Social isolation and outcomes post stroke

OBJECTIVE: To assess the relationship between social isolation and stroke outcomes in a multiethnic cohort. METHODS: As part of the Northern Manhattan Stroke Study, the authors prospectively followed a cohort of patients with stroke for 5 years. Baseline data including social isolation were collected. At follow-up, the authors documented outcome events as defined by the first occurrence of myocardial infarction (MI), stroke recurrence, or death. Cox hazard models were used to calculate the hazard ratio (HR, 95% CI) for prestroke predictors of post stroke outcomes. RESULTS: The authors followed 655 ischemic stroke cases for a mean of 5 years. The cohort was 55% women; 17% white, 27% African American, 54% Hispanic; mean age 69 +/- 12 years. There were 265 first outcome events. In univariate analysis, coronary artery disease (OR 1.3, 1.0 to 1.7), age > 70 years (OR 1.9, 1.5 to 2.5), atrial fibrillation (AF) (OR 1.8, 1.3 to 2.5), race-ethnicity (white vs Hispanic) (OR 1.7, 1.1 to 2.9), physical inactivity (OR 1.3, 1.1 to 2.6), help at home (OR 1.8, 1.4 to 2.4), and social isolation (OR 1.4, 1.2 to 1.6) were associated with increased risk of an outcome event. No association was seen for hypertension, diabetes, education, sex, insurance, occupation, marital status, or primary care physician. In the multivariable model controlling for age, AF (OR 1.9, 1.5 to 2.5), help at home (OR 1.5, 1.1 to 2.0), and social isolation (OR 1.4, 1.1 to 1.8) predicted outcome events. CONCLUSION: Prestroke social isolation is a predictor of outcome events post stroke. Lack of social support may contribute to poorer outcomes due to poor compliance, depression, and stress.