The association between PM2.5 and stillbirth and infant mortality in 1940 in the US

Abstract

IntroductionFine particulate matter (PM2.5) has been associated with adverse birth outcomes, including stillbirth and infant mortality, in studies conducted mostly in recent decades. Few studies have investigated the effect of PM2.5 on adverse birth outcomes prior to 1970, when air pollution profile was different and was mostly attributable to coal and petroleum-based sources. We leveraged modeled exposure data to estimate the association between PM2.5, traffic-related nitrogen oxide (NOx), and fossil fuel power plants on stillbirth and infant mortality in 1940 in the US. Methods We used ecological cross-sectional design with county as the unit in the contiguous US. Data on exposure in 1940 were obtained from our collaborators at George Mason University, and data on outcome and covariates were obtained from the 1940 census. We created separate models for stillbirth and infant mortality, and adjusted for county-level female education, urban residence, and geographic region. We included an interaction term for proportion non-White population to explore potential effect modification by area-level racial composition. Results In 1940, the median stillbirth rate was 27 per 1000 births, and the median infant mortality rate was 45 per 1000 live births. There was 73% higher stillbirth rate (RR = 1.73, 95% CI: 1.68, 1.79) for 1 IQR (3.94 µg/m3) higher PM2.5 level in counties with mean non-White population, which diminished by 4% (RR = 0.96, 95% CI: 0.94, 0.98) for every 10% higher non-White population. Similarly, the effects estimates for traffic-related NOx on stillbirth were 1.003 (95% CI: 1.003, 1.004) with an effect modification of 0.997 (95% CI: 0.997, 0.998) for every 10% higher non-White population, and the effect estimates for fossil fuel power plants on stillbirth were 1.38 (95% CI: 1.30, 1.45) with no effect modification by racial composition (RR: 1.01, 95% CI: 0.97, 1.06). There was 39% higher infant mortality rate (RR = 1.39, 95% CI: 1.35, 1.44) for 1 IQR higher PM2.5 level in counties with mean non-White population, which diminished by 11% (RR = 0.89, 95% CI: 0.87, 0.91) for every 10% higher non-White population. Similarly, the effect estimates for traffic-related NOx on infant mortality were 1.003 (95% CI: 1.003, 1.004) with an effect modification of 0.998 (95% CI: 0.998, 0.999) for every 10% higher non-White population, and the effect estimates for fossil fuel power plants on infant mortality were 1.36 (95% CI: 1.28, 1.43) with an effect modification of 1.08 (95% CI: 1.03, 1.14) for every 10% higher non-White population. Conclusion Total PM2.5 was strongly associated with stillbirth and infant mortality in the US in 1940. The two major sources of PM2.5 in that period, fossil fuel power plants and transport-related NOx emissions, respectively, had strong and weak association with the outcomes. The positive association between PM2.5 and adverse birth outcomes observed in 1940 is consistent with the findings from studies conducted in recent decades. In addition, area-level racial composition appeared to modify the exposure-outcome relationship, although the direction of the effect was opposite to what the current literature suggests.