Aung Max T, Yu Youfei, Ferguson Kelly K, Cantonwine David E, Zeng Lixia, McElrath Thomas F, Pennathur Subramaniam, Mukherjee Bhramar, Meeker John D
Department of Biostatistics, University of Michigan (U-M) School of Public Health, Ann Arbor, Michigan, USA.
Epidemiology Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, North Carolina, USA.
Environ Health Perspect. 2021 Mar;129(3):37007. doi: 10.1289/EHP7396. Epub 2021 Mar 24.
Humans are exposed to mixtures of toxicants that can impact several biological pathways. We investigated the associations between multiple classes of toxicants and an extensive panel of biomarkers indicative of lipid metabolism, inflammation, oxidative stress, and angiogenesis.
We conducted a cross-sectional study of 173 participants (median 26 wk gestation) from the LIFECODES birth cohort. We measured exposure analytes of multiple toxicant classes [metals, phthalates, phenols, and polycyclic aromatic hydrocarbons (PAHs)] in urine samples. We also measured endogenous biomarkers (eicosanoids, cytokines, angiogenic markers, and oxidative stress markers) in either plasma or urine. We estimated pair-wise associations between exposure analytes and endogenous biomarkers using multiple linear regression after adjusting for covariates. We used adaptive elastic net regression, hierarchical Bayesian kernel machine regression, and sparse-group LASSO regression to evaluate toxicant mixtures associated with individual endogenous biomarkers.
After false-discovery adjustment (), single-pollutant models yielded 19 endogenous biomarker signals associated with phthalates, 13 with phenols, 17 with PAHs, and 18 with trace metals. Notably, adaptive elastic net revealed that phthalate metabolites were selected for several positive signals with the cyclooxygenase (), cytochrome p450 (), and lipoxygenase () pathways. Conversely, the toxicant classes that exhibited the greatest number of negative signals overall in adaptive elastic net were phenols () and metals ().
This study characterizes cross-sectional endogenous biomarker signatures associated with individual and mixtures of prenatal toxicant exposures. These results can help inform the prioritization of specific pairs or clusters of endogenous biomarkers and exposure analytes for investigating health outcomes. https://doi.org/10.1289/EHP7396.
人类暴露于可能影响多种生物途径的有毒物质混合物中。我们研究了多类有毒物质与一系列广泛的生物标志物之间的关联，这些生物标志物可指示脂质代谢、炎症、氧化应激和血管生成。
我们对来自LIFECODES出生队列的173名参与者（妊娠中位数为26周）进行了横断面研究。我们测量了尿液样本中多类有毒物质的暴露分析物[金属、邻苯二甲酸盐、酚类和多环芳烃（PAHs）]。我们还测量了血浆或尿液中的内源性生物标志物（类花生酸、细胞因子、血管生成标志物和氧化应激标志物）。在调整协变量后，我们使用多元线性回归估计暴露分析物与内源性生物标志物之间的成对关联。我们使用自适应弹性网络回归、分层贝叶斯核机器回归和稀疏组套索回归来评估与个体内源性生物标志物相关的有毒物质混合物。
在进行错误发现调整后，单污染物模型产生了19个与邻苯二甲酸盐相关的内源性生物标志物信号、13个与酚类相关的信号、17个与多环芳烃相关的信号和18个与痕量金属相关的信号。值得注意的是，自适应弹性网络显示，邻苯二甲酸酯代谢物被选为与环氧化酶（）、细胞色素p450（）和脂氧合酶（）途径相关的几个阳性信号。相反，在自适应弹性网络中总体显示出最多阴性信号的有毒物质类别是酚类（）和金属（）。
本研究描述了与产前有毒物质暴露的个体及混合物相关的横断面内源性生物标志物特征。这些结果有助于为调查健康结果时确定特定的内源性生物标志物对或簇以及暴露分析物的优先级提供信息。https://doi.org/10.1289/EHP7396 。
