Abstract Body (Do not enter title and authors here): Background: Stroke is a leading cause of death in the US and is associated with long term disability. Social risk factors such as access to health care, living in poverty, and economic mobility confer significant risk for stroke, however limited work has examined the association between key social risk factors and prevalence of stroke at the population level. This study examined the association between multiple social risk factors and prevalence of stroke in the US population.

Methods: Census tract level data was combined to create a final analytic dataset of 11,457 census tracts across 157 counties, within 38 states. Prevalence of stroke in each census tract was based on CDC PLACES data. Social risk factors measured at census tract levels included incarceration (% incarcerated), poverty (% below poverty line), housing (average rent), education (% with college degree), employment (% employed), job environment (no. high paying jobs within 5 miles), economic mobility (probability of reaching top 1% income, and % born outside US), and healthcare access (% uninsured, and % with annual check-up). Regression models using standardized beta to allow comparison between social risks were run in Stata v17, controlling for population of each census tract.

Results: Mean prevalence of stroke was 3.6% across census tracts, ranging from 0.3% to 14% in individual census tracts. In the standardized regression model, health care access measured by annual check-ups (0.73, p<0.001) and insurance status (0.33, p<0.001), poverty (0.12, p<0.001), incarceration (0.07, p<0.001), economic mobility (0.04, p<0.001), and job environment (0.02, p<0.001) were associated with higher prevalence of stroke in a census tract. Economic mobility (-0.14, p<0.001), education (-0.10, se 0.07, p<0.001), employment (-0.08, p<0.001), and housing (-0.05, p<0.001) was associated with lower prevalence of stroke in a census tract.

Conclusions: Multiple social risk factors were associated with the prevalence of stroke in a census tract. The strongest associations were with health care access, economic mobility, and poverty, suggesting the importance of targeting both health system-based and economic empowerment-based strategies simultaneously to address geographic variation in stroke prevalence.