Skip to main content
NCBI home page
Maternal & Child Nutrition logoLink to Maternal & Child Nutrition
. 2024 Jun 24;20(4):e13682. doi: 10.1111/mcn.13682

Effects of exposure to multiple metallic elements in the first trimester of pregnancy on the risk of preterm birth

Ting Wu 1,2, Chuan Luo 3, Tao Li 1,2, Chen Zhang 3, Hui‐Xi Chen 1,2, Yi‐Ting Mao 3, Yan‐Ting Wu 3,4,✉, He‐Feng Huang 1,2,3,4,5,6,7,✉
PMCID: PMC11574644  PMID: 38925571

Abstract

Exposure to certain heavy metals has been demonstrated to be associated with a higher risk of preterm birth (PTB). However, studies focused on the effects of other metal mixtures were limited. A nested case‒control study enrolling 94 PTB cases and 282 controls was conducted. Metallic elements were detected in maternal plasma collected in the first trimester using inductively coupled plasma‒mass spectrometry. The effect of maternal exposure on the risk of PTB was investigated using logistic regression, least absolute shrinkage and selection operator, restricted cubic spline (RCS), quantile g computation (QGC) and Bayesian kernel machine regression (BKMR). Vanadium (V) and arsenic (As) were positively associated with PTB risk in the logistic model, and V remains positively associated in the multi‐exposure logistic model. QGC analysis determined V (69.42%) and nickel (Ni) (70.30%) as the maximum positive and negative contributors to the PTB risk, respectively. BKMR models further demonstrated a positive relationship between the exposure levels of the mixtures and PTB risk, and V was identified as the most important independent variable among the elements. RCS analysis showed an inverted U‐shape effect of V and gestational age, and plasma V more than 2.18 μg/L was considered a risk factor for shortened gestation length. Exposure to metallic elements mixtures consisting of V, As, cobalt, Ni, chromium and manganese in the first trimester was associated with an increased risk of PTB, and V was considered the most important factor in the mixtures in promoting the incidence of PTB.

Keywords: Bayesian kernel machine regression, LASSO regression, preterm birth, quantile g computation, restricted cubic spline, vanadium

Exposure to metallic elements mixtures consisting of vanadium (V), arsenic (As), cobalt (Co), chromium (Cr) and manganese (Mn) in the first trimester was associated with an increased risk of preterm birth (PTB), and V was considered the most important factor in the mixtures in promoting the incidence of PTB.

graphic file with name MCN-20-e13682-g004.jpg

Key messages

  • The positive joint effect of a metal mixture consisting of vanadium (V), arsenic, cobalt, chromium and manganese on preterm birth (PTB) was observed by applying different interdisciplinary statistical methods.

  • V was defined as the most important risk factor for PTB among the metal mixture.

  • Maternal plasma concentration of V of more than 2.18 μg/L was considered a risk factor for shortened gestation length.

1. INTRODUCTION

Preterm birth (PTB), defined as the birth before 37 weeks of gestation, is a significant global public health problem given its high morbidity and associated mortality of neonatal and children below the age of 5 years (Romero et al., 2014). Every year, approximately 15 million babies are born preterm worldwide, estimating about 11% global PTB rate (Walani, 2020). It is a syndrome caused by multiple pathologic processes, and several risk factors, including multiple pregnancies, chorioamnionitis, genetic factors, pre‐existing maternal disease and uterine abnormalities, have been well recognized (Green & Arck, 2020). However, the precise etiology of PTB has not been illustrated clearly yet. With the increasingly serious environmental pollution, much attention has been focused on the adverse effect of environmental chemicals on PTB risk. Metals are ubiquitously detected in the environment, and a growing number of epidemic studies have explored the effects of prenatal exposure to certain toxic metals on the risk of PTB (Khanam et al., 2021). Other kinds of metals also remain further studied for their effect on the human body, which may play important biological roles in normal circumstances but become toxic in different oxidation states and doses, or when interacting with other metals. For instance, vanadium (V) has been demonstrated an essential role in the metabolism of carbohydrates, lipids, phospholipids and cholesterol, but can also serve as a strong pro‐oxidant and pro‐apoptotic factor, especially when at its highest oxidation state (+5) (Ścibior et al., 2020). manganese (Mn) and chromium (Cr) exhibit both toxic and nutritional properties, depending on the level of exposure. Both adverse and protective effects of nickel (Ni) have been reported in relation to fetal growth (Deyssenroth et al., 2018; Cabrera‐Rodríguez et al., 2018; Jalali & Koski, 2018; Sun et al., 2018).

Humans are exposed to a complex mixture of various metallic elements rather than a single metal in daily life. However, when exploring the health effects of metallic elements mixtures by using traditional regression model, the results may be highly unstable if including several highly correlated (collinear) metallic elements in the model. Recent epidemiological studies have moved the research focus to study the co‐exposure health effects of metals by using various statistical methods (Yu, 2022). A large Chinese birth cohort study involving 7291 pregnant women suggested higher maternal urinary concentrations of Ni were positively associated with the risk of preterm delivery (Chen et al., 2018). Maternal exposure to Mn and Cr detected in the placenta was suggested to positively affect gestational age by the Environment and Childhood (INMA) Project in Spain which involved 327 mother–infant pairs (Freire et al., 2019). However, a USA cohort study found higher exposure level of Mn was associated with a higher risk of PTB and shorter gestational age (Ashrap et al., 2020). A nested case–control study conducted in China demonstrated the metal mixture consisting of 18 metals detected in maternal urine was positively associated with PTB and considered V as the most significant risk factor of PTB in the metal mixture (Liu et al., 2022). In addition, more attention should be paid to the association between prenatal exposure to Co and the risk of PTB, and the health effects of As on PTB need further proof.

Located at the mouth of the Yangtze River, Shanghai is the largest, richest and most developed region in China, the Yangtze River Delta has experienced increasingly severe environmental pollution of metals produced by industrial and agricultural activities (Zhuang & Zhou, 2021). Therefore, our study aimed to explore the relationship between maternal exposure level of the abovementioned six metallic elements detected in maternal plasma before 14 weeks of pregnancy and the risk of PTB in Shanghai, China by using four statistical methods, and the non‐linear relationship between metallic elements exposure and gestational age were also illustrated.

2. MATERIALS AND METHODS

2.1. Study population

The participants enrolled in the present case–control study were part of a prospective study that recruited pregnant women between November 2020 and February 2021, from the International Peace Maternal and Child Hospital (IPMCH). A total of 2069 pregnant women agreed to participate in the study and blood samples were collected between 8 and 14 weeks of pregnancy when they conducted their first prenatal examination.

Finally, 1777 pregnant women gave birth in our hospital with complete information on birth outcomes, we excluded those with multiple births (n = 52), those diagnosed with chronic hypertension before pregnancy (n = 17), or with other serious medical diseases (n = 12). In addition, there were 79 participants with insufficient blood samples and one participant with missing pre‐pregnancy body mass index (BMI).

Among the 1617 pregnant women finally included, 94 who delivered before 37 gestational weeks were defined as PTB cases. Controls were matched by the ratio of 1:3 according to maternal age and the newborn sex (the nearest), and 282 pregnant women were included in the control group.

2.2. Data collection and outcomes

The data used in this study were collected from the electronic medical records, including maternal demographic and socio‐economic information (e.g., maternal age, maternal height, maternal pre‐pregnancy weight, education level, ethnicity and annual household income level), lifestyle factors (e.g., smoking and alcohol consumption history), history of family and personal disease, and data on delivery. Maternal pre‐pregnancy BMI was calculated using the formula BMI = weight (kg)/height (m2) and was categorized as <18.5, 18.5–24.9 or ≥25 kg/m2 corresponding to underweight, normal weight, and overweight or obesity, respectively. Maternal education level was classified as high school and lower (<10 years), junior and college (10–12 years) and university and higher (≥13 years). Annual household income level was categorized as <0.1 million CNY, 0.1–0.3 million CNY or ≥0.3 million CNY. Almost all pregnant women in the study have no or almost no history of smoking and drinking. Gestational age was calculated based on the self‐reported first day of the last menstrual period and the gestational week of delivery, and PTB was defined as a live birth before 37 weeks.

2.3. Exposure assessment

Concentrations of maternal metallic elements were measured in plasma collected between 8 and 14 weeks of pregnancy, with a median of 12.57 weeks. The plasma was collected into EDTA tubes from blood after the centrifugation (2000 rpm, 20 min) and stored at −80°C. For the preparation of the analysis, plasma was transferred to a refrigerator of 4°C the night before the test. About 100 μL plasma was equilibrated to room temperature before the test, diluted with 1.9 mL sample diluent (1% TMAH1 + % nitric acid), and homogenized by shaking. Metallic elements were measured via inductively coupled plasma‒mass spectrometry (ICP‐MS) using a NexION 300X device (PerkinElmer).

Several element standard solutions (PerkinElmer) diluted in proportion were measured to prepare multiple standard curves before the sample test. The value of the limit of detection (LOD) of each metallic element was calculated according to each standard curve, the value of LOD/√2 was used to present the concentration of the sample which concentration is below the value of LOD. Standard samples were redetected after the detection of every 30 samples to verify the detection accuracy.

2.4. Statistical method

Using the propensity score matching (PSM) method to select the control for the PTB cases by the ratio of 1:3 based on maternal age and the newborn sex. The plasma concentrations of each metallic element were natural log‐transformed [Ln(X)] to reduce the impact of extreme values. We also visualized the pairwise correlations among the six metallic elements analyzed through Spearman's rank correlation by a correlation‐matrix heatmap. For a statistical description of the characteristics and exposure level, N (%) represents categorical variables, and median and interquartile range (IQR) represent continuous variables. Wilcoxon rank‐sum test (for continuous variables) or chi‐square (χ2) test (for categorical variables) were used to conduct comparisons between case and control groups.

2.4.1. Single‐metal model

Conditional logistic regression was adopted to evaluate the association between maternal metal exposure levels and the risk of PTB represented by odds ratios (ORs) and 95% confidence intervals (CIs). Maternal metal exposure levels were represented by Ln‐transformed metal which is defined as continuous variables. The single‐metal models were adjusted by pre‐pregnancy BMI, ethnicity (Ethnic Han or others), education level (<10, 10–12 and ≥13 years), annual household income level (<0.1, 0.2–0.3 and >0.3 million), parity (nulliparous and multiparous) and pregnancy‐induced hypertension (PIH) (yes or no). In addition, the potential nonlinearity of the association between Ln‐transformed metal with gestational age was further examined by restricted cubic spline (RCS) with three knots fixed at the 25, 50 and 75 percentiles (P25, P50 and P75).

2.4.2. Mixture analysis models

We fitted the multi‐exposure model by using the least absolute shrinkage and selection operator (LASSO), quantile g computation (QGC) and Bayesian kernel machine regression (BKMR) to explore the effects of metallic element mixture on PTB risk and identify important mixture components. LASSO regression is a shrinkage (penalized regression) method that pushes the minimums of coefficients to exactly zero by directly shrinking the sum of the absolute values of coefficients (Yu, 2022). Six potential confounding factors, including maternal BMI level, ethnicity, educational level, household income level, parity and pregnant women whether have PIH, as well as six metallic elements, were introduced into the LASSO regression model, and the independent variables with a greater contribution to PTB risk were selected when the regression coefficient reached zero. These screened‐out metallics elements were simultaneously included in the multi‐exposure logistic model adjusted or not adjusted by selected covariates including ethnic group, educational level, household income level, and with or without PIH. QGC is a new method that combines the advantage of weighted quantile sum (WQS) and g computation. It can estimate the effect of the multi‐metal mixture on PTB risk and evaluate the positive or negative weights of each metal on PTB risk. BKMR is also capable of assessing the cumulative effects of metal mixture on PTB risk and determining the contribution of each metal. Posterior inclusion probabilities (PIPs) that represented the relative importance of each metal in the mixture were obtained, with the index range from 0 (least important) to 1 (most important), and the metal with a PIP greater than 0.5 was usually considered significant. In addition, nonlinear relationships between individual metals and potential interactions among metals were also identified by BKMR.

Statistical analyses were conducted using SPSS 26.0 and R version 4.2.0 software. All significant levels were set to p value (two‐tailed) <0.05. RCS, LASSO, QGC and BKMR were performed with the R packages ‘rms’, ‘glmnet’, ‘qgc’ and ‘bkmr’, respectively.

2.5. Ethical statement

This study was approved by the IPMCH Ethics Committee ((GKLW) 2019‐51). All participants in the study have signed the informed consent form.

3. RESULTS

3.1. Study population

The baseline information of 376 participants is presented in Table 1 . The study population has a median age of 31 years and has received a good education, with about 71.2% of women having an education level of university or above. Primiparous women account for 71.20%. The median gestational age of infants in the PTB group was 35.40 weeks. The characteristics between the two groups were comparable and no pregnant women had a smoking or drinking history during the pregnancy.

Table 1.

Characteristics of the participants.

Total (n = 376) Control (n = 282) PTB (n = 94) p Value
Maternal age 31.00 (29.00–34.00) 31.00 (29.00–34.00) 31.00 (29.00–34.00) 0.863
Pre‐BMI level
<18.5 44 (11.7%) 33 (11.7%) 11 (11.7%) 0.523
18.55–23.9 270 (71.8%) 206 (73.0%) 64 (68.1%)
>24 62 (16.5%) 43 (15.2%) 19 (20.2%)
Ethnic group
Han 366 (97.3%) 273 (96.8%) 93 (98.9%) 0.459
Others 10 (2.7%) 9 (3.2%) 1 (1.1%)
Education level
High school and lower 40 (10.7%) 16 (17.0%) 24 (8.5%) 0.070
Junior and college 68 (18.1%) 16 (17.0%) 52 (18.5%)
University and higher 267 (71.2%) 62 (66.0%) 205 (73.0%)
Household income level
<0.1 million 26 (7.0%) 17 (6.1%) 9 (9.6%) 0.293
0.1–0.3 million 229 (61.2%) 169 (60.4%) 60 (63.8%)
>0.3 million 119 (31.8%) 94 (33.6) 25 (26.6%)
Parity
Nulliparous 267 (71.2%) 203 (72.2%) 64 (68.1%) 0.441
Multiparous 108 (28.8%) 78 (27.8%) 30 (31.9%)
Gestational diabetes
No 295 (78.5%) 67 (71.3%) 228 (80.9%) 0.051
Yes 81 (21.5%) 27 (28.7%) 54 (19.1%)
Pregnancy‐induced hypertension
No 337 (89.6%) 78 (83.0%) 259 (91.8%) 0.015*
Yes 39 (10.4%) 16 (17.0%) 23 (8.2%)
Gestational weeks 38.50 (36.70–39.40) 39.10 (38.30–39.60) 35.40 (34.30–36.40) <0.001***
Infant sex
Male 222 (59.0%) 166 (58.9%) 56 (59.6%) 0.904
Female 154 (41.0%) 116 (41.1%) 38 (40.4%)
Sampling time 12.57 (12.04–13.14) 12.57 (12.14–13.14) 12.43 (12.00–13.00) 0.286
Open in a new tab

Abbreviations: BMI, body mass index; PTB, preterm birth.

*

p < 0.05

***

p < 0.001.

3.2. Plasma concentration of metals and PTB risk

The profiling of the metallic elements in maternal plasma is shown in Supporting Information S5: Table S1. The detection ratio of metals was 100% for Cr and Co, higher than 95% for V (97.61%) and As (99.73%), 94.15% for Mn and 88.03% for Ni. Compared with pregnant women with term delivery, there were significantly higher maternal plasma levels of V and As and significantly lower levels of Cr in the PTB group. No differences were observed in plasma Mn, Co and Ni (Table 2).

Table 2.

Exposure levels of metallic elements in the case–control group.

Total (n = 376) Control (n = 282) PTB (n = 94) p Value
V (μg/L) 2.47 (1.56–3.97) 2.14 (1.41–3.11) 5.10 (2.98–7.62) <0.001***
Cr (μg/L) 413.35 (238.36–611.04) 437.69 (246.69–625.28) 343.11 (231.12–530.47) 0.010**
Mn (μg/L) 5.31 (2.96–8.38) 5.40 (3.13–8.62) 4.94 (2.49–7.31) 0.267
Co (μg/L) 53.66 (38.07–75.78) 53.95 (38.85–76.43) 51.44 (33.34–71.59) 0.281
Ni (μg/L) 29.34 (16.22–46.05) 29.50 (18.37–47.05) 28.42 (12.86–40.67) 0.086
As (μg/L) 12.56 (9.78–15.00) 11.84 (9.42–14.19) 14.27 (11.25–17.53) <0.001***
Open in a new tab

Abbreviations: As, Arsenic; Co, Cobalt; Cr, Chromium; Mn, Manganese; Ni, Nickel; PTB, preterm birth; V, Vanadium.

**

p < 0.01

***

p < 0.001.

After adjusting for maternal pre‐pregnancy BMI level, ethnic group, educational level, household income level, parity, and PIH, plasma V (OR = 5.07 [95% CI = 3.15–8.16]) and As (OR = 7.90 [95% CI = 3.30–18.89]) have shown positive associations with the risk of PTB. In contrast, there was a negative association between the level of plasma Ni and the risk of PTB; for each unit increase in Ln‐Ni, there was a 25% (OR = 0.75 [95% CI = 0.59–0.96]) decrease in the risk of PTB (Table 3).

Table 3.

Associations between prenatal metallic element exposure and PTB risk.

Single‐exposure OR (95% CI) Co‐exposure OR (95% CI)
Ln‐metals Crude model Adjusted model Crude model Adjusted model
V (μg/L) 5.30 (3.33–8.45)*** 5.07 (3.15–8.16)*** 3.97 (2.21–7.13)*** 3.50 (1.90–6.43)***
Cr (μg/L) 0.69 (0.47–1.01) 0.71 (0.48–1.07) / /
Mn (μg/L) 0.89 (0.66–1.19) 0.89 (0.66–1.22) / /
Co (μg/L) 0.83 (0.56–1.22) 0.84 (0.56–1.26) 0.59 (0.30–1.15) 0.62 (0.31–1.24)
Ni (μg/L) 0.77 (0.61–0.98)* 0.75 (0.59–0.96)* 0.71 (0.47–1.08) 0.67 (0.43–1.03)
As (μg/L) 7.18 (3.16–16.33) 7.90 (3.30–18.89)*** 3.74 (0.80–17.58) 4.75 (0.89–25.30)
Open in a new tab

Note: The single‐exposure adjusted model was adjusted by ethnicity, pre‐pregnancy BMI level, education level, parity, household income level and pregnancy‐induced hypertension during pregnancy; The co‐exposure adjusted model was adjusted by household income level.

Abbreviations: As, Arsenic; BMI, body mass index; CI, confidence interval; Co, Cobalt; Cr, Chromium; Mn, Manganese; Ni, Nickel; OR, odds ratio; PTB, preterm birth; V, Vanadium.

*

p < 0.05

***

p < 0.001.

LASSO regression considered V, Co, Ni and As as significant risk factors of PTB risk (Figure S2). But when we simultaneously bring V, Co, Ni and As into the logistic regression model adjusted by confounding factors screened by the LASSO regression including ethnic group, educational level, household income level and PIH, only V showed positive associations with the risk of PTB (OR = 3.50 [95% CI = 1.90–6.43]).

3.3. Dose–response relationships between V, Cr, Mn exposure and the gestational age

The potential nonlinear relationships between Ln‐V (p overall < 0.001, p nonlinearity < 0.001), Ln‐Cr (p overall = 0.020, p nonlinearity = 0.015) and Ln‐Mn (p overall = 0.028, p nonlinearity = 0.014) with the gestational age were observed in the RCS analysis. Figure 1a illustrated an inverted U‐shape between Ln‐V and the infant's gestational age, the gestational age of infants was significantly decreased as the increased level of V when it was higher than 2.18 μg/L. Figure 1b illustrated an S‐shape between Ln‐Cr and the infant's gestational age, the gestational age of infants was increased as the increased level of Cr when it was between 310.78 and 633.45 μg/L. Similarly, an S‐shape was also observed between Ln‐Mn and gestational age, and the critical values were 1.64 and 6.57 μg/L.

Figure 1.

Figure 1

Open in a new tab

RCS analysis for ln‐transformed V, Cr, Mn and gestational age. Knots were placed at Ln's P25, P50, and P75 (plasma V, Cr and Mn). Results were adjusted by ethnicity, pre‐pregnancy BMI level, education level, parity, annual household income level and PIH. BMI, body mass index; PIH, pregnancy‐induced hypertension; RCS, restricted cubic spline. (a) RCS analysis for ln‐transformed V and gestational age. (b) RCS analysis for ln‐transformed Cr and gestational age. (c) RCS analysis for ln‐transformed Mn and gestational age.

3.4. Quantile g computation analyses

QGC analysis demonstrated that the risk of PTB was positively associated with the increasing exposure level of the metallic element mixture (OR = 2.34, 95% CI = 1.49–3.68, p < 0.001). The individual impact of each mixture component on PTB risk is presented in Figure 2, with V (69.42%), As (18.43%), Co (9.77%) and Mn (2.37%) having positive contributions to PTB risk, and the negative weight of Ni and Cr was 70.30% and 29.70%, respectively.

Figure 2.

Figure 2

Open in a new tab

Quantile g computation analyses diagram. The estimated positive or negative effects of each metallic element in the mixture.

3.5. BKMR analyses

The PIP value of each metallic element varied between 0.2 and 1 (Supporting Information S5: Table S2), and V was considered important to the PTB risk as its PIP was higher than 0.5. Figure 3a showed the cumulative effect of exposure to metallic element mixture on the PTB risk, and with the elevated of all metallic element exposure levels from the 25th percentile to their 75th percentile, PTB risk increased significantly.

Figure 3.

Figure 3

Open in a new tab

The joint effect of element mixtures on PTB risk by using the BKMR model. (a) The cumulative effects of metallic element mixtures were estimated at 95% CI) by comparing the value of the exposure–response relationships when all the elements were fixed at several certain percentiles and when all were at their 50th percentile. (b) The effects of single exposure with an individual metallic element fixed at P75 as compared to P25, when all other metallic elements were fixed at different quantiles (P25, P50 and P75). The results were adjusted by ethnicity, pre‐pregnancy BMI level, education level, parity, annual household income level and PIH. BKMR, Bayesian kernel machine regression; BMI, body mass index; CI, confidence interval; PIH, pregnancy‐induced hypertension; PTB, preterm birth.

Figure 3b shows the estimated contribution of each element exposure to the joint effect by observing the change of PTB risk when the exposure level of a single element increased (25th–75th), with all of the other elements fixed to a certain quantile, such as P25, P50 and P75. Visually, V showed a positive association with the risk of PTB, and this effect was hardly changed by the different exposure levels of other metallic elements, whether fixed at their P25, P50 or P75.

Univariate exposure–response functions were also estimated by the BKMR models (Figure 3); approximately negative linear relationships between PTB risk with Co, Ni and As were observed, and for Cr with PTB risk, there is a positive linear relationship. V exhibited a U‐shaped non‐linear relationship. Besides, bivariate exposure–response analysis is displayed in Figure 3; when V was fixed at the P90, the trends of other metals and PTB were changed compared to when V was at its 10th quantile, suggesting the potential existing interaction between V and other remaining metals in the mixture.

4. DISCUSSION

In this study, we investigated the associations between the concentrations of six metallic elements in maternal plasma before 14 weeks of pregnancy and PTB risk by fitting several statistical models, and evaluated the contribution of each metallic element to the outcome.

In conditional logistic regression analyses, there were positive associations between Ln‐V and Ln‐As and the PTB risk, while Ni was negatively associated. Both QGC and BKMR demonstrated the positive cumulative effect of the metallic element mixture on PTB risk. V was found positively associated with PTB risk in all these methods and was identified as the most crucial independent effect. In addition, RCS depicted the non‐linear dose–response relationship between Cr (S‐shaped), Mn (S‐shaped) and V (inverted U‐shaped) with gestational age.

Both V and As were found significantly higher in the PTB group and the conditional logistic model showed V and As were positively associated with PTB risk, both indicating maternal exposure to V and As a risk factor of PTB. In addition, both the BKMR model and QGC analysis considered V as the most important contributor to the incidence of PTB among the metallic element mixture, which is consistent with the previous findings from the same research team (Hu et al., 2017; Liu et al., 2022). Hu et al. (2017) measured V in maternal urine from 7297 women and indicated a non‐linear dose–response relationship between Ln‐V concentration with risk of PTB (S‐shaped, p < 0.001). Wang et al. (2022) determined concentrations of 21 elements in cord whole blood and also found V was positively associated with PTB risk. The design of our study is similar to that of the study conducted by Liu et al. (2022) and V was also demonstrated as the most important risk factor of PTB risk in mixtures consisting of 18 metals by the use of the BKMR model.

Similarly, the BKMR model in the present study also observed an approximately U‐shaped non‐linear relationship. Besides, a non‐linear relationship between Ln‐V concentration and gestational age was also shown in RCS analysis from our findings (inverted U‐shaped, p < 0.001); the gestational age of infants was significantly decreased as the increased level of V when it was higher than 2.18 μg/L. To our knowledge, this is the first study to indicate the dual effect of maternal V exposure on gestation length and provide a safe V exposure level (2.18 μg/L) of potential clinical significance, which is close to the median value of the V exposure level of our participants (median: 2.47 μg/L).

V is a transition element that can be ubiquitously detected in the environment. Human exposure to V mainly from daily diet and atmosphere. We detected a relatively higher plasma V concentration (median: 2.47 μg/L) than those reported from previous studies (median: 0.191 μg/L [Li et al., 2017], 0.80 μg/L [Li et al., 2022]). V can form various compounds, plays important biological functions, and has a wide range of applications. There also be a hazard to human health when exposed to V, especially when V is at its highest oxidation state (+5) (Ścibior et al., 2020). The reproductive toxicity of V is not yet clear and should be multifactorial. Animal research has shown that oral administration of V can cross the placental barrier and accumulate in the fetal membrane (Paternain et al., 1990). Recent studies indicated that V can increase the production of reactive oxygen species (ROS), leading to membrane damage and steatosis (Soares et al., 2007; Ścibior et al., 2006). In addition, V compounds can activate key transcription factors in the inflammatory response and induce the secretion of several pro‐inflammatory cytokines, such as IL‐6 and IL‐8 (Soares et al., 2007; Ye et al., 1999; Zwolak, 2014). Both oxidative stress and inflammatory response are involved in the environmental exposure‐related premature rupture of membranes (PROM) (Wallace et al., 2016), which is the main factor inducing premature birth in pregnant women. Experimental data have also shown that V compounds can upregulate the expression and activity of cyclooxygenase‐2 (COX‐2) (Korbecki et al., 2015) as pro‐inflammatory agents, an enzyme involved in prostaglandin synthesis in human myometrial cells and elevated at the onset of labour in the fetal membrane (Hirst et al., 1995). Besides, V compounds can activate the mitogen‐activated protein kinase (MAPK) cascades and the nuclear factor kappa‐light‐chain enhancer of activated B cells (NF‐κB) signalling at low concentrations (Chen et al., 1999), and the activating of MAPK and NF‐κB, as well as ROS, are involved in regulating the COX. In conclusion, V‐induced ROS and inflammatory reaction, as well as the upregulating of COX‐2, may be the underlying mechanism that contributes to premature deliveries. Thus, more attention should be paid to the adverse effect of maternal exposure to V on PTB risk.

However, the positive association between As and PTB risk is not as stable as those of V, with no association found between As and PTB risk when we brought Co, Ni and V into multiple logistic models. Similarly, BKMR estimated no association between As and PTB risk in the univariate exposure–response functions. The study of the effect of exposure to As and PTB has been well‐established (Khanam et al., 2021). A Chinese cohort study measured maternal serum As at the beginning and middle of the pregnancy (4–27 weeks of gestation) in 3194 pregnant women and indicated that the incidence of moderate‐to‐late PTB was elevated in the high As exposure group as compared with the low As exposure group (Wang et al., 2018). Cord blood As level was also reported to be significantly associated with PTB in a prospective birth cohort in Bangladesh (Huang et al., 2021). However, two studies from Chinese and Spain (Freire et al., 2019; Wang et al., 2022), which measured Arsenic in cord whole blood and placental tissue, respectively, did not reach the same result. Considering the As were found significantly higher in the PTB group and were positively associated with PTB risk in logistic models and QGC, we would like to recommend reducing the exposure to As for pregnant women, especially during the first period of pregnancy.

Epidemiological studies focused on the effects of maternal exposure to Ni and gestational age are limited. Xu et al. (2022) measured metallic elements in pregnant women's serum collected at 4–22 gestational weeks and did not find a relationship between Ni exposure and risk of spontaneous preterm birth (SPB). Another study (Wang et al., 2022) from China that measured metallic elements in cord blood also did not find such a relationship. However, a Chinse study (Guo et al., 2010) measured Ni concentration in the placenta and found a significant correlation between Ni exposure level with shortened gestational length. It was reported that Ni can trigger oxidative stress (Käkelä et al., 1999), induce inflammatory reactions (Viemann et al., 2007), and directly lead to lipid peroxidation damage of the placental membrane (Lin, 1998). Therefore, we predicted that Ni exposure would promote the occurrence of PTB in the study, but the result was the opposite. Ni showed a negative association with PTB in logistic regression and was determined as a critical protection factor of PTB in QGC analysis. The discrepancy of study samples and the plasma Ni level (median (IQR) 29.34 [17.34–48.58] μg/L) between our study and previous studies (median: 2.48–6.484 μg/L) (Wang et al., 2019; Li et al., 2021; Yuan, 2018; Yüksel et al., 2021) may largely account for the inconsistent findings.

Cr and Co are essential trace elements for humans, and Mn can act as an essential element and a potential toxicant based on the exposure level. No associations were found in the present study between exposure to Cr, Mn and Co with PTB risk in the logistic model. Interestingly, S‐shaped non‐linear relationships between Cr, Mn and gestation length were both observed in RCS analysis.

Studies exploring associations between prenatal exposure to Cr, Mn and PTB risk have shown varied findings. As the study mentioned before, Xu et al. (2022) did not find a relationship between Cr exposure and the risk of SPB, but the concentrations of Mn in maternal serum during early pregnancy were positively associated with the risk of SPB. However, Wang et al. (2022) did not find relationships between Cr, Mn in the cord whole blood and the risk of PTB. While, a study based on Spanish mother‐child pairs indicated that in‐utero exposure to Cr and Mn, as measured in the placenta, has a positive effect on gestational length (Freire et al., 2019). RCS graph in our results showed the negative association between gestation length with Cr and Mn levels, when the plasma concentration of Mn was higher than 6.57 μg/L, and Cr higher than 633.45 μg/L.

It is noteworthy that the plasma concentration of Mn in our study where relatively lower (median [IQR] 5.31 [2.96–8.38] μg/L) than previously reported (median: 6.52–21.85 μg/L) (Ashrap et al., 2020; Callan et al., 2013; Shan et al., 2016; Soomro et al., 2019; Zhu et al., 2021), which may be the main reason we fail to observe the potential high‐dose Mn‐related toxicities as indicated in previous findings. Different from Mn, the median value of the plasma concentration of Cr measured in our study population was significantly higher than those of previous studies (median: 0.2–3.97 μg/L) (Chen et al., 2017; Lin et al., 2016; Zhu et al., 2021). Although Cr was significantly lower in the PTB group, further study is needed to investigate the health effect of maternal Cr exposure on the PTB risk.

Few studies have focused on the association between Co exposure and the risk of PTB. Our present results appear to support the findings of Xu et al. (2022), who did not find a relationship between Co exposure and the risk of SPB. However, Wang et al. (2022) and Li et al. (2019) indicated that Co levels in the cord whole blood and umbilical cord serum, respectively, were associated with the increased risk of PTB. The results of associations between Cr, Co and PTB risk were inconsistent between QGC analysis and those of the BKMR model, which may be attributed to the different statistical models, especially when these related trends are not obvious.

It is worth mentioning that inconsistent findings of maternal exposure to metallic elements and their association with PTB risk can partly be attributed to geographic or demographic variations. Two studies explored the associations between element concentrations in cord blood and PTB risk based on the population in rural Bangladesh (Huang et al., 2021) and Guangdong, China (Wang et al., 2022), respectively. Both studies used ICP‐MS for element detection, and the detection rates and concentrations of As were comparable, but did not reach the same conclusion. Most recent epidemiological studies were conducted in wealthy countries/regions in North America and Asia (Issah et al., 2024), mainly in the United States and China. The different environmental contexts and socioeconomic factors between these countries and other middle and lower‐income countries/regions may introduce geographical biases and may limit the widespread applicability of research results on a global scale. Further research was imperative to be conducted in the middle and lower‐income countries to deepen the understanding of metallic elements' exposure impact on PTB risk.

In addition, more advancements should be made to enhance our understanding of the potential genotoxic and reprotoxic effects of V or its compounds and facilitate more informed and evidence‐based decision‐making. Several studies have reported the occupational exposure toxicity of V, including causing oxidation of DNA bases, affecting DNA repair (Ehrlich et al., 2008), and altering neurobehavioral outcomes (Li et al., 2013; Ścibior et al., 2023), which can give rise to the onset of genetic syndromes, fetal malformations, and adversely affects the neural development of offspring. In addition, apart from PTB, higher levels of V exposure and other metallic elements were also reported to be associated with increased odds of delivering LBW (low birthweight) infants and might impair fetal growth (Issah et al., 2024; Li et al., 2021). Policymakers should pay attention to the potential impact of V exposure on worker health, strengthen the construction of protective measures, and enhance worker awareness of protection through publicity and education.

What is more, the long‐term health effects of prenatal metallic elements exposure on children cannot be ignored. Results from recent studies showed several associations between levels of metallic elements during gestation and higher blood pressure (Howe et al., 2021), affected growth trajectory pattern (Yao et al., 2023), impacted lung function (Signes‐Pastor et al., 2023), impaired neuro and behaviour development (Ma et al., 2023) in children. Prenatal V exposure was reported to hurt early‐childhood growth and cellular immunity (Yao et al., 2023; Zhang et al., 2021). There are evidence that relates As and Mn exposure with neurodevelopmental problems in children (Rodríguez‐Barranco et al., 2013; Skogheim et al., 2021), especially the As. Wang et al. (2018) provided evidence of inverse association between low‐level prenatal As exposure and neurobehavioral performance of newborns. Liang et al. (2020) used the Ages and Stages Questionnaire of China (ASQ‐C) to assess the status of children's development and behaviour and found low‐level As exposure in utero could have an adverse domain‐specific effect on children's development at 6 months of age, particularly among females. Chen et al. (2023) found inverse associations between prenatal arsenic exposure, especially in early pregnancy, and neurodevelopment of children at 2 years old. Such impairment was also reported in children of 4–5 years of age living in Spain (Freire et al., 2018). Therefore, the impact of prenatal exposure to As on the neurodevelopment of offspring may persist from birth to the preschool stage.

The major strength of our study was that the sample collection period was in early pregnancy, which is a dangerous period of environmental exposure and may suggest the causal relationship between metallic elements exposure and the occurrence of PTB. Besides, more accurate continuous variables rather than categorical variables were used to present the exposure level. Third, we adopted several statistical models to confirm the metal mixture's positive joint effect on the risk of PTB. The main limitation was the relatively small sample size. What is more, plasma may not well reflect the body exposure level of some metallic elements with a short half‐life period, such as Ni.

5. CONCLUSION

Our study revealed the significantly positive joint effect metal mixture consisting of V, As, Co, Ni, Cr and Mn on PTB by applying different statistical methods. We further confirmed that V was the most important risk factor for PTB among the metal mixture, and maternal plasma concentration of V of more than 2.18 μg/L was considered a risk factor for shortened gestation length. Further study is warranted to explore the underlying mechanism of V act on the occurrence of the PTB.

AUTHOR CONTRIBUTIONS

Ting Wu designed the study, perfomed the experiment, analysed the data, and writed the original manuscript. Chuan Luo collected the data, performed the experiment. Tao Li collected the biological sample Chen Zhang contributed to statistical analysis. Hui‐Xi Chen contributed to the data collecting. Yi‐Ting Mao contributed to the biological sample collecting. Yan‐Ting Wu contributed to the design and implementation of the study, review and editing the manuscript. He‐Feng Huang contributed to the funding acquisition, supervision, review and editing. All authors have read and agreed to the published version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

Supplementary Figure 1. Spearman's rank correlation between the six Ln‐transformed metallic elements. Correlation coefficients and the heatmap were presented. Blue circles represent a positive correlation. * P < 0.05, ** P < 0.01, *** P < 0.001.

MCN-20-e13682-s001.pdf (192.1KB, pdf)

Supplementary Figure 2. LASSO regression analysis diagram. Covariates including ethnicity, pre‐pregnancy BMI level, education level, parity, household income level, PIH, and six metallic elements were included in the LASSO regression model for analysis. (a) The cross‐validation results. (b) LASSO coefficient profiles of the 12 variables.

MCN-20-e13682-s002.pdf (255.2KB, pdf)

Supplementary Figure 3. The effect of individual metallic element exposure on PTB risk by using the BKMR model. Univariate dose‐response relationships of each metallic element (95% CI) with other exposures were fixed at their 50th percentile. The results were adjusted by ethnicity, pre‐pregnancy BMI level, education level, parity, annual household income level, and PIH.

MCN-20-e13682-s004.pdf (247.6KB, pdf)

Supplementary Figure 4. The bivariate cross‐section effects of the exposure‐response relationship of an individual metallic element when another element was fixed at P25, P50, and P75.

MCN-20-e13682-s005.pdf (22.4KB, pdf)

Supporting information.

MCN-20-e13682-s003.docx (49.8KB, docx)

ACKNOWLEDGEMENTS

The authors would like to thank the participants and the medical staff of the International Peace Maternal and Child Hospital (IPMCH). This work is supported by the National Natural Science Foundation of China (82088102), CAMS Innovation Fund for Medical Sciences (2019‐I2M‐5‐064), Collaborative Innovation Program of Shanghai Municipal Health Commission (2020CXJQ01), Key Discipline Construction Project (2023‐2025) of Three‐Year Initiative Plan for Strengthening Public Health System Construction in Shanghai (GWVI‐11.1‐35), Shanghai Clinical Research Center for Gynecological Diseases (22MC1940200), Shanghai Urogenital System Diseases Research Center (2022ZZ01012) and Shanghai Frontiers Science Research Center of Reproduction and Development.

Wu, T. , Luo, C. , Li, T. , Zhang, C. , Chen, H.‐X. , Mao, Y.‐T. , Wu, Y.‐T. , & Huang, H.‐F. (2024). Effects of exposure to multiple metallic elements in the first trimester of pregnancy on the risk of preterm birth. Maternal & Child Nutrition, 20, e13682. 10.1111/mcn.13682

Ting Wu and Chuan Luo should be considered joint first authors.

Contributor Information

Yan‐Ting Wu, Email: yanting_wu@163.com.

He‐Feng Huang, Email: huanghefg@hotmail.com.

DATA AVAILABILITY STATEMENT

The data supporting this study's findings are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

REFERENCES

  1. Ashrap, P. , Watkins, D. J. , Mukherjee, B. , Boss, J. , Richards, M. J. , Rosario, Z. , Vélez‐Vega, C. M. , Alshawabkeh, A. , Cordero, J. F. , & Meeker, J. D. (2020). Maternal blood metal and metalloid concentrations in association with birth outcomes in Northern Puerto Rico. Environment International, 138, 105606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Cabrera‐Rodríguez, R. , Luzardo, O. P. , González‐Antuña, A. , Boada, L. D. , Almeida‐González, M. , Camacho, M. , Zumbado, M. , Acosta‐Dacal, A. C. , Rial‐Berriel, C. , & Henríquez‐Hernández, L. A. (2018). Occurrence of 44 elements in human cord blood and their association with growth indicators in newborns. Environment International, 116, 43–51. [DOI] [PubMed] [Google Scholar]
  3. Callan, A. C. , Hinwood, A. L. , Ramalingam, M. , Boyce, M. , Heyworth, J. , McCafferty, P. , & Odland, J. Ø. (2013). Maternal exposure to metals—Concentrations and predictors of exposure. Environmental Research, 126, 111–117. [DOI] [PubMed] [Google Scholar]
  4. Chen, F. , Demers, L. M. , Vallyathan, V. , Ding, M. , Lu, Y. , Castranova, V. , & Shi, X. (1999). Vanadate induction of NF‐κB involves IκB Kinase β and SAPK/ERK Kinase 1 in macrophages. Journal of Biological Chemistry, 274(29), 20307–20312. [DOI] [PubMed] [Google Scholar]
  5. Chen, H. , Zhang, H. , Wang, X. , Wu, Y. , Zhang, Y. , Chen, S. , Zhang, W. , Sun, X. , Zheng, T. , Xia, W. , Xu, S. , & Li, Y. (2023). Prenatal arsenic exposure, arsenic metabolism and neurocognitive development of 2‐year‐old children in low‐arsenic areas. Environment International, 174, 107918. [DOI] [PubMed] [Google Scholar]
  6. Chen, S. , Jin, X. , Shan, Z. , Li, S. , Yin, J. , Sun, T. , Luo, C. , Yang, W. , Yao, P. , Yu, K. , Zhang, Y. , Cheng, Q. , Cheng, J. , Bao, W. , & Liu, L. (2017). Inverse association of plasma chromium levels with newly diagnosed type 2 diabetes: A case‐control study. Nutrients, 9(3), 294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen, X. , Li, Y. , Zhang, B. , Zhou, A. , Zheng, T. , Huang, Z. , Pan, X. , Liu, W. , Liu, H. , Jiang, Y. , Sun, X. , Hu, C. , Xing, Y. , Xia, W. , & Xu, S. (2018). Maternal exposure to nickel in relation to preterm delivery. Chemosphere, 193, 1157–1163. [DOI] [PubMed] [Google Scholar]
  8. Deyssenroth, M. A. , Gennings, C. , Liu, S. H. , Peng, S. , Hao, K. , Lambertini, L. , Jackson, B. P. , Karagas, M. R. , Marsit, C. J. , & Chen, J. (2018). Intrauterine multi‐metal exposure is associated with reduced fetal growth through modulation of the placental gene network. Environment International, 120, 373–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ehrlich, V. A. , Nersesyan, A. K. , Hoelzl, C. , Ferk, F. , Bichler, J. , Valic, E. , Schaffer, A. , Schulte‐Hermann, R. , Fenech, M. , Wagner, K. H. , & Knasmüller, S. (2008). Inhalative exposure to vanadium pentoxide causes DNA damage in workers: Results of a multiple end point study. Environmental Health Perspectives, 116(12), 1689–1693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Freire, C. , Amaya, E. , Gil, F. , Fernández, M. F. , Murcia, M. , Llop, S. , Andiarena, A. , Aurrekoetxea, J. , Bustamante, M. , Guxens, M. , Ezama, E. , Fernández‐Tardón, G. , & Olea, N. (2018). Prenatal co‐exposure to neurotoxic metals and neurodevelopment in preschool children: The Environment and Childhood (INMA) Project. Science of the Total Environment, 621, 340–351. [DOI] [PubMed] [Google Scholar]
  11. Freire, C. , Amaya, E. , Gil, F. , Murcia, M. , LLop, S. , Casas, M. , Vrijheid, M. , Lertxundi, A. , Irizar, A. , Fernández‐Tardón, G. , Castro‐Delgado, R. V. , Olea, N. , & Fernández, M. F. (2019). Placental metal concentrations and birth outcomes: The Environment and Childhood (INMA) project. International Journal of Hygiene and Environmental Health, 222(3), 468–478. [DOI] [PubMed] [Google Scholar]
  12. Green, E. S. , & Arck, P. C. (2020). Pathogenesis of preterm birth: Bidirectional inflammation in mother and fetus. Seminars in Immunopathology, 42(4), 413–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Guo, Y. , Huo, X. , Li, Y. , Wu, K. , Liu, J. , Huang, J. , Zheng, G. , Xiao, Q. , Yang, H. , & Wang, Y. (2010). Monitoring of lead, cadmium, chromium and nickel in placenta from an e‐waste recycling town in China. Science of the Total Environment, 408(16), 3113–3117. [DOI] [PubMed] [Google Scholar]
  14. Hirst, J. J. , Hirst, J. J. , Teixeira, F. J. , Zakar, T. , & Olson, D. M. (1995). Prostaglandin endoperoxide‐H synthase‐1 and ‐2 messenger ribonucleic acid levels in human amnion with spontaneous labor onset. Journal of Clinical Endocrinology & Metabolism, 80(2), 517–523. [DOI] [PubMed] [Google Scholar]
  15. Howe, C. G. , Margetaki, K. , Vafeiadi, M. , Roumeliotaki, T. , Karachaliou, M. , Kogevinas, M. , McConnell, R. , Eckel, S. P. , Conti, D. V. , Kippler, M. , Farzan, S. F. , & Chatzi, L. (2021). Prenatal metal mixtures and child blood pressure in the Rhea mother‐child cohort in Greece. Environmental Health, 20(1), 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hu, J. , Xia, W. , Pan, X. , Zheng, T. , Zhang, B. , Zhou, A. , Buka, S. L. , Bassig, B. A. , Liu, W. , Wu, C. , Peng, Y. , Li, J. , Zhang, C. , Liu, H. , Jiang, M. , Wang, Y. , Zhang, J. , Huang, Z. , Zheng, D. , … Xu, S. (2017). Association of adverse birth outcomes with prenatal exposure to vanadium: A population‐based cohort study. The Lancet. Planetary Health, 1(6), e230–e241. [DOI] [PubMed] [Google Scholar]
  17. Huang, H. , Wei, L. , Chen, X. , Zhang, R. , Su, L. , Rahman, M. , Golam Mostofa, M. , Qamruzzaman, Q. , Zhao, Y. , Yu, H. , Wei, Y. , Christiani, D. C. , & Chen, F. (2021). Cord serum elementomics profiling of 56 elements depicts risk of preterm birth: Evidence from a prospective birth cohort in rural Bangladesh. Environment International, 156, 106731. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Issah, I. , Duah, M. S. , Arko‐Mensah, J. , Bawua, S. A. , Agyekum, T. P. , & Fobil, J. N. (2024). Exposure to metal mixtures and adverse pregnancy and birth outcomes: A systematic review. Science of the Total Environment, 908, 168380. [DOI] [PubMed] [Google Scholar]
  19. Jalali, L. M. , & Koski, K. G. (2018). Amniotic fluid minerals, trace elements, and prenatal supplement use in humans emerge as determinants of fetal growth. Journal of Trace Elements in Medicine and Biology, 50, 139–145. [DOI] [PubMed] [Google Scholar]
  20. Käkelä, R. , Käkelä, A. , & Hyvärinen, H. (1999). Effects of nickel chloride on reproduction of the rat and possible antagonistic role of selenium. Comparative Biochemistry and Physiology. Part C, Pharmacology, Toxicology & Endocrinology, 123(1), 27–37. [DOI] [PubMed] [Google Scholar]
  21. Khanam, R. , Kumar, I. , Oladapo‐Shittu, O. , Twose, C. , Islam, A. A. , Biswal, S. S. , Raqib, R. , & Baqui, A. H. (2021). Prenatal environmental metal exposure and preterm birth: A scoping review. International Journal of Environmental Research and Public Health, 18(2), 573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Korbecki, J. , Baranowska‐Bosiacka, I. , Gutowska, I. , & Chlubek, D. (2015). Vanadium compounds as pro‐inflammatory agents: Effects on cyclooxygenases. International Journal of Molecular Sciences, 16(6), 12648–12668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li, C. , Wu, C. , Zhang, J. , Li, Y. , Zhang, B. , Zhou, A. , Liu, W. , Chen, Z. , Li, R. , Cao, Z. , Xia, W. , & Xu, S. (2021). Associations of prenatal exposure to vanadium with early‐childhood growth: A prospective prenatal cohort study. Journal of Hazardous Materials, 411, 125102. [DOI] [PubMed] [Google Scholar]
  24. Li, H. , Zhou, D. , Zhang, Q. , Feng, C. , Zheng, W. , He, K. , & Lan, Y. (2013). Vanadium exposure‐induced neurobehavioral alterations among Chinese workers. Neurotoxicology, 36, 49–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Li, X. , Zhu, Y. , Yin, J. , Li, B. , Li, P. , Cao, B. , Wang, Q. , Xu, J. , & Liu, L. (2022). Inverse association of plasma vanadium concentrations with gestational diabetes mellitus. Nutrients, 14(7), 1415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Li, X. T. , Yu, P. F. , Gao, Y. , Guo, W. H. , Wang, J. , Liu, X. , Gu, A. H. , Ji, G. X. , Dong, Q. , Wang, B. S. , Cao, Y. , Zhu, B. L. , & Xiao, H. (2017). Association between plasma metal levels and diabetes risk: A case‐control study in China. Biomedical and Environmental Sciences, 30(7), 482–491. [DOI] [PubMed] [Google Scholar]
  27. Li, Z. , Liang, C. , Xia, X. , Huang, K. , Yan, S. , Tao, R. , Pan, W. , Sheng, J. , Tao, Y. , Xiang, H. , Hao, J. , Wang, Q. , Tong, S. , & Tao, F. (2019). Association between maternal and umbilical cord serum cobalt concentration during pregnancy and the risk of preterm birth: The Ma'anshan birth cohort (MABC) study. Chemosphere, 218, 487–492. [DOI] [PubMed] [Google Scholar]
  28. Li, Z. , Long, T. , Wang, R. , Feng, Y. , Hu, H. , Xu, Y. , Wei, Y. , Wang, F. , Guo, H. , Zhang, X. , & He, M. (2021). Plasma metals and cancer incidence in patients with type 2 diabetes. Science of the Total Environment, 758, 143616. [DOI] [PubMed] [Google Scholar]
  29. Liang, C. , Wu, X. , Huang, K. , Yan, S. , Li, Z. , Xia, X. , Pan, W. , Sheng, J. , Tao, R. , Tao, Y. , Xiang, H. , Hao, J. , Wang, Q. , Tong, S. , & Tao, F. (2020). Domain‐ and sex‐specific effects of prenatal exposure to low levels of arsenic on children's development at 6 months of age: Findings from the Ma'anshan birth cohort study in China. Environment International, 135, 105112. [DOI] [PubMed] [Google Scholar]
  30. Lin, C.‐C. , Tsweng, G. J. , Lee, C. F. , Chen, B. H. , & Huang, Y. L. (2016). Magnesium, zinc, and chromium levels in children, adolescents, and young adults with type 1 diabetes. Clinical Nutrition, 35(4), 880–884. [DOI] [PubMed] [Google Scholar]
  31. Lin, C. Y. C. T. H. (1998). Nickel toxicity to human term placenta: In vitro study on lipid peroxidation. Journal of Toxicology and Environmental Health. Part A, 54(1), 37–47. [DOI] [PubMed] [Google Scholar]
  32. Liu, J. , Ruan, F. , Cao, S. , Li, Y. , Xu, S. , & Xia, W. (2022). Associations between prenatal multiple metal exposure and preterm birth: Comparison of four statistical models. Chemosphere, 289, 133015. [DOI] [PubMed] [Google Scholar]
  33. Ma, J. , Geng, S. , Sun, Q. , Zhang, X. , Han, L. , Yao, X. , Zhang, B. , Zhu, L. , & Wen, J. (2023). Exposure to metal mixtures and young children's growth and development: A biomonitoring‐based study in Eastern China. Ecotoxicology and Environmental Safety, 268, 115726. [DOI] [PubMed] [Google Scholar]
  34. Paternain, J. L. , Domingo, J. L. , Gómez, M. , Ortega, A. , & Corbella, J. (1990). Developmental toxicity of vanadium in mice after oral administration. Journal of Applied Toxicology, 10(3), 181–186. [DOI] [PubMed] [Google Scholar]
  35. Rodríguez‐Barranco, M. , Lacasaña, M. , Aguilar‐Garduño, C. , Alguacil, J. , Gil, F. , González‐Alzaga, B. , & Rojas‐García, A. (2013). Association of arsenic, cadmium and manganese exposure with neurodevelopment and behavioural disorders in children: A systematic review and meta‐analysis. Science of the Total Environment, 454‐455, 562–577. [DOI] [PubMed] [Google Scholar]
  36. Romero, R. , Dey, S. K. , & Fisher, S. J. (2014). Preterm labor: One syndrome, many causes. Science, 345(6198), 760–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ścibior, A. , Llopis, J. , Dobrakowski, P. P. , & Męcik‐Kronenberg, T. (2023). CNS‐related effects caused by vanadium at realistic exposure levels in humans: A comprehensive overview supplemented with selected animal studies. International Journal of Molecular Sciences, 24(10), 9004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ścibior, A. , Pietrzyk, Ł. , Plewa, Z. , & Skiba, A. (2020). Vanadium: Risks and possible benefits in the light of a comprehensive overview of its pharmacotoxicological mechanisms and multi‐applications with a summary of further research trends. Journal of Trace Elements in Medicine and Biology, 61, 126508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ścibior, A. , Zaporowska, H. , Ostrowski, J. , & Banach, A. (2006). Combined effect of vanadium(V) and chromium(III) on lipid peroxidation in liver and kidney of rats. Chemico‐biological Interactions, 159(3), 213–222. [DOI] [PubMed] [Google Scholar]
  40. Shan, Z. , Chen, S. , Sun, T. , Luo, C. , Guo, Y. , Yu, X. , Yang, W. , Hu, F. B. , & Liu, L. (2016). U‐shaped association between plasma manganese levels and type 2 diabetes. Environmental Health Perspectives, 124(12), 1876–1881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Signes‐Pastor, A. J. , Notario‐Barandiaran, L. , Guill, M. , Madan, J. , Baker, E. , Jackson, B. , & Karagas, M. R. (2023). Prenatal exposure to metal mixtures and lung function in children from the New Hampshire birth cohort study. Environmental Research, 238(Pt 2), 117234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Skogheim, T. S. , Weyde, K. V. F. , Engel, S. M. , Aase, H. , Surén, P. , Øie, M. G. , Biele, G. , Reichborn‐Kjennerud, T. , Caspersen, I. H. , Hornig, M. , Haug, L. S. , & Villanger, G. D. (2021). Metal and essential element concentrations during pregnancy and associations with autism spectrum disorder and attention‐deficit/hyperactivity disorder in children. Environment International, 152, 106468. [DOI] [PubMed] [Google Scholar]
  43. Soares, S. S. , Martins, H. , Duarte, R. O. , Moura, J. J. G. , Coucelo, J. , Gutiérrez‐Merino, C. , & Aureliano, M. (2007). Vanadium distribution, lipid peroxidation and oxidative stress markers upon decavanadate in vivo administration. Journal of Inorganic Biochemistry, 101(1), 80–88. [DOI] [PubMed] [Google Scholar]
  44. Soomro, M. H. , Baiz, N. , Huel, G. , Yazbeck, C. , Botton, J. , Heude, B. , Bornehag, C. G. , & Annesi‐Maesano, I. (2019). Exposure to heavy metals during pregnancy related to gestational diabetes mellitus in diabetes‐free mothers. Science of the Total Environment, 656, 870–876. [DOI] [PubMed] [Google Scholar]
  45. Sun, X. , Jiang, Y. , Xia, W. , Jin, S. , Liu, W. , Lin, X. , Liu, H. , Chen, X. , Peng, Y. , Li, H. , Lu, B. , Xu, S. , Li, Y. , & Shen, X. (2018). Association between prenatal nickel exposure and preterm low birth weight: Possible effect of selenium. Environmental Science and Pollution Research, 25(26), 25888–25895. [DOI] [PubMed] [Google Scholar]
  46. Viemann, D. , Schmidt, M. , Tenbrock, K. , Schmid, S. , Müller, V. , Klimmek, K. , Ludwig, S. , Roth, J. , & Goebeler, M. (2007). The contact allergen nickel triggers a unique inflammatory and proangiogenic gene expression pattern via activation of nf‐κb and hypoxia‐inducible factor‐1α. The Journal of Immunology, 178(5), 3198–3207. [DOI] [PubMed] [Google Scholar]
  47. Walani, S. R. (2020). Global burden of preterm birth. International Journal of Gynecology & Obstetrics, 150(1), 31–33. [DOI] [PubMed] [Google Scholar]
  48. Wallace, M. E. , Grantz, K. L. , Liu, D. , Zhu, Y. , Kim, S. S. , & Mendola, P. (2016). Exposure to ambient air pollution and premature rupture of membranes. American Journal of Epidemiology, 183(12), 1114–1121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang, B. , Liu, J. , Liu, B. , Liu, X. , & Yu, X. (2018). Prenatal exposure to arsenic and neurobehavioral development of newborns in China. Environment International, 121(Pt 1), 421–427. [DOI] [PubMed] [Google Scholar]
  50. Wang, H. , Li, J. , Zhang, X. , Zhu, P. , Hao, J. H. , Tao, F. B. , & Xu, D. X. (2018). Maternal serum arsenic level during pregnancy is positively associated with adverse pregnant outcomes in a Chinese population. Toxicology and Applied Pharmacology, 356, 114–119. [DOI] [PubMed] [Google Scholar]
  51. Wang, Y. , Zhang, P. , Chen, X. , Wu, W. , Feng, Y. , Yang, H. , Li, M. , Xie, B. , Guo, P. , Warren, J. L. , Shi, X. , Wang, S. , & Zhang, Y. (2019). Multiple metal concentrations and gestational diabetes mellitus in Taiyuan, China. Chemosphere, 237, 124412. [DOI] [PubMed] [Google Scholar]
  52. Wang, Z. , Huang, S. , Zhang, W. , Zeng, X. , Chu, C. , Li, Q. , Cui, X. , Wu, Q. , Dong, G. , Huang, J. , Liu, L. , Tan, W. , Shang, X. , Kong, M. , & Deng, F. (2022). Chemical element concentrations in cord whole blood and the risk of preterm birth for pregnant women in Guangdong, China. Ecotoxicology and Environmental Safety, 247, 114228. [DOI] [PubMed] [Google Scholar]
  53. Xu, R. , Meng, X. , Pang, Y. , An, H. , Wang, B. , Zhang, L. , Ye, R. , Ren, A. , Li, Z. , & Gong, J. (2022). Associations of maternal exposure to 41 metals/metalloids during early pregnancy with the risk of spontaneous preterm birth: Does oxidative stress or DNA methylation play a crucial role? Environment International, 158, 106966. [DOI] [PubMed] [Google Scholar]
  54. Yao, L. , Liu, L. , Dong, M. , Yang, J. , Zhao, Z. , Chen, J. , Lv, L. , Wu, Z. , Wang, J. , Sun, X. , Self, S. , & Bhatti, P. (2023). Trimester‐specific prenatal heavy metal exposures and sex‐specific postpartum size and growth. Journal of Exposure Science & Environmental Epidemiology, 33(6), 895–902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Ye, J. , Ding, M. , Zhang, X. , Rojanasakul, Y. , Nedospasov, S. , Vallyathan, V. , Castranova, V. , & Shi, X. (1999). Induction of TNFalpha in macrophages by vanadate is dependent on activation of transcription factor NF‐kappaB and free radical reactions. Molecular and Cellular Biochemistry, 198(1–2), 193–200. [DOI] [PubMed] [Google Scholar]
  56. Yu, L. , Liu, W. , Wang, X. , Ye, Z. , Tan, Q. , Qiu, W. , Nie, X. , Li, M. , Wang, B. , & Chen, W. (2022). A review of practical statistical methods used in epidemiological studies to estimate the health effects of multi‐pollutant mixture. Environmental Pollution, 306, 119356. [DOI] [PubMed] [Google Scholar]
  57. Yuan, Y. , Xiao, Y. , Yu, Y. , Liu, Y. , Feng, W. , Qiu, G. , Wang, A. , Liu, B. , Wang, J. , Zhou, L. , Liu, K. , Xu, X. , Yang, H. , Li, X. , Qi, L. , Zhang, X. , He, M. , Hu, F. , Pan, A. , & Wu, T. (2018). Associations of multiple plasma metals with incident type 2 diabetes in Chinese adults: The Dongfeng‐Tongji Cohort. Environmental Pollution, 237, 917–925. [DOI] [PubMed] [Google Scholar]
  58. Yüksel, B. , Arıca, E. , & Söylemezoğlu, T. (2021). Assessing reference levels of nickel and chromium in cord blood, maternal blood and placenta specimens from Ankara, Turkey. Journal of the Turkish‐German Gynecological Association, 22(3), 187–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhang, W. , Chen, H. , Zeng, Q. , Xu, S. , Xia, W. , & Li, Y. (2021). Prenatal and postnatal exposure to vanadium and the immune function of children. Journal of Trace Elements in Medicine and Biology, 67, 126787. [DOI] [PubMed] [Google Scholar]
  60. Zhu, G. , Zheng, T. , Xia, C. , Qi, L. , Papandonatos, G. D. , Ming, Y. , Zeng, Z. , Zhang, X. , Zhang, H. , & Li, Y. (2021). Plasma levels of trace element status in early pregnancy and the risk of gestational diabetes mellitus: A nested case‐control study. Journal of Trace Elements in Medicine and Biology, 68, 126829. [DOI] [PubMed] [Google Scholar]
  61. Zhuang, W. , & Zhou, F. (2021). Distribution, source and pollution assessment of heavy metals in the surface sediments of the Yangtze River Estuary and its adjacent East China Sea. Marine Pollution Bulletin, 164, 112002. [DOI] [PubMed] [Google Scholar]
  62. Zwolak, I. (2014). Vanadium carcinogenic, immunotoxic and neurotoxic effects: A review of in vitro studies. Toxicology Mechanisms and Methods, 24(1), 1–12. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figure 1. Spearman's rank correlation between the six Ln‐transformed metallic elements. Correlation coefficients and the heatmap were presented. Blue circles represent a positive correlation. * P < 0.05, ** P < 0.01, *** P < 0.001.

MCN-20-e13682-s001.pdf (192.1KB, pdf)

Supplementary Figure 2. LASSO regression analysis diagram. Covariates including ethnicity, pre‐pregnancy BMI level, education level, parity, household income level, PIH, and six metallic elements were included in the LASSO regression model for analysis. (a) The cross‐validation results. (b) LASSO coefficient profiles of the 12 variables.

MCN-20-e13682-s002.pdf (255.2KB, pdf)

Supplementary Figure 3. The effect of individual metallic element exposure on PTB risk by using the BKMR model. Univariate dose‐response relationships of each metallic element (95% CI) with other exposures were fixed at their 50th percentile. The results were adjusted by ethnicity, pre‐pregnancy BMI level, education level, parity, annual household income level, and PIH.

MCN-20-e13682-s004.pdf (247.6KB, pdf)

Supplementary Figure 4. The bivariate cross‐section effects of the exposure‐response relationship of an individual metallic element when another element was fixed at P25, P50, and P75.

MCN-20-e13682-s005.pdf (22.4KB, pdf)

Supporting information.

MCN-20-e13682-s003.docx (49.8KB, docx)

Data Availability Statement

The data supporting this study's findings are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Articles from Maternal & Child Nutrition are provided here courtesy of Wiley

RESOURCES