New England is one of three areas in the United States with the highest annual deposition of mercury, an established environmental pollutant with a variety of health effects. We measured the mercury content in toenails of 27 individuals in New Hampshire who participated as controls in a health study in 1994–95. The mean total toenail mercury concentration was 0.27 mcg/g (median 0.16; SD 0.27; range 0.04–1.15 mcg/g). The best predictor of toenail mercury levels was the mean combined fish and shellfish consumption measured using four simple questions from a validated food frequency questionnaire. Toenail total mercury content was significantly correlated with the mean average weekly consumption of finfish and shellfish (Spearman correlation coefficient 0.48, P=0.012). Multivariate models confirmed that toenail total mercury concentration was best predicted by total finfish and shellfish consumption.
Mercury is an established environmental pollutant with a variety of serious health effects in humans (Risher et al., 1999). Numerous attempts have been made to identify and characterize useful biomarkers of exposure to mercury, including blood, urine, hair, fingernails and toenails, both as correlates of environmental exposures and also in relation to diseases such as coronary heart disease (e.g., Sinclair et al., 1980; MacIntosh et al., 1997; Guallar et al., 2002; Mortada et al., 2002; Yoshizawa et al., 2002). Toenail mercury concentrations are reasonably stable indicators of exposure over time (Garland et al., 1993) and toenails have the advantage over hair and fingernails of being less susceptible to external contamination. Toenail mercury content is significantly associated with fish consumption; in fact, knowledge of fish consumption alone can be sufficient to characterize mercury exposure as measured in toenails (Garland et al., 1993; MacIntosh et al., 1997).
New England is one of three areas in the United States with the highest annual deposition of mercury (Rice et al., 1997). Since 1994, the New Hampshire Department of Health and Human Services has issued a fish consumption advisory because of mercury contamination (New Hampshire Department of Health and Human Services Bureau of Health Risk Assessment, 2001) and studies of wildlife have demonstrated high mercury levels in eight states in north America (Evers et al., 2003). To our knowledge, no published studies have described biomarkers of mercury exposure in New Hampshire's human population. The aims of this pilot study were to analyze mercury levels in toenails taken from individuals in New Hampshire who participated as controls in a previous health study, and to determine their correlation with three dietary measures of seafood intake.
We analyzed toenail clippings obtained from control subjects during a case–control study conducted in 1994–95 in New Hampshire, which was reported previously (Karagas et al., 1998; Karagas et al., 2001). The original study investigated the relationship between the occurrence of basal and squamous cell carcinomas identified through a population-based surveillance mechanism, and the concentration of arsenic in toenail clippings. The control group was chosen from New Hampshire residents, frequency matched on age and sex to the combined distribution of cases. Selection was made using lists obtained annually from the Department of Transportation for those aged 25–64 years, and the Health Care Financing Administration's Medicare Program for those aged 65–74 years. The study was approved by the Committee for the Protection of Human Subjects at Dartmouth College.
Enrollees were sent a food frequency questionnaire (Willett et al., 1985) that covered usual diet over the previous year including four seafood consumption questions (Appendix 1). Participants also completed a water and seafood consumption diary during the 3 days preceding a scheduled interview (Appendix 1). At interview, more detailed information was collected about the type and frequency of seafood consumed, including locally caught freshwater fish (Appendix 1). We also collected sociodemographic and lifestyle data and documented physical characteristics such as body weight and height. Because the original study focused on arsenic rather than mercury, we did not have information about dental amalgams or other possible exposures to mercury.
Participants were instructed to save toenail clippings from all 10 toes. These were either returned at interview or by mail and subsequently stored in sealed vials until testing in 2003. They were prepared and analyzed for mercury content at the Trace Element Analysis Core Facility at Dartmouth College. Toenail samples were washed with acetone, Triton X-100 and three times with DI-water and analyzed by Cold Vapor High Resolution Inductively Coupled Mass Spectrometry. This procedure has a detection limit of 0.10 ng/g (0.10 ppb). All sample preparations and analyses were carried out in a trace-metal clean HEPA-filtered-air environment. Each batch of analyses included a certified reference material quality control sample (DC 73347, Human hair, CNAC, Beijing, China). Our determinations (0.32 μg/g±0.03, n=6) were in good agreement with the certified content of 0.36±0.05 μg/g.
Dietary exposure to mercury through seafood consumption was estimated in three ways (Appendix 1). Each method was used to estimate the proportion of participants who ate fish at least once per week and the mean weekly fish consumption. To do so, we replaced fish consumption estimates of less than once per month with 0/week; 1–3/month with 0.5/week; 2–4/week with 3/week; and 5–6/week with 5.5/week.
Method 1. Participants kept a 3-day diary of water and seafood consumption, which was used to calculate average weekly consumption of finfish and shellfish.
Method 2. We used the semiquantitative food frequency questionnaire (Willett et al., 1985) to calculate average weekly consumption of tuna, dark fish, shellfish and other fish. The Nutrition Questionnaire Service Center affiliated with the Harvard School of Public Health processed these questionnaires to provide a detailed breakdown of dietary nutrients for each participant. This included an estimate for omega 3 intake, which was derived principally from information on fish consumption and dietary supplements (cod liver oil and omega 3 fish pills).
Method 3. We used detailed questionnaires to estimate the weekly consumption of each fish species. We then multiplied the amount of each finfish or shellfish species consumed by the corresponding published mean mercury level for that species (U.S. Department of Health and Human Services and U.S. Environmental Protection Agency, 2004; U.S. Environmental Protection Agency National Fish and Wildlife Contamination Program, 2004). We summed these totals to produce a “weighted” estimate of seafood consumption.
We performed statistical tests using Stata 9.0 (Stata Corporation, 4905 Lakeway Drive, College Station, Texas 77845, USA). Mercury measurements were not normally distributed and, after fitting multivariate linear regression models to identify predictors of mercury measurements, we also found that the residuals for the final model were not normally distributed. We log-transformed toenail mercury concentrations to approximately normal distributions and repeated the modeling process. We found no evidence that the residuals in the final model deviated significantly from a normal distribution. For comparison, we also performed the regression analyses using untransformed data. We used t-tests with a significance level of 0.05 to compare mercury concentrations in relation to age group, body mass index (BMI), years of education, and gender. We used Spearman correlation coefficients to describe the relation between mercury concentration and single variables.
Toenail mercury measurements and adequate food frequency data were available for 28 subjects. One participant's daily calorie intake estimated using Method 2 was <300 kCal; this participant's dietary estimations were assumed to be unreliable and the participant was dropped from further analyses. The remaining 27 participants consisted of 16 males and 11 females with a mean age of 59 years (median 63; SD 11.4; range 37–73); a mean BMI of 27.7 (median 26.8; SD 4.1; range 22.2–36.6); and a mean of 2.9 (median 2.0; SD 2.8; range 0–8) years of education beyond high school. The mean total toenail mercury concentration was 0.27 mcg/g (median 0.1561; SD 0.2720; range 0.0398–1.1522 mcg/g). Assuming an estimated portion size of 4 ounces (115 g), the median daily fish intakes were 25 g/day of finfish and 8 g/day of shellfish. A plot of mercury levels revealed marked skew, and the residuals in the regression model were not normally distributed. The dependent variable was log transformed before performing further regression analyses.
Of the 27 participants analyzed, 17 (63%) had not eaten any finfish within the past 3 days. The numbers of participants who reported eating any finfish at least once per week were 26/27 (96%) in the simpler, semiquantitative food frequency questionnaire (Method 2) and 18/27 (67%) in the detailed fish consumption interview (Method 3). The mean weekly finfish consumptions estimated by Methods 2 and 3 were 1.9 (SD 1.2, range 0.5–5) and 1.2 (SD 0.9, range 0–4) 4-ounce portions respectively. Only 4/27 (15%) participants reported having eaten fish from local sources within the preceding year.
In univariate analyses, significant correlations were identified between toenail mercury levels and various measures of fish consumption derived from the semiquantitative food frequency questionnaire (Method 2) (Table 1). No significant correlations were identified between fish consumption (Method 2) and years of education (r=0.13; P=0.53), BMI (r=0.07; P=0.72) or age (r=−0.05; P=0.82). Omega-3 fatty acid consumption was significantly correlated with toenail mercury concentration (r=0.46; P=0.016). The estimate of omega 3 intake was derived from the semiquantitative food frequency questionnaire, and consequently it was highly correlated with total fish consumption (Method 2) (r=0.74, P<0.0001) because fish constitutes the major source of omega 3 in the diet. Among the four fish consumption variables derived from Method 2, the only significant correlation was found between dark fish and shellfish consumption (r=0.39, P=0.46).
After log transforming the mercury measurements, we used multivariate regression analyses to identify predictors of toenail mercury concentrations. In the modeling process, we excluded omega 3 and included the fish consumption estimates described in Table 1. Regression analyses were performed first after excluding one individual for whom BMI was unknown. Because BMI did not significantly affect the model, analyses were repeated after including this individual (n=27). In the final model, log mercury concentrations were best predicted by the mean total consumption of finfish and shellfish (P=0.016) estimated by Method 2, with a coefficient of 0.361 (SE 0.139; 95% CI 0.074–0.648). Similar results were seen when we used the untransformed mercury concentration in regression models. Other variables such as age, gender, education, and estimates of fish consumption using Methods 1 and 3 (weighted or unweighted) did not contribute usefully to the model. These results indicate that, on average, we would expect to see a 36% increase in toenail mercury for every additional fish meal regularly consumed per week.
In this small pilot study, we confirmed previous observations that toenail mercury content and fish consumption are significantly correlated (MacIntosh et al., 1997). Having done so, we investigated the correlations between various measures of dietary fish intake and toenail total mercury content. The highest Spearman correlation coefficient that we identified was between mercury concentration and mean weekly combined finfish and shellfish consumption (r=0.48). Multivariate models confirmed that toenail mercury concentration was best predicted by total finfish and shellfish consumption.
The potential advantages of toenails as a long term biomarker of exposure to trace elements include representation of several weeks or months of exposure; relatively low susceptibility to external contamination; and convenience of collection (Garland et al., 1993). Reproducibility of toenail mercury levels over time and correlation of toenail mercury with dietary intake of fish have both been demonstrated (Garland et al., 1993). However, there is currently no gold standard against which to compare either toenail mercury concentrations or dietary assessments. Of the methods we used in this study, we found the best correlation between toenail mercury levels and the combined fish/shellfish consumption measured by a previously validated food frequency questionnaire (Method 2). However, the study tools on which Methods 1 and 3 were based were intended for the original study which was focused on diet in general rather than mercury intake specifically. It remains unclear whether the differences in the correlations we observed using each dietary assessment resulted from deficiencies in the dietary assessment method, the biomarker, or both. As our more detailed but unvalidated assessments of dietary fish intake (Methods 1 and 3) failed to correlate significantly with toenail mercury, we can only conclude that Method 2, which has previously been used successfully in relation to mercury, performed best of those we assessed.
Long term exposure to fish is thought to account for almost all methyl mercury measured in the population (World Health Organization, 1990), but exposure to sources other than fish can contribute to the body's total mercury load. Blood studies have shown that the average total mercury blood level in the population is ∼1–8 mg/l, but a background level of 2 mg/l can be found among individuals who do not consume fish (Gerhardsson & Brune 1989; World Health Organization, 1990). Mercury in nails (Table 2) has been shown to reflect both methyl and elemental mercury exposure. For example, in one study, inorganic mercury comprised 9.9% of the total mercury measured in toenails among volunteers without occupational exposure to mercury. In contrast, toenail mercury levels among dentists exposed to elemental mercury in amalgam were twice as high as levels among non-dentist controls; both occupation and fish consumption contributed significantly to a multivariate regression analysis to identify the predictors of toenail mercury in that study (Joshi et al., 2003). Because our study participants were not occupationally exposed, we would expect a small proportion of their total toenail mercury to result from inorganic sources such as dental amalgam, perhaps at the level of around 10% described by Suzuki et al. (1989). Exposure to mercury through dental amalgam was not measured in this study and could account for some non-dietary variability in the toenail mercury levels (Table 3).
Our attempts to estimate dietary mercury exposures more accurately using detailed seafood consumption questionnaires (Methods 1 and 3) were unsuccessful, as judged by their correlation with toenail mercury. Because we observed a significant correlation between toenail mercury and dietary fish intake measured by Method 2, we suspect deficiencies in our two unvalidated methods as the reason for their lack of correlation with toenail mercury. This may have been owing to the relatively infrequent consumption of each fish species. For example, there was a substantial discrepancy between the proportions who reported eating fish at least once per week by Method 2 (96%) and Method 3 (67%). The main difference between these methods is that Method 2 uses broad categories of fish, whereas Method 3 focuses on individual species. It seems likely that the more detailed questionnaire is less accurate because it categorizes and sums very infrequent events. A participant who eats fish once per week completes Method 2 easily. Closer questioning about individual fish species (Method 3) may pose more difficulties in recall, and the errors associated with each guess would be summed to give an answer that will likely deviate from once per week. Thus the detailed method (Method 3) may be more susceptible to misclassification than the simpler method (Method 2). Furthermore, we would expect four of every seven individuals who usually eat fish once per week to be misclassified as non-consumers by a 3-day diary (Method 1). The response category “never, or less than once per month” might be better divided for future studies so that individuals who never eat fish are clearly identified. Consumers of fish known to be high in mercury may follow published guidelines and consume them less frequently than once per month; in the questionnaires used, these occasional high mercury exposures would be classified along with zero consumption, which would account for substantial misclassification in our study which is already limited by its small sample size.
Our results suggest that the short-term dietary assessments were of variable use in predicting the toenail biomarker levels, and the best correlation resulted from the semiquantitative food frequency questionnaire. However, both the dietary and toenail estimates reflecting mercury exposure are each subject to errors, and there is no gold standard measure of mercury exposure. Our fish intake assessment tools addressed usual (Methods 2 and 3) or recent (Method 1) patterns of fish consumption, which may not precisely reflect the growth interval of the toenail clippings. The clippings from all 10 toes result from several weeks of nail growth that represents exposure over the previous 3–12 months or more (Longnecker et al., 1993). Therefore we view our dietary assessments and exposure biomarker (i.e. toenails) as a means of estimating usual fish intake rather than intake during a specified exposure period. Variability in fish consumption over time is also possible, and in future studies it would be useful to determine whether participants had made significant changes to their fish consumption over time. Our small sample proved inadequate to shed light on the possible relation between consumption of locally caught fish and toenail mercury levels.
The levels of mercury found in our study were comparable to a similarly designed, recent study using “control” participants in eight European studies (Guallar et al., 2002) but lower than many early other studies (Suzuki et al., 1989; Garland et al., 1993; MacIntosh et al., 1997; Mortada et al., 2002; Yoshizawa, 2002; Joshi et al., 2003, p. 4). Temporal trends are one possible explanation; however, a larger study would be necessary to identify such trends in mercury levels. In one of these studies, median fish intake reported for each quintile of toenail mercury concentration ranged from 21 to 51 g/day (Yoshizawa et al., 2002), compared with overall medians in our study of 25 g/day for fin fish and an additional 8 g/day of shellfish. The mean mercury toenail concentration in that study was 0.45 mcg/g among participants without occupational exposure, compared with 0.27 mcg/g in ours. Unfortunately, for most of the studies of toenail mercury listed in Table 2, fish consumption data are not reported in a format that can be compared with our own. According to the United States Environmental Protection Agency, the mean per capita consumption of uncooked finfish and shellfish by adults, including non-consumers, is 19.9g/day (uncooked) (United States Environmental Protection Agency, 2002).
In conclusion, we report the use of toenail mercury measurements as a biomarker for mercury exposure through dietary fish consumption in a New Hampshire study. Our findings suggest that, in their current form, our two investigational methods of assessing dietary exposure to mercury (a 3-day diary and detailed food frequency questionnaire) did not provide additional benefit in characterizing the toenail mercury biomarker as an indicator of mercury exposure through fish consumption. Rather, toenail mercury levels were best described by the combined fish and shellfish consumption measured using four simple questions from a validated standardized food frequency questionnaire.
Evers D.C., Taylor K.M., Major A., Taylor R.J., Poppenga R.H., and Scheuhammer A.M. Common loon eggs as indicators of methylmercury availability in North America. Ecotoxicology 2003: 12 (1–4): 69–81.
Garland M., Morris J.S., Rosner B.A., Stampfer M.J., Spate V.L., Baskett C.J., Willett W.C., and Hunter D.J. Toenail trace element levels as biomarkers: reproducibility over a 6-year period. Cancer Epidemiol Biomarkers Prev 1993: 2 (5): 493–497.
Gerhardsson L., and Brune D. Mercury in dentistry. In: Brune DK, and Edling C. (Eds.). Occupational Hazards in the Health Profession. CRC Press Inc: Boca Raton, FL, 1989 pp 307–321.
Guallar E., Sanz-Gallardo M.I., van't Veer P., Bode P., Aro A., Gomez-Aracena J., Kark J.D., Riemersma R.A., Martin-Moreno J.M., and Kok F.J. Mercury, fish oils, and the risk of myocardial infarction. N Engl J Med 2002: 347 (22): 1747–1754.
Joshi A., Douglass C.W., Kim H.D., Joshipura K.J., Park M.C., Rimm E.B., Carino M.J., Garcia R.I., Morris J.S., and Willett W.C. The relationship between amalgam restorations and mercury levels in male dentists and nondental health professionals. J Public Health Dent 2003: 63 (1): 52–60.
Karagas M.R., Stukel T.A., Morris J.S., Tosteson T.D., Weiss J.E., Spencer S.K., and Greenberg E.R. Skin cancer risk in relation to toenail arsenic concentrations in a US population-based case-control study. Am J Epidemiol 2001: 153 (6): 559–565.
Karagas M.R., Tosteson T.D., Blum J., Morris J.S., Baron J.A., and Klaue B. Design of an epidemiologic study of drinking water arsenic exposure and skin and bladder cancer risk in a U.S. Population Environ Health Perspect 1998: 106 (Suppl 4): 1047–1050.
Longnecker M.P., Stampfer M.J., Morris J.S., Spate V., Baskett C., Mason M., and Willett W.C. A 1-y trial of the effect of high-selenium bread on selenium concentrations in blood and toenails. Am J Clin Nutr 1993: 57: 408–413.
MacIntosh D.L., Williams P.L., Hunter D.J., Sampson L.A., Morris S.C., Willett W.C., and Rimm E.B. Evaluation of a food frequency questionnaire-food composition approach for estimating dietary intake of inorganic arsenic and methylmercury. Cancer Epidemiol Biomarkers Prev 1997: 6 (12): 1043–1050.
Mortada W.I., Sobh M.A., el-Defrawy M.M., and Farahat S.E. Reference intervals of cadmium, lead, and mercury in blood, urine, hair, and nails among residents in Mansoura city, Nile delta, Egypt. Environ Res 2002: 90 (2): 104–110.
New Hampshire Department of Health Human Risk Assessment. Services Bureau of Health Mercury in New Hampshire's Fish, Vol. 2003. Department of Health and Human Risk Assessment: New Hampshire, 2001.
Rice G., Amrose R., Bullock O., and Swartout J. Mercury Study Report to Congress. Volume III: Fate and Transport of Mercury in the Environment. Environmental Protection Agency: United States, 1997 ES-4.
Risher J.F., De Rosa C.T., Jones D.E., and Murray H.E. Summary report for the expert panel review of the toxicological profile for mercury. Toxicol Ind Health 1999: 15 (5): 483–516.
Sinclair P.M., Turner P.R., and Johns R.B. Mercury levels in dental students and faculty measured by neutron activation analysis. J Prosthet Dent 1980: 43 (5): 581–585.
StataCorp. Stata Statistical Software, Release 7.0 ed. Stata Corporation: College Station, TX, 2001.
Suzuki T., Watanabe S., and Matsuo N. Comparison of hair with nail as index media for biological monitoring of mercury. Sangyo Igaku 1989: 31 (4): 235–238.
U.S. Department of Health and Human Services. Mercury Levels in Commercial Fish and Shellfish., Vol. 2004. US: Department of Health and Human Services and U.S. Environmental Protection Agency, 2004.
United States Environmental Protection Agency. Estimated Per Capita Fish Consumption in The United States. Environmental Protection Agency: United States, 2002 4–18, 4–20, 5–18, 5–20.
U.S. Environmental Protection Agency National Fish and Wildlife Contamination Program. Average Tissue Mercury Concentrations in Noncommercial Fish, Vol. 2004. Environmental Protection Agency National Fish and Wildlife Contamination Program: US, 2004.
Willett W.C., Sampson L., Stampfer M.J., Rosner B., Bain C., Witschi J., Hennekens C.H., and Speizer F.E. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985: 122 (1): 51–65.
WHO. Methyl Mercury, Vol. 101 World Health Organization, International Program on Chemical Safety: Geneva, Switzerland, 1990.
Yoshizawa K., Rimm E.B., Morris J.S., Spate V.L., Hsieh C., Spiegelman D., Stampfer M.J., and Willett W.C. Mercury and the risk of coronary heart disease in men. N Engl J Med 2002: 347 (22): 1755–1760.
This work was supported by NIH Grant P42ES07373 and the Center for Environmental Health Sciences at Dartmouth.
Appendix 1. Questionnaires used in the study
Appendix 1. Questionnaires used in the study
Method 1. Water and seafood intake record
INSTRUCTIONS: Please record your water and seafood intake for the 3 days before your interview.
Method 2. Fish consumption questions from the semiquantitative food frequency questionnaire (Willett et al., 1985)
The fish consumption questions are included within a large comprehensive dietary assessment. The possible responses are: Never, or less than once per month; 1 per mo.; 1 per week; 2–4 per week; 5–6 per week; 1 per day; 2–3 per day; 4–5 per day; 6+ per day.
Please fill in your average use during the last year, of each specified food. Please try to average your seasonal use of foods over the entire year.
Canned tuna fish (3–4 oz.)
Dark meat fish, for example, mackerel, salmon, sardines, bluefish, swordfish (3–5 oz)
Shrimp, lobster, scallops as a main dish
Method 3. Detailed fish consumption questionnaire. Have you eaten fish from a local pond, lake or river in the last year?
If Subject reports eating “dark meat fish, for example, mackerel, salmon, etc.” or “other fish” one or more times per month: What types of fish do you eat? How often?
If Subject reports eating “shrimp, lobster, scallops, etc. as a main dish” one or more times per month: What types of shellfish do you eat? How often?
About this article
Prenatal exposure to mercury in relation to infant infections and respiratory symptoms in the New Hampshire Birth Cohort Study
Environmental Research (2019)
Using hair and fingernails in binary logistic regression for bio-monitoring of heavy metals/metalloid in groundwater in intensively agricultural areas, Thailand
Environmental Research (2018)
Muscle & Nerve (2018)
Assessment of saliva, hair and toenails as biomarkers of low level exposure to manganese from drinking water in children
Associations between Diet and Toenail Arsenic Concentration among Pregnant Women in Bangladesh: A Prospective Study