Background: Previous studies on environmental pollutant exposure during pregnancy have mostly focused on individual chemical substances or single urine measurements. Thus, our understanding of the potential cumulative or interactive effects of exposure is limited. Objective: We aimed to ascertain the characteristics and predictors of exposure to environmental chemicals over three trimesters among pregnant women.
Methods: We measured the concentrations of 34 chemicals in spot urine samples provided by 745 participants in their early, middle, and late pregnancy. We calculated Spearman correlation coefficients (SCC) between exposure levels of multiple chemicals in each trimester. K-means clustering and principal components analysis (PCA) were applied to classify the populations and reduce data dimensionality. We used generalized linear models (GLM) to confirm predictors of each cluster and principal component.
Results: SCC showed that the correlations of chemical concentrations from the same classes were higher than those among concentrations of different classes. Cluster analysis categorized participants into three clusters, and each cluster represented different chemical concentrations. We restricted the principal components to six, which explained more than 50% of the data variations. Several physiological, socio-demographic factors, and behavior patterns were related to different clusters and principal components.
Conclusion: Distinct exposure patterns and dominant exposure components of multiple environmental chemicals among pregnant women might help research the potential health effects of exposure to chemical mixtures and develop relevant public health interventions.
Authors: Huan Chen, Wenxin Zhang, Yanqiu Zhou, Jiufeng Li, Hongzhi Zhao, Shunqing Xu, Wei Xia, Zongwei Cai, Yuanyuan Li
; Full Source: The Science of the total environment 2020 Sep 3;754:142167. doi: 10.1016/j.scitotenv.2020.142167.