Metabolic effects of environmental pollutants in pregnant women -an exposome approach

Abstract

Introduction: Environmental pollutants are suspected of adversely affecting child development through, among other mechanisms, disruption of metabolic processes. Few studies have investigated the impact of combined exposures to environmental pollutants on chemical fingerprints that cellular processes leave behind among human population-based samples, with pregnant women representing a particularly important subgroup. Aims: To assess the impact of the exposome, measured by multiple pollutant biomarkers, on the metabolome, measured by 1H NMR urinary metabolic profiling among a large cohort of pregnant women. Methods: Women were enrolled into the INMA birth cohort at the 1st trimester routine antenatal visit in Sabadell, Spain during 2004 -2007. Biomarkers of 34 exposures were measured in blood (organochlorine pesticides, PCBs, PFAS, mercury) and twice in urine (metals, phthalates, BPA) collected at 12 and 32 weeks of pregnancy. 1H NMR metabolic profiles were acquired in the same urine samples (N = 394 at 12 weeks and 469 at 32 weeks). Untargeted exposome-metabolome-wide association were investigated by Spearman and partial correlation in each trimester accounting for confounding factors such as demographic and lifestyle behaviours. Results: 44 exposure-metabolite associations passed correction for effective number of tests in the 1st trimester. Among the strongest of these associations, replicated in the 3rd trimester, were between arsenic and trimethylamine-oxide, a human-microbial co-metabolite (r = 0.39, p= 6 x 10-15) and dimethylamine (r=0.25, p= 1 x 10-6). A novel biomarker associated with arsenic was detected and assigned, possibly related to shellfish consumption. Conclusions: A number of novel metabolic markers associated with environmental pollutants were identified in preliminary analyses. Further work includes replication in the INMA Gipuzkoa subcohort and grouping exposures by their overall metabolic effects using multivariate methods such as O2-PLS.