Single blood-Hg samples can result in exposure misclassification: temporal monitoring within the Japanese community (United States)

Prenatal mercury exposure and the secondary sex ratio: The Japan Environment and Children's Study

Prior research into the association between prenatal mercury (Hg) exposure and the secondary sex ratio has yielded inconclusive and conflicting results. Notably, no study has used cord blood Hg measurement in this context. Also, the differences in Hg species and the potential modifying role of selenium (Se) on this association remain unexplored. Using data from the Japan Environment and Children's Study, we analyzed mother-child pairs with available data for concentrations of total mercury (THg) and Se in maternal blood during late pregnancy, and THg, inorganic mercury (IHg), methylmercury (MeHg), and Se in cord blood. Logistic regression models were employed to examine the association between Hg and Se biomarkers and the secondary sex ratio. Out of the total sample of 3698 children, 1877 (50.8 %) were male, corresponding to an overall secondary sex ratio of 1.03. After adjusting for maternal age and parity, no significant associations were observed between THg concentrations of maternal blood and the secondary sex ratio. Nevertheless, we identified that two-fold increases in THg, IHg, and MeHg concentrations in cord blood were positively associated with increased odds of having a male child, yielding adjusted odds ratios of 1.13 (95 %CI: 1.04, 1.22), 1.12 (1.03, 1.21), and 1.12 (1.03, 1.22), respectively. When stratified by the median Se concentrations, no apparent differences were detected in the associations between Hg concentrations and the secondary sex ratio. In summary, elevated Hg concentrations in cord blood, but not maternal blood, were associated with an increased probability of male births.

Hair to blood mercury concentration ratios and a retrospective hair segmental mercury analysis in the Northwest Territories, Canada

Concentrations of total mercury were measured in blood and hair samples collected as part of a human biomonitoring project conducted in First Nations communities of the Mackenzie Valley, Northwest Territories, Canada. Hair (n = 443) and blood (n = 276) samples were obtained from six communities in the Dehcho region and three communities in the Sahtú region of the Mackenzie Valley. The aim of this paper was to calculate hair to blood mercury ratios (for matched samples) and determine if: 1) ratios differed significantly between the two regions; 2) ratios differed from the 250:1 ratio proposed by the WHO; and, 3) point estimates of hair to blood mercury ratios could be used to estimate blood mercury concentrations. In addition, this paper aims to determine if there were seasonal patterns in hair mercury concentrations in these regions and if so, if patterns were related to among-season variability in fish consumption. The majority of mercury levels in hair and blood were below relevant health-based guidance values. The geometric mean hair (most recent segment) to blood mercury ratio (stratified by region) was 619:1 for the Dehcho region and 1220:1 for the Sahtú region. Mean log-transformed hair to blood mercury ratios were statistically significantly different between the two regions. Hair to blood ratios calculated in this study were far higher (2-5 times higher) than those typically reported in the literature and there was a large amount of inter-individual variation in calculated ratios (range: 114:1 to 4290:1). Using the 250:1 ratio derived by the World Health Organisation to estimate blood mercury concentrations from hair mercury concentrations would substantially over-estimate blood mercury concentrations in the studied regions. However, geometric mean site-specific hair to blood mercury ratios can provide estimates of measures of central tendency for blood mercury concentrations from hair mercury concentrations at a population level. Mercury concentrations were determined in segments of long hair samples to examine exposure of participants to mercury over the past year. Hair segments were assigned to six time periods and the highest hair mercury concentrations were generally observed in hair segments that aligned with September/October and November/December, whereas the lowest hair mercury concentrations were aligned with March/April and May/June. Mean log-transformed hair mercury concentrations were statistically significantly different between time periods. Between time periods (e.g., September/October vs. March/April), the geometric mean mercury concentration in hair differed by up to 0.22 μg/g, and the upper margins of mercury exposure (e.g., 95th percentile of hair mercury) varied by up to 0.86 μg/g. Results from self-reported fish consumption frequency questionnaires (subset of participants; n = 170) showed total fish intake peaked in late summer, decreased during the winter, and then increased during the spring. Visual assessment of results indicated that mean hair mercury concentrations followed this same seasonal pattern. Results from mixed effects models, however, indicated that variability in hair mercury concentrations among time periods was not best explained by total fish consumption frequency. Instead, seasonal trends in hair mercury concentrations may be more related to the consumption of specific fish species (rather than total wild-harvested fish in general). Future work should examine whether seasonal changes in the consumption of specific fish species are associated with seasonal changes in hair mercury concentrations.

Temporal variation of total mercury levels in the hair of pregnant women from the Maternal-Infant Research on Environmental Chemicals (MIREC) study

Prenatal exposure to total mercury (T-Hg) comes from both natural and anthropogenic sources. T-Hg can cross the blood-brain and placental barriers, and may be associated with future neurological and physiological dysfunctions. Scalp hair is an optimal and non-invasive indicator of chronic T-Hg exposure. As part of the Maternal-Infant Research on Environmental Chemicals (MIREC) Study, hair samples from 350 women were collected within weeks after giving birth, to determine temporal variations in T-Hg levels from preconception to delivery, and to compare these levels to corresponding levels measured in other matrices (maternal and umbilical cord blood, and infant's meconium). A maximum of 12 one-cm hair segments were cut starting at the scalp; segments closer to the scalp reflected recent exposure (within the last month). For proper comparison, the hair segments were matched with the collection dates for other matrices. GM hair T-Hg levels greatly decreased during pregnancy, from 0.26 μg g^-1 (preconception or full-length hair) to 0.18 μg g^-1 (at delivery or segments closer to the scalp). A similar decreasing trend was found for T-Hg in maternal blood: 1st trimester (0.60 μg L^-1) to 3rd trimester (0.47 μg L^-1). The median hair-to-blood ratios of T-Hg levels varied from 364 (1st trimester), to 408 (3rd trimester), to 229 (cord blood). Very low T-Hg levels were detected in meconium. Mercury levels in blood and hair correlated with consumption of large predatory fish.

Determination of Mercury Daily Intake and Hair-to-Blood Mercury Concentration Ratio in People Resident of the Coast of the Persian Gulf, Iran

The objectives of this study were to understand the mercury daily intake and hair-to-blood mercury ratio in fishermen and non-fishermen families in the coast of the Persian Gulf in Iran. The mean mercury concentration in the hair of fishermen and non-fishermen families was 5.76 and 2.27 μg/g, respectively. The mean mercury concentrations of RBCs were obtained for fishermen families and non-fishermen families: 35.96 and 17.18 μg/L, respectively. Hair mercury concentrations in 17% of people were higher than 10 μg/g, the No Observed Adverse Effects Level set by the World Health Organization. 78% of people had a blood mercury value > 5.8 μg/L, the standard level set by the U.S. Environmental Protection Agency. A significant correlation (r = 0.94, p = 0.000) was seen between log hair and RBCs mercury concentrations. The mean mercury daily intake for fishermen and non-fishermen families was 0.42 and 0.20 µg/kg BW per day, respectively. The mean mercury daily intake of fishermen families was higher than the provisional tolerable daily intake (0.23 µg/kg BW per day) suggested by the Joint Expert Committee on Food Additives. Mercury daily intake significantly correlated with fish consumption (r = 0.50, p = 0.000) and log hair mercury (r = 0.88, p = 0.000). The total mean of hair-to-blood mercury concentration ratio was 306. We conclude that the use of mercury concentrations in the hair and RBCs could have been suitable biomarkers for predicting mercury exposure of people with a high rate of fish consumption.

Variability and epistemic uncertainty in water ingestion rates and pharmacokinetic parameters, and impact on the association between perfluorooctanoate and preeclampsia in the C8 Health Project population

We recently utilized a suite of environmental fate and transport models and an integrated exposure and pharmacokinetic model to estimate individual perfluorooctanoate (PFOA) serum concentrations, and also assessed the association of those concentrations with preeclampsia for participants in the C8 Health Project (a cross-sectional study of over 69,000 people who were environmentally exposed to PFOA near a major U.S. fluoropolymer production facility located in West Virginia). However, the exposure estimates from this integrated model relied on default values for key independent exposure parameters including water ingestion rates, the serum PFOA half-life, and the volume of distribution for PFOA. The aim of the present study is to assess the impact of inter-individual variability and epistemic uncertainty in these parameters on the exposure estimates and subsequently, the epidemiological association between PFOA exposure and preeclampsia. We used Monte Carlo simulation to propagate inter-individual variability/epistemic uncertainty in the exposure assessment and reanalyzed the epidemiological association. Inter-individual variability in these parameters mildly impacted the serum PFOA concentration predictions (the lowest mean rank correlation between the estimated serum concentrations in our study and the original predicted serum concentrations was 0.95) and there was a negligible impact on the epidemiological association with preeclampsia (no change in the mean adjusted odds ratio (AOR) and the contribution of exposure uncertainty to the total uncertainty including sampling variability was 7%). However, when epistemic uncertainty was added along with the inter-individual variability, serum PFOA concentration predictions and their association with preeclampsia were moderately impacted (the mean AOR of preeclampsia occurrence was reduced from 1.12 to 1.09, and the contribution of exposure uncertainty to the total uncertainty was increased up to 33%). In conclusion, our study shows that the change of the rank exposure among the study participants due to variability and epistemic uncertainty in the independent exposure parameters was large enough to cause a 25% bias towards the null. This suggests that the true AOR of the association between PFOA and preeclampsia in this population might be higher than the originally reported AOR and has more uncertainty than indicated by the originally reported confidence interval.

Impact of Exposure Uncertainty on the Association between Perfluorooctanoate and Preeclampsia in the C8 Health Project Population

BACKGROUND

Uncertainty in exposure estimates from models can result in exposure measurement error and can potentially affect the validity of epidemiological studies. We recently used a suite of environmental models and an integrated exposure and pharmacokinetic model to estimate individual perfluorooctanoate (PFOA) serum concentrations and assess the association with preeclampsia from 1990 through 2006 for the C8 Health Project participants.

OBJECTIVES

The aims of the current study are to evaluate impact of uncertainty in estimated PFOA drinking-water concentrations on estimated serum concentrations and their reported epidemiological association with preeclampsia.

METHODS

For each individual public water district, we used Monte Carlo simulations to vary the year-by-year PFOA drinking-water concentration by randomly sampling from lognormal distributions for random error in the yearly public water district PFOA concentrations, systematic error specific to each water district, and global systematic error in the release assessment (using the estimated concentrations from the original fate and transport model as medians and a range of 2-, 5-, and 10-fold uncertainty).

RESULTS

Uncertainty in PFOA water concentrations could cause major changes in estimated serum PFOA concentrations among participants. However, there is relatively little impact on the resulting epidemiological association in our simulations. The contribution of exposure uncertainty to the total uncertainty (including regression parameter variance) ranged from 5% to 31%, and bias was negligible.

CONCLUSIONS

We found that correlated exposure uncertainty can substantially change estimated PFOA serum concentrations, but results in only minor impacts on the epidemiological association between PFOA and preeclampsia.

CITATION

Avanasi R, Shin HM, Vieira VM, Savitz DA, Bartell SM. 2016. Impact of exposure uncertainty on the association between perfluorooctanoate and preeclampsia in the C8 Health Project population. Environ Health Perspect 124:126-132; http://dx.doi.org/10.1289/ehp.1409044.

Prenatal mercury exposure, autism, and developmental delay, using pharmacokinetic combination of newborn blood concentrations and questionnaire data: a case control study

BACKGROUND

Methylmercury (MeHg), known for well over a century as a neurotoxin in adults, has more recently been studied for potential detrimental effects during early brain development. While several studies have estimated mercury exposure, they usually rely on either a single biomarker or questionnaire data, each of which has limitations. The goal of this paper was to develop a toxicokinetic model that incorporates both biomarker and questionnaire data to estimate the cumulative exposure to MeHg through seafood consumption using data collected from the Childhood Autism Risks from Genetics and the Environment (CHARGE) study.

METHODS

We utilized a previously described discrete-time model that estimates blood MeHg concentration given a piecewise-constant ingestion rate and single-compartment pharmacokinetics. We measured newborn bloodspot Hg concentrations and obtained information pertaining to maternal fish consumption using a questionnaire. Using MeHg concentration estimates from the toxicokinetic model, cumulative MeHg exposure was estimated in children with autism, children with developmental delay, and typically developing children. Median estimated cumulative MeHg was compared among diagnostic groups using the Kruskal-Wallis Test. Multinomial logistic regression models were constructed to assess the association between cumulative MeHg concentration and the risk of autism and developmental delay (vs. typical development).

RESULTS

The estimated average MeHg concentration of for all fish species consumed by mothers was 42 ppb. Median cumulative MeHg over gestation was similar across diagnostic groups (p-values raged from 0.91 to 0.98). After adjusting for potential confounding, we found no association between cumulative MeHg exposure and the risk of autism (OR = 0.95, 95% CI: 0.95, 1.12) or developmental delay (OR = 1.00, 95% CI: 0.89, 1.13).

CONCLUSIONS

The toxicokinetic model described in this paper yielded fish MeHg concentration estimates that are consistent with fish species containing lower levels of MeHg. Overall, cumulative MeHg exposure does not appear to detectably elevate the risk of autism or developmental delay. Based on the regression standard error for the association between ASD and TD, we would have reported statistical significance for an adjusted odds ratio of 1.09 or larger. This method can easily be extended to other epidemiologic studies in which there is a biomarker measurement and questionnaire data regarding exposure.

The complexity of hair/blood mercury concentration ratios and its implications

BACKGROUND

The World Health Organization (WHO) recommends a mercury (Hg) hair-to-blood ratio of 250 for the conversion of Hg hair levels to those in whole blood. This encouraged the selection of hair as the preferred analyte because it minimizes collection, storage, and transportation issues. In spite of these advantages, there is concern about inherent uncertainties in the use of this ratio.

OBJECTIVES

To evaluate the appropriateness of the WHO ratio, we investigated total hair and total blood Hg concentrations in 1333 individuals from 9 First Nations (Aboriginal) communities in northern Québec, Canada.

METHODS

We grouped participants by sex, age, and community and performed a 3-factor (M)ANOVA for total Hg in hair (0-2 cm), total Hg in blood, and their ratio. In addition, we calculated the percent error associated with the use of the WHO ratio in predicting blood Hg concentrations from hair Hg. For group comparisons, Estimated Marginal Means (EMMS) were calculated following ANOVA.

RESULTS

At the community level, the error in blood Hg estimated from hair Hg ranged -25% to +24%. Systematic underestimation (-8.4%) occurred for females and overestimation for males (+5.8%). At the individual level, the corresponding error range was -98.7% to 1040%, with observed hair-to-blood ratios spanning 3 to 2845.

CONCLUSIONS

The application of the ratio endorsed by the WHO would be unreliable for determining individual follow-up. We propose that Hg exposure be assessed by blood measurements when there are human health concerns, and that the singular use of hair and the hair-to-blood concentration conversion be discouraged in establishing individual risk.

Association of body burden of mercury with liver function test status in the U.S. population

The majority of mercury (Hg) exposure in the US population is from consumption of fish contaminated with methylmercury (MeHg). Since inorganic Hg is the predominant form excreted in the feces and urine, hepatic biotransformation is a critical step in its normal clearance. This study was set to test the hypothesis that compromised liver function is associated with body burden of Hg as indirectly reflected by Hg sampled in blood and urine. From the National Health and Nutrition Examination Survey (NHANES, 2003-2008), 3769 adults aged 20 years and above were selected for analysis. Hepatic function was inferred from the three standard serum liver-related enzyme activities, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and γ-glutamyltransferase (GGT). Multivariate regression models were used to examine the associations of interest. Although urinary Hg was significantly correlated with serum Hg, the blood-urinary Hg relationship was influenced by liver function, which is also a function of demographic and lifestyle factors (e.g., gender). Although the results were only marginally significant for examined enzymes (p=0.06-0.08), urinary Hg tended to be lower among subjects with elevated liver enzymes, as compared to those with normal enzyme measurements. Conversely, MeHg generally represents a higher fraction of the total circulating Hg among those with elevated liver enzyme levels, especially among participants with elevations in all three enzymes (p=0.01). In conclusion, this population-based study identified an association between liver function, serum Hg and urinary Hg. Urinalysis may not be the optimal approach to monitor Hg elimination toxicokinetics or Hg exposure, since the majority of Hg excretion is fecal and the fidelity of urinary excretion may depend on healthy liver function. Future prospective studies are warranted to expand these findings.

Prenatal methylmercury exposure through maternal rice ingestion: insights from a feasibility pilot in Guizhou Province, China

Maternal hair and blood were investigated as biomarkers for prenatal methylmercury (MeHg) exposure among seventeen mothers recruited at parturition in Wanshan, Guizhou Province, China, where rice ingestion was the primary MeHg exposure pathway and atmospheric mercury (Hg) levels were elevated. For all three trimesters (n = 51), hair total Hg (THg) and MeHg concentrations ranged from 0.27 to 4.9 μg/g (median: 0.96 μg/g) and 0.077 to 2.3 μg/g (median: 0.43 μg/g), respectively, while blood THg levels ranged from 1.7 to 11 μg/L (median: 3.0 μg/L, n = 17). Despite adequate hair washing procedures, median %MeHg (of THg) was 37% (range: 14-89%, n = 51), indicating exogenous inorganic Hg(II) contamination or incorporation of elemental Hg (Hg(o)) into the hair shaft were important. Rice MeHg levels (n = 17) were highly correlated with blood THg (r(2) = 0.66) compared to hair MeHg (r(2) = 0.31) (when variables were log10-transformed), suggesting blood THg was a more preferable biomarker for prenatal MeHg exposure within this population.

Chronologically matched toenail-Hg to hair-Hg ratio: temporal analysis within the Japanese community (U.S.)

BACKGROUND

Toenail-Hg levels are being used as a marker of methylmercury (MeHg) exposure in efforts to associate exposure with effects such as cardiovascular disease. There is a need to correlate this marker with more established biomarkers that presently underlie existing dose-response relationships in order to compare these relationships across studies.

METHODS

As part of the Arsenic Mercury Intake Biometric Study, toenail clippings were collected at three time points over a period of one year amongst females from within the population of Japanese living near Puget Sound in Washington State (US). Variability in temporal intra-individual toenail-Hg levels was examined and chronologically matched hair and toenail samples were compared to more accurately define the toxicokinetic variability of Hg levels observed between the two compartments.

RESULTS

Mean toenail-Hg values (n=43) for the 1st, 2nd and 3rd visits were 0.60, 0.60 and 0.56 ng/mg. Correlations were as follows: r=0.92 between 1st and 2nd clinic visits, r=0.75 between 1st and 3rd visits and r=0.87 between 2nd and 3rd visits. With few exceptions, toenail-Hg values from any visit were within 50-150% of the individual's mean toenail-Hg level. Nearly all participants had less than a two-fold change in toenail-Hg levels across the study period. A regression model of the relationship between toenail-Hg and hair-Hg (n = 41) levels representing the same time period of exposure, gave a slope (Hg ng/mg) of 2.79 for hair relative to toenail (r=0.954).

CONCLUSIONS

A chronologically matched hair-Hg to toenail-Hg ratio has been identified within a population that consumes fish regularly and in quantity. Intra-individual variation in toenail-Hg levels was less than two-fold and may represent dietary-based fluctuations in body burden for individuals consuming various fish species with different contaminant levels. The chronologically matched ratio will be useful for relating MeHg exposure and dose-response derived from toenail-Hg measurements to those derived from hair-Hg measurements in other studies, and may be useful in future investigations as an indicator of stable MeHg body burden within a population.