Seasonality of preterm birth in the Collaborative Perinatal Project: demographic factors.

Marked circannual variation in the incidence of preterm birth, adjusted for the normal seasonal fertility rate, was observed in data from the Collaborative Perinatal Project collected at 12 urban university centers during 1959-1966 (p less than 10(-5). The variation in monthly preterm birth rates ranged from a trough of 64% in May to a peak of 144% in September. This is considerably greater seasonal variation than found in most previous studies, which used differing definitions of prematurity. The seasonal trend was present for maternal age groups 22-26 years (p less than 0.005) and greater than 26 years (p less than 0.005), for Bureau of the Census socioeconomic index 3.6-4.9 (p less than 10(-4) and greater than 4.9 (p less than 0.01), for those living in northern states (p less than 10(-4), for married gravidas (p less than 10(-6), for blacks (p less than 0.05) and whites (p less than 0.001), but not for those less than age 22, those with socioeconomic index less than 3.6, those living in southern states, and those not married. Thus, preterm birth was seasonal in the population studied, and the seasonal factor was more evident in demographic groups less predisposed to deliver preterm. These findings may have implications for the pathogenesis of a portion of premature deliveries, and should be considered in the design of studies related to the pathogenesis of preterm birth.