Association between prenatal exposure to trace elements mixture and visual acuity in infants: A prospective birth cohort study

Abstract

Background: Prenatal environmental factors may affect the development of the offspring and can bring long lasting consequences to the offspring's health. To date, only few studies have reported inconclusive association between prenatal single trace element exposure and visual acuity, and no studies have investigated the association between prenatal exposure to trace elements mixture and visual acuity in infants.

Methods: In the prospective cohort study, grating acuity in infants (12 ± 1 months) was measured by Teller Acuity Cards II. Concentrations of 20 trace elements in maternal urine samples collected in early-trimester were measured by Inductively Coupled Plasma Mass Spectrometry. Elastic net regression (ENET) was applied to select important trace elements. Nonlinear associations of the trace elements levels with abnormal grating were explored using the restricted cubic spline (RCS) method. The associations between selected individual elements and abnormal grating acuity were further appraised using the logistic regression model. Then Bayesian Kernel Machine Regression (BKMR) was used to estimate the joint effects of mixture and interactions between trace elements combining with NLinteraction.

Results: Of 932 mother-infant pairs, 70 infants had abnormal grating acuity. The ENET model produced 8 trace elements with non-zero coefficients, including cadmium, manganese, molybdenum, nickel, rubidium, antimony, tin and titanium. RCS analyses identified no nonlinear associations of the 8 elements with abnormal grating acuity. The single-exposure analyses using logistic regression revealed that prenatal molybdenum exposure possessed a significantly positive association with abnormal grating acuity (odds ratio [OR]: 1.44 per IQR increase, 95% confidence interval [CI]: 1.05, 1.96; P = 0.023), while prenatal nickel exposure presented with a significantly inverse association with abnormal grating acuity (OR: 0.64 per IQR increase, 95% CI: 0.45, 0.89; P = 0.009). Similar effects were also observed in BKMR models. Moreover, the BKMR models and NLinteraction method identified potential interaction between molybdenum and nickel.

Conclusions: We established that prenatal exposure to high concentration of molybdenum and low concentration of nickel was associated with the increased risk of abnormal visual acuity. Potential interaction may exist between molybdenum and nickel on abnormal visual acuity.

Keywords: Bayesian kernel machine regression (BKMR); Mixed exposure; Molybdenum; Nickel; Trace element; Visual acuity.