Risk indicators for periodontal disease in a racially diverse urban population

A cross-sectional study of 117 subjects from a dental clinic serving a diverse population (i.e., Whites, African-Americans, Native-Americans, and Asians) was performed to evaluate risk indicators of periodontal disease. Gingival crevicular fluid (GCF) and subgingival plaque were taken at the same visit from 4 posterior sites of the most diseased sextant in each subject. Age, smoking packyears, beta-glucuronidase (beta G), neutrophil elastase (NE), myeloperoxidase (MPO), Fusobacterium nucleatum (F. nucleatum), and Porphyromonas gingivalis (P. gingivalis) were significantly (p < 0.05-0.005) correlated with attachment loss. Probing depth was significantly correlated with smoking packyears, beta G, NE, MPO, F. nucleatum and Prevotella intermedia (P. intermedia) (p < 0.05-0.005). Mean NE value of Whites was lower than the mean NE values of African-Americans, Native-Americans and Asians (p < 0.05). Whites had a lower mean beta G value compared to African Americans, and a lower mean MPO value compared to African Americans and Native Americans. The %s of patients positive for F. nucleatum, P. intermedia and Eikenella corrodens (E. corrodens) were higher in Native Americans compared to Whites. Step-wise multiple regression analysis was performed to construct models for the estimation of probing depth and attachment loss. The most parsimonious regression models which had the best R2 values included the following variables and accounted for the indicated % of variability: models 1 and 2: beta G, race, and F. nucleatum accounted for 50% of the variability in mean probing depth and 39% of the variability in a single site (first molar) for probing depth, respectively; model 3: age, beta G, and F. nucleatum accounted for 53% of the variability in mean attachment loss; model 4: age, NE, and F. nucleatum explained 35% of the variability in a single site (first molar) for attachment loss. The results suggest that age, race, smoking packyears, beta G, NE, MPO, F. nucleatum, P. gingivalis and P. intermedia are risk indicators for periodontal disease in this racially diverse urban population. Regression models which include multiple variables (i.e., demographic factors, GCF enzymes and periodontopathic bacteria) can be used to estimate periodontal disease status.