Over 500 residents of Tecumseh, Michigan, were dentally examined in 1959 as part of a community-wide health study. In 1987, the dental examinations were repeated, with use of the same criteria as in 1959, for 167 dentate persons from the original group. Another 28 reported by telephone that they had become edentulous since 1959. This report uses a historical cohort analysis for exploration of the risk factors for tooth loss, both total and partial, over the 28-year period. Over that time, the edentulous lost an average of 18.0 teeth (95% confidence interval 15.5, 20. 7), whereas the age-matched 90 dentate persons lost only 3.2 (2.2, 4.2) teeth each. Descriptive data showed the edentulous to have higher baseline scores for plaque, calculus, and gingivitis, and a higher proportion of them smoked, though only loss of periodontal attachment (LPA) of 4 mm or more, early loss of first molars, and educational attainment were significant risk factors in regression analysis. Odds ratios for these three variables were 4.0 (1.2, 12.8), 2.0 (1.3, 3.1), and 0.6 (0.4, 0.9), respectively. The strongest risk factors for partial tooth loss among 116 dentate persons were baseline gingivitis (which was correlated with LPA of 4 mm or more) and the baseline number of teeth present, with odds ratios of 2.4 (1.2, 5.2) and 0. 8 (0. 7, 1.0), respectively. While the analysis had to be carried out without caries data, it was concluded that total tooth loss is a social-behavioral issue as much as it is disease-related. Social-behavioral factors were less clearly related to partial tooth loss in dentate persons; oral disease characteristics were the most prominent risk factors for partial tooth loss. © 1990, SAGE Publications. All rights reserved.
