Background and purpose: Approximately one fourth of stroke occur during sleep. Despite the clinical and radiological similarities between wake-up stroke (WUS) and non-WUS, the functional outcomes of WUS are largely unknown.
Methods: This retrospective analysis reviewed 2289 consecutive patients with acute ischemic stroke who were admitted between November 2002 and December 2009. We used 3 end-point analytic techniques to evaluate the association between WUS and functional outcomes: dichotomized analysis for "functional dependency" (a discharge modified Rankin Scale [mRS] score ≥2 regardless of initial stroke severity), severity-adjusted responder analysis for "unfavorable outcome" (a discharge mRS ≥1 for an admission National Institutes of Health Stroke Scale score 0 to 7; mRS ≥2 for National Institutes of Health Stroke Scale 8 to 14; or mRS ≥3 for National Institutes of Health Stroke Scale ≥15), and shift analysis for changes in overall distributions of discharge mRS scores.
Results: The initial National Institutes of Health Stroke Scale score of patients with WUS was significantly higher than that of their non-WUS counterparts (median [interquartile range]; 4 [2 to 7] versus 3 [1 to 6]; P<0.01). The dichotomized analysis strategy failed to detect a significant association between WUS and functional dependency at discharge (adjusted OR, 0.99; 95% CI, 0.76 to 1.28). However, the responder analysis showed that patients with WUS were more likely to have "unfavorable outcomes" (adjusted OR, 1.33; 95% CI, 1.02 to 1.72), and the shift analysis also detected significant effect of WUS on the mRS score distributions toward increased dependency (adjusted OR, 1.22; 95% CI, 1.01 to 1.48).
Conclusions: From our study, we documented that WUS was associated with worse short-term outcomes after ischemic stroke. Careful selection of appropriate analytic techniques may help to detect modest associations in observational studies.