Abstract P414: Modification of Diet-Metabolite Associations by Race and Sex in the Coronary Artery Risk Development in Young Adults Study

Objective: There is known heterogeneity in the relationship between diet quality and cardiovascular disease by self-reported race and sex, which may underlie inequities in cardiovascular health. Methods: We used data from 2,832 participants in the Coronary Artery Risk Development in Young Adults (CARDIA) study (2005-06, aged 37-55, 57% women, 45% Black). Using fasted blood samples, we obtained ultra-high-performance liquid chromatography high resolution mass spectrometry untargeted metabolomics and derived an a priori diet quality score using a validated diet history questionnaire, based on classification of 46 food groups with potential beneficial (n=20), adverse (n=13), or neutral (n=13) implications for cardiovascular health. We tested effect modification of associations between metabolites and diet quality by race and sex, separately, using multivariable-adjusted linear regression for 7,522 metabolite peaks, accounting for multiple comparisons, adjusted for batch, field center, demographics, lifestyle behaviors, total energy intake, hypertension and diabetes status, medication use, and BMI. We also used race- and sex-stratified multivariate Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA) analyses to examine how metabolites distinguish by diet quality 1 st and 4 th quartiles. We identified differing metabolite pathways by race and sex through biochemical pathway analysis (Mummichog version 1.0.10) from regression and OPLS-DA models. Results: In linear regression, race significantly modified metabolite-diet quality associations for 231 metabolite peaks; sex modified associations for 1 peak (FDR<0.1). OPLS-DA pathway analysis identified cysteine, methionine, histidine, and alpha-linoleic metabolism and terpenoid backbone biosynthesis as significantly associated with diet quality in Black, but not White, race. Linear regression and OPLS-DA pathway analysis found differences (p<0.05) in arginine biosynthesis by sex. OPLS-DA also found aminoacyl-tRNA and arginine biosynthesis, nitrogen, porphyrin, and chlorophyll metabolism were significantly associated with diet quality in men only. Conclusions: Metabolite pathways were statistically different by self-reported race, and to a lesser extent, by sex. Differences in metabolite-diet quality associations may reflect differences in contextual or social variables, and potentially, heterogeneity in metabolism.