Zonal determinants of healthy aging: An exposomic analysis of determinants of healthy cardiovascular aging in Ohio

BACKGROUND: Emerging evidence supports a role for social, behavioral, and environmental variables in explaining significant disparities in cardiovascular health and aging, beyond conventional risk factors. OBJECTIVES: This study sought to identify and characterize regional "Healthy Aging" zones by identifying census-tract clusters with high rates of healthy coronary artery calcium (CAC) scores as a precison marker of cardiovascular aging in older populations. METHODS: In this prospective cohort study, 31,544 older patients (age ≥ 65) across 1,141 census tracts in Ohio enrolled in a no-charge CAC screening program (CLARIFY registry) were grouped as "healthy" (CAC = 0, n = 7,932) and "non-optimal" aging (CAC > 0, n = 23,612). Cluster analysis was performed using Local Moran's I on census-tracts and adjusted Z-scores of Local Moran's I values were used to rank clusters from most significant "Non-optimal Aging" (low-low Moran's I) to most significant "Healthy Aging" (high-high Moran's I) zones. An eXtreme Gradient Boosting (XGBoost) model followed by SHAP (SHapley Additive exPlanations) analysis was used to determine the importance of 100 socioeconomic/environmental features in determining census-tract adjusted Z-scores. RESULTS: A total of 412 "High-High" and 356 "Low-Low" census-tracts were identified, which formed 4 distinct "Healthy Aging" clusters in northeast Ohio. "Healthy Aging" clusters were associated with less pollution, less noisy environments, lower rates of depression and disability, higher vegetation index, greater healthcare utilization, and greater economic/financial opportunity. CONCLUSIONS: Our analysis implies geospatial clustering of healthy cardiovascular aging and provides evidence of a "healthy" census-tract phenotype, supporting the importance of exposomic factors in determining cardiovascular health.