Variability of urinary phenols and phthalate metabolites in school-age children of 5 European countries

Abstract

Introduction: Improvement in exposure assessment and reduction of measurement error remains an important challenge in studies evaluating the potential health effects of non-persistent pollutants. We assessed the between-day and -season variability of 7 phenols (4 parabens, bisphenol-A, benzophenone-3, and triclosan) and 8 phthalate metabolites (4 di-2-ethylhexyl phthalate metabolites, mono-benzyl phthalate and 3 low molecular weight phthalate metabolites) in school-age children. Methods: We collected all the first morning and bedtime voids during 1 week in 2 different seasons of 152 children aged 6-9 years of age from 5 European birth cohorts. We analyzed non-persistent pollutants in 4 consecutive days’ pools (first morning + bedtime) and in the 2 weekly pools. All compounds were analyzed using liquid chromatography–mass spectrometry with limits of quantification ranging from 0.2 to 2.0ng/mL. We calculated intraclass-correlation coefficient (ICC) of the 4 consecutive days to assess between-day variability and of the 2 weekly pools to assess between-season variability. Results: Geometric mean concentrations ranged from 0.14ng/mL for n-butyl paraben to 61.3ng/mL for the low molecular mono-ethyl phthalate. Among phenols, bisphenol-A presented the lowest ICCs (between-day: 0.52, between-season: 0.18) whereas benzophenone-3 presented the highest ICCs (between-day: 0.87, between-season: 0.57). Among phthalates, the ICCs for between-day variability ranged from 0.57 for mono-2-ethyl-5-hydroxyhexyl phthalate (MEHHP) to 0.67 for mono-iso-butyl phthalate (MiBP) and for seasonal variability from 0.51 for monoethyl phthalate (MEP) to 0.73 for mono-2-ethylhexyl phthalate (MEHP). Conclusions: This study suggests that phenols, except benzophenone-3, have higher between-day and seasonal variability than phthalate metabolites. Results from this study will serve to characterize the degree of exposure misclassification when using fewer biomarker measurements in the larger cohorts.