Associations Between Climate, Latitude, Fertility and the Decline of the US Sex Ratio at Birth

The US sex ratio at birth (SRB) has declined since 1970, while ambient temperatures have been increasing. This study examines the temporal and spatial variation of the US SRB from 1979–2002 in association with fertility rates and climate variables. Approximately 62.8 million birth records from the National Center for Health Statistics were linked to monthly climate division data and county level socioeconomic variables to evaluate the association of SRB and environmental conditions at or near the time of conception.

Seasonal variation in US SRB is detectable in time series analysis, and is somewhat in phase with variation in fertility. Logistic regression analysis shows that temperature in the month before conception is significantly positively correlated with the likelihood of a male birth when birth order, maternal age, maternal education, plurality, gestation length, race, and Hispanic origin are controlled. This association was significant in models that include all births from 1979–1988, non-Hispanic white births from 1979–1988, and all births in US large counties from 1979–2002. Geographic nonstationarity of US SRB was found in smoothed rate climate division maps for 1979–1988, with higher SRB in latitudes below 40 degrees N, especially in the southeastern US. However, both the overall rates of summer conception and the likelihood of summer male conception are reduced in lower latitudes relative to higher ones.

A logistic regression model was also fit using only non-Hispanic births from US large counties from 1989–2002. In addition to a significant positive association of sex ratio and temperature in the month before conception, deviation from normal monthly temperature during the month of conception, compared to the 1971–2000 baseline temperature, is significantly associated with sex ratio variation. In this population, fewer males were conceived when temperature extremes were significantly above normal; more males were conceived when temperatures were significantly below normal. In both high and low latitude zones over this period, the peak of male conceptions shifted to earlier in the year. Variation in SRB is potentially a sentinel health event and this research suggests that the association between temperature and SRB should be integral to any study of SRB variation across large geographic areas or long time periods.