Mixture analysis of air pollution and endocrine disrupting chemicals in relation to Vitamin D concentrations in pregnancy

Background and aim: Over two-thirds of pregnant women in the U.S. have insufficient Vitamin D, which can adversely impact fetal health. Several environmental pollutants may affect Vitamin D concentrations, but they have not been examined as mixtures. Methods: We analyzed data from 336 general population controls from the case-control Early Markers for Autism study of children born in Southern California in 2000-2003. Maternal serum was assayed for Vitamin D and environmental chemicals; eleven polychlorinated biphenyls (PCBs), two organochlorine pesticides, six per- and polyfluoroalkyl substances, and six brominated flame retardants (BFRs) were detected in ≥ 60% of specimens and included in the study. Vitamin D was analyzed as a continuous variable, and concentrations <75 nMol/L were dichotomized as insufficient. Particulate matter <=10 microns (PM10) and <=2.5 microns, nitric oxide, nitrogen dioxide, and ozone data for 30 days preceding specimen collection were downloaded from monitoring stations near maternal addresses. Bayesian Kernel Machine Regression and Bayesian Hierarchical Modeling were used to assess an overall mixture effect of all chemicals and associations accounting for co-exposures, respectively, adjusting for demographic characteristics. Results: In the context of the mixture, PM10, BFRs as a group, and organochlorine trans-nonachlor were associated with lower continuous Vitamin D concentrations, while ozone was associated with higher concentrations. BFRs were also associated with increased odds of insufficient Vitamin D. Higher quantiles of combined exposures were associated with lower mean Vitamin D with wide credible intervals. Several null associations were also found. Conclusions: We found associations of environmental pollutants with both lower and higher Vitamin D, though cumulative exposure was associated with lower concentrations. Some findings differed from non-mixture regressions, highlighting the importance of mixtures approaches for estimating real-world associations. Keywords: Mixtures, Vitamin D, air pollution, PM10, organochlorines, PBDE, PCB, PFAS