Metabolomic and genomic expansion of type 2 diabetes subtyping in multiethnic populations

Project Summary/Abstract Type 2 diabetes (T2D) involves derangement in multiple tissues, yet the physiological drivers of disease will vary and are difficult to discern for a given individual. This variability in T2D pathophysiology contributes to heterogeneity in T2D complications and in which treatment strategy will offer optimum benefit. Together, these

pose challenges for clinical decisions on the optimal intervention. We and others have helped disentangle this heterogeneity by advancing subclassification of T2D and identifying subtypes with different clinical risk, genetic susceptibility, and response to interventions. However, multiple clinical gaps remain before T2D subtypes can

be used in routine clinical care, including a lack of racial/ethnic diversity in study populations, understanding of T2D subtypes across the age spectrum, and integration of multiple different data types to provide comprehensive understanding of the physiological drivers of heterogeneity in T2D. This application is in response to RFA-DK-

23-019, which seeks to establish a consortium to bring together investigative teams to improve classification of T2D. We propose to contribute to this consortium and provide data from the Atherosclerosis Risk in Communities Study, Coronary Artery Risk Development in Young Adults Study, Diabetes Heart Study (DHS), African American

DHS, Hispanic Community Health Study/Study of Latinos, Insulin Resistance Atherosclerosis Study (IRAS), IRAS Family Study, Action for Health in Diabetes Trial, Multi-Ethnic Study of Atherosclerosis, and SEARCH for Diabetes in Youth. These 10 populations total 55,766 individuals (25% prevalent T2D, 8% incident T2D) and

reflect the diversity of the US population at risk for and with T2D, spanning ages 10 to 86-years at enrollment, and include large numbers of females (57%), Black (21%), and Hispanic (37%) individuals. We have a multidisciplinary team with expertise in biostatistics, clinical diabetes care, data science, epidemiology, genetics,

metabolomics, and physiology. We have led developments in diabetes subgrouping and novel computational methods to estimate missing data that will be used as a framework for broader application in this proposal. Aim 1 will assemble, harmonize, and integrate clinical, genetic, metabolomic, social determinants of health (SDoH),

and disease outcomes data across these cohorts. Aim 2 will build and cross-validate data-type distinct and aggregate models for classification of T2D subtypes based on clinical, genetic, and metabolomics data. Aim 3 will assess T2D subtypes with risk for cardiovascular disease (and subtypes coronary heart disease,

cerebrovascular disease, and heart failure), chronic kidney disease, retinopathy, and mortality. We will also assess the influence of SDoH on these relationships. Successful completion of this proposal will generate evidence needed to establish physiological drivers of and precise definitions for T2D subtypes. This work furthers

our long-term goal of developing more efficient precision medicine-based prevention and treatment strategies for T2D that maximize benefits and minimize risk for adverse outcomes.