4

DOSE ESTIMATION

IN assessing the risks of exposure to MeHg, quantitative exposure assessments are required to derive dose-response relationships from epidemiological data. A quantitative exposure assessment also allows risk assessment of an exposed population by comparing actual exposures to a reference dose (or similar benchmark) derived from critical studies. In contrast to experimental animal studies, in which the dose can be closely controlled, the dose in population-based epidemiological studies is not controlled and is therefore viewed as a random variable distributed across the study population. Three metrics for retrospective dose estimation and reconstruction are available for MeHg: dietary assessment, hair analysis, and blood analysis. Each metric has advantages and disadvantages. Ponce et al. (1998) proposed an approach for examining the relative uncertainties of those metrics.

DIETARY ASSESSMENT

With the exception of intakes through breast milk, which is less well characterized, exposure to MeHg occurs almost entirely from a single dietary category — fish (IPCS 1990, 1991). For that reason, the task of assessing dietary intake or assessing ongoing intake in populations with uncontrolled exposures is relatively straight forward compared to assessment of multiple types of food. There are several basic approaches to the estimation of MeHg exposure from dietary intake: collection of



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 105
Toxicological Effects of Methylmercury 4 DOSE ESTIMATION IN assessing the risks of exposure to MeHg, quantitative exposure assessments are required to derive dose-response relationships from epidemiological data. A quantitative exposure assessment also allows risk assessment of an exposed population by comparing actual exposures to a reference dose (or similar benchmark) derived from critical studies. In contrast to experimental animal studies, in which the dose can be closely controlled, the dose in population-based epidemiological studies is not controlled and is therefore viewed as a random variable distributed across the study population. Three metrics for retrospective dose estimation and reconstruction are available for MeHg: dietary assessment, hair analysis, and blood analysis. Each metric has advantages and disadvantages. Ponce et al. (1998) proposed an approach for examining the relative uncertainties of those metrics. DIETARY ASSESSMENT With the exception of intakes through breast milk, which is less well characterized, exposure to MeHg occurs almost entirely from a single dietary category — fish (IPCS 1990, 1991). For that reason, the task of assessing dietary intake or assessing ongoing intake in populations with uncontrolled exposures is relatively straight forward compared to assessment of multiple types of food. There are several basic approaches to the estimation of MeHg exposure from dietary intake: collection of

OCR for page 105
Toxicological Effects of Methylmercury duplicate portions of foods consumed; food-consumption diaries, in which daily fish intakes are recorded quantitatively; recall methods, such as 24-hr recall of fish consumption; diet histories of usual consumption at various meals; and food-frequency measures of usual frequency of consumption of fish and shellfish. Duplicate-diet collections and food-consumption diaries are prospective approaches, and the others are retrospective approaches. General considerations for duplicate-diet studies were recently discussed by Berry (1997) and Thomas et al. (1997). In duplicate-diet studies, participants collect an identical portion of the food they consume and provide it to the investigator for laboratory analyses. In theory, duplicate-diet studies have the potential to provide the most accurate information on the ingested dose of MeHg, because the mass of fish and other nutrients and contaminants, in addition to MeHg, can be measured directly. The fact that only the fish portion of any given meal will contain MeHg simplifies the burden of duplicate-diet collection. In practice, however, this approach is limited by the demands it makes on the participants, the difficulty in identifying individuals who are willing to carry out such a study, the influence exerted by investigator observation, and the potential change in diet resulting in response to the burden of food collection. Thomas et al. (1997), working with nine highly motivated households, was able to collect duplicate samples for 97% of meals and 94% of snacks over a 7-day period. The number of uncollected meals, however, tripled after the first 3 days, and participants strongly recommended that future studies be limited to a maximum of 3-4 days. When such studies are confined to fish consumption, 3-4 days of collection might be useful only for populations with very frequent and highly regular patterns of fish consumption. Because of the practical limits on the length of the collection period, the authors recommended that duplicate-diet studies for risk-assessment purposes should be done over multiple intervals of time. Moreover, when the calorie content of collected food was compared with the estimated energy requirements of participants, duplicate portions were found to be underestimated. Duplicate-diet studies have been specifically applied to the estimation of MeHg exposure by Sherlock et al. (1982) and Haxton et al. (1979). Sherlock et al. (1982) carried out a 1-week duplicate-diet study with 98 participants selected on the basis of frequent fish consumption. In

OCR for page 105
Toxicological Effects of Methylmercury addition, a 1-month dietary diary was kept by the participants; the last week of the diary corresponded to the duplicate-diet collection. No indication is provided of the completeness of the duplicate-diet collection, but the weight of fish calculated from the diary during the week of duplicate-diet collection corresponded closely to the weight of fish measured from the duplicate samples. The authors noted however, indirect evidence of undercollection of duplicate-diet portions relative to consumed portions. It should also be noted that the preselection of subjects with frequent fish consumption increased the likelihood of collecting a meaningful number of samples over a 1-week period. A similar study with a randomly selected study sample would be less likely to provide adequate representation of infrequent consumers. Haxton et al. (1979) conducted a 1-week duplicate-diet study with 174 subjects selected from fishermen and their families in coastal communities to obtain a population with high fish-consumption rates. No simultaneous diaries were kept, but the characteristic intake for each individual was identified from pre-collection interviews. No estimate of the completeness of the duplicate-diet collection was provided. However, the authors noted that the measured weight of weekly fish intake from the duplicate-diet samples was lower than that calculated from the interviews, and all measured intakes were below the calculated mean intake. The authors suggested that the discrepancy resulted from misidentification of characteristic intake in the interviews rather than from undercollection of dietary samples. No data are provided to support that assertion. As with the Sherlock et al. (1982) study, the preselection of subjects with frequent fish consumption made the relatively short collection period feasible. Multiple-day food records (food-consumption diaries) are often used in conjunction with duplicate-diet studies ( Sherlock et al. 1982, Thomas et al. 1997). This method, if conducted appropriately, has the advantage of recording information prospectively with little reliance on recall. It also requires less effort from participants than the duplicate-diet approach. However, daily recording of foods eaten at each meal requires a continuous and significant time commitment. Because fish are consumed relatively infrequently, the duration of the recording period might require many weeks to adequately capture infrequent consumers as well as variability in consumption among more frequent consumers. Furthermore, the design must be such that possible seasonal patterns of

OCR for page 105
Toxicological Effects of Methylmercury consumption are observable. The determination of the mass of food consumed when using food-consumption diaries can be made by weighing samples or by participants' estimating portion size. The former is preferable but more invasive, especially when foods are consumed away from home. Participants' estimation of portion size introduces a degree of measurement error not seen with duplicate-diet methods. Furthermore, if the diary approach is used without duplicate-diet collection, analysis of Hg concentration in each fish meal consumed cannot be made directly but must be based on the characteristic Hg concentration in each reported species. Such studies must, therefore, rely on participants for correct identification of species. Incorrect species identification can lead to errors in estimation of MeHg intake. Consumers, as well as the markets from which they purchase fish, might not know or correctly identify the species that was bought and consumed. The data from the Continuing Survey of Food Intake by Individuals (CSFII) generated by the U.S. Department of Agriculture from 1989 to 1995 rely on self-administered food consumption diaries for the second and third days of its 3 days of reporting (discussed in EPA 1997). The CSFII data have been used by the U.S. EPA to estimate fish consumption in the U.S. population (Jacobs et al. 1998) and to estimate MeHg intake (EPA 1997). The National Purchase Diary conducted by the Market Research Corporation used dietary diaries over 1-month periods between 1973 and 1974 (discussed in EPA 1997). The fish-consumption portions of these diary data were used to estimate MeHg exposure in the U.S. population (Stern 1993, EPA 1997). Retrospective dietary-assessment methods are simpler and less expensive than prospective and duplicate-diet methods, and therefore are used more often as the basis of dietary exposure assessments. Food-frequency studies take the form of participants identifying their typical fish consumption (e.g., “How many times per week/month do you usually eat fish A?”). Diet histories involve recollection of specific meals over a specific time (e.g. 24-hr or 1-week periods). In the studies mentioned above, Sherlock et al. (1982) and Haxton et al. (1979) used retrospective assessment of typical consumption to preselect subjects. In a recent study of MeHg exposure among pregnant women in New Jersey (Stern et al. 2000), participants were asked to identify their typical consumption frequency and typical portion size of 17 species of fish and

OCR for page 105
Toxicological Effects of Methylmercury fish dishes (e.g., fish sticks). MeHg intake was estimated as the product of the characteristic MeHg concentration for each fish species, the self-reported yearly frequency of consumption, and the self-reported average portion size. The yearly MeHg intake estimated in that manner was poorly correlated with the Hg concentration in hair from the same individuals. The authors attributed the discrepancy to the relatively infrequent consumption of fish in general. Therefore, the hair segments might have been too short to provide an adequate sample of the yearly intake. Uncertainty in the reporting of characteristic consumption frequency and portion size was also suspected as a contributing factor to the poor correlation. The usefulness of studies using dietary recollection over a specific period depends on the participants' ability and willingness to recall information about fish meals over the target period. Recall of fish consumption seems to be much better than recall of other dietary items or of food intake in general. However, short-term recall methods of dietary assessment will tend to underrepresent the consumption characteristics of infrequent consumers (Whipple et al. 1996; Stern et al. 1996). If the species of fish (and thus the characteristic Hg concentration in the fish) consumed by frequent and infrequent consumers differ, or if the average portion size consumed by each group differs, the estimate of MeHg intake in the overall population will not be accurate. The CSFII data used by EPA to estimate fish consumption (Jacobs et al. 1998) and MeHg exposure (EPA 1997) nationally are, as noted above, based on 1 day of recall and 2 succeeding days of diary entries. The National Health and Nutrition Examination Surveys (NHANES III) dietary data, generated from 1-day recall, were also used by the EPA to generate estimates of MeHg in the U.S. population (EPA 1997). Stern et al. (1996) used data from a fish-consumption-specific telephone survey of New Jersey residents. The survey elicited a 7-day recall. Relatively short-term recall studies can miss long-term patterns of variability in consumption and might not adequately capture consumption patterns of infrequent consumers. To address those issues, information on respondents' usual frequency of fish consumption was also elicited. That information allowed identification of infrequent consumers of fish in the sample. The information was used to investigate reweighting of the data to estimate the distribution of consumption frequency represented

OCR for page 105
Toxicological Effects of Methylmercury in a hypothetical 1-year recall study. Interestingly, the reweighting of the data using several different approaches resulted in only minor differences in estimates of fish consumption and MeHg exposure. Retrospective dietary data and diary data on fish consumption have frequently been used to stratify a study population into broad classes of MeHg intake before more quantitative estimation of exposure by measurement of Hg in biomarkers. Such data have also been used to provide a rough validation of biomarker analyses (e.g., Dennis and Fehr 1975; Skerfving 1991; Grandjean et al. 1992; Holsbeek et al. 1996; Vural and Ünlü 1996; Mahaffey and Mergler 1998). Less frequently, retrospective and diary data on fish consumption are used directly in quantitative estimations of MeHg exposure (Buzina et al. 1995; Stern et al. 1996; Chan et al. 1997). Such estimates, however, generally require species-specific Hg concentrations (microgram of Hg per gram of fish), which are combined with the reported consumption frequency (grams of fish per day), to yield an Hg intake rate (micrograms Hg per day). The assignment of species-specific concentration data is a potential source of error in such studies for several reasons. First, the identity of the species on the part of the retailer or the consumer is often ambiguous. Second, Hg concentrations characteristic of a given species in local and regional markets or waters might differ from the characteristic concentrations identified on the basis of nationwide sampling. Finally, characteristic Hg concentrations derived from data that are often decades old might not be valid today. In the United States, data on Hg concentration in commercial fish are largely available from two sources: (1) the National Marine Fisheries Service (NMFS) study, which sampled fish that were intended for human consumption and which were landed in the United States in the early to mid-1970s (Hall et al. 1978); and (2) the U.S. Food and Drug Administration (FDA) sampling conducted in the early 1990s (FDA 1992). Both data bases represent samples of fish collected from landings and markets in various parts of the United States but do not identify the locations at which samples were obtained or sold. The NMFS data were collected more systematically, represent more species, and generally contain considerably more samples for each species than the FDA data. However, the FDA data are about 20 years more recent than the NMFS data. Analysis of species represented in both data bases by at least three samples (n=15) indicates that, in almost all cases, the Hg concentration

OCR for page 105
Toxicological Effects of Methylmercury reported by FDA in a particular species is significantly lower than the concentration reported by NMFS. The most likely explanation for the discrepancy is the decreased availability of large fish due to overfishing (Stern et al. 1996). Those data cannot be used, therefore, to reflect potentially important local or regional differences in the characteristic concentrations of Hg by species. Studies addressing smaller populations with fewer varieties of fish (e.g., Buzina et al. 1995) can generate population-specific estimates of Hg concentrations by species. Given the variability in concentration within species, the assignment of a single representative value of Hg concentration is another potential source of error in such studies. The NMFS data base provides no estimate of such variability in U.S. commercial supplies. The FDA data base provides concentration ranges as well as average concentrations by species, but the ability to assess intraspecies variability in Hg concentration is limited by the generally small sample sizes. A study of Hg concentration in canned tuna (Yess 1993) indicates coefficients of variation of 55-120% across the several types of tuna commonly sold in cans. In general, sparse data on fish from commercial sources in the United States (FDA 1992) and data on food fish from noncommercial sources (e.g., Schuhmacher et al. 1994; Castilhos et al. 1998) often show a 2-3-fold difference between the mean and the maximum concentrations of Hg. Interspecies variability can be considerably larger. When data are available on intraspecies variability in MeHg concentrations, the variability can be integrated into estimates of intake through Monte Carlo probabilistic analysis (Chan et al. 1997). Intraspecies variability in Hg concentration might be less of a source of error in studies of frequent consumers of that species. With repeated consumption of a species of fish, total MeHg intake by consumers will approach the average concentration in that species. However, for populations with infrequent or sporadic consumption of a species, the effect of ignoring intraspecies variability in Hg concentrations could be significant. BIOMARKERS OF EXPOSURE MeHg lends itself to assessment of exposure through direct measurement in blood and hair. Assessment of Hg exposure through analysis of nail clippings has also been done (Pallotti et al. 1979; MacIntosh et al.

OCR for page 105
Toxicological Effects of Methylmercury 1997), but its correlation with fish consumption has yet to be clearly established. Hg exposure through breast milk has also been investigated (Pitkin et al. 1976; Fujita and Takabatake 1977; Skerfving 1988; Grandjean et al. 1995; Oskarsson et al. 1996). Compared with whole blood, breast milk (which is derived from maternal plasma) contains a much higher proportion of inorganic Hg (Skerfving 1988; Oskarsson et al. 1996). Therefore conclusions regarding the exposure of infants to MeHg from breast milk should use MeHg-specific analysis. Finally, there have been no reports of measurement of Hg in the fetal brain, the ultimate target of MeHg developmental neurotoxicity, although Cernichiari et al. (1995a) reported on a small set of measurements of Hg in infant brain. The relationship among the several possible indicators of exposure is shown in Figure 4-1. Some of the indicators, such as the biomarkers of hair and blood Hg concentration, are commonly measured directly, whereas others, particularly fetal brain Hg concentration, are assumed to be correlated with the directly measured quantities. For the purposes of risk assessment, biomarker concentrations of MeHg serve two functions. First, a biomarker concentration is used as a surrogate for the unknown biologically relevant dose of MeHg in the developing fetal brain. That permits the development of a “dose”-response relationship in which the dose is represented by the biomarker concentration. Second, once such a dose-response relationship has been established, the biomarker concentration identified as the critical (e.g., benchmark) concentration must be translated into an estimate of the ingested dose. At that point, public-health interventions and regulatory measures can be guided by that estimate. The translation of the biomarker concentration to the ingested dose involves the use of toxicokinetic modeling to recapitulate the steps that precede the measured biomarker compartment in Figure 4-1 (see Chapter 3). Methylmercury in Blood The detection limit for total Hg in blood is generally in the range of 0.1 to 0.3 µg/L (ppb) (Grandjean et al. 1992; Girard and Dumont 1995; Oskarsson et al. 1996; Mahaffey and Mergler 1998). The mean concentration reported in U.S. studies in which high-fish-consuming populations were not specifically selected appears to be in the range of 1 to 5

OCR for page 105
Toxicological Effects of Methylmercury FIGURE 4-1Relationship among the various indicators of MeHg exposure. Maternal ingestion of MeHg refers to the ingested dose, the magnitude of which depends on the amount of fish consumed and the concentration of MeHg in the fish. The concentration of Hg measured in the maternal blood, fetal blood, cord blood, maternal nails, and maternal hair are all biomarkers of exposure. Concentrations of Hg in the fetal brain, if available, would be considered the effective dose. (Note: The Fetal Brain box is shown with a dotted line because no direct data are available on Hg concentrations in the fetal brain.) µg/L (Humphrey 1975; Brune et al. 1991; Nixon et al. 1996; EPA 1997; Kingman et al. 1998; Stern et al. 2000). Thus, current methods for blood MeHg determination appear to be adequate for fully characterizing population distributions of MeHg exposure and are, in practice, limited only by the volume of blood that can be obtained. Fish and other seafood, including marine mammals, are the only significant source of MeHg exposure (IPCS 1990). Therefore, the blood Hg concentrations in populations with little or no fish consumption should reflect exposure to inorganic Hg. The mean blood Hg concentration in such populations was reported to be about 2 µg/L (standard deviation (SD) = 1.8 µg/L) (Brune et al. 1991). Blood Hg concentrations in populations with high fish consumption are usually considerably higher than that value. For example, median cord-blood concentration in a cohort with high fish consumption in the Faroe Islands was 24 µg/L (Grandjean et al. 1992). Therefore, the measurement of the concentration of total Hg in blood is generally a good surrogate for the concentration of MeHg in blood in populations with high fish consumption. In populations with relatively low fish consumption, inorganic Hg concentration might constitute a larger fraction of total Hg concentrations. Therefore, for such popula-

OCR for page 105
Toxicological Effects of Methylmercury tions, estimates of MeHg exposure from cord blood might be unreliable. Adult blood Hg concentration has frequently been used as a biomarker of adult MeHg exposure, and it has been used to assess dose-response relationships in adult neurotoxicity (e.g., Hecker et al. 1974; Dennis and Fehr 1975; Gowdy et al. 1977; Palotti et al. 1979; Skerfving 1991; Mahaffey and Mergler 1998). Cord-blood or maternal-blood Hg concentrations have also been used with some frequency in assessing exposure to the developing fetus (Dennis and Fehr 1975; Pitkin et al. 1976; Fujita and Takabatake 1977; Kuhnert et al. 1981; Kuntz et al. 1982; Sikorski et al. 1989; Grandjean et al. 1992; Girard and Dumont 1995; Oskarsson et al. 1996). In assessing the appropriateness of a particular biomarker of exposure, it is important to consider three factors: (1) how well the biomarker of exposure (i.e., the concentration of Hg in hair or blood) correlates with the ingested dose of MeHg; (2) how well the biomarker of exposure correlates with the Hg concentration in the target tissue; and (3) how well the variability over time in the biomarker of exposure correlates with changes in the effective dose at the target tissue over time. For developmental neurotoxicity, the target organ is the developing fetal brain. The kinetics of MeHg transport among compartments is subject to interindividual variability at each step, and therefore, the more closely a compartment is kinetically related to the target tissue, the more closely the concentration measured in that compartment is likely to correlate with the concentration in the target tissue. As shown in Figure 4-1, the fetal and cord-blood compartment is one compartment removed from the fetal-brain compartment. Thus, the cord-blood Hg concentration might be a reasonable surrogate for the biologically relevant dose to the fetal brain. Having determined a critical concentration of Hg in the blood, it is then necessary to back-calculate the ingested dose (micrograms of Hg per kilogram of body weight per day) corresponding to the critical concentration in blood (Stern 1997). Just as the kinetic proximity of the biomarker compartment to the target tissue increases the correlation between biomarker concentration and dose to the target tissue, the kinetic closeness of the biomarker compartment to the ingested dose will increase the correlation between the critical biomarker concentration and the estimated intake. The cord-blood Hg concentration is more closely linked to the fetal-brain compartment than

OCR for page 105
Toxicological Effects of Methylmercury to the ingested dose. Maternal-blood Hg concentration is more closely linked to the ingested dose than to the fetal-brain compartment. Thus, with the use of blood as a biomarker of MeHg exposure, there is a trade-off between the precision in the derivation of the dose-response relationship and the precision in the estimate of the corresponding ingested dose. The mean half-life of total MeHg in blood in humans is about 50 days (Stern 1997; EPA 1997), but much longer half-lives (more than 100 days) are observed. Blood Hg concentration, therefore, reflects relatively short-term exposures relative to the total period of gestation. However, the Hg blood concentration at any given time reflects both the decreasing concentration from earlier exposures and the increase in concentration from recent exposures. Individuals with frequent and regular patterns of fish consumption achieve, or approximate, steady-state blood Hg concentrations (IPCS 1990). At steady state, the daily removal of Hg from the blood equals the daily addition to the blood from intake. Under such conditions, an individual's blood Hg concentration at any given time provides a good approximation of the mean blood Hg concentration over time. For individuals with infrequent or irregular fish consumption, however, recent fish consumption will result in peaks in blood Hg concentration. A single blood sample showing an elevated concentration, without additional exposure information, does not provide a temporal perspective and does not permit differentiation between increasing peak concentrations, decreasing peak concentrations, and steady-state exposure. Conversely, a single blood sample obtained between peak exposures and showing a low blood Hg concentration provides no evidence of peak exposures. That result can introduce error into dose-response and risk assessment in adult populations in whom short-term peak exposures might be relevant to chronic toxicity. The blood Hg concentration can correlate well with the dose presented to the brain at the time of sampling, but such information cannot necessarily be extrapolated to dose at the target tissue at other times. A blood Hg measurement that might be adequate to reflect exposure over time can be determined to some extent by obtaining dietary intake data that corresponds to several half-lives preceding the Hg measurement. In assessing exposure and dose-response relationship in utero, the temporal considerations associated with the use of blood Hg concentration as a biomarker are further complicated by two additional factors: (1) the

OCR for page 105
Toxicological Effects of Methylmercury better predict deficits in fine-motor function. The authors attributed those qualitative differences to the different periods of development reflected by each of the measurements. That conclusion is consistent with the idea that discrete windows of vulnerability in the developmental toxicity of MeHg are differentially represented by hair- and cord-blood Hg measurements. However, the lack of uniformity in the lengths of the hair segments analyzed in the Faroe Islands study (Grandjean et al. 1999) make a clear interpretation of such differences somewhat problematic. Therefore, the uncertainties and limitations in the various biomarkers that are used for MeHg exposure assessment could result in exposure misclassification in the dose-response assessment. Misclassification of exposure in these studies could take several forms. Those include incorrectly considering exposures that occurred during developmental periods during which there is little or no vulnerability of the observed developmental endpoints to MeHg; failing to identify peak concentrations that might be more toxicologically relevant than the measured average concentrations; and using portions of hair with Hg concentrations that accumulated before or after pregnancy. Generally, exposure misclassification biases to the null — that is, use of an incorrect exposure level in a regression analyses of outcome data leads to decreased power to detect a real effect. Thus, the likely implication of the uncertainties and limitations in the dose metrics used in the Seychelles and the Faroe Islands studies is that the probability of observing true associations of dose and response will be reduced. In addition, the magnitude of those observed associations may be underestimated. Therefore, the existence of uncertainties and limitations notwithstanding, those statistically significant dose-response associations observed with any of the dose metrics are likely to reflect (perhaps indirectly) true associations (if other sources of bias have been adequately addressed). Failure to observe statistically significant dose-response associations could well be due to exposure misclassification resulting from one or more of the uncertainties and limitations discussed above. SUMMARY AND CONCLUSIONS Duplicate diet data can potentially provide accurate data on MeHg intake, although interindividual pharmacokinetic variability creates

OCR for page 105
Toxicological Effects of Methylmercury uncertainty in the use of such data to estimate the dose to the fetal brain. The collection of duplicate dietary data places demands on study participants. This approach is, therefore, generally limited to short periods of observation that might not capture critical intake variability in populations with high intraindividual variability in intake of fish. Retrospective dietary data (diary and recall) are relatively simple to collect, but diary-based data are subject to participant errors in species identification, portion estimation, and assignment of MeHg concentration by species. The number of days of dietary-intake data collected needs to be long enough to characterize adequately the frequent fish consumer and to differentiate the levels of less frequent consumption. Recall-based data are additionally subject to recall errors. Such data might be useful in stratifying exposure and in temporal calibration of hair strands. On the other hand, prospective data on all sources of Hg exposure, such as vaccines and dental amalgams and, in particular, dietary intakes of MeHg are essential to understanding the effects of environmental Hg exposures on any outcomes. Quantitative dietary intake data on intakes of all marine food sources can and should be collected in any serious study of this contaminant. Such data are essential for quantifying exposures, separating out the effect modifiers that account for the differences between exposures and target tissue concentrations. Intake data are also essential for identifying possible confounding factors, such as other contaminants or nutrients that are abundant in some of these food sources but not in others. Cord-blood Hg concentration is closely linked kinetically to the fetal-brain compartment and should correlate closely with the concentrations at the target organ near the time of delivery. Cord-blood Hg is less closely linked to the ingested dose. That separation can introduce uncertainty into the back-calculation of a reference dose from cord-blood-based dose-response data. Cord-blood Hg measurement cannot show temporal variability in exposure. It can provide data on a limited portion of gestation whose duration is somewhat uncertain but which occurs late in gestation. That portion of gestation might not correspond to the periods of greatest fetal sensitivity to MeHg neurotoxicity.

OCR for page 105
Toxicological Effects of Methylmercury Maternal-hair Hg concentration is less closely linked kinetically to the fetal-brain compartment than is cord-blood Hg concentration, and the kinetic distance between the maternal-hair and fetal-brain compartments might be a significant source of statistical error in dose-response assessment. Hair Hg measurement can potentially provide a range of dose metrics. Analysis of longer strands corresponding to all or part of gestation will provide average exposure data but no information on temporal variability in exposure. Segmental analysis can isolate specific periods of gestation, but peak exposures might be inadequately represented. Continuous single-strand analysis is a powerful technique that can recapitulate MeHg exposure during the entire period of gestation with accurate representation of peak exposures. This approach presents a range of dose metrics that can be investigated in assessing dose response. Because of intraindividual and interindividual variability in hair-growth rates, attempts to identify hair Hg concentrations corresponding to specific time periods during gestation might be subject to significant error which can result in exposure misclassification in dose-response assessment. The temporal calibration of Hg measurements along a hair strand can be aided by consideration of corresponding dietary intake data for Hg. Each of the dose metrics — dietary records, cord blood, and hair — provides different exposure information. Use of data from two or more of these metrics will increase the likelihood of uncovering a true dose-response relationship. In the Seychelles studies, dose was estimated from the average Hg concentration in a length of hair assumed to represent the first 8 months of pregnancy. That approach precluded observing any intraindividual variability in exposure over the course of gestation. Fish-consumption data for the Seychelles cohort established the generally high level of fish consumption but could not provide any data on intraindividual variability in exposure. In the Faroe Islands studies, dose was estimated from cord-blood and single-sample maternal-hair Hg concentration. The cord-blood Hg data cover exposures over an indeterminate period late in gestation. The hair Hg samples appear not to have been of uniform length and therefore do not necessarily reflect comparable periods

OCR for page 105
Toxicological Effects of Methylmercury of gestation. Those differences in length are relevant only if there is significant variability in MeHg exposure during gestation. Neither of these metrics has the ability to show intraindividual variability over the course of gestation. Fish-consumption data for the Faroe cohort indicated a high rate of consumption of fish with low Hg concentrations, and less frequent consumption of pilot whale containing high concentrations of MeHg. Such a diet suggests a pattern of peaking exposures. Exposure modeling suggests that as reflected by accumulation in hair such peaks might represent a moderate increase above baseline concentrations. The uncertainties and limitations in exposure assessment in these studies can result in exposure misclassification, which will lessen the ability to detect significant dose-response associations and might result in inaccuracies in the derivation of dose-response relationships. If exposure misclassification occurred in the studies of MeHg, such misclassification would tend to obscure any true effect. Therefore, statistically significant dose-response associations are likely to reflect true dose-response relationships, assuming that other sources of bias are adequately addressed. Dose-response assessments using either cord-blood or maternal-hair Hg concentrations are adequate to support the derivation of an RfD. RECOMMENDATIONS Quantitative dietary intake data on patterns of consumption of the primary sources of MeHg including all marine food sources, should be collected in all prospective studies of MeHg exposure. Estimates of exposures will improve dose-response analyses that have implications for regulatory purposes. In future studies, data on maternal fish intake by species and by meal should be collected along with Hg biomarker data. Those data should be used to provide estimates of temporal variability in MeHg intake during pregnancy. Future studies should collect data on maternal-hair, blood, and

OCR for page 105
Toxicological Effects of Methylmercury cord-blood Hg concentrations. All three dose metrics should be considered in attempting to identify dose-response relationships. Data are needed that reliably measure both Hg intake and biomarkers of Hg exposure to clarify the relationship between the different dose metrics. NHANES IV data should be examined when it becomes available to determine if it satisfies those needs. To detect exposure variability, archived hair strands from both the Seychelles and the Faroe Islands studies should be analyzed by continuous single-strand XRF analysis. The possible dose metrics that can be derived from XRF analysis should be examined in the dose-response assessment. Such considerations should also be addressed in future studies. REFERENCES Airey, D. 1983. Total mercury concentrations in human hair from 13 countries in relation to fish consumption and location. Sci. Total Environ. 31(2):157-180. ATSDR (Agency for Toxic Substances and Disease Registry). 1999. Toxicological Profile for Mercury (Update). U.S. Department of Health and Human Services, Agency for Toxic Substances and Disease Registry . Atlanta, GA. Batista, J., M. Schuhmacher, J.L. Domingo, and J. Corbella. 1996. Mercury in hair for a child population from Tarragona Province, Spain . Sci. Total Environ. 193(2):143-148. Berry, M.R. 1997. Advances in dietary exposure research at the United States Environmental Protection Agency-National Exposure Research Laboratory. J. Expo. Anal. Environ. Epidemiol, 7(1):3-16. Boischio, A.A.P., and E. Cernichiari. 1998. Longitudinal hair mercury concentration in riverside mothers along the Upper Madeira river (Brazil). Environ. Res. 77(2):79-83. Bruhn, C.G., A.A. Rodríguez, C. Barrios, V.H. Jaramillo, J. Becerra, U. Gonzáles, N.T. Gras, O. Reyes, and Seremi-Salud. 1994. Determination of total mercury in scalp hair of pregnant women resident in fishing villages in the Eighth Region of Chile. J. Trace Elem. Electrolytes Health Dis. 8(2):79-86. Brune, D., G.F. Nordberg, O. Vesterberg, L. Gerhardsson, and P.O. Wester. 1991. A review of normal concentration of mercury in human blood. Sci. Total Environ. 100(spec No):235-282. Buzina, R., P. Stegnar, S.K. Buzina-Suboticanec, M. Horvat, I. Petric, and T.M. Farley

OCR for page 105
Toxicological Effects of Methylmercury . 1995. Dietary mercury intake and human exposure in an Adriatic population . Sci. Total Environ. 170(3):199-208. Castilhos, Z.C., E.D. Bidone, and L.D. Lacerda. 1998. Increase of the background human exposure to mercury through fish consumption due to gold mining at the Tapajos River region, Para State, Amazon. Bull. Environ. Contam. Toxicol. 61(2):202-209. Cernichiari, E., T.Y. Toribara, L. Liang, D.O. Marsh, M.W. Berlin, G.J. Myers, C. Cox, C.F. Shamlaye, O. Choisy, P. Davidson, et.al. 1995. The biological monitoring of mercury in the Seychelles study. Neurotoxicology 16(4):613-628. Cernichiari, E., R. Brewer, G.J. Myers, D.O. Marsh, L.W. Lapham, C. Cox, C.F. Shamlaye, M. Berlin, P.W. Davidson, and T.W. Clarkson. 1995a. Monitoring methylmercury during pregnancy: Maternal hair predicts fetal brain exposure. Neurotoxicology 16(4):705-710. Chan, H.M., P.R. Berti, O. Receveur, and H.V. Kuhnlein. 1997. Evaluation of the population distribution of dietary contaminant exposure in an Arctic population using Monte Carlo statistics. Environ. Health Perspect. 105(3):316-321. Choi, B.H. 1989. The effects of methylmercury on the developing brain. Prog. Neurobiol. 32(6):447-470. Clewell, H.J., J.M. Gearhart, R. Gentry, T.R. Covington, C.B. VanLandingham, K.S. Crump, and A.M. Shipp. 1999. Evaluation of the uncertainty in an oral Reference Dose for methylmercury due to interindividual variability in pharmacokinetics. Risk Anal. 19(4):547-558. Cox, C., T.W. Clarkson, D.O. Marsh, L. Amin-Zaki, S. Tikriti, and G.G. Myers. 1989. Dose-response analysis of infants prenatally exposed to methyl mercury: An application of a single compartment model to single-strand hair analysis. Environ. Res. 49(2):318-322. Davidson, P.W., G.J. Myers, C. Cox, C. Axtell, C. Shamlaye, J. Sloane-Reeves, E. Cernichiari, L. Needham, A. Choi, Y. Wang, M. Berlin, and T.W. Clarkson. 1998. Effects of prenatal and postnatal methylmercury exposure from fish consumption on neurodevelopment: Outcomes at 66 months of age in the Seychelles Child Development Study. JAMA 280(8):701-707. Davidson, P.W., G.J. Myers, C. Cox, E. Cernichiari, T.W. Clarkson, and C. Shamlaye. 1999. Effects of methylmercury exposure on neurodevelopment. [Letter]. JAMA 281(10):896-897. Dennis, C.A., and F. Fehr. 1975. Mercury levels in whole blood of Saskatchewan residents. Sci. Total Environ. 3(3):267-274. EPA (U.S. Environmental Protection Agency). 1997. Mercury Study Report to Congress. Vol. IV: An Assessment of Exposure to Mercury in the United States .

OCR for page 105
Toxicological Effects of Methylmercury EPA-452/R-97-006. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, and Office of Research and Development. FDA (U.S. Food and Drug Administration). 1992. Compilation of methylmercury results by species for FY=91 and FY=92 . Office of Seafood, Washington, DC. Fujita, M., and E. Takabatake. 1977. Mercury levels in human maternal and neonatal blood, hair and milk . Bull. Environ. Contam. Toxicol. 18(2):205-209. Gaggi, C., F. Zino, M. Duccini, and A. Renzoni. 1996. Levels of mercury in scalp hair of fishermen and their families from Camara de Lobos-Madeira (Portugal): A preliminary study. Bull. Environ. Contam. Toxicol. 56(6):860-865. Gilbert, S.G., and K.S. Grant-Webster. 1995. Neurobehavioral effects of developmental methylmercury exposure. Environ. Health Perspect. 103(Suppl. 6):135-142. Ginsberg, G.L., and B.F. Toal. 2000. Development of a single meal fish consumption advisory for methyl mercury. Risk Anal. 20(1):41-48. Giovanoli-Jakubczak, T., and G.G. Berg. 1974. Measurement of mercury in human hair. Arch. Environ. Health 28(3):139-144. Girard, M., and C. Dumont. 1995. Exposure of James Bay Cree to methylmercury during pregnancy for the years 1983-91. Water Air Soil Pollut. 80:13-19. Gowdy, J.M., R. Yates, F.X. Demers, and S.C. Woodward. 1977. Blood mercury concentration in an urban population. Sci. Total Environ. 8(3):247-251. Grandjean, P., and P. Weihe. 1993. Neurobehavioral effects of interuterine mercury exposure; potential sources of bias. Environ. Res. 61(1):176-183. Grandjean, P., E. Budtz-Jørgensen, R.F. White, P. Weihe, F. Debes, and N. Keiding. 1999. Methylmercury exposure biomarkers as indicators of neurotoxicity in children aged 7 years. Am. J. Epidemiol 150(3):301-305. Grandjean, P., P. Weihe, P.J. Jørgensen, T. Clarkson, E. Cernichiari, and T. Viderø. 1992. Impact of maternal seafood diet on fetal exposure to mercury, selenium, and lead. Arch. Environ. Health 47(3):185-195. Grandjean, P., P. Weihe, L.R. Needham, V. W. Burse, D.G. Patterson, Jr., E.J. Sampson, P.J. Jørgensen, and M. Vahter. 1995. Relation of a seafood diet to mercury, selenium, arsenic, and polychlorinated biphenyl and other organochlorine concentrations in human milk. Environ. Res. 71(1):29-38.. Grandjean, P., P. Weihe, R.F. White, N. Keiding, E., Budtz-Jørgensen, K. Murato, and L. Needham. 1998. Prenatal exposure to methylmercury in the Faroe Islands and neurobehavioral performance at age seven years. Response to workgroup questions for presentation on 18-20 November, 1998. In: Scientific Issues Relevant to Assessment of Health Effects from Exposure

OCR for page 105
Toxicological Effects of Methylmercury to Methylmercury. Appendix II-B.- Faroe Islands Studies. National Institute for Environmental Health Sciences. [Online]. Available: http://ntp-server.niehs.nih.gov/Main_Pages/PUBS/MethMercWkshpRpt. html Hall, R.A., E.G. Zook, and G.M. Meaburn. 1978. Survey of Trace Elements in the Fisheries Resource. Technical Report NMFSSSRF-721. National Oceanic and Atmospheric Administration. National Marine Fisheries Service. 313pp. Haxton, J., D.G. Lindsay, J.S. Hislop, L. Salmon, E.J. Dixon, W.H. Evans, J.R. Reid, C.J. Hewitt, and D.F. Jeffries. 1979. Duplicate diet study on fishing communities in the United Kingdom: Mercury exposure in a “critical group”. Environ. Res. 18(2):351-368. Hecker, L.H., H.E. Allen, and B.D. Dinman. 1974. Heavy metals in acculturated and unacculturated population. Arch. Environ. Health 29(4):181-185. Hislop, J.S., T.R. Collier, G.P. White, D.T. Khathing, and E. French. 1983. The use of keratinised tissues to monitor the detailed exposure of man to methylmercury from fish. Pp. 145-148 in Clinical Toxicology and Clinical Chemistry of Metals, S.S. Brown, ed. New York: Academic Press. Holsbeek, I., H.K. Das, and C.R. Joiris. 1996. Mercury in human hair and relation to fish consumption in Bangladesh . Sci. Total Environ. 186(3):181-188. Humphrey, H.E.B. 1975. Mercury concentrations in humans and consumption of fish containing methylmercury. Pp. 33 in Heavy Metals in the Aquatic Environment Proceedings of the International Conference held in Nashville, TN, December, 1973, P.A. Krenkel, ed. Oxford: Pergamon Press. Hopps, H.C. 1977. The biologic basis for using hair and nail for analyses of trace elements. Sci. Total Environ. 7(1):71-89. IPCS (International Programme on Chemical Safety). 1990. Environmental Health Criteria Document 101 - Methylmercury. Geneva: World Health Organization. IPCS (International Programme on Chemical Safety). 1991. Environmental Health Criteria Document 118 -Inorganic Mercury. Geneva: World Health Organization. Jacobs, H.L., H.D. Kahn, K.S. Stralka, and D.B. Phan. 1998. Estimates of per capita fish consumption in the U.S. based on the continuing survey of food intake by individuals (CSFII). Risk Anal. 18(3):283-291. Katz, S., and A. Chatt. 1988. Pp. 6-12 in Hair Analysis: Application in the Biomedical and Environmental Sciences . New York: VCH Publishers. Kingman, A., T. Albertini, and L.J. Brown. 1998. Mercury concentrations in urine and whole blood associated with amalgam exposure in a U.S. military population. J. Dent. Res. 77(3):461-471. Kjellström, T., P. Kennedy, S. Wallis, and C. Mantell. 1986. Physical and

OCR for page 105
Toxicological Effects of Methylmercury Mental Development of Children with Prenatal Exposure to Mercury from Fish. Stage I: Preliminary Tests at Age 4. National Swedish Environmental Protection Board Report 3080. Solna, Sweden. Kjellström, T., P. Kennedy, S. Wallis, A. Stewart, L. Friberg, B. Lind, T. Wutherspoon, and C. Mantell. 1989. Physical and Mental Development of Children with Prenatal Exposure to Mercury from Fish. National Swedish Environmental Protection Board Report No. 3642. Kuhnert, P.M, B.R. Kuhnert, and P. Erhard. 1981. Comparison of mercury levels in maternal blood, fetal cord blood, and placental tissues. Am. J. Obstet. Gynecol. 139(2):209-213. Kuntz, W.D., R.M. Pitkin, A.W. Bostrom, and M.S. Hughes. 1982. Maternal and cord blood background mercury levels: A longitudinal surveillance. Am. J. Obstet. Gynecol. 143(4):440-443. López-Artiguez, M., A. Grilo, D. Martinez, M.L. Soria, L. Nuñez, A. Ruano, E. Moreno, F. García-Fuente, and M. Repetto. 1994. Mercury and methylmercury in population risk groups on the Atlantic coast of southern Spain. Arch. Environ. Contam. Toxicol. 27(3):415-419. Lubin, B.H. 1987. Reference values in infancy and childhood. Pp. 1683 in Hematology of Infancy and Childhood, 3rd Ed., D.G. Nathan, and F.A. Oski, eds. Philadelphia: W.B. Saunders. MacIntosh, D.L, P.L. Williams, D.J. Hunter, L.A. Sampson, S.C. Morris, W.C. Willett, and E.B. Rimm. 1997. Evaluation of a food frequency questionnaire-food composition approach for estimating dietary intake of inorganic arsenic and methylmercury . Cancer Epidemiol. Biomarkers Prev. 6(12):1043-1050. Mahaffey, K.R., and D. Mergler. 1998. Blood levels of total and organic mercury of residents of the upper St. Lawrence River basin, Quebec: Association with age, gender, and fish consumption. Environ. Res. 77(2):104-114. Marsh, D.O., T.W. Clarkson, C. Cox, G.J. Myers, L. Amin-Zaki, and S. Al-Tikriti. 1987. Fetal methylmercury poisoning. Relationship between concentration in single strands of maternal hair and child effects. Arch. Neurol. 44(10):1017-1022. Nixon, D.E., G.V. Mussmann, and T.P. Moyer. 1996. Inorganic, organic and total mercury in blood and urine: Cold vapor analysis with automated flow injection sample delivery. J. Anal. Toxicol. 20(1):17-22. Oskarsson, A., B.J. Lagerkvist, B. Ohlin, and K Lundberg. 1994. Mercury levels in the hair of pregnant women in a polluted area in Sweden. Sci. Total Environ. 151 (1):29-35. Oskarsson, A., A. Schütz, S. Skerfving, I.P. Hallén, B. Ohlin, and B.J. Lagerkvist. 1996. Total and inorganic mercury in breast milk and blood in relation to fish consumption and amalgam fillings in lactating women. Arch. Environ. Health 51(3):234-241. Pallotti, G., B. Bencivenga, and T. Simonetti. 1979. Total mercury levels in

OCR for page 105
Toxicological Effects of Methylmercury whole blood, hair and fingernails for a population group from Rome and its surroundings. Sci. Total Environ. 11(1):69-72. Pecoraro, V., J.M. Barman, and I. Astore. 1967. The normal trichogram of pregnant women. Pp. 203-210 in Advances in Biology of Skin, Vol. IX, Hair Growth. Proceedings of the University of Oregon Medical School Symposium on the Biology of Skin. Pub. No. 277, Oregon Regional Primate Research Center, W. Montagna, and R.L. Dobson, eds. Oxford: Pergamon Press. Pitkin, R.M, J.A Bahns, L.J. Filer, Jr., and W.A. Reynolds. 1976. Mercury in human maternal and cord blood, placenta and milk. Proc. Soc. Exp. Biol. Med. 151(3):565-567. Ponce, R.A., S.M. Bartell, T.J. Kavanagh, J.S. Woods, W.C. Griffith, R.C. Lee, T.K. Takaro, and E.M. Faustman. 1998. Uncertainty analysis methods for comparing predictive models and biomarkers: A case study of dietary methyl mercury exposure. [Review]. Regul. Toxicol. Pharmacol. 28(2):96-105. Schuhmacher, M., J. Batiste, M.A. Bosque, K.L. Domingo, and J. Corbella. 1994. Mercury concentrations in marine species from the coastal area of Tarragona Province, Spain. Dietary intake of mercury through fish and seafood consumption. Sci. Total Environ. 156(3):269-273. Shamlaye, C.F., D.O. Marsh, G.J. Myers, C. Cox, P.W. Davidson, O. Choisy, E. Cernichiari, A. Choi, M.A. Tanner, and T.W. Clarkson. 1995. The Seychelles child development study on neurodevelopmental outcomes in children following in utero exposure to methylmercury from a maternal fish diet: background and demographics. Neurotoxicology 16(4):597-612. Sherlock, J.C., D.G. Lindsay, J.E. Hislop, W.H. Evans, and T.R. Collier. 1982. Duplicate diet study on mercury intake by fish consumers in the United Kingdom. Arch. Environ. Health 37(5):271-278. Sikorski, R., T. Poszkowski, P. Slawinski, J. Szkoda, J. Zmudzki, and S. Skawinski. 1989. The intrapartum content of toxic metals in maternal blood and umbilical cord blood. Ginekol. Pol. 60(3):151-155. Skerfving, S. 1988. Mercury in women exposed to methylmercury through fish consumption and in their newborn babies and breast milk. Bull. Environ. Contam. Toxicol. 41(4):475-482. Skerfving, S. 1991. Exposure to mercury in the population. Pp. 411-425 in Advances in Mercury Toxicology, T. Suzuki, N. Imura, and T.W. Clarkson, eds. New York: Plenum Press. Smith, J.C., P.V. Allen, and R. Von Burg. 1997. Hair methylmercury levels in U.S. women. Arch. Environ. Health 52(6):476-480. Stern, A.H. 1993. Re-evaluation of the reference dose for methylmercury and assessment of current exposure levels. Risk Anal 13(3):355-364. Stern, A.H. 1997. Estimation of the interindividual variability in the one-

OCR for page 105
Toxicological Effects of Methylmercury compartment pharmacokinetic model for methylmercury: implications for the derivation of a reference dose. Regul. Toxicol. Pharmacol. 25(3):277-288. Stern, A.H., L.R. Korn, and B.E. Ruppel. 1996. Estimation of fish consumption and methylmercury intake in the New Jersey population. J. Expo. Anal. Environ. Epidemiol. 6(4):503-525. Stern, A.H., and M. Gochfeld. 1999. Effects of methylmercury exposure on neurodevelopment. [Letter]. JAMA 281(10):896-897. Stern, A.H., M. Gochfeld, C. Weisel, and J. Burger. 2000. Mercury and methylmercury exposure in the New Jersey pregnant population . Arch. Environ. Health. In press. Thomas, K.W., L.S. Sheldon, E.D. Pellizzari, R.W. Handy, J.M. Roberds, and M.R. Berry. 1997. Testing duplicate diet sample collection methods for measuring personal dietary exposures to chemical contaminants. J. Expo. Anal. Environ. Epidemiol. 7(1):17-36. Vural, N., and H. Ünlü. 1996. Methylmercury in hair of fisherman from Turkish coasts. Bull. Environ. Contam. Toxicol. 57:315-320. Whipple, C., L. Levin, and C. Seigneur. 1996. Sensitivity of mercury exposure calculations to physical and biological parameters of fish consumption. Presented at Fourth International Conference on Mercury as a Global Pollutant, Hamburg, August 4-8. Yess, N.J. 1993. U.S. Food and Drug Administration survey of methyl mercury in canned tuna. J. AOAC Int. 76(1):36-38.