Etiologic models for incident periodontal attachment loss in older adults.

Etiologic factors for incident periodontal attachment loss (ALOSS) have not been conclusively identified. The purpose of this study was to develop etiologic models for ALOSS in older adults. Data on 697 older blacks and whites were obtained from 5 sequential examinations over 7 years in the Piedmont 65+ dental study, a complex random sample of older adults in North Carolina. Multivariable Poisson regression models were fit for average number of ALOSS events per person or site month at risk. In models for whites, molar sites, sites with adjacent caries, sites in persons who had: Porphyromonas gingivalis (P.g.) at least 2% of total microbial count (TMC), never had a dental checkup, more depression symptoms, fewer than 12 years of education, higher BANA score, or smoked, had significantly higher rates of ALOSS. In a separate model for blacks, interproximal and molar sites, sites in blacks who had: P.g. at least 2% of TMC, higher BANA score, never had a dental checkup, lower socioeconomic status (SES), or smoked, had significantly higher rates of ALOSS. These results confirm a multifactorial etiology for ALOSS in older adults and indicate that interventions aimed at infection, smoking, and preventive dental care utilization, may be most useful.