Within-mother analysis of seasonal patterns in health at birth

A large literature describes relationships between month of birth, birth weight, and gestation. These relationships are hypothesized to reflect the causal impact of seasonal environmental factors. However, recent work casts doubt on this interpretation by showing that mothers with lower socioeconomic status are more likely to give birth in months that are associated with poorer birth outcomes. Seasonality in the numbers of conceptions in different months can also induce a mechanical correlation between preterm birth and month of birth. This paper analyzes the seasonality of health at birth using a large sample of 647,050 groups of US siblings representing 1,435,213 children. By following the same mother over time, we eliminate differences in fixed maternal characteristics as an explanation for seasonal differences in health at birth. We find a sharp trough in gestation length among babies conceived in May, which corresponds to an increase in prematurity of more than 10%. Birth weight conditional on gestation length, however, is found to be strongly hump-shaped over the year, with 8-9 additional g for summer conceptions. We examine several potential mechanisms for explaining seasonality in birth outcomes that have generally been dismissed in the literature on seasonality in rich countries, notably disease prevalence and nutrition. The May trough in gestation length coincides with a higher influenza prevalence in January and February, when these babies are nearing full term, whereas the hump shape in birth weight is associated with a similar pattern in pregnancy weight gain.