DSpace at VNU: Dioxins and related compounds in human breast milk collected around open dumping sites in Asian developing countries: Bovine milk as a potential source

In this study, concentrations of dioxins and related compounds DRCs—such as polychlorinated dibenzo-p-diox-ins, polychlorinated dibenzofurans, and coplanar polychlori-nated biphenyls—we

Dioxins and Related Compounds in Human Breast Milk Collected Around Open Dumping Sites in Asian Developing Countries: Bovine Milk as a Potential Source

T Kunisue,1M Watanabe,1H Iwata,1A Subramanian,1,2I Monirith,1T B Minh,1R Baburajendran,2T S Tana,3

P H Viet,4M Prudente,5S Tanabe1

1 Center for Marine Environmental Studies, Ehime University, Bunkyo-cho 2-5, Matuyama 790-8577, Japan

2 Center of Advanced Study in Marine Biology, Annamalai University, Tamil Nadu, India

3 Social and Culture Observation Unit, Cabinet of the Council of Minister, Kingdom of Cambodia

4 Center for Environmental Technology and Sustainable Development, Hanoi National University, Hanoi, Vietnam

5 Science Education Department, De La Salle University, Manila, Philippines

Received: 11 August 2003 /Accepted: 14 March 2004

Abstract In this study, concentrations of dioxins and related

compounds (DRCs)—such as polychlorinated

dibenzo-p-diox-ins, polychlorinated dibenzofurans, and coplanar

polychlori-nated biphenyls—were found in human breast milk from

women living near dumping sites of municipal waste and

reference sites in India, Cambodia, Vietnam, and the

Philip-pines during 1999 to 2000 DRCs were detected in all human

breast milk samples analyzed, demonstrating that residents in

these Asian developing countries have been exposed to these

contaminants In India, the concentrations of DRCs in human

breast milk from women living near the investigated dumping

site were notably higher than those from women living near

reference sites and from women in other Asian developing

countries Toxic equivalent quantity (TEQ) levels of DRCs

were comparable with or higher than those reported in the

general populations of developed countries since 1990 In

contrast, levels of these contaminants in human breast milk in

women from Cambodia and Vietnam were not significantly

different between milk from women living near the dumping

and reference sites These results indicate that significant

pol-lution sources for DRCs are present in Indian dumping sites

and that residents there have been exposed to relatively higher

levels of these contaminants TEQ levels in human breast milk

from the dumping site in India tended to decrease with an

increase in the number of previous deliveries by mothers,

whereas no significant relationship was observed in Cambodia,

Vietnam, or the Philippines This suggests that mothers who

have been exposed to relatively high levels of DRCs transfer

greater amounts of these contaminants to the first infant than

later ones through breast-feeding, which in turn implies that the

first children of these mothers might be at higher risk from

DRCs When the residue levels of DRCs in bovine milk

col-lected from the Indian dumping site and reference sites were

examined, TEQ levels in bovine milk from the dumping site

were higher than those from reference sites This result

sug-gests that bovine milk is a potential source of DRCs for residents living near the dumping site in India To our knowl-edge, this is the first comprehensive study on exposure to DRCs of residents living in proximity to open dumping sites of municipal waste in Asian developing countries

Polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated

dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs) are lipophilic-stable contaminants of great concern with respect

to their toxic effects on humans and wildlife In developed countries, the residue levels of these contaminants in various environmental media and biota, including humans, have gen-erally decreased in recent years (Alcock and Jones 1996;

Brad-ley 2000; LaKind et al 2001; Nore´n and Meironyte´ 2000)

because of highly efficient incinerators and strict regulations on production, use, and transportation of various chemicals In contrast, smaller numbers of studies have reported contamina-tion status and temporal trends of these chemicals in develop-ing countries, especially regarddevelop-ing human exposure, than

de-veloped countries (LaKind et al 2001).

Asian developing countries—such as India, Cambodia, Viet-nam, and the Philippines—located in the tropical region have large open dumping sites of municipal waste in the suburbs of major cities In these sites, varieties of municipal waste are dumped continuously and burned under low temperature by spontaneous combustion or intentional incineration It can be anticipated that dioxins and related compounds (DRCs) would

be formed by such low-temperature combustion practices and that the surrounding environment may be polluted by these contaminants In addition, it can also be anticipated that PCBs would leach out from electric appliances dumped in these dumping sites We previously conducted a survey of DRCs in soils collected from these Asian developing countries and found that the residue levels of DRCs were higher in soils from the dumping sites than from the agricultural and urban soils collected far from these areas, indicating that the dumping sites

Correspondence to: S Tanabe; email: shinsuke@agr.ehime-u.ac.jp

E

a n d T oxicology

© 2004 Springer Science ⴙBusiness Media, Inc.

are potential sources of DRCs (Minh et al 2003) From this we

presumed that residents living near the dumping sites might be

exposed to these contaminants because most of them obtain

their livelihood by doing some dumpsite-dependent labor

Es-pecially, it is expected that in utero and lactational exposure to

DRCs may adversely affect the brain development and immune

systems of infants and children (Koopman-Esseboom et al.

1994; Nagayama et al 1998a, b; Porterfield et al 1994;

Weis-glas-Kuperus et al 1995, 2000) To our knowledge, however,

no study has reported on the exposure of residents living

around the open dumping sites of municipal waste in Asian

developing countries to these contaminants

This study attempted to elucidate the contamination status of

DRCs in human breast milk collected from the women living in

proximity to dumping sites of municipal waste in India,

Cam-bodia, Vietnam, and the Philippines and to assess the risk of

exposure of infants to these contaminants We compared the

present data with those from reference sites in the respective

countries and also in the general populations of countries

reported elsewhere so that we might understand the magnitude

of contamination in human breast milk from the dumping sites

In addition, we also examined the contamination of DRCs in

bovine milk collected from cows around the dumping and

reference sites in India, which was considered as one potential

source of these chemicals to humans

Materials and Methods

Sample Collection

Human breast milk samples (one sample from each mother) were

collected from mothers living near open dumping sites of municipal

waste in Perungudi, Chennai, India during August 2000 (n⫽ 11), in

Meanchey, Phnom Penh, Cambodia during November 1999 and

De-cember 2000 (n⫽ 19), in Tay Mo, Hanoi, Vietnam during April 2000

(n⫽ 8), and in Payatas, Quezon, Philippines during February 2000

(n⫽ 9) The characteristics and the situations of the dumping sites in

these countries have been reported previously (Minh et al 2003).

Samples were also collected from mothers in Chennai (n⫽ 8), Phnom

Penh (n ⫽ 16), and Hanoi (n ⫽ 10) on the same dates in locations at

least 10 km away from the dumping sites (reference sites) We

ob-tained informed consent from all the donors of milk samples, the

details of whom are listed in Table 1 All of these biological

charac-teristics are not significantly different between the dumping and

ref-erence sites (p⬎ 0.1) Bovine milk samples were collected from cows

near dumping sites (cow n ⫽ 2, buffalo n ⫽ 3) and from cows near

reference sites (cow n ⫽ 3, buffalo n ⫽ 2) in India All of the samples

were collected in chemically cleaned containers and stored at ⫺20°C

until analysis.

Chemical Analysis

During chemical analysis, extraction of human breast milk was

con-ducted per the method reported by Hirai et al (2001) Cleanup and

separation processes of human breast milk and chemical analyses of

bovine milk were performed per the method recommended by the

Ministry of Health, Labor, and Welfare of Japan with some

modifi-cations The following were spiked to 50 g human breast milk and

bovine milk samples as internal standards: 13 C12-labeled PCDD/DFs

as well as non- and mono-ortho PCBs (2,3,7,8-TetraCDD/

DF; 1,2,3,7,8-PentaCDD/DF; 1,2,3,6,7,8-HexaCDD/DF; 1,2,3,7,8,9-HexaCDF; 1,2,3,4,6,7,8-HeptaCDD/DF; OctaCDD/DF; TetraCB77; TetraCB81; PentaCB118; PentaCB126; HexaCB156; HexaCB167; HexaCB169; and HeptaCB189) The human breast milk samples were added onto diatomaceous earth (Extrelut NT, Merck, Germany) packed in a glass column and extracted with diethyl ether The bovine milk samples were added into a glass separating funnel with saturated sodium oxalate solution ethanol, diethyl ether, and hexane and ex-tracted twice Lipid in the extract was removed by gel permeation chromatography packed Bio-Bead S-X 3(Bio-Rad) Fifty percent di-chloromethane in hexane was used as moving phase, and flow rate was set at 5 ml/min First fraction containing lipid was discarded, and the next timed fraction containing DRCs was concentrated and passed through activated silica gel (Wako-Gel S-1; Wako Pure Chemical, Japan) packed in a glass column DRCs were eluted with hexane After concentration, the extract was spiked onto activated alumina (alumi-nium oxide 90 active basic, Merck) packed in a glass column The first

fraction eluted with hexane contained most of the mono-ortho PCBs,

and the second fraction eluted with 50% dichloromethane in hexane

contained the remaining mono-ortho PCBs, the non-ortho PCBs, and

the PCDD/DFs Then the second fraction was passed through activated carbon-dispersed silica gel (Kanto Chemical, Japan) packed in a glass column The first fraction was eluted with 25% dichloromethane in

hexane to obtain the remaining mono-ortho PCBs and combined with the first fraction separated by alumina column Non-ortho PCBs and

PCDD/DFs were eluted with toluene as the second fraction Both fractions were concentrated to near dryness, and 13 C12-labeled Pen-taCB105, HexaCB157, and HeptaCB180 in decane were added to the combined first fraction; 13 C12-labeled 1,2,3,4-TetraCDD and 1,2,3,7,8,9-HexaCDD in decane were added to the second fraction, all

as injection spikes To determine lipid content in human breast and bovine milk, another 10-g sample was extracted, dried at 80°C, and weighed.

Identification and quantification were performed using a gas chro-matograph (GC, Agilent 6890 series) with an autoinjection system and

a bench-topped, double-focusing mass selective detector (MS, JEOL GC-Mate II) with a resolving power⬎3000 for mono-ortho PCBs and

a high-resolution MS (JEOL JMS-700D) with a resolving power

⬎10,000 for non-ortho PCBs and PCDD/DFs Both pieces of

equip-ment were operated at an electron ionization energy of 38 – 40 eV, and the ion current was 600 ␮A DRCs were monitored by selective ion

monitoring mode at the two most intensive ions of the molecular ion cluster among [M]⫹, [M ⫹ 2] ⫹ , and [M ⫹ 4] ⫹ , except P5CDD, which was monitored at [M]⫹ and [M ⫹ 2] ⫹ All of the congeners were quantified using an isotope dilution method to the corresponding

13 C12-congeners when the isotope was within 15% of the theoretical ratio and the peak area was more than 10 times of noise Recoveries for the 13 C12-labeled PCDD/DFs and coplanar PCBs were within 60% to 110% Toxic equivalent quantities (TEQs) were estimated based on human/mammal toxic equivalency factors (TEFs) proposed by the

World Health Organization (WHO) (Van den Berg et al 1998).

Statistical Analysis

The Mann-Whitney U test was employed to detect the differences in concentrations of DRCs in human breast milk and characteristics of the mothers between the dumping and reference sites as well as the differences in concentrations affected by the number of deliveries per

mother A p value ⬍ 0.05 was considered statistically significant.

These analyses were performed using StatView software (version 4.51.1; Abacus Concepts).

Results and Discussion

Residue Levels in Human Breast Milk

DRCs were detected in all of the samples of human breast milk

analyzed in this study (Table 2), demonstrating that residents

living near open dumping sites of municipal waste and

refer-ence sites in India, Cambodia, Vietnam, and the Philippines

have been exposed to these contaminants The concentrations

of PCDDs in human breast milk from dumping sites in

differ-ent countries in decreasing order were as follows: India (mean

[range] 290 [150 –780] pg/g lipid wt) ⬎ Philippines (190

[29 –730] pg/g lipid wt)⬎ Cambodia (49 [14–170] pg/g lipid

wt)ⱖ Vietnam (32 [10–81] pg/g lipid wt) At the same time,

the concentrations of PCDFs in decreasing order were as

fol-lows: India (50 [15–130] pg/g lipid wt) ⬎ Philippines (21

[5.9 – 44] pg/g lipid wt) ⱖ Vietnam (20 [7.3–42] pg/g lipid

wt)ⱖ Cambodia (15 [5.2–55] pg/g lipid wt) Furthermore, in

India the concentrations of PCDD/DFs in human breast milk

from the dumping site were higher than those from reference

sites, whereas levels of these contaminants in human breast

milk from Cambodia and Vietnam were not significantly

dif-ferent between the dumping and reference sites (Figure 1)

These results indicate that significant pollution sources of

PCDD/DFs are present in the dumping site in India and that the

residents living near them have been exposed to relatively

higher levels of these contaminants than residents in the other

countries evaluated in this study

The concentrations of non-ortho (mean [range] 260 [30 –

610] pg/g lipid wt) and mono-ortho PCBs (38,000 [2500 –

170,000] pg/g lipid wt) in human breast milk collected from

mothers living near the dumping site in India were also higher

than those of Vietnam women (non-ortho PCBs 62 [17–100]

pg/g lipid wt, mono-ortho PCBs 24,000 [4200 – 46,000] pg/g

lipid wt); Philippine women (non-ortho PCBs 76 [26 –160]

pg/g lipid wt, mono-ortho PCBs, 8800 [1700 –28,000] pg/g

lipid wt), and Cambodian women (non-ortho PCBs 51 [29 –

130] pg/g lipid wt; mono-ortho PCBs 8000 [820 –28,000] pg/g

lipid wt) (Table 2) In addition, the levels of non- and

espe-cially mono-ortho PCBs in human breast milk from the

dump-ing site of India were notably higher than those from reference

sites (Figure 1) As in the case of PCDD/DFs described above,

this fact indicates that pollution sources of non- and

mono-ortho PCBs are also present in the dumping site and that

residents living near there have been exposed to these contam-inants It was previously reported that the concentrations of PCBs in various foods from urban and rural regions of India

were relatively low (Kannan et al 1992b), which supports our

finding In human breast milk from Vietnam, relatively high

concentrations of mono-ortho PCBs were detected, but the

levels were lower than those from the dumping site in India In

a previous investigation of various foodstuffs from Vietnam, relatively high levels of PCBs were noted and the older trans-formers and capacitors imported from Russia and France were

implicated as possible sources (Kannan et al 1992a).

To understand the magnitude of contamination in human breast milk from dumping sites in India, Cambodia, Vietnam, and the Philippines, TEQ levels were compared with values for human breast milk from the general populations of other coun-tries since 1990, which were selected from publications in which concentrations of all the isomers were reported (Figure 2) Because international TEFs were mainly used to calculate TEQs, the reported data were recalculated using WHO TEFs for comparison The levels of TEQs in human breast milk from India (38 pg TEQs/g lipid wt) were comparable with or higher

than those from developed countries (Becher et al 1995; Dawailly et al 1992; Fu¨rst et al 1994; Gonzalez et al 1996; Kiviranta et al 1999; Liem et al 1995; Ministry of Health, Labor, and Welfare 1999; Schecter et al 1990a; Schuhmacher

et al 1999) and Russia (Schecter et al 1990b) This suggests

that residents living near the dumping site in India have been exposed to comparable levels of DRCs as the general popula-tions of developed countries In contrast, the levels of TEQs in human breast milk from Cambodia (9.2 pg TEQs/g lipid wt), the Philippines (12 pg TEQs/g lipid wt), and Vietnam (13 pg TEQs/g lipid wt) were lower than those from developed tries and comparable with those from other developing

coun-tries (Schecter et al 1990a, 1994; Paumgartten et al 2000) In

this international comparison, however, there were some un-certainties such as age and parity of the mothers, sampling period, sample number, and accuracy of the analytical tech-niques involved In addition, very few data are available on

non- and mono-ortho PCBs in the literature Because of such

uncertainties, it was difficult to draw any firm conclusions using the information shown in Figure 2 However, the obser-vation that TEQs of DRCs in human breast milk from the dumping site of India were comparable with or higher than those from some developed countries, including Japan, is

note-Table 1 Details of the breast milk donors

Characteristics

Mean (Range)

Dumping Site

(n⫽ 11) Reference Site(n⫽ 8) Dumping Site(n⫽ 19) Reference Site(n⫽ 16) Dumping Site(n⫽ 8) Reference Site(n⫽ 10) Dumping Site(n⫽ 9)

Age (yr) 25.2 (20–34) 23.9 (19–29) 29.1 (19–46) 26.7 (18–38) 31.6 (22–42) 27.9 (22–34) 27.0 (17–44) Height (cm) Not measured Not measured 154.7 (144–165) 154.9 (145–160) 154.8 (145–162) 156.4 (150–160) 158.3 (152–165) Weight (kg) Not measured Not measured 50.7 (39–68) 51.6 (41–60) 48.6 (45–60) 50.7 (40–60) 47.1 (35–68) BMI (kg/m 2 ) Not calculated Not calculated 21.1 (17.5–27.2) 21.5 (17.9–26.7) 20.3 (16.4–22.9) 20.7 (16.6–25.0) 18.8 (15.1–27.4)

No previous

deliveries

BMI: Body mass index.

a (range)

Dioxins 2,3,7,8-T

O8

Furans 2,3,7,8-T

O8

T4

P5

H6

H7

O8

worthy In developed countries, it is claimed that the residue

levels of DRCs in human breast milk decreased recently

(La-Kind et al 2001) because of the installation of highly efficient

incinerators and strict regulations on the production and usage

Fig 1 Comparison of the concentrations of DRCs in human breast milk from dumping (d) and reference (r) sites The circles and bars represent

mean and range values, respectively.ⴱp ⬍ 0.05 ⴱⴱp ⬍ 0.01

Fig 2 Comparison of TEQ levels in

human breast milk from dumping sites in Asian developing countries with those from other countries a Pre-sent study, b PCDD/DFs only;

1Schecter et al (1994);2 Paumgartten

et al (2000);3Schecter et al (1990a);

4Schecter et al (1990b);5

Schuhma-cher et al (1999);6 Ministry of Health Labor, and Welfare (1999);

7Gonzalez et al (1996);8Dewailly et

al (1992);9Fu¨rst et al (1994);

10Becher et al (1995);11Kiviranta et

al (1999);12Liem et al (1995)

Ref-erence data were recalculated by

us-ing WHO TEF (Van den Berg et al.

1998)

of various chemicals In contrast, in Asian developing countries

it can be anticipated that the pollution caused by DRCs may

increase further, and hence residue levels in human breast milk

may also increase in the future because the release of these

contaminants are not at all controlled, even now

Variation with Number of Deliveries and Risk Assessment

for Infants

It has been reported that concentrations of DRCs varied with

factors such as number of deliveries by mothers and extent of

breast-feeding (Beck et al 1994; Bates et al 1994; Hooper et

al 1999; Iida et al 1999; LaKind et al 2001; Schecter et al.

1998) In this study, we examined the relationship between

number of deliveries by the mothers and TEQs in human breast

milk from India, Cambodia, Vietnam, and the Philippines In

India, data from the dumping site only were examined because

significantly different levels of TEQs were observed between

the dumping and reference sites (Figure 1) TEQ levels in

human breast milk from the dumping site in India tended to

decrease with an increase in the number of deliveries (Figure

3) In fact, one of the primipara donors had exceptionally high

level of TEQs (140 pg/g lipid wt) In contrast, in Cambodia,

Vietnam, and the Philippines no significant relationship was

observed (Figure 3) These results suggested that the mothers

who have been exposed to relatively high levels of DRCs, as in

case of the mothers in developed countries, may transfer higher

amounts of these contaminants to the first infant through

breast-feeding than to infants born afterward fed in the same

manner, hence the first-born children might be at higher risk

from DRCs In previous investigations in developed countries

(Beck et al 1994; Iida et al 1999), it was reported that

concentrations of DRCs in human breast milk from primiparas

were higher than those in multiparas Beck et al (1994) found

that TEQ levels in German multiparas with three deliveries

were 43% lower than those in primiparas

To understand the magnitude of exposure to DRCs by

in-fants, we estimated daily intake (DI) from the levels of TEQs

in human breast milk observed in this study based on the

assumption that an infant ingests 700 ml milk/d and that the

weight of an infant is 5 kg (Hooper et al 1997) As expected,

the highest DI (500 pg TEQs/kg/d) was observed in the infants

of mothers living near the dumping site in India, and DIs in all

the cases exceeded 1 to 4 pg TEQs/kg/d, which is the tolerable

daily intake (TDI) proposed by WHO (Van Leeuwen et al.

2000) (Table 3) It has been reported that 1–3-month-old

in-fants absorb ⬎90% of 2,3,7,8-substituted PCDD/DF isomers

(except hepta- and octa-CDD/Fs) contained in their mother’s

milk (Dahl et al 1995; Korer et al 1993; McLachlan 1993;

Pluim et al 1993), hence they may be exposed to relatively

high levels of DRCs during this period If an infant’s

consump-tion is to remain below the TDI, a decrease in human breast

milk contamination must be sought In case of the subjects in

this study, if the daily intake of TEQs should be⬍4 pg, an

infant can ingest only 1% to 57% of the necessary amount of

milk (700 ml/d) to stay below the TDI value Furthermore, in

some Asian developing countries, water used for many formula

milk preparations may contain various infectious organisms

and environmental contaminants, which may adversely affect

infant health (Carpenter et al 2000) Considering all of the

above, breast-feeding cannot be avoided for the infants in Asian developing countries and so finding ways to decrease levels of DRCs in human breast milk has become mandatory to save infants from possible toxic effects Because the majority

of DRCs in human breast milk comes through mobilization from adipose tissue, and only a small amount (14%) comes from dietary sources (Koppe 1995), it is imperative that dioxin exposure should be decreased using urgent control and regu-lation of DRC pollution sources

Residue Levels in Bovine Milk—A Potential Source

Although residue levels of DRCs in soils collected from open dumping sites in Asian developing countries were apparently

greater than those from the control sites (Minh et al 2003), the

levels of DRCs in human breast milk from residents living near the dumping sites in Cambodia and Vietnam were not signif-icantly higher than those from reference sites However, resi-due levels of these contaminants in Indian samples around the dumping site were notably higher (Table 2 and Figure 1) This implies that residents living in proximity to the dumping sites

in Cambodia and Vietnam have not been greatly exposed to DRCs originating from the dumping sites

In the case of humans, it has been reported that the food chain, especially meat and dairy products, accounts for 98.8%

of exposure to dioxins and that water, soil, and air are not major sources (Travis and Hattemer-Frey 1991) In addition, it has been suggested that residue levels and composition of DRCs in

human tissues generally reflect those in ingested foods (Cole et

al 1997; Domingo et al 1999; Fiedler et al 1997; Goldman et

al 2000; Hooper et al 1999; Johansen et al 1996) In India,

buffaloes reared near the waste-dumping site mainly feed on dumped leftovers, whereas cows reared near the dumping site feed mainly on pastures In addition, residents living near the dumping site constantly drink the milk collected from these bovines In contrast, in Cambodia, Vietnam, and the Philip-pines, livestock such as buffaloes and cows are not reared near dumping sites To elucidate whether bovine milk is a potential source of DRCs for residents in proximity to the dumping site

in India, residue levels of these contaminants in buffalo milk and cow milk collected in and around the site were estimated

In all of the bovine milk samples analyzed, DRCs were detected (Table 4), demonstrating that bovines in India have been exposed to these contaminants Concentrations of DRCs

in buffalo milk collected near the dumping site (mean TEQs 16 pg/g lipid wt) were higher than those in cow milk collected near the dumping site (mean TEQs 7.3 pg/g lipid wt) and in bovine milk collected from reference sites (mean TEQs buffalo 2.1 pg/g lipid wt and cow 3.8 pg/g lipid wt) (Table 4) This indicates that buffaloes feeding at the dumping site in India consume greater amounts of DRCs through contaminated soils, waterweeds, and leftovers In addition, TEQ levels in cow milk collected near the dumping site were slightly higher than those from reference sites, implying that pastures near the dumping site are also contaminated by DRCs, probably formed in the dumping site and transferred via runoff or/and atmosphere to these pastures

Compositions of PCDD/DFs in bovine milk showed

differ-ent patterns depending on the area of collection In bovine milk

collected from the dumping site, some low chlorinated

conge-ners—such as 2,3,7,8-T4CDD; 1,2,3,7,8-P5CDD; and

2,3,4,7,8-P5CDF—were predominant, whereas the residue

lev-els of 1,2,3,4,6,7,8-HCDD and OCDD were relatively higher

in those from reference sites (Figure 4) Furthermore, higher levels of T4, P5, and H6CDD/DFs were noted in bovine milk from the dumping site than that from reference sites (Figure 5)

In particular, concentration ratios of these congeners of dioxins and furans observed in buffalo milk were higher than in cow milk, indicating that notable sources of T4, P5, and H6CDD/ DFs are present in the dumping site, and hence buffalos feeding there have been exposed to these contaminants Higher ratio of

T4, P5, and H6CDD/DFs were observed in cow milk collected around the dumping site than those from reference sites, im-plying that pastures near the dumping site have been contam-inated by these compounds In a previous study, we reported that concentrations of T4, P5, and H6CDD/DFs in soils from the Indian dumping site were higher than those from reference sites

(Minh et al 2003) These results indicate that T4, P5, and

H6CDD/DFs are formed in the dumping site in India, possibly

by combustion of municipal wastes, and that buffaloes and cows feeding in and around these areas accumulate higher amounts of these compounds through contaminated soils, wa-terweeds, leftovers and pastures

In soils collected in and around the Indian dumping site, however, 1,2,3,4,6,7,8-H7CDD and O8CDD were predominant

among all the 2,3,7,8-substituted congeners (Minh et al 2003).

In addition, Thomas et al (2002) reported that the estimated

Table 3 Estimated daily intakes of TEQs from human breast

milk a

Country

Daily Intake (pg TEQs/kg/d)

India

Cambodia

Vietnam

Philippines

a Estimated based on the assumption that an infant ingests 700 ml

milk/d and that the weight of an infant is 5 kg (Hooper et al 1997).

Fig 3 Relationship between concentrations of

TEQs in human breast milk from Asian developing countries and number of previous deliveries (A) Indian dumping site, (B) Cambodian sites, (C) Vietnamese sites, (D) Philippines dumping sites.

ⴱp ⬍ 0.05

average percent contribution of high-chlorinated DD/DFs

found in pastures affected through soil particle adhesion was

greater than that of low-chlorinated DD/DFs and that during

summer—the period of high atmospheric temperature— uptake

of PCDD/DFs by pasture from vapor phase increased with

increasing degree of chlorination (increasing KOA)

Further-more, Alcock et al (2002) showed that PCDD/DF pollution in

cow milk reflected not only intake from pastures but also from

ingestion of contaminated soils These facts show that intake of high-chlorinated DD/DFs—such as 1,2,3,4,6,7,8-H7CDD and

O8CDD— by buffaloes and cows in and around the dumping site in India are greater than low-chlorinated DD/DFs How-ever, in bovine milk from the dumping site, higher levels of low-chlorinated DD/DFs—such as T4, P5, and H6CDD/DFs— than high-chlorinated DD/DFs were observed This indicates that buffaloes and cows in and around the dumping site in India

Table 4 Meana (range) concentrations (pg/g lipid wt) of DRCs in bovine milk from dumping and reference sites in India

Compound b

Dioxins

1,2,3,7,8,9-H6CDD 0.87 (0.35–1.5) 0.52 (0.51–0.54) 0.23 (0.10–0.37) 0.28 (0.11–0.46)

Furans

1,2,3,7,8-P5CDF 0.35 (0.14–0.60) 0.22 (0.21–0.22) 0.065 ( ⬍0.052–0.091) 0.081 ( ⬍0.065–0.17)

1,2,3,4,6,7,8-H7CDF 0.62 (0.34–0.96) 0.29 (0.26–0.31) 0.32 (0.19–0.46) 0.58 (0.18–0.91)

Non-ortho PCBs

Mono-ortho PCBs

2,3 ⬘,4,4⬘,5-P 5 CB (118) 5300 (3300–6500) 1400 (1300–1400) 420 (270–560) 2400 (230–3800)

Total mono-ortho PCBs 8800 (5500–11,000) 2300 (2200–2400) 750 (500–1000) 3600 (410–5600)

Mono-ortho PCBs-TEQsa 1.3 (0.75–1.6) 0.33 (0.32–0.35) 0.12 (0.075–0.16) 0.56 (0.064–0.91)

a The concentrations below detection limits were treated as zero for calculation of arithmetic mean and TEQ values.

b T4: tetra, P5: penta, H6: hexa, H7: hepta, and O8: octa.

c All the samples were below detection limit.

DRCs: dioxins and related compounds, PCBs: polychlorinated biphenyls, PCDDs: polychlorinated dibenzo-p-dioxins, PCDFs: polychlorinated

dibenzofurans, TEQ: toxic equivalent quantity.

preferentially transfer greater amounts of low-chlorinated DD/

DFs to their milk Fries et al (1999, 2002) investigated the

mass balance of PCDD/DFs in cows after administration of

pentachlorophenol-treated wood and reported that transfer to

milk and storage in body fat increased with decreasing degree

of chlorination, whereas excretion in feces increased with

in-creasing degree of chlorination These observations indicate

notable pollution sources of low-chlorinated DD/DFs such as

T4, P5, and H6CDD/DFs in and around the Indian dumping site

Buffaloes and cows feeding there accumulate high amounts of

these contaminants and transfer them to their milk, whereas in

reference sites, comparatively low levels of pollution sources

of T4, P5, and H6CDD/DFs were present

Levels of non-ortho PCBs in bovine milk collected from the

dumping site were also higher than those from reference sites,

with the exception of H6CB 169 in cow milk (Table 4 and

Figure 4) This indicates that notable pollution sources of

non-ortho PCBs are present at the dumping site in India and

that buffaloes and cows obtain greater amounts of these

con-taminants via feeding, especially T4CB 81 and P5CB 126, and

then transfer them to their milk In addition, levels of

mono-ortho PCBs in buffalo milk collected from the dumping site

were higher than those from reference sites (Table 4 and Figure

4), whereas levels of mono-ortho PCBs in cow milk collected

from the dumping site were comparable with or lower than

those from reference sites (Table 4 and Figure 4) Although a clear and plausible reason could not be assigned, it seems that some cows feeding near reference sites in India might have

been exposed to levels of mono-ortho PCBs comparable with those near the dumping site Non-ortho PCBs are formed by combustion of municipal wastes (Sakai et al 2001), whereas they are poorly included in technical PCB mixtures (Schulz et

al 1989; Takasuga et al 1995) In contrast, mono-ortho PCBs are abundant in technical PCB mixtures (Schulz et al 1989; Takasuga et al 1995) although less formed by combustion process (Sakai et al 2001) Further investigations regarding

pollution and transfer of coplanar PCBs in bovine milk are needed

Concentrations of all the DRCs in buffalo milk collected from the dumping site were higher than that collected from reference sites, indicating that daily intake of bovine milk by residents living near the dumping site in India is one of the possible reasons why TEQ levels in human breast milk col-lected from the dumping site were significantly higher than that collected from reference sites Furthermore, it was observed that not only buffaloes but cows near the dumping site also transfer higher amounts of low-chlorinated DD/DFs and

non-ortho PCBs to their milk than those near reference sites,

implying that residents living near the dumping site in India are

at greater health risk because of these highly toxic

contami-Fig 4 Compositions of PCDD/DFs in bovine milk collected from dumping and reference sites in India (A) Buffalo milk from the dumping site.

(B) Cow milk from the dumping site (C) Buffalo milk from reference sites (D) Cow milk from reference sites

nants In India, consumption of dairy products is generally

higher than in other countries, and average consumption of

milk in India by one person per day increased from 135 g in

1980 to 176 g in 1990 (John et al 2001) The residents near the

dumping site in India constantly drink the milk collected from

buffaloes and cows reared near by Assuming that an adult

weighing 60 kg drinks 176 g of the buffalo or cow milk

investigated in this study/d, estimated daily intake of TEQs

from bovine milk collected from the dumping site would range

from 1 to 4 pg TEQs/kg/d—the TDI range proposed by WHO

(Van Leeuwen et al 2000)—and only in one buffalo milk

sample did the value exceed the TDI (Figure 6) Although the

values are within the TDI, the residents living near the

dump-ing site in India are exposed to considerably high levels of

DRCs and hence may be at greater risk of exposure to these

contaminants by way of bovine milk

Conclusion

To our knowledge, this is the first comprehensive study on

exposure to DRCs in residents living near open dumping sites

of municipal waste in India, Cambodia, Vietnam, and the Philippines In this study, we showed that residents living near dumping sites in these Asian developing countries have been exposed to DRCs In particular, our results suggest that resi-dents near the dumping site in India have been exposed to relatively high levels of these contaminants, possibly through intake of bovine milk In addition, TEQ levels in human breast milk from residents living near the Indian dumping site tended

to decrease with an increase in the number of previous deliv-eries by the mothers, suggesting that the primiparae living there transfer greater amounts of these contaminants to their infants through breast-feeding than do multiparae This implies that first-born infants might be at higher risk from DRCs In open dumping sites of municipal waste in Asian developing coun-tries, it is anticipated that pollution by DRCs may further increase and that residue levels in human breast milk may increase in the future because even now the sources of these contaminants are not regulated at all Control measures to regulate the pollution sources of DRCs in open dumping sites

in Asian developing countries are urgently needed Further investigations on the effect of pollution and temporal trends on

Fig 5 Concentration ratios

(dump-ing site to reference sites) of DRCs

in bovine milk collected from dump-ing and reference sites in India DRC: Dioxins and related com-pounds NC: not calculated because values were below detection limits