Circadian rhythmicity of stroke onset. Intracerebral and subarachnoid hemorrhage.

BACKGROUND AND PURPOSE Our purpose was to describe and further understand the determinants of the time of onset of parenchymatous intracerebral hemorrhage and subarachnoid hemorrhage in patients enrolled in the Stroke Data Bank. METHODS We analyzed the observed times of onset of intracerebral hemorrhage (n = 237 patients) and subarachnoid hemorrhage (n = 243 patients) compared with expected times of onset if the probability of onset was constant across all time intervals. We also analyzed the role of clinical features (if any) in explaining the findings. RESULTS For intracerebral hemorrhage, 52.5% of patients reported onset times between 0600 hours and 1400 hours, with peak onset between 1000 and 1200 hours (chi 2 = 62.94, df = 11, p less than 0.001). Patients with subarachnoid hemorrhage were more likely to lack a history of hypertension compared with patients who had intracerebral hemorrhage (chi 2 = 23.3, df = 1, p less than 0.001). Patients with subarachnoid hemorrhage were more likely to have more uniform onset time throughout the day (chi 2 = 12.92, df = 7, p = 0.074). However, subarachnoid hemorrhage patients with a history of hypertension were more likely to have peak onset times in mid-to-late morning compared with patients without such a history (chi 2 = 35.25, df = 10, p less than 0.001). The nonuniformity of onset times for intracerebral hemorrhage persisted even if patients with unknown onset times were treated as through their onset times were randomly distributed between 0000 and 0800 hours. Seasonal periodicity and the relation between initial systolic or diastolic blood pressure and time of onset for either type of hemorrhage were not observed. CONCLUSIONS Our data suggest that the time of onset for both intracerebral hemorrhage and subarachnoid hemorrhage patients with a history of hypertension is similar to the diurnal variation in blood pressure.