Air Pollution and Preterm Birth: Do Air Pollution Changes over Time Influence Risk in Consecutive Pregnancies among Low-Risk Nulliparous Women?

Abstract

Both air pollution, generally decreasing in the U.S., and having a prior preterm birth (PTB) are risk factors for subsequent pregnancy PTB. To investigate whether air pollution exposures would have a potential differential effect on PTB risk in the first and second pregnancy among low-risk women, we estimated exposures based on modified Community Multiscale Air Quality models linked to the NICHD Consecutive Pregnancy Study. Electronic medical records for consecutive delivery admissions were available for 27,137 nulliparous women with singleton births in 20 Utah-based hospitals between 2002-10. We categorized whole pregnancy exposures based on percentiles as high (>75), moderate (25-75) and low (<25). Modified Poisson regression with generalized estimating equations estimated PTB risk in the second pregnancy associated with persistent high and moderate exposure, and increasing or decreasing exposure, compared to persistent low exposure. Analyses were adjusted for prior PTB, interpregnancy interval and other demographic and clinical characteristics. Pollution levels generally decreased over time, reducing the number of women with high second pregnancy exposure. Second pregnancy PTB risk was increased when exposure increased for sulfur dioxide (34%), ozone (80%), nitrogen dioxides (NO2; 38%), and carbon monoxide (CO; 33%) versus stayed consistently low. Similar trends were observed for consistently high versus consistently low exposure. We also observed significant interaction by prior PTB status. Women with no prior PTB had significantly higher risk associated with increasing levels of several pollutants as compared to women with prior PTB, including NO2 (increase of 63% vs. decrease of 7%, p-interaction=0.004) and CO (increase of 57% vs. no change, p-interaction=0.031). These findings suggest area-level changes in air pollution exposure have important consequences in repeated pregnancies even among low risk women at moderate levels of exposure in the U.S.