Abstract WP520: The Role of Neighborhood Deprivation in Stroke Risk in Two Countries: France and the United States

Abstract

Background and Aims: The role of socioeconomic status in stroke has been difficult to study due to differences at both the individual and neighborhood level. We therefore sought to examine the role of neighborhood socioeconomic status on risk of stroke in two countries: the United States (US) and France. Methods: We used the REasons for Geographic and Racial Differences in Stroke (REGARDS) study in the US and the 3-City (3C) study in France. In both studies, we used a factor analysis to combine various socioeconomic indicators collected by the governments of the two countries (in France IRIS and in the US the census) into a score indicating the level of neighborhood disadvantage. We limited the analysis to those free from stroke at baseline and used combined outcome of ischemic or hemorrhagic stroke. Cox models were used to model the association between neighborhood deprivation and risk of stroke; all models were adjusted for components of the Framingham Stroke Risk Score (REGARDS also adjusted for race), dietary pattern, individual income and education. Results: There were 802 strokes in REGARDS in 17996 participants and 326 strokes in 3C in 7647 participants. In both studies neighborhood deprivation was associated with lower individual income and educational attainment as well as higher BMI and systolic blood pressure. Only in the REGARDS study was neighborhood deprivation associated with greater risk of stroke. In REGARDS, comparing the 5 th quintile to the 1 st quintile the hazard ratio (HR) and 95% CI was 1.36 (1.09, 1.72). In the 3C study comparing the 5 th quintile to the 1 st quintile, the HR (95% CI) was 1.04 (0.73, 1.47). Conclusion: In the REGARDS study neighborhood deprivation was associated with risk of stroke but these findings were not seen in the 3C study. Future work to further evaluate the role of diet and individual socio-economic status in mediating the findings observed in France could help to better explain the differences in these two studies.