This proposal is for a Physician Scientist Award to support research training for a clinician in the basic science discipline of biostatistics as applied to research problems in periodontology. Training and experience will be supervised by a primary sponsor who is an established biostatistician and a secondary sponsor, established in research in periodontology. Phase I, anticipated to be two years in duration, includes didactic course work in biostatistics, training in methods of measuring periodontal diseases, and new analyses of existent data representing laboratory and clinical findings from a range of periodontal disease forms and health. Data include clinical assessments of plaque levels, gingival inflammation, suppuration, attachment levels, probeable depths, and gingival bleeding, and laboratory assessments of neutrophil chemotaxis, spontaneous peripheral lymphocyte proliferation in vitro, serum antibody levels to a panel of oral microorganisms, presence and relative proportion of individual bacterial species in periodontal microfloras, and serum C-reactive protein. Cluster and multivariate discriminant analyses will be performed to define a) two putatively separate disease forms, or b) a putative indicator of disease activity, based on combinations of data other than those that were used to establish semi-arbitrary subject-disease groups prior to collection of the data. In Phase II, a longitudinal study will be conducted to test the hypothesis a) that the two putative disease forms behave differently over time or b) that the putative indicator of disease activity accurately predicts that disease will progress.
