Skssk

Natural History of Periodontal Disease in Adults:
Findings from the Tecumseh Periodontal Disease Study, 1959-87
A.I. ISMAIL, E.C. MORRISON', B.A. BURT2, R.G. CAFFESSE', and M.T. KAVANAGH
Faculty of Dentistry, Dalhousie University, Halifax, NS, Canada B3H 3J5; 'Department of Periodontics, Dental Branch, University of Texas
Health Science Center, Houston, Texas 77225; and 2Program in DentalPublic Health, School of Public Health, Ann Arbor, Michigan 48109
The purpose of this epidemiological study was to estimate the degree
of change in periodontal attachment level in a sample of adults examined in 1959 and 1987 in Tecumseh, Michigan. Out of 526 individuals between the ages of five and 60 years in 1959, a sample of
325 resided within an 80-km-radius area in 1987. Of those, 167 werere-examined. Loss of periodontal attachment (LPA) was determined
with a Michigan #0 probe on four tooth sites (disto-buccal, midbuccal, mesio-buccal, mid-lingual) for all teeth present. Of the individuals contacted, 28 had lost all their teeth during the 28 years. Of
the 167 adults examined, two refused periodontal probing. Out of the
165 adults with LPA measurements in 1987, only 22 (13.3%) hadan
average increased loss of 2 mm or more per person between 1959
and 1987; five adults (3.0%) had an average LPA increase of 3 mm
or more, and only two adults (1.2%) had an average LPA increase
of 4 mm or more. The attachment level in 59.3% of all the tooth sites
examined in 1959 in the 165 individuals either did not change or
changed within ± 1.0 mm. On the basis of bivariate analyses,the
individuals with high LPA increase (.2 mm) had the following characteristics significantly different from those with low LPA increase:
They were older, smoked, had tooth mobility at baseline, higher gingivitis, plaque, calculus, and tooth mobility scores at follow-up, lower
education level, and irregular dental attendance. However, in logistic
regression analyses, only the following riskmarkers remained significantly associated with high LPA increase: age, smoking, and presence of tooth mobility.
J Dent Res 69(2):430-435, February, 1990

Introduction.
In oral epidemiology, most published work has concentrated
on describing the distribution of oral conditions in populations

and identifying the risk markers associated with these conditions [a risk marker is a factor associatedwith an increased
probability of having a disease, but is not necessarily a causal
factor (Last, 1983)]. Even for the most widely studied oral
condition, dental caries, few studies of the natural history of
the disease can be identified (Hughes et al., 1982). The progression of periodontal disease in American adults has received
even less investigation (Hughes et al., 1982; Feldman et al.,1987). Data from the North Carolina Studies (Hughes et al.,
1982) are not easily interpreted because they are expressed as
Periodontal Index (Russell, 1956) averages.
The most comprehensive studies of the natural history of

periodontal disease to date were carried out by JJe et al. (1978,
1986) in Sri Lanka and Norway. In each country, a group of
adults was followed longitudinally. Thefirst detailed analysis
showed that the Norwegians had a lower severity of loss of
periodontal attachment (LPA), compared with the Sri Lankan
sample (Lde et al., 1978). In the Norwegians, 40 years of age,
Received for publication May 29, 1989
Accepted for publication September 19, 1989
This investigation was supported in part by USPHS Research Grant
DE-08051 from the National Institute ofDental Research, National
Institutes of Health, Bethesda, MD 20892.
Data were collected when the first author was at the Faculty of
Dentistry, McGill University, Montreal.
430

the mean annual rates of attachment loss were 0.08 and 0.10
mm for interproximal and buccal surfaces, respectively. In the
Sri Lankans, 40 years of age, the mean annual rates of attachment loss were 0.30 and 0.20 mm...