Concurrent exposures to nonylphenol, bisphenol A, phthalates, and organophosphate pesticides on birth outcomes: A cohort study in Taipei, Taiwan

Abstract

Prenatal exposure to phenols, phthalates (PAEs), and organophosphate (OP) pesticides may increase the risk of abnormal birth outcomes. However, many previous studies have examined exposure to a limited number of chemical classes or exposure profiles limited to a specific stage of pregnancy. This study aims to characterize the concurrent exposure scenario throughout pregnancy by simultaneously monitoring internal doses of several endocrine-disrupting compounds (EDCs), including 2 phenols (nonylphenol (NP) and bisphenol A (BPA)), 9 PAEs, and 6 OP pesticide metabolites and to assess the relationships between concurrent exposure to EDCs and infant birth weight, length, and head and chest circumference. One hundred and sixty two women provided three spot urine samples at approximately 11 and 26weeks gestation and at delivery. We applied multivariable linear regression and ridge regression models to estimate the effects of separate and correlated exposures. Multivariable linear regression models revealed that women with short birth-length infants had significantly higher urinary second-trimester NP levels (50th percentile, 5.03μg/g creatinine) (β=−0.47cm; 95% CI=−0.93 to −0.01). Similarly significant relationships were observed between second-trimester mono-methyl phthalate (MMP) exposure and short birth length, second-trimester ΣPAEs and short birth length, second-trimester ΣPAEs exposure and reduced head and chest circumference, second-trimester diethylphosphate (DEP) exposure and reduced birth weight and length, and second-trimester ΣDEPs exposure and short birth length. Women with urinary BPA above the 75th percentile or ΣPAEs levels above the 50th percentile in the third trimester had infants with significantly reduced head circumference. These observations suggest that the second trimester may be the critical stage of susceptibility for fetal development. In ridge regression models, for which women with fewer measures for exposure to NP, BPA, MMP, ΣPAEs, DEP and ΣDEPs simultaneously were available, no relationships were found with infant size at birth. Additional studies with larger sample sizes are warranted.