DOI: 10.1161/circ.148.suppl_1.14510 ISSN: 0009-7322

Abstract 14510: Precision Phenotyping of Social Determinants of Health and Risk of Coronary Disease, Heart Failure, and Stroke: The Jackson Heart Study

Rani Zierath, Victoria Lamberson, Yimin Yang, Brian Claggett, Michael E Hall, Tanya M Spruill, James S Floyd, Michelle C Odden, Mario Sims, Amil M Shah
  • Physiology (medical)
  • Cardiology and Cardiovascular Medicine

Introduction: While social determinants of health (SDOH) are associated with cardiovascular disease (CVD) risk, limited data exist regarding their interrelated impact on distinct CVD events.

Methods: Among 4,164 Black adults in the Jackson Heart Study (JHS) free of prevalent coronary disease (CHD), heart failure (HF), and stroke at baseline (2000-2004), we used consensus clustering - an unsupervised class discovery and validation approach - to identify subgroups of participants based on 9 SDOH measures reflecting the Healthy People 2030 framework: income, education, discrimination, food insecurity, healthy food stores within 1 mile, healthcare access, neighborhood violence, neighborhood disadvantage, and air pollution (PM2.5). We used multivariable Cox proportional hazards models to relate cluster membership to risk of incident CHD, HF, and stroke adjusting for age, sex, hypertension, diabetes, obesity, and chronic kidney disease.

Results: Mean age was 55 ± 13 years and 63% were women. We identified 4 participant clusters with differential exposure to SDOH: Cluster 1 (n=1359) had the most favorable SDOH profile with highest income and lowest neighborhood disadvantage and violence; Cluster 2 (n=915) had highest PM 2.5 and lowest frequency of food stores; Cluster 3 (n=998) had lowest education level and highest neighborhood disadvantage and violence; and Cluster 4 (n=820) had lowest income and healthcare access and highest prevalence of food insecurity. Over a median of 11 [IQR 10 - 11] years, 255 participants developed HF, 159 CHD, and 167 stroke. Compared to Cluster 1, Clusters 3 and 4 were associated with higher risk of incident HF while Clusters 2 and 3 were associated with higher risk of stroke ( Figure ). Cluster membership was not associated with risk of CHD.

Conclusions: Multiparametric SDOH assessment identifies unique clusters of individuals differentially burdened by individual-level social determinants and at heightened risk for distinct CVD events.

More from our Archive