Project Summary The overall goal of the proposed research is to provide evidence for individualizing care for older adults with prediabetes (PD) and diabetes (DM) by investigating their risks and development of CVD and disability. Older adults with PD and DM suffer disproportionality from diabetes morbidity, particularly CVD and also disability, an outcome that really matters to patients. Our study will use longitudinal data to observe CVD and disability development in PD/DM patients over time. These analyses will lead to practical population based methods to stratify people with DM for their individual risks of CVD and disability, helping to individualize care. Our research will test whether CVD and disability development in PD/DM follows several distinct trajectories with different predictors and risks that will differ by age. We hypothesize that CVD explains the risk of disability more in people with PD/DM < 75 than in those older, and that we can develop risk models to better guide clinical decision-making. The specific aims are to: 1), investigate heterogeneity in the risk of developing CVD In adults from middle to old age aging with PD/DM; 2) investigate heterogeneity in their risk of developing disability; and 3) specify risk models for CVD and disability in people aging with PD/DM, including people with diverse ethnicities and the oldest old. To accomplish our goal we will perform secondary data analysis of the Health and Retirement Survey (HRS) a nationally representative health interview survey that has been collecting data for over 20 years. Besides extensive health interview information, the HRS also features biomarkers (HBA1c, total and HDL cholesterol, CRP, cystatin), physical assessment information (blood pressure, weight, waist circumference, walking speed), prescription medications for a subsample of participants, and Medicare claims linkages. The HRS allows us to evaluate key variables longitudinally, to account for ethnicity and cohort changes in people aging with PD/DM, and to consider management and its changes over many years. The HRS has enrolled nearly 6,000 people with diabetes aged 51 to over 100. We will employ sophisticated statistical analyses as important tools to test our ideas about dynamic processes, time-dependent risks and age-cohort differences. We believe this work will provide evidence to guide individualized DM care for older patients and will will lead to risk models for outcomes that include people of advanced age, diverse ethnicities and complex health status.