Mercury (Hg) contamination of the marine environment is an important public health concern globally [1
]. Inorganic mercury is released into the atmosphere from a variety of industrial activities and subsequently deposited into aquatic and marine ecosystems through rainfall and by dry deposition [2
]. Inorganic Hg is converted into methylmercury (MeHg) by sulfate-reducing anerobic bacteria in sediments. MeHg is biomagnified through trophic levels in the marine food web and accumulates in apex predators like dolphins and sharks [3
]. Human exposure to mercury occurs primarily through the diet by consumption of seafood, particularly large predatory fish such as swordfish, shark, and albacore tuna [4
The ability of MeHg to pass through the placenta and cross the blood brain barrier of the fetus resulted in the well-known disaster at Minamata Bay, Japan, where fetal abnormalities, blindness, and severe physical and developmental retardation were reported in the offspring of pregnant women who consumed highly contaminated, locally obtained seafood [5
]. The sensitivity of the developing brain to the effects of mercury deposition has been shown in studies of pregnant women exposed through the consumption of seafood, even at relatively low levels of prenatal Hg [6
Populations living in proximity to marine ecosystems may be at increased risk of exposure to MeHg due to the availability and increased consumption of locally obtained fish and shellfish [7
]. In Florida the average adult consumes almost ten times as many grams of seafood per day compared to the general population of the United States [8
], potentially increasing the risk of Hg exposure above safe limits. The Indian River Lagoon (IRL), an estuary which extends >250 km and traverses 40% of the eastern coastline in Florida, is an example of a highly impacted environment [7
]. Atmospheric deposition of Hg is consistently higher in Florida than in other regions of the country [9
]. The increased deposition is compounded by the high capacity for methylation in estuarine waters with low turnover rates and flushing as exists in the IRL [10
]. As a result, apex predators like resident bottlenose dolphins (Tursiops truncatus
) have been reported to have some of the highest concentrations of total mercury (THg) in blood and skin described in the species worldwide [11
]. A series of epidemiological studies was conducted to examine THg concentrations in dolphin prey species [12
] and dolphins [11
These studies led us to conduct a study of hair THg concentrations among recreational anglers and coastal residents to assess human exposure and to “Close the Loop” between the wildlife sentinel and human health [7
]. The THg concentrations in the hair of 135 participants (mean 1.53 ± 1.89 μg/g) was higher than that previously reported for similar populations in the United States [7
]. Additionally, individuals who reported consuming locally caught seafood three times a week or more had a threefold risk of having a hair THg concentration above 1.0 µg/g, the US Environmental Protection Agency (EPA) exposure guideline [7
]. The data raised concern about the risk to the subpopulation most vulnerable to mercury exposure, pregnant women. Therefore, the objective of the current study was to assess mercury concentrations in the hair of pregnant women living in coastal Florida and to determine the relationships between hair THg concentrations, fish consumption, sources of seafood, knowledge of the risks of mercury exposure, and seafood consumption during pregnancy.
Despite the fact that southern Florida is an area of selective deposition of atmospheric mercury, and that mercury is bioaccumulated in local fish species and apex predators, the mean concentration of THg in this sample of pregnant women (0.31µg/g + 0.54) was similar to or lower than those reported elsewhere in the United States. In the National Health and Nutrition Examination Survey (NHANES), a representative probability sample of the US population, the mean hair Hg concentration among 1726 women of child-bearing age was 0.47 (95% CI = 0.35–0.58) µg/g [16
]. Fish consumption during the previous 30 days was directly related to hair Hg concentration. In a population-based cohort study of pregnant women in Massachusetts [17
], the mean hair mercury concentration was 0.45 µg/g; a majority of participants reported consuming more than two fish servings weekly, higher seafood consumption than reported in the current study. The concentrations reported here are also lower than those reported for women of childbearing age in other studies from Florida; i.e., 0.37 µg/g [18
] and 0.56 µg/g [19
]. In our earlier study of coastal Florida residents, the mean hair concentration of THg among women was 0.96 ± 0.74 µg/g [7
]. Twenty-five of 62 women (40.3%) had hair THg concentrations > 1.0 µg/g, which corresponds to an EPA reference dose for Hg in hair THg [14
]. In that study [7
], 56% of participants reported consuming seafood three times a week or more. In contrast, only 8.3% of women sampled in the current study had concentrations above 1.0 µg/g and 16.6% reported eating seafood with that frequency.
Fish consumption patterns and hair mercury concentrations may vary geographically depending on access to local sources of seafood and cultural habits. From an international perspective, the concentrations of THg in hair in this sample of pregnant Florida women were also lower than those reported from countries where seafood consumption is likely to be high. Pregnant women from coastal Italy, Greece, and Taiwan had mean concentrations of 1.06, 136, and 1.73 µg/g in hair, respectively [20
]. Thus, it appears that differences in patterns of seafood consumption are likely to explain differences in mean THg concentrations across populations, even in the same general areas.
Patterns of seafood consumption are also likely to explain higher concentrations of THg in hair among Asian women described above (Table 1
). Despite the small sample size, this finding is consistent with other reports on Hg in hair of women of childbearing age in multiple studies conducted in the United States [23
]. Fish consumption among Asian women is commonly reported as being higher than among other races in the United States, thus resulting in higher Hg concentrations [10
Socio-economic status, typically measured as level of achieved education or income, has been reported to be a significant risk factor for Hg exposure among women in previous studies [24
]. In the current study, level of educational achievement was directly related to hair mercury concentration in a step-wise manner. Women who had completed post-graduate education were 11 times more likely to be in the highest tertile of THg concentration. Level of educational achievement may be related to knowledge of the beneficial effects of fish consumption on cardiovascular health and other endpoints. An annual income of >$
75,000 was significantly associated with Hg concentrations in earlier studies in Florida, USA [18
]. Income is likely related to the affordability of seafood and relatively expensive fish species. Thus, both measures of socio-economic status appear to be related to total seafood consumption and exposure to Hg.
The relationship between seafood consumption and hair mercury concentration has been described in multiple studies. In the current study, the highest concentration of THg in hair was found among women who reported eating seafood three times a week (0.62 µg/g). This concentration was almost four times as high as that found in women who reported consuming no seafood during the previous three months. The consumption of fish and seafood by pregnant women is directly related to the Hg concentrations in the unborn fetus [29
]. A dose response relationship between seafood consumption before and during pregnancy, Hg concentration in maternal blood and Hg concentration in cord blood [30
], or newborn hair [29
], has been previously demonstrated.
We also examined smoking and alcohol use during pregnancy as indicators of health awareness. Participants who continued to smoke and drink alcohol had higher hair THg compared to those who stopped. However, neither behavior remained statistically significant after adjusting for confounding factors in multivariate analyses. These behaviors also did not influence the relationship between seafood consumption and hair Hg concentration in a previous study from Canada [30
]. In contrast, in a cohort study in Spain, smoking was associated with THg among pregnant women [31
]. It seems plausible that high risk behaviors, such as smoking, consuming alcohol, and consuming foods that pose a health risk during pregnancy, are related, but the relationships are complex and poorly understood.
An important component of this study was an examination of the relationships between knowledge of the risks of seafood consumption and the potential effects of mercury on the fetus, the behavioral practices of this sample of pregnant women, and the concentrations of THg in hair. The majority of participants (85%) reported being aware that Hg may be harmful to the unborn child and 89% were aware that fish can contain high levels of Hg. These data suggest that the educational efforts and advisories circulated by the Florida Department of Health [32
] and others are reaching their target audience. However, awareness has not always been effective in reducing hair Hg among women [33
]. In the current study, there was no significant difference in the THg in hair between women who reported being aware of the risks of Hg in fish compared to those who were not although the concentrations were lower in the former group. Similarly, although the concentrations among those who stopped consuming seafood during pregnancy (n
= 19) were nearly half of those who continued, the difference was also not statistically significant. These results are in contrast to previous research supporting the positive impact of dietary intervention among women with elevated hair Hg [33
It should be noted that the relationships between fish consumption during pregnancy, exposure to mercury, and markers of neuro-development, cognition, fine motor skills, and other neuro-behavioral outcomes are complex and have led to conflicting results in epidemiologic studies [34
]. Neurotoxic effects of mercury on the developing brain have been demonstrated in longitudinal studies of population groups with high levels of seafood consumption, such as those in the Faroe Islands where consumption of seafood consisting mainly of pilot whale containing high levels of methylmercury was associated with decrements in verbal skills, memory, and attention at seven years of age [35
] and at older ages [36
]. Conversely, associations between maternal seafood consumption and childhood cognition were not confirmed in studies of populations in the Seychelles [37
], Great Britain [40
], United States [41
], and elsewhere [42
The discrepancies may be explained by the fact that while seafood can be a source of fetal exposure to MeHg, it is also the major source of omega-3 fatty acids which play an essential role in neuro-development. Fish consumption during pregnancy may have a beneficial effect on markers of neuro-development and cognition as shown in several studies [43
]. Maternal fish consumption had a beneficial effect on language and communication skills in a study of children in Norway but a detrimental effect of MeHg in the highest exposure group was also noted [45
]. A threshold effect for maternal hair mercury concentration at 1 µg/g or greater was suggested for inattention, impulsivity, and hyperactive disorders at eight years of age in a cohort study in the USA; however, a protective effect for the same outcomes was associated with maternal consumption of more than two fish meals per week [17
]. Therefore, it is critical to balance the risks of mercury exposure with the benefits of seafood consumption during pregnancy and to enhance the understanding of the knowledge, attitudes, and practices of pregnant women concerning this subject.
Several limitations of this study deserve mention. Thirty percent of the samples were below the limit of detection of the assay, thus limiting statistical power. The number of participants in some groups, e.g., women who were unaware of the risks of Hg consumption, as well as the small number of Asian participants, limits the conclusions for these subpopulations. Non-differential misclassification of behaviors was likely to have occurred, likely contributing to the wide confidence intervals for some variables. Questions regarding knowledge of the risks of mercury exposure via seafood consumption were asked at the end of the interview and thus were unlikely to have created response bias. Importantly, we did not assess the level of knowledge regarding beneficial effects of seafood consumption during pregnancy, nor did we assess the proportion of women who took fish oil or other omega-3 supplements during pregnancy and could not assess the effects of these factors on hair mercury concentrations.