Predictors of tooth loss in two US adult populations.

OBJECTIVES

This study determines tooth loss rate over a 10-year period and identifies predictors of tooth loss in two separate US adult longitudinal study populations.

METHODS

Subjects from the Baltimore Longitudinal Study of Aging (BLSA), consisting of 47 men and 47 women, ages ranging from 30 to 69 years, were compared to subjects from the VA Dental Longitudinal Study (VADLS) in Boston, MA, consisting of 481 men in the same age range. Baseline and follow-up examinations were performed on each cohort over a 10-year period. Using multivariate regression models, significant predictors of tooth loss were identified.

RESULTS

A mean rate of tooth loss of 1.5 teeth lost per 10 years was noted in the VADLS cohort compared to 0.6 teeth lost per 10 years in the BLSA (P < .001). Combining subjects from both populations, significant predictors of tooth loss were baseline values of: percent of teeth with restorations, mean probing pocket depth score, age, tobacco use, alcohol consumption, number of teeth present, and male sex. However, the set of significant predictor variables differed between the two populations and sexes. In BLSA men, number of teeth present, percent of teeth with restorations, mean probing pocket depth score, and alcohol consumption, but not age, were significant, while in BLSA women, only age was a significant predictor.

CONCLUSIONS

Over a 10-year period, the incidence of tooth loss, the rates of tooth loss, and the predictors of tooth loss were found to vary by population and by sex. These results illustrate the limits of generalizing tooth loss findings across different study cohorts and indicate that there may exist important differences in risk factors for tooth loss among US adult populations.