Abstract WP234: Seasonal Differences in Sleep and Incident and Recurrent Strokes: Evidence From the Stroke Warning Information and Faster Treatment (SWIFT) Study

Abstract

Introduction: Studies have documented seasonal patterns in stroke outcomes. While results are equivocal most report increased rates of these outcomes in the winter-spring and hypothesize this is explained by ambient temperature. Sleep disturbance, which also varies by season, may also explain this pattern. In this analysis we leverage a unique cohort to test whether incident or recurrent strokes vary by season, and whether this co-varies with sleep disturbance. Methods: Data comes from SWIFT, a stroke preparedness RCT conducted in northern Manhattan. Season of stroke was classified as winter (W; Dec - Feb), spring (Sp; Mar - May), summer (Su; Jun - Aug), fall (F; Sept - Nov). We restricted our analysis to the period Feb 2005 - Jan 2009. Incident strokes were individuals enrolled at baseline without radiological or clinical evidence of prior stroke. Recurrent strokes were captured as part of follow-up protocol. Sleep was measured with the MOS sleep scale, a 12 item questionnaire that produces 8 validated scales. For brevity, we report results using the sleep problem index 2 (9 questions). We assessed seasonal variation in stroke incidence using a one-way goodness of fit chi-square test assuming a distribution based on the number of days per season. The relationship between seasonal variation in stroke and sleep was assessed using ANOVA. Results: A total of 835 incident strokes were enrolled (25% white, 15.5% black, 53.1% Hispanic, 6.4% other; 49.8% female; mean age 63.4±15.6). Incident stroke was not significantly different by season (24.3% W; 22.8% Sp; 28.6% Su; 24.3% F; p=0.09). Mean sleep problems did not vary by season (32.3 W; 30.8 Sp; 31.9 Su; 31.4 F; p=0.86) for incident strokes. A total of 120 first recurrent strokes were captured (20.8% white, 17.5% black, 5% Hispanic, 6.7% other; 49.2% female; mean age 66.0±15.0). Recurrent stroke was significantly different by season (21.7% W; 14.2% Sp; 36.7% Su; 27.5% F; p<0.01). Mean sleep problems did not vary by season (32.4 W; 29.9 Sp; 31.7 Su; 35.1 F; p=0.70) for recurrent strokes. Conclusion: We find no variation in incident strokes but an increased proportion of first recurrent strokes in the summer. Neither incident nor recurrent strokes co-varied with sleep problems. Future studies in non-urban samples are warranted.