BACKGROUND In previous studies, exposures to heavy metals such as Pb and Cd have been associated with adverse birth outcomes; however, knowledge on effects at low levels of exposure and of other elements remain limited. METHOD We examined individual and mixture effects of metals and metalloids on birth outcomes among 812 pregnant women in the Puerto Rico Testsite for Exploring Contamination Threats (PROTECT) cohort. We measured 16 essential and non-essential metal(loid)s in maternal blood collected at 16-20 and 24-28 weeks gestation. We used linear and logistic regression to independently examine associations between geometric mean (GM) concentrations of each metal across visits and gestational age, birthweight z-scores, preterm birth, small for gestational age (SGA), and large for gestational age (LGA). We evaluated effect modification with infant sex*metal interaction terms. To identify critical windows of susceptibility, birth outcomes were regressed on visit-specific metal concentrations. Furthermore, average metal concentrations were divided into tertiles to examine the potential for non-linear relationships. We used elastic net (ENET) regularization to construct Environmental Risk Score (ERS) as a metal risk score and Bayesian Kernel Machine Regression (BKMR) to identify individual metals most critical to each outcome, accounting for correlated exposures. RESULTS In adjusted models, an interquartile range (IQR) increase in GM lead (Pb) was associated with 1.63 higher odds of preterm birth (95%CI = 1.17, 2.28) and 2 days shorter gestational age (95% CI = -3.1, -0.5). Manganese (Mn) and zinc (Zn) were also associated with higher odds of preterm birth and shorter gestational age; the associations were strongest among the highest tertile for Mn and among females for Zn. Mercury (Hg) was associated with higher risk of preterm birth at the later window of pregnancy. Ni measured later in pregnancy was associated with lower odds of SGA. ENET and BKMR models selected similar metals as "important" predictors of birth outcomes. The association between ERS and preterm birth was assessed and the third tertile of ERS was significantly associated with an elevated odds ratio of 2.13 (95% CI = 1.12, 5.49) for preterm birth compared to the first tertile. CONCLUSION As the PROTECT cohort has lower Pb concentrations (GM = 0.33 μg/dL) compared to the mainland US, our findings suggest that low-level prenatal lead exposure, as well as elevated Mn and Zn exposure, may adversely affect birth outcomes. Improved understanding on environmental factors contributing to preterm birth, together with sustainable technologies to remove contamination, will have a direct impact in Puerto Rico and elsewhere.

中文翻译：

---

波多黎各北部母亲血液金属和类金属浓度与出生结果的关系。


背景 在之前的研究中，接触铅和镉等重金属与不良出生结果有关。然而，关于低水平暴露和其他元素的影响的知识仍然有限。方法 我们研究了波多黎各污染威胁探索试验场 (PROTECT) 队列中 812 名孕妇的金属和类金属对出生结局的单独和混合影响。我们测量了妊娠 16-20 周和 24-28 周时采集的母体血液中的 16 种必需和非必需金属（类）。我们使用线性和逻辑回归来独立检查每次访视期间每种金属的几何平均 (GM) 浓度与胎龄、出生体重 z 分数、早产、小于胎龄 (SGA) 和大于胎龄 (LGA) 之间的关联。我们评估了婴儿性*金属相互作用项的效应修正。为了确定易感性的关键窗口，根据特定就诊的金属浓度对出生结果进行了回归。此外，平均金属浓度被分为三等分，以检查非线性关系的可能性。我们使用弹性网 (ENET) 正则化构建环境风险评分 (ERS) 作为金属风险评分，并使用贝叶斯核机器回归 (BKMR) 来识别对每个结果最关键的单个金属，并考虑相关暴露。结果 在调整后的模型中，GM 铅 (Pb) 的四分位数范围 (IQR) 增加与早产几率增加 1.63 (95% CI = 1.17, 2.28) 和孕龄缩短 2 天相关 (95% CI = -3.1， -0.5）。 锰（Mn）和锌（Zn）也与较高的早产几率和较短的胎龄有关；这种关联在最高三分位的锰和女性中与锌的关联最强。汞 (Hg) 与怀孕后期早产的风险较高相关。妊娠后期测量的 Ni 与 SGA 发生率较低相关。 ENET 和 BKMR 模型选择了类似的金属作为出生结果的“重要”预测因子。对 ERS ​​和早产之间的关联进行了评估，与第一个三分位数相比，ERS 的第三个三分位数与早产的比值比升高 2.13 (95% CI = 1.12, 5.49) 显着相关。结论 由于与美国大陆相比，PROTECT 队列的铅浓度较低（GM = 0.33 μg/dL），我们的研究结果表明，低水平的产前铅暴露以及较高的锰和锌暴露可能会对出生结果产生不利影响。提高对导致早产的环境因素的了解，加上消除污染的可持续技术，将对波多黎各和其他地方产生直接影响。