Abstract 07: Spatial Modeling of Social Determinants of Hypertension Within the RESTORE Network

• Published in: Circulation (New York, N.Y.) Vol. 149; no. Suppl_1
• Main Authors: Teshome, Biruk, Parikh, Karan, Ibukun, Folahan A, Wheeler-Martin, Katherine, Dennison Himmelfarb, Cheryl R, Cooper, Lisa A, Commodore-Mensah, Yvonne, Ogungbe, Oluwabunmi V
• Format: Journal Article
• Language: English
• Published: 19-03-2024
• ISSN: 0009-7322; 1524-4539
• DOI: 10.1161/circ.149.suppl_1.07

Abstract only
Introduction:
Multiple environmental and neighborhood factors impact hypertension (HTN) prevalence. Applying spatial methods to explore drivers of HTN risks can reveal complex social and ecologic drivers.
Objectives:
To examine ecological associations between social determinants of health (SDoH), space, and HTN across the American Heart Association-funded RESTORE (Add
RE
ssing
S
ocial
D
eterminants
TO
p
R
event Hypertension) Network.
Methods:
We used City Health Dashboard data, which contains city and county-level aggregate 2019 data for over 950 US cities. We included populated census tracts and counties in RESTORE Network states (Alabama, Maryland, Massachusetts, Michigan, and New York). Primary outcome was HTN prevalence: proportion of adults ≥18 years with diagnosed HTN. Covariates were indicators of clinical care, health behaviors, health outcomes, and physical environment. Using Bayesian hierarchical Besag-York-Mollié (BYM) models with Gaussian distribution by integrated nested Laplace approximation, we estimated associations between social determinants and HTN, accounting for spatial autocorrelation.
Results:
Of the 5 states assessed, 4828 tracts were in New York, 2745 tracts in Michigan, 1175 tracts in Alabama, 1463 tracts in Massachusetts and 1388 tracts in Maryland, totaling 11,599 census tracts. The observed spatial variation in the model was 91% in New York, 83% in Michigan, 78% in Alabama, 79% in Massachusetts, and 74% in Maryland. After adjusting for known HTN risk factors at the tract level, results showed a moderate-high spatial random effect, i.e., residual HTN prevalence remained in several tracts in the 5 states. Patterns of variation of this spatial random effect also differed between states (
Figure
).
Conclusions:
These findings can be used to identify areas that have higher rates of HTN after accounting for other SDoH. Targeted local health policies may reduce the burden of HTN across the 5 states and advance cardiovascular health equity.