Assessing and Managing Methylmercury Risks Associated With Power Plant Mercury Emissions in the United States

Gail Charnley, PhD

In This Article

Health Risks From Methylmercury Exposure

Developmental Neurotoxicity

Methylmercury is a known developmental neurotoxicant when ingested at high levels. During the 1950s-1970s, methylmercury poisoning incidents occurred in Japan and Iraq. In Japan, women who were consuming fish that were heavily contaminated with methylmercury from industrial discharges into Minimata Bay gave birth to children with severe psychomotor retardation. Some adults were also affected.[23] In Iraq, both children and adults experienced toxicity, including neurologic damage, after consuming methylmercury-treated seed grains. More than 6500 people were hospitalized and 459 died.[24] In Japan, the average fish methylmercury concentrations ranged from 9 ppm to 24 ppm, and the amount of fish that was consumed was about 20 times greater than the amount consumed by recreational fishers in the United States.[25] In Iraq, those exhibiting symptoms of neurotoxicity were estimated to have consumed between 200 mg and 1000 mg of methylmercury over a period between 1 and 3 months (about 2-33 mg/day).[26] These poisoning incidents are difficult to extrapolate to the much lower mercury exposure levels commonly associated with fish consumption in the United States (mean, 0.0008 mg/day; see Table 1 ), and the extent to which mercury is a developmental neurotoxicant at US environmental levels of exposure remains controversial.

Several, recent prospective epidemiologic studies have attempted to characterize the effects of methylmercury at environmental levels of exposure by evaluating subtle neurologic changes among children exposed prenatally and postnatally in societies relying heavily on fish and marine mammals as sources of protein. The results of those studies are conflicting. In the Seychelles Islands, a study involving 711 mother-child pairs found no positive associations between mercury exposure as determined by hair total mercury level and the neurodevelopmental outcomes evaluated at 66 months of age.[27] The study authors concluded that exposure to methylmercury from a diet that is high in ocean fish (maternal average, 12 fish meals per week) posed no neurodevelopmental risks through 66 months of age. A follow-up of the same cohort at 9 years of age also found no support for the hypothesis that there is a neurodevelopmental risk from prenatal methylmercury exposure resulting from ocean fish consumption,[28] although subtle effects that remained undetected cannot be ruled out. The mean prenatal methylmercury exposure was 6.9 parts per billion.

In the Faroe Islands, 917 7-year-old children were tested for their performance on tasks that were associated with the neuropathologic abnormalities seen in the earlier methylmercury poisoning incidents in Japan and Iraq.[29,30] The tests evaluated fine motor tasks, attention measures, executive function tests, language tests, short-term memory tests, and visuospatial tasks. Mercury exposure was characterized by measuring mercury levels in maternal hair when the child was born, the child's hair at 12 months and 7 years of age, the child's cord blood, and the child's blood at 7 years of age. Multiple regression analyses found that cord blood methylmercury concentrations showed the clearest associations with decreased attention, memory, and language.

The Faroe Islands study results have been questioned due to potential confounding by the relatively high levels of polychlorinated biphenyls (PCBs) to which this population is also exposed from consuming pilot whale blubber and to which children were exposed in utero and postnatally via breast milk.[8,31] PCBs are considered developmental neurotoxicants by the EPA. Reanalyses of the data have concluded that the adverse effects observed were not confounded by prenatal PCB exposure, although postnatal PCB exposure was not evaluated.[8] Steuerwald and coworkers[32] evaluated postnatal exposure to PCBs from breast milk and reported an absence of confounding. However, their analysis evaluated PCB effects only at 2 weeks of age and would have missed any later impacts. Figure 2 compares Faroe and Seychelles adult methylmercury intakes from fish and/or whales, Faroe adult PCB intake, and Faroe child PCB and mercury intakes from breast milk. The level at which PCBs have produced developmental neurotoxicity in infant monkeys fed PCB-contaminated breast milk[33] is also shown. Faroe infant exposure to PCBs from breast milk is almost twice the level that is toxic to infant monkeys and almost 1000 times higher than the EPA limit on exposure that is considered safe. In contrast, the Seychelles study, which reported no impact of methylmercury exposure on neurodevelopment, also detected no exposure to PCBs.[27] Toxicology suggests that the developmental effects seen in Faroe children are more likely to be attributable to PCBs in breast milk than to methylmercury (or to a combination of the 2),[31] but further analysis is needed to support that conclusion.

Figure 2.

Contaminant intakes in the Faroe and Seychelles Islands. Polychlorinated biphenyl (PCB) intake from breast milk exceeds the dose that is developmentally neurotoxic in infant monkeys and almost 1000 times higher than the EPA reference dose for PCBs (source: Dourson and coworkers[31]).

In New Zealand, general intelligence, academic attainment, language development, social adjustment, and fine and gross motor coordination were assessed in a group of 230 6- and 7-year-old children.[34,35] The study authors concluded that there was an "apparent consistent association" between prenatal exposure to high levels of methylmercury (based on either "higher" or "lower" maternal hair mercury levels) and decreased performance on scholastic and psychological tests. Because the New Zealand cohort was small, however, and because of technical problems with the study, it is less reliable than the Seychelles or Faroe Islands cohorts.

All 3 epidemiologic studies of the developmental effects of methylmercury are concerned with effects that are very small compared with normal variation, leading to several problems. First, because the signal-to-noise ratio is very low, it is difficult to detect or measure the size of the effect, which may explain some of the differences between the studies. Second, it is very easy for a secondary correlation to increase or diminish the apparent effect, which is why correlation does not prove causation. In addition to PCBs in the Faroe Islands, other potential confounding factors include urban vs rural upbringing (ie, the children with lower exposures in the Faroe Islands tended to live in town) and the potential negative correlation from beneficial effects from fish consumption, which may have been realized to a lesser extent in the Faroe Islands than in the Seychelles Islands or New Zealand because many of the high methylmercury exposures were from whale consumption, not fish. Because it is not possible to draw any strong conclusions from these studies, the fact that the effects observed in the Iraqi poisoning incident are much larger relative to background variation makes that study worthy of consideration when evaluating risk at environmental exposures despite the data collection problems associated with that study.[36,37]

A recent study evaluated the relationship between blood mercury levels and neurobehavioral function in an aging US population participating in a study of cognitive decline with age. For the 474 participants, aged 50-70 years, no evidence of an association between mercury levels and worse neurobehavioral performance was found.[38]

Coronary Heart Disease

Many studies have demonstrated the clear cardiovascular benefits of regular fish consumption, despite the potential for mercury exposure. For example, among the high fish-consuming native peoples of James Bay in Quebec, Canada, a period of high methylmercury exposure from fish was related to a period of low coronary heart disease risk.[39,40] Eating broiled or baked fish (but not fried), including tuna, has been associated with a lower incidence of irregular heartbeat among the elderly.[41] Eating oily fish, such as salmon, tuna, or bluefish, at least twice a week can prevent sudden cardiac death.[42] The American Heart Association recommends that individuals consume 2 servings of a variety of fish weekly, both for the benefits of omega-3 fatty acids and because fish tends to be low in saturated fats, which contribute to elevated cholesterol levels.

Recently, however, a study of Finnish men found an association among the highest third of hair mercury content and an approximately 60% greater prevalence of coronary heart and cardiovascular diseases compared with men with the lower two thirds of hair mercury content.[43] The men who were least likely to experience heart problems were those who had both low levels of hair mercury and high blood levels of fatty acids found in fish that are known to reduce the risk for heart disease. Attempts to correlate hair mercury content with fish consumption were tenuous, with one third of the men in the highest hair mercury group reporting higher fish consumption than the other study participants. No information was provided on whether high- or low-mercury-containing types of fish were consumed. Contrary to the large body of epidemiologic evidence showing a negative correlation between fish consumption and heart disease, the population of Eastern Finland has a high rate of heart disease regardless of high fish consumption,[44] suggesting that factors other than methylmercury are responsible for elevated risk.[45] The Finnish results were considered preliminary by the American Heart Association, which has concluded that when consumed according to the established FDA/EPA guidelines, the cardiovascular benefits of eating fish far outweigh the risks for middle-aged and older men and women after menopause.[46]

Numerous other factors may contribute to the discrepancies among studies, such as differences in contaminant levels and in the omega-3 fat content of the fish consumed, the extent of variability in fish consumption in the population, background risk, the nature of the endpoints examined,[47] and differential exposure to inorganic or methylmercury (both of which are reflected in total hair mercury levels).

The mechanisms of mercury cardiotoxicity, if any, are not known. Multiple mechanisms are plausible, however, including altered cardiac sodium handling and modified responses to viral infections.[48] Methylmercury can accumulate in heart tissue.[49] Thus, although the evidence that mercury is associated with coronary heart disease risk is contradictory and there is insufficient evidence to conclude that mercury is associated with risk, the suggestive positive findings and the plausible biological modes of action warrant additional research.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.