To the Editor.
—The article by Marsh et al1 in the November 1990 issue of the Archives and the letter to the editor by Alberts2 and the reply by Marsh3 in the November 1990 and August 1991 issues of the Archives, respectively, were of much concern to me. I realize that to determine the time of stroke onset is difficult. However certain sociotraditional factors may aid in resolving this problem.In a study that I undertook in Haydarpasa Numune (State) Hospital, Istanbul, Turkey, the time intervals of infarction occurrence and percentages of patients (N=110) involved were as follows: midnight to 6 AM, 13.1% of patients; 6 AM to noon, 34.0%; noon to 6 PM, 38.4%; and 6 PM to midnight, 14.2% of patients. In the majority of the patients, the infarction had occurred between 6AM and 6 PM, and there was no peak regarding any specific time
INCE B. Circadian Variation in Stroke. Arch Neurol. 1992;49(9):900. doi:10.1001/archneur.1992.00530330018006