JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No.

1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

Correlation Between Mucosal Inflammation and

Marginal Bone Loss Around Hydroxyapatite-Coated
Implants: A -Year Cross-Sectional Study
Eduardo Rolim Teixeira , DDS/Yuuji Sato, DDS, PhD/Yasumasa Akagawa, DDS,
PhD/Tomohide Kimoto , DDS

In this study, a possible correlation between peri-implant mucosal inflammation and marginal
bone loss in a group of patients who received two-stage hydroxyapatite-coated dental implants
was evaluated years after implantation. Thirty-two implants were evaluated as to their Plaque
Index, Gingival Index, and crevicular fluid volume. Patients were divided into three groups: a
control group with no inflammatory signs; a borderline group with moderate inflammation;
and a gingivitis group with an established inflammatory process. Three dental radiographs per
implant were analyzed at months, months, and years after implantation by measuring bone
height at the mesial and distal sides. Preliminary results indicated increased bone resorption
in the gingivitis group compared with the control and borderline groups at both the mesial and
distal surfaces (P < .05), suggesting correlation between mucosal inflammation and marginal
resorption observed around these implants. Statistical analysis of correlation between the
assessed mucosal parameters and bone loss indicated significance for Gingival Index and
crevicular fluid volume (P < .05).
(INT J ORAL MAXILLOFAC IMPLANTS 1997;12:74–81)
Key words: dental implant, hydroxyapatite, marginal bone loss, peri-implant mucosal inflammation

T he success of osseointegration in implant dentistry appears to be linked to the cellular response

at the implant interface,1,2 which is expected to be an inflammation-free zone demonstrating the
direct apposition of a calcified matrix or soft tissue cells to the implant surface. The presence of an
inflammatory process at this zone could induce an inappropriate environment for connective tissue
cells and thus interfere with collagen fiber apposition and the mineralization process at the implant
surface during osseointegration.

It has been recognized through human3 and animal trials4 that plaque accumulation leads to mucosal
inflammation around osseointegrated implants. Also, it has been shown that the histologic arrangement
of peri-implant soft tissues resembles the one observed in natural teeth.5,6 Based on these facts, it could
be assumed that an inflammatory process starting at the peri-implant mucosa might evolve into a
destructive reaction, leading to bone resorption in the peri-implant zone, and thus endangering the
achieved integration between bone and implant. However, a review of the dental literature revealed no
consensus as to whether such an inflammatory process would actually constitute a cause leading to loss
of bone anchorage around osseointegrated dental implants.

In support of a correlation concept, Kirsch and Mentag7 reported in a multicenter clinical evaluation
that among those verified cases of IMZ implant failure, 75% were directly related to poor oral hygiene
and subsequent plaque accumulation on the abutment surface. Also, the importance of careful patient

Article Text 1
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

selection in implant dentistry was emphasized, suggesting that those patients with lack of motivation
and/or ability to perform oral hygiene procedures should not receive implants. Block et al8 found a
marked correlation between marginal bone loss around hydroxyapatite-coated (HA-coated) implants and
poor oral hygiene, especially in the posterior mandible area. They associated 18% and 9% of the verified
failures with mucosal inflammation involving implants placed into the mandible and maxilla,
respectively9. Furthermore, in a prospective multicenter study using the Brånemark System (Nobel
Biocare AB, Göteborg, Sweden) Henry and coworkers10 reported that implant failures were
concentrated in patients with a higher degree of plaque accumulation, also pointing out that although the
pathogenesis of bacterial plaque in implant dentistry is still uncertain, it should be controlled.

In contrast to these findings, Apse et al11 stated in a review of patients who received 238 implants in
the Toronto study that no evidence was found linking plaque accumulation and loss of osseointegration
during a 4- to 9-year follow-up evaluation. It was noted that the role of peri-implant mucosal response in
the clinical longevity of dental implants is still unclear. Lekholm and coworkers 12 also found results
similar to those of the Toronto study, stating that verified deep pockets and gingivitis were not correlated
to marginal resorption in 125 observed implants. Smith and Zarb13 proposed that the peri-implant
mucosal response should not be included in the criteria for implant success because it has not been shown
to be an important factor in achieving or maintaining osseointegration.
In the present cross-sectional study, the preliminary results of an investigation specifically designed
to examine a possible correlation between peri-implant soft tissue inflammation and marginal bone loss
around HA-coated dental implants are reported. The term gingivitis is used here specifically to
characterize the assessed inflammatory state of tissues related to a group of implants, and not to generally
describe peri-implant mucosal inflammation, since gingiva has been recognized as a descriptive term for
the periodontal mucosa, presenting a characteristic distribution of collagen fibers attached to the tooth
surface. Studies focused on the anatomy of peri-implant mucosa have not demonstrated the presence of
such an arrangement. Therefore, use of the term gingiva referring to peri-implant mucosa is not
warranted.
Materials and Methods
From September 1988 to April 1992, a total of 110 two-stage HA-coated titanium alloy (Ti-6Al-4V)
dental implants (Kyocera, Kyoto, Japan) were placed in 42 patients (17 men, 25 women). The applied
system presents two basic components, implant and post, connected by two types of lock-screws,
indicated according to the superstructure characteristics. The patients, completely or partially edentulous,
were subjected to clinical and laboratory examinations and to a roentgenologic and computerized
tomographic (CT) analysis to determine bone quality and quantity at the expected implantation sites
(maxilla, n = 25 implants; mandible, n = 85 implants). Prostheses (n = 39) connecting neighboring
implants, single crowns (n = 7), and overdentures (n = 8) were designed and fabricated according to the
requirements and aspirations of each patient. All fixed superstructures were made using gold-palladium
alloy, and hard acrylic resin was applied to the occlusal table to minimize the impact of masticatory
forces. A success rate of 90.9% was verified using a clinical and radiographic assessment after a 1-year
postplacement period.14
Group Selection and Evaluation Protocol. During the first semester of , a routine recall evaluation
of 31 patients (n = 84 implants) was performed in the clinic of the Department of Removable
Prosthodontics, Hiroshima University Dental Hos pital. Peri-implant mucosal condition involving the

Article Text 2
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

implants was investigated through Plaque Index15 (PI) and Gingival Index16 (GI) recordings, as well as
through crevicular fluid volume (Periotron score [PS]) measured with a Periotron device (Harco Elec
tronics, Winnipeg, Manitoba, Canada). To verify the marginal bone level and general status of
osseointegration, periapical radiographs of the implanted area were taken at the end of the examination.
The implants assessed in this study were selected based on their past treatment homogeneity. At first,
a subject profile was conceived based on the most usual characteristics found among the implanted
group. The analysis of these characteristics suggested that a woman subject with a free-standing fixed
prosthesis supported by implants in the posterior man dible area could be the typical subject. This applied
selection process based on matching features such as sex, implantation site, and type of superstructure
was chosen to eliminate bias originated by different clinical characteristics among the studied groups.
Also, to evaluate the selected implants regarding success and to determine their inclusion in this study,
the following criteria were applied:
1. Absence of any complaints such as pain, dysesthesia, or paresthesia at the implant site
2. Absence of recurring peri-implant infection and/or suppuration
3. Implants presenting no perceptible mobility
4. Implants presenting no radiolucencies at the implant-bone junction
The degree of inflammation involving the peri-implant mucosa was evaluated through the registering
of GI and PI scores at mesial and distal sites, and also through the PS, which was determined by the
mean value of two measurements made at the buccal side of each implant using the two-strip
technique.17 During the data examination, the implant was chosen as the unit of analysis, sometimes
causing the inclusion of the same patient in two different groups according to the peri-implant mucosal
indexes of each implant. To allow comparison of the implant surfaces clinically and radiographically,
only the mesial and distal sides of all implants were observed.
Thirty-two implants were divided into three groups according to their peri-implant mucosal indexes,
as follows:
1. Control group (CG): PS < 5 and GI d 1 at mesial or distal surfaces
2. Borderline group (BG): 5 d PS < 16 and GI = 1 or 2 at mesial or distal surfaces
3. Gingivitis group (GG): PS t 16 and GI t 2 at mesial or distal surfaces
Table 1 shows characteristics regarding number of patients and implants in CG, BG, and GG, and
mean values of evaluated peri-implant mucosal parameters. Mean values regarding age, time of
implantation, and loading period for each examined group are shown in Table 2. A mean of 3.5 years for
implantation period and years for loading period was verified, depicting similar values between the
studied groups.
Radiographic Analysis. Three radiographs per implant were obtained and analyzed with regard to
the marginal bone level at three different time points (Figs 1 and 2). The mean time, verified after the
first surgical procedure, at which the first, second, and recall radiographs were taken, was 2.3, 8.9, and
25.6 months, respectively. The first radiograph revealed the implant position related to the crest of the
alveolar ridge after implantation. From this film, measurements in millimeters of the amount of bone
above or beneath the edge of the implants were made. Subsequently, the second and third radiographs

Article Text 3
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

were measured to verify the amount of marginal bone loss for each period, always measuring from the
same landmark at the edge of the implant. All measurements were performed with a magnifying ocular
lens with a 0.1-mm scale to obtain a precise measurement.
To assess the quality of the observed radiographs, the degree of distortion and angulation was
investigated by measuring, in all examined radiographs, the length of the implant from the landmark at
the edge to its lowest portion. Later, these obtained values were cross-checked with the actual length of
the respective implant. Two distinct groups were formed, one group composed by values of measured
lengths and another group of values determined by the actual implant length. The difference between
these two groups was determined by statistical analysis.
Statistical Analysis. The differences between mesial and distal bone loss values within a group and
also between different groups were calculated by applying a nonparametric method (Mann-Whitney U
test). Parametric tests were not chosen in this study to avoid the occurrence of type 1 (false positive)
statistical error because PI and GI are considered to be nonparametric. In addition, PI, GI, and PS were
separated into groups according to their observed scores: PI = 0 and 1 (two groups); GI = 0 or 1 and 2
(two groups); and PS < 5, 5 d PS < 16, and PS t 16 (three groups). Later, the differences between each
group, in relation to mean values of bone loss, were analyzed statistically by the Mann-Whitney U test.
Results
Bone Loss Tendency. A decline in marginal bone height from the first to the second radiograph was
observed in all groups, illustrating that bone remodeling began around the implants just after
implantation. However, a comparison between changes in bone height observed in CG, BG, and GG after
a mean period of 8.9 months postimplantation denoted differences in resorption patterns between these
groups. At this point, higher values indicating marginal bone loss were verified in GG, compared to CG
and BG. From the second to the third radiograph, greater differences in bone loss were verified in GG
compared to CG and BG mean values during a mean period of 26.7 months after the second
measurement.
Bone Height Measurements. Table 3 depicts mean and standard deviations (SD) of all
measurements for the mesial and distal sides of each analyzed group. Positive values appearing in the
first radiograph after first-stage surgery in all implanted groups were the result of deep placement of the
im plants. Also, similar mean values were seen be tween mesial and distal surfaces. The second
radiograph was taken at the superstructure placement appointment, and it showed the resorption pattern
after a postimplantation mean period of 8.9 months. At this time, negative values were observed in
almost all groups, indicating that the bone found above the implant edge at the time of the first
radiograph was lost, regardless of the presence of a superstructure that would promote greater loading of
the implants. Nevertheless, at this point the greatest amount of bone loss was observed in GG. The third
radiograph, taken during a recall evaluation of all implants, did show negative values for both analyzed
surfaces in all examined groups.
The column in Table 3 on bone loss shows the total amount of bone lost between the first and third
radiographs. It confirms considerable bone loss of GG compared to BG and CG data for both mesial and
distal surfaces. No statistically significant differences were found between mesial and distal sides. The
recorded marginal bone loss was greater in both surfaces of GG compared to other groups (P < .05). No
statistically significant differences were observed between CG and BG (P > .05). Also, analysis of the
distortion degree and angulation on those observed radiographs calculated through implant length
measurements presented no statistically significant differences when compared to the real length of the

Article Text 4
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

implant (P > .05).

Relationship of PI, GI, and PS to Bone Loss. Figure 3 reveals the correlation between bone loss
found at the distal surfaces of the examined implants and their respective PI, GI, and PS values, gathered
in groups according to their respective scores. Statistical analysis showed a significant correlation
between groups of implants presenting different GI and PS values (P < .05), indicating their significance
among the evaluated factors and justifying their application in the group formation criteria. No
statistically significant differences were found between groups presenting different PI scores in relation
to bone loss.
Discussion
In the present study, the possible effects of peri-implant mucosal inflammation on marginal bone loss in a
group of patients who received newly developed two-stage HA-coated dental implants were studied. A
review of the dental literature disclosed divergent opinions regarding the role of peri-implant
inflammation in the achievement or maintenance of osseointegration.7-13 The review disclosed a variety
of dental implant systems composed of different biomaterials that were evaluated concerning marginal
bone loss and mucosal inflammation. These studies used different assessment parameters that made any
comparison or correlation between them impractical. Also, the applicability of periodontal indexes used
in these studies to evaluate mucosal inflammation around dental implants was questioned.18 Never
theless, supporting data corroborating the concept application are available.3
The present study provides the preliminary results of an investigation designed to follow the
progression of bone loss around HA-coated implants. The re search design for this evaluation considered
the implant as the unit of analysis, based on our belief that peri-implant mucosal response can vary
significantly even between implants placed in the same jaw. Therefore, the selection of the patient as the
unit of analysis could be erroneous. The 3-year results suggested correlation between mucosal
inflammation of the peri-implant and marginal bone loss. Calculated mean values and resorption
tendencies denoted different standards between groups with and without mucosal inflammation.
However, because of a relatively short observation period and the unpredictability of future
measurements of peri-implant indexes and marginal bone height, the use of conclusive affirmation in
outcome interpretation is probably not appropriate.

It has been said that there are basically five possible causes of bone loss around dental implants.18
They are (1) effects of surgical trauma during implant placement, (2) inadvertent stress distribution by de-
fective prostheses, (3) unfavorable maxillomandibular relation or parafunctional habits, (4) physiologic
resorption caused by edentulousness, and (5) peri-implant mucosal inflammation. The three groups
analyzed in the present study were formed based on their past treatment homogeneity with the intent of
excluding bias originated by different clinical characteristics and maintaining only peri-implant mucosal
inflammation as the possible cause of bone loss. The patient groups were operated in the same hospital
by the same staff of surgeons, stressing the importance of all procedures recommended in a surgical
protocol.18 Sterilized sharp instruments and drills, gentle preparation of the bony socket, low rotational
speed (ie, no more than 800 rpm), and constant irrigation with a saline solution were used to avoid any
damage to the adjacent bone caused by stress or overheating.
Preoperative diagnostic procedures included, among other examinations, the careful investigation of
any parafunctional activities or malocclusion that, if present, would contraindicate implant treatment. All
fabricated fixed prostheses were designed following the same protocol, presenting hygienic contours,

Article Text 5
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

narrow occlusal tables made in acrylic resin, low cusp height, and shallow occlusal sulci to avoid any
potentially damaging superstructure design for the marginal bone. Moreover, all occlusal adjustments
were performed uniformly, applying a one-point contact per element with no guidance pattern, reducing
stresses on the implants that might be caused by the incidence of nonparallel forces during lateral
mandibular movements. Natural or physiologic resorption of the edentulous jaw was addressed through
the group selection process in which an attempt was made to establish the same clinical characteristics
that could influence resorptive patterns in all groups.

The findings of the present study are in accordance with the results of Block et al8 and Block and
Kent,9 which suggested correlation between mucosal inflammation and marginal bone loss among the
examined dental implants. Also, analysis of the correlation between the assessed mucosal parameters and
bone loss presented statistically significant differences between groups having distinct Gingival Index
and crevicular fluid volume values, confirming their significance in association with bone loss in this
study. This calculated correlation confirmed the in clusion of GI and PS in the criteria for group
formation and excluded PI because of the verified poor correlation results. This result presumably was ob-
tained because of the fact that the examined patients were aware that an evaluation of the implants would
be performed, and they may have intensified oral hygiene procedures just before the recall appointment.

An important similarity between our evaluation and two aforementioned studies8,9 was that all
analyses were performed on HA-coated implants, establishing similar rough surfaces on which plaque
could adhere, proliferate, and possibly result in the same pathogenesis. Also, the rough characteristic of
the HA surface prevents an effective cleaning or scaling procedure, and it creates an environment that
could enhance plaque attachment and growth. In those investigations showing divergent results, the
chosen implant systems employed different designs and biomaterials, raising the hypothesis that implant
characteristics such as chemical composition, surface energetic properties, morphology, and texture may
influence the capacity of the dental plaque to attach and spread on its surfaces. However, scientific
evidence in support of these postulates is not available.
The radiographs examined and measured in this study were taken using a freehand technique,
allowing the occurrence of distortions and different horizontal and vertical angulations. Nevertheless, the
difference found between the measured implant length compared with the actual size of the implant was
not statistically significant (P < .05). This finding suggests that there was no substantial image distortion
among the radiographs examined, although one should still consider the differences in horizontal
angulation that could not be assessed.
Another important consideration is whether peri-implant inflammation definitely constitutes the
cause of the bone loss observed around the examined implants. The fact that different values of bone
height were verified between the groups, at the time of the second radiograph, suggested the influence of
mucosal inflammation on bone remodeling during the initial postimplantation period, because no
superstructure was present to promote a specific loading on the implants. From the time interval between
the second and third radiograph, this aforementioned influence was present, as evidenced by the
differences among group measurements. However, the use of conclusive statements might be erroneous
because of the fact that other elements still not considered as being related to bone loss, or the natural
resorption of the edentulous jaw, may modify the observed results.
Summary
Preliminary results indicated increased bone resorption in the gingivitis group (established inflammatory

Article Text 6
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

process) compared with the control group (no inflammatory signs) and the borderline group (moderate
inflammation) at both the mesial and distal surfaces (P < .05), suggesting correlation between mucosal
inflammation and marginal resorption observed around these implants.

Tomohide Kimoto

Graduate Student, Department of Removable

Prosthodontics, Hiroshima University, School of
Dentistry, Hiroshima, Japan.

Eduardo Rolim Teixeira

Graduate Student, Department of Removable

Prosthodontics, Hiroshima University, School of
Dentistry, Hiroshima, Japan.

Yuuji Sato

Associate Professor, Department of Removable

Prosthodontics, Hiroshima University, School of
Dentistry, Hiroshima, Japan.

Yasumasa Akagawa

Professor and Chairman, Department of

Removable Prosthodontics, Hiroshima University,
School of Dentistry, Hiroshima, Japan.

FIGURES

Footnotes 7
Figure 1a

Fig. 1a Radiograph taken 2 months after implantation in the control group. Arrows indicate sites
from which measurements were made.

Figure 1b

Fig. 1b Radiograph taken 9 months after implantation in the control group.

Figures 8
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

Figure 1c

Fig. 1c Radiograph taken 3 years after implantation in the control group.

Figure 2a

Fig. 2a Radiograph taken 2 months after implantation in the gingivitis group. Arrow indicates
marginal bone height.

Figures 9
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

Figure 2b

Fig. 2b Radiograph taken 9 months after implantation in the gingivitis group.

Figure 2c

Fig. 2c Radiograph taken 3 years after implantation in the gingivitis group. Arrow indicates
marginal bone loss.

Figures 10
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

Figure 3

Fig. 3 Correlation between bone loss (in millimeters) and PI, GI, and PS, divided in groups
according to the verified scores. *Statistical significance (P<.05).

TABLES

Table 1

Tables 11
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81): Correlation Between Mucosal Inflammation and Marginal Bone L

Table 2

Table 3

Correlation Between Mucosal Inflammation and Marginal Bone Loss Around

1. Brunski JB, Moccia AF Jr, Pollack SR, Korostoff E, Trachten berg D. The influence of functional use of
endosseous dental implants on the tissue-implant interface. I. Histological aspects. J Dent Res
1979;58:1953–1969.
2. Brunski JB, Moccia AF Jr, Pollack SR, Korostoff E, Trachtenberg D. The influence of functional use of
endosseous dental implants on the tissue-implant interface. II. Clinical aspects. J Dent Res 1979;58:1970–1980.
3. Akagawa Y, Takata T, Matsumoto T, Nikai H, Tsuru T. Correlation between clinical and histological
evaluations of the peri-implant gingiva around the single-crystal sapphire endosseous implant. J Oral Rehabil
1989;16:581–587.

References 12
JOMI on CD-ROM (1997 © Quintessence Pub. Co.), 1997 Vol. 12, No. 1 (74 - 81):

4. Akagawa Y, Matsumoto T, Hashimoto M, Tsuru T. Clinical evaluation of the gingiva around single-crystal
sapphire endosseous implant after experimental ligature-induced plaque accumulation in monkeys. J Prosthet
Dent 1992;68:111–115.
5. Hashimoto M, Akagawa Y, Nikai H, Tsuru H. Single-crystal sapphire endosseous dental implants loaded with
functional stress: Clinical and histological evaluation of peri-implant tissues. J Oral Rehabil 1987;15:65–76.
6. Hashimoto M, Akagawa Y, Nikai H, Tsuru H. Ultrastructure of the peri-implant junctional epithelium on
single-crystal sapphire endosseous dental implant loaded with functional stress. J Oral Rehabil
1989;16:261–270.
7. Kirsch A, Mentag P. The IMZ endosseous two phase implant system: A complete oral rehabilitation treatment
concept. J Oral Implantol 1986;12:576–589.
8. Block MS, Kent JN, Finger IM. Factors associated with tissue compromise with endosseous dental implants
[abstract 1268]. J Dent Res 1990;69(special issue):267.
9. Block MS, Kent JN. Long-term follow-up on hydroxylapatite-coated cylindrical dental implants: A comparison
between developmental and recent periods. J Oral Maxillofac Surg 1994;52:937–943.
10. Henry PJ, Tolman DE, Bolender C. The applicability of osseointegrated implants in the treatment of partially
edentulous patients: Three-year results of a prospective multicenter study. Quintessence Int 1993;24:123–129.
11. Apse P, Zarb GA, Schmitt A, Lewis DW. The longitudinal effectiveness of osseointegrated dental implants.
The Toronto study: Peri-implant mucosal response. Int J Periodont Rest Dent 1991;11:95–111.
12. Lekholm U, Adell R, Lindhe J, Brånemark P-I, Eriksson B, Rockler B, et al. Marginal tissue reactions at
osseointegrated fixtures. (II) A cross-sectional retrospective study. Int J Oral Maxillofac Surg 1986;15:53–61.
13. Smith DE, Zarb GA. Criteria for success of osseointegrated endosseous implants. J Prosthet Dent
1989;62:567–572.
14. Akagawa Y, Ichikawa Y, Kubo T, Sagara M, Nagakane K, et al. Clinical evaluation of the newly developed
two-stage hydroxyapatite-coated dental implant in humans [in Japanese]. Hiroshima Daigaku Shigaku Zasshi
1992;24:227–234.
15. Silness J, Löe H. Periodontal disease in pregnancy II. Correlation between oral hygiene and periodontal
condition. Acta Odontol Scand 1964;22:121–135.
16. Löe H, Silness J. Periodontal disease in pregnancy I. Prevalence and severity. Acta Odontol Scand
1963;21:533–551.
17. Garnick JJ, Pearson R, Harrell D. The evaluation of the Periotron. J Periodontol 1979;50:424–426.
18. Adell R, Leckholm B, Rockler B, Brånemark P-I. A 15-year study of osseointegrated implants in the treatment
of the edentulous jaw. Int J Oral Surg 1981;10:387–416.

References 13