Background
Prenatal exposure to methylmercury is associated with adverse neurologic development in children. We examined total blood mercury concentrations and predictors of higher blood mercury concentrations in pregnant and nonpregnant women.
Methods
We analyzed data from 1183 pregnant and 5587 nonpregnant women aged 16-49 years from the 1999-2006 National Health and Nutrition Examination Survey (NHANES). We estimated geometric mean blood mercury concentrations and characteristics associated with higher mercury concentrations (≥3.5 μg/L) in crude and adjusted linear and logistic regression models.
Results
After adjusting for age and race/ethnicity, geometric mean blood mercury concentrations were clinically similar but significantly lower for pregnant (0.81 μg/L; 95% confidence interval [CI], 0.71–0.91) and nonpregnant women of childbearing age (0.93 μg/L; 95% CI, 0.87–0.99); 94% of pregnant and 89% of nonpregnant women had blood mercury concentrations below 3.5 μg/L. The most significant predictor of higher blood mercury concentrations for both pregnant and nonpregnant women was any seafood consumption vs no consumption in the last 30 days (odds ratio, 18.7; 95% CI, 4.9–71.1; odds ratio, 15.5; 95% CI, 7.5–32.1, respectively). Other characteristics associated with ≥3.5 μg/L blood mercury concentrations were older age (≥35 years), higher education (greater than high school), and higher family income to poverty ratio (3.501+) for both pregnant and nonpregnant women.
Conclusion
Pregnancy status was not strongly associated with blood mercury concentrations in women of childbearing age and blood mercury concentrations above the 3.5 μg/L cut were uncommon.
Mercury is a naturally-occurring element that is widespread in the environment and exists as elemental, inorganic, and organic mercury (methylmercury; MeHg). Inorganic and elemental mercury, measured in urine, are usually associated with dental amalgams, occupational exposures, and herbs and medicines adulterated with inorganic mercury. MeHg exposure, which can be estimated by measuring concentrations in blood and hair comes almost exclusively from consumption of fish or shellfish. Studies have consistently reported that increasing frequency of seafood intake is the single most influential predictor of blood MeHg concentrations.
Fetuses are a high-risk group for MeHg exposure because of the increased susceptibility of the developing brain to environmental insults. Three long-term studies, the Seychelles Island study, the Faroe Islands study, and a third study conducted in New Zealand, have investigated MeHg in children who were prenatally exposed to MeHg through maternal seafood consumption. MeHg dose-related deficits in tests of memory, attention, and language were observed in children of different ages. Given that permanent damage to the developing brain can occur, MeHg exposure in pregnant women is a source for concern. The US Environmental Protection Agency (EPA) set the reference dose (RfD)—a dose which is unlikely to have deleterious effects—for MeHg at 0.1 μg/Kg/day.
Total blood mercury (BHg) includes all form of mercury and is used in biomonitoring as a proxy for MeHg, although the distribution of mercury types in blood can vary. Although the RfD varies by body weight, the value 5.8 μg/L (ppb) for BHg has been widely used in place of a weight-specific value in studies of mercury concentrations and health outcomes. Originally, it was thought that cord blood and maternal blood mercury levels were equivalent when calculating the RfD ; however, studies comparing maternal and cord blood concentrations of MeHg have found that cord-blood mercury is higher than maternal blood mercury. According to the EPA, a review of the literature identified 21 studies that reported cord-blood mercury and maternal blood mercury data; these data indicated that cord-blood mercury concentrations are higher than maternal mercury concentrations, with a ratio of approximately 1.7. Studies have therefore suggested and used a lower RfD reflecting exposures equivalent to RfD of 3.5 μg/L total blood mercury concentration.
Two previous studies have examined and compared BHg concentrations in pregnant and nonpregnant women; 1 reported no significant differences between the 2 groups using hair total mercury concentrations from NHANES 1999-2000, and the other found significantly lower concentrations of total BHg in pregnant women aged 17-39 years using NHANES 2003-2010. The objective of our study was to assess whether predictors of BHg concentrations were the same in pregnant and nonpregnant women.
Materials and Methods
Data source
We used the National Health and Nutrition Examination Survey (NHANES) data obtained from 1999 through 2006. NHANES is a stratified, multistage probability sample of the civilian, noninstitutionalized population of the US conducted by the National Center for Health Statistics (NCHS) of the US Centers for Disease Control and Prevention (CDC). The NCHS Research Ethics Review Board approved the NHANES protocol. The consent form to participate in the survey as well to store specimens of their blood for future research were signed by all participants of the survey. NHANES includes an in-home questionnaire and an examination at a mobile examination center.
Study sample
Our study included all women 16-49 years of age, who completed both the interview and examination portions of NHANES. The NHANES variable RIDEXPREG was used to determine pregnancy status for this analysis. Women who were identified as pregnant through a positive laboratory pregnancy test or who self-reported as pregnant at the time of the interview were considered pregnant and those who specified that they were not pregnant at the time of the interview and who did not test positive in the laboratory pregnancy test were considered nonpregnant. Women with missing pregnancy status and those for whom pregnancy status could not be ascertained were not included in this analysis. NHANES 1999-2006 were used for this study because pregnant women aged 15-39 years were oversampled during these years. We excluded NHANES 2007-2010 because pregnant women were no longer oversampled during this time and therefore they would have been disproportionately represented in the analysis, with much less stable estimates able to be produced from these later data. In addition, beginning in 2007 the age ranges included in the public release dataset for pregnant women were restricted to 20 to 44 years of age, although for previous years there was essentially no bound (variable available for ages 8 to 59 years).
Blood mercury measurement
Total blood mercury (BHg) was measured in micrograms per liter (μg/L) of whole blood using cold vapor atomic absorption spectrometry for NHANES 1999-2002, and inductively-couple plasma dynamic reaction cell mass spectrometry for NHANES 2003-2006. There is no documentation of differences because of methodologic changes in BHg measurement. Analysis of blood samples was done at the CDC’s Division of Laboratory Sciences, National Center for Environmental Health (NCEH). Other than measurement technique, no changes in laboratory procedures or equipment were reported over the study period. The BHg limit of detection for both methods was 0.14 μg/L; 58 pregnant and 197 nonpregnant women in this study had BHg concentrations below 0.14 μg/L (weighted percentages: 5.6 and 3.2, respectively). According to the NHANES, results below the limit of detection are replaced with a value equal to the detection limit divided by the square root of 2. These values in NHANES cycles for this analysis were 0.07 and 0.1 and were used for our analysis. We excluded pregnant (n = 77) and nonpregnant (n = 261) women with missing blood mercury concentrations. Because we were primarily interested in the potential for fetal exposure to MeHg, we defined maternal methyl mercury concentrations at or above 3.5 μg/L as high exposure concentrations.
Mercury intake via fish and shellfish consumption and other covariates
Survey participants, including all women aged 16 to 49 years of age, were asked about fish and shellfish consumption during the previous 30 days. Respondents were asked whether they consumed fish and/or shellfish in the past 30 days, and if so, the frequency of consumption during that time. Participants were then asked about consumption of different types of fish and shellfish including a category for “other and unknown” fish or shellfish. No information was obtained about portion sizes or preparation methods.
Demographic information including race and ethnicity, educational attainment, and income and family size were self-reported at the time of the interview. We categorized race and ethnicity into the following categories: nonHispanic white, nonHispanic black, Mexican American, and other race/ethnicity, which includes other Hispanics, Asians, Pacific Islanders, Native Americans, Alaskan Natives, and those with multiple races and/or ethnicities. NHANES data do not allow representative estimates to be made for racial/ethnic subgroups other than nonHispanic white, nonHispanic Black, and Mexican Americans, therefore, these subgroups were combined. Family income to poverty ratio (FIPR) is the total household income divided by the poverty threshold for the year of the interview. The poverty threshold is determined annually by the US Census Bureau, taking into account geographic location, rate of inflation, and family size.
Statistical analysis
Analyses were conducted using SAS (version 9.3; SAS Institute, Cary, NC) and SUDAAN (version 10.0; Research Triangle Institute, Research Triangle Park, NC). Mobile examination center examination sample weights and the appropriate sample design variables were used in the analysis to account for the complex survey design, oversampling, and differential nonresponse and noncoverage to obtain nationally representative estimates of the US civilian noninstitutionalized population.
The distribution of BHg concentrations in pregnant and nonpregnant women were described through the calculation of percentiles and geometric means. SUDAAN’s Taylor series linearization method was used to calculate 95% confidence intervals (CIs) for the estimated prevalences, and χ 2 statistics were used to compare pregnant with nonpregnant women. We estimated both crude and adjusted geometric means of mercury concentration, with age group and race/ethnicity controlled for in the adjusted estimate. Logistic regression was used to examine the potential associations between having high BHg concentrations and selected characteristics including age, race/ethnicity, education, socioeconomic status, and seafood consumption in pregnant and, separately, nonpregnant women.
Results
We analyzed data on 6770 women: 1183 pregnant and 5587 nonpregnant women from 1999 to 2006 who had completed interviews, examinations, and valid pregnancy data. Geometric mean BHg concentrations were significantly lower in pregnant women compared with nonpregnant women (0.71 and 0.92 μg/L, respectively; P < .0001). After adjustment for age and race/ethnicity, the difference in geometric mean BHg concentration between the groups was smaller, 0.81 and 0.93 μg/L, respectively, although still statistically significant ( Table 1 ). We also examined geometric mean BHg concentrations by trimester of pregnancy alone as well as adjusting for age and found no significant differences by trimester of pregnancy ( P = .88). There were statistically significant differences between pregnant and nonpregnant women by age ( P ≤ .0001) and race/ethnicity ( P ≤ .0001). Among pregnant women, 91% were younger than 35 years of age, compared with approximately 53% of nonpregnant women; and a greater percentage of pregnant women were Mexican American compared with nonpregnant women, approximately 16% and 9%, respectively ( Table 1 ). Pregnant women were less likely to have reported consuming any seafood (fish or shellfish) compared to nonpregnant (73.0% and 78.1%, respectively; P = .05), and fish specifically (62.5% and 68.8%, respectively; P = .03) in the last 30 days but not shellfish. There were no significant differences observed between pregnant and non-pregnant women in education level or family income to poverty ratio ( Table 1 ).
| Variable | Pregnant women | Nonpregnant women | P value a | ||
|---|---|---|---|---|---|
| Total mean blood mercury, μg/L | 0.71 (0.63–0.79) | 0.92 (0.85–0.99) | < .0001 | ||
| Total mean blood mercury adjusted for age and race, μg/L | 0.81 (0.71–0.91) | 0.93 (0.87–0.99) | .0001 | ||
| Total, n b | 1183 | 5587 | |||
| Age, n, % b (95% CI) c | < .0001 | ||||
| 16-25 | 551 | 41.6 (37.0–46.4) | 2386 | 25.9 (24.4–27.5) | |
| 26-35 | 558 | 49.4 (43.8–55.0) | 1209 | 26.6 (25.0–28.2) | |
| 36-49 | 74 | 9.0 (5.9–13.6) | 1992 | 47.5 (45.5–49.5) | |
| Race, n, % b (95% CI) c | < .0001 | ||||
| NonHispanic white | 526 | 56.3 (50.3–62.1) | 2189 | 66.8 (63.5–70.0) | |
| NonHispanic black | 183 | 15.1 (11.5–19.6) | 1405 | 12.7 (10.8–15.0) | |
| Mexican American | 350 | 16.0 (13.0–19.5) | 1494 | 8.7 (7.2– 10.4) | |
| Other race and/or multiracial | 124 | 12.7 (8.8–18.0) | 499 | 11.8 (9.7–14.2) | |
| Education level, n, % b (95% CI) c | .07 | ||||
| Less than high school graduate | 374 | 23.1 (19.4–27.2) | 2017 | 20.3 (18.8–21.9) | |
| High school graduate or GED | 259 | 18.7 (15.2–22.7) | 1218 | 22.9 (21.3–24.5) | |
| Greater than high school | 549 | 58.2 (53.3–36.0) | 2347 | 56.7 (54.4–59.0) | |
| Missing | 1 | — d | 5 | — d | |
| Family income to poverty ratio, e n, % b (95% CI) c | .09 | ||||
| 0-1.3 | 399 | 25.6 (21.7–29.9) | 1849 | 23.6 (21.4–25.9) | |
| 1.301-3.5 | 389 | 35.5 (31.7–39.6) | 1853 | 32.8 (30.7–35.0) | |
| 3.501+ | 319 | 32.1 (27.0–37.7) | 1516 | 38.2 (35.5–40.9) | |
| Missing | 76 | — d | 369 | 5.4 (4.5–6.5) | |
| Fish or shellfish consumption in the past 30 days, n, % b (95% CI) c | .05 | ||||
| Yes | 849 | 73.0 (67.2–78.0) | 4160 | 78.1 (76.0–80.1) | |
| No | 292 | 23.6 (19.2–28.6) | 1226 | 18.7 (17.0–20.6) | |
| Missing | 42 | — d | 201 | 3.2 (2.5–4.0) | |
| Fish consumption in the past 30 days, n, % b (95% CI) c | .03 | ||||
| Yes | 700 | 62.5 (56.8–67.8) | 3545 | 68.8 (66.6–71.0) | |
| No | 441 | 34.1 (29.4–39.2) | 1843 | 28.1 (26.1–30.2) | |
| Missing | 42 | — d | 199 | 3.1 (2.5–4.0) | |
| Shellfish consumption in the past 30 days, n, % b (95% CI) c | .13 | ||||
| Yes | 545 | 46.1 (41.0–51.4) | 2769 | 51.4 (48.6–54.2) | |
| No | 596 | 50.4 (45.5–55.4) | 2615 | 45.4 (42.7–48.1) | |
| Missing | 42 | — d | 203 | 3.2 (2.6–4.1) | |
d Estimates suppressed because minimum degrees of freedom (12) for strata not met
e Family income to poverty ratio is the total household income divided by the poverty threshold for the year of the interview.
In examining the factors associated with having a high BHg concentration (at or above 3.5 μg/L), we observed that pregnancy status was not a strong predictor after adjusting for other factors (odds ratio [OR] for pregnant vs nonpregnant, 0.7; 95% CI, 0.5–1.0). Reported seafood consumption was the strongest predictor of high BHg concentrations for both pregnant (OR, 18.7; 95% CI, 4.9–71.1) and nonpregnant (OR, 15.5; 95% CI, 7.5–32.1) women. Among nonpregnant women, other factors significantly associated with high BHg concentrations were increasing age (OR for 26-35 vs 16-25 years, 1.8; 95% CI, 1.3–2.5; OR for 36-49 vs 16-25 years: 2.0; 95% CI, 1.5–2.7), other/multiracial ethnicity (OR compared to nonHispanic white, 2.2; 95% CI, 1.4–3.4), greater than high school education (OR compared with less than high school, 1.4; 95% CI, 1.0–1.9), and higher family income to poverty ratio (OR for FIPR >3.5 compared with FIPR ≤1.3: 95% CI, 2.1; 1.4–3.3) for nonpregnant women. Similar, but nonsignificant patterns were observed for pregnant women, for whom the sample size was smaller ( Table 2 ).
| Variable | All women | Pregnant women | Non-pregnant women |
|---|---|---|---|
| OR (95% CI) n = 621 | OR (95% CI) n = 55 | OR (95% CI) n = 566 | |
| Pregnancy status | |||
| No | 1.0 (Ref) | N/A | N/A |
| Yes | 0.70 (0.47–1.03) | ||
| Age, y | |||
| 16-25 | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| 26-35 | 2.00 (1.46–2.75) | 2.25 (0.72–7.02) | 1.79 (1.28–2.49) |
| 36-49 | 2.33 (1.71–3.19) | 3.78 (0.68–20.90) | 1.98 (1.46–2.68) |
| Race | |||
| NonHispanic white | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| NonHispanic black | 1.39 (0.97–1.99) | 0.44 (0.19–1.02) | 1.23 (0.85–1.78) |
| Mexican American | 0.66 (0.45–0.95) | 0.31 (0.10–0.98) | 0.64 (0.43–0.95) |
| Other/multiracial | 2.13 (1.43–3.19) | 1.19 (0.23–6.21) | 2.21 (1.44–3.38) |
| Education level | |||
| Less than high school graduate | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| High school graduate or GED | 0.99 (0.70–1.40) | 2.95 (0.77–11.37) | 0.95 (0.66–1.37) |
| Greater than high school | 1.39 (1.03–1.89) | 1.66 (0.50–5.53) | 1.39 (1.03–1.87) |
| Family income to poverty ratio a | |||
| 0-1.3 | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| 1.301-3.5 | 1.09 (0.75–1.60) | 0.91 (0.30–.80) | 1.13 (0.77–1.66) |
| 3.501+ | 2.04 (1.33–3.13) | 6.59 (1.51–28.69) | 2.14 (1.41–3.26) |
| Fish and shellfish consumption in the last 30 days | |||
| No | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| Yes | 15.70 (7.83–31.48) | 18.69 (4.91–71.13) | 15.50 (7.48–21.2) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
