Background: Inorganic arsenic (iAs) is a ubiquitous metalloid and drinking water contaminant. Prenatal exposure is associated with birth outcomes across multiple studies. During metabolism, iAs is sequentially methylated to mono- and di-methylated arsenical species (MMAs and DMAs) to facilitate whole body clearance. Inefficient methylation (e.g., higher urinary % MMAs) is associated with increased risk of certain iAs-associated diseases. One-carbon metabolism factors influence iAs methylation, modifying toxicity in adults, and warrant further study during the prenatal period. The objective of this study was to evaluate folate, vitamin B12, and homocysteine as modifiers of the relationship between biomarkers of iAs methylation efficiency and birth outcomes.

Methods: Data from the Biomarkers of Exposure to ARsenic (BEAR) pregnancy cohort (2011-2012) with maternal urine and cord serum arsenic biomarkers and maternal serum folate, vitamin B12, and homocysteine concentrations were utilized. One-carbon metabolism factors were dichotomized using clinical cutoffs and median splits. Multivariable linear regression models were fit to evaluate associations between each biomarker and birth outcome overall and within levels of one-carbon metabolism factors. Likelihood ratio tests of full and reduced models were used to test the significance of statistical interactions on the additive scale (α = 0.10).

Results: Among urinary biomarkers, % U-MMAs was most strongly associated with birth weight (β = - 23.09, 95% CI: - 44.54, - 1.64). Larger, more negative mean differences in birth weight were observed among infants born to women who were B12 deficient (β = - 28.69, 95% CI: - 53.97, - 3.42) or experiencing hyperhomocysteinemia (β = - 63.29, 95% CI: - 154.77, 28.19). Generally, mean differences in birth weight were attenuated among infants born to mothers with higher serum concentrations of folate and vitamin B12 (or lower serum concentrations of homocysteine). Effect modification by vitamin B12 and homocysteine was significant on the additive scale for some associations. Results for gestational age were less compelling, with an approximate one-week mean difference associated with C-tAs (β = 0.87, 95% CI: 0, 1.74), but not meaningful otherwise.

Conclusions: Tissue distributions of iAs and its metabolites (e.g., % MMAs) may vary according to serum concentrations of folate, vitamin B12 and homocysteine during pregnancy. This represents a potential mechanism through which maternal diet may modify the harms of prenatal exposure to iAs.