SummaryType 2 diabetes is a common and well-researched chronic condition, though research in large prospective cohort studies has predominantly focussed on cross-sectional associations of exposures with prevalent diabetes, while research on incident disease (which would limit the risk of reverse causation) in the German National Cohort (NAKO) remains sparse. We therefore aim to leverage the different sources of follow-up data from primary and secondary health care in the NAKO and UK Biobank, to establish an algorithmically-defined incident diabetes outcome, compare participant characteristics across case ascertainment group, and explore associations with conventional and novel markers of body composition, which is considered a key risk factor for type 2 diabetes and other cardiometabolic diseases. The development of a robust and valid type 2 diabetes endpoint would ensure that studies can accurately estimate associations of risk factors with risk of incident type 2 diabetes.

Keywords Diabetes Electronic-health epidemiology

InstitutionsUniversity of Oxford, Universitätsmedizin Greifswald, Helmholtz Zentrum München, Deutsches Forschungszentrum für Gesundheit und Umwelt (GmbH), Institut für Epidemiologie, University of Oxford - UK Biobank, Leibniz Institut für Präventionsforschung und Epidemiologie - BIPS, Otto-von-Guericke-Universität Magdeburg, Medizinische Fakultät, Univeristy of Oxford - UK Biobank, University of Oxford - Oxford Population Health, Deutsches Institut für Ernährungsforschung Potsdam-Rehbrücke, Universitätsklinikum Freiburg, Deutsches Diabetes-Zentrum