Exposure to phenols during pregnancy and the first year of life in a new type of couple-child cohort relying on repeated urine biospecimens

Distribution of the phenol urinary concentrations adjusted for specific gravity in the SEPAGES cohort during pregnancy (periods second (NT2 = 477) and third trimesters ((NT3 = 456), and infancy (2 months (NM2 = 152) and 1 year (NY1 = 100). Compounds with less than 4% detection are not shown (bisphenols F, B, AF and triclocarban). For the bisphenol A infant samples (2 months and 1 year) the concentrations of the conjugated form instead of the total form are presented as external contamination was suspected. Abbreviations: LOD = limit of detection, LOQ = limit of quantification. [Display omitted] •Phenol concentrations were assessed accurately through the collection of repeated samples.•Pregnancy and early infancy, sensitive yet rarely analyzed periods were studied.•High detection rates of bisphenol A, parabens and triclosan in mothers and infants.•Concentrations for several phenols were higher at 12 months than during pregnancy.•Cashiers had higher bisphenol S but not bisphenol A levels than unemployed women. Parabens, bisphenol A and triclosan have been forbidden or restricted in specific types of consumer goods in Europe and France. Limited biomonitoring data are available in France since the implementation of these regulations, and exposure data on infants is scarce worldwide. Understanding the predictors of phenol urinary concentrations will help identify potential targets for prevention. We described levels, variability and predictors of exposure to 12 phenols in pregnant women and infants recruited between 2014 and 2017 in a French couple-child cohort. Among 479 pregnant women and 150 of their infants, we studied phenol urinary concentrations in within-subject, within-period pools of repeated urine samples collected during the second and third trimesters of pregnancy (up to 42 samples per woman), at 2 months and 12 months (up to 14 samples per infant). Time trends and associations with demographic, protocol, occupational and behavioral factors were studied using interval censored models to accommodate for undetected and unquantified urine concentrations. Detection rates were above 90% for bisphenol A, ethylparaben, methylparaben, benzophenone-3 and triclosan and below 5% for bisphenol AF, B, F and triclocarban. Median levels of bisphenol A, bisphenol S, methylparaben, ethylparaben and propylparaben at 12 months were similar or higher than during pregnancy. For pregnant women all phenols but benzophenone-3 and bisphenol S showed a linear decrease betwe