Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
Advisories on fish intake for pregnant women have had to weigh the potential benefits gained from the nutrients in fish against the potential risks associated with contaminants.1,2 Fish is a common source of exposure to methylmercury and other contaminants in varying amounts. Methylmercury is a known neurotoxicant at high levels; it crosses the placenta and accumulates in the fetus at higher concentrations than in the mother.3,4 The possible relation between mercury exposure and children's neurodevelopment has been studied extensively in populations consuming large quantities of fish.5–16 The potential for beneficial effects from fish among populations with relatively low levels of contamination, however, has been relatively unexplored.
Fish intake during pregnancy has the potential to improve fetal development because it is a good source of iron and long chain omega fatty acids, which are necessary for proper development and function of the nervous system.17–20 Clinical trials have shown that fish oil supplementation of infant formula improves visual acuity at 4 months and may improve other aspects of infant growth and cognitive development.21,22 In addition, maternal fish intake during pregnancy has been associated with improved stereoacuity at age 3.5 years.23 Whether maternal fish intake during pregnancy has additional benefits for children's cognitive development is unclear.
The United Kingdom offers a unique setting for assessing the possible beneficial relation between maternal fish intake during pregnancy and the child's cognitive development. Although fish consumption is high, mercury levels measured in commercially available fish have not exceeded levels designated as safe by U.K. regulatory agencies.24 Furthermore, the United Kingdom has been less contaminated by polychlorinated biphenyls than most other developed countries.25 We investigated the relation between maternal fish intake during pregnancy and cognitive development in English children while accounting for prenatal exposure to mercury and postnatal intake of fish. Participants were in the Avon Longitudinal Study of Parents and Children (ALSPAC).
All children born to mothers residing in Bristol and surrounding areas between April 1991 and December 1992 were eligible for the ALSPAC study.26,27 Expectant mothers were recruited during routine prenatal health visits and completed an enrollment questionnaire. An estimated 85% of the eligible population participated. The present investigation was restricted to singleton, term (≥37 completed weeks gestation) children whose mothers provided questionnaires throughout pregnancy and developmental assessments for their children at 15 and 18 months of age (n = 10,092). The ALSPAC Ethics and Law Committee established ethical standards for human subjects, which were approved by the Institute's review committee. Mothers provided consent for participation.
Mothers returned questionnaires by mail at various times before and after delivery to provide information on their diet and lifestyle. Maternal fish intake during pregnancy was assessed by food frequency questionnaire completed during week 32 of gestation. Mothers reported the frequency with which they ate white fish (cod, haddock, plaice, fish fingers, and so on) and oily fish (pilchards, sardines, mackerel, tuna, herring, kippers, trout, salmon, and so on).
We also considered the infants’ intake of fish and breast milk (both of which can be sources of mercury and nutrients) during their first year and maternal dental treatment during pregnancy (a source of elemental mercury). Breastfeeding practices were reported by questionnaire 15 months after the child's birth. The child's fish intake was assessed by questionnaires at 6 and 15 months after birth. Dental treatment of the mother was reported by questionnaire at 33 months after the child's birth. Maternal demographic and lifestyle factors were assessed by questionnaire during week 32 of pregnancy; these factors included age, parity, education level, and use of alcohol or cigarettes.
Total mercury concentration was analyzed in umbilical cord tissue taken at birth from a convenience sample of 1225 children, as study funds allowed. Samples were stored at −20°C and analyzed using cold vapor fluorescence spectrometry at Sheffield Hallam University using reference samples and standard laboratory practices to ensure validity and reliability.28 Before analyses, 1 cm cord tissue samples were washed with distilled water to remove cord blood, weighed, and digested by closed system microwave digestion using nitric acid and hydrogen peroxide, made to 10 mL solution. Mercury concentrations were measured as μg/10 mL digest and divided by the cord weight to give μg/g of cord (wet weight).
The developmental assessments of the child were completed by the mothers and returned by mail. The ALSPAC adaption of the MacArthur Communicative Development Inventory (MCDI) was completed when the child was 15 months of age. The MCDI is a parent-completed assessment used for clinical research on language and communication development.22,29,30 The MCDI assessed the child's vocabulary comprehension and social activity. We evaluated the age-adjusted raw scores for each of these components. The ALSPAC adaptation of the Denver Developmental Screening Test (DDST) was completed by the mother when the child was 18 months of age.26,31 The DDST has been used extensively to screen for developmental differences among groups in research settings,32,33 including those evaluating the effects of mercury. We evaluated the individual DDST components that assessed the child's social and language skills, as well as the DDST total score (which aggregates scores for the child's language, social, fine, and gross motor skills).
The MCDI and DDST scores were initially evaluated as continuous variables. For each score, a dichotomous variable was also created to indicate a low developmental score (a score falling in the lowest 15 percentile, adjusted for age in weeks), and a high developmental score (a score falling in the highest 15 percentile, adjusted for age in weeks).
Maternal fish intake, combining oily and white fish intake, was categorized as follows: rarely or never, once per 2 weeks, 1 to 3 times per week, and 4 or more times per week. We assumed that each fish meal averaged 4.5 ounces to estimate the median ounces of fish eaten per week; this resulted in a corresponding ordinal variable with the values 0, 2.25, 9, and 18 ounces of fish per week for use in the evaluation of trend. Infant fish intake was categorized as at least once per week at age 6 months (yes/no) and at least once per week at age 12 months (yes/no). Child breastfeeding was dichotomized as ever/never. Duration of breastfeeding was examined but was not found to be important in these data. Maternal dental treatment during pregnancy was also dichotomized as yes/no.
We used generalized linear models to estimate the children's mean developmental scores for each level of maternal fish intake during pregnancy and infant fish intake at 6 and 12 months. We estimated odds ratios (ORs) and 95% confidence intervals (CIs) by logistic regression to assess the relation of exposures to low and high developmental scores. We conducted tests for trend by including in regression analysis the ordinal variable representing ounces of fish eaten per week for maternal fish intake. We assessed effect measure modification by considering differences in the magnitude of associations across strata of covariates, and by assessing the statistical significance of the product terms and the improvement in model fit by the likelihood ratio test.
We estimated geometric mean mercury levels in umbilical cord for each level of maternal fish intake during pregnancy. The relation between neurodevelopment scores and total mercury concentration was assessed among the subset of children with umbilical cord mercury measures. Mercury concentration was used as a continuous variable for tests of trend.
All models were adjusted for the child's sex, age at testing (weeks), breastfeeding, maternal receipt of dental treatment, age (years), prenatal smoking (yes/no), prenatal alcohol use (yes/no), birth order (first born/nonfirst born), level of maternal educational achievement according to the English system (CSE = low, Vocational = moderately low, O level = moderate, A level = moderately high, Degree = high and missing), and the quality of the parent and home environment, represented by a self-completion adaptation by ALSPAC of the Home Observation for Measurement of the Environment (HOME) score34 (continuous).
For 7421 of the eligible children, at least 1 of the developmental tests was conducted within 4 months of target test date and complete data for maternal prenatal fish intake and covariates were available. Children were excluded if both developmental tests took place more than 4 months after the target test date (n = 47) or if there were missing data on maternal dental history (n = 1833), breastfeeding (n = 258), parity (n = 254), infant fish intake (n = 209), HOME score (n = 145), or alcohol use (n = 25). Among those with umbilical cord mercury data, 1054 had complete information on developmental outcomes and covariates. The number included in the models varied slightly as a result of variation in response rate and completeness of each developmental questionnaire; none of the models was missing more than 5% of the children.
The characteristics of subjects included in the present analysis were similar to others in the ALSPAC study with respect to maternal fish intake, dental treatment during pregnancy, children's developmental scores, and most covariates (data not shown). However, the mothers in the present analysis were slightly older than other ASLPAC subjects (mean age, 29 vs. 27 years), more likely to have had a university degree (16% vs. 6%), and to have breastfed their child (67% vs. 55%), and less likely to have smoked during pregnancy (15% vs. 28%). The characteristics of the 1054 children with umbilical cord samples for analysis were essentially the same as those without such samples.
Eighty-eight percent of the women in this study ate fish during pregnancy (Table 1). Among those who ate fish, 80% ate fish at least once per week and 65% ate both white and oily fish varieties. Among all women, those who ate white fish were similar those who did not in level of education, but more likely to have breastfed their infant (68% vs. 61%); those who ate oily fish were more likely than those who did not to have had higher education (20% vs. 8%) and to have breastfed (73% vs. 57%).
The mean developmental assessment scores subtly but consistently increased with increased maternal fish intake during pregnancy (Table 2). The largest effect was for the MCDI comprehension score, which was nearly 5 points (7%) higher among children whose mothers ate fish at least once per week during pregnancy compared with those whose mothers did not eat fish. For the MCDI social activity score, the greatest differences occurred between no fish and any fish, with the incremental increase in fish intake only slightly strengthening the association. The DDST total score was 2% higher among children whose mothers ate fish 1 to 3 times per week compared with those whose mothers ate no fish. Most developmental scores increased only fractions of a point with each increase in fish intake during pregnancy.
The odds of scoring high or low were also subtly associated with maternal fish intake. Again, the MCDI scores appeared to have the strongest association with maternal fish intake. Fish intake during pregnancy was associated with a decreased probability of low MCDI social activity score (Table 3) and an increased probability of high MCDI vocabulary comprehension and social scores (Table 4). High and low DDST language scores followed a similar pattern, but other DDST scores were not as strongly associated with maternal fish intake.
These associations were not altered by the child's fish intake (unadjusted data not shown); however, the infants’ fish intake was independently associated with an increase in most neurodevelopment assessment scores. Approximately 43% of the children ate fish at least once per week at 6 months of age, and most (81%) ate fish at least once per week at 12 months of age. All mean MCDI scores were slightly higher among children who ate fish at least once per week at 6 months and at 12 months of age. Relations between the infants’ fish intake and DDST scores followed a similar pattern but were of smaller magnitude.
When maternal intake of each type of fish (oily and white) was assessed separately, the results bore a similar positive relation to the child's developmental assessment scores (data not shown). Breastfeeding was more frequent among those who ate oily fish (73%) than among those who age white fish (68%). Breastfeeding was not strongly associated with most developmental scores and did not confound the association between maternal or child fish intake and neurodevelopment. The effects of maternal fish intake in stratified analysis were stronger and more precise among those who breastfed, although this interaction did not reach statistical significance.
Factors associated with higher developmental assessment scores included the child being female or first born, a higher HOME score, higher maternal education, and mother abstaining from alcohol during pregnancy; however, these factors did not modify relations between maternal fish intake and developmental scores. Maternal report of dental treatment during pregnancy was associated with subtle decreases in developmental scores, but did not confound or modify the effect of the relation between fish and neurodevelopment (data not shown).
Maternal fish intake during pregnancy was associated with increased umbilical cord mercury concentrations (Fig. 1). Overall, however, the mercury levels were low in this population (median = 0.01 μg/g wet weight) and not associated with developmental scores (Table 5). The relation between maternal fish intake and cognitive development remained similar when mercury levels were included in the model for this subset of children; the positive association with fish intake remained but was not strengthened when adjusted for mercury.
Maternal fish intake during pregnancy and infants’ fish intake during their first year were independently associated with modestly but consistently higher developmental scores for language comprehension and social activity components of the MCDI at 15 months of age and DDST scores at 18 months of age. Although small changes on these assessments may have little clinical significance, they could be important when considering subtle shifts in development at the population level.35
The relation between fish and neurodevelopment was strongest for those eating fish 1 to 3 times per week. Our results suggest that this optimal fish intake during pregnancy had modest beneficial effects on the children's cognitive development, specifically communication skills. Fish consumption by pregnant women at this frequency (averaging ≤4 ounces per serving 1–3 times per week) is within the limit of 12 ounces of fish that are low in mercury per week advised by the U.S. Food and Drug Administration and U.S. Environmental Protection Agency in March 2004.2 Thus, this advisory does not interfere with the potential benefits of fish consumption suggested by this study.
In this population, fish intake provided subtle beneficial rather than detrimental effects on cognitive development. Further investigation is needed to determine what aspects of fish consumption might explain this association. Fish intake may be an indirect sign of better caregiving or better diet overall, which could in turn improve cognitive development. Fish more directly provides a dietary source of lean protein, iron, and long-chain omega fatty acids, which may compliment growth and development.
Much of the recent research related to the nutritional contribution of fish intake has focused on the long-chain omega 3 fatty acids.17–23 Fish, especially oily fish, is a dietary source of eicosapentaenoic and docosahexaenoic acids (DHA), which are important in the structural and functional development of the brain in utero and through the first year after birth.17,18,36,37 The concentration of DHA in fetal brain increases rapidly during the third trimester of gestation.38 Preformed DHA (C22:6n-3) is provided to the fetus by placental transfer and to the infant through breast milk.38,39 The infant also metabolizes 18-carbon fatty acids in their own diet to the form needed by tissues (C22:6n-3); however, it is unclear at what stage the infant's synthesis of DHA becomes sufficient and independent of the mother's supply.38–41 Most clinical trials of DHA supplementation have provided DHA to infants immediately after birth rather than to their pregnant mothers, and few have evaluated cognitive development in children older than 1 year of age. In a small study of subjects in the ALSPAC study, maternal DHA levels were associated with improved visual stereoacuity among offspring at 3.5 years of age.23 The observed benefits of fish intake in our data are difficult to compare with those observed in clinical trials of DHA supplementation because of differences in the measured exposure dose and timing.
Although total cord mercury levels increased with maternal fish intake, our data did not suggest adverse developmental effects associated with mercury. However, the mercury levels measured appeared low in this population (median umbilical cord concentration 0.01 μg/g wet weight, with only 33 children as high as or above 0.1 μg/g). These levels were considerably lower than the median reported for Faroese children (0.3 μg/g dry weight),41 which would roughly translate to 0.04 μg/g if adjusted for the water content of umbilical cord (85–90%) as measured in these samples from the United Kingdom.42 Although calibration data for the mercury assays we used were not available, subsequent analyses of samples from other ALSPAC subjects using a different method gave even lower values, further supporting our judgment that mercury levels were low in this population. Few other studies have measured mercury in umbilical cord tissue.43 Mercury concentrations in hair samples taken in the United Kingdom were 1.6 ppm,44 much lower than concentrations from hair samples from the Faroe Islands (4.3 ppm) where adverse effects of prenatal exposure to methylmercury were clear.5 As noted here, the reported mercury concentrations measured in the U.K.'s consumer fish supplies during the years that the ALSPAC study specimens were collected measured below the safety standards set by the World Health Organization.24
The lack of adverse effect associated with fish intake in this population could be the result of factors other than low levels of exposure to methylmercury or polychlorinated biphenyls.25 The developmental tests in this study were completed by the mother and focused primarily on her report of her child's language and communication skills.45,46 Other developmental domains such as motor development and behavior may be more sensitive to the adverse effects of mercury. Likewise, other instruments provide greater sensitivity for detecting the subtle effects of neurotoxicants in young children and are less reliant on maternal report. These children were evaluated when less than 2 years of age; developmental evaluations tend to be more stable among school-aged children than toddlers.47
For most of the outcomes, we noted a threshold effect for the relation between fish and cognitive development, indicating benefit from eating fish at least once every 2 weeks, but not incremental increase in benefit with more frequent fish consumption. This threshold could indicate that some fish, but not large amounts of fish, are needed to benefit development. We looked for evidence that nonfish-eating women differ in other ways such as socioeconomic or lifestyle factors. In stratified analyses, we found no patterns to indicate that the relation between fish and cognitive development differed by strata of maternal education, parity, breastfeeding, or alcohol intake, although results in some strata were unstable.
Each of the questionnaires was completed by approximately 85% of the target population. Many women did not provide information for all the exposures, outcomes, and covariates needed for these analyses. Based on their partial data, women with incomplete data were similar to those included in this analysis, but slightly less likely to have breastfed their child or to have higher education. Although these and other factors related to missing data could have affected the relations between fish consumption and cognitive development, we have no reason to suspect that the associations differed among those not included.
In general, fish intake tends to be reported with high reliability and validity in food frequency questionnaires.47–49 As mentioned here, fish intake is frequent in Great Britain. We found that maternal fish intake, as reported in the questionnaire, correlated with infant cord mercury levels from a small sample of this population; similar correlations have been reported with maternal prenatal serum DHA levels measured in another sample from the larger ALSPAC study population (P <0.004).23 These findings validate our use of reported fish intake as a surrogate for both DHA and mercury in this population.
In summary, maternal and infant fish intake may subtly enhance the child's early development of language and communication skills. ALSPAC is ongoing and will continue collecting data on many aspects of neurodevelopment as the children age, which will allow further investigation of the long-term neurodevelopment effects related to fish intake and low-level total mercury levels. The balance between the benefits of fish itself and the adverse effects of mercury contamination in relation to neurodevelopment remains unclear for populations in which mercury levels are higher than in this population but still within the World Health Organization guidelines.50
We are extremely grateful to all the mothers who took part and to the midwives for their cooperation and help in recruitment. The whole ALSPAC study team comprises interviewers, computer technicians, laboratory technicians, clerical workers, research scientists, volunteers, and managers who continue to make the study possible. The ALSPAC study is part of the WHO-initiated European Longitudinal Study of Pregnancy & Childhood. We also thank David Umbach, Lars Bjorkman, David Shore, Ruth Little, and Donna Baird for their critical feedback on early versions of the manuscript.
1. Center for Food Safety and Applied Nutrition, US Food and Drug Administration. Consumer Advisory—Risk of Mercury in Fish. An Important Message for Pregnant Women and Women of Childbearing Age Who May Become Pregnant About the Risks of Mercury in Fish.
T01-04 January 12, 2001.
2. US Department of Health and Human Services and US Environmental Protection Agency. 2004 EPA and FDA Advice for: Women Who Might Become Pregnant, Women Who Are Pregnant, Nursing Mothers, Young Children
. March 2004, EPA-823-R-04-005.
3. Choi BH, Lapham LW, Amin-Zaki L, et al. Abnormal neuronal migration, deranged cerebral cortical organization, and diffuse white matter astrocytosis of human fetal brain: a major effect of methylmercury poisoning in utero. J Neuropathol Exp Neurol
4. Clarkson TW. The toxicology of mercury. Crit Rev Clin Lab Sci
5. Grandjean P, Weihe P, White RF, et al. Cognitive deficit in 7-year old children with prenatal exposure to methylmercury. Neurotoxicol Teratol
6. Grandjean P, Weihe P, White RF, et al. Cognitive performance of children prenatally exposed to ’safe’ levels of methylmercury. Environ Res
7. Crump KS, Kjellstrom T, Shipp AM, et al. Influence of prenatal mercury exposure upon scholastic and psychological test performance: benchmark analysis of a New Zealand Cohort. Risk Analysis
8. Myers GH, Davidson PW, Cox C, et al. Prenatal methylmercury exposure from ocean fish consumption in the Seychelles child development study. Lancet
9. Myers GJ, Marsh DO, Davidson PW, et al. Main neurodevelopmental study of Seychellois children following in utero exposure to methylmercury from a maternal fish diet: outcome at six months. Neurotoxicology
10. Myers GJ, Davidson PW, Shamlaye CF, et al. Effects of prenatal methylmercury exposure from a high fish diet on developmental milestones in the Seychelles Child Development Study. Neurotoxicology
11. Myers GJ, Davidson PW. Prenatal methylmercury exposure and children: neurologic, developmental and behavioral research. Environ Health Perspect
. 1998;106(suppl 3):841–847.
12. Myers GJ, Davidson PW, Palumbo D, et al. Secondary analysis from the Sechelles Child Development Study: the Child Behavior Checklist. Environ Res
13. Davidson PW, Meyers G, Cox C, et al. Longitudinal neurodevelopmental study of Seychellois children following in utero exposure to methylmercury from a maternal fish ingestion: Outcomes at 19 and 29 months. Neurotoxicology
14. Davidson PW, Meyers G, Cox C, et al. Effects of prenatal and postnatal methylmercury exposure from fish consumption on neurodevelopment. JAMA
15. Davidson PW, Palumbo D, Myers GJ, et al. Neurodevelopmental outcomes of Seychellois children from the pilot cohort at 108 months following prenatal exposure to methylmercury from a maternal fish diet. Environ Res
16. Myers GH, Davidson PW, Cox C, et al. Prenatal methylmercury exposure from ocean fish consumption in the Seychelles child development study. Lancet
17. Carlson SE, Neuringer M. Polyunsaturated fatty acid status and neurodevelopment: a summary and critical analysis of the literature. Lipids
18. Yehuda S. Tabinovitz S, Mostofsky DI. Essential fatty acids are mediators of brain biochemistry and cognitive functions. J Neurosci Res
19. Olsen SF, Secher NJ. Low consumption of seafood in early pregnancy as a risk factor for preterm delivery: prospective cohort study. BMJ
20. Elias SL, Innis SM. Infant plasma trans, n-6, and n-3 fatty acids and conjugated linoleic acids are related to maternal plasma fatty acids, length of gestation, and birth weight and length. Am J Clin Nutr
21. SanGiovanni JP, Berkey CS, Dwyer JT, et al. Dietary essential fatty acid, and visual resolution acuity in healthy fullterm infants: a systematic review. Early Hum Dev
22. SanGiovanni JP, Parra-Cabrera S, Colditz GA, et al. Meta-analysis of dietary essential fatty acids and long-chain polyunsaturated fatty acids as they relate to visual resolution acuity in healthy preterm infants. Pediatrics
23. Williams C, Birch EE, Emmett PM, et al. Stereoacuity at age 3. 5 y in children born full-term is associated with prenatal and postnatal dietary factors: a report from a population-based cohort study. Am J Clin Nutr
25. Dwarka S, Harrison DJ, Hoodless RA, et al. Organochlorine compound residues in human milk in the United Kingdom 1989–1991. Hum Exp Toxicol
27. Golding J, Pembrey M, Jones R, ALSPAC Study Team. ALSPAC—The Avon Longitudinal Study of Parents & Children. I. Study methodology. Paediatr Perinat Epidemiol
28. Winfield SA, Boyd ND, Vimy MJ, et al. Measurement of total mercury in biological specimens by cold vapor atomic fluorescence spectrometry. Clin Chem
29. Feldman HM, Dollaghan CA, Campbell TF, et al. Measurement properties of the MacArthur communicative development inventories ages one and two years. Child Dev
30. Fenson L, Dale PS, Reznick JS, et al. The MacArthur Communicative Development Inventories: User's Guide and Technical Manual
. San Diego: Singular Publishing Group; 1993.
31. Frankenburg WK, Dodds JB, Fandal AW, et al. Denver Developmental Screening Test Reference Manual, Revised 1975 Edition
. University of Colorado Medical Center; 1975.
32. First LR, Palfrey JS. The infant or young child with developmental delay. N Engl J Med
33. Glascoe FP, Byrne KE, Ashford LF, et al. Accuracy of the Denver-II in developmental screening. Pediatrics
34. Caldwell B, Bradley R. Home Observation of the Environment. Administration Manual.
Little Rock, AR: University of Arkansas at Little Rock; 1984.
35. Needleman HL. Correction: lead and cognitive performance in children. N Engl J Med
36. Neuringer M, Reisbick S, Janowsky J. The role of n-3 fatty acids in visual and cognitive development: current evidence and methods of assessment. J Pediatr
37. Simopoulos AP. Human requirement for N-3 polyunsaturated fatty acids. Poult Sci
38. Clandinin MT. Brain development and assessing the supply of polyunsaturated fatty acid. Lipids
39. Nettleton JA. Are n-3 fatty acids essential nutrients for fetal and infant development? J Am Diet Assoc
40. Crawford MA. The role of essential fatty acids in neural development: implications for perinatal nutrition. Am J Clin Nutr
41. Dalgard C, Grandjean P, Jorgensen PJ, et al. Mercury in the umbilical cord: implications for risk assessment for Minamata disease. Environ Health Perspect
42. Golding J, Cooke M. Unpublished data for Final Report to MAFF. February 26, 1998.
43. Akagi H, Grandjean P, Takizawa Y, et al. Methylmercury dose estimation from umbilical cord concentrations in patients with Minimata disease. Environ Res
44. Airey D. Total mercury concentrations in human hair from 13 countries in relation to fish consumption and location. Sci Total Environ
45. Meisels SJ. Can developmental screening tests identify children who are developmentally at risk? Pediatrics
46. Grandjean P, Weihe P. Neurobehavioral effects of interuterine mercury exposure: potential sources of bias. Environ Res
47. Hu FB, Rimm E, Smith-Warner SA, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr
48. Erkkola M, Karppinen M, Javanainen J, et al. Validity and reproducibility of a food frequency questionnaire for pregnancy Finnish women. Am J Epidemiol
49. Shatenstein B, Kosatsky T, Nadon S, et al. Reliability and relative validity of fish consumption data obtained in an exposure assessment study among Montreal-area sportfishers. Environ Res
50. World Health Organization. Evaluation of Certain Food Additives and Contaminants.
Forty-first Report of the Joint FAO/WHO Expert Committee on Food Additives. Technical Report Series No. 837. Geneva: WHO; 1993.
Supplemental Digital Content
© 2004 Lippincott Williams & Wilkins, Inc.