Abstract
Maternal exposure to some individual rare earth elements and trace elements is associated with preterm birth, but few elements have been studied and little is known about the potential effect of simultaneous exposure to multiple elements. We examined individual and mixture effects of elements on preterm birth among 745 pregnant women in a prospective birth cohort in Bangladesh (2008-2011). We measured 56 elements in umbilical cord blood collected during delivery using inductively coupled plasma-mass spectrometry. Using elastic net (ENET) regularization and multivariate logistic regression, we examined independent associations between element concentrations and preterm birth. Bayesian kernel machine regression identified mixture effects of elements most critical to preterm birth, accounting for correlated exposure and interaction. ENET identified titanium (Ti), arsenic (As), and barium (Ba) as the most important predictors of shortened gestational age and preterm birth. In adjusted models, cord blood Ti (OR = 2.52; 95% CI: 1.08-5.93; P = 0.033), As (odds ratio (OR) = 1.34; 95% CI: 1.04-1.73; P = 0.023), and Ba (OR = 1.18; 95% CI: 1.02-1.38; P = 0.029) were significantly associated with preterm birth. Bayesian kernel machine regression suggested an interaction effect between As and Ba. Further, we constructed an element risk score (ERS) using estimated weights from a multivariate regression model for Ti, As, and Ba and regressed preterm birth by this score (OR = 2.72, 95% CI: 1.57-4.69; P = 3.35 × 10(-4)). Additionally, we observed a significant modification effect of child marriage on ERS, which means marriage before the age of 18 (P(interaction) = 0.0438). This study identified element exposures profiles in cord blood and constructed metal risk score that are jointly associated with the risk of preterm birth. Ti, As, and Ba exposure may adversely affect birth outcomes as well as child marriage may be a modifiable factor potentially affecting environmental element exposure and preterm birth.