Study Design

To assess blood levels of toxic metals in pregnant women, we analyzed blood samples submitted to the North Carolina Department of Health and Human Services (NCDHHS) State Laboratory of Public Health for blood group typing. Residual volumes of blood originally submitted for group typing were analyzed for As, Cd, Hg, and Pb. Data were provided under the context of a data use agreement for a limited data set using a data custodian as an intermediary to protect identifiable information. This study was reviewed by the UNC Public Health-Nursing IRB (#09-1399) and was determined to be exempt from review according to regulatory category 4 under the National Institutes of Health description on Research on Human Specimens (44 CFR 46.101(b)). Similar strategies that assess metal levels in samples collected under the auspices of existing public health monitoring programs have been used to establish environmental contaminant levels in individuals [30].

Women in this study were North Carolina residents in their third trimester of pregnancy and receiving prenatal care at local health departments. The mothers' venous blood samples were collected between October 2009 and January 2011. Bladen, Cumberland, Richmond, Stanly, Union, and Wake Counties were selected for analysis based on trends in private well water metal levels reported in publicly available data [31]. Throughout the remainder of the manuscript county identities are coded and designated as A–F, using random assignment. Data collected included: (1) pregnant women's (n = 211) whole blood level of As, Cd, Hg, and Pb, (2) county of residence, (3) age, and (4) race. No additional demographic information on subjects was requested in accordance with the limited data use agreement.

Blood Analysis

The North Carolina Department of Health and Human Services (NCDHHS) for blood typing group is certified by the U.S. DHHS as compliant with the 1988 Clinical Laboratory Improvement Amendment (CLIA). In accordance with CLIA certification, the laboratory adheres to quality assessment/quality control (QA/QC) requirements with oversight by the Centers for Medicare and Medicaid Services. Blood samples from non-fasting pregnant women were collected into phlebotomy tubes containing ethylenediaminetetraacetic acid (EDTA) as an anticoagulant and stored at 4°C prior to processing at the Division of Public Health State Laboratory. Whole blood analysis was performed via Inductively Coupled Plasma Dynamic Reaction Cell Mass Spectrometry (ICP/DRC/MS) according to the US CDC laboratory methods for metals in whole blood, DLS Method code CTL-TMS 3.01 [32], [33]. The samples were analyzed on a Perkin Elmer Elan DRC II ICP/MS instrument. Arsenic was measured in blood using the DRC mode to remove isobaric interference from Argon Chloride at mass 75. Iridium was used as an internal standard for all elements. The limits of detection (LODs) were 0.23 µg-As/L, 0.11 µg-Cd/L, 0.23 µg-Hg/L, and 0.13 µg-Pb/dL. The results represent levels of total metal concentrations; speciation of metals was not performed. The ICP-MS analyses for Cd, Hg, and Pb were similar to those used in the Fourth National Report on Human Exposure to Environmental Chemicals (NHANES IV).

Statistical analysis

Statistical analysis was conducted with the statistical package SAS 9.2 (SAS Institute Inc., Cary, North Carolina). Blood samples with metal levels below the LOD were assigned an imputed value equal to LOD/√2, as in NHANES IV [34]. The data were log-transformed to address the non-normal distribution of metal concentrations. Spearman's correlation coefficient estimates among metal pairs and corresponding p-values were calculated. Linear regression was performed on every pairwise metal combination by adjusting for maternal age to test for associations between blood metal levels and race (NHW served as the referent group). Additional multivariate analyses were performed adjusting for maternal age, race, and categorized county of residence to assess the relationship between metal levels and residence in individual counties. Furthermore, differences between the mean metal levels of individual counties were compared using Scheffe's test (α = 0.05).

Pregnant women are exposed to toxic metals

The aim of this study was to assess toxic metal exposure in pregnant mothers in North Carolina. The data suggest that indeed pregnant women in North Carolina are exposed to toxic metals. For each of the four toxic metals studied, more than 50% of the women had detectable blood levels. Specifically, depending on the metal detectable blood levels were found in 57–100% of the samples. Most of the metal measurements revealed levels of exposure that did not fall outside regulated limits. For some individuals, however, blood levels of toxic metals exceeded the 95^th percentile of Cd, Hg, and Pb reported in the U.S. population. Moreover, a few women had metal levels above current guideline levels adopted for health protection. Specifically, some women had blood metal levels above reference values indicative of a threshold for adverse health effects: five women (2.3%) had blood Hg levels above 3.5 µg/L and two women (0.94%) had Pb levels greater than 5.0 µg/dL (the CDC blood lead advisory). Since no current threshold levels exist for As in blood we compared our findings with a previous study by Concha et al [2]. Three women in our study showed levels similar to those reported among As-exposed pregnant women served by contaminated public water supplies containing up to 200 µg/L As (blood range: 5.6–13 µg/L) [2]. In addition, thirty-seven women had blood levels greater than 1 µg/L, a level previously associated with known exposure to As in drinking water [36].

As and Hg are positively correlated in pregnant mothers' blood

Environmental exposures in the general public more commonly occur as mixtures than as single-contaminant exposures. Contaminant mixtures can contribute to an individuals' susceptibility and potentially have synergistic effects resulting in disease manifestation [37], [38]. Here we examined the correlation of the four toxic metals among the pregnant women. Notably, we report a relationship between maternal As and Hg blood levels. An implication of this finding is that maternal and fetal body burden may result from combined or interactive effects of more than one metal. Thus, concomitant exposure is an important consideration in future studies that assess in utero health outcomes from metal exposure.

Non-Hispanic Black maternal race is associated with increased Cd

The findings also highlight associations between blood metal levels and maternal race. Our data suggest that across the sample population NHB maternal race was associated with increased Cd exposure compared to NHW women. To our knowledge this association has not been previously reported. Poor birth outcomes in North Carolina and across the U.S. disproportionately affect infants born to NHB mothers [39], [40], despite the lack of a definitive causal mechanism. Our findings suggest that the role of maternal race in birth outcome etiology may be confounded by metal exposure, and further investigation of social, environmental, and genetic factors disproportionately affecting NHB mothers is needed.

County of residence is associated with mothers' metal levels

This initial study suggests that on average, women residing in various North Carolina counties experience differential exposure to toxic metals during pregnancy. We show that on average, pregnant women residing in Counties C and F were likely to have higher levels of As when compared to women residing in the other counties. Exposure to As primarily occurs through ingestion of food and drinking water [41]. In public drinking water supplies, allowable levels of As can range up to the EPA Maximum Contaminant Level (MCL) of 10 µg/L. As such, individuals can be exposed to detectable levels of As that are still within regulatory guidelines. We have shown recently that in North Carolina, some counties have elevated levels of As in unregulated private well water [23]. Here we show that women living in Counties C and F have elevated blood As, two counties that were not identified as particular outliers in the well water analysis. This could suggest an alternate route of exposure among residents, perhaps through diet. In addition, this study did not select for individuals based on their type of drinking water source. Taken together, these results suggest that a future study of women supplied by private drinking water sources may be prudent.

We also found that on average, pregnant women residing in Counties D and F also had elevated levels of blood Hg. Most Hg exposures in the U.S. are attributable to diet, particularly fish consumption [42]; however, dental amalgams [4] and some industrial processes also contribute to human exposure [43]. In North Carolina, a statewide freshwater fish advisory for mercury (specifically in largemouth bass) has been in effect since 2009. A recent study in Durham County, North Carolina found that the type of fish consumption and corresponding Hg content varied among different demographic groups of pregnant women [24]. There are no reports of public or private well drinking water Hg levels above acceptable EPA MCL standard of 2 µg/L in either of these counties.

Cadmium exposure commonly occurs through food consumption or inhalation of cigarette smoke, although minor exposure can occur from byproducts of industrial processes [44]. Previous studies have shown Cd levels in maternal and cord blood are higher in smokers [45], and that Cd levels among smokers were reportedly 4–5 times higher than nonsmokers [46]. Moreover, Cd levels in pregnant women exposed to second–hand smoke were double that of nonsmokers [6]. In our study, residence in County A was associated with increased maternal blood Cd levels; yet, the smoking rate in this county is lower than the statewide rate (20.9% of adults) [47]. Detectable levels of Cd in public drinking water can range up to 5.0 µg/L (the EPA MCL) and may potentially contribute to low-level exposure in the general population. Further research is needed to establish the source of maternal Cd exposure.

Exposure to Pb can result from various environmental sources such as contaminated dust or soil, drinking water, lead-based paint, cigarette smoke, as well as byproducts of industrial processes [48]. Elevated maternal blood levels of Pb were associated with residence in County C. Similar to the other metals, Pb levels in public drinking water supplies can range up to the 15 µg/L and up to 10% of samples can exceed this level under EPA guidelines. Currently, no evidence indicates that North Carolina public water sources contain endemic levels of lead. While drinking water is a potential route of lead exposure, rural counties may experience additional risk factors for Pb exposure such as older housing and low socioeconomic status that may put mothers at risk [49].

Although the design of this study did not allow us to ascertain exposure sources, both anthropogenic and environmental sources can contribute to an individuals' toxic metal exposure. Body weight, race, occupational factors, duration of residence, socioeconomic status, smoking status, gestational time during pregnancy, and genetic factors can affect the body burden of metals. It is likely that multiple exposure routes and sources may have contributed to the metal levels assessed in this study and warrant future follow-up.

Biomonitoring programs may reduce the risk of prenatal metal exposure

Despite evidence that adverse health effects are attributable to in utero metal exposure, currently Minnesota, New York City, and New York State are the only jurisdictions that have active guidelines for monitoring maternal blood lead levels in the U.S. [26]. The initiation of statewide metal biomonitoring programs in the U.S. may be accomplished through collaborations with state programs to better protect public health, particularly in susceptible populations such as pregnant women and children. This study demonstrates that detectable metal levels were present in maternal blood and suggests that further biomonitoring of blood levels of toxic metals in susceptible North Carolina populations is prudent. Furthermore, in lieu of existing targeted biomonitoring programs, public health strategies to prevent exposure such as raising awareness among pregnant mothers may be vital to protecting the health of newborns.

In conclusion, this study demonstrates that the pregnant women under study were exposed to four toxic metals at detectable levels. Specifically, we show evidence of elevated levels of As, Hg, and Pb among some women in the third trimester of pregnancy. In addition, we identified statistically significant relationships between maternal As and Hg in blood levels. Furthermore, we report associations between maternal blood metal levels and maternal race as well as county of residence. The increased levels of one or more of these metals are associated with detrimental neurological and physiological developmental outcomes. Informing pregnant women of the sources of toxic metals such as diet (including drinking water), cigarette smoke, and other environmental sources may help to reduce exposures. Targeted national, statewide, and community level programs promoting mothers' awareness of metals exposure and in utero effects would each help to protect human health and lessen the economic burden of metal exposure. Biomonitoring programs in North Carolina and other at-risk areas can help to protect the health of expectant mothers and their children.