Abstract T MP69: Do Neighborhood Characteristics Elevate Post-Stroke Mortality?

Abstract

Background: Neighborhood context predicts long-term mortality of both stroke patients and stroke-free general community populations. If the association of neighborhood characteristics and mortality is stronger among stroke survivors than in the general population, it would implicate stroke-care-specific mechanisms. We hypothesized that neighborhood risk factors predict post-stroke mortality, and that these associations are stronger than in stroke free populations. Methods: Health and Retirement Study participants age 50+ free of stroke at baseline (n=15,974) were followed up to 12 years for incident stroke (1,809 events) and mortality (5,578 Deaths). We compared mortality hazard ratios associated with both objective neighborhood measures (census tract poverty, racial composition, residential stability, and foreign born) and self-reported neighborhood social integration (social ties and informal socializing), among stroke-free and post-stroke individuals, using covariate adjusted Cox models. Results: Stroke predicted elevated mortality (HR= 1.98; 95% CI: 1.85, 2.13), as did residence in census tracts in the highest quartile of % non-white (HR= 1.12; 1.22, 1.04). Percent non-white was more strongly related to mortality among stroke patients (HR= 1.32; 1.52, 1.14) than among stroke-free individuals (HR= 1.08; 1.00, 1.18; p for interaction=.013). Low neighborhood social integration also significantly predicted mortality, but the relative effects were similar for stroke survivors and stroke-free individuals. Conclusions: In this large cohort, neighborhood racial composition and social integration predicted survival. Neighborhood racial composition had larger effects on post-stroke survival than on survival in the general population. Post-stroke care available to residents of non-white neighborhoods may be less effective than care available to residents of predominantly white neighborhoods.