A multilevel analysis model for dental caries determinants in independently-living elderly

To evaluate the prevalence of untreated caries and its association with biological, individual, and environmental variables in independently-living elderly people. Methods: This cross-sectional study included 72 elderly (≥60 years) patients of a university dental clinic in Belo Horizonte, Brazil. Sociodemographic data, systemic diseases, medications, and free sugar intake were collected. Visible plaque, Decayed, Missing, and Filled Teeth (DMFT), and Decayed and Filled Root (DFR) indexes were assessed through clinical examination. Unstimulated saliva was collected to determine salivary flow, pH, and buffering capacity. Descriptive analysis and multilevel logistic regression analysis were performed following a dental caries theoretical model (p <0.05, 95% CI). Results: The mean DMFT and DFR were 24.44 (SD=4.59) and 3.21 (SD=2.93), respectively. The prevalence of untreated caries was 61.11%. In the adjusted multilevel regression model involving 1639 teeth, untreated dental caries was significantly associated with the presence of biofilm (OR = 1.84; 95% CI: 1.24­2.74), salivary buffering capacity (OR = 0.87; 95% CI: 0.77­0.99) and per capita income (OR = 0.06; 95% CI: 0.004­0.74). Conclusion: The experience of dental caries was widespread among independently-living elderly patients, and its variability was best explained by the presence of biofilm, reduced salivary buffering capacity, and low per capita income. A comprehensive assessment is needed of the biological, individual, and environmental factors related to the presence of dental caries in independently-living elderly people