Breast Milk Lead and Cadmium Levels in Suburban Areas of Nanjing,China

Kang-sheng Liu,Jia-hu Hao,Yu-qing Xu,Xiao-qi Gu,Juan Shi,Chun-fang Dai,Fei Xu,and Rong Shen＊

1State Key Laboratory of Reproductive Medicine,Department of Clinical Laboratory,Nanjing Maternity and Child Health Care Hospital Affiliated to Nanjing Medical University,Nanjing 210029,China

2Department of Maternity and Child Health Care,School of Public Health,Anhui Medical University,Hefei 230032,China

3Director’s Office,4Child Health Care Department,5Obstetrical Department,Nanjing Maternity and Child Health Care Hospital Affiliated to Nanjing Medical University,Nanjing 210004,China

6Department of Clinical Laboratory,Nanjing Children’s Hospital,Nanjing 210008,China

THE rapid industrialization and urbanization have increased the exposure to environmental pollutants.1Lead (Pb) and cadmium (Cd) have been reported as among the top 10 toxic metals in the Priority List of Hazardous Substances announced by the Agency for Toxic Substances and Disease Registry.2Pb poses a public health problem due to its adverse effects,mainly those affecting the central nervous system in the most vulnerable populations,such as pregnant and lactating women and children.In adults,Pb exposure has been found associated with anxiety and depression.3Because of the absence of a safe exposure limit in children and the ability of Pb to accumulate in the body for a long time,a great interest in evaluating its adverse effects in low concentrations has emerged.4Cd has also been found to cause neurotoxicologic and behavioral changes in both human and experimental animal studies.5Cd exposure may be implicated in some neurological disorders including hyperactivity and increased aggressiveness in humans.6Nonetheless,the observational clinical data on the relationship between anxiety or depression and Cd exposure are inadequate.

Pb can negatively influence the growth of newborns.Previous studies have reported an inverse relationship between anthropometric measurements of newborns and the Pb levels in placenta/umbilical cord.7As Pb is readily transferred to the fetus through the placenta,8its interference can be observed on early embryonic development and during the last months of pregnancy.9The primary sources of Pb in breast milk are diet and bone lead.When up to 5% of bone mass is mobilized as a source of calcium during pregnancy and lactation,the lead accumulated in bone from previous exposure is concomitantly released into the blood and excreted into breast milk.10Detectable Pb level in breast milk has been documented in population studies of women with no current environmental or occupational Pb exposures.11

The levels of toxic metals in breast milk can indicate the exposure on the part of the infants and the possible prenatal exposure on the part of the mothers.12A limited number of studies have been conducted to determine the toxic metal concentrations in breast milk.4In the present study,we detected the Pb and Cd levels in breast milk in the second postpartum month,investigated the relationship between some sociodemographic parameters and the concentrations of Pb and Cd,and analyzed the influence of these toxic metals on the infants’ growth and on the mothers’ postpartum depression.

MATERIALS AND METHODS

Study population

This is a cross-sectional study conducted from November 2009 to December 2010.A total of 170 voluntarily enrolled healthy mothers were evaluated at Nanjing Maternity and Child Health Care Hospital,and all of them provided breast milk samples at the second postpartum month.These mothers were all living in a suburban but not non-industrial area of Nanjing,and not occupationally exposed to toxic metals.

All the mothers completed a questionnaire regarding their occupation,smoking habits,childbearing history,history of anemia at any time,and intake of iron and vitamin supplementation during pregnancy and within 2 months after delivery.The level of maternal hemoglobin on the first postpartum day was taken from the hospital records.All the studied mothers were evaluated using the Edinburgh Postpartum Depression Scale (EPDS) to identify the risk of postpartum depression,those with a score ≥13 were considered at the risk of depression.13The infants of these mothers were examined,and their z scores of weight for age (WAZ),length for age (LAZ),head circumference for age (HCZ),and body mass index for age (BAZ) were calculated based on recent World Health Organization Growth References.14The study was approved by the Ethical Committee of Nanjing Maternity and Child Health Care Hospital.The mothers were informed about the purpose of the study and written informed consents were obtained from all the participants.

Analysis of the breast milk

The breast milk samples were collected 2 hours after the latest feeding session in the morning.Every mother expressed 10 mL milk directly into clean polyethylene tubes.All the samples were frozen immediately after collection and kept in -20°C until analyzed.

In order to minimize the polyatomic interferences,reagent water equivalent to ASTM Type 1 (ASTM D 1193,＞18 mΩ cm resistivity),0.5% (v/v) nitric acid,a tuning solution (for sensitivity tuning∶10 μg/L for each of the metals Li,Y,and Tl in 0.1% HNO3),and a single-element standard stock solution for Pb and Cd were used.Breast milk sample (1 g) was accurately weighed,put into a dry XP1500 vessel,into which 5 mL of 0.1% HNO3and 5 mL of distilled water were added.For digestion,this mixture was placed in a microwave oven at 105°C for 10 minutes.Following digestion,the samples were diluted with 25 mL distilled water.The standard stock solution and the milk samples in the polyethylene tubes were placed in the automatic sequencer of the Inductively Coupled Plasma Mass Spectrometer (Yokogawa Analytical Systems,Tokyo,Japan).The lowest limits of quantitation (LOQ) for Pb and Cd were both 0.2 μg/L.

Statistical analysis

The data were analyzed using SPSS 14.0.The results were presented as means±SD (range),and medians (25th-75th percentiles).The Mann-Whitney U test was applied to compare the Pb and Cd levels between subgroups divided according to the characteristics listed in the questionnaire.The Pearson correlation coefficient was used for the relationship between breast milk Pb and Cd concentrations and numerical variables (anthropometric measurements,EPDS scores,etc.).P＜0.05 was considered statistically significant.

RESULTS

General information

The mean maternal age was 26±4 years (18-42 years),7(4.1%) were over 34 years.One hundred and ten (64.7%)of the mothers had been educated for no more than 8 years.Only 37 (21.8%) mothers were working.The monthly family income was less than 2000 yuan in 20 (11.8%),2000 yuan in 40 (23.5%),and over 2000 yuan in 110(64.7%).Mean gravidity was 2.1±0.9,and mean parity 1.7±0.8 (Table 1).The birth interval was less than 2 years in 10 (5.3%) mothers who had parity ＞ 1.The mean maternal hemoglobin level was 11.4±1.4 g/dL (6.1-14.8 g/dL).

Seventy-five (44.1%) of the infants were female and 95 (55.9%) were male.Nine (5.3%) of the infants were＜ 2500 g at birth and 21 were born ＜ 37 weeks gestation.The mean birth weight was 3216±345 g (2131-4110 g)and the mean head circumference at birth was 34.1±1.7 cm (31.0-39.0 cm).In the second month after birth,the mean weight for age was 4788±634 g (2670-7100 g),the mean length for age was 54.17±2.71 cm (41.5-67.0 cm),and the mean head circumference for age was 37.9±1.4 cm (33.0-42.0 cm) (Table 1).

Pb and Cd levels in breast milk

The Pb level was above the LOQ in 168 (98.8%) breast milk samples,and ＞ 5 μg/L in 164 (96.5%) samples.The median (25th-75th percentile) Pb level of the 170 samples was 40.6 μg/L (22.1-59.2 μg/L),which was confirmed by a duplicate analysis.

The Cd level in breast milk was above LOQ in 87(51.2%) samples,and ＞ 1 μg/L in 54 (31.8%) samples.The median Cd level was 0.67 μg/L (＜ LOQ-1.26 μg/L)(Table 2).The highest Cd level in the samples was 43.0 μg/L.The mother was 19 years old,and this was her first pregnancy.The baby was delivered by vaginal birth,weighing 3400 g.There was no history of chronic disease.She was diagnosed with anemia during the pregnancy but did not take iron supplements.The maternal hemoglobin level on the first postpartum day was 9.1 g/dL.She had lived in a suburban area since childhood.The heating in her former childhood house was provided by a coal stove.Her family currently lived in a 20-year-old house,heated by a coal stove as well,and close to the main road (＜ 200 m).She and her husband were both non-smokers and the mother was not exposed to cigarette smoke during her childhood in her family environment.

Table 1.Characteristics of the studied mothers and infants

Relationship between maternal sociodemographic characteristics and breast milk Pb/Cd levels

The maternal age and education level,monthly family income,parity,and birth interval were not related with the breast milk Pb and Cd levels.The employment status of the mothers did not affect the Pb level in breast milk,while the median breast milk Cd level was higher in the unemployed mothers than that in the working mothers (0.71 μg/Lvs.0.01 μg/L,P=0.025).

Active and passive smoking during pregnancy and in 2 postpartum months was analyzed.The mean daily number of cigarettes in active smokers was ＜ 5 and thus active and passive smokers were combined for analysis(active and passive smokers versus non-smokers).Smoking during pregnancy or in 2 postpartum months had no effect on the breast milk Pb level,while the median breast milk Cd level in the active and passive smokers during pregnancy was significantly higher than that in non-smokers (0.88 μg/Lvs.0.00 μg/L,P=0.025,Table 3).Smoking in 2 postpartum months had no significant effect on the breast milk Cd level.

The mothers with a history of anemia at any time had higher breast milk Pb level than those without anemia(41.1 μg/Lvs.37.9 μg/L,P=0.050,Table 3).According to the Spearman correlation analysis,no significant correlation was observed between the maternal Hb level and the breast milk levels of Pb and Cd (Pb∶rs=-0.031,P=0.644;Cd∶rs=-0.117,P=0.160).The use of iron and vitamin supplements during pregnancy or in 2 postpartum months did not affect the breast milk Pb level.The mothers who did not take iron and vitamin supplements for 2 months postpartum had higher breast milk Cd level than those who did (iron supplement;0.74 μg/Lvs.00 μg/L,P=0.025;vitamin supplement∶0.78 μg/Lvs.0.00 μg/,P=0.005;Table 3).Nonetheless,the use of iron and vitamin supplements during pregnancy did not affect the breast milk Cd level.

Twenty-six mothers (15.3%) were at risk of postpartum depression at the second postpartum month.The breast milk Pb and Cd levels in the mothers at risk of depression were not significantly different from those in the mothers not at risk (Pb∶r=-0.055,P＞ 0.05;Cd∶r=-0.051,P＞ 0.05).

Relationship between infant factors and breast milk Pb/Cd levels

Breast milk Pb or Cd levels did not change with birth weight,gestational age,or gender of infants (Table 4).In the infants,61 had crying attacks of colic type and 75 had irregular sleep pattern.Infants with irregular sleep pattern and colic crying had similar breast milk Cd and Pb levels compared with infants without these problems.A total of 117 infants (68.8%) were exclusively breastfed at the second postpartum month.In the comparison between exclusively breastfed infants and mixed fed infants,the Pb and Cd levels in their mothers’ breast milk at the second postpartum month showed no significant difference.

Breast milk Cd level at the second postpartum month was negatively correlated with HCZ and WAZ at birth in girls (HCZ∶r=-0.248,P=0.042;WAZ∶r=-0.241,P=0.024;Table 5).Breast milk Pb and Cd levels were not correlated with other anthropometric measurements.

DISCUSSION

The WHO has reported that 2-5 μg/L of Pb may exist in the breast milk in 3 postpartum months under normal conditions based on its research conducted in 1989 in 6 countries (Guatemala,Hungary,Nigeria,Philippines,Sweden,and Zaire.).15Pb level in breast milk varies with regions(0.5-126.6 μg/L).16-18The difference in the breast milk Pb level may depend on various factors such as the time of sampling (morning or night),the time of lactation (colostrum/transient/mature milk or foremilk/hindmilk),the method of sampling (pump or manual),maternal factors(parity and maternal Pb burden),as well as environmental factors (place of residence and exposure level/duration).However,several other factors (methods of analysis,contamination of the samples,etc.) might interfere in the final results as well.19-21Considering the previously reported range,the present study detected high levels of Pb in breast milk (40.6 μg/L),consistent with the results of some previous studies,in which there were risk factors such as occupational exposure,urban living,living close to ore processing plants.22,23The participants in the present study,in contrast,lived in a suburban area,had no occupational exposure,and were not close to ore processing plants.

Table 2.Levels of lead (Pb) and cadmium (Cd) in breast milk at the second postpartum month (μg/L,n=170)

Table 3.Maternal sociodemographic characteristics and Pb/Cd levels in breast milk

Table 4.Relationship between infant factors and Pb/Cd levels in breast milk at the second postpartum month

Table 5.Correlations between breast milk Pb/Cd levels and z scores of anthropometric measurements at birth and the second postpartum month

The median breast milk Cd level at the second postpartum month was 0.67 (＜ LOQ-1.26) μg/L in this study.The level reported in the literature is 0.06-24.6 μg/L.24,25Only 5 breast milk samples in the present study showed a Cd level higher than the minimal risk level (5 μg/L) stated by Abadin et al.11The breast milk Cd level at the second postpartum month in the present study was lower than that reported by Turan et al.26

The present study did not found maternal age and education level,monthly family income,parity,and birth interval correlated with the breast milk Pb and Cd levels.However,Younes et al27detected lower breast milk Pb levels in younger mothers (≤20 years) than in older mothers (≥36 years) (5.1±1.4 μg/Lvs.13.4±3.5 μg/L).Several studies have shown that maternal age and parity are not correlated with breast milk Cd concentrations.4,28In the present study,mothers who were housewives had significantly higher breast milk Cd concentrations than the working mothers.A possible explanation is that Cd may bind to the dust particles in indoor air.It has been reported that the amount of Cd in household dust particles was higher than that in outdoor air.29The inadequate ventilation and the presence of cigarette smoke in the house may cause the high breast milk Cd level in housewife mothers.

Some studies reported a relation between iron deficiency and high blood Pb level in humans.30In the present study,mothers with a history of anemia at any time had higher breast milk Pb level than those without anemia history (P=0.050).Nonetheless,this study revealed no significant relation between breast milk Pb level and maternal Hb level on the first postpartum day postpartum,or between breast milk Pb level and iron supplementation during pregnancy or in 2 postpartum months.Similarly,Koyashiki et al31observed no correlation between Hb and lead level in milk/blood samples taken between the 15 th and 210 th day after delivery.

Vahter et al32reported an increase in Cd absorption and its toxicity in the presence of decreased iron stores and iron deficiency.Satarug et al33reported that non-smoking mothers with insufficient iron stores had three times more body Cd load than the mothers with sufficient iron stores.However,the present study showed that the breast milk Cd level at the second postpartum month did not change with maternal anemia status,which might be explained by the severity and duration of anemia.

In this study,the breast milk of the mothers not taking vitamin and iron supplements at the second postpartum month contained higher levels of Cd than those of the mothers taking supplements (P=0.005 and 0.025).Similarly,Gundacker et al34identified lower breast milk Cd concentrations at 2-14 postpartum days in the non-smoking mothers who took trace element/vitamin supplements.In contrast,Leotsinidis et al35reported that taking trace element/vitamin supplements did not affect the Cd levels in breast milk.

One of the most well-known routes of exposure to Pb and Cd is through cigarette smoking.36In this study,mothers with active and/or passive smoking exposure during pregnancy had significantly higher breast milk Cd level than non-smoking mothers;however,maternal smoking habit in the postpartum period did not affect breast milk Pb level.Although some studies reported a relation between smoking and breast milk Pb levels,34other studies have shown no such relationship.37This might be explained by the number of cigarettes consumed or other associated environmental contaminants.Ursinyova et al37found higher breast milk Cd level in mothers actively smoking before and during pregnancy.

Some clinical studies have been conducted on anxiety and depression status with exposure to Pb.38Dang et al38assessed the mental health status of the residents in a mining area in Hubei using the Symptom Checklist 90(SCL-90) and State-Trait Anxiety Inventory (STAI).They found that the mental health status of the residents was affected,having higher scores on the S-AI and T-AI scales than those living in a non-mining area.This study found no correlation between the breast milk Pb/Cd levels and EPDS scores.To our knowledge,this is the first study investigating the association between breast milk Pb/Cd levels and maternal mental health in the postpartum period.Further studies are necessary to confirm the finding of the present study.

The present study found no relationship between the breast milk Pb/Cd levels at the second postpartum month and infant characteristics such as birth weight,gestational age,and gender.

Irregular sleep pattern or colic crying in infants were not associated with different breast milk Pb and Cd levels in their mothers at the second postpartum month in the present study.On the other hand,there has been no published study exploring the effect of toxic metals on infantile colic.

The clinical presentation with toxic metals might differ with gender.39In this study,breast milk Cd level was negatively correlated with HCZ and WAZ at birth in girls.The breast milk Pb level in the second postpartum month was not correlated with the anthropometric measurements of the infants at birth or 2 months after birth in either sex in the present study.Kordas et al40reported a negative correlation between the blood Pb levels and the head circumference measurements of 602 infants aged 8 years and living near a casting plant in Mexico.The possible explanation for these differences might be the level of intoxication,associated environmental conditions,genetic factors,and follow-up duration.40

In conclusion,Pb level in breast milk samples in this study were much higher than the currently recommended safety limits.Human milk could be used for research on lead exposure,with the goal to evaluate the risk to children’s health using a non-invasive biological procedure.Investigations are urgently required to explore factors such as environment,diet,lifestyle,and/or cultural habits contributing to maternal and fetal exposures to Pb and Cd.Preventive measures must be taken accordingly to reduce environmental exposure and manage anemia in women in all age groups.Furthermore,periodic breast milk monitoring programs could help in evaluating maternal exposure due to ongoing exposure during the course of lactation.

ACKNOWLEDGEMENTS

This study was supported by Nanjing Maternity and Child Health Care Hospital.The authors are grateful to the women who volunteered to participate in this study.

1.Wong SL,Lye EJ.Lead,mercury and cadmium levels in Canadians.Health Rep 2008;19∶31-6.

2.Agency for Toxic Substances &Disease Registry.Priority List of Hazardous Substances.Atlanta∶Agency for Toxic Substances and Disease Registry.2007 [cited 2012 Jul 1].Available from∶http∶//www.atsdr.cdc.gov/SPL/index.html

3.Schnaas L,Rothenberg SJ,Flores MF,et al.Reduced intellectual development in children with prenatal lead exposure.Environ Health Perspect 2006;114∶791-9.

4.Dursun A.The levels of Pb,Hg,Cd in cord blood,breast milk,and newborn hair.Ankara∶Thesis,Hacettepe University∶2008.

5.Sinha M,Manna P,Sil PC.Cadmium-induced neurological disorders∶prophylactic role of taurine.J Appl Toxicol 2008;28∶974-86.

6.Matés JM,Segura JA,Alonso FJ,et al.Role of dioxins and heavy metals in cancer and neurological diseases by ROS-mediated mechanisms.Free Radic Biol Med 2010;49∶1328-41.

7.Llanos MN,Ronco AM.Fetal growth restriction is related to placental levels of lead and arsenic but not with antioxidant activities.Reprod Toxicol 2009;27∶88-92.

8.Ballew C,Khan LK,Kaufmann R,et al.Blood lead concentration and children's anthropometric dimensions in the Third National Health and Nutrition Examination Survey (NHANES III),1988-1994.J Pediatr 1999;134∶623-30.

9.Rossipal E,Krachler M,Li F,et al.Investigation of the transport of trace elements across barriers in humans∶studies of placental and mammary transfer.Acta Paedriatr 2000;89∶1190-5.

10.Ettinger AS,Téllez-Rojo MM,Amarasiriwardena C,et al.Influence of maternal bone lead burden and calcium intake on levels of lead in breast milk over the course of lactation.Am J Epidemiol 2006;163∶48-56.

11.Abadin HG,Hibbs BF,Pohl HR.Breast-feeding exposure of infants to cadmium,lead and mercury∶a public health viewpoint.Toxicol Ind Health 1997;13∶495-517.

12.Järup L.Hazards of heavy metal contamination.Br Med Bull 2003;68∶167-82.

13.Engindeniz N.Validation of Turkish form of Edinburgh Postpartum Depression Scale.Thesis,Ege University,İzmir∶1996.

14.WHO Multicentre Growth Reference Study Group.WHO Child growth standards based on length/height,weight and age.Acta Paediatr Suppl 2006;450∶76-85.

15.Parr RM,DeMaeyer EM,Iyengar VG,et al.Minor and trace elements in human milk from Guatemala,Hungary,Nigeria,Philippines,Sweden,and Zaire.Results from a WHO/IAEA joint project.Biol Trace Elem Res 1991;29∶51-75.

16.Guidi B,Ronchi S,Ori E,et al.Lead concentrations in breast milk of women living in urban areas compared with women living in rural areas.Pediatr Med Chir 1992;14∶611-6.

17.Gulson BL,Mizon KJ,Palmer JM,et al.Longitudinal study of daily intake and excretion of lead in newly born infants.Environ Res 2001;85∶232-45.

18.Gundacker C,Pietschnig B,Wittmann KJ,et al.Smoking,cereal consumption,and supplementation affect cadmium content in breast milk.J Expo Sci Environ Epidemiol 2007;17∶39-46.

19.Casey CE,Smith A,Zhang P.Microminerals in human and animal milk.In∶Jensen RC,editor.Handbook of milk composition.San Diego∶Academic Press;1995.

20.Ettinger AS,Téllez-Rojo MM,Amarasiriwardena C,et al.Levels of lead in breast milk and their relation to maternal blood and bone lead levels at one month postpartum.Environ Health Perspect 2004;112∶926-31.

21.Iyengar V,Woittiez J.Trace elements in human clinical specimens∶evaluation of literature data to identify reference values.Clin Chem 1988;34∶474-81.

22.Al-Saleh I,Shinwari N,Mashhour A.Heavy metal concentrations in the breast milk of saudi women.Biol Trace Elem Res 2003;96∶21-37.

23.Souad C,Farida Z,Narda L,et al.Trace element level in infant hair and diet,and in the local environment of the Moroccan city of Marrakech.Sci Total Environ 2006;370∶337-42.

24.Hallén IP,Jorhem L,Lagerkvist BJ,et al.Lead and cadmium levels in human milk and blood.Sci Total Environ 1995;166∶149-55.

25.Sternowsky HJ,Wessolowski R.Lead and cadmium in breast milk.Higher levels inurban vs rural mothers during the first 3 months of lactation.Arch Toxicol 1985;57∶41-5.

26.Turan S,Saygi S,Kiliç Z,et al.Determination of heavy metal contents in humancolostrum samples by electrothermal atomic absorption spectrophotometry.J Trop Pediatr 2001;47∶81-5.

27.Younes B,Al-Meshari AA,Al-Hakeem A,et al.Lead concentration in breast milk of nursing mothers living in Riyadh.Ann Saudi Med 1995;15∶249-51

28.Honda R,Tawara K,Nishijo M,et al.Cadmium exposure and trace elements in human breast milk.Toxicology 2003;186∶255-9.

29.Rasmussen PE,Subramanian KS,Jessiman BJ.A multi-element profile of house dust in relation to exterior dust and soils in the city of Ottawa,Canada.Sci Total Environ 2001;267∶125-40.

30.Wright RO,Tsaih SW,Schwartz J,et al.Association between iron deficiency and blood lead level in a longitudinal analysis of children followed in an urban primary care clinic.J Pediatr 2003;142∶9-14.

31.Koyashiki GA,Paoliello MM,Matsuo T,et al.Lead levels in milk and blood from donors to the Breast Milk Bank in Southern Brazil.Environ Res 2010;110∶265-71.

32.Vahter M,Berglund M,Akesson A,et al.Metals and women’s health.Environ Res 2002;88∶145-55.

33.Satarug S,Ujjin P,Vanavanitkun Y,et al.Influence of body iron store status and cigarette smoking on cadmium body burden of healthy Thai women and men.Toxicol Lett 2004;148∶177-85.

34.Gundacker C,Pietschnig B,Wittmann KJ,et al.Lead and mercury in breast milk.Pediatrics 2002;110∶873-8.

35.Leotsinidis M,Alexopoulos A,Kostopoulou-Farri E.Toxic and essential trace elements inhuman milk from Greek lactating women∶association with dietary habits and other factors.Chemosphere 2005;61∶238-47.

36.Gundacker C,Pietschnig B,Wittmann KJ,et al.Smoking,cereal consumption,and supplementation affect cadmium content in breast milk.J Expo Sci Environ Epidemiol 2007;17∶39-46.

37.Ursinyova M,Masanova V.Cadmium,lead and mercury in human milk from Slovakia.Food Addit Contam 2005;22∶579-89.

38.Dang WM,Ma WJ,Wang S,et al.Investigation on mental health of residents living in a mineral area in Hubei province.Zhonghua Lao Dong Wei Sheng Zhi Ye Bing Za Zhi 2008;26∶457-60.

39.Vahter M,Akesson A,Lidén C,et al.Gender differences in the disposition and toxicity of metals.Environ Res 2007;104∶85-95.

40.Kordas K,Lopez P,Rosado JL,et al.Blood lead,anemia,and short stature are independently associated with cognitive performance in Mexican school children.J Nutr 2004;134∶363-71.