Infections and systemic disease: Methodological issues for periodontal infection models and an application to type 2 diabetes mellitus risk factor epidemiology

Purpose. To investigate optimality characteristics of periodontal disease exposure definitions and the potentially bi-directional relationship between periodontal disease and type 2 diabetes mellitus (DM).; Methods. The comparative utility of several candidate clinical periodontal disease definitions is determined by assessing correlations between these definitions and: (1) the underlying microbiology of periodontal disease, as well as (2) age and tooth loss. Candidate clinical definitions were based on extent (degree to which sites are affected in the mouth) and severity (relating to depth of pockets or attachment loss in mm). Results from these analyses are used to inform an investigation of the bi-directionality of the periodontal disease-DM relationship. These studies were carried out in two populations: (1) The Oral Infections and Vascular Disease Epidemiology Study and (2) The Study of Health in Pomerania.; Results. Periodontal exposure definitions based on low or no severity cut points, utilizing relative measures of pocket depth had the strongest relationships with the microbiology of periodontal disease. Similarly, low or no severity cut points, utilizing relative measures of attachment loss had the strongest correlations with age and tooth loss. Our data indicate that the relationships between aging, periodontal disease and tooth loss are contextual and can influence appropriate interpretation of various periodontal disease definitions. Finally, we observed an approximate 0.06% absolute increase in five-year hemoglobin A1c across tertiles of periodontal disease (p<0.05). A gender interaction was noted in which men realized a five year change of approximately 0.12% (p<0.05) whereas the relationship was weak and non-significant among women. Participants in the highest textile of periodontal disease had a non-significant 60% increase in the odds of incident diabetes. Participants with uncontrolled diabetes experienced an absolute increase in clinical periodontal disease of ∼8% over five years of follow up when compared to non-diabetic participants.; Conclusion. Definitions of periodontal disease are potentially population dependent while both current and historical definitions with low severity cut points appear to be the most informative. The observed association between periodontal disease and diabetes is suggestive of a bi-directional relationship between the two conditions.