作者: Hui Hu , Jinying Zhao , David A. Savitz , Mattia Prosperi , Yi Zheng
DOI: 10.1016/J.ENVINT.2020.105797
关键词: Biology 、 Exposome 、 Pregnancy 、 Social environment 、 Replication (statistics) 、 Logistic regression 、 Multivariable calculus 、 Elastic net regularization 、 Demography 、 Bonferroni correction
摘要: Abstract It is widely recognized that exogenous factors play an important role in the development of hypertensive disorders pregnancy (HDP). However, only a few external environmental have been studied, often separately, with no attempt to examine totality environment, or exposome. We conducted exposome-wide association study (ExWAS) using Florida Vital Statistics Birth Records including 819,399 women live births 2010–2013. A total 5784 characterizing women’s surrounding natural, built, and social environment during from 10 data sources were collected, harmonized, integrated, spatiotemporally linked based on periods 250 m buffers around their geocoded residential addresses. random 50:50 split divided into discovery replication sets, 3-phase procedure was used. In phase 1, associations between HDP individual examined, Bonferroni adjustment performed. 2, elastic net model used perform variable selection among significant variables 1. 3, multivariable logistic regression all selected by fitted. Variables both sets retained. Among 528 490 identified Phase 232 224 67 48 remained statistically 3 respectively. 12 air toxicants (e.g., 2,2,4-trimethylpentane), meteorological omega vertical velocity at 125mb pressure level), neighborhood crime safety burglary rate), sociodemographic status urbanization). This first large exposome HDP. confirmed some previously reported generated unexpected predictors within may warrant more focused evaluation.