Is there any circadian variation consequence on acute case fatality of stroke? Takashima Stroke Registry, Japan (1990-2003)

Abstract

Circadian periodicity in the onset of stroke has been reported. However, it is unclear whether this variation affects the acute stroke case fatality. Time of the day variation in stroke case fatality was examined using population-based stroke registration data. Stroke event data were acquired from the Takashima Stroke Registry, which covers a stable population of approximately 55,000 in Takashima County in central Japan. During the period of 1990-2003, there were 1080 (549 men and 531 women) cases with classifiable stroke onset time. Stroke incidence was categorized as occurring at night (midnight-6 a.m.), morning (6 a.m.-noon), afternoon (noon-6 p.m.), and evening (6 p.m.-midnight). The 28-day case fatality rates and 95% confidence intervals (95% CI) were calculated by gender, age, and stroke subtype across the time blocks. After adjusting for gender, age at onset, and stroke severity at onset, the hazard ratios for fatal strokes in evening, night, and morning were calculated, with afternoon serving as the reference. For all strokes, the 28-day case fatality rate was 23.3% (95% CI:19.4-27.6) for morning onset, 16.9% (95% CI:13.1-21.6) for afternoon onset, 18.3% (95% CI:13.6-24.1) for evening onset, and 21.0% (95% CI:15.0-28.5) for the night onset stroke. The case fatality for strokes during the morning was higher than the case fatality for strokes during afternoon. This fatality risk excess for morning strokes persisted even after adjusting for age, gender, and stroke severity on onset in multivariate analysis. In the examination of circadian variation of stroke case fatality, 28-day case fatality rate tended to be higher for the morning strokes.