Serum Antibodies to Periodontal Bacteria as Diagnostic Markers of Periodontitis

Objectives: Determination of periodontal status in epidemiologic studies requires a clinical periodontal examination. In a nationally representative sample, we investigated the ability of serum IgG antibodies to selected periodontal bacteria to reflect prevalent periodontitis.

Methods: We used checkerboard immunoblotting to perform assessments of serum IgG levels to a panel of 19 species, including established and putative periodontal pathogens and health-associated bacteria, in 5,722 dentate subjects ³40 year old, participants in the third National Health and Nutrition Examination Survey (NHANES III, 1988-94) with available periodontal data. Periodontitis was classified using three alternative definitions based on combinations of pocket depth (PD) and attachment loss (AL), as follows: Definition A: ³2 teeth with interproximal AL ³4mm or ³2 teeth with interproximal PD ³5 mm; Definition B: ³1 sites with concomitant AL ³3mm and PD ³4mm; and Definition C: interproximal AL ³5mm in ³30% of the teeth present. Optimized “elevated” titer thresholds and corresponding sensitivities and specificities for each periodontitis definition were calculated. Titers significantly associated with periodontitis were identified in univariable and multivariable logistic regression models. Parsimonious models were subsequently developed after using age, gender, education, smoking and diabetes status.

Results: In unadjusted models, high titers to Porphyromonas gingivalis were most strongly associated with periodontitis in all three definitions (OR 2.05-2.34, p<0.05). In parsimonious models including additional covariates, high P. gingivalis and low Eubacterium nodatum titers were consistently associated with periodontitis across all definitions. ROC curves for the parsimonious models showed that the area under the curve (C-value) ranged between 0.70 and 0.77.

Conclusion: Serum IgG titers to selected periodontal species, used in combination with demographic and behavioral characteristics, result in moderately accurate predictions of periodontal status in epidemiologic studies. The external validity of these findings must be examined further.

Supported by the American Heart Association (Grant-In-Aid #256205T).