ABSTRACT Type 2 diabetes disproportionately affects those with low socioeconomic status (SES), while unfavorable neighborhood factors — such as lack of physical activity resources, limited healthy food options, socioeconomic disadvantage and barriers to health care — often intersect with low individual SES, compounding disparity. This intersectionality of multiple levels of influence (individual, neighborhoods, society) results in highly vulnerable populations at greater risk for developing diabetes, and likely contributes to marked regional variations in diabetes risk. Given that approximately 88 million adults in the U.S. have prediabetes, and most structured diabetes prevention approaches (e.g. lifestyle modification and medications) require significant time and financial investment, efforts that enable health care providers to prioritize the most high-risk patients would optimize benefit. Diabetes risk prediction models can be used to identify individuals at high-risk for progression to diabetes; however, traditional models include clinical parameters with little integration of social factors, ignoring the multiple levels of influence on disease prevention. Thus, integrating social determinants of health (SDoH) data into risk stratification has the potential to identify high-risk individuals based on clinical and social vulnerabilities, facilitating better targeted interventions and reductions in disparities. Importantly, risk stratification approaches that utilize SDoH in the electronic medical record (EMR) may provide an avenue to improve diabetes outcomes and address disparity at the population level. Moreover, understanding how to recruit and engage high-risk, socially vulnerable patients, as well as how to individualize prevention efforts — such as the Diabetes Prevention Program — has potential to improve diabetes outcomes and health equity at the population level. Therefore, the goals of this K01 proposal are to: (1) evaluate the addition of SDoH to a validated diabetes risk prediction model — the cardiometabolic disease staging (CMDS) — to determine improvement of risk classification in two population-based cohorts; (2) determine the prevalence of adults at high-risk for diabetes, both clinically and socially, in the UAB Health System using risk stratification; and (3) identify strategies to engage high-risk, socially vulnerable individuals in diabetes prevention using stakeholder engagement. Conducting this research, in combination with the training and mentoring plan proposed, will help me to obtain skills and experience in health disparities and SDoH measurement; stakeholder engagement and qualitative methods; and diabetes clinical outcome measurement. This award will allow me to develop my independent research path focusing on utilizing social determinants of health (SDoH) data to inform the design of better tailored initiatives for the prevention of cardiometabolic disease. This study will provide the groundwork to inform a future trial to assess the effectiveness of delivering the Diabetes Prevention Program, based on clinical and SDoH factors, to ultimately decrease disparities.

PROJECT NARRATIVE There are marked disparities in type 2 diabetes prevalence among socioeconomically disadvantaged and socially vulnerable populations. Tackling disparities in diabetes incidence requires innovative strategies that address both clinical parameters and underlying social and environmental determinants of health. Incorporating social determinants of health data into risk prediction models to identify those at high risk utilizing data already available in the electronic medical record is a promising avenue to prioritize and tailor local diabetes prevention initiatives to socioeconomically disadvantaged, socially vulnerable individuals, with potential to improve diabetes outcomes and health equity at the population level.