Abstract
To characterize temporal variations of live births in Japan, we analyzed data on the 1,203,147 births of 1998. In hospitals, with 20+ beds, the daily average of live births was significantly lower at weekends and national holidays (mean=1,433, SD = 100) than on weekdays (mean = 1,957, SD = 126). Hourly distributions of live births showed a single sharp peak at 1:00-2:59 pm on weekdays with a small peak at an earlier hour on Saturdays, Sundays and national holidays. The results in clinics, with no bed or less than 20 beds, were similar to those in hospitals except on Saturdays. The difference in the daily average of live births between Saturdays and weekdays was smaller in clinics than that found in hospitals, and hourly distributions on Saturdays resembled those of weekdays but not Sundays or national holidays. Maternity homes showed no differences in the mean number of daily live births over the days of the week including national holidays, and no clear peak of percentage distributions of hourly live births on each day of the week. The present study suggests that the weekly and hourly variations observed in hospitals and clinics are not due to a biological rhythm of labor, but to obstetric intervention in the timing of delivery, either through induction of labor or elective cesarean section.