Phthalate Exposure Across Pregnancy: Can We Use a Single Measure to Stand in for Exposure?

Abstract

Listen

Translate article

BACKGROUND AND AIM: Exposure to non-persistent chemicals remains difficult to assess, as spot samples only reflect exposure in a short time frame. Often researchers will only have a single spot sample to estimate exposure across pregnancy and to estimate associations with outcomes in the infant and child. Using a large nationally representative birth cohort with measures of phthalate metabolites across pregnancy, we evaluate whether these measures are reliable across pregnancy. METHODS: In a cohort of 2,040 nulliparous women, we measured 16 urinary phthalate metabolites up to three times during pregnancy (each trimester). We assessed reliability of the metabolite concentrations during pregnancy in two ways: using the intraclass correlation coefficient and using Spearman correlation coefficients. RESULTS:Over 70% of samples had values less than the limit of detection for MCOH, MHNCH, MINP and MCIOP, and they will not be considered in these calculations. For the other metabolites, the ICCs between adjacent trimesters ranged from values close to zero for DEHP metabolites to moderate values (e.g., 0.3 to 0.6) for MiBP and MBzP. In contrast, Spearman correlations coefficients were expectedly uniformly higher, and ranged from 0.2 to 0.7; MiBP and MBzP had the highest values of the Spearman correlation coefficients. CONCLUSIONS:Spot measure of phthalate metabolites during pregnancy likely do not reflect exposure and should not be used to estimate associations between ‘timing of exposure’ and health outcomes. Even for measures taken in adjacent trimesters, the values of the ICCs were low, possibly reflecting different patterns of product use during pregnancy. Values of Spearman correlation coefficients were somewhat higher (although still in the moderate to good range) and reflect the rank ordering of exposure in these women. We urge caution in interpreting results from studies which have one (or even two) measures of these metabolites. KEYWORDS: Phthalates, Pregnancy outcomes, Environmental epidemiology