3e-2: Human exposure and health effects of mercury: epidemiology and population studies

The Centers for Disease Control and Preventions Inorganic and Radiation Analytical Toxicology Laboratory measures total mercury, inorganic mercury, ethyl mercury, and methyl mercury in people age one and older in the U.S. population as part of the National Health and Nutrition Examination Survey (NHANES). Analysis of specific mercury species began in 2011 and has allowed us to determine baseline levels of inorganic, methyl, and ethyl mercury for the United States population. A number of studies have looked at either total mercury or total and inorganic mercury in specific demographic groups, but we have four years of data that looks at those forms of mercury and additional forms across several demographic groups. We evaluated all available NHANES data for total, methyl mercury, and inorganic mercury for survey years 2003 through 2014. With new data from NHANES 2013-14, we have a more comprehensive picture of the trends of mercury in the U.S population over time.

We were able to statistically evaluate the data for the 95th percentile of the population and identify the key demographic characteristics including gender, ethnicity, education and household income that have an impact on the likelihood of an individual being placed into the 95th percentile. This statistical evaluation was completed for total, inorganic, and methyl mercury to further assess the differences that exist. Beginning in survey year 2011, the survey started oversampling Asians which affords us the opportunity to provide estimates of Asians in addition to Non-Hispanic White, Non-Hispanic Black and all Hispanic race/ethnicity groups. There is a statistically significant difference in the prevalence of mercury in the Asian segment of the population that was surveyed as opposed to other ethnic groups. This correlates to dietary trends (i.e. higher fish consumption). Household income is another demographic variable in survey years 2011 through 2014 where an increase in income has led to an increase in mercury exposure, possibly through higher seafood consumption. NHANES mercury data provides a unique opportunity to evaluate mercury trends over time in a representative sampling of the U.S. population. An overview and trends of the NHANES speciated mercury results from 2003 to 2014 will be presented along with highlights of significant findings of different demographic characteristics.

India is a major consumer of coal. There are 132 coal-fired power plants (431 units) in India, which meet >51% of India’s commercial energy demand. In this work, we investigated the possible role of power plants in increasing the local exposure to mercury. We also develop baseline for several demographics. A new category, oldsmiths, different to ASGM workers, is also included.

We analysed mercury in hair from 668 volunteers in three cities in South India (Hyderabad (n = 103) in interior, Vasco da Gama (n = 326) on the west coast and Nellore (n = 239) on the east coast that has 4 active coal-fired power plants). Baseline Hg levels in some demographic categories were: subsistence fishermen/women (n = 66) [Geometric Mean (GM): 0.38 μg/g (95% CI: 0.34-0.42 μg/g)], goldsmiths (n = 21) [GM: 0.22 μg/g (95% CI: 0.18-0.28 μg/g)]. Women of child-bearing age had Hg levels of (0.14 µg/g (95% CI: 0.12-0.16 µg/g), similar to US NHANES and European DEMOCOPHES studies after accounting for diet and age.

The GM of hair Hg levels of people from Nellore (n = 239) [0.19 μg/g (95% Confidence Interval (CI): 0.09-0.37 μg/g)] are statistically (p < 0.05) higher compared to both Vasco da Gama (n = 326) [0.13 μg/g (95% CI: 0.05-0.37 μg/g)] and Hyderabad (n = 103) [0.08 μg/g (95% CI: 0.04-0.15 μg/g)]. This was true for different fish consumption frequencies (rare, once a week, thrice a week), even though fishes consumed were from similar trophic levels. Majority of volunteers consumed rice and which was primarily locally grown. Vegetarians (who never ever consumed fish) from Nellore (n = 37) [0.09 μg/g (95% CI: 0.03-0.13 μg/g)] had higher (p < 0.05) GM hair Hg values compared to vegetarians from Vasco da Gama (n = 47) [0.03 μg/g (95% CI: 0.01-0.06 μg/g)]. Preliminary results indicate that rice samples from Nellore appear to have higher concentration of Hg than rice samples from other sites.

Our results suggest that exposure to mercury may be higher in Indian regions that have active mercury sources in their vicinity. Since there are more than 100 power plants in India, the implications to mercury exposure may be substantial. We are now studying mercury isotopes in hair samples from our study population to further investigate this hypothesis, and results will be presented in the conference.

Mercury (Hg) is a heavy metal pollutant with high toxicity. It is being released into aquatic ecosystems of Colombia by gold mining activities, where it is used to extract the precious metal. One of these ecosystems is the Magdalena River, which near the Caribbean Coast, it provides freshwater to the Ciénaga Grande de Santa Marta (CGSM), the largest and most productive coastal lagoon complex in Colombia. To evaluate mercury pollution status in the CGSM, total mercury (T-Hg, fresh weight) was measured in the hair of 158 habitants from two stilt-house towns, Nueva Venecia and Buena Vista. Total Hg was also measured in the muscle and liver of several common fish species. The measurements were performed using a direct mercury analyzer, DMA-80. Low T-Hg levels were detected in hair from humans, with average concentrations of 0.71±0.04 and 0.70±0.08 ppm, for Nueva Venecia and Buena Vista, respectively. Males had greater T-Hg concentrations (0.81±0.05 ppm) than females (0.63±0.04 ppm). Occupation is also a factor that determines T-Hg in hair, with the largest average detected in fishermen (0.87±0.06 ppm), when compared to other examined groups, such as housewives or students. Low levels of T-Hg were found in fish muscle, although there were significant inter-species differences in Hg content: Cathorops mapale (0.094±0.003 ppm, n=36) > Notarius bonillai (0.083±0.001 ppm, n=2) > Megalops atlanticus (0.081±0.002 ppm, n=22) > Mugil incilis (0.079±0.003 ppm, n=34) > Elops saurus (0.078±0.003 ppm, n=17). Total Hg in liver showed not significant differences between Mugil incilis (0.082±0.004 ppm, n=13) and Megalops atlanticus (0.088±0.004 ppm, n=22). These results indicate that consumption of fish from CGSM is not resulting in Hg accumulation in humans. Therefore, the risk of developing Hg-related health problems is considered negligible. Much more important is the fact that in some places of Colombia, it is still possible to find aquatic habitats where Hg is not a concern for human health and welfare.

Background. World-wide current recommendations for pregnant women are to eat fish with caveats to avoid species with high levels of mercury resulting in a reduction in seafood consumption in some pregnant women. There is conflicting evidence concerning the harm from mercury in regard to offspring cognitive outcomes if the woman eats fish.

Methods. This study uses the ALSPAC study of pregnancies in 1990­­­–92. Mercury levels were measured in whole blood in early pregnancy, and outcomes for 2062 children were the verbal, performance and total IQ measured at age 8 treated as (a) continuous and (b) as the lowest 25% of the distribution. Multiple and logistic regression analyses took account of social and demographic variables, and stratified analyses considered children of fish eaters separately.

Findings. Before adjustment, the mean IQs increased with increasing mercury (change with 1SD of mercury = +2·02; 95%CI +1·40, +2·64 IQ points); after adjustment the effect size was reduced although still positive (+0·61: -0·06, +1·29 IQ points). However when mothers who ate fish were considered separately, the adjusted positive relationship was stronger (+0·84: +0·13, +1·56 IQ points), in comparison with the outcomes for the women who did not eat fish where the adjusted relationship was negative (-2·22: -5·00, +0·56 IQ points). Results for the binary outcome showed a similar pattern with the adjusted OR for non-fish-eaters 1.79[95%CI 1.10,2.93], significantly different from that for the fish consumers [0.94; 95%CI 1.10,2.93].

Interpretation. This suggests that the relationship between intrauterine exposure to mercury is benign in regard to offspring IQ provided the mother consumed fish. On the present evidence, pregnant women should be assured that consumption of fish is likely to be more beneficial than harmful to the developing child.

Background: Artisanal and small-scale gold mining (ASGM) is the major source for mercury emissions globally. Toxic mercury causes negative health effects especially for miners working directly with mercury such as panners and amalgam smelters. Good and reliable data is difficult to obtain, because ASGM activities are largely informal and take place in remote mining areas. Thus, the global burden of disease due to mercury used in ASGM is unknown.

Objective: To calculate a first rough estimate of disability-adjusted life years (DALYs) caused by chronic metallic mercury vapor intoxication (CMMVI) in miners worldwide resulting from the use of mercury in ASGM.

Methods: A primary data set was set up including all obtainable studies where data was sufficient to determine the prevalence rate of CMMVI with a diagnostic tool. Further studies reporting only mercury levels in urine were reanalyzed using the primary dataset to obtain a valid prevalence rate. The prevalence rate was multiplied with the number of miners for each country taken from a literature search which supplemented available reviews and the disability weight (DW) for moderate CMMVI. The DW contains information about the severity of the intoxication in comparison to other diseases. Mortality was ruled out as outcome. Uncertainty intervals (UI) were quantified to account for uncertainties in the input data (number of miners, DW).

Results: Concentrating on moderate CMMVI, the burden of 14 to 19 million gold miners ranges from 1.22 (UI 0.87-1.61) to 2.39 (UI 1.69-3.14) million DALYs due to moderate CMMVI. This estimate is based on human biomonitoring data from 3,194 gold miners. A number of 62 countries were taken into account of which most countries with available data belong to the African region. Subject to the quantification, nearly every fourth up to every third miner is expected to be mercury intoxicated.

Conclusions: Our results give a first rough estimate of global DALYs, based on country estimates of the use of mercury in ASGM. A major limiting factor is the limited data availability regarding human biomonitoring data, health assessments, and the number of affected miners. We estimate the disease burden of about 3.3 to 6.5 million intoxicated miners. Our results show, that ASGM is a global environmental health hazard. Further research is needed to improve the model input parameters.

This work has received funding from Pure Earth and the European Unions Seventh Programme for research, technological development and demonstration under grant agreement No 603946 (HEALS).

Conflict of interest: The authors declare no competing financial interests.

Workers within artisanal and small-scale gold mining (ASGM) communities are amongst the highest Hg-exposed groups worldwide. While formalization of this sector has been suggested as a means to reduce exposures and improve health and safety, we are unaware of empirical evidence that supports this notion. The objective of this study was to compare mercury exposure profiles among miners working in licensed versus un-licensed ASGM sites. To achieve this, 404 small-scale miners were recruited in 2014 from 9 mining sites in Tarkwa (Ghana), of which 5 were licensed and 4 were not licensed. Miners were interviewed, and urine and hair samples were taken for measurements of total Hg content in a subset (n=316). Sociodemographic characteristics of miners from the two groups were relatively similar (33.8 yrs old, 92% male, 29% completed high school, 66% living with a partner). Those currently working in an un-licensed mine have higher mean levels of total Hg in urine (110.7 vs 16.0 ug/L; p<0.001) and hair (4.5 vs. 2.0 ug/g; p=0.09) compared to workers from the licensed mines. A range of other information was collected showing differences between these groups of miners such as work history and injury episodes, and these will be discussed. These findings advance our understanding of mercury exposure (and other differences) amongst ASGM workers especially important differences between miners working in a licensed versus and un-licensed site.

In May 2016, the Peruvian Ministry of Health (MINSA) declared a State of Emergency due to high mercury levels reported in Madre de Dios (MDD), Peru. This declaration stemmed from three interrelated projects studying the impacts of road construction, gold mining, and gas extraction on human and environmental health that recorded elevated mercury levels in human and environmental samples. Here, study designs, exposure assessments, and preliminary findings associated health outcomes to mercury exposure are reported. The studies include: (1) Interoceanic Highway Study (IOH); (2) Rio MDD Study (RIO); and (3) Amarkaeri Cohort Study (AmCS). The IOH and RIO studies were population-based samples (a 2-stage selection design and a stratified sampling design, respectively), while the AmCS was a census of rural communities and a 50-75% sample of three peri-urban communities all surrounding the Amarakaeri Reserve. Total mercury contents in hair were evaluated in all household members for the IOH and RIO studies, but only in sentinel groups for the AmCS (women of child-bearing age, their children and spouse). In total, hair mercury was evaluated in 724, 231, and 2308 persons from the IOH, RIO and AmCS, respectively. Among all samples, median mercury levels were high (1.47 ug/g, IQR: 0.71-3.01), with 57.5% and 39.2% exceeding the USEPA (1.2ug/g) and PERU/WHO (2.0 ug/g) exposure thresholds, respectively. High hair mercury was correlated with increased fish consumption in RIO participants, particularly high trophic level fish, negatively correlated with quinoa and kiwicha consumption (Wyatt et.al in-review). Spatial interpolation of mercury levels identified the highest exposures in communities 90-120 km upstream of ASGM (n=216), where 96.8% and 90.7% exceeded the USEPA and PERU/WHO guidelines (20% exceeded 9.0 ug/g). In addition, fish, sediment, and water were obtained from river sections near, upstream and downstream of ASGM (Diringer et.al 2015): The highest total Hg in fish and total Hg and methylmercury in sediments were observed within and downstream from ASGM concession areas. River sites upstream of mining did not have high Hg in sediment or fish. These two findingslow environmental mercury, but high human exposureare in direct contrast that may be due to river transport of Hg-containing sediment, behavioral differences in diet, or ecosystem services affecting methylation.Preliminary analyses evaluating health impacts suggest significant associations between mercury levels and mitochondrial DNA copy number (RIO), anemia risk (AmCS), impaired neurocognitive function in children (AmCS) and impaired kidney function (AmCS).

The aim of this work was to study mercury accumulation in inhabitants from smelting area, and their health risk due to mercury exposure. The investigation included the: (a) distributions of Hg in diet, dust and water, (b) mercury accumulation and influencing factors in hair and blood of inhabitants, and (c) assessment of health risks due to mercury exposure. Smelting activities from Huludao zinc plant in Liaoning province, northeast China had seriously contaminated the surrounding environment. In the smelting district of Huludao, THg contents of the seeds of maize, soybean, and broomcorn are 0.008, 0.006, and 0.057 mg/kg, respectively, exceeding the maximum level of contaminant in food (GB2762-2005) by 4.7 times. The edible parts of vegetables are also contaminated with a range of mercury contents of 0.001–0.147 mg/kg (dry weight). The average and maximum mercury daily intake (DI) of adult around the Huludao zinc plant via consuming vegetables are 0.015 and 0.051 lg/kg/d, respectively, and those of children are 0.02 and 0.07 lg/kg/d, respectively. The average and maximum weekly intakes of total mercury for adult are 2.1 and 7.1%, respectively, of the provisional tolerable weekly intake (PTWI), and 2.8 and 9.7%, respectively, of the PTWI for children. The maximum Hg, contents in dust were 5.324 mg/kg, and were 144 times as high as the background values in soil. Children were also experiencing the potential health risk from Cd and Pb exposure from dust near HZP, not include Hg. The average mercury content in hair of Huludao city inhabitants were 0.43 mg/kg. Effect of different age, occupation, gender, and living area of Huludao city residents was different to mercury in hair. The concentrations of Hg in different age groups decreased in the following order: young and middle aged > Youth >adolescent > the elderly (Hg). The gradual degradation of the digestive system of the elderly may make contribution to the reduction of heavy metals content in the hair. The average content of Hg in male resident’s hair was roughly equal to that of female residents, but the content of male dispersion range was less than that of women. The average mercury content in blood of women was 1.75µg/L while these of children was 3.26 µg/L. Moreover, effect from Hg was much less than Pb and Cd for women and children.