Three-year coronal caries incidence in older Canadian adults.

This paper describes the incidence of coronal caries in a sample of older adults. A 3-year follow-up study was conducted of 493 community-dwelling adults aged 50 years and over in Ontario, Canada. The incidence of coronal caries was 57.0%, and the mean net DFS increment was 1.9 surfaces. In bivariate analysis, several variables were significantly associated with incidence and/or mean DFS increment. These included: age, marital status, baseline coronal DFS, number of teeth at baseline, mean periodontal attachment loss of 4 mm or more, and wearing partial dentures. In logistic regression analysis only four factors had significant independent effects. These were level of education, marital status, mean periodontal attachment loss and number of teeth at baseline. The predictive ability of this model was fair: accuracy 65.7%, sensitivity 80.2%, and specificity 46.2%. When logistic analysis was repeated separately for two age groups, different predictors had significant independent effects, and sensitivity and specificity values differed substantially. These findings indicate predictive models for caries incidence should include both clinical and non-clinical variables because both types of variables may help to explain different aspects of coronal caries experience. Further research is required to identify other factors associated with coronal caries in older adults.