Protective effect of high zinc levels on preterm birth induced by mercury exposure during pregnancy: A birth cohort study in China.

Objective: The aims of our study were to determine whether prenatal mercury levels are associated with the risk of preterm birth (PTB) and whether high maternal serum zinc (Zn) levels alleviate any negative effects of maternal mercury (Hg) exposure regarding PTB.

Methods: Serum concentrations of Zn and Hg were measured in 3025 pregnant women from the Ma'anshan Birth Cohort. Before the collection of blood samples, they underwent examinations via the completion of questionnaires. The delivery records of the women were obtained from a series of medical records. We divided the study population into tertiles according to the participants' Hg levels: the low-Hg group (the first tertile, <0.30 μg/L), the medium-Hg group (the second tertile, 0.30-0.43 μg/L) and the high-Hg group (the third tertile, ≥0.43 μg/L). The associations of Hg exposure with both the risk of PTB and gestational age (weeks) at birth were estimated using a binary logistic regression model and multivariable linear regression analysis, respectively. Afterwards, we conducted a repeated analyses test after the participants were stratified according to their Zn levels, using the 75th percentile division method.

Results: Overall, the medians and the interquartile ranges of Hg and Zn in the second trimester were 0.36 (0.27, 0.48) μg/L and 812.34 (731.26, 896.59) μg/L, respectively. Hg levels were associated with PTB [adjusted odds ratio (OR) and 95% confidence interval (95% CI): 1.91 (1.17, 3.12) for the third tertile vs. the first tertile of the serum Hg levels]. In the stratification analysis of the participants in the low-Zn group, the high-Hg group exhibited a significant odds ratio of PTB [adjusted OR (95% CI): 1.87 (1.08, 3.24)], compared to the low-Hg group. However, in the participants from the high-Zn group, the high-Hg group exhibited a non-significant OR of PTB [adjusted OR (95% CI): 2.32 (0.73, 7.42)]. In the multivariate linear regression analysis, gestational age (weeks) at delivery was significantly and inversely associated with the ln-transformed Hg concentrations [adjusted β (95% CI): -0.16 (-0.26, -0.06)]. Similarly, after the stratification analysis in the high-Zn group, there were no significant associations between PTB and the Hg levels [adjusted β (95% CI): -0.12 (-0.33, 0.09)].

Conclusion: Prenatal Hg exposure adversely affected PTB, and high Zn levels alleviate this effect, which indicates that a more stringent control of Hg and a sufficient intake of Zn are necessary to help birth outcomes.