Risk assessment in relation to neonatal metal exposure† Agneta Oskarssona, Ira Palminger Hall�enb, Johanna Sundbergc and Kierstin Petersson Graw�ec a Department of Food Hygiene, Faculty of Veterinary Medicine, Swedish University of Agricultural Sciences, Box 7009, SE-75007 Uppsala, Sweden b Safety Assessments 681/RSR, Astra AB, SE-15185 S�odert�alje, Sweden c Toxicology Division, National Food Administration, Box 622, SE-75126 Uppsala, Sweden Rapid changes in organ development and function occur during the neonatal period.During this period the central nervous system is in a rapid growth rate and highly vulnerable to toxic effects of, e.g., lead and methylmercury. Furthermore, the kinetics of many metals is age-specific, with a higher gastrointestinal absorption, less effective renal excretion as well as a less effective blood-brain barrier in newborns compared to adults. Due to their low body weight and high food consumption per kg of body weight, the tissue levels of contaminants can reach higher levels in newborns than in adults.Generally, there is a low transfer of toxic metals through milk when maternal exposure levels are low. However, knowledge is limited about the lactational transport of metals and the potential effects of metals in the mammary gland on milk secretion and composition. There are some data from rodents on the lactational transfer and the uptake in the neonate of inorganic mercury, methylmercury, lead and cadmium.Metal levels in human breast milk and blood samples from different exposure situations can give information on the correlation between blood and milk levels. If such a relationship exists, milk levels can be used as an indicator of both maternal and neonatal exposure. Better understanding of the neonatal exposure, including kinetics in the lactating mother and in the newborn, and effects of toxic metals in different age groups is needed for the risk assessment.Interactions with nutritional factors and the great beneficial value of breast-feeding should also be considered. Keywords: Lead; mercury; cadmium; neonate; infant; developmental effects; breast milk; risk assessment As an end-point of risk assessment of food additives and contaminants, the acceptable or tolerable daily intake is established. The definition by the Joint FAO/WHO Expert Committee on Food Additives of acceptable daily intake1 is ‘an estimate of the amount of a food additive, expressed on a body weight basis, that can be ingested daily over a lifetime without appreciable health risk’.For toxic metals the concept of ‘provisional tolerable weekly intake’ (PTWI) is used to express that these compounds have no intended function, are accumulated over time and the evaluation is tentative due to the paucity of reliable data.1 For risk assessment of metals during infancy it is important to recognise that the child is not a small adult. The early postnatal period is characterized by rapid growth and development, the child has unique metabolic and physiological pathways and has a different exposure pattern than the adult.2,3 With this in mind, it is surprising that newborns and lactating women are generally not recognized as risk groups in risk assessment in a similar way as fetuses and pregnant women are.The susceptibility to toxic agents depends on the substance, the exposure and the age. For risk assessment of a particular substance it would be preferable to differentiate between age groups for exposure assessment, kinetics and adverse health effects. WHO2 has recommended the following definitions for infants and young children: the neonatal period extends from birth to 4 weeks, infancy from 4 weeks to 1 year and young childhood from 1 to 5 years.The present review will focus on exposure patterns and biological differences between neonates/infants and adults with impact on risk assessment (Table 1).Results, mainly from our own studies on the toxic metals lead, mercury and cadmium, will be used as examples. It should be kept in mind, that also other elements, such as arsenic, copper and aluminium, are of concern regarding exposure and effects in neonates and infants. Neonatal exposure to toxic elements The main source of exposure during the neonatal and first part of infancy periods is breast milk. Thus, it is important to study the lactating mother and the special conditions determining the kinetics and transport of toxic elements into breast milk.Kinetics of metals during lactation Lactation represents a stage of altered physiological conditions. The disposition and clearance of chemicals during lactation are major factors determining the breast milk concentration. We have studied the influence of lactation on the kinetics of lead during 10 days after a single intravenous injection in lactating and non-lactating mice.4 The elimination from plasma was more rapid in lactating mice than in those non-lactating.Half lives in plasma were 26 h in lactating and 65 h in non-lactating mice during the terminal phase. Significant differences were revealed with the mean plasma clearance and the volume of distribution being about 3 times higher in lactating mice. Partly, the higher clearance can be explained by an additional route of excretion, namely milk. About 30% of the administered dose † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997.Table 1 Biological characteristics in neonates/infants with implication to risk assessment Exposure— Milk Lacting mother Infant formula/drinking water Kinetics— Absorption Distribution Elimination Effects— CNS Other effects Analyst, January 1998, Vol. 123 (19–23) 19was excreted in milk; however, this accounts only for 1/3 of the total plasma clearance. Increased excretion via the bile during lactation is suggested as a possible explanation.The increase in mammary gland and blood volume probably contributes to the increase in plasma volume of distribution in lactating mice, as well as the exchangeable pool of lead in the bone. In whole blood there was no difference in half lives, demonstrating that more useful data, revealing the kinetic characteristics, are obtained from lead in plasma than from whole blood. In the epidemic of methylmercury poisoning in Iraq, blood clearance half-times of total mercury were significantly shorter in lactating women than in non-lactating females or in males.5 Kinetic parameters for mercury in mice have been determined after intravenous injection of either methylmercuric chloride or inorganic mercuric chloride.6 The demethylation of methylmercury was taken into consideration when estimating the toxicokinetic parameters by determination of the ratio between mercury in plasma and whole blood.Preliminary results indicate that the kinetics can be described by a three compartment model. Significant differences between lactating and non-lactating mice were present for methylmercury, with a higher plasma clearance and a larger volume of distribution in lactating than in non-lactating mice. This may be explained by increased biliary excretion, increased blood/plasma volume and lower content of plasma proteins during lactation. For inorganic mercury there were no evident differences in the kinetics between lactating and non-lactating females.The milk excretion of mercury in mice during 9 days was approximately 4 and 8%, of the administered dose of methylmercury and inorganic mercury, respectively.6 Transport of metals to milk Milk production and secretion is a complex process. Even if the composition of milk differs from one species to another, the mechanisms by which the various milk components are secreted seems to be remarkably similar. There are mainly five pathways for the secretion of milk components.7 1.The exocrine pathway is quantitatively the most important and is similar to other secretory cells. The major milk proteins casein, lactabumin and possibly lactoferrin are synthesized on ribosomes and inserted across the membranes of the rough endoplasmic reticulum and transferred to the Golgi system. Casein combines with calcium and phosphah stabilizes the molecule, and form large aggregates, micelles.The Golgi vesicles with micelles migrate to the apical membrane and release the micelles into the milk alveoli via exocytosis. Lactose, milk citrate and phosphate are also secreted via this pathway. 2. Milk lipids are secreted via this pathway, unique to the mammary gland, involving triglyceride synthesis from precursor fatty acids transported from the plasma or synthesized within the alveolar cell. The triglycerides form large lipid droplets surrounded by a membrane in a complex known as the milk fat globuli, which are extruded from the cell. 3. Small molecules, including sodium, potassium, chloride and glucose can pass across the apical membrane. 4. Immunoglobulins bind to a receptor at the basolateral membrane and are transported by a transcytotic pathway. Probably this pathway also is responsible for the secretion of peptide hormones (prolactin and insulin) and some plasma proteins. 5. Some components may enter the milk via a paracellular pathway, which is tightly closed during full lactation.The transport of xenobiotics into milk is supposed to follow the same pathways as do milk components. However, except for medical drugs, few data are available on the transport of chemicals into milk.8,9 For many drugs, passive diffusion is believed to be the major transport mechanism but also binding to carrier proteins in plasma may play a role. Small, uncharged molecules as well as high lipid solubility and low plasma protein binding are factors that generally favour the transport of chemicals into milk.We have investigated the binding of radioactively labelled lead, cadmium and mercury in milk fractions after administration to rats and mice (Table 2). Each metal is distributed in a characteristic way between the milk fractions. Lead is almost exclusively found in the casein fraction in rat milk, while the highest proportions of cadmium and methylmercury are found in fat whilst inorganic mercury is found mostly in whey fractions.10–12 Toxic elements could be expected to enter milk in a similar way as essential trace elements, but there is limited information available also on the transport of essential elements in milk.13 Iron and copper are believed to be excreted in breast milk by binding to specific or non-specific carrier proteins in plasma.About one-third of the iron in human milk is associated with the low molecular mass fraction, one-third with the milk fat, mainly the outer fat globule membrane and of the remainder, about 10% is found in the casein fraction and a fraction with the unique milk glycoprotein, lactoferrin.14 Zinc and copper in human milk are approximately 75% associated to the whey fraction, 15% to fat and the rest present in casein micelles. The major copper- and zinc-binding protein in whey appears to be serum albumin and some may also be associated with citrate and free amino acids.A pathway, by which lead is transported into milk through the mammary gland has been suggested.15 Studies by X-ray microanalysis of mouse mammary cells showed that lead was associated with casein micelles both inside the secretory cells and in the milk alveolar lumen.Lead has a high affinity to casein and in rat milk 95% and in human milk 60–80% of the lead was bound to casein.10 It is known that calcium is secreted into milk within casein micelles by exocytosis of secretory vesicles derived from the Golgi secretory system and lead is probably transported in a similar way as calcium.The high binding of lead to casein may also explain the differences in the milk excretion of lead between different species. Thus, there is a high excretion of lead in the casein-rich milk from rats and mice and a low excretion in human milk with a low content of casein. Breast milk normally contains trace levels of metals but elevated levels are found after high maternal exposure.16 Blood and milk levels of lead, mercury and cadmium were determined at six weeks after delivery in women living in the surroundings of a copper and lead smelter and in a control area in the north of Sweden (Table 3).In general, the levels were low and for lead and cadmium there was no significant correlation between the levels in milk and blood.17 However, significant relationships Table 2 Distribution of metals in milk fractions after administration of metal salts to lactating rats or mice10–12 Metal Fat (%) Casein (%) Whey (%) Pb 2 96 2 Cd 49 43 7 Hg, inorganic 15 31 41 MeHg 39 11 34 Table 3 Concentrations* of Pb, Cd and Hg in milk and blood in women at 6 weeks after delivery (mg l21)17,18 Milk Blood Pb 0.7 ± 0.4 32 ± 8 Cd 0.06 ± 0.04 0.9 ± 0.3 Hg 0.6 ± 0.4 2.3 ± 1.0 * Mean ± s, n = 75 for Pb and Cd; n = 30 for Hg. 20 Analyst, January 1998, Vol. 123were found for both total and inorganic mercury, but not for organic mercury, in milk and blood.18 The average total mercury level in milk was 27% of the level in whole blood and for inorganic mercury the average level in milk was 55% of the levels in blood.Total and inorganic mercury in blood and milk were correlated with the number of amalgam fillings. The concentrations of total and organic mercury in blood (but not in milk) were correlated with the estimated recent intake of methylmercury via fish consumption. The results indicate an efficient transfer of inorganic mercury from blood to milk and even if the levels in milk were low the exposure to the nursing infant corresponds to approximately one-half the TDI for adults recommended by WHO.19 At higher exposure levels of methylmercury, as in a population in the Faroe Islands, the hair mercury concentrations in infants were found to increase with duration of nursing period.20 Other sources of neonatal exposure to metals For infants that are not breastfed, exposure to toxic metals may occur from infant formula and drinking water.Any toxicant present in water used to make up infant formula will be taken in by an infant in larger quantities per unit body weight than by an older child or adult using the same water supply. Water may be contaminated at the source, during treatment and transport. Lead in drinking-water from lead pipes in combination with corrosive water is a problem in many areas in Europe. Also, the occurrence of high copper levels from domestic piping has recently been recognized and evaluated for health effects especially in infants.21 There are also possibilities of dermal exposure, e.g., some cosmetics and folk medicine used on newborns in some parts of the world.A high skin absorption may occur especially under occlusive conditions, such as is the case for diapers. The cadmium intake in infants from breast milk and infant formula has been compared. Calculated for a 4 months old infant, with a body weight of 6.6 kg and a daily intake of 1000 ml, the cadmium intake from infant formula powder (not including the contribution from drinking water) is 6 times higher than the intake from breast milk.17,22 Cadmium intake from soya based formula, which is recommended for infants with cow milk allergy, is about 20 times higher than the intake from breast milk.The bioavailability of cadmium in these diets in the newborn is not known. Low-dose effects of cadmium in the developing CNS have been shown (see under Toxic effects). Kinetics in the newborn There are significant differences in absorption, distribution, metabolism and excretion demonstrated in neonates and adults.23 In the newborn, gastric pH is high (pH 6–8).Adult values are reached by 3 months of age. The difference in distribution between children and adults are mainly determined by the relative proportion of total body water, that decreases from approximately 75% in infants to about 55% by 12 years of age and by the low levels of plasma proteins, resulting in a higher amount of unbound chemicals in the newborn.Most metabolic enzymes are present at birth and their activities increase with age. Adult levels of most enzyme systems are achieved in humans by 2–3 months of age. Renal function increases over the first year of life. At birth glomerular filtration is 30–40% of adult and tubular filtration is also lower in infants than in adults. The ability to concentrate urine is initially low and may be associated with lower sensitivity to nephrotoxic agents compared to adults.Newborns, as shown in laboratory animals, also have a lower capacity to excrete compounds in the bile. The age-related differences in excretion result in decreased body clearance in the newborn. This has been shown for cadmium, mercury and lead in animals and for lead also in humans. Newborns of all species absorb metals to a higher extent than adults. Thus, the bioavailability of lead ranges from 10% in adults to 40% in children below the age of 6 years.24 The uptake of metals in neonates is dependent on the bioavailability of the metals from milk diets.Most infants are breastfed during the first months but the availability from infant formula is also of importance. There is conflicting evidence in the literature concerning the effect of milk on the absorption of lead in sucklings with some studies reporting an inhibiting effect of milk on lead absorption in the sucklings25 whereas others have seen a stimulating effect.26 We have found the highest and most rapid absorption of lead in suckling rats from diets with a low casein content, that is human milk.10 These rats also had the highest accumulation of lead in the duodenum, which is the principal absorption site for lead in adults.Lead in casein-rich milk, like rat and mice milk, was absorbed to a lower extent and more slowly with the highest accumulation in the ileal mucosa.A delayed absorption of lead was shown in suckling pups, which can be explained by pinocytosis of the retained caseinbound lead in the ileum.27 The maximal concentrations in tissues of the sucklings were not reached until 2 to 3 days after oral administration. The delay in absorption and measurements at different timepoints after administration may partly explain the contradictory results on the effect of milk on lead absorption. Thus, when measurements are performed after short periods of time, there is a lower absorption of lead from milk, whereas after very long periods of time even higher absorption may be reached from milk diets.There are indications of higher bioavailability of lead from human milk than from infant formula in the Glasgow duplicate study, where similar blood levels were found in breastfed infants as in bottle-fed, although the dietary intake of lead was much lower in breastfed infants.28 Methylmercury is well absorbed in the gastrointestinal tract of infants and in adults.In experimental animals there is a low demethylating ability of the intestinal microflora and a low excretion rate before weaning.29 Thus, it seems likely that there is a higher retention of methylmercury in young infants than in older children and adults. The blood-brain barrier is not fully developed until around 6 months after birth in man. Metals are retained in the brain more readily during infancy than during adulthood. This has been shown in rats for lead,30 cadmium31,32 and mercury.33 Toxic effects Animal studies have shown that toxic effects of certain chemicals are different at different stages of development.In addition, a perinatal exposure of a chemical may produce delayed effects later on in life. Age-dependent differences in toxicological response can be due to differences in kinetics and be both quantitative and qualitative. Also the toxicological response can depend on the development of receptors to various chemicals, that are developed at various stages.The brain is especially vulnerable during the brain growth spurt, that is the period when the brain undergoes several fundamental changes, like dendritic and axonal growth, synaptogenesis, rapid myelination, etc.34 In many species, such as mice, rats, dogs and man this period is postnatal. In humans, the full number of neurons is not reached until about 2 years of age and myelination is not complete until adolescence.The CNS toxicity of lead and methylmercury in early developmental stages is well known. High dose exposure causes pronounced effects whereas the effects in the low dose range are subtle and many confounding factors such as socio-economic status, intellectual stimulation and nutrition, have to be taken into account when evaluating the effects. For lead, the effect on Analyst, January 1998, Vol. 123 21cognitive ability seems to have no threshold and the most likely size of the effect is a decrement of between 1 and 3 IQ points for each 10 mg per 100 ml increment in blood lead.35 For methylmercury there are conflicting evidence from the studies in the Seychelles and Peru, where there were no significant developmental effects in children36,37 and the studies from the Faroe Islands, where significant effects recently have been reported.38 A few studies in experimental animals have shown neurotoxic and behavioural effects after early cadmium exposure.39,40 Relatively high doses and single administrations have been used.In a recent study with a continuous low dose cadmium exposure (5 ppm) via drinking water to lactating dams, we found neurochemical disturbances of the serotonergic system in the offspring.41 There were no detectable levels of cadmium in the brain of the suckling pups. Either the serotonergic system is very sensitive to cadmium during this period or the effects are indirectly caused by a disturbance on milk production and composition.The kidney function is immature in the newborn and there are indications that some compounds are less toxic to the immature kidney than to the mature. We found an effect on plasma urea nitrogen in offspring exposed to cadmium via drinking water after weaning but not in the pups exposed both during lactation and postweaning.41 This may be explained by induction of metallothionein in the pups that were continuously exposed to low doses during lactation resulting in a protection of kidney function.Newborns do not have a fully competent immune defence system. Profound effects on the immune system have been induced after exposure of chemicals during the development of the lymphoid system. No specific data are available for human infants. We studied the immunomodulating effects in mice after perinatal exposure to methylmercury (prior to mating, during gestation and lactation). There was an effect on thymocyte development and a stimulation of certain mitogen- and antigeninduced lymphocyte activities.42 Effects on the developing endocrine system is a subject of high concern for chemicals, e.g., those with androgenic or estrogenic activity such as chlorinated environmental pollutants.Not much is known about metals in this respect. Benefits of breastfeeding Breastfeeding is the most ideal form of infant nutrition. Among the benefits are protection against infection and prevention of development of allergic disease.43 Recently it was reported that a certain component of milk, multimeric alfa-lactalbumin, is a potent apoptosis-inducing agent with a selective activity on transformed, embryonic and lymphoid cells, thereby possibly contributing to mucosal immunity.44 Several studies have shown that children who were breastfed gain higher scores on intelligence tests than those who were bottlefed.The suggested explanation is a higher content in breast milk of essential fatty acids, essential for brain development.45 However, the interpretation of such studies is complicated due to the current association between breastfeeding and higher social class.46 Thus, it was shown in a Danish study that 79% of the well educated women but only 29% of the less-educated women were still breastfeeding their infants six month after parturition. 47 Conclusions and recommendations From this work the following conclusions can be drawn.Neonates, infants and lactating women should be considered as unique groups for risk assessment. Exposure, kinetics and toxic effects should be characterized in the the various age groups. Qualitative as well as quantitative differences in response to a toxic metal should be studied at various stages of development. The great beneficial effects of breastfeeding and the influence of nutritional status on toxicity should also be taken into account. References 1 IPCS, Environmental Health Criteria 70, Principles for the Safety Assessment of Food Additives and Contaminants in Food, World Health Organization, Geneva, 1987. 2 IPCS, Environmental Health Criteria 59, Principles for Evaluating Health Risks From Chemicals During Infancy and Early Childhood: The Need for a Special Approach, World Health Organization, Geneva, 1986. 3 Similarities and Differences Between Children and Adults. Implications for Risk Assessment, ed. Guzelian, P. S., Henry, C. J., and Olin, S.S., ILSI Press, Washington, DC, 1992. 4 Palminger Hall�en, I., J�onsson, S., Karlsson, M. O., and Oskarsson, A., Toxicol. Appl. Pharmacol., 1996, 136, 342. 5 Greenwood, M. R., Clarkson, T. W., Doherty, R. A., Gates, A. H., Amin-Zaki, L., Elhassani, S., and Majeed, M. A., Environ. Res., 1978, 16, 48. 6 Sundberg, J., J�onsson, S., Karlsson, M. O., Palminger Hall�en, I., and Oskarsson, A., unpublished work. 7 Lactation: Physiology, Nutrition and Breast-feeding, ed.Neville, M. C., and Neifert, M. R., Plenum Press, New York, 1983. 8 Wilson, J. T., Drug Metab. Rev., 1983, 14, 619. 9 WHO Working Group in Drugs and Human Lactation, ed. Bennett, P. N., Elsevier, Amsterdam, 1988, pp. 27–48. 10 Palminger Hall�en, I., and Oskarsson, A., Biometals, 1995, 8, 231. 11 Petersson, K., and Oskarsson, A., unpublished work. 12 Sundberg, J., Ersson, B., and Oskarsson, A., unpublished work. 13 Casey, C. E., Smith, A., and Zhang, P., in Handbook of Milk Composition, ed.Jensen, R. G., Academic Press, San Diego, California, 1995, pp. 622–674. 14 Fransson, G.-B., and L�onnerdal, B., Ped. Res., 1983, 17, 912. 15 Palminger Hall�en, I., Norrgren, L., and Oskarsson, A., Arch. Toxicol., 1996, 70, 237. 16 Oskarsson, A., Palminger Hall�en, I., and Sundberg, J., Analyst, 1995, 120, 765. 17 Palminger Hall�en, I., Jorhem, L., Json Lagerkvist, B., and Oskarsson, A., Sci. Total Environ., 1995, 166, 149. 18 Oskarsson, A., Sch�utz, A., Skerfving, S., Palminger Hall�en, I., Ohlin, B., and Json Lagerkvist, B., Arch.Environ. Health, 1996, 51, 234. 19 WHO/FAO, Toxicological Evaluation of Certain Food Additives and Contaminants, World Health Organization Food Additives Series: 24, Cambridge University Press, 1989. 20 Grandjean P., Jørgensen, P. J., and Weihe, P., Environ. Health Perspect., 1994, 102, 74. 21 IPCS, Environmental Health Criteria, Copper, World Health Organization, Geneva, in the press. 22 Eklund, G., Helmersson, S., and Oskarsson, A., unpublished results. 23 Milsap R. L., Hill, M. R., and Szefler, S. J., in Applied Pharmacokinetics: Principles of Therapeutic Drug Monitoring, ed. Evans, W. E., Scentag, J. J., and Jusko, W. J., Applied Therapeutics, Vancouver, Washington, 3rd edn., 1992, pp. 10-1-32. 24 Davis, J. M., and Grant, L. D., in Similarities and Differences Between Children and Adults. Implications for Risk Assessment, ed. Guzelian, P. S., Henry, C. J., and Olin, S. S., ILSI Press, Washington, DC, 1992, pp. 150–162. 25 Henning, S. J., and Cooper, L. C., Proc. Soc. Exp. Biol. Med., 1988, 187, 110. 26 Kello, D., and Kostial, K., Environ. Res., 1973, 6, 355. 27 Palminger Hall�en, I., J�onsson, S., Karlsson, M. O., and Oskarsson, A., Toxicol. Appl. Pharmacol., 1996, 140, 13. 28 Department of the Environment Central Directorate on Environmental Pollution, The Glasgow Duplicate Diet Study (1979/80), Pollution Report No. 11, 1982. 29 Rowland, I. R., Robinson, R.D., Doherty, R. A., and Landry, T. D., in Reproductive and Developmental Toxicity of Metals, ed. Clarkson, T. W., Nordberg, G. F., and Sager, P. R., Plenum Press, New York, 1983, p. 745. 30 Momcilovic, B., and Kostial, K., Environ. Res., 1974, 8, 214. 22 Analyst, January 1998, Vol. 12331 Kostial, K., Kello, D., Jugo, S., Rabar, I., and Maljkovic, T., Environ. Health Perspect., 1978, 25, 81. 32 Wong, K. L., and Klaassen, C., Toxicol. Appl. Pharmacol., 1980, 53, 343. 33 Jugo, S., Environ. Res., 1980, 21, 336. 34 Davison, A. N., and Dobbing, J., in Applied Neurochemistry, ed. Davison, A. N., and Dobbing, J., Blackwell Scientific Publications, Oxford, 1968, pp. 253–282. 35 IPCS, Environmental Health Criteria 165, Lead, World Health Organization, Geneva, 1995. 36 Myers, G. J., Davidson, P. W., Cox, C., Shamlaye, C. F., Tanner, M. A., Marsh, D. O., Cernichiari, E., Lapham, L. W., Berlin, M., and Clarkson, T. W., Neurotoxicol., 1995, 16, 711. 37 Marsh, D. O., Turner, M. D., Crispin Smith, J., Allen, P., and Richdale, N., Neurotoxicol., 1995, 16, 717. 38 Grandjean, P., Weihe, P., White, R. F., Debes, F., Araki, S., Yokoyama, K., Murata, K., Sørensen, N., Dahl, R., Budtz- Jørgensen, E., and Jørgensen, P. J., Abstract presented at the International Conference on Human Health Effects of Mercury Exposure, T�orshavn, Faroe Islands, June 22–26, 1997. 39 Wong, K. L., and Klaassen, C. D., Toxicol. Appl. Pharmacol., 1982, 63, 330. 40 Smith, M. J., Pihl, R. O., and Farrell, B., Neurobehav. Toxicol. Teratol., 1985, 7, 19. 41 Andersson, H., Petersson-Graw�e, K., Lindqvist, E., Luthman, J., Oskarsson, A., and Olson, L., Neurotoxicol. Teratol., 1997, 19, 105. 42 Thuvander, A., Sundberg, J., and Oskarsson, A., Toxicology, 1997, 114, 163. 43 Forsyth, J. S., Proc. Nutr. Soc., 1995, 54, 407. 44 Håkansson, A., Zhivotovsky, B., Orrenius, S., Sabharwal, H., and Svanborg, C., Proc. Natl. Acad., Sci., 1995, 92, 8064. 45 Uauy, R., and Andraca, I., J. Nutr. 1995, 125, 2278S. 46 Gale, C. R., and Martyn, C. N., Lancet, 1996, 347, 1072. 47 Michaelsen, K. F., Larsen, P. S., Thomsen, B. L., and Samuelson, G., Acta Paed., 1994, 83, 565. Paper 7/05136K Received July 17, 1997 Accepted October 1, 1997 Anal