Gestational exposure to PM2.5, NO2, and sex steroid hormones: Identifying critical windows of exposure in the Rochester UPSIDE Cohort.
BACKGROUND: Sex steroid hormones are critical for maintaining pregnancy and optimal fetal development. Air pollutants are potential endocrine disruptors that may disturb sex steroidogenesis during pregnancy, potentially leading to adverse health outcomes. METHODS: In the Environmental influences on Child Health Outcomes Understanding Pregnancy Signals and Infant Development pregnancy cohort (Rochester, NY), sex steroid concentrations were collected at study visits in early-, mid-, and late-pregnancy in 299 participants. Since these visits varied by the gestational age at blood draw, values were imputed at 14, 22, and 30 weeks gestation. Daily NO2 and PM2.5 concentrations were estimated using random forest models, with daily concentrations from each 1-km2 grid containing the subject's residence. Associations between gestational week mean NO2 and PM2.5 concentrations and sex steroid concentrations were examined utilizing distributed lag nonlinear models. RESULTS: Each interquartile range (IQR = 9 ppb) increase in NO2 during weeks 0-5 was associated with higher early-pregnancy total testosterone levels (cumulative β = 0.45 ln[ng/dl]; 95% CI = 0.07, 0.83), while each IQR increase in NO2 during weeks 12-14 was associated with lower early-pregnancy total testosterone levels (cumulative β = -0.27 ln[ng/dl]; 95% CI = -0.53, -0.01). Similar NO2 increases during gestational weeks 0-14 were associated with higher late-pregnancy estradiol concentrations (cumulative β = 0.29 ln[pg/ml]; 95% CI = 0.10, 0.49), while each IQR increase in NO2 concentrations during gestational weeks 22-30 was associated with lower late-pregnancy estradiol concentrations (cumulative β = -0.18 ln[pg/ml]; 95% CI = -0.34, -0.02). No associations with PM2.5 were observed, except for an IQR increase in PM2.5 concentrations (IQR = 4 µg/m3) during gestational weeks 5-11 which was associated with lower late-pregnancy estriol levels (cumulative β = -0.16 ln[ng/ml]; 95% CI = -0.31, -0.00). CONCLUSIONS: Residential NO2 exposure was associated with altered sex steroid hormone concentrations during pregnancy with some indication of potential compensatory mechanisms.
