Abstract TMP37: Social Network Size at Stroke Onset is Associated With 3 and 6-Month Stroke Outcomes

Abstract

Introduction: Social mechanisms involved in stroke recovery are important and underutilized. For at least three decades, research has shown that social factors such as living alone negatively influence outcomes after stroke. Prior studies relied on social support measures without study of the network social structure. Our study uses higher-resolution social network analysis to examine structure (e.g., size and connectivity) and composition (e.g., proportion kin) in relation to stroke outcomes at 3 and 6 months. Hypothesis: Social network structural traits, size and connectivity, at stroke onset is related to the Patient Recovery Outcome Measurement Information System (PROMIS) physical function score at 3 and 6 months. Methods: In a cohort study, 160 patients with a first-time ischemic stroke were enrolled at stroke hospitalization. Social network characteristics were assessed using a bedside survey interview. The PROMIS physical function score was assessed on the phone at 3 and 6 months after stroke. Unadjusted and adjusted analyses were completed using Spearman Correlation and multivariate linear regression. Results: 144 patients (90%) at 3 months and 120 patients (75%) at 6 months completed the study. The mean age was 62, and NIH stroke scale was 3. Social network size at stroke onset was positively associated with 3- (spearman correlation=0.22, p<0.01) and 6-month (spearman correlation=0.21, p=0.02) PROMIS physical function score. Connectivity measures and the proportion of kin in the network were not associated with the outcome. After control of age, stroke severity, and depression, network size remained strongly associated to physical function at 3 and 6-months, beta=0.74 [0.51,0.97], p<0.01. Conclusions: Of social network traits, network size at stroke onset is independently associated with physical function after stroke. Small networks are an important risk factor for poor recovery, likely equivalent to traditional risk factors.