Risk factors for total tooth loss in the United States; longitudinal analysis of national data.

The NHANES I Epidemiologic Follow-up Study (NHEFS) of 1982-84 collected longitudinal data from 10,523 individuals initially seen during the first National Health and Nutrition Examination Survey (NHANES I) of 1971-75. Among this additional data was information on the incidence of total tooth loss during the 10 years between the surveys, which could then be added to NHANES I data to identify risk factors. In this analysis, a series of bivariate analyses were carried out, followed by logistic regression analysis to assess the simultaneous effect of major variables. Results showed that 7.4 percent of dentate Americans aged 25-74 at NHANES I became edentulous over the next 10 years. In bivariate analyses, the incidence of edentulism was correlated with baseline measures of lower income and education status, poorer oral health, self-perceptions of poor general health and oral health, absence of a regular dentist, and a lower number of remaining teeth at baseline. No correlation was found with gender and geographic region, nor with self-reported diabetes and arthritis, and age was not a factor when the number of remaining teeth at baseline were taken into account. In a logistic regression model assessing the effect of these variables simultaneously, none of the demographic variables retained significance; the only variable statistically significant in both age groups was the number of teeth remaining at baseline. Other significant variables in younger persons were higher periodontal disease scores, perceived poor dental health, perceived need for extractions, history of smoking, and low ascorbic acid intake. Some of these variables were reflections of negative health behavior and attitudes rather than direct correlates. Principal findings from this study were the importance of early tooth loss in eventual edentulism and the virtual disappearance of gender and age as determinants of total tooth loss.