Diurnal variation in stroke onset in atrial fibrillation.

Diurnal Variation in Stroke Onset in Atrial Fibrillation To the Editor: We read with interest the article by Engström et al1 on cardiac arrhythmias and stroke and agree that apart from atrial fibrillation (AF), little scientific attention has been given to the associations between cardiac arrhythmias and stroke onset. Even in AF, we still need to know much more on the relation to stroke onset, as this arrhythmia is usually present in about 15% to 20% of patients with acute stroke and is associated with a 1.5to 3.0-fold higher mortality than that for stroke patients who are in sinus rhythm.2 Strokes attributed to AF also tend to be more severe, with greater disability, longer hospital stay, and lower rate of patient discharge to own home.3,4 A diurnal variation in the onset of cardiac events and stroke is well recognized. Although the diurnal variation of stroke onset has generally been established in patients who are in sinus rhythm, we are unaware of any published data on diurnal variation of stroke onset in patients with AF. To investigate this further, we assessed the time of the stroke onset in 60 patients (21 men, mean6SD age 76610 years) with first-onset stroke who were in AF on admission to our city center teaching hospital to determine the existence of any diurnal variation. Furthermore, antithrombotic therapy prescribed before admission and after discharge was noted. There was a circadian rhythm of stroke onset among patients with atrial fibrillation, with a significantly higher number of strokes occurring between 6 AM and 6 PM and the least number of strokes occurring between midnight and 6 AM (x test, P50.025; Figure, Table). CT scan reports were available for 37 patients (62%): 34 reported acute cerebral infarcts while 3 were reported to be normal. Two of the 34 cerebral infarcts showed areas of hemorrhagic transformation. None of the available CT scans reported primary hemorrhagic infarcts. Only 44% were taking antithrombotic therapy on admission: 8 were taking warfarin (13%), 16 aspirin (27%) and 2 both aspirin and warfarin (3%). Six patients died in hospital and 54 patients were discharged alive. On discharge, only 19 were taking warfarin (35%), 24 aspirin (44%) and 3 both aspirin and warfarin (6%). Most of the reports on circadian variations suggest a peak in the incidence of events between 6 AM and noon. For example, most strokes occurred between 6 AM and noon,5 with the pattern being observed for both ischemic and hemorrhagic stroke as well as stroke subgroups.5–7 A recent meta-analysis8 reported a 49% increase in stroke of all types between 6 AM and noon, which was a 79% increase over the normalized risk of the other 18 hours of the day; however, there were 29% fewer strokes between midnight and 6 AM, a 35% decrease compared with the other 18 hours of the day. Thus, a significant clustering of stroke was observed between 6 AM and noon, although these data are mostly based on patients in sinus rhythm. Though AF is an independent risk factor for stroke, we are not aware of any studies examining the circadian variation of stroke onset among patients with AF per se. Moreover, there appears to be lack of diurnal variation of hemostatic factors or platelet activation among patients with chronic AF, representing a persistently prothrombotic state associated with this arrhythmia.9 In conclusion, this pilot study demonstrates the existence of a circadian variation in stroke onset among patients with AF, with most strokes between 6 AM and 6 PM. Despite the evidence from randomized trials,10 antithrombotic therapy use in this high-risk group still remains suboptimal.