Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 18.
Published in final edited form as: Sci Total Environ. 2024 Apr 7;928:172295. doi: 10.1016/j.scitotenv.2024.172295

Associations of maternal blood metal concentrations with plasma eicosanoids among pregnant women in Puerto Rico

Christine Kim a, Amber L Cathey a, Seonyoung Park a, Deborah J Watkins a, Bhramar Mukherjee b, Zaira Y Rosario-Pabón c, Carmen M Vélez-Vega c, Akram N Alshawabkeh d, José F Cordero e, John D Meeker a,*
PMCID: PMC12912272  NIHMSID: NIHMS2138766  PMID: 38588744

Abstract

Background/aim:

Heavy metals are known to induce oxidative stress and inflammation, and the association between metal exposure and adverse birth outcomes is well established. However, there lacks research on biomarker profiles linking metal exposures and adverse birth outcomes. Eicosanoids are lipid molecules that regulate inflammation in the body, and there is growing evidence that suggests associations between plasma eicosanoids and pregnancy outcomes. Eicosanoids may aid our understanding of etiologic birth pathways. Here, we assessed associations between maternal blood metal concentrations with eicosanoid profiles among 654 pregnant women in the Puerto Rico PROTECT birth cohort.

Methods:

We measured concentrations of 11 metals in whole blood collected at median 18 and 26 weeks of pregnancy, and eicosanoid profiles measured in plasma collected at median 26 weeks. Multivariable linear models were used to regress eicosanoids on metals concentrations. Effect modification by infant sex was explored using interaction terms.

Results:

A total of 55 eicosanoids were profiled. Notably, 12-oxoeicosatetraenoic acid (12-oxoETE) and 15-oxoeicosatetraenoic acid (15-oxoETE), both of which exert inflammatory activities, had the greatest number of significant associations with metal concentrations. These eicosanoids were associated with increased concentrations of Cu, Mn, and Zn, and decreased concentrations of Cd, Co, Ni, and Pb, with the strongest effect sizes observed for 12-oxoETE and Pb (β:−33.5,95 %CI:−42.9,−22.6) and 15-oxoETE and Sn (β:43.2,95 %CI:11.4,84.1). Also, we observed differences in metals-eicosanoid associations by infant sex. Particularly, Cs and Mn had the most infant sex-specific significant associations with eicosanoids, which were primarily driven by female fetuses. All significant sex-specific associations with Cs were inverse among females, while significant sex-specific associations with Mn among females were positive within the cyclooxygenase group but inverse among the lipoxygenase group.

Conclusion:

Certain metals were significantly associated with eicosanoids that are responsible for regulating inflammatory responses. Eicosanoid-metal associations may suggest a role for eicosanoids in mediating metal-induced adverse birth outcomes.

Keywords: Metals, Eicosanoids, Pregnancy, Puerto Rico

Graphical Abstract

graphic file with name nihms-2138766-f0001.jpg

1. Introduction

Eicosanoids are lipid molecules that play important roles in multiple processes in the human body, such as immune responses (Tassoni et al., 2008). Eicosanoids are metabolized from polyunsaturated fatty acids, including arachidonic acid (AA), through multiple enzymatic pathways, such as cyclooxygenases (COX), lipoxygenases (LOX), and cytochrome P450 (CYP450) (Funk, 2001; Yuan et al., 2018). Because of their roles in regulating immune responses, the dysregulation of eicosanoids can result in chronic inflammation and lead to the progression of various pathologies (Wang et al., 2021).

Eicosanoids regulate immune responses by tightly interacting with immunological signaling molecules (Yuan et al., 2018), and they are also involved in processes of the human reproductive system, including ovulation, corpus luteum function, fertilization, and control of the uterus in parturition (Novy and Liggins, 1980; Strauss and FitzGerald, 2019). Consistently, there is increasing evidence that demonstrates a link between maternal plasma eicosanoids and birth outcomes (Welch et al., 2020; Aung et al., 2019; Meagher and FitzGerald, 1993; Gouveia-Figueira et al., 2017; Long et al., 2016). For example, Welch et al. observed that eicosanoids metabolized from AA via the CYP450 and LOX pathways are positively associated with small for gestational age infants (n = 90) from the LIFECODES prospective birth cohort study (Boston, Massachusetts) (Welch et al., 2020). Moreover, from different subset of women within the same birth cohort study, Aung et al. demonstrated that both pro- and anti-inflammatory eicosanoids were associated with increased risk for overall and spontaneous preterm birth (n = 173), highlighting the importance of evaluating these eicosanoids collectively to predict preterm birth (Aung et al., 2019). Together, these studies suggest that changes in maternal eicosanoid profiles can lead to pregnancy complications.

Pregnant women are readily exposed to various metals from all possible routes, including ingestion. Several previous studies have demonstrated associations between heavy metal exposures and birth outcomes, such as gestational age at delivery, birth weight, and birth length (Liu et al., 2018; Cheng et al., 2017a; Xia et al., 2016; Cheng et al., 2017b; Wai et al., 2017). Consistently, Ashrap et al. showed associations between maternal blood concentrations of Pb, Hg, Mn, and Zn and higher odds of preterm delivery in the Puerto Rico PROTECT cohort (n = 812) (Ashrap et al., 2020a). Although eicosanoids play important roles in the human biological system and have the potential to serve as a powerful tool to predict adverse birth outcomes, there is a paucity of research on these potentially important biologic pathways in the relationship between metal exposures and adverse birth outcomes.

The overarching goal of this study was to investigate associations between metal exposures and eicosanoid profiles among pregnant women in the ongoing PROTECT birth cohort taking place in Puerto Rico and identify eicosanoids that may provide a better understanding of the relationship between metal exposures and adverse birth outcomes. We believe this study will further contribute to identifying mechanisms by which metal exposures induce preterm birth or other adverse pregnancy outcomes and could inform future efforts aimed at developing diagnostic tools or interventions.

2. Methods

We used data from a subset of the Puerto Rico PROTECT birth cohort. The PROTECT cohort began recruitment in 2010 in the Northern Karst Region of Puerto Rico through funding from the National Institute of Environmental Health Sciences Superfund Research Program. Briefly, women were recruited at approximately 14 ± 2 weeks of gestation from seven prenatal clinics and hospitals throughout Northern Puerto Rico during 2010–2019. Inclusion criteria for recruitment included: participant age between 18 and 40 years; residence in the Northern Karst aquifer region; disuse of oral contraceptives three months before pregnancy; disuse of in vitro fertilization; and no indication in medical records for major obstetrical complications, including pre-existing diabetes.

2.1. Blood biomarker measurements

Blood samples were collected at up to two study visits per participant (median 18 and 26 weeks gestation). Samples were collected in metal-free tubes, divided into aliquots, frozen at −80C, and shipped on dry ice to NSF International (Ann Arbor, MI, USA) for analysis.

2.1.1. Blood metals

Blood concentrations of metals were measured in samples using a Thermo Fisher (Waltham, MA, USA) ICAPRQ inductively coupled plasma mass spectrometry (ICPMS) and CETAC ASX-520 autosampler, as described previously (Kim et al., 2018). Standards of known purity and identity were used during the preparation of the calibration, quality control, and internal standards. The ICPMS was calibrated with a blank and a minimum of 4 standards for each element of interest. The calibration curve response versus concentration was evaluated for the goodness of fit. All validated analyte correlation coefficients (R2) were ≥ 0.995. Initial checks of calibration curves were conducted at three points within the curve, and continuing calibration checks and blanks were utilized throughout the analytical run after every 10 samples to ensure maintenance of acceptable performance. Twenty-one metals were included in the analytical panel including arsenic (As), barium (Ba), Beryllium (Be), cadmium (Cd), cobalt (Co), chromium (Cr), cesium (Cs), copper (Cu), mercury (Hg), manganese (Mn), molybdenum (Mo), nickel (Ni), lead (Pb), platinum (Pt), antimony (Sb), tin (Sn), thallium (Tl), uranium (U), vanadium, (V), tungsten, (W), and zinc (Zn). Metals which were measured above the limit of detection for <60 % of samples were removed from the analysis (As, Ba, Be, Cr, Pt, Sb, Tl, U, V, and W). Thus, the final set of metals included Cd, Co, Cs, Cu, Hg, Mn, Mo, Ni, Pb, Sn, and Zn, all of which were measured in ng/mL except for Pb which was measured in ug/dL. Metal concentrations below the limit of detection (LOD) were imputed with LOD/sqrt2.

2.1.2. Plasma eicosanoids

Eicosanoids from plasma samples at the later study visit (median 26 weeks) were quantified using a 6490 triple quadruple mass spectrometer (Agilent, New Castle, DE, USA). Sixty-three eicosanoids were included in the analytical panel, 55 of which were measured in enough samples to be included our analysis. Table 2 shows each included analyte according to its subclass and abbreviations for each analyte. Details on extraction of eicosanoids from plasma, HPLC/MS/MS analysis, and quality control have been previously described (Park et al., 2023). Briefly, plasma samples were diluted and spiked with internal standards, and eicosanoids were extracted using Starta-X Polymeric SPE columns (Phenomenex, Inc., Torrance, CA). All eicosanoid standards were purchased from Cayman Chemical (Ann Arbor, MI). Reverse phase chromatography was employed using a Luna C18 LC column, followed by mass spectrometry using an Agilent 6490 Triple Quadrupole system. Eicosanoid quantification was achieved by comparing peak areas of the analyte of interest with corresponding internal standards. Sequential dilution of each internal standard was done in duplicate to establish linearity, identify the lower limit of detection, and estimate the coefficient of variation. Drift in measurements over time, as well as batch-to-batch variability, was assessed by running a pool of study samples at the beginning of each batch and then after every ten samples during mass spectrometry. All eicosanoids were measured in nmol/L except for arachidonic acid, docosahexaenoic acid, eicosapentaenoic acid, linoleic acid, and α-linoleic acid, which were measured in μmol/L. We utilized machine read values when eicosanoids were measured below the limit of detection, unless the measured concentration was negative, in which case the measurement was removed from the analysis. Imputation with the LOD divided by the square root of two was not utilized to prevent having many data points with identical values, which can adversely affect distributions and standard errors (Schisterman et al., 2006).

Table 2.

A panel of 55 eicosanoids in plasma that we measured according to their subgroups and abbreviations for each eicosanoid among 654 women in PROTECT.

Plasma eicosanoids: Cyclooxygenase pathway Plasma eicosanoids: Lipoxygenase pathway Plasma eicosanoids: Cytochrome P450 pathway Plasma lipid parent compounds
Bicyclo prostaglandin E1 [BCPGE1] Leukotriene B4 [LTB4] 11, 12-Dihydroxy-eicosa-trienoic acid [11, 12-DHET] Arachidonic acid [AA]
Bicyclo prostaglandin E2 [BCPGE2] Leukotriene C4 methyl-ester [LTC4-ME] 11, 12-Epoxy-eicosatrienoic acid [11, 12-EET] α-Linolenic ACID [αLA]
15-Deoxy-12,14-prostaglandin J2 [DeoPGJ2] Leukotriene D4 [LTD4] 12, 13-Dihydroxy-octadece-noic acid [12, 13-DiHOME] Docosahexaenoic acid [DHA]
13,14-Dihydro-15-keto prostaglandin D2 [DiPGD2] Leukotriene E4 [LTE4] 12, 13-Epoxy-octadecenoic acid [12-EpoME] Eicosapentaenoic acid [EPA]
13,14-Dihydro-15-keto prostaglandin e2 [DiPGE2] Resolvin D1 [RVD1] 14, 15-Epoxy-eicosatrienoic acid [14(15)-EET] Linoleic acid [LA]
13,14-Dihydro-15-keto prostaglandin F2 [DiPGF2] Resolvin D2 [RVD2] 16-Hydroxy-eicosatetraenoic acid [16-HETE]
13,14-Dihydro-15-keto prostaglandin J2 [DiPGJ2] 12-Hydroxy-eicosatetraenoic acid [12-HETE] 17-Hydroxy-eicosatetraenoic acid [17-HETE]
Prostaglandin A2 [PGA2] 12-Oxoeicosatetraenoic acid [12-oxoETE] 18-Hydroxy-eicosatetraenoic acid [18-HETE]
Prostaglandin B2 [PGB2] 13-Oxooctadeca-dienoic acid [13-oxoODE] 20-carboxy Arachidonic Acid [20-CAA]
Prostaglandin D2 [PGD2] 13S-Hydroxy-octadecadienoic acid [13S-HODE] 5, 6-Dihydroxy-eicosatrie-noic acid [5, 6-DHET]
Prostaglandin D3 [PGD3] 15-Hydroxy-eicosatetraenoic acid [15-HETE] 5, 6-Epoxy-eicosatrienoic acid [5, 6-EET]
Prostaglandin E1 [PGE1] 15-Oxoeicosatetraenoic acid [15-oxoETE] 8, 9-Dihydroxy-eicosatrie-noic acid [8, 9-DHET]
Prostaglandin E2 [PGE2] 5-Hydroxy-eicosatetraenoic acid [5-HETE] 8, 9-Epoxy-eicosatrienoic acid [8, 9-EET]
Prostaglandin E3 [PGE3] 5-Oxoeicosatetraenoic acid [5-oxoETE] 9, 10-Dihydroxy-octadece-noic acid [9, 10-DiHOME]
Prostaglandin J2 [PGJ2] 8-Hydroxy-eicosatetraenoic [8-HETE] 9-Epoxy-octadecenonic acid [9-EpoME]
Thromboxane B2 [TXB2] acid 9S-Hydroxy-octadecadienoic acid [9S-HODE]
9-Oxooctadeca-dienoic acid [9-oxoODE] 20-Hydroxy-eicosatetraenoic acid [20-HETE]
11-Hydroxy-eicosatetraenoic acid [11-HETE]
Open in a new tab

2.2. Statistical analysis

Our study population initially consisted of 809 women (providing 1098 samples) for whom we had biomarker data on either blood metals or eicosanoids. Various potential covariates were explored among this subset of women: maternal age, education level attained, marital status, employment, annual household income, smoking and exposure to secondhand smoke, alcohol use, parity, pre-pregnancy body mass index (BMI), and infant sex. We employed a forward stepwise procedure to introduce potential covariates into statistical models. Covariates were retained in the models if they resulted in a change in the main effect by at least 10 %. The resulting models were adjusted for continuous maternal age and categorical forms of maternal education, exposure to environmental tobacco smoke, and pre-pregnancy BMI. After removing women with incomplete data on selected covariates, our final sample consisted of 654 women.

Distributions of blood metal concentrations were assessed at each study visit. Intraclass correlation coefficients (ICCs), which describe between- and within-individual variability in biomarker concentrations across study visits, were also calculated. ICCs range between 0 (no reproducibility) and 1 (perfect reproducibility) and reflect a degree of reliability (<0.40: poor reliability, 0.40 < 0.75: moderate to good reliability, >0.75: excellent reliability) (Rosner, 2015). ICC estimates and their 95 % confidence intervals were calculated in R (version 4.0.4) using the package ICC. All blood biomarkers were log-normally distributed and thus were natural log-transformed for all subsequent analyses.

We used multiple linear regression to estimate associations between individual blood metal concentrations across pregnancy (averaged between two visits) and eicosanoid concentrations. We also employed sensitivity analyses which assessed associations between visit-specific metals concentrations and eicosanoids to explore possible windows of susceptibility to metals exposure. The significance level was set to alpha = 0.05. We utilized the Benjamini and Hochberg method to account for false discovery due to multiple comparisons (FDR q-value), using each metal exposure biomarker as a family of tests (Benjamini, 1995).

Subsequently, we conducted sensitivity analyses to explore possible effect modification by infant sex on the associations between metals and eicosanoids by using metal*sex interaction terms in regression models. All reported results can be interpreted as the percent change in blood eicosanoids concentration associated with an interquartile range (IQR) increase in blood metal concentration.

3. Results

The demographic and health characteristics of the study participants are presented in Table 1. The mean age of the participants at the time of enrollment was 26.3 years (standard deviation of 5.5 years). A majority of the participants had earned tertiary education (71.6 %), had an annual household income of less than $50,000 (78.4 %), and approximately 55 % of the participants were employed at the time of enrollment. About half of the study participants had a pre-pregnancy BMI of <25 kg/m2 (46.3 %) and most of the participants did not smoke (97.9 %) or drink alcohol (95.6 %) during pregnancy. Fetal sex was evenly distributed between female (n = 148) and male infants (n = 149).

Table 1.

Demographic and other relevant health information on 654 women in PROTECT.

Mean (SD)
Maternal age (years) 26.3 (5.5)
N %
Maternal age (years)
 18–24 272 41.6 %
 25–29 201 30.7 %
 30–34 111 17.0 %
 35–41 70 10.7 %
 Missing 0
Maternal education
 GED or less 182 27.8 %
 Some college 226 34.6 %
 Bachelors or higher 242 37.0 %
 Missing 4 0.6 %
Marital status
 Single 126 19.3 %
 Married 286 43.7 %
 Cohabitating 235 35.9 %
 Missing 7 1.1 %
Currently employed
 No 289 44.2 %
 Yes 360 55.0 %
 Missing 5 0.8 %
Annual household income
 <10k 257 39.3 %
 10k-<30k 144 22.0 %
 30k-<50k 112 17.1 %
 ≥50k 56 8.6 %
 Missing 85 13.0 %
Smoking status
 Never 566 86.5 %
 Ever 74 11.3 %
 Current 14 2.1 %
 Missing 0
Alcohol use
 Never 383 58.6 %
 Yes, before pregnancy 236 36.1 %
 Yes, currently 29 4.4 %
 Missing 6 0.9 %
Number of children
 0 300 44.1 %
 1 212 31.1 %
 2–5 139 20.4 %
 Missing 30 4.4 %
BMI
 (0, 25] 303 46.3 %
 (25, 29.9] 178 27.2 %
 (29.9, 51] 143 21.9 %
 Missing 30 4.6 %
Fetal sex
 Female 264 40.4 %
 Male 259 39.6 %
 Missing 131 20 %
Open in a new tab

Table 2 presents a panel of 63 eicosanoids in plasma that we measured according to their subgroups and abbreviations for each eicosanoid, and Table S1 shows the distributions of the eicosanoid concentrations. Table 3 shows the distributions of maternal blood metal concentrations including Cd, Co, Cs, Cu, Hg, Mn, Mo, Ni, Pb, Sn, and Zn. The majority of metals analyzed were above the LOD in at least 85 % of samples, except for Cd, Ni, and Sn, which were measured above the LOD in 55 % of samples (65 % at visit 1 and 56 % at visit 3; 80 % at visit 1 and 64 % at visit 3; 65 % at visit 1 and 56 % at visit 3, respectively). Most metals showed low to moderate within-person variability, given by ICCs ranging from 0.50 to 0.76. In contrast, several metals showed greater within-person variability, including Co (ICC = 0.21), Mo (ICC = 0.22), Ni (ICC = 0.10), and Sn (ICC = 0.24).

Table 3.

Distributions of blood metal concentrations, in μg/La, among 654 women in PROTECT.

Metal Visit N LOD % > LOD Min P25 Med P75 P90 Max GM SD IQR ICC (95 % CI)
Cadmium (Cd) 1 553 0.1 65 % 0.07 0.07 0.13 0.20 0.29 22.0 0.13 1.87 0.12 0.50 (0.44, 0.56)
3 649 56 % 0.07 0.07 0.11 0.17 0.25 1.34 0.12 1.77 0.10
Cobalt (Co) 1 553 0.2 97 % 0.14 0.30 0.37 0.47 0.58 2.67 0.37 1.46 0.17 0.21 (0.13, 0.29)
3 649 98 % 0.14 0.33 0.40 0.49 0.60 1.51 0.40 1.37 0.16
Cesium (Cs) 1 553 0.04 100 % 0.50 1.02 1.29 1.57 1.84 2.69 1.27 1.35 0.55 0.58 (0.52, 0.63)
3 649 100 % 0.39 0.83 1.06 1.30 1.55 3.89 1.05 1.37 0.46
Copper (Cu) 1 553 9 100 % 761 1392 1533 1730 1927 2847 1549 1.19 338 0.58 (0.52, 0.63)
3 649 100 % 735 1427 1606 1795 2011 3798 1602 1.19 368
Mercury (Hg) 1 553 0.2 97 % 0.14 0.56 0.86 1.32 1.98 4.80 0.84 1.99 0.77 0.73 (0.69, 0.77)
3 649 98 % 0.14 0.58 0.89 1.29 1.89 5.54 0.85 1.93 0.71
Manganese (Mn) 1 553 2 100 % 4.98 8.48 10.4 12.5 14.9 24.2 10.4 1.32 4.05 0.59 (0.54, 0.64)
3 649 100 % 6.27 10.1 12.2 14.5 17.3 48.2 12.3 1.31 4.40
Molybdenum (Mo) 1 340 0.3 92 % 0.21 0.43 0.55 0.73 0.96 8.11 0.55 1.63 0.30 0.22 (0.11, 0.32)
3 381 95 % 0.21 0.41 0.53 0.73 0.89 5.07 0.55 1.56 0.31
Nickel (Ni) 1 553 0.5 80 % 0.35 0.77 1.30 2.10 2.67 22.8 1.19 2.19 1.34 0.10 (0.01, 0.18)
3 649 64 % 0.35 0.35 0.82 1.55 2.41 16.9 0.83 2.20 1.19
Lead (Pb)a 1 553 0.02 100 % 0.07 0.22 0.28 0.37 0.49 1.95 0.29 1.53 0.15 0.76 (0.72, 0.79)
3 649 100 % 0.09 0.22 0.28 0.36 0.48 3.35 0.29 1.55 0.15
Tin (Sn) 1 340 0.1 65 % 0.07 0.07 0.14 0.25 0.86 10.8 0.17 2.64 0.18 0.24 (0.14, 0.35)
3 381 56 % 0.07 0.07 0.12 0.21 0.51 7.21 0.14 2.33 0.13
Zinc (Zn) 1 553 24 100 % 1969 4500 5024 5538 6075 9369 4986 1.18 1038 0.56 (0.51, 0.62)
3 649 100 % 2600 4630 5138 5673 6155 8632 5097 1.17 1043
Open in a new tab

GM: geometric mean; GSD: geometric standard deviation; IQR: interquartile range; ICC: intraclass correlation coefficient; CI: confidence interval.

a

Units for Pb were μg/dL.

The associations between eicosanoids and pregnancy average maternal blood metal concentrations after adjusting for false discovery rate (FDR) adjusted p-values (q-values) (p-value <0.05, q-value <0.1) are shown in Table S2 and Fig. 1. Overall, most eicosanoids had significant negative associations with toxic blood metals (non-essential metals) such as Cd, Hg, and Pb; however, essential metals such as Co, Cu, Mn, Mo, and Zn tended to show positive associations (Fig. 1). In particular, the majority of the parent compounds had significant negative associations with the non-essential metals (Fig. 1). Similarly, the associations between prostaglandins and the essential metals were predominantly in the opposite direction as the associations between prostaglandins and non-essential metals (Fig. 1). It is important to note that 12-oxoETE and 15-oxoETE in the lipoxygenase group had the greatest number of significant associations with maternal blood metal concentrations, and they were consistent in direction (Table S2 and Fig. 1). Interestingly, metals associations with 12-HETE, which is a precursor of 12-oxoETE, tended to be in the opposite direction from the metals associations with 12-oxoETE (Table S2 and Fig. 1). The fewest number of significant associations were found with eicosanoid biomarkers from the cyclooxygenase pathway. Among metals, Pb had the highest number of significant associations with the eicosanoid biomarkers.

Fig. 1.

Fig. 1.

Open in a new tab

Heatmap depicting associations between blood metals, averaged over pregnancy, and eicosanoids. Yellow colored boxes indicate a positive effect estimate while blue colored boxes indicate a negative effect estimate, and the intensity of the color signifies the magnitude of the effect estimate.

Models were adjusted for continuous maternal age and categorical forms of maternal education, exposure to environmental tobacco smoke, and pre-pregnancy BMI.

*p < 0.05.

**p < 0.01.

We also observed infant sex-specific differences in the associations between maternal blood metal concentrations and eicosanoids shown in Fig. 2 and Table S3. Notably, Cs and Mn showed the greatest number of infant sex-specific significant associations with the eicosanoids. Cs demonstrated inverse associations with 11-HETE (cytochrome p450 group), 14,15-EET (cytochrome p450 group), 12-oxoETE (lipoxygenase group), and 15-HETE (lipoxygenase group) among female infants, while corresponding associations among male infants were positive but null. Three eicosanoids in the cyclooxygenase pathway (PGA2, PGD2, PGE3) were positively associated with Mn among only female infants. Additionally, 13S_HODE (lipoxygenase group) was inversely associated with Mn among female infants, but positively among male infants, though neither association reached statistical significance.

Fig. 2.

Fig. 2.

Open in a new tab

Selected associations between blood metals and eicosanoids which significantly differed between fetal sexes. Estimates depict the percent change in eicosanoids with an IQR increase in blood metal concentrations. Blue circles indicate male fetuses and red squares indicate female fetuses.

Models were adjusted for continuous maternal age and categorical forms of maternal education, exposure to environmental tobacco smoke, and pre-pregnancy BMI.

Differences in associations between eicosanoids and metals exposure at study visits 1 and 3 are shown in Fig. S1. While overall trends were consistent with pregnancy average metals findings, positive associations tended to be stronger in magnitude with visit 1 metals than with visit 3 metals, particularly for eicosanoids in the parent compound group and the cytochrome p450 group. Specifically, among exposures to essential metals, Co showed the most change from associations at visit 1 compared to visit 3; these associations mostly remained in the same direction but were stronger in magnitude and significance at visit 1. Among exposures to toxic metals, Ni showed the most change between visits with most changing associations being positive at visit 1 and inverse at visit 3, particularly among cytochrome p450 eicosanoids. Cd and Cs showed strong positive associations with parent compounds only at visit 1. Conversely, Cd associations at visit 1 with cytochrome p450 eicosanoids were mostly significant and positive, while those with Cs at visit 1 were mostly inverse. Cs also showed significant and inverse associations with lipoxygenase eicosanoids only at visit 3. Finally, Sn showed many changes in associations between visits, most of which observed with cyclooxygenase eicosanoid associations becoming inverse and significant at visit 3.

4. Discussion

The pregnant women participating in the ongoing Puerto Rico PROTECT birth cohort are exposed to a variety of environmental contaminants, including metals. Although there is growing evidence showing associations between plasma eicosanoids and pregnancy outcomes, there is a paucity of research on the role that eicosanoids may play in the relationship between metal exposures and adverse birth outcomes. The current study observed numerous associations between prenatal metal exposure and eicosanoid biomarkers among participants in PROTECT, as well as infant sex-specific differences in several of these associations.

We have previously compared distributions of metal concentrations among women aged 18–40 years from NHANES 2009–2016 to those in PROTECT (Ashrap et al., 2020b) and have found that metal concentrations tend to be higher among PROTECT women. Specifically, concentrations of Mn, Co, Cs, Hg, Mn, Mo, Sn, and Zn were between 1.3 and 13 times greater in PROTECT than in NHANES. Thus, women in Puerto Rico may be more highly susceptible to toxic effects of metals exposure during pregnancy than women in the general population.

Notably, 12-oxoETE and 15-oxoETE, both synthesized via the lipoxygenase pathway, had the greatest number of associations with maternal blood metal concentrations. More specifically, both 12-oxoETE and 15-oxoETE were negatively associated with Cd, Co, and Pb, but positively associated with Cu, Mn, and Zn (Table S2 and Fig. 1). While limited information is available on the role of 12-oxoETE, it may function as an agonist of the LTB4 receptor of human neutrophils and exert anti-inflammatory effects (Naccache et al., 1991). Similarly, previous animal studies have indicated that overexpression of 15-LO-1, an enzyme involved in the metabolism of AA to 15-HETE, was associated with an anti-inflammatory response, and that 15-oxoETE showed anti-inflammatory activity (Wei et al., 2009). The positive associations between Zn and both 12-oxoETE and 15-oxoETE suggest the activation of anti-inflammatory responses related to Zn exposure. In contrast, the negative associations of Pb with both oxoETEs suggest suppression of anti-inflammatory responses associated with Pb exposure.

Previous studies have reported beneficial effects of Zn for prevention of inflammatory responses, oxidative stress, and pregnancy complications (Iqbal and Ali, 2021; Olechnowicz et al., 2018), while other studies have shown positive associations between Pb exposure and proinflammatory cytokines, oxidative stress, and elevated risk of spontaneous abortion, preterm birth, lower infant birth weight and length, and neurological dysfunctions (Cheng et al., 2017b; Dorea, 2019; Zhang et al., 2015; Habibian et al., 2022). While there is significant evidence showing toxic effects of Pb exposure (Vigeh et al., 2011), there are inconsistencies in the literature regarding the impacts of Zn on inflammation or oxidative stress during pregnancy. Some studies have shown potential activation of inflammatory cytokines (Tsou et al., 2011) wherein both Zn overload and Zn deficiency were shown to induce oxidative stress and cell death (Truong-Tran et al., 2001; Hao and Maret, 2005). In our previous studies we have shown positive associations between Zn and markers of oxidative stress (Ashrap et al., 2021) and inflammation (Kim et al., 2022). Additionally, we have observed that Pb, Zn, Mn, and Hg were associated with shorter gestational age at birth and increased odds of preterm delivery in the PROTECT cohort (Ashrap et al., 2020a). Given the complexities of the potential impacts of metals exposures on eicosanoid concentrations, oxidative stress, and inflammation, and the potential interactions between these pathways, further research is clearly needed to elucidate the effect of metals exposure on preterm birth and related biological pathways.

Previous work in vitro has demonstrated that treatment with 12-oxoETE induces apoptosis of the trophoblast (Kamada, 2023), which is the main cellular component of the placenta, and that this process was mediated by matrix metalloproteinases (MMPs). During normal pregnancy, trophoblast cells undergo rapid division and remodeling to create and maintain the interface between mother and fetus and to remodel uterine spiral arteries (Straszewski-Chavez et al., 2005). This process relies on coordinated apoptosis of trophoblast cells as the placenta develops throughout pregnancy. However, increased incidence of trophoblast apoptosis has been observed in cases of preeclampsia and intrauterine growth restriction (Straszewski-Chavez et al., 2005). Specifically, excessive trophoblast apoptosis early in pregnancy may contribute to insufficient spiral artery formation, which is often observed in abnormal pregnancies (Reister et al., 2001; DiFederico et al., 1999). Interestingly, in our previous study, we observed certain metals including Cs, Mn, and Zn to be significantly associated with MMPs that are responsible for uterine remodeling and healthy pregnancy outcomes (Kim et al., 2022). Together, these observations suggest that metals play a role in modulating eicosanoid expression levels, which may have subsequent impacts on placental physiology and adverse pregnancy outcomes.

We also observed infant sex-specific differences in associations between maternal blood metals and eicosanoid levels. Cs and Mn had the greatest number of infant sex-specific significant associations with the eicosanoids, which were driven by female infants (Fig. 2). Importantly, Cs was negatively associated with the eicosanoids, specifically with 12-oxoETE, 15-HETE, and 11-HETE, which all exert inflammatory activities. Consistently, in our previous study, we observed significant negative associations between Cs and MMP1 and MMP2 predominantly in mothers carrying female fetuses (Kim et al., 2022). Together, these results suggest that maternal blood Cs is inversely associated with inflammatory eicosanoids, as well as other inflammatory regulators, such as MMPs, in an infant sex-specific manner. Additionally, Mn showed strong positive associations with PGD2 and PGE3, from the cyclooxygenase pathway, among only female fetuses. A previous animal study has shown that PGD2 is the dominant prostanoid in the uterus during late pregnancy and functions to stimulate contractions (Liu et al., 2017). Importantly, in our previous work we have shown differential associations between environmental exposures and timing of delivery by fetal sex (Ashrap et al., 2020a; Cathey et al., 2022; Siwakoti et al., 2023). Sexual dimorphism in the associations between metals exposure and cyclooxygenase prostaglandins may help explain some of the differences in preterm delivery risk in pregnancies with males versus females. Clearly, more work is needed to fully understand the role of infant sex in the associations between prenatal metal exposures and eicosanoids.

While limited information is available regarding the sexual dimorphism of metals exposure impacts on eicosanoids, previous animal studies suggest the role of sex hormones on different eicosanoid related enzymatic activities (Pace et al., 2017). For example, in animal studies, estrogen increased expression of 15-LO and 5-LO, the lipoxygenases involved in metabolism of AA to 15-HETE and lipoxin. In contrast, progesterone was found to rapidly down-regulate biosynthesis of 5-LO (Pergola et al., 2015). In addition, estrogen was demonstrated to significantly enhance COX activity and the production of prostaglandins (Lai et al., 2019). Moreover, it has also been reported that testosterone has inhibitory effects on the cyclooxygenase pathway of AA metabolism by inhibiting prostaglandin secretion (ElAttar et al., 1982). Reproductive hormones play crucial roles in pregnancy and parturition, during which time their concentrations are consistently in flux (Mesiano, 2001). In our previous study, we observed that the impact of gestational metals exposure on maternal reproductive hormone levels varied by fetal sex (Rivera-Nunez et al., 2021). However, the limited number of epidemiological studies assessing the impact of metals on reproductive hormone levels during pregnancy calls for further research to fully understand the impact of reproductive hormones on the associations between metals exposure and eicosanoids.

There are important limitations in our study that need to be addressed. One limitation is that the metal blood samples in this study were only obtained from up to two study visits (median 18 and 26 weeks gestation), and eicosanoid blood samples were only obtained from one study visit (26 weeks gestation). Because inflammatory regulators change dramatically during pregnancy, more blood samples would have allowed a more detailed picture of the dynamic changes in eicosanoids over the course of pregnancy in our participants. Also, our study focused on an underrepresented community in the U.S., therefore, the generalizability of these observations to other populations may be limited.

Despite these limitations, it is important to highlight the strengths of our study. There is limited research to date on the associations between metal exposure and eicosanoids, which might uncover new information about pathways through which metals may impact pregnancy. Furthermore, we observed infant sex-specific differences in associations between maternal blood metal concentrations and eicosanoids. Though future work is warranted to dissect the role of infant sex in the associations between prenatal metal exposure and eicosanoids, our observation suggests the importance of considering infant sex as an independent factor in studying adverse birth outcomes. Additionally, we were able to explore possible windows of susceptibility to changing eicosanoid levels with metals exposure at two different time points during pregnancy. Though that part of this analysis was exploratory, we were able to show that the direction of some associations changed with exposure to various metals earlier versus later in pregnancy, which can help to inform future intervention efforts. Finally, we conducted this preliminary analysis in an established and well-characterized birth cohort among an underrepresented population of pregnant women at risk for elevated environmental exposures as well as adverse birth outcomes.

5. Conclusions

In this study, we present new information that may contribute to our understanding of relevant biological pathways for adverse birth outcomes in pregnant women who are exposed to environmental contaminants. We observed both positive and negative associations between maternal blood metal levels and eicosanoid profiles in pregnant mothers in the PROTECT cohort. We also reported the impacts of infant sex on metal associations with eicosanoid levels. Taken together, these findings highlight the need to further assess not only eicosanoid-metal associations but also warrant future work to explore their associations with other inflammatory regulators and their relation to adverse birth outcomes.

Supplementary Material

Supplementary Material

HIGHLIGHTS.

  • There are significant associations between maternal blood metal levels and eicosanoids.

  • 12-oxoETE and 15-oxoETE, both having inflammatory activity, had the greatest number of significant associations with metals.

  • There are infant sex-specific differences in the associations between maternal blood metal concentrations and eicosanoids.

Acknowledgments

We thank the nurses and research staff who participated in cohort recruitment and follow up, as well as the Federally Qualified Health Centers (FQHC) and clinics in Puerto Rico who facilitated participant recruitment, including Morovis Community Health Center (FQHC), Prymed: Ciales Community Health Center (FQHC), Camuy Health Services, Inc. (FQHC), and the Delta OBGyn (Prenatal Clinic).

Funding

This study was supported by the Superfund Research Program of the National Institute of Environmental Health Sciences, National Institutes of Health (P42ES017198). Additional support was provided from NIEHS grant numbers T32ES007062, P50ES026049, R01ES032203, and P30ES017885 and the Environmental influences on Child Health Outcomes (ECHO) program grant number UH3OD023251.

This study was approved by the research and ethics committees of the University of Michigan School of Public Health, University of Puerto Rico, Northeastern University, and participating hospitals and clinics. The patients/participants provided their written informed consent to participate in this study.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.scitotenv.2024.172295.

Footnotes

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Christine Kim: Writing – original draft, Formal analysis. Amber L. Cathey: Writing – review & editing, Methodology. Seonyoung Park: Writing - review & editing. Deborah J. Watkins: Writing – review & editing, Conceptualization. Bhramar Mukherjee: Supervision, Funding acquisition, Conceptualization. Zaira Y. Rosario-Pabón: Project administration, Data curation. Carmen M. Vélez-Vega: Project administration, Data curation. Akram N. Alshawabkeh: Funding acquisition, Conceptualization. José F. Cordero: Funding acquisition, Conceptualization. John D. Meeker: Supervision, Funding acquisition, Conceptualization.

Data availability

Data will be made available on request.

References

  1. Ashrap P, Watkins DJ, Mukherjee B, Boss J, Richards MJ, Rosario Z, et al. , 2020a. Maternal blood metal and metalloid concentrations in association with birth outcomes in Northern Puerto Rico. Environ. Int. 138, 105606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ashrap P, Watkins DJ, Mukherjee B, Boss J, Richards MJ, Rosario Z, et al. , 2020b. Predictors of urinary and blood metal(loid) concentrations among pregnant women in Northern Puerto Rico. Environ. Res. 183, 109178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ashrap P, Watkins DJ, Milne GL, Ferguson KK, Loch-Caruso R, Fernandez J, et al. , 2021. Maternal urinary metal and metalloid concentrations in association with oxidative stress biomarkers. Antioxidants (Basel) 10 (1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Aung MT, Yu Y, Ferguson KK, Cantonwine DE, Zeng L, McElrath TF, et al. , 2019. Prediction and associations of preterm birth and its subtypes with eicosanoid enzymatic pathways and inflammatory markers. Sci. Rep. 9 (1), 17049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Benjamini YHY, 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B Methodol. 57 (1), 289–300. [Google Scholar]
  6. Cathey AL, Watkins DJ, Rosario ZY, Velez C, Mukherjee B, Alshawabkeh AN, et al. , 2022. Biomarkers of exposure to phthalate mixtures and adverse birth outcomes in a Puerto Rico birth cohort. Environ. Health Perspect. 130 (3), 37009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cheng L, Zhang B, Zheng T, Hu J, Zhou A, Bassig BA, et al. , 2017a. Critical windows of prenatal exposure to cadmium and size at birth. Int. J. Environ. Res. Public Health 14 (1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cheng L, Zhang B, Huo W, Cao Z, Liu W, Liao J, et al. , 2017b. Fetal exposure to lead during pregnancy and the risk of preterm and early-term deliveries. Int. J. Hyg. Environ. Health 220 (6), 984–989. [DOI] [PubMed] [Google Scholar]
  9. DiFederico E, Genbacev O, Fisher SJ, 1999. Preeclampsia is associated with widespread apoptosis of placental cytotrophoblasts within the uterine wall. Am. J. Pathol. 155 (1), 293–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dorea JG, 2019. Environmental exposure to low-level lead (Pb) co-occurring with other neurotoxicants in early life and neurodevelopment of children. Environ. Res. 177, 108641. [DOI] [PubMed] [Google Scholar]
  11. ElAttar TM, Lin HS, Tira DE, 1982. Testosterone inhibits prostaglandin formation by human gingival connective tissue: relationship to 14C-arachidonic acid metabolism. Prostaglandins Leukot. Med. 9 (1), 25–34. [DOI] [PubMed] [Google Scholar]
  12. Funk CD, 2001. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294 (5548), 1871–1875. [DOI] [PubMed] [Google Scholar]
  13. Gouveia-Figueira S, Martens DS, Nawrot TS, Nording ML, 2017. Cord blood eicosanoid signatures and newborn gestational age. Prostaglandins Other Lipid Mediat. 133, 123–127. [DOI] [PubMed] [Google Scholar]
  14. Habibian A, Abyadeh M, Abyareh M, Rahimi Kakavandi N, Habibian A, Khakpash M, et al. , 2022. Association of maternal lead exposure with the risk of preterm: a meta-analysis. J. Matern. Fetal Neonatal Med. 35 (25), 7222–7230. [DOI] [PubMed] [Google Scholar]
  15. Hao Q and Maret W, Imbalance between pro-oxidant and pro-antioxidant functions of zinc in disease. J. Alzheimers Dis, 2005. 8(2): p. 161–70; discussion 209–15. [DOI] [PubMed] [Google Scholar]
  16. Iqbal S, Ali I, 2021. Effect of maternal zinc supplementation or zinc status on pregnancy complications and perinatal outcomes: an umbrella review of meta-analyses. Heliyon 7 (7), e07540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kamada H, 2023. 12-Oxoeicosatetraenoic acid, a candidate signal for placenta separation, activates matrix metalloproteinase and induces apoptosis in bovine trophoblast cells. Anim Biosci 36 (3), 429–440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kim SS, Meeker JD, Carroll R, Zhao S, Mourgas MJ, Richards MJ, et al. , 2018. Urinary trace metals individually and in mixtures in association with preterm birth. Environ. Int. 121 (Pt 1), 582–590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kim C, Cathey AL, Watkins DJ, Mukherjee B, Rosario-Pabon ZY, Velez-Vega CM, et al. , 2022. Maternal blood metal concentrations are associated with matrix metalloproteinases (MMPs) among pregnant women in Puerto Rico. Environ. Res. 209, 112874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lai ZZ, Yang HL, Ha SY, Chang KK, Mei J, Zhou WJ, et al. , 2019. Cyclooxygenase-2 in endometriosis. Int. J. Biol. Sci. 15 (13), 2783–2797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu B, Yang J, Luo W, Zhang Y, Li J, Li H, et al. , 2017. Prostaglandin D(2) is the major cyclooxygenase-1-derived product in prepartum mouse uteri where it mediates an enhanced in vitro myometrial contraction. Eur. J. Pharmacol. 813, 140–146. [DOI] [PubMed] [Google Scholar]
  22. Liu H, Lu S, Zhang B, Xia W, Liu W, Peng Y, et al. , 2018. Maternal arsenic exposure and birth outcomes: a birth cohort study in Wuhan. China. Environ Pollut 236, 817–823. [DOI] [PubMed] [Google Scholar]
  23. Long A, Ma S, Li Q, Lin N, Zhan X, Lu S, et al. , 2016. Association between the maternal serum levels of 19 eicosanoids and pre-eclampsia. Int. J. Gynaecol. Obstet. 133 (3), 291–296. [DOI] [PubMed] [Google Scholar]
  24. Meagher EA, FitzGerald GA, 1993. Disordered eicosanoid formation in pregnancy-induced hypertension. Circulation 88 (3), 1324–1333. [DOI] [PubMed] [Google Scholar]
  25. Mesiano S, 2001. Roles of estrogen and progesterone in human parturition. Front. Horm. Res. 27, 86–104. [DOI] [PubMed] [Google Scholar]
  26. Naccache PH, Leblanc Y, Rokach J, Patrignani P, Fruteau de Laclos B, Borgeat P, 1991. Calcium mobilization and right-angle light scatter responses to 12-oxo-derivatives of arachidonic acid in neutrophils: evidence for the involvement of the leukotriene B4 receptor. Biochim. Biophys. Acta 1133 (1), 102–106. [DOI] [PubMed] [Google Scholar]
  27. Novy MJ, Liggins GC, 1980. Role of prostaglandins, prostacyclin, and thromboxanes in the physiologic control of the uterus and in parturition. Semin. Perinatol. 4 (1), 45–66. [PubMed] [Google Scholar]
  28. Olechnowicz J, Tinkov A, Skalny A, Suliburska J, 2018. Zinc status is associated with inflammation, oxidative stress, lipid, and glucose metabolism. J. Physiol. Sci. 68 (1), 19–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pace S, Sautebin L, Werz O, 2017. Sex-biased eicosanoid biology: impact for sex differences in inflammation and consequences for pharmacotherapy. Biochem. Pharmacol. 145, 1–11. [DOI] [PubMed] [Google Scholar]
  30. Park S, Cathey AL, Hao W, Zeng L, Pennathur S, Aung MT, et al. , 2023. Associations of phthalates, phthalate replacements, and their mixtures with eicosanoid biomarkers during pregnancy. Environ. Int. 178, 108101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pergola C, Schaible AM, Nikels F, Dodt G, Northoff H, Werz O, 2015. Progesterone rapidly down-regulates the biosynthesis of 5-lipoxygenase products in human primary monocytes. Pharmacol. Res. 94, 42–50. [DOI] [PubMed] [Google Scholar]
  32. Reister F, Frank HG, Kingdom JC, Heyl W, Kaufmann P, Rath W, et al. , 2001. Macrophage-induced apoptosis limits endovascular trophoblast invasion in the uterine wall of preeclamptic women. Lab. Investig. 81 (8), 1143–1152. [DOI] [PubMed] [Google Scholar]
  33. Rivera-Nunez Z, Ashrap P, Barrett ES, Watkins DJ, Cathey AL, Velez-Vega CM, et al. , 2021. Association of biomarkers of exposure to metals and metalloids with maternal hormones in pregnant women from Puerto Rico. Environ. Int. 147, 106310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rosner B, 2015. Fundamentals of Biostatistics. Nelson Education. [Google Scholar]
  35. Schisterman EF, Vexler A, Whitcomb BW, Liu A, 2006. The limitations due to exposure detection limits for regression models. Am. J. Epidemiol. 163 (4), 374–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Siwakoti RC, Cathey A, Ferguson KK, Hao W, Cantonwine DE, Mukherjee B, et al. , 2023. Prenatal per- and polyfluoroalkyl substances (PFAS) exposure in relation to preterm birth subtypes and size-for-gestational age in the LIFECODES cohort 2006–2008. Environ. Res. 237 (Pt 2), 116967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Straszewski-Chavez SL, Abrahams VM, Mor G, 2005. The role of apoptosis in the regulation of trophoblast survival and differentiation during pregnancy. Endocr. Rev. 26 (7), 877–897. [DOI] [PubMed] [Google Scholar]
  38. Strauss JF, FitzGerald GA, 2019. Chapter 4-steroid hormones and other lipid molecules involved in human reproduction. In: Strauss JF, Barbieri RL. (Eds.), Yen and Jaffe’s Reproductive Endocrinology, Eighth edition. Elsevier, Amsterdam, pp. 75–114.e7. [Google Scholar]
  39. Tassoni D, Kaur G, Weisinger RS, Sinclair AJ, 2008. The role of eicosanoids in the brain. Asia Pac. J. Clin. Nutr. 17 (Suppl. 1), 220–228. [PubMed] [Google Scholar]
  40. Truong-Tran AQ, Carter J, Ruffin RE, Zalewski PD, 2001. The role of zinc in caspase activation and apoptotic cell death. Biometals 14 (3–4), 315–330. [DOI] [PubMed] [Google Scholar]
  41. Tsou TC, Chao HR, Yeh SC, Tsai FY, Lin HJ, 2011. Zinc induces chemokine and inflammatory cytokine release from human promonocytes. J. Hazard. Mater. 196, 335–341. [DOI] [PubMed] [Google Scholar]
  42. Vigeh M, Yokoyama K, Seyedaghamiri Z, Shinohara A, Matsukawa T, Chiba M, et al. , 2011. Blood lead at currently acceptable levels may cause preterm labour. Occup. Environ. Med. 68 (3), 231–234. [DOI] [PubMed] [Google Scholar]
  43. Wai KM, Mar O, Kosaka S, Umemura M, Watanabe C, 2017. Prenatal heavy metal exposure and adverse birth outcomes in Myanmar: a birth-cohort study. Int. J. Environ. Res. Public Health 14 (11). [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wang YN, Hu HH, Zhang DD, Wu XQ, Liu JL, Guo Y, et al. , 2021. The dysregulation of eicosanoids and bile acids correlates with impaired kidney function and renal fibrosis in chronic renal failure. Metabolites 11 (2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wei C, Zhu P, Shah SJ, Blair IA, 2009. 15-oxo-Eicosatetraenoic acid, a metabolite of macrophage 15-hydroxyprostaglandin dehydrogenase that inhibits endothelial cell proliferation. Mol. Pharmacol. 76 (3), 516–525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Welch BM, Keil AP, van’t Erve TJ, Deterding LJ, Williams JG, Lih FB, et al. , 2020. Longitudinal profiles of plasma eicosanoids during pregnancy and size for gestational age at delivery: a nested case-control study. PLoS Med. 17 (8), e1003271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Xia W, Hu J, Zhang B, Li Y, Wise JP Sr., Bassig BA, et al. , 2016. A case-control study of maternal exposure to chromium and infant low birth weight in China. Chemosphere 144, 1484–1489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yuan ZX, Majchrzak-Hong S, Keyes GS, Iadarola MJ, Mannes AJ, Ramsden CE, 2018. Lipidomic profiling of targeted oxylipins with ultra-performance liquid chromatography-tandem mass spectrometry. Anal. Bioanal. Chem. 410 (23), 6009–6029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhang B, Xia W, Li Y, Bassig BA, Zhou A, Wang Y, et al. , 2015. Prenatal exposure to lead in relation to risk of preterm low birth weight: a matched case-control study in China. Reprod. Toxicol. 57, 190–195. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Material
NIHMS2138766-supplement-Supplementary_Material.docx (640.5KB, docx)

Data Availability Statement

Data will be made available on request.

RESOURCES