Abstract P20: Exploring Psychosocial Pathways Between Neighborhood Characteristics and Stroke in Older Adults: the Cardiovascular Health Study

Background: Studies have not investigated whether psychosocial pathways mediate the effect of neighborhood exposures on stroke. Methods: We used data from the Cardiovascular Health Study, a longitudinal population-based cohort study of adults >= 65 years, to construct race-stratified Cox proportional hazard models for time to first ischemic stroke. The key explanatory variable in the baseline model was an index of neighborhood socioeconomic status (NSES), constructed at the census-tract level from six measures of population income, education, employment and wealth. Covariates were individual sociodemographic characteristics (age, gender, marital status, education, and income). In subsequent models, we added psychosocial (depression, social support, social networks, and stressful life events), behavioral (smoking, physical activity, and alcohol), and biologic factors (EKG abnormalities, subclinical cardiovascular disease, hypertension, diabetes, and hyperlipidemia) separately and simultaneously to assess mediation of the effect of NSES on stroke. Results: Of the 3834 participants with no prior stroke, 548 had an incident ischemic stroke over the 11.5-year follow-up. Among the psychosocial factors, only depression was independently associated with higher stroke hazard (Hazard Ratio [HR]=1.26; 95% CI:1.02-1.59). However, addition of psychosocial factors to the baseline model slightly reduced the HRs among residents in the lowest relative to highest NSES quartile from 1.32 (1.01-1.74) to 1.30 (0.98-1.71). Depression ceased to be statistically significant in the model with behavioral and biological factors adjusted. We found no association between NSES and stroke incidence among African Americans. Conclusions: Psychosocial factors played a minimal role in mediating the effect of NSES on stroke incidence among whites.