Total Cholesterol as Risk Indicator for Periodontitis

Objectives: Objective:
To describe the cross-sectional relationship between obesity related measures and periodontitis in orally examined participants of The Danish Health Examination Survey (DANHES).
Methods: Methods:
DANHES is a baseline study carried out in Denmark in 2007-08. In total, 18,065 participated in the general health examination. Of those, 4,402 individuals aged 18-96 years had an oral examination in 2008-09. The periodontal examination was based on half-mouth registration at 6 sites per tooth and included level of gingival margin, probing pocket depth (PPD) and calculation of clinical attachment level (CAL). Periodontitis was defined according to severe periodontitis (Page & Eke 2007), specific case definition (Tonetti & Claffey 2005), meanCAL≥2.55mm (corresponding to 75-percentile), CAL-tertile definition (Beck et al. 2001), and PPD-CAL definition (Lopez et al. 2002). A multivariable logistic regression model was used to fit the cross-sectional association of body mass index, body fat, waist hip ratio, total cholesterol, triglycerides, and c-reactive protein with periodontitis adjusting for the effects of relevant covariates i.e. age, gender, number of teeth, smoking, alcohol consumption, diabetes, educational level, respiratory infection, resting heart rate, and physical activity.
Results: Results:
Forty-five participants were excluded due to edentulism and missing data, leaving 4,357 participants for statistical analyses. Multivariable logistic regression analyses showed that total cholesterol [odds ratio, 1.13, (95%confidence interval, 1.00-1.29); 1.11, (1.01-1.21) and 1.12, (1.03-1.22)] was associated with periodontitis defined as severe periodontitis, meanCAL≥2.55mm, and PPD-CAL definition, respectively.
Conclusions: Conclusion:
The results of the present cross-sectional study indicate that total cholesterol is associated with periodontitis. It is important to consider that there are numerous ways to define a periodontitis case, and the analyses of associations obviously show variations dependent on case definition.