with light treatment. In a randomized controlled
trial, CBT was administered to groups of SAD
subjects in twice-weekly 1.5-hour sessions over
6 weeks. Three other randomly assigned groups
received 6 weeks of light treatment, combined
CBT and light, or delayed treatment with minimal
contact [80]. Results of this trial were similar to
an earlier feasibility study showing that combina-
tion CBT and light treatment resulted in a remission
rate of 73%, significantly higher than control group
remission (20%). The light-treatment condition
resulted in a 50% remission rate, similar to results
of many other studies of light treatment for SAD.
A promising treatment direction for SAD suf-
ferers was discovered when negative ion generators
were used as putative placebos in light-therapy
studies. The degree of response noted to negative
ions led to studies of the treatment potential of neg-
ative ions for SAD. Treatment with high-density
negative ions has been superior to treatment with
low-density negative ions, and high-density nega-
tive ion treatment has resulted in improvement in
50% of cases [28]. A study of high-density negative
air ionization presented to subjects while asleep
resulted in therapeutic benefits that were not signif-
icantly different from those of standard post-
awakening bright-light treatment provided to
subjects in the same study. Low-density negative
ions provided little benefit [29]. No side effects to
treatment with air ionization for SAD have been
identified, which may recommend it as an alterna-
tive treatment or a companion treatment to light.

It remains unknown how air ions improve SAD
symptoms.
Summary
Light is universally acknowledged as a basic ele-
ment of life as we know it, from the ‘‘dawn of crea-
tion,’’ and the field of contemporary research is
shedding light on light’s healing power. As applied
to SAD, a proliferation of studies of this disorder, of
its response to light, and of its psychologic and
psychophysiological correlates leads to a view of
SAD as multidimensional, manifested in individual
patterns that have both common elements and also
highly divergent aspects. The importance of light as
a corrective treatment with results superior to pla-
cebo has been established, and standards for the
safe and effective practice of light therapy have
been propounded [2,29,70,72–74]. Meanwhile,
research continues on strategies for optimizing the
effectiveness of light timing, intensity and spectrum
for SAD and related circadian disorders.
Acknowledgments
The authors gratefully acknowledge John Hanifin,
Claudia Penrose, and Mike Jablonski for their excel-
lent assistance with referencing and seeking copy-
right permissions. In addition, we appreciate Kat
West for her thorough editorial comments on the
final draft.
References
[1] Rosenthal NE. Winter blues: seasonal affective
disorder: what it is and how to overcome it.
New York: The Guilford Press; 1993.
[2] Lam RW, Levitt AJ, editors. Canadian consensus
guidelines for the treatment of seasonal affective
disorder. Vancouver (BC): Clinical and Academic
Publishing; 1999. p. 13–6.
[3] Lewy AJ, Wehr TA, Goodwin FK, et al. Light
suppresses melatonin secretion in humans.
Science 1980;210(4475):1267–9.
[4] Lewy AJ, Kern HE, Rosenthal NE, et al. Bright
artificial light treatment of a manic-depressive
patient with a seasonal mood cycle. Am J Psychi-
atry 1982;139(11):1496–8.
[5] Czeisler CA, Allan JS, Strogatz SH, et al. Bright
light resets the human circadian pacemaker inde-
pendent of the timing of the sleep-wake cycle.
Science 1986;233(4764):667–71.
[6] Lewy AJ, Sack RL, Miller LS, et al. Antidepressant
and circadian phase-shifting effects of light.
Science 1987;235(4786):352–4.
[7] Rosenthal NE, Sack DA, Gillin JC, et al. Seasonal
affective disorder. A description of the syndrome
and preliminary findings with light therapy. Arch
Gen Psychiatry 1984;41(1):72–80.
Seasonal Affective Disorder and Light Therapy 313
[8] American Psychiatric Association. Diagnostic
and statistical manual of mental disorders. 4th
edition. Washington, DC: American Psychiatric
Association; 2000.
[9] Sohn CH, Lam RW. Treatment of seasonal affec-
tive disorder: unipolar versus bipolar differences.
Curr Psychiatry Rep 2004;6(6):478–85.
[10] Rosenthal NE, Bradt GH, Wehr TA. Seasonal pat-
tern assessment questionnaire. Bethesda (MD):
National Institute of Mental Health; 1987.
[11] Mersch PP, Middendorp HM, Bouhuys AL, et al.
Seasonal affective disorder and latitude: a review
of the literature. J Affect Disord 1999;53(1):
35–48.
[12] Wehr TA, Giesen HA, Schulz PM, et al. Contrasts
between symptoms of summer depression and
winter depression. J Affect Disord 1991;23(4):
173–83.
[13] Magnusson A. An overview of epidemiological
studies on seasonal affective disorder. Acta Psy-
chiatr Scand 2000;101(3):176–84.
[14] Schwartz PJ, Brown C, Wehr TA, et al. Winter
seasonal affective disorder: a follow-up study of
the first 59 patients of the National Institute of
Mental Health Seasonal Studies Program. Am J
Psychiatry 1996;153(8):1028–36.
[15] Thompson C, Raheja SK, King EA. A follow-up
study of seasonal affective disorder. Br J Psychia-
try 1995;167(3):380–4.
[16] Graw P, Gisin B, Wirz-Justice A. Follow-up study
of seasonal affective disorder in Switzerland. Psy-
chopathology 1997;30(4):208–14.
[17] Reiter RJ. Pineal gland: interface between the
photoperiodic environment and the endocrine
system. Trends Endocrinol Metab 1991;1:13–9.
[18] Arendt J. Melatonin and the pineal gland: influ-
ence on mammalian seasonal and circadian
physiology. Rev Reprod 1998;3:13–22.
[19] Wehr TA. The durations of human melatonin
secretion and sleep respond to changes in day-
length (photoperiod). J Clin Endocrinol Metab
1991;73(6):1276–80.
[20] Wehr TA, Moul DE, Barbato G, et al. Conserva-
tion of photoperiod-responsive mechanisms in
humans. Am J Physiol 1993;265(4 Pt 2):
R846–57.
[21] Potkin SG, Zetin M, Stamenkovic V, et al. Sea-
sonal affective disorder: prevalence varies with
latitude and climate. Clin Neuropharmacol
1986;9(Suppl 4):181–3.
[22] Young MA, Meaden PM, Fogg LF, et al. Which
environmental variables are related to the onset
of seasonal affective disorder? J Abnorm Psychol
1997;106(4):554–62.
[23] Brainard GC, Hanifin JP, Greeson JM, et al.
Action spectrum for melatonin regulation in
humans: evidence for a novel circadian photore-
ceptor. J Neurosci 2001;21(16):6405–12.
[24] Lockley SW, Brainard GC, Czeisler CA. High sen-
sitivity of the human circadian melatonin
rhythm to resetting by short wavelength light.
J Clin Endocrinol Metab 2003;88(9):4502–5.
314 Byrne & Brainard
[25] Lewy AJ, Lefler BJ, Emens JS, et al. The circadian
basis of winter depression. Proc Natl Acad Sci
U S A 2006;103(19):7414–9.
[26] Eastman CI, Young MA, Fogg LF, et al. Bright
light treatment of winter depression: a placebo-
controlled trial. Arch Gen Psychiatry 1998;
55(10):883–9.
[27] Lewy AJ, Bauer VK, Cutler NL, et al. Morning vs.
evening light treatment of patients with winter de-
pression. Arch Gen Psychiatry 1998;55(10):890–6.
[28] Terman M, Terman JS, Ross DC. A controlled
trial of timed bright light and negative air ioniza-
tion for treatment of winter depression. Arch
Gen Psychiatry 1998;55(10):875–82.
[29] Terman M, Terman J. Light therapy for seasonal
and nonseasonal depression: efficacy, protocol,
safety, and side effects. CNS Spectr 2005;10(8):
647–63.
[30] Terman M, Terman JS, Quitkin FM, et al. Light
therapy for seasonal affective disorder. A review
of efficacy. Neuropsychopharmacology 1989;
2(1):1–22.
[31] Thompson C, Rodin I, Birtwhistle J. Light ther-
apy for seasonal and nonseasonal affective
disorder: a Cochrane meta-analysis. Society
of Light Treatment and Biological Rhythms
Abstracts 1999;11:11.
[32] Sohn CH, Lam RW. Update on the biology of
seasonal affective disorder. CNS Spectr 2005;
10(8):635–46.
[33] Terman JS, Terman M, Lo ES, et al. Circadian
time of morning light administration and thera-
peutic response in winter depression. Arch Gen
Psychiatry 2001;58(1):69–75.
[34] Wirz-Justice A, Graw P, Krauchi K, et al. Light
therapy in seasonal affective disorder is indepen-
dent of time of day or circadian phase. Arch Gen
Psychiatry 1993;50(12):929–37.
[35] Wirz-Justice A, Graw P, Krauchi K, et al. ‘Natural’
light treatment of seasonal affective disorder.
J Affect Disord 1996;27(2–3):109–20.
[36] Doghramji K, Gaddy JR, Stewart KT, et al. Two-
versus four-hour evening phototherapy of
seasonal affective disorder. J Nerv Ment Dis
1990;178(4):257–60.
[37] Wehr TA, Jacobsen FM, Sack DA, et al. Photo-
therapy of seasonal affective disorder: time of
day and suppression of melatonin are not critical
for antidepressant effects. Arch Gen Psychiatry
1986;43(9):870–5.
[38] Oren DA, Levendosky AA, Kasper S, et al. Circa-
dian profiles of cortisol, prolactin, and thyrotro-
pin in seasonal affective disorder. Biol Psychiatry
1996;39(3):157–70.
[39] Lewy AJ, Bauer VK, Cutler NL, et al. Melatonin
treatment of winter depression: a pilot study.
Psychiatry Res 1998;77(1):57–61.
[40] Lewy AJ, Lefler BJ, Hasler BP, et al. Plasma
DLMO10 Zeitgeber time 14: the therapeutic
window for phase-delayed winter depressives
treated with melatonin [abstract]. Chronobiol
Int 2003;20:1215–6.
[41] Young MA, Watel LG, Lahmeyer HW, et al. The
temporal onset of individual symptoms in win-
ter depression: differentiating underlying mecha-
nisms. J Affect Disord 1991;22(4):191–7.
[42] Lam RW, Tam EM, Yatham LN, et al. Seasonal
depression: the dual vulnerability hypothesis
revisited. J Affect Disord 2001;63(1–3):123–32.
[43] Kasper S, Rogers SL, Yancey A, et al. Photother-
apy in individuals with and without subsyndro-
mal seasonal affective disorder. Arch Gen
Psychiatry 1989;46(9):837–44.
[44] Lambert GW, Reid C, Kaye DM, et al. Effect of
sunlight and season on serotonin turnover in
the brain. Lancet 2002;360(9348):1840–2.
[45] Lam RW, Zis AP, Grewal A, et al. Effects of rapid
tryptophan depletion in patients with seasonal
affective disorder in remission after light therapy.
Arch Gen Psychiatry 1996;53(1):41–4.
[46] Neumeister A, Turner EH, Matthews JR, et al.
Effects of tryptophan depletion vs. catechol-
amine depletion in patients with seasonal
affective disorder in remission with light therapy.
Arch Gen Psychiatry 1998;55(6):524–30.
[47] Neumeister A, Konstantinidis A, Praschak-
Rieder N, et al. Monoaminergic function in the
pathogenesis of seasonal affective disorder. Int J
Neuropsychopharmacol 2001;4(4):409–20.
[48] Allen JM, Lam RW, Remick RA, et al. Depressive
symptoms and family history in seasonal and
nonseasonal mood disorders. Am J Psychiatry
1993;150(3):443–8.
[49] Sher L. Genetic studies of seasonal affective
disorder and seasonality. Compr Psychiatry
2001;42(2):105–10.
[50] Kaplan HI, Sadock BJ. Mood disorders. In: Syn-
opsis of psychiatry. 5th edition. Baltimore
(MD): Williams & Wilkins; 1988. p. 288–309.
[51] Madden PA, Heath AC, Rosenthal NE, et al. Seao-
nal changes in mood and behavior. The role of
genetic factors. Arch Gen Psychiatry 1996;
53(1):47–55.
[52] Jang KL, Lam RW, Livesley WJ, et al. Gender
differences in the heritability of seasonal mood
change. Psychiatry Res 1997;70(3):145–54.
[53] Sher L. Alcoholism and seasonal affective disor-
der. Compr Psychiatry 2004;45(1):51–6.
[54] Lam RW, Buchanan A, Clark CM, et al. Ultravio-
let versus non-ultraviolet light therapy for
seasonal affective disorder. J Clin Psychiatry
1991;52(5):213–6.
[55] Waxler M, James RH, Brainard GC, et al. Reti-
nopathy and bright light therapy [letter]. Am J
Psychiatry 1992;149(11):1610–1.
[56] Wehr TA, Skwerer RG, Jacobsen FM, et al. Eye
versus skin phototherapy of seasonal affective
disorder. Am J Psychiatry 1987;144:753–7.
[57] Brainard GC. The healing light: interface of phys-
ics and biology. In: Lam RW, editor. Seasonal
affective disorder and beyond: light treatment
for SAD and non-SAD conditions. Washington,
DC: American Psychiatric Press, Inc.; 1998. p.
1–44.

[58] Brainard GC, Rosenthal NE, Sherry D, et al. Ef-
fects of different wavelengths in seasonal affec-
tive disorder. J Affect Disord 1990;20(4):209–16.
[59] Oren DA, Brainard GC, Joseph-Vanderpool JR,
et al. Treatment of seasonal affective disorder
with green light versus red light. Am J Psychiatry
1991;148(4):509–11.
[60] Stewart KT, Gaddy JR, Byrne B, et al. Effects of
green or white light for treatment of seasonal
depression. Psychiatry Res 1991;38(3):261–70.
[61] Thapan K, Arendt J, Skene DJ. An action spec-
trum for melatonin suppression: evidence for
a novel non-rod, non-cone photoreceptor
system in humans. J Physiol 2001;535(1):261–7.
[62] Brainard GC, Hanifin JP. Photons, clocks and
consciousness. J Biol Rhythms 2005;20(4):
314–25.
[63] Glickman G, Byrne B, Pineda C, et al. Light ther-
apy for seasonal affective disorder with blue
narrow-band light-emitting diodes (LED). Biol
Psychiatry 2006;59(6):502–7.
[64] Avery DH, Bolte MA, Dager SR, et al. Dawn simu-
lation treatment of winter depression: a controlled
study. Am J Psychiatry 1993;150(1):113–7.
[65] Society for Light Treatment and Biological
Rhythms. Available at: www.sltbr.org. Accessed
March 18, 2008.
[66] Center for Environmental Therapeutics. Avail-
able at: www.cet.org. Accessed March 18, 2008.
[67] Terman M, Terman JS. Morningness-
eveningness, circadian phase and the timing of
sleep in patients with seasonal affective disorder.
Society of Light Treatment and Biological
Rhythms Abstracts 2001;13.
[68] Terman M, White TM, Jacob J. Automated
morningness-eveningness questionnaire. New
York: State Psychiatric Institute; 2002. Available
at: www.cet.org. Accessed March 18, 2008.
[69] Sliney D, Wolbarsht M. Safety with lasers and
other optical sources. New York: Plenum Press;
1980. p. 1–1035.
[70] American National Standards Institute and Illu-
minating Engineering Society of North America.
Recommended practice for photobiological
safety for lamps and lamp systems—general
Seasonal Affective Disorder and Light Therapy 315
requirements RP-27.1. New York: Illuminating
Engineering Society of North America; 2005.
[71] Matthes R, Sliney D, Didomenico S, et al, editors.
Measurements of optical radiation hazards.
Munchen (Germany): ICNIRP; 1998. p. 1–762.
[72] International Commission on Non-Ionizing
Radiation Protection. Guidelines on limits of
exposure to broad-band incoherent optical
radiation (0.38 to 3 microM). Health Phys
1997;73:539–54.
[73] Commission Internationale de l’Eclairage. Pho-
tobiological safety of lamps and lamp systems,
CIE Publication No. S 009/E:2002; Vienna:
Austria; 2002.
[74] American Conference of Governmental Indus-
trial Hygienists. Documentation of the threshold
limit values and biological exposure indices.
Cincinnati (OH): American Conference of
Governmental Industrial Hygienists; 2006.
[75] Kogan AO, Guilford PM. Side effects of short-
term 10,000-lux light therapy. Am J Psychiatry
1998;155(2):293–4.
[76] Levitt AJ, Lam RW, Levitan R. A comparison of
open treatment of seasonal major and minor
depression with light therapy. J Affect Disord
2002;71(1–3):243–8.
[77] Krauchi K, Wirz-Justice A, Graw P. High intake of
sweets late in the day predicts a rapid and persis-
tent response to light therapy in winter depres-
sion. Psychiatry Res 1993;46(2):107–17.
[78] Williams JBW, Link MJ, Rosenthal NE, et al.
Structured interview guide for the Hamilton
depression rating scale, seasonal affective
disorder version (SIGH-SAD). New York: State
Psychiatric Institute; 2002.
[79] Williams JBW, Link MJ, Terman M. Structured
interview guide for the Hamilton depression
rating scale, seasonal affective disorder ver-
sion—self-rating version (SIGH-SAD-SR). New
York: State Psychiatric Institute; 1998.
[80] Rohan KJ, Roecklein KA, Tierney Lindsey K, et al.
A randomized controlled trial of cognitive-
behavioral therapy, light therapy, and their
combination for seasonal affective disorder.
J Consult Clin Psychol 2007;75(3):489–500.