Association of whole-blood arsenic speciation with hypertensive disorders of pregnancy and birth weight: a prospective birth cohort study

Evidence on the associations of arsenic species during pregnancy with hypertensive disorders of pregnancy (HDP) and fetal growth remains limited, particularly under low-level environmental exposure. We investigated the associations of maternal whole-blood arsenic species with HDP and neonatal birth weight and explored heterogeneity in arsenic metabolism. This prospective birth cohort included 437 singleton pregnant women recruited at 10–14 weeks of gestation in eastern Yunnan, China. Whole-blood arsenic species were quantified using HPLC-ICP-MS. Multivariable logistic regression and restricted cubic splines were used to assess associations between individual arsenic species and HDP. Unsupervised clustering was applied to characterize arsenic metabolic patterns, and Bayesian kernel machine regression (BKMR) was used to explore the joint associations of arsenic species with birth weight. Whole-blood As(V) concentrations were higher among women who developed HDP than among those who did not (geometric means 2.10 vs. 1.34 ng/mL; mean ± SD 3.96 ± 4.00 vs. 3.07 ± 4.15 ng/mL; P = 0.020). Each doubling of As(V) concentration was associated with a modest 24% increase in the odds of HDP (adjusted OR = 1.24, 95% CI: 1.03–1.49). Women in the highest As(V) tertile had higher odds of HDP than those in the lowest tertile (OR = 3.04, 95% CI: 1.39–6.62; Ptrend < 0.005). Restricted cubic spline analysis suggested an overall association (Poverall = 0.027), without significant evidence of nonlinearity (Pnonlinear = 0.112). Exploratory clustering identified arsenic-speciation profiles differing mainly in MMA and methylation indices. Birth-weight analyses suggested species-specific and mixture-related associations, with MMA showing relatively greater importance in the arsenic mixture. Maternal whole-blood As(V) was associated with modestly higher odds of HDP under LOD-substitution approaches, but this association was not statistically significant in a detected-only analysis (OR = 1.17, 95% CI 0.78–1.75; P = 0.462); residual confounding related to maternal BMI cannot be excluded. The clinical relevance of this association requires cautious interpretation. Heterogeneity in arsenic methylation, particularly involving MMA and methylation indices, may contribute to interindividual susceptibility. These findings highlight the importance of arsenic speciation and metabolism in evaluating maternal and fetal health effects and warrant confirmation in larger prospective studies. Not applicable.