Exposure to PM₂.₅ during pregnancy and ultrasound parameters of fetal growth in Massachusetts, USA

Abstract

BACKGROUND AND AIM: Prior studies have examined the association between particulate matter less than 2.5 µm (PM₂.₅) and fetal growth with inconsistent results. In this study, we examined the association between PM₂.₅ exposure in the first 16 weeks of pregnancy and fetal growth in a pregnancy cohort using routine ultrasound measurements and spatiotemporally resolved PM₂.₅ in Massachusetts, USA. METHODS: We used prenatal and obstetric data from 8,811 pregnancies that delivered at Beth Israel Deaconess Medical Center from 2011-2016. There were 21,874, 21,788, 21,776, and 21,738 ultrasound measurements for biparietal diameter (BPD), head circumference (HC), femur length (FL), and abdominal circumference (AC), respectively. We used generalized additive mixed models to examine the exposure-response relationship between PM₂.₅ in the first 16 weeks of pregnancy and fetal ultrasound parameters in distinct developmental windows (weeks 14-20, weeks 21-26, weeks 27-32, weeks 33+) adjusted for sociodemographic characteristics, maternal comorbidities, season, and temperature. RESULTS:For all fetal ultrasound parameters, we observed a nonlinear exposure-response relationship in gestational weeks 14-20. Below 7 µg/m³, higher PM₂.₅ exposure was associated with reduced ultrasound measurements, while at higher concentrations, it was either not associated with fetal growth (BPD and AC) or associated with larger ultrasound measurements (HC and FL). After week 27, PM₂.₅ exposure was linearly associated with reduced BPD and AC, but these associations were weaker than those earlier in pregnancy. For example, an increase in PM₂.₅ from 0 to 7 µg/m³ was associated with BPD z-score reductions of 0.3, 0.2, and 0.1 in weeks 14-20, 27-32, and 33+, respectively. CONCLUSIONS:Gestational exposure to PM₂.₅ was associated with impaired fetal growth at levels below the current national standards, where early pregnancy appears to be a particularly sensitive window of development during which a fetus is susceptible. Nonlinearities in the exposure-response relationship in weeks 14-20 could potentially be attributed to live-birth bias. KEYWORDS: air pollution, causal inference, particulate matter, pregnancy outcomes