Content uploaded by Mine Tezal
Author content
All content in this area was uploaded by Mine Tezal
Content may be subject to copyright.
J Periodontol • July 2005
1123
* Department of Oral Biology, University at Buffalo, Buffalo, NY.
† Department of Social and Preventive Medicine, Epidemiology and Community Health,
University at Buffalo.
‡ Department of Mathematics and Statistics, University of North Carolina Charlotte,
Charlotte, NC.
I
n the United States, about $1.5 billion
is spent annually for replacement of
missing teeth, which represents only a
fraction of the actual amount that could
be spent considering that not all missing
teeth are replaced.
1,2
Tooth loss dimin-
ishes quality of life by causing loss of
function and esthetics and, consequently,
self-esteem.
Distribution of tooth loss is highly
skewed and only a small percentage of
the population is responsible for the
majority of tooth loss.
3
Summary data
may mask differences between subpop-
ulations that are exposed to different
biological and social exposures. Post-
menopausal women represent a subpop-
ulation with unique factors. Estrogen
deficiency after menopause
2,4
and con-
sequent loss of bone mineral density
5,6
have been shown to be associated with
increased rate of tooth loss. These rela-
tionships may be explained by increased
severity of periodontal disease
7-12
and
decreased bone mineral density
13,14
in
estrogen deficiency.
The role of baseline periodontal status
on the incidence of tooth loss in post-
menopausal women has not been assessed
in previous studies. The aims of this study
were to describe the distribution of incident
tooth loss over 10.6 to 13.3 years of follow-
up and determine to what extent the inci-
dence of tooth loss is associated with the
baseline periodontal variables in a popu-
lation of postmenopausal women.
MATERIALS AND METHODS
Study Population
The population of this cohort study was
derived from individuals who participated
Periodontal Disease and the Incidence of
Tooth Loss in Postmenopausal Women
Mine Tezal,*
†
Jean Wactawski-Wende,
†
Sara G. Grossi,* Jacek Dmochowski,
‡
and Robert J. Genco*
Background: The role of periodontal disease as a predictor
of incident tooth loss in postmenopausal women has not been
determined. The aim of this cohort study was to determine the
extent of the association between baseline periodontal status
and incident tooth loss in a population of postmenopausal
women.
Methods: The study population included 106 dentate white
postmenopausal women who participated in a cross-sectional
study between 1989 and 1991 who were willing and eligible to
have a repeat examination after 10 to 13 years. At baseline,
full-mouth assessment of periodontal status was performed clin-
ically and radiographically. Assessment of tooth loss during fol-
low-up was assessed clinically by a periodontist. Odds ratio
(OR) and its 95% confidence interval (CI) for each periodontal
variable was obtained from separate multiple logistic regression
analyses adjusting for the effect of age, household income,
smoking, hormone therapy, snack consumption, and number
of decayed teeth.
Results: Sixty-one (57.5%) subjects lost at least one tooth
during follow-up. Mean tooth loss per person was 1.81 ± 2.77.
After adjusting for confounders, each millimeter of alveolar bone
loss at baseline increased the risk of tooth loss 3-fold (OR =
3.26; 95% CI: 1.60 to 6.64). The risk of tooth loss also increased
2.5 times for each millimeter of clinical attachment loss (OR =
2.50; 95% CI: 1.24 to 5.07). Probing depth (OR = 2.53; 95% CI:
0.98 to 6.53), gingival bleeding (OR = 1.99; 95% CI: 0.21 to
18.94), calculus (OR = 2.05; 95% CI: 0.91 to 4.61), and plaque
(OR = 0.70; 95% CI: 0.13 to 3.34) were not significantly asso-
ciated with incident tooth loss.
Conclusion: Periodontal disease, especially measured by alve-
olar bone loss, is a strong and independent predictor for inci-
dent tooth loss in postmenopausal women. J Periodontol
2005;76:1123-1128.
KEY WORDS
Periodontal diseases; postmenopause; risk factors;
tooth loss.
40205.qxd 7/7/05 8:12 AM Page 1123
1124
Periodontal Disease and Tooth Loss in Postmenopausal Women Volume 76 • Number 7
in the cross-sectional Erie County Study
15,16
on risk
factors for periodontal disease between July 1989 and
December 1991. Subjects were recruited using vari-
ous methods including random selection from census
tract data (N = 365), convenience patients who pre-
sented for treatment at the University at Buffalo,
School of Dental Medicine (N = 325), and respondents
to advertisements in local newspapers (N = 736). Sub-
jects who had a history of invasive cancer or required
antibiotic prophylaxis for subacute bacterial endocardi-
tis were excluded at baseline. A total of 1,426 den-
tate men and women were enrolled in the Erie County
Study.
Follow-up data for the current study was obtained
between May 2002 and June 2003. Inclusion criteria
were female gender, permanent menopause at base-
line, age at menopause ≥40 years, and presence of
≥ six natural teeth at baseline. Contacted subjects who
developed cancer during follow-up were excluded. The
cohort was restricted to white women due to an insuf-
ficient number of black women for statistical analy-
ses. Information on menopausal status was not
obtained at baseline. Therefore, only women aged 45
or older at baseline examination were targeted and
information on menopausal status and age at
menopause was obtained at follow-up retrospectively.
A total of 303 subjects met the inclusion criteria for the
follow-up study and were targeted for recruitment.
Among the subjects targeted for recruitment, 34
(11.2%) had died, 23 (7.6%) moved out of town, 14
(4.6%) were unable to participate due to health rea-
sons, 59 (19.5%) refused to participate, 32 (10.6%) did
not respond after five or more telephone call attempts,
and 35 (10.6%) could not be contacted due to dis-
connected or wrong telephone numbers. The final study
population consisted of 106 subjects (35% of the tar-
get population).
Definition of Variables
Baseline clinical examination included assessments of
clinical attachment loss (CAL), probing depth (PD),
gingival bleeding (GB), supragingival plaque, and cal-
culus and was performed on all existing teeth except
the third molars by a team of nine trained and cali-
brated dentists according to the National Institutes of
Health guidelines.
17
Supragingival plaque and gingival
bleeding were assessed on three surfaces per tooth
(buccal, mesio-buccal and lingual). The criteria for a
positive plaque score was the visualization of plaque
regardless of the amount. The presence of gingival
bleeding was assessed in response to a periodontal
probe (Michigan O probe) run along the gingival sul-
cus. For calculus, a single score was assigned for each
tooth utilizing a no. 17 dental explorer according to
the following scale: 0: calculus absent; 1: presence of
supragingival calculus but no subgingival calculus; or
2: supragingival and subgingival calculus or subgingi-
val calculus alone. CAL was measured on six surfaces
per tooth (disto-buccal, buccal, mesio-buccal, disto-
lingual, lingual, and mesio-lingual) and was defined
as the distance between the cemento-enamel junction
(CEJ) and base of the gingival sulcus. The distance
from CEJ to gingival margin (GM) was measured
with Michigan O probe, and PD at the same site
was measured with a constant-force (20 g) electronic
probe.
18
CAL was automatically calculated by a
computer program according to the formula: CAL =
PD – (GM-CEJ).
Alveolar bone loss (ABL) was determined from six
anterior periapical and four posterior vertical bite-wing
radiographs taken with a Rinn alignment system.
E-speed films, size 0 for the anterior periapicals and
size 2 for the posterior vertical bite-wings, were uti-
lized. The radiographs were taken using an x-ray unit
operating at 70 kVp and 15 mA setting. The patients
wore a lead lined apron with a thyroid collar. An auto-
matic processor was utilized for film processing. ABL
was measured on mesial and distal surfaces of all teeth
present, except the third molars and the canines, using
a computer program working with digitized radio-
graphic images.
19,20
The program allows measure-
ment of the distance from CEJ to bone crest in a line
parallel to the long axis of the tooth.
Information on covariates was obtained by ques-
tionnaires and included age (years), household income
(<$10,000; $10,000 to 29,999; ≥$30,000), smoking
status (never, former, current), hormone therapy
(never, former, current), number of snacks per day,
and number of decayed teeth. For both smoking sta-
tus and hormone therapy, current was defined as pres-
ence of the exposure at baseline regardless of the
amount (dosage); former was defined as quitting the
exposure before the baseline examination regardless
of the time since quitting.
At follow-up, missing teeth, caries, and restorations
were assessed clinically by one periodontist, using the
same criteria as the baseline examination. Subjects
were also asked about the reasons for tooth loss and
these were grouped as periodontal disease, caries,
failed endodontic treatment, orthodontic reasons, frac-
ture or accident, unerupted, and congenitally missing.
Reliability of the Measurements
During the study period, replicate measurements of
ABL, CAL, and PD were obtained on randomly selected
20% of the subjects by nine calibrated examiners.
Replicate measurements were performed on randomly
selected quadrants and the examiner had no previous
knowledge of the scheduled second measurements at
the time of the first measurements. Overall, intraexaminer
mean ± SD of differences between replicate measure-
ments for PD was 0.57 ± 0.16 mm; for CAL, 0.76 ±
40205.qxd 7/7/05 8:12 AM Page 1124
1125
J Periodontol • July 2005 Tezal, Wactawski-Wende, Grossi, Dmochowski, Genco
33 mm; and for ABL, 0.30 ± 38 mm. Interexaminer
mean ± SD of differences between replicate measure-
ments for PD was 0.73 ± 0.12 mm; for CAL, 0.95 ±
0.18 mm; and for ABL, 0.34 ± 0.42 mm.
19-21
There
was a perfect (100%) agreement on repeated assess-
ment of tooth loss at follow-up.
Statistical Analyses
Descriptive statistics included means, standard devia-
tions, frequencies, and proportions and were used to
describe the study population and the distribution of tooth
loss. Correlation coefficients, chi square tests, t tests, and
crude odds ratios were used to select variables for the
multivariate models. Variables that showed associations
with incident tooth loss in unadjusted analyses (P <0.20)
were entered into multivariate models. Over 50 baseline
variables including demographic, lifestyle, health, pro-
fessional, and self dental care characteristics were eval-
uated as potential confounders. Separate multiple logistic
regression analyses were used to determine the inde-
pendent effect of each periodontal variable on incident
tooth loss after adjusting for the effects of confounders.
Incident tooth loss was defined as ≥1 tooth loss between
baseline and follow-up examinations. Odds ratios and
their 95% CI were obtained. Possible interactions between
independent variables were tested in multivariate models.
Baseline characteristics of the study population and
the population lost to follow-up were compared with
chi square and t tests to assess the presence of sys-
tematic differences between the two populations as a
source of possible bias.
RESULTS
Sixty-one (57.5%) subjects lost at least one tooth dur-
ing follow-up. The mean number of teeth lost was 1.81
(range: 0 to 17). Among those who lost teeth, the aver-
age number of teeth lost was 3.15; the majority (35
[57.4%]) lost one or two teeth, 19 (31.1%) subjects lost
three to five teeth, and the remaining seven (11.5%)
subjects lost six to 17 teeth. None of the subjects became
totally edentulous. During the follow-up period, a total
of 192 teeth were lost. According to subject self-report,
119 (62%) teeth were lost due to periodontal disease
and 73 (38%) teeth were lost due to caries or failed
endodontic treatment. Eighteen (30%) subjects lost teeth
due to periodontal disease (mean: 6.61 per person); 34
(55%) subjects lost teeth due to caries or failed endodon-
tic treatment (mean: 2.15 teeth per person), and nine
(15%) subjects lost teeth due to both periodontal disease
and caries (mean: 8.11 teeth per person).
The average follow-up time was 11.7 years and ranged
between 10.6 and 13.3 years. The baseline age ranged
from 45 to 73 years (mean: 58.03). All subjects had
reached menopause at baseline, with an average age at
menopause of 48.8 years. Slightly more than half of the
study population (54.8%) had high school or higher edu-
cation and the majority of the study population (59.4%)
had a household income between $10,000 and $29,999.
More than a quarter of the study population (26.4%) had
a history of hormone therapy (HT) and 22.6% were still
on HT at baseline. Only two (1.9%) women reported hav-
ing osteoporosis at baseline. The percentages of current
and former smokers at baseline were 17.9% and 34%,
respectively. About half of the study population (47.2%)
consumed two or more snacks a day. The study popu-
lation had an average of 22.78 remaining teeth and 1.26
decayed teeth at baseline (Table 1).
The subjects had mean ABL of 2.68 mm, mean CAL
of 2.24 mm, mean PD of 2.17 mm, mean gingival bleed-
ing of 0.34, mean calculus of 1.02 and mean plaque of
0.64 (Table 2). Those who lost teeth had significantly
higher mean baseline ABL (2.99 versus 2.25, P = 0.001)
compared to those who did not lose teeth. They also had
higher mean CAL (2.37 versus 2.07, P = 0.058), PD
Ta b le 1.
Description of the Study Population at
Baseline (N
==
106)
Follow-up time (years) 11.68 ± 0.66 (10.60-13.30)*
Age (years) 58.03 ± 7.58 (45.00-73.00)
Age at menopause (years) 48.79 ± 4.43 (40.00-60.00)
Education (years)
≤12 47 (45.2)
†
>12 57 (54.8)
Household income
<$10,000 15 (14.2)
$10,000-29,999 63 (59.4)
≥$30,000 28 (26.4)
Hormone therapy
Never 78 (73.6)
Former 4 (3.8)
Current 24 (22.6)
Osteoporosis
No 104 (98.1)
Ye s2 (1.9)
Smoking status
Never 51 (48.1)
Former 36 (34.0)
Current 19 (17.9)
N snacks
<2/day 56 (52.8)
≥2/day 50 (47.2)
N remaining teeth 22.78 ± 4.92 (7.00-28.00)
N decayed teeth 1.26 ± 2.07 (0.00-12.00)
* Mean ± SD (range).
† N (%).
40205.qxd 7/7/05 8:12 AM Page 1125
1126
Periodontal Disease and Tooth Loss in Postmenopausal Women Volume 76 • Number 7
Ta b le 2.
Description of the Study Population by
Baseline Periodontal Variables
Variable Mean ± SD (range)
ABL (mm) 2.68 ± 1.02 (0.93-6.34)
CAL (mm) 2.24 ± 0.79 (0.92-5.79)
PD (mm) 2.17 ± 0.52 (1.16-4.01)
Gingival bleeding (0-1) 0.34 ± 0.19 (0.00-1.00)
Calculus (0-2) 1.02 ± 0.61 (0.00-2.00)
Plaque (0-1) 0.64 ± 0.29 (0.11-1.00)
Ta b le 4.
Crude Odds Ratios of Baseline Periodontal
Variables for Incident Tooth Loss (N
==
106)
Variable OR 95% CI P
ABL (per mm) 2.56 1.49-4.40 0.001
CAL (per mm) 1.72 0.97-3.05 0.066
PD (per mm) 1.48 0.69-3.18 0.320
Gingival bleeding (0-1) 1.77 0.24-12.83 0.573
Calculus (0-2) 1.64 0.85-3.15 0.141
Plaque (0-1) 1.11 0.30-4.14 0.873
(2.21 versus 2.11, P = 0.38), gingival bleeding (0.35
versus 0.33, P = 0.57), calculus (1.09 versus 0.92, P =
0.14), and plaque (0.64 versus 0.63, P = 0.87), but these
differences were not statistically significant (Table 3).
The risk of tooth loss significantly increased by 2.5-
fold for each millimeter of alveolar bone loss at baseline
(OR = 2.56; 95% CI: 1.49 to 4.40). Tooth loss also
increased with increasing levels of baseline CAL (OR =
1.72; 95% CI: 0.97 to 3.05), PD (OR = 1.48; 95% CI:
0.69 to 3.18), gingival bleeding (OR = 1.77; 95% CI:
0.24 to 12.83), and calculus (OR = 1.64; 95% CI: 0.85
to 3.15), but these associations did not reach statistical
significance. There was no relationship between base-
line plaque levels and incident tooth loss (OR = 1.11;
95% CI: 0.30 to 4.14) (Table 4).
After adjusting for age, household income, smoking,
hormone therapy, number of snacks per day, and num-
ber of decayed teeth, the risk of tooth loss significantly
increased by 3-fold for each millimeter of ABL at base-
line (OR = 3.26; 95% CI: 1.60 to 6.64) and by 2.5 times
for each millimeter of CAL at baseline (OR = 2.50; 95%
CI: 1.24 to 5.07). The risk of tooth loss also increased
Ta b le 5.
Adjusted* Odd Ratios of Baseline
Periodontal Variables for Incident Tooth
Loss
NOR 95% CI P
ABL (per mm) 106 3.26 1.60-6.64 0.001
CAL (per mm) 106 2.50 1.24-5.07 0.011
PD (per mm) 106 2.53 0.98-6.53 0.054
Gingival bleeding (0-1) 106 1.99 0.21-18.94 0.550
Calculus (0-2) 106 2.05 0.91-4.61 0.083
Plaque (0-1) 106 0.70 0.13-3.34 0.624
* Age, income, smoking, hormone therapy, number of snacks per day, and
number of decayed teeth.
with increasing levels of baseline PD (OR = 2.53; 95%
CI: 0.98 to 6.53), gingival bleeding (OR = 1.99; 95%
CI: 0.21 to 18.94), and calculus (OR = 2.05; 95% CI:
0.91 to 4.61), but these associations were not statis-
tically significant. There was no relationship between
baseline plaque levels and the risk of incident tooth loss
(OR = 0.70; 95% CI: 0.13 to 3.34) (Table 5).
DISCUSSION
In this study, roughly 5% of the study population per year
lost at least one tooth during follow-up, with a mean
number of 1.81 teeth per person (0.16 teeth per year).
Two previous cohort studies have reported incidence of
tooth loss in postmenopausal women.
4,6
In a retrospec-
tive cohort study of 7 years with 189 healthy post-
menopausal women with a mean age of 60 years, 24%
of the subjects (3.4% per year) lost at least one tooth.
6
Ta b le 3.
Description of Incident Tooth Loss by
Baseline Periodontal Variables
Tooth Loss
Variable 0 ≥1 P*
ABL (mm) 2.25 ± 0.55
†
2.99 ± 1.10 0.001
CAL (per mm) 2.07 ± 0.69 2.37 ± 0.84 0.058
PD (per mm) 2.11 ± 0.47 2.21 ± 0.55 0.322
Gingival bleeding (0-1) 0.33 ± 0.21 0.35 ± 0.18 0.576
Calculus (0-2) 0.92 ± 0.64 1.09 ± 0.57 0.141
Plaque (0-1) 0.63 ± 0.31 0.64 ± 0.28 0.874
* t tests.
† Mean ± SD.
40205.qxd 7/7/05 8:12 AM Page 1126
1127
J Periodontol • July 2005 Tezal, Wactawski-Wende, Grossi, Dmochowski, Genco
In that study, tooth loss was assessed by questionnaires.
In the second study with 42,171 postmenopausal regis-
tered nurses, aged 56 to 71 years, from 11 states across
the United States, 23% (11.5% per year) of the popula-
tion reported having lost at least one tooth in the last
2 years by mailed questionnaires.
4
Therefore, incidence
of tooth loss in our population was within the range
of previous studies that examined a postmenopausal
women population.
This study was the first one that assessed the role
of baseline periodontal disease on the incidence of
tooth loss in postmenopausal women. The three previous
cohort studies
4-6
of tooth loss examining postmenopausal
women populations did not adjust for or assess the effect
of baseline periodontal disease levels. In our study, among
periodontal variables, ABL was the strongest predictor of
incident tooth loss. After adjusting for confounders, each
millimeter of ABL at baseline increased the risk of future
tooth loss by more than three times. CAL and PD were
also associated with incident tooth loss; however, the point
estimates for these associations were somewhat more
modest than for ABL. The reason for their weaker asso-
ciations is partly due to lower precisions of these mea-
surements compared to ABL measurements. In addition,
some systemic factors in postmenopausal women such
as reduced bone mineral density after menopause may
have a more direct relationship with ABL compared to PD
and CAL. After the addition of ABL, neither CAL nor PD
contributed significantly to the model. Therefore, of the
three periodontal measures, ABL was the best predictor
of the incidence of tooth loss.
Four previous cohort studies assessed the effect of
baseline ABL on the risk of incident tooth loss in other
populations.
22-26
In a retrospective cohort study of 0.33
to 15.17 years with 100 consecutive adult maintenance
patients (no age, gender, or race information was
reported) from a clinician’s appointment book who had
initially moderate to severe periodontitis, 1% alveolar
bone loss at baseline was associated with a 3% increase
in the risk of tooth loss (relative risk = 1.03, P = 0.0003)
after adjusting for probing depth, furcation involvement,
mobility, parafunctional habit without a bite-guard, and
smoking.
22
Another retrospective cohort study
23
of 23
years follow-up in 690 predominantly white veteran men
(97%) aged 21 to 75 years at baseline, showed that
mean ABL at baseline (r = 0.12, P <0.001) and percent
sites with ABL progression (r = 0.11, P <0.001) were
significantly associated with the number of teeth lost
during follow-up after adjusting for age, smoking, edu-
cation, probing depth, and number of remaining teeth
at baseline. ABL was measured at two sites per tooth
from periapical radiographs with Schei ruler with 20%
increments.
23
In a retrospective cohort study of 1.6 to
5.2 years with 415 predominantly white (95.6%) men
and women with mild or no periodontal disease and
aged 25 to 75 years at baseline, subjects who lost at least
one tooth during follow-up had significantly higher base-
line ABL compared to subjects who did not lose teeth
(2.43 mm versus 1.95 mm, P = 0.0001).
24
Finally, in a
prospective cohort study of 20 years with 515 white
men and women from Stockholm, aged 18 to 65 years
at baseline, mean ABL at baseline was significantly
related to the number of teeth lost during follow-up both
in unadjusted analyses (r = 0.49, P <0.001) and after
adjusting for plaque, number of missing teeth at base-
line, age, and education (β=0.25, P <0.001).
25,26
It is
not possible to directly compare our results to previous
studies that employed other populations due to differ-
ent measurement methods, statistical analyses, and
follow-up periods. However, the predictive ability of ABL
for incident tooth loss seems to be stronger in post-
menopausal women compared to other populations.
This may be explained by the high rate of systemic bone
loss in postmenopausal women.
A limitation of our study is that no information was
available on systemic bone loss (i.e., bone density)
and only two women reported having osteoporosis at
baseline; therefore, we could not evaluate this para-
meter. Since systemic bone loss has been shown to be
a significant factor for tooth loss,
5,6
future studies of
postmenopausal women with larger samples sizes are
needed for an estimate of ABL effect size adjusted by
systemic bone loss.
Another limitation of this study was the exclusion of
third molars from clinical examination and the exclu-
sion of canines from ABL measurements at baseline.
Exclusion of third molars is a potential source of bias
because loss of these teeth is not counted in the inci-
dence rates. However, the effect of bias on the esti-
mated rate of tooth loss is likely to be low. In a study
of extractions with a random sample of dental prac-
tices, only 1.9% of adults lost third molars and half of
these subjects lost other teeth as well.
27
In our study,
this bias is expected to be even lower due to the older
age of the subjects. The bias from excluding the canines
from ABL measurements is also negligible since the
mean ABL was used as the independent variable.
Loss to follow-up in our study was high (65%) due
to a long follow-up period and the older age of subjects.
The study population was compared to the population
lost to follow-up by baseline variables to evaluate
whether the observed results were biased. The popu-
lation lost to follow-up had a significantly lower base-
line number of missing teeth (21.44 versus 22.78, P =
0.03) and years of education (12.86 versus 13.49, P =
0.05) compared to the study population. Therefore, the
effects of these two variables on incident tooth loss
may be underestimated in our study. However, there
were no significant differences in periodontal variables
between the two populations. Therefore, effects of perio-
dontal variables on incident tooth loss were not biased.
In previous literature, loss to follow-up for 9- to 15-year
40205.qxd 7/7/05 8:12 AM Page 1127
1128
Periodontal Disease and Tooth Loss in Postmenopausal Women Volume 76 • Number 7
cohort studies ranged between 42% and 86%.
28-31
Therefore, the follow-up rate of our study was within the
range of previous studies with similar follow-up periods.
The long follow-up period, reliable and highly sen-
sitive methodology to measure periodontal status by
trained and calibrated examiners, and a well-defined
baseline population are advantages of this study and
allowed us to test the study hypotheses reliably.
We can conclude that periodontal status at baseline,
especially measured by alveolar bone loss, is a strong
independent predictor for incident tooth loss in post-
menopausal women. Control of periodontal disease can
significantly reduce tooth loss in postmenopausal women.
ACKNOWLEDGMENTS
This study was supported by USPHS grant DE04898,
DSA grant DE00158, NIDCR grants 1R01-DE13505 and
T32-DE07034, and Army grant DAMD 17-96-1-6319.
REFERENCES
1. Miller Y, Locker D. Correlates of tooth loss in a Canadian
adult population. J Can Dent Assoc 1994;60:549-555.
2. Paganini-Hill A. The benefits of estrogen replacement
therapy on oral health. The Leisure World cohort. Arch
Internal Med 1995;155:2325-2329.
3. Marcus SE, Drury TF, Brown LJ, Zion GR. Tooth reten-
tion and tooth loss in the permanent dentition of adults:
United States, 1988-1991. J Dent Res 1996;75:684-695.
4. Grodstein F, Colditz GA, Stampfer MJ. Tooth loss and
hormone use in postmenopausal women. Compend Con-
tin Educ Dent 1998;(Suppl. 22):S9-S16.
5. Krall EA, Dawson-Hughes B, Papas A, Garcia R. Tooth
loss and skeletal bone density in healthy postmenopausal
women. Osteoporos Int 1994;4:104-109.
6. Krall EA, Garcia RI, Dawson-Hughes B. Increased risk
of tooth loss is related to bone loss at the whole body,
hip and spine. Calcif Tissue Int 1996;59:433-437.
7. Reinhardt RA, Payne JB, Maze CA, Patil KD, Gallagher
SJ, Mattson JS. Influence of estrogen and osteope-
nia/osteoporosis on clinical periodontitis in post-
menopausal women. J Periodontol 1999;70:823-828.
8. Mealey BL, Moritz AJ. Hormonal influences: Effects of
diabetes mellitus and endogenous female sex steroid
hormones on the periodontium. Periodontology 2000
2003;32:59-81.
9. Sooriyamoorthy M, Gower DB. Hormonal influences on
gingival tissue: Relationship to periodontal disease. J Clin
Periodontol 1989;16:201-208.
10. Grossi SG. Effect of estrogen supplementation on perio-
dontal disease. Compend Contin Educ Dent 1998;(Suppl.
22):S30-S36.
11. Genco RJ, Grossi SG. Is estrogen deficiency a risk fac-
tor for periodontal disease? Compend Contin Educ Dent
1998;(Suppl. 22):S23-S29.
12. Tezal M, Wactawski-Wende J, Grossi SG, Ho AW, Dunford
R, Genco RJ. The relationship between bone mineral
density and periodontitis in postmenopausal women.
J Periodontol 2000;71:1492-1498.
13. Kribbs PJ. Comparison of mandibular bone in normal and
osteoporotic women. J Prosthet Dent 1990;63:218-222.
14. Payne JB, Zachs NR, Reinhardt RA, Nummikoski PV, Patil
K. The association between estrogen status and alveolar
bone density changes in postmenopausal women with a
history of periodontitis. J Periodontol 1997;68:24-31.
15. Grossi SG, Zambon JJ, Ho AW, et al. Assessment of
risk for periodontal disease. I. Risk indicators for attach-
ment loss. J Periodontol 1994;65:260-267.
16. Grossi SG, Genco RJ, Machtei EE, et al. Assessment of
risk for periodontal disease. II. Risk indicators for alve-
olar bone loss. J Periodontol 1995:66;23-29.
17. Diagnostic criteria for dental examinations. In: Oral
Health of United States Adults. Epidemiology and Oral
Disease Prevention Program. Bethesda, MD: National
Institute of Dental Research; 1987:159-167. NIH publi-
cation no. 87-2868.
18. Gibbs CH, Hirschfelt JW, Lee JG, et al. Description and
clinical evaluation of a new computerized probe – the
Florida probe. J Clin Periodontol 1988;15:137-144.
19. Hausmann E, Allen K, Dunford R, Christersson L. A reli-
able computerized method to determine the level of the
radiographic alveolar crest. J Periodontal Res 1989;24:
368-369.
20. Hausmann E, Allen K, Carpio L, Christersson LA,
Clerehugh V. Computerized methodology for detection
of alveolar crestal bone loss from serial intraoral radio-
graphs. J Periodontol 1992;63:657-662.
21. Machtei EE, Christersson LA, Grossi SG, Dunford R,
Zambon JJ, Genco RJ. Clinical criteria for the definition of
“established periodontitis.” J Periodontol 1992;63:207-215.
22. McGuire M, Nunn M. Prognosis versus actual outcome.
III. The effectiveness of clinical parameters in accurately
predicting tooth survival. J Periodontol 1996;67:666-
674.
23. Krall E, Garvey A, Garcia R. Alveolar bone loss and
tooth loss in male cigar and pipe smokers. J Am Dent
Assoc 1999;130:57-64.
24. Machtei EE, Hausmann E, Dunford R, et al. Longitudinal
study of predictive factors for periodontal disease and
tooth loss. J Clin Periodontol 1999;26:374-380.
25. Jansson L, Lavstedt S, Zimmerman M. Prediction of
marginal bone loss and tooth loss – a prospective study
over 20 years. J Clin Periodontol 2002;29:672-678.
26. Jansson L, Lavstedt S. Influence of smoking on mar-
ginal bone loss and tooth loss – a prospective study over
20 years. J Clin Periodontol 2002;29:750-756.
27. Locker D, Ford J, Leake JL. Incidence of and risk fac-
tors for tooth loss in a population of older Canadians.
J Dent Res 1996;75:783-789.
28. Nordstrom G, Bergman B, Borg K, Nilsson H, Tillberg
A, Wenslov J. A 9-year longitudinal study of reported
oral problems and dental and periodontal status in 70-
and 79-year-old city cohorts in northern Sweden. Acta
Odontol Scand 1998;56:76-84.
29. Eklund SA, Burt BA. Risk factors for tooth loss in the
United States. Longitudinal analysis of national data.
J Public Health Dent 1994;54:5-14.
30. Holm G. Smoking as an additional risk for tooth loss.
J Periodontol 1994;65:996-1001.
31. Warren JJ, Watkins CA, Cowen HJ, Hand JS, Levy SM,
Kuthy RA. Tooth loss in the very old: 13-15 year inci-
dence among elderly Iowans. Community Dent Oral Epi-
demiol 2002;30:29-37.
Correspondence: Dr. Mine Tezal, 4433 Chestnut Ridge Rd.,
Amherst, NY 14228. E-mail: mtezal@buffalo.edu.
Accepted for publication November 23, 2004.
40205.qxd 7/7/05 8:12 AM Page 1128