Abstract TH913: Serum Metabolomics and the Association between Social Determinants of Health and Heart Failure: The Atherosclerosis Risk in Communities (ARIC) Study

Introduction: Adverse social determinants of health (SDOH) are associated with greater heart failure (HF) risk, but data regarding underlying biologic mechanisms are limited. Methods: Among 3,321 HF-free Atherosclerosis Risk in Communities (ARIC) study participants at baseline (1987-1989), we used exploratory factor analysis to derive a SDOH-based latent factor comprised of income, education level, and area deprivation index (ADI), and assessed its association with incident HF using Cox proportional hazards regression. Due to differential distributions of SDOH between race groups, we stratified this analysis by self-reported race. Linear regression was used to relate 334 metabolites to the SDOH factor in race strata adjusting for demographics (age, sex) and batch at false discovery rate (FDR) p<0.05. We then assessed the association between SDOH-associated metabolites with incident HF at FDR p<0.05 using Cox regression adjusting for demographics, comorbidities (BMI, smoking, hypertension, diabetes, chronic kidney disease (CKD), and coronary artery disease (CAD)), and batch. We performed causal mediation analysis for SDOH and HF-associated metabolites in both race strata using FDR p<0.05 to define significant mediation effects. Results: Mean age was 54±6 years, 59% were female, and 60% identified as Black race. Over 20 [IQR 16-20] years, 201 White and 308 Black participants developed HF. The SDOH latent factor was associated with incident HF in Black (HR 1.47 [95% CI 1.26-1.72] and White (HR 1.49 [95% CI 1.27-1.76]) participants. In Black participants, the SDOH latent factor was cross-sectionally associated with 92 metabolites, 28 also associated with incident HF. In White participants, 87 metabolites associated with the SDOH factor, 8 of which associated with incident HF. We identified four metabolites including pseudouridine, a marker of RNA degradation, two γ-glutamyl amino acids, and 1-lineoleoylglycerophosphocholine (18:2n6), a marker of phospholipid remodeling, which mediated the association between the SDOH factor and HF in both strata. Proportion of the observed association that was mediated by these metabolites ranged from 2-10%. Conclusions: Large-scale metabolomic profiling identified metabolites linking adverse SDOH and HF risk reproducibly across race groups, suggesting perturbations in phospholipid remodeling and RNA turnover processes. These findings strengthen our understanding of the biologic pathways connecting SDOH to CV risk.