Abstract
BACKGROUND: Cognitive impairment after stroke is common and is present in up to 60% of survivors. Stroke severity, indicated by both volume and location, is the most consequential predictor of cognitive impairment, with severe strokes predicting higher chances of cognitive impairment. The current investigation examines the associations of 2 stroke severity ratings and a caregiver-report of poststroke functioning with longitudinal cognitive outcomes. METHODS AND RESULTS: One hundred fifty-seven caregivers and stroke survivor dyads participated in the CARES (Caring for Adults Recovering From the Effects of Stroke) project, an ancillary study of the REGARDS (Reasons for Geographic and Racial Differences in Stroke) national cohort study. The Glasgow Outcome Scale and modified Rankin Scale scores collected at hospitalization discharge were included as 2 primary predictors of cognitive impairment. The number of caregiver-reported problems and impairments at 9 months following stroke were included as a third predictor. Cognition was measured using a biennial telephone battery and included the domains of learning, memory, and executive functioning. Multiple cognitive assessments were analyzed up to 5 years poststroke, controlling for prestroke cognition and demographic variables of the stroke survivor. Separate mixed models showed significant main effects of the Glasgow Outcome Scale (b=0.3380 [95% CI, 0.14-0.5]; P=0.0009), modified Rankin Scale (b=-0.2119 [95% CI, -0.32 to -0.10]; P=0.0002), and caregiver-reported problems (b=-0.0671 [95% CI, -0.09 to -0.04]; P<0.0001) on longitudinal cognitive scores. In a combined model including all 3 predictors, only caregiver-reported problems significantly predicted cognition (b=-0.0480 [95% CI, -0.08 to -0.03]; P<0.0001). CONCLUSIONS: These findings emphasize the importance of caregiver feedback in predicting cognitive consequences of stroke.