Urinary Concentrations of Dialkylphosphate Metabolites of Organophosphate pesticides in the Study of Asian Women and their Offspring’s Development and Environmental Exposures (SAWASDEE)

Background: Measurements of urinary dialkyl phosphate (DAP) metabolites are often used to characterize exposures to organophosphate (OP) insecticides; however, some challenges to using urinary DAP metabolites as an exposure measure exist. OP insecticides have short biological half-lives with measurement in a single urine sample typically only reflecting recent exposure within the last few days. Because of the field staff and participant burden of longitudinal sample collection and the high cost of multiple measurements, typically only one or two urine samples have been used to evaluate OP insecticide exposure during pregnancy, which is unlikely to capture an accurate picture of prenatal exposure. Methods: We recruited pregnant farmworker women in Chom Thong and Fang, two districts of Chiang Mai province in northern Thailand (N = 330) into the Study of Asian Women and their Offspring’s Development and Environmental Exposures (SAWASDEE) from 2017 to 2019. We collected up to 6 serial urine samples per participant during gestation and composited the samples to represent early, mid, and late pregnancy. We measured concentrations of urinary DAP metabolites in the composited urine samples and evaluated the within- and between-participant variability of these levels. We also investigated predictors of OP insecticide exposure. Results: DAP metabolite concentrations in serial composite samples were weakly to moderately correlated. Spearman correlations indicated that composite urine samples were more highly correlated in Fang participants than in Chom Thong participants. The within-person variances (0.064–0.65) exceeded the between-person variances for DETP, DEP, ∑DEAP, DMP, DMTP, ∑DMAP, ∑DAP. The intraclass correlations (ICCs) for the volume-based individual metabolite levels (ng/mL) ranged from 0.10 to 0.66. For ∑DEAP, ∑DMAP, and ∑DAP the ICCs were, 0.47, 0.17, 0.45 respectively. We observed significant differences between participants from Fang compared to those from Chom Thong both in demographic and exposure characteristics. Spearman correlations of composite samples from Fang participants ranged from 0.55 to 0.66 for the ∑DEAP metabolite concentrations in Fang indicating moderate correlation between pregnancy periods. The ICCs were higher for samples from Fang participants, which drove the overall ICCs. Conclusions: Collecting multiple (∼6) urine samples during pregnancy rather than just 1 or 2 improved our ability to accurately assess exposure during the prenatal period. By compositing the samples, we were able to still obtain trimester-specific information on exposure while keeping the analytic costs and laboratory burden low. This analysis also helped to inform how to best conduct future analyses within the SAWASDEE study. We observed two different exposure profiles in participants in which the concentrations and variability in data were highly linked to the residential location of the participants