Chemical Contaminants in Human Milk:An OverviewBabasaheb R. SonawaneU. S. Environmental Protection Agency, Washington, DC

This review contains a succinct overview of the nature and extent of the problem of contamination of human milk with environmental and occupa-tional chemicals, excluding drugs. Factors influencing the levels of contaminants in breast milk are discussed. Also, data on major chemicals of con-cern with potential health risk(s) to the general population and risk-benefit considerations are dealt with briefly. Based on the available data on thesubject, research needs have been identified and policy recommendations are suggested. - Environ Health Perspect 103(Suppl 6):197-205 (1995)

Key words: breast feeding, pesticides, organohalogens, lipophilic chemicals, infants and children, Food and Drug Administration (FDA),organochlorine pesticides

Introduction

In recent years, there has been a renewedinterest in breast feeding; both medical andpsychological studies have emphasized theconsiderable benefits of breast feedinginfants. At present, about 65% of babies inthe United States are breast fed when theyare discharged from the hospital. TheCanadian and American academies of pedi-atrics have published a position paper urg-

ing a return to breast feeding as the bestnutrition for infants for the first 6 monthsof life (1).

With increased interest in breast feed-ing, there has been a parallel increase inconcern over the excretion of drugs andenvironmental chemicals into breast milkand contaminants found in human milk.The following brief discussion on milkconsumption and the diet of infants andchildren and the possible mechanisms ofexcretion of chemicals into breast milk isincluded to introduce the reader to thesignificance of the chemical excretion inhuman milk.

This paper was presented at the Symposium onPreventing Child Exposures to EnvironmentalHazards: Research and Policy Issues held 18-19March 1994 in Washington, DC. Manuscript received:December 5, 1994; accepted: May 15, 1995.

The views expressed in this paper are those of theauthor and do not necessarily reflect the views orpolicies of the U.S. Environmental Protection Agency.The U.S. Government has the right to retain a nonex-clusive, royalty-free license in and to any copyrightcovering this article.

Address correspondence to Dr. BabasahebSonawane, Office of Health and EnvironmentalAssessment, Research and Development, U.S.Environmental Protection Agency (8602), 401 MStreet, SW, Washington, DC 20460. Telephone (202)260-1495. Fax (202) 260-8719.

A number of studies demonstrate thatthe volume of milk intake among healthy,exclusively breast-fed infants ranges widely.After the first 4 to 5 months, the varianceis even greater. For infants who were breastfed for at least 12 months and given solidfoods beginning at 4 to 7 months, milkintake averaged 769 g/day (range,335-1144 g/day) at 6 months, 637 g/day(range, 205-1185 g/day) at 9 months, and445 g/day (range, 27-1154 g/day) at 12months (2).

Milk intake is most often determinedby weighing the infant before and afterfeeding. This method leads to underesti-mations of intake ranging from approxi-mately 1 to 5% (3) because of water lossthrough evaporation from the infantbetween weighings. Newer techniquesbased on stable isotopes have been devel-oped to measure breast-milk intake (4),but few data have been generated by thismethod to date. For a more detailed reviewof these data, see Nutrition DuringLactation (5).

Milk predominates in the diet of 1- to6-year-old children; the values for nonfatand fat milk solids are 30.4 and 13.4%,respectively. Any assessment of dietaryexposures of nursing infants is complex.Human milk is the major food and sourceof essential nutrients consumed by infantsduring their first year of life, but a vastrange of variables must be considered: theage at which supplementary foods areintroduced, the selection of foods given tothem, and the volume of human milk con-sumed. Factors greatly affecting the feedingpatterns of infants include economic status;ethnic background; and the mother'snutrition, age, marital status, educationallevel, parity, and employment.

Infant formula is the sole source offood for nonnursing infants for the first 3months of life. Milk or milk-based foodremains the predominant source of energyand nutrients for all infants throughouttheir first year of life. Averaged over thefirst 12 months, nonfat and fat milk solidsprovide 44.2 and 10.4% energy, respec-tively, of their diet. The diets of infantsand children are less diverse than those ofadults. Caloric consumption by infants perunit of body weight is approximately 25times higher for the very young infant (6)than that for adults. Therefore, comparingthe consumption data for infants andadults on the basis of grams per kilogramof body weight results in an elevated valuefor infants. It is thus important to monitorboth the percentage of total diet and themultiple of the national average consump-tion for each food and age group to iden-tify areas relative to dietary exposure topesticides and other environmental chemi-cals of concern.

The composition of human milk andits immunological properties providesignificant advantages to human milk asthe sole nutrient source in early infancy,even in industrially developing countries.However, in addition to the nutritionaland immunological benefits of humanmilk, pediatricians and scientists havebegun to recognize a wide range of chemi-cal contaminants found in human milkand their potential adverse health effects onchildren.

The most widely recognized group ofchemical contaminants are the fat-soluble,environmentally persistent organohalogencompounds such as dichlorodiphenyl-trichloroethane (DDT), polychlorinatedbiphenyls (PCBs), and dioxins. In

Environmental Health Perspectives 197

B. R. SONAWANE

addition, other contaminants such as heavymetals, pesticides other than DDT, andvarious organic solvents have been foundin human milk.

This brief overview attempts to focus onthe problem of environmental chemicalcontaminants (excluding drugs that arefound in human milk), time trends in lev-els, factors influencing the levels of contam-inants in breast milk, and associated adversehealth effects in breast-fed infants. Finally,some research gaps are identified, togetherwith recommendations for future work.

Recently, the American Academy ofPediatrics Committee on Drugs revisedthe list of agents (primarily drugs) trans-ferred into human milk and describedtheir possible known effects on the infantor on lactation (7). This review is not nec-essarily inclusive of all published literature,but it focuses on selected classes of envi-ronmental chemicals.

Pesticides in Human MilkIn general in the U.S. population, lowlevels of organochlorine pesticides arefound in human tissues. These pesticidesare fat soluble and bioaccumulate, andelimination from body fat stores is veryslow. Excretion of these compounds viahuman milk exposes breast-feeding infantsto a variety of organochlorine pesticides,especially in agricultural areas where theyare most often used. Occupational expo-sure to organochlorine pesticides mayoccur during manufacturing, distribution,use in agriculture, recreation gardening,etc. Ongoing pesticide surveillance and thefood residue level monitoring program bythe Food and Drug Administration (FDA)indicate that pesticide concentrations inhuman milk continue to decline over time;however, data are difficult to interpretbecause the primary intent of the studies isfor regulatory compliance and enforce-ment. Although data on pesticide residuesin food are extensive and infant formulasand processed baby foods are routinelymonitored to ascertain pesticide residuelevels, many uncertainties exist with regardto uniformity of sampling, analytical tech-niques, quality control, etc., thus raisingquestions about the usefulness of the datain actual risk estimations. In general, dueto the decreasing use of this class of pesti-cides in the United States and other coun-tries, declining concentrations in humanmilk have been observed. Potential adversehealth effects due to low-level exposure indeveloping infants and children areunknown.

As stated earlier, many of these lipid-soluble compounds bioaccumulate and arenot cleared rapidly. Human milk is oneroute of elimination for the mother's bodyburden, but unfortunately that route alsoincreases the exposure of infants (8). Therehave been many surveys of pesticides inhuman milk-some in response toepisodes of food or dairy product contami-nation. The data from these surveys havebeen used to compare pesticide concentra-tions in human milk to establish allowabledaily intakes.

In general, the more recent surveys ofpesticides in human milk have demon-strated that the concentrations are lowerthan those observed in previous surveys.Despite this decrease in concentrations,more effort is needed to characterize thepotential adverse effects of the low concen-trations of chlorinated pesticides found inhuman milk. Initial attempts at estimatingsuch effects associated with low concentra-tions have recently been described byMattison (9-11) and Rogan et al. (12).

DDT and Its MetabolitesDDT has a long history of use worldwideas an effective pesticide for controllingmosquitoes and other pests. Concern overreproductive effects of DDT and itsmetabolite dichlorodiphenyldichloroethyl-ene (DDE) in birds and its long biologicalpersistence led to the cessation ofDDT usein the United States approximately 20years ago. Despite the ban, this pesticideand its metabolites continue to be found inhuman milk in the United States atdecreasing concentrations over time,demonstrating the remarkable biologicalpersistence. The range of means of p~p-DDT and pp-DDE reported in surveys ofhuman milk in the United States before1986 varied from 0.2 to 4.3 ppm and 1.2to 14.7 ppm in milk fat, respectively. Themean concentrations among quantifiablesamples from the 1986 survey in Arkansaswere at the low end of these means.Among all samples, the mean concentra-tions were considerably lower than thosepreviously reported. The mean concentra-tion ofpp'-DDT in all samples assayed inArkansas was 0.039 ppm; the highestquantified level was 0.203 ppm (13).Among those with quantifiable concentra-tions, the mean was 0.954 ppm. Thisappears to show a continued decrease inDDT concentrations in human milk overthe years.

In developing countries DDT and itsmetabolites are often the most widespread

contaminants in human milk, found aspp'- DDT and p,p'-DDE. In general,DDT levels are relatively high in develop-ing countries where DDT still is, or untilrecently has been, used extensively in agri-culture and public health (14). Typicalaverage background levels are now around30 ppb total DDT in whole milk and 1ppm in milk fat. In developing countries,levels 10 to 100 times higher may still befound in some areas. There are obviouslylarge differences in levels in human milkbetween countries and regions of devel-oped countries; these levels are directlyrelated to the recent use of DDT. Womenimmigrants from developing countriesoften have far more DDT in their breastmilk than women in the local populationof developed countries (15).

Dieldrin, Aldrin, and EndrinThese are very persistent insecticides thathave been banned in the United States, butthey remain in use in some developingcountries (16). Dieldrin is a metabolite ofaldrin that persists in adipose tissue.Previous surveys conducted in the UnitedStates have identified detectable levels in0.04 to 100% of the human milk samplesanalyzed (8). Mean dieldrin concentrationsranged from 0.0.5 ppm to 0.24 ppm inmilk fat (8). The mean concentration in2% of Arkansas samples with quantifiablelevels was 0.071 ppm. Recently, high lev-els of dieldrin have been detected in milksamples from the Middle East, SouthAmerica, and Australia (17,18). Aldrinand endrin have been occasionallyreported in human milk.

LindaneLindane is a mixture of various isomers ofhexachlorocyclohexanes (HCH) and hasbeen used as a substitute for DDT. In pre-vious surveys conducted in the UnitedStates, HCH isomers were found inquantifiable concentrations in 4 to 68% ofthe human milk samples analyzed (8). Forexample, the f-HCH isomer was found in27% of the human milk samples testedfrom Arkansas women (13). Among thosewith quantifiable levels, the mean concen-tration was 0.12 ppm, and among all sam-ples, the concentration was 0.03 ppm.Lindane's agricultural uses in the UnitedStates have been virtually eliminated bychanges in regulations over the past 20years. The recent levels of HCH isomersreported for most European countries aregenerally low (average 0.2 ppm in fat)compared with those reported from Asia,

Environmental Health Perspectives198

CHEMICALS INHUMANMILK

especially India and the People's Republicof China (average 6 ppm in fat).

HexachlorobenzeneHexachlorobenzene (HCB) is a persistentchemical with a variety of sources, includ-ing its previous use as a pesticide and itspresence as an impurity in several otherpesticide formulations. This compoundcan disrupt porphyrin metabolism (8).Fatal cases of infant poisoning from inges-tion of highly contaminated human milkhave been reported. Because of persistenceand solubility, hexachlorobenzene has beendetected in many surveys of adipose tissuesin the United States; few studies haveexplored the presence of this chemical inhuman milk. Among previous surveys con-ducted in the United States, the mean con-centration was 0.04 ppm in milk fat(range, 0.018-0.063). Among the 6% withquantifiable levels in Arkansas, the meanwas 0.03 ppm; and among all samples, themean concentration was 0.002 ppm; bothare lower than in earlier reports (8).Hexachlorobenzene is no longer registeredfor agricultural use, and its occurrence as aformulation impurity in other registeredproducts has been greatly curtailed.

In the late 1950s, about 500 peoplewere fatally mass intoxicated in Turkey andabout 4000 became sick from eating breadsmade from HCB-treated wheat. HCB con-tamination caused skin lesions due toaltered porphyrin metabolism. Childrenunder 2 years of age who were breast fed bymothers exposed to HCB died of the con-dition known as "pembe yara" (pink sore).Elevated levels of HCB in human milkwere still observed in the area 20 to 30 yearsafter the accident (19,20).

Cyclodiene PesticidesHeptachlor, chlordane, and their metabo-lites (heptachlor epoxide, oxychlordane,trans-nonachlor) are closely related cyclo-diene pesticides. Surveys conducted in theUnited States have demonstrated thatbetween 25 and 100% of human milksamples analyzed had quantifiable concen-trations of heptachlor or heptachlor epox-ide ranging from 0.035 to 0.13 ppm (8). Asomewhat greater proportion of samples(46-100%) had quantifiable concentra-tions of chlordane and oxychlordane(range, 0.05-0.12 ppm), perhaps reflectingfrequent use as a termiticide in houses(16). Among samples surveyed in Arkansas,5% had quantifiable concentrations of hep-tachlor (mean, 0.03 ppm) and 74% hadquantifiable concentrations of heptachlor

epoxide (mean, 0.06 ppm) (13). Two per-cent of the samples in that study containedquantifiable concentrations of chlordane;77 and 84% had quantifiable concentra-tions of trans-nonachlor and oxychlordane,respectively. The mean concentrationsamong quantifiable samples measured inArkansas for trans-chlordane, cis-chlordane,and oxychlordane were 0.18, 0.15, and0.06 ppm in milk fat, respectively. In moststudies, heptachlor and heptachlor epoxidewere also detected in human adipose tissuesamples (21,22).

One study of approximately 1500women (23) explored regional differencesin the pesticide content of human milk. Inthe southeast region of the United States,including Arkansas, 76% of the samplestested had detectable levels of heptachlorepoxide. The distribution of heptachlorepoxide concentrations in samples was alsohigher in the southeast region. Only 23%of the samples tested contained trace orundetectable concentrations. Half of thesamples (52%) had heptachlor epoxideconcentrations ranging from 0.001 ppm to0.1 ppm. The remainder (approximately25%) contained concentrations above 0.1ppm. This was the highest concentrationamong all regions in the United States.The mean concentration of heptachlorepoxide in these samples with detectablelevels was 0.128 ± 0.209 ppm, which alsowas the highest mean level for all theregions surveyed in the United States.

Similar studies of human milk inPennsylvania (24) and in Missouri (25)have demonstrated mean heptachlor epox-ide concentrations of 0.16 and 0.0027ppm, respectively. Studies conducted inHawaii (26,27), where inhabitants havealso been exposed to heptachlor and hep-tachlor epoxide in dairy products, havedemonstrated levels ranging from 0.001 to0.067 ppm (mean 0.036 ppm) amongwomen on Oahu, and from 0.0 15 to 0.052ppm (mean 0.031 ppm) among women onneighboring islands.

With the exception of endosulfan,which does not exhibit the persistence andbioconcentration characteristics of otherchemicals in the group, virtually all agricul-tural uses of the cyclodiene pesticides havebeen eliminated or greatly restricted byregulatory actions over the past 20 years.

Other PersistentOrganohalogensDioxins and DibenzofiuansChlorinated dioxins and dibenzofurans are

highly toxic chemicals derived from com-mercial sources such as 2,4,5-T and pen-tachlorophenol found in hazardous wastesites and bleaching paper or generated dur-ing industrial processes (e.g., incinerationof municipal waste and combustion ofleaded gasoline). The main source of expo-sure to these compounds is food, especiallymeat, dairy products, and fish (oil).Inhalation is only a minor source of expo-sure. Dioxins have been detected in humanmilk at relatively high concentrations(28-31). Dioxins are very lipophilic andare mainly stored in adipose tissue(32-34). The concentration of dioxins inbreast milk decreases with duration of lac-tation and with the number of breast-fedchildren (30). A recent study indicates thatdioxin exposure of the breast-fed infant isvery high and exceeds the accepted.dailyintake more than 20-fold at the age of 4weeks (31). It was estimated that dailyintake from milk of a cow grazing in theneighborhood of an incinerator was about200 times the intake by inhalation of theambient air (35). The bioavailability ofdioxins and dibenzofurans from breastmilk is high, and no obvious change infecal excretion of dioxins with age wasfound. Furthermore, no effects of dioxinexposure could be observed on physicaland neurological development of thebreast-fed infant. However, several clinicallaboratory parameters were affected ondioxin exposure.

Polychlorinated BiphenylsPolychlorinated biphenyls are produced astechnical mixtures with different degrees ofchlorination-usually with a chlorine con-tent between 40 and 60% (e.g., Aroclor1242 and 1260).

In 1968, a serious mass intoxicationoccurred in Japan from a large-scale PCBcontamination of rice-bran oil due to aleak in a heat transfer installation. Morethan 1700 people became ill and about 20died. The main symptoms of this so-called"Yusho disease" were severe dermatologicalabnormalities, including chloracne (36).Infants born to women in Yusho hadabnormally dark brown skin and otherabnormalities. The average PCB concen-tration in whole blood from Yusho patientsat the time of the incident was about 60ppm; the concentration of PCBs in Yushomilk fat may have been more than 10 ppm(37,38).A similar mass poisoning, called Yu-

Chen, occurred in Taiwan in 1979, withmore than 2000 identified victims. It is

Volume 103, Supplement 6, September 1995 199

B. R. SONAWANE

now accepted that polychlorinated diben-zofurans, which were present as impuritiesin the used PCB liquid, were the majoretiologic factor (39).

PCBs are widely distributed in humanmilk from industrialized countries, butPCB levels are mostly below the detectionlimit in milk from Third World countries.The average concentrations of total PCBsin human milk fat are typically between0.5 and 2 ppm, depending on the place ofsampling and the analytical methods used.

In a Canadian study, the highest levelof PCBs (4.3 ppm) was found in milk fatfrom a mother who had lived in an indus-trialized area close to a municipal incinera-tor for 5 years (40). Occasionally, extremelyhigh levels of PCBs (< 10 ppm) in humanmilk fat have been found in lactatingwomen living at farms where the silos havebeen treated with PCB-containing paints(41). In many industrialized countries,women immigrants from less-developedcountries often have far more DDT andless PCBs in their breast milk than nativecitizens. Of the 209 PCB congeners, abouta dozen can be separated in human milk bythe most common analytical methods. Ingeneral, relatively more of the higher chlo-rinated congeners are found in humanmilk. However, the pattern of chlorinesubstitution is also important.

At present, there is no universallyaccepted procedure for determining thetotal PCB content of human tissue sam-ples. In fact, the average background levelsof PCBs in human milk do not differ verymuch between industrialized countries ifthe quantitation procedures are nearly thesame. Typical average levels are 0.5 to 1.5ppm PCBs in extractable fat. The difficul-ties in investigating trends are illustrated byRogan et al. (12,43). They found some-what higher PCB levels in milk fat thanthose previously reported in the UnitedStates by Schwarz et al. (44), and these dif-ferences may be explained by analyticalapproaches.

Polychorinated TerphenylsThe chemical and biological properties ofpolychlorinated terphenyls (PCTs) areclosely related to those of PCBs. PCTshave only been found in human milk sam-ples from Japan (45). The average level ofPCTs in human milk fat was about 0.02ppm; this corresponds to 1/60 of that ofPCBs in the same sample. In contrast tothe situation with PCBs, the levels of PCTsin milk fat were significantly lower thanthose found in adipose tissue. PCTs were

not detected in human milk from Canada(46). In Europe, PCTs have not beeninvestigated in human milk, but PCTswere found some years ago in two adiposetissue samples (0.5-0.8 ppm) from theNetherlands (47).

Polybrominated BiphenylsThe hazards of polybrominated biphenyls(PBBs) were discovered after the pollutionepisode in Michigan in 1973. Animal feedswere accidentally contaminated with FireMaster BP-6-a mixture of PBBs mainlyconsisting of hexabromobiphenyls, whichwere normally used as a flame retardant inpolymers. Later, PBBs were detectedwidely in domestic animals, foodstuffs, andhuman tissues. Although no acute orchronic effects of PBBs have beenidentified in humans, the possibility oflong-term effects cannot be ruled out (48).

Most of Michigan's inhabitantsreceived measurable quantities of PBBs intheir body tissues. In 1976, 96% of 53samples and 43% of 12 samples of humanmilk from two areas of Michigan contained0.01 to 1.2 ppm PBBs on a fat basis. Themedian value was 0.068 ppm (49). In alarger investigation of 2986 breast milksamples obtained between May 1976 andDecember 1978 from all of Michigan,PBBs were detectable in 88% of the sam-ples. The maximum level in milk fat was 2ppm, and the mean and median were 0.1ppm and 0.06 ppm, respectively (50). Inbreast milk fat from 32 directly exposedwomen farmers, a mean value of 3.6 ppmPBBs, with a maximum of 92 ppm, wasdetected. The mean ratio of milk to serumPBB values in 21 individual women show-ing detectable levels was 122:1, while themean ratio of adipose fat to serum PPB was362:1 (51,52). Weil et al. (53) reportedthat 42% of the PBB-exposed womenbreast fed their children compared with85% of controls. Furthermore, the dura-tion of breast feeding was longer amongthe controls (mean, 29.6 weeks) thanamong the PBB-exposed women (14.8weeks). Some preliminary results suggestthe existence of an inverse relationshipbetween body levels of PBBs and somedevelopmental effects in 2- to 4-year-oldchildren (54). In May 1974, a PBB toler-ance of 1 ppm in human milk fat wasestablished, but in November of the sameyear it was reduced to 0.3 ppm (55).

A more recent study on lactatingwomen concludes that PBBs are very per-sistent, and no significant decrease hasbeen found in the levels in the population

during a decade; it was estimated that 47%of all breast milk samples would still havedetectable levels of PBBs by the year 2000(50). PBBs have not been reported inhuman milk from outside Michigan, e.g.,in a survey in 1977 to 1978 in Alberta,Canada (56).

Other OrganohalogensChlorobenzenesSeveral chlorobenzenes were detected inhuman milk (57). The most abundant wasp-dichlorobenzene (58). The sources ofthese chlorobenzenes have not beenidentified, but it is known that o-dichloro-benzene is used as a bactericide, and p-dichlorobenzene is used in mothballs and asa deodorant in toilets. Other chloroben-zenes are chemical intermediates or conta-minants in commercial chemicals, e.g.,pentachlorobenzene is an impurity in and adegradation product of hexachlorobenzene,indicating a potential for a wider occurrencein human milk.

PentachiorophenolIn recent investigations, trace amounts ofpentachlorophenol (PCP) were detected inhuman milk. Free PCP has also beenfound in human adipose tissue at levelsfrom 4 to 250 ppb, together with itspalmitic acid ester (59). This ester mayalso be present in human milk, butthis has never been investigated.Pentachlorophenol is an important woodpreservative in many countries.

MMxxMirex is a persistent, fully chlorinated,cyclic hydrocarbon. Until 1978, it waswidely used as a pesticide for fire ant con-trol in the southeastern United States andalso as a flame retardant. The use of mirexas a pesticide caused contamination of cat-tle milk (60), and considerable mirex pol-lution was discovered some years ago inLake Ontario (61). A few investigationshave indicated the occurrence of traces ofmirex in human milk from North America(46,62).

ToxheneThe pesticide toxaphene is a mixture ofpolychlorinated terpenes. It is a globalpollutant spread by long-range air trans-portation. Although the pesticide hashardly been used, it was detected inwildlife and in two pooled human milksamples from Sweden. The toxaphene levelcalculated was 0.1 ppm in milk fat (63).

Environmental Health Perspectives200

CHEMICALS INHUMANMILK

ChloroethersBis(2,3,3,3-tetrachloropropyl)ether, whichis used in Japan as a synergist in pyrethruminsecticides for mosquitoes, has beendetected in human milk (64).

Polychiorinated NaphthalenesPolychlorinated naphthalenes (PCNs) areused as insulating materials for cables andin lubricants. In a recent investigationPCNs were detected in human milk fromLos Angeles and Sweden. The concentra-tions were between 1.7 and 3 ppb PCNs inmilk fat (65).

Nonhalogenated Organic CompoundsMost nonhalogenated organic chemicalsare not very persistent, either in the envi-ronment or in the human body. Thus,detectable levels of such substances inhuman milk are usually found when theexposure is high and long-term in nature,as in the occupational environment.

OrganophosphatesIn Taiwan, extremely high levels ofmalathion were found in human blood andbreast milk samples from 1974 to 1975.The average malathion level in 12 milksamples was 1.88 ppm in whole milk (66).Furthermore, 0.1 ppm of anotherorganophosphate pesticide, dimethyl-dichlor-vinyl phosphate (DDVP), wasfound in a single sample. In nine humanmilk samples from women living in theSanta Clara Valley of California, an arearepeatedly sprayed with malathion, nomalathion could be detected in any of themilk samples. The detection limit was < 5ppb in whole milk (67).

Polycyclic Aromatic HydrocarbonsPolycyclic aromatic hydrocarbons (PAHs)are emitted on heating or burning oforganic matter, including food. In 10German breast milk samples, 14 PAHcomponents were detected, including thecarcinogens benz[a] anthracene andbenzo[a]pyrene. The average total level was0.1 ppb in whole milk (68).

NitrosanunesIn a recent investigation of 51 milk sam-ples from 13 nursing mothers in theUnited States, 16 samples contained mea-surable levels (0. 1-1.1 ppb) of N-nitroso-dimethylamine. In certain individuals,eating a meal of bacon and a vegetable highin nitrate occasionally resulted in higherlevels of nitrosoamine in their milk (69).

NicotineRecently it was discovered that passivesmoking may result in measurable nicotineconcentrations (mean about 12 ppb) inbreast milk (70).

Heavy MetalsBreast milk normally contains trace levelsof most metals and other elements. Bothinorganic and organic compounds of themetals are found in human milk, but notassociated with milk fat. The processes bywhich the metals are excreted through themammary glands are not fully known, butthey are probably different from those forthe lipophilic organohalogens.

In risk evaluations, it has to be borne inmind that the absorption of heavy metalsin infants is generally higher when they areon a milk diet, probably due to binding toreadily absorbed milk proteins.

LeadAt present, average background levels oflead in human milk from industrializedcountries are probably between 5 and 20ppb. In heavily polluted areas they may beup to 20 times higher (71). Directly toxiclevels have been seen in occupational set-tings or in lead poisoning cases (72). Leadin milk is better absorbed into the bodythan lead present in other dietary compo-nents. Lead levels in breast milk are nor-mally lower than lead levels in milk-basedinfant formulas (73). Lamm and Rosen(74) noted higher blood lead levels in for-mula-fed infants than in breast-fed infants.

CadmiumFewer investigations have been carried outon cadmium in human milk, and thereported levels vary widely. Average back-ground levels seem to be <2 ppb (jig/l),although levels 10 times higher are oftenquoted. Cadmium levels in human milkcorrespond to those in cow's milk (75).Compared to other foods, cadmium inmilk seems to be more bioavailable and itconcentrates more in the kidneys (76). Incows, cadmium in the feed has been shownto decrease milk production (72).

MercuryThe levels of mercury in human milk aretypically lower than those of lead and cad-mium. Average background levels are < 1ppb (jig/l). The highest levels weredetected in milk from fish eaters, andabout 20% of the total mercury contentwas in the form of the more toxic methylmercury (77).

During pollution episodes in Japan,such as the outbreak of Minamata disease,mercury levels of about 50 ppb in humanmilk were found (78). With the tragicmercury poisoning in Iraq in 1972 due toingestion of homemade bread preparedfrom wheat treated with a methylmercuryfungicide, the predominant route of infantintake of mercury was via breast milk. Themercury content in human milk wasreported to be up to around 200 ppb, andabout 60% was methyl mercury (79).

Other Metals and Trace ElementsNumerous studies on the presence of met-als and trace elements in human milk havebeen published. Most are difficult to com-pare because of the use of different analyti-cal techniques and the lack of adequatequality assurance (80). However, for sev-eral of these trace elements, the data base isvery limited to ascertain risk to infants.Occupational exposure to these elementsmay result in much higher and even toxiclevels in breast milk.

Considerations of SpecificSubpopulationsCertain populations of infants and childrenmay be more sensitive to the effects ofchemical exposure through breast feedingbecause of physiological and biochemicalfactors such as genetic predisposition,chronic medical conditions, interactions ofchemicals with medications, nutritionalstatus of mother, and general health status.Other factors that may make certaininfants and children more susceptibleinclude increased exposure through farmwork or parental occupational exposureand low socioeconomic status.

Children living in poverty may consti-tute an additional sensitive population.One factor contributing to compromisedhealth status is poor nutrition. Preschoolchildren of low socioeconomic status havebeen found to have lower dietary intakes,lower biochemical indices, and smallerphysical size for their age than children ofhigher socioeconomic status. Because ofcompromised health status, infants andchildren of low socioeconomic status areprobably more susceptible to any toxicinsult, including chemical exposure.Infants and children of poor families aremore likely to live in highly pollutedneighborhoods and thus have greater expo-sure to environmental toxicants. Therefore,one might expect that adverse effects ofchemicals, whether acute or chronic, mightbe magnified in this subpopulation.

Volume 103, Supplement 6, September 1995 201

B. R. SONAWANE

Uncertainties in Risks toInfants from Chemicalsin Human MilkBreast feeding has substantial benefit,including psychological, immunological,and general health promotion. In many

countries, breast feeding confers measur-

able benefits such as decreased rates ofinfectious disease and increased rates ofgrowth and development. Despite the con-

centrations of chemicals found in humanmilk, no major studies have demonstratedthat these chemical concentrations have ledto adverse health outcomes in the childrenexposed through breast feeding, with some

exceptions as pointed out earlier.Therefore, although there is concern thatexposure to chemicals in human milk may

carry some potential for adverse healtheffects for the mother and for the nursinginfant, it is important to recognize the

benefits of breast feeding. Furthermore,surveys conducted in the United Statesover the past four decades have shown thatthe number of samples with detectablechemical concentrations has been falling-even when improved analytical methodswith increased sensitivity have been used.

The FDA monitors certain pesticideresidues in all food other than meat, milk,and eggs. The FDA's monitoring program

is not designed to determine dietary expo-

sure to all pesticides or to other chemicalsof concern; rather, its objectives are to

enforce compliance with U.S. EPAtolerance levels.

It is also important to recognize thewide variation in sampling and analyticalmethods, size, overall design, and objectivesof existing residue testing programs. Themost comprehensive and best recorded dataon pesticides are those collected throughFDA's market basket sampling survey andanalysis, but no single data bank providesideal residue values for other toxic chemi-cals. Chemical residue analyses are complex,difficult to perform, and expensive. All datashould be judged within this perspective.Methods for uniform sampling, collection,and reporting to reflect the adequacy of thedata's quality are needed.

Fat-soluble chemicals, such as chlori-nated halogens, heavy metals, and organicsolvents, are excreted in human milk. It isevident that human milk is contaminatedwith organic and inorganic chemicals inthe environment, and exposure can occur

to breast-feeding infants through mother'smilk and other food sources. However, no

overt clinical illness has occurred in infants

from such exposures except in cases of acci-dental poisoning outbreaks. It is not clearwhether overall in utero or lactational expo-sure of the mother has an impact on thebody burden of infants. Furthermore, thecontribution of chemical contamination inhuman milk, compared with dietary intakeother than through mother's milk, and itspotential impact on adverse toxicologicaleffects are unknown, and no adequate eval-uation has been conducted for most chemi-cals. Current risk assessment methodsgenerally do not make specific allowancesfor chemical exposures via mother's milk toinfants and children. Species conversion ofdietary intakes per unit of body weight isnormally based on adult body weights andfood consumption data.

Lactating women may be exposed tolipophilic chemicals from various sources,including air, food, water, cosmetics, andoccupational and household environments;and they may have substantial stores ofsome chemicals that can be mobilized dur-ing lactation. Within a few months ofbirth, the concentration of organochlorinecompounds in the body fat of the breast-fed infant may be equivalent to that in thefat of the lactating mother. Although theperiod of breast feeding is short comparedto the total life span, the lipophilic chemi-cals accumulated in the body during theseearly months may be retained for years.

Another source of exposure to chemicalsis drinking water and water used for mixingmilk formulas. Although water intake isconsidered, neither nondietary exposuresnor exposures in drinking water are consid-ered in deriving risk estimates for totalchemical exposure in infants' milk. Becauseof this limitation, total exposures of infantsand children may be underestimated.

ConclusionsThe evaluation of potential risks to infantsand children due to chemical residues pre-sent in human milk and diet requires con-sideration of a number of factors. Inparticular, the level of risk depends on eachindividual's food consumption patterns,nature and levels of chemical residues inmilk and other food as consumed by lactat-ing mothers and breast-fed infants andchildren, and their toxicological potency. Acomprehensive analysis of the potentialhealth risks to infants and childrenexposed to chemicals in their milk anddiets requires consideration of all these fac-tors as well as any unique characteristics ofinfants and children relative to adults. Inaddition, it is known that socioeconomic,

nutritional, and health status influence thevulnerability of human infants and chil-dren to environmental toxicants andshould be accounted for in an estimationof health risks.

Infants and children may exhibitunique susceptibility to the toxic effects ofchemicals because they are undergoingrapid tissue growth and development, butempirical evidence to support this is mixed.Infants and children also consume muchgreater quantities of milk fat and certainfoods than do adults on a body weightbasis and thus may be subjected to higherlevels of exposure to certain chemicals thanadults. These exposures occurring earlier inlife can predispose infants and children to agreater or a lower risk of chronic toxiceffects than exposures occurring later in life.Therefore, traditional approaches to toxico-logical risk assessment may not always ade-quately protect infants and children.

Although current uncertainty factorsused to extrapolate toxicological data tohumans account for 10-fold variationsbetween species and within the humanpopulation, additional protection may berequired as justified, depending on thetoxicant of interest and the amount ofdietary residue monitoring and testing thathas been conducted. It must be recognizedthat there exist only limited data on theresidue levels of chemicals in milk andfood consumption patterns of infants andchildren that are appropriate for use in riskassessment.

Recommendations* Additional data are needed on the

chemical contaminants in mother'smilk and other foods and food con-sumption patterns of infants and chil-dren. Current data indicate that infantsand children consume significantlygreater amounts of milk and certainfoods on a body weight basis thanadults do. Because such higher expo-sures can lead to higher risks, it isimportant to have accurate data on thechemical contaminants in the milk andfood consumption patterns of infantsand children. The available data arebased on relatively small samples andmay not reflect current trends in foodconsumption by infants and children.A centralized data base for chemicalresidue data and standardized analyticalprocedures may be needed to improvethe ability to characterize chemicalexposures.

Environmental Health Perspectives202

CHEMICALS INHUMAN MILK

* All sources of exposure to chemicals-dietary and nondietary-need to beconsidered when evaluating the poten-tial risks to infants and children. Thetotal intake from all sources of foods onwhich residue may be present should becalculated when estimating exposure ofinfants and children. Chemicals alsomay be present in drinking water andwater used in preparing milk formula

due to contamination of sources ofwater.

* Physiological and biochemical charac-teristics of infants and children thatinfluence pharmacokinetics of xenobi-otics need to be considered in riskassessment. Physiological parameterssuch as tissue growth rates and bio-chemical parameters such as enzymeinduction may affect the response of

infants and children to chemicalresidues in milk, milk products, andother foods. Pharmacokinetic modelsthat provide for the unique physiologiccharacteristics of infants and childrenshould be developed.

* There is a need for validated animalmodels for predicting the risks of chem-ical contaminants in milk to infantsand children.

REFERENCES

1. Nutrition Committee of the Canadian Pediatric Society. Breastfeeding. Pediatrics 62:511-601 (1978).

2. Dewey KG, Heining MJ, Nommsen LA, Lonnerdal B. Lowenergy intakes and growth velocities of breast-fed infants: Arethere functional consequence? In: Activity, Energy Expenditure,and Energy Requirements of Infants and Children (Schurch B,Scrimshaw N, eds). Lausanne, Switzerland:InternationalDietary Energy Consulting Group-Nestle Foundation,;35-44.

3. Woolridge MW, Butte N, Dewey KG, Ferris AM, Garza C,Kelley RP. Methods for measurement of milk volume intake ofthe breast-fed infants. In: Human Lactation: Milk Componentsand Methodologies (Jensen RG, Neville MC, eds). New York:Plenum Press, 1985;5-21.

4. Butte NF, Wong WW, Patterson BW, Garza C, Klein PD.Human-milk intake measured by administration of deuteriumoxide to the mother: a comparison with the test-weighing tech-nique. Am J Clin Nutr 47:815-821 (1988).

5. NRC. Nutrition During Lactation. Washington:NationalAcademy Press, 1991.

6. NRC. Recommended Dietary Allowances, 10th ed.Washington:National Academy Press, 1989.

7. American Academy of Pediatrics. The transfer of drugs andother chemicals into human milk. Pediatrics 1:137-150(1994).

8. Jensen AA, Slorach SA. Chemical Contaminants in HumanMilk. Boca Raton, FL:CRC Press, 1991.

9. Mattison DR. Pesticides in Human Breast Milk: Population-Based Estimates of Excess Lifetime Cancer Risk. Society forRisk Analysis, Annual Meeting, 7-10 October 1990, NewOrleans.

10. Mattison DR. Estimate of Cancer Risk from HalogenatedPesticides in Human Breast Milk: Arkansas 1986. Society ofRisk Analysis, Annual Meeting, 7-10 October 1990, NewOrleans.

11. Mattison DR. Pesticides in Human Breast Milk: ChangingPatterns and Estimates of Cancer Risk. Washington:Society forGynecologic Investigation, 1991.

12. Rogan WJ, Blanton PJ, Portier CJ, Stallard E. Should the pres-ence of carcinogens in breast milk discourage breast feeding?Regul Toxicol Pharmacol 13:228-240 (1991).

13. Mattison DR, Wohleb J, To T, Lamb YL, Faitak S, BrewsterMA, Walls RC, Selevan SG. Pesticide concentrations inArkansas breast milk. J Ark Med Soc 88:553-557 (1992).

14. WHO. Environmental Health Criteria 38, Heptachlor.Geneva:World Health Organization, 1984.

15. Skaare JU, Tuveng JM, Sande HA. Organochlorine pesticidesand polychlorinated biphenyls in maternal adipose tissue,blood, milk, and cord blood from mothers and their infants liv-ing in Norway. Arch Environ Contam Toxicol 17:55-71(1988).

16. Baker SR, Wilkinson CF, eds. The Effects of Pesticides onHuman Health. Advances in Modern EnvironmentalToxicology Series, Vol XVIII. Princeton, NJ:PrincetonScientific Publishing, 1990.

17. Stacey CI, Perriman WS, Whitney S. Organochlorine pesticideresidue levels in human milk: Western Australia, 1979-1980.Arch Environ Health 40:102-108 (1985).

18. Stacey CI, Tatum T. House treatment with organochlorinepesticides and their levels in human milk: Perth, WesternAustralia. Bull Environ Contam Toxicol 35:202-208 (1985).

19. Peters HA, Gocmen A, Cripps DJ, Bryan GT, Dogramaci I.Epidemiology of hexachlorobenzene-induced porphyria inTurkey. Arch Neurol 39:744 (1982).

20. Cripps DJ, Peters HA, Gocmen A, Dogramaci I. Porphyria tur-cica due to hexachlorobenzene: a 20-30 year follow-up studyon 204 patients. Br J Dermatol 111:413 (1984).

21. Curley AV, Burse W, Jennings RW, Villanuvea EC, TomatiaL, Akazaki K. Chlorinated hydrocarbon pesticides and relatedcompounds in adipose tissue from people of Japan. Nature242:338-340 (1973).

22. Mussalo-Ranhamaa H, Pyysalo H, Miolanen R. Influence ofdiet and other factors on the levels of organochlorine com-pounds in human adipose tissue in Finland. J Toxicol EnvironHealth 13:689-704 (1984).

23. Savage EP, Keefe TJ, Tessari JD, Wheeler HW, ApplehansFM, Goes EA, Ford SA. National study of chlorinated hydro-carbon insecticide residues in human milk. Am J Epidemiol113:413-422 (1981).

24. Kroger M. Insecticide residues in human milk. J Pediatr80:401-405 (1972).

25. Jonsson V, Liu GJK, Liu J, Armbruster LL, Kettelhut LL,Drucker B. Chlorohydrocarbon pesticide residues in humanmilk in greater St. Louis, Missouri, 1977. Am J Clin Nutr30:1106-1109 (1977).

26. Takahashi W, Saidin D, Takei G, Wong L. Organochlorinepesticide residues in human milk in Hawaii, 1979-1980. BullEnviron Contain Toxicol 27:506-511 (1981).

27. Takahashi W, Parks LH. Organochlorine pesticide residues inhuman tissues, Hawaii, 1968-1980. Hawaii Med J41:250-251 (1982).

28. Rappe C, Bergquist P-A, Hansson M, Kjeller L-O, LindstromG, Marklund S, Nygren M. Chemistry and analysis of poly-chlorinated dioxins and dibenzofurans in biological samples.In: Biological Mechanisms of Dioxin Action, Banbury Report18. Cold Spring Harbor, NY:Cold Spring Harbor LaboratoryPress, 1984;17-25.

29. Jensen AA. Polychloribiphenyls (PCB), polychlorodibenzo--dioxins (PCDDs), and polychlorodibenzofurans (PCDFs) inhuman milk, blood and adipose tissue. Sci Total Environ64:259-293 (1987).

30. Beck H, Ekhart K, Mathar W, Wittkowski R. PCDD andPCDF body burden from food intake in the Federal Republicof Germany. Chemosphere 18:417-424 (1989).

31. Pluim HJJ. Dioxins. Pre- and Postnatal Exposure in HumanNewborn. Ph.D. Thesis, University of Amsterdam,Amsterdam, The Netherlands, 1993.

32. Beck H, Drofg A, Mathar W. 11th International Symposiumon Chlorinated Dioxins and Related Compounds (Dioxin '91),September 1991, Research Triangle Park, NC.

Volume 103, Supplement 6, September 1995 203

B. R. SONAWANE

33. Graham M, Hileman F, Kirk D, Wending J, Wilson J.Background human exposure to 2,3,7,8-TCDD. Chemosphere14:925-928 (1985).

34. Schecter A, Purst P, Kruger C, Meemken HA, Groebel W,Constable JD. Levels of polychlorinated dibenzofurans, diben-zodioxins, PCBs, DDT, and DDE, hexachlorobenzene, diel-drin, hexachlorocyclohexanes and oxychlordane in humanbreast milk from the United States, Thailand, Vietnam andGermany. Chemosphere 18:445-454 (1989).

35. Connect P, Webster T. An estimation of the relative humanexposure to 2,3,7,8-TCDD emissions via inhalation and inges-tion of cow's milk. Chemosphere 16:2079-2084 (1987).

36. Kuratsune M. Yusho, with reference to Yu-Cheng. In:Halogenated Biphenyls, Terphenyls, Naphthalenes,Dibenzodioxins, and Related Products, 2nd fully revised ed.Topics in Environmental Health, Vol. 4 (Kimbrough RD,Jensen AA, ed). Amsterdam:Elsevier, 1989;381-400.

37. Funatsu I, Yamashita F, Ito Y, Tsugawa S, Funatsu T,Yoshikane T, Hayashi M. Polychlorobiphenyls (PCB) inducedfetopathy. I. Clinical observation. Kurume Med J 19:43-51(1972).

38. Harada M. Intrauterine poisoning: clinical and epidemiologicalstudies of the problem. Bull Inst Const Med, Kumamoto Univ25:1-60 (1976).

39. Hsu S-T, Ma C-L, Hsu S-K, Wu S-S, Hsu N-H, Yeh C, WuS-B. Discovery and epidemiology of PCB poisoning in Taiwan:a four-year follow-up. Environ Health Perspect 59:5-10(1985).

40. Dillon JC, Martin GB, O'Brien HT. Pesticide residues inhuman milk. Food Cosmet Toxicol 19:437-442 (1981).

41. Humphrey HEB. Population studies of PCBs in Michigan resi-dents. In: PCBs: Human and Environmental Hazards (D'ItriFM, Kamrin MA, eds). Lansing MI:Buttersworth,1983;299-310.

42. Rogan WJ, Gladen BC, Jung K-L, Koong S-L, Shih L-Y,Taylor JS, Wu Y-C, Yang D, Ragan NB, Hsu C-C. Congenitalpoisoning by polychlorinated biphenyls and their contaminantsin Taiwan. Science 241:334-336 (1988).

43. Rogan WJ, Gladen BC, McKinney JD, Carreras N, Hardy P,Thullen J, Tingelstad J, Tully M. Polychlorinated biphenyls(PCBs) and dichlorodiphenyl dichloroethane (DDE) in humanmilk: effect of maternal factors and previous lactation. Am JPublic Health 76:172-177 (1986).

44. Schwartz DA, Newsum LA, Heifetz RM. Parental occupationand birth outcome in an agricultural community. Scand JWork Environ Health 12:51-54 (1986).

45. Minagawa J, Takigawa Y, Sakai H, Sasagawa I. Polychlorinatedterphenyls in human fat and mother's milk. Nippon EiseigakuZasshi 28:543 (1974).

46. Mes J, Davis DJ. Presence of polychlorinated biphenyl andorganochlorine pesticide residues and the absence of polychlori-nated terphenyls in Canadian human milk samples. BullEnviron Contain Toxicol 21:381-387 (1979).

47. Freudenthal J, Greve PA. Polychlorinated terphenyls in theenvironment. Bull Environ Contam Toxicol 10:108-111(1973).

48. Fries GF. The PBB episode in Michigan: an overall appraisal.CRC Crit Rev Toxicol 16:105-156 (1985).

49. Brilliant LB, Wilcox K, Van-Amburg G, Eyster J, Isbister J,Bloomer AW, Humphrey H, Price H. Breast-milk monitoringto measure Michigan's contamination with polybrominatedbiphenyls. Lancet 2(8091):643-646 (1978).

50. Miller FD, Brilliant LB, Copeland R. Polybrominatedbiphenyls in lactating Michigan women: persistence in the pop-ulation. Bull Environ Contam Toxicol 32:125-133 (1984).

51. Cordle F, Corneliussen P, Jelinek C, Hackley B, Lehman R,McLaughlin J, Rhoden R, Shapiro R. Human exposure topolychlorinated biphenyls and polybrominated biphenyls.Environ Health Perspect 24:157-172 (1978).

52. Landrigan PJ, Wilcox KR, Silva J, Harrold E, Humphrey B,Kauffman C, Heath CW. Cohort study of Michigan residentsexposed to polybrominated biphenyls: epidemiologic and

immunologic findings. Ann NYAcad Sci 320:284-294 (1979).53. Weil WB, Spencer M, Benjamin D, Seagull E. The effect of

polybrominated biphenyl on infants and young children. JPediatr 98:47-51 (1981).

54. Seagull EAW. Developmental abilities of children exposed topolybrominated biphenyls (PBB). Am J Public Health73:281-285 (1983).

55. Poland RL, Cohen SN. The contamination of the food chainin Michigan with PBB: the breast-feeding question. In: Drugand Chemical Risks to the Fetus and Newborn. New York:Alan R. Liss, 1980; 1-166.

56. Currie RA, Kadis VM, Breitkreitz WE, Cunningham CB,Bruns GW. Pesticide residues in human milk. Alberta, Canada.Pest Monit J 13:52-55 (1979).

57. Jan J, Komar M, Sancin A. Rozidua polikloriranih bifenilov vmaterinem mleku, v Sloveniji. Zurav Vestn 43:145 (1974).

58. Kodric-Smit M. Gas-chromatographic Determination ofOrganochlorine Contaminants in Human Milk. Thesis,University of Zagreb, Zagreb, 1980 (in Croatian).

59. Ansari GAS, Britt SG, Reynolds ES. Isolation and characteriza-tion of palmitolypentachlorophenol from human fat. BullEnviron Contam Toxicol 34:661-667 (1985).

60. Kontek M, Kubacki S, Pradowski S, Wierzchowiecka B.Chlorinated insecticides in human milk. Pediatr.Pol 46:183(1971) (in Polish).

61. Kaiser KLE. The rise and fall of Mirex. Environ Sci Technol12:520-528 (1978).

62. Bush B, Snow J, Connor S, Rueckert L, Han Y, Dymerski P,Hilker D. Mirex in human milk in upstate New York. ArchEnviron Contain Toxicol 12:739-746 (1983).

63. Vaz R, Blomkvist G. Traces of toxaphene components inSwedish breast milk analyzed by capillary. GC using ECD,electron impact and negative ion chemical ionization MS.Chemosphere 14:223-231 (1985).

64. Miyazaki T. Residues of the synergist S-421 in human milkcollected from the Tokyo metropolitan area. Bull EnvironContam Toxicol 29:566-569 (1982).

65. Heath RG, Harless RL, Gross ML, Lyon PA, Dupuy AE Jr,McDaniel DD. Determination of 2,3,7,8-tetrachlorodibenzo-p-dioxin in human milk at 0.1 to 10-parts-per-trillion level:method validation and survey results. Anal Chem 58:463-468(1986).

66. Yeh CY, Kwo P-H, Tsai S-T, Wang G-Y, Wang Y-T. A studyon pesticide residues in umbilical cord blood and maternalmilk. J Formosan Med Assoc 75:463-470 (1976).

67. Lonnerdal B, Asquith MT. Malathion not detected in breastmilk of women living in aerial spraying areas (letter). N Engl JMed 307:439 (1982).

68. Deutsche Forschungsgemeinschaft. Ruckstande undVerunreinigungen in Frauenmilch, Mitteilung Xl I derKommission zur Prufung von Ruckstande in Lebensmitteln.Weinheim:Verlag Chemie, 1984.

69. Lakritz L, Pensabene JW. Survey of human milk for volatile N-nitrosamines and the influence of diet on their formation. FoodChem Toxicol 22:721-724 (1984).

70. Trundle JI, Skellern GG. Gas chromatographic determinationof nicotine in human breast milk. J Clin Hosp Pharm8:289-291 (1983).

71. Ghelberg NW, Ricket I, Straus H. The lead content of milk inan area with non-ferrous metallurgical industry. Igeina Lekarski40:525 (1972).

72. Kroger M. General environmental contaminants occurring inmilk. In: Lactation: A Comprehensive Treatise (Larson BL,Smith VR, eds). New York:Academic Press, 1974; 135.

73. Walker B. Lead content of milk and infant formula. J FoodProt 43:178-179 (1980).

74. Lamm SH, Rosen JF. Lead contamination in milk fed toinfants: 1972-1973. Pediatrics 53:137-141 (1974).

75. Larson B, Slorach SA, Hagman U, Hofvander Y. WHO collab-orative breast feeding study. II. Levels of lead and cadmium inSwedish human milk, 1978-1979. Acta Paediatr Scand70:281-284 (1981).

204 Environmental Health Perspectives

CHEMICALS INHUMAN MILK

76. Bhattacharyya MH. Bioavailability of orally administered cad-mium and lead to the mother, fetus, and neonate during preg-nancy and lactation: an overview. Sci Total Environ28:327-342 (1983).

77. Skerfving S. Mercury in women exposed to methylmercurythrough fish consumption, and in their newborn babies andbreast milk. Bull Environ Contam Toxicol 41:475-482 (1988).

78. Fujita M, Takabatake E. Mercury levels in human maternaland neonatal blood, hair and milk. Bull Environ Contam

Toxicol 18:205-209 (1977).79. Amin-Zaki L, Majeed MA, Greenwood MR, Elhassani SB,

Clarkson TW, Doherty RA. Methylmercury poisoning in theIraqi suckling infant: a longitudinal study over five years. JAppl Toxicol 1:210-214 (1981).

80. WHO. Minor and Trace Elements in Breast Milk. Report of aJoint WHO/IAEA Collaborative Study. Geneva:World HealthOrganization, 1989.

Volume 103, Supplement 6, September 1995 205