Association between periodontitis and serum antioxidants depends upon case definition

Epidemiological studies exploring associations between periodontitis and risk indicators employ inconsistent case definitions, limiting data comparisons (Borrell & Papapanou 2005). Objective: To determine associations between serum antioxidant concentrations and periodontitis using dual case definitions. Methods: Data from 11,895 adults from NHANES III was analysed. Mild periodontitis (n=1567) was defined as ≥one site with both clinical attachment loss (CAL) ≥4mm and probing depth (PD) ≥4mm (Tomar & Asma 2000) and severe disease (n=609) as ≥2 mesiobuccal sites with CAL ≥5mm and ≥one mesiobuccal sites with PD ≥4mm (modified from CDC). Multiple logistic regression was employed to determine associations between serum concentrations of vitamin C, bilirubin and derived total antioxidant capacity (TAOC) and periodontitis. Three models were evaluated: simple model (adjusting for age, gender, race/ethnicity), full model (age, gender, race, smoking, female hormones, diabetes, poverty-income ratio, education) and full model restricted to never-smokers. Results: Significant inverse associations were found between serum vit-C and periodontitis prevalence in all models and for both case definitions. This was also true for bilirubin in simple and full models, but not in never-smokers. In mild periodontitis there was no association between serum TAOC except in never-smokers, but significant inverse associations were found for severe periodontitis in all models. Inverse associations were stronger in the more severe case definition for all antioxidants and in all models except bilirubin in never-smokers. The odds ratios (OR) and 95% confidence intervals (CI) for TAOC in simple, full and never-smoker models respectively were: mild 0.83 (0.68, 1.01), 0.83 (0.68, 1.03), 0.7 (0.5, 0.98) and severe 0.57 (0.43, 0.76), 0.63 (0.47, 0.85), 0.55 (0.33, 0.93). Conclusions: The outcome of association studies depends upon the case definition employed. Significant protective effects of increasing serum total antioxidant concentrations were demonstrated for severe periodontitis but only in never smokers with mild periodontitis.