Long-term biomonitoring of polychlorinated biphenyls and organochlorine pesticides in human milk from mothers living in northern Germany

ARTICLE IN PRESS

Int. J. Hyg. Environ. Health 211 (2008) 624–638 www.elsevier.de/ijheh

Long-term biomonitoring of polychlorinated biphenyls and organochlorine pesticides in human milk from mothers living in northern Germany Bjo¨rn P. Zietz, Michael Hoopmann, Markus Funcke, Rene´ Huppmann, Roland Suchenwirth, Edith Gierden Governmental Institute of Public Health of Lower Saxony (Niedersa¨chsisches Landesgesundheitsamt), Division of Environmental Medicine and Environmental Epidemiology, Roesebeckstraße 4-6, D-30449 Hannover, Germany Received 11 January 2008; received in revised form 25 March 2008; accepted 1 April 2008

Abstract Polychlorinated biphenyls (PCBs) and organochlorine pesticides are persistent organic pollutants that have a widespread distribution in the environment. Human biomonitoring is a suitable tool to assess the burden of humans with these substances. Over a time span of 8 years, a free analysis of their milk was offered to lactating mothers residing in the state of Lower Saxony, Germany. The human milk was analyzed for a number of organic chemicals including polychlorinated biphenyls (PCBs), 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (DDT), hexachlorobenzene (HCB) and b-hexachlorocyclohexane (b-HCH). Factors that may influence these levels were investigated using a questionnaire. In total, 4314 samples were collected in the years 1999–2006 and analyzed for their content of these persistent organic pollutants (POPs). A clear downward trend of median total PCB, DDT, b-HCH and HCB values in all participants and also in different selected subgroups could be observed. The median values of calculated total PCB in the year 2006 including all participants was 0.1825 mg/kg lipid, that of DDT 0.0815 mg/kg lipid, b-HCH 0.0116 mg/ kg lipid and of HCB 0.0229 mg/kg lipid. There were reductions between 40.9% and 47.1% compared to the year 1999. Among other influencing factors, median concentrations of total PCB, DDT, b-HCH and HCB showed a clear rise with increasing age of mothers whereas an increasing number of breastfed infants per mother led to a decrease. The proportions of other measured substances exceeding limits of quantification were as follows: dieldrin 68.6%, a-HCH 1.3%, g-HCH 60.1%, heptachlor epoxide 41.5%, musk xylene 15.6%, musk ambrette 0.4%. We conclude that the known declining trend of important xenobiotic substances in human milk of German mothers has continued. r 2008 Elsevier GmbH. All rights reserved. Keywords: Persistent organic pollutants; Human milk; Biomonitoring; Polychlorinated biphenyls; PCB; Organochlorine pesticides; DDT; Lower saxony; Germany

Introduction

Corresponding author. Tel.: +49 511 4505 328; fax: +49 511 4505 140. E-mail address: [email protected] (B.P. Zietz).

1438-4639/$ - see front matter r 2008 Elsevier GmbH. All rights reserved. doi:10.1016/j.ijheh.2008.04.001

Human milk is the natural food for infants having the optimal balance of fats, carbohydrates, and proteins. It provides many immunologic, developmental, psychological, economic, and practical advantages (Pronczuk et al., 2002, 2004; Solomon and Weiss, 2002). On the

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

other hand, there is concern about the potential risks posed by the presence of chemical contaminants in human milk. Many nonpharmaceutical xenobiotics, particularly those that are lipophilic such as polyhalogenated chemicals, can occur in breast milk (Solomon and Weiss, 2002; LaKind et al., 2001). Polyhalogenated chemicals tend to degrade slowly in the environment, to bioaccumulate in the food chain, and to have long halflives in human beings. Different adverse health and reproductive outcomes have been attributed to these chemicals in animals and in also humans (Needham and Wang, 2002; Solomon and Weiss, 2002; Landrigan et al., 2002). In the following, some basic facts of the most important substances that were measured in this study are described. It should be noted that these anthropogenic substances are ubiquitous contaminants in most environmental and biotic compartments (AMAP, 2004). PCBs have been used for decades in different applications such as in plasticizers, surface coatings, inks, adhesives, flame retardants and paints. Because of their general chemical inertness and heat stability, they have been used in dielectric fluids in transformers and capacitors (WHO, 2003a; ATSDR, 2000). The environmental concentrations of the organochlorine insecticide DDT (1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane) and its metabolites have been decreasing since the ban on DDT was instituted in the United States and most of the world in 1972. The ban excludes selected uses in some parts of the world especially for controlling malaria vectors. The substances 1,1-dichloro-2,2-bis (p-chlorophenyl)ethane (DDD) and 1,1-dichloro-2,2bis(p-chlorophenyl)ethylene (DDE) are both degradation products of DDT. DDD was also manufactured and used as an insecticide (ATSDR, 2002a). Hexachlorocyclohexane (HCH) is a synthetic chemical that exists in eight isomers. g-HCH (lindane) was produced and used as an insecticide in huge quantities and is still available in prescription medicines to treat ectoparasites. Technical-grade HCH is a mixture of several isomers of HCH and was once also used as an insecticide (ATSDR, 2005). The fully chlorinated hydrocarbon industrial chemical hexachlorobenzene (HCB) was used as a pesticide and to make pyrotechnic materials, and synthetic rubber. It is also an unintentional by-product or a minor part of other chemical products (ATSDR, 2002b). Possible health effects of low-level chronic exposure with these substances were reviewed in detail by the WHO (2003b). Human biomonitoring can be an efficient and costeffective tool to evaluate internal exposure of humans with different chemical substances. In many cases it can be an integral part of risk assessment and risk management of a substance (Angerer et al., 2007; Jakubowski and Trzcinka-Ochocka, 2005). Some persistent organic pollutants have been measured in human milk since

625

several decades, but only a few countries have systematic human milk monitoring programs that have tested considerable numbers of women over time using consistent sampling methods (UNEP, 2004; Solomon and Weiss, 2002). Aim of the human milk monitoring program in Lower Saxony was a biomonitoring of important persistent organic pollutants (POPs) offered to all interested mothers residing in this federal state of Germany. Beside the individual information and assessment for the mothers, the data was collected to monitor the general contamination trend of human milk in Lower Saxony with these substances.

Materials and methods The human milk monitoring program for mothers living in Lower Saxony, Germany, was based on voluntary donors. Chemical analysis of the milk was free of charge for the mothers. Interested mothers were requested to sign a written declaration of consent. Information about the monitoring program was distributed via local physicians (gynecologists, pediatricians), midwives and obstetrical wards using leaflets. Additionally, in the last years detailed information about the program was posted on the department’s homepage. Regularly press announcements of the ongoing project and selected results were made by the Ministry of Social Affairs, Women, Family and Health of Lower Saxony. Sampling kits, comprising a cleaned 250 ml glass bottle with screw cap (NUK/MAPA, Zeven, Germany), a sampling instruction, a cooling gel package and a questionnaire, were sent to those mothers willing to participate. The return of the package by mail was free for the mothers. Participants were asked to collect at least 100 ml of milk, but lower volumes were also accepted. The mothers were requested to wash their hands thoroughly only using a non-perfumed soap before sampling. The breast should only be rinsed with copious amounts of water. The milk should be expressed either manually or by means of a cleaned breast pump, if possible after finishing with breastfeeding. A questionnaire served to asses factors of interest regarding personal characteristics of the mother and the infant. Additionally, different lifestyle factors (e.g. smoking, eating habits, profession, chemical/industrial plants in proximity to home, home moving) had been asked. Samples were collected between January 1999 and December 2006. Milk samples were stored in their original glass sample bottles at a temperature of 18 1C. The fat content of the breast milk samples was determined by the Gerber method following ISO 488:1983-12. First 10 ml H2SO4 (p.a., 91%; Fluka/ Riedel-de Hae¨n, Seelze, Germany) were filled into a

ARTICLE IN PRESS 626

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

Gerber butyrometer (No. 3134; Funke-Gerber, Berlin, Germany). Then a 10.75 ml specimen was delivered by means of a milk pipette (No. 3431; Funke-Gerber). Finally, 1 ml isoamyl alcohol (Merck, Darmstadt, Germany) was added, the butyrometer was placed in a casing and the layers were mixed. The butyrometer in the casing was then centrifuged at 350  g and 65 1C for 10 min (centrifuge Nova-Safety, Funke-Gerber). The fat content was read off the scale of the butyrometer in mass percent. Centrifugation and reading off was repeated once. To extract persistent organic pollutants 10 g of the milk specimens were homogenized with 15 g of silica gel 60 (Merck), activated at 450 1C for 3 h. The mixture was then filled into glass columns (i.d. 2.2 cm; length 52 cm) preloaded with 15 g of freshly deactivated silica gel 60 (Merck). Silica gel was deactivated by adding 10% water to the activated silica gel. Analytes were eluted using a mixture of petroleum ether (boiling range 40–60 1C) and dichloromethane (80:20) (both Rotisolv Pestilyse; Roth, Karlsruhe, Germany or before Promochem, Wesel, Germany). The extraction process follows the method described by Steinwandter (1982). The eluate was evaporated to near dryness and the residues dissolved in n-hexane (Labscan, Dublin, Ireland or before Promochem). As an internal standard 50 ml aldrin 10 ng/ml dissolved in cyclohexane (Dr. Ehrenstorfer, Augsburg, Germany) was used. A custom-made LGC Promochem standard mix (22 components) (LGC Promochem, Wesel, Germany) was utilized as an external standard. POP concentrations were determined using a gas chromatograph Hewlett-Packard (now Agilent Technologies; Waldbronn, Germany) GC 6890 with microelectron capture detector (mECD) and a cold injection system (CIS4; GERSTEL, Mu¨lheim an der Ruhr, Germany). The used capillary column was HP-5MS (Agilent; length 60 m, i.d. 0.25 mm, film thickness 0.25 mm). Carrier gas was helium 5.0 with a constant nitrogen 5.0 makeup flow for the detector (both Linde AG, Wiesbaden, Germany). The gas chromatography method matches DIN 38407 part 2. The following substances were measured: PCB congeners 28, 52, 101, 138, 153, 180; a-/b-/g-HCH, HCB, cis-/trans-heptachlor epoxide (HCE [CAS 1024-57-3]), DDT (as a sum of: o,p0 -/p,p0 -DDE, o,p0 -/p,p0 -DDD, o,p0 -/p,p0 -DDT), dieldrin and synthetic musk fragrances (musk xylene [CAS 81-15-2], musk ambrette [CAS 83-66-9]). Limit of quantification (LOQ) of these measured substances was 0.5 pg/ml each. Concentrations of these lipophilic substances were specified as a ratio to the fat content of the human milk samples (mg/kg milk fat). A total PCB content was calculated by multiplying the sum of the concentrations of PCB congeners 138, 153 and 180 with the factor of 1.64 (Schulte and Malisch, 1984). These are the most characteristic congeners representing about 61% of the PCBs in biological samples.

The laboratory has successfully participated in regular external quality assurance and is now accredited according to ISO/IEC 17025 by the AKS Hannover, Germany. Milk samples were regularly exchanged with other laboratories for quality control purposes. Data management and statistical analysis of the collected data was performed with the software MS Excel 2000s (Microsoft, Unterschleißheim, Germany); MS Access 2000s (Microsoft); WinStat 2007.1s (R. Fitch Software, Bad Krozingen, Germany) and Systat 11s (Systat Software, San Jose´, USA). Measured concentrations in human milk below the LOQ were included in the data as zero (deviating from other investigations, cf. discussion below). Only values of participants that also returned a questionnaire were included in the data analysis.

Results During the 8 years of monitoring, a total of 4314 samples were analyzed. Of these, 361 samples (8.4%) were returned from mothers that were born outside of Germany. A few samples from mothers living outside Lower Saxony were analyzed (e.g. some that had moved after agreeing to participate). The number of participating mothers was between 280 in the year 2002 and 791 in 2000 (1999: 739; 2001: n ¼ 345; 2003: n ¼ 662; 2004: n ¼ 405; 2005: n ¼ 565; 2006: n ¼ 523). The age distribution of the participating mothers can be found in Fig. 1. Compared to the age distribution of mothers in Lower Saxony that gave birth to a child in the year 2005, the distribution of program participants is shifted to a higher age with a peak at the year of 31. Younger mothers, especially those under the age of 26, are therefore underrepresented. Median age of participating mothers was 32 years (mean: 31.55 years). Mean number of breastfed children was 1.60. Type of diet was gathered in a condensed variable between 2002 and 2005 (n ¼ 1911) and had the following results: 94.6% mixed diet, 5.2% vegetarian diet, 0.3% vegan diet. Mean BMI of the mothers at the time of sampling was 24.3 kg/m2 (median 23.6). Over the years there was an increase in median BMI values from 23.2 in 1999 to 23.9 in 2006. Some 58.3% of the mothers stated that they had never smoked, 36.3% quoted to have quit smoking and 4.9% declared that they were smokers (no statement 0.6%). The median fat content of the human milk was 3.90% (mean: 4.02%; 95th percentile: 6.90%; 5th percentile: 1.50%). The relative distribution of the fat content in the milk samples can be found in Fig. 2. An overview of descriptive statistical values of measured POPs can be found in Table 1A including all participants and in Table 1B, including participants of the year 2006.

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

627

10 Monitoring program 9

Lower Saxony - year 2005

8

Relative frequency in %

7 6 5 4 3 2 1 0 until 15

17

19

21

23

25

27

29 31 Age classes

33

35

37

39

41

43

45

Fig. 1. Age distribution (years) of all program participants (years 1999–2006) in comparison to the age distribution of mothers in Lower Saxony that gave birth to a child in the year 2005 (Data source: NLS, 2006).

Concerning outliers of measured substances there were two mothers with remarkably high DDT concentrations. Both were born in South America (Chile, Ecuador) and had lived there for many years. Both participated in the year 2000. The mother with the highest DDT value found in this investigation (18.66 mg/kg milk lipid) was 27 years old at the time of sampling, born in Ecuador and she breastfed her first child. A control sample about 2 months later contained 15.21 mg/kg lipid. She consumed a mixed diet and moved to Germany in the year before sampling. The other case with a DDT value of 15.44 mg/kg lipid was a 38-year-old mother born in Chile who lived in Germany for 8 years before sampling. She had breastfed one child before. Both mothers stated that they had not worked in the agricultural sector.

Trend of contamination and situation in the years 2005 and 2006 A clear downward trend of median total PCB, DDT, b-HCH and HCB values in all participants and all subgroups of interest could be observed (from 1999 to 2006). Comparing the median values of the years 1999 and 2006 including all participants, there was a reduction of total PCB values of 43.2% to 0.1825 mg/kg lipid.

In the same time, there was a reduction of DDT of 40.9% to 0.0815 mg/kg lipid, of b-HCH of 47.1% to 0.0116 mg/kg lipid and of HCB of 42.7% to 0.0229 mg/kg lipid, each including all participants in that year. The time trend for each of these substances in all participants and different subgroups can be found in Fig. 3A–D. Median values of PCB congeners 28, 52 and 101 were below the LOQ in all cases. Descriptive statistical information on the most recent contamination levels of the measured substances are depicted in a Boxand-Whisker plot in Fig. 4 (participants of the years 2005 and 2006). To homogenize the collective, only mothers born in Germany and only first breastfed children were included (subgroup 2). Because the measured values of total PCB, DDT, b-HCH and HCB (all participants) were not normally distributed in the test of Kolmogorov–Smirnov, analytical results of participants from the years 1999 and 2006 were compared using the U-test (Mann–Whitney). All of these POPs had a highly significant difference between these years (po0.001). Taking into account only mothers born in Germany and only the first breastfed children the difference was still highly significant (po0.001). Forming a subgroup that is similar to the cohort inclusion criteria of the Fourth WHO coordinated survey of human milk for persistent organic pollutants (WHO,

ARTICLE IN PRESS 628

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

14

12

Relative frequency in %

10

8

6

4

2

e

10

10

5 9.

9

8

5 8.

7.

5

7

6

5 6.

5.

5

5

5 4.

4

5

3

3.

2.

5

2

5 1.

ab

ov

up

to

1

0

Fat content - classes in %

Fig. 2. Relative frequency of fat content in human milk samples in mass percent including all participants (n ¼ 4314).

Table 1A.

Basic descriptive statistical values including all participants (n ¼ 4314)

Substance

Mean

Median

75th percentile

95th percentile

Maximum

Standard deviation

Total PCB DDT b-HCH HCB Dieldrin g-HCH HCE

0.2784 0.1688 0.0268 0.0384 0.0038 0.0037 0.0043

0.251 0.113 0.017 0.031 0.004 0.003 oLOQ

0.348 0.176 0.025 0.044 0.006 0.006 0.008

0.550 0.430 0.052 0.084 0.009 0.011 0.016

2.36 18.66 1.597 0.879 0.031 0.079 0.098

0.1503 0.4131 0.0608 0.0333 0.0033 0.0048 0.0063

oLOQ, below limit of quantification. Units are in mg/kg lipid.

Table 1B.

Basic descriptive statistical values including participants of the year 2006 (n ¼ 523)

Substance

Mean

Median

75th percentile

95th percentile

Maximum

Standard deviation

Total PCB DDT b-HCH HCB Dieldrin g-HCH HCE

0.2022 0.1257 0.0209 0.0268 0.00018 0.00166 0.00011

0.183 0.081 0.012 0.023 oLOQ oLOQ oLOQ

0.252 0.132 0.017 0.031 oLOQ 0.0036 oLOQ

0.413 0.342 0.056 0.048 oLOQ 0.0073 oLOQ

1.40 3.14 0.783 0.281 0.0165 0.0196 0.0123

0.1139 0.1840 0.0504 0.0191 0.00128 0.00302 0.00103

oLOQ: below limit of quantification. Units are in mg/kg lipid.

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

629

The proportions of other measured substances exceeding LOQ including all participants were as follows: dieldrin 68.6%, a-HCH 1.3%, g-HCH 60.1%, heptachlor epoxide 41.5%, musk xylene 15.6%, musk ambrette 0.4%. A clear time trend based on the 95th percentiles for other measured substances in subgroup 2 that could be detected at least regularly (Dieldrin, g-HCH, musk xylene and heptachlor epoxide (HCE)) could not be found. There was no straight downward trend concerning these chemicals. In the year 2006, three substances had a 95th percentile below the LOQ (Dieldrin, musk xylene, HCE). It has to be taken into account that these trends were founded on a much smaller data basis than the median trends, because most

2007), results in lower median values of the mentioned substances. The subgroup (n ¼ 35) was formed using data from mothers with an age below 30 years that participated in the year 2006, that were born in Germany, breastfed their first child and took the sample between the third and eights week of breastfeeding. In this group, median concentrations and standard deviations were as follows: total PCB 0.1336 mg/kg lipid (mean 0.1472, S.D. 0.0691), DDT 0.0600 mg/kg lipid (mean 0.0841, S.D. 0.0671), b-HCH 0.0089 mg/kg lipid (mean 0.0102, S.D. 0.0051) and HCB 0.0196 mg/kg lipid (mean 0.0212, S.D. 0.0076). The much younger age of the mothers in this subgroup is probably one of the main factors contributing to the lower concentrations. 0.35 0.33

PCB in mg/kg lipid

0.31 0.29 0.27 0.25 0.23 All participants

0.21

Subgroup 1

0.19

Subgroup 2 Subgroup 3

0.17 0.15 1999

2000

2001

2002

2003

2004

2005

2006

2003 Year

2004

2005

2006

Year 0.15 0.14

DDT in mg/kg lipid

0.13 0.12 0.11 0.1 All participants Subgroup 1

0.09

Subgroup 2 Subgroup 3

0.08 0.07 1999

2000

2001

2002

Fig. 3. (A–D) Time trend of median total PCB, DDT, b-HCH and HCB values of all participants and different subgroups. Subgroup 1: only participants born in Germany [n ¼ 3947]. Subgroup 2: only participants born in Germany and only the first breastfed children [n ¼ 2119]. Subgroup 3: only participants born in Germany with first breastfed children and age of mother only between 27 and 34 years [n ¼ 1437]. Scale origin of y-axis was modified to better demonstrate the trend.

ARTICLE IN PRESS 630

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

0.026

β-HCH in mg/kg lipid

0.024 0.022 0.02 0.018 0.016 0.014 0.012

All participants Subgroup 1 Subgroup 2 Subgroup 3

0.01 1999

2000

2001

2002

2003

2004

2005

2006

2003

2004

2005

2006

Year 0.045

HCB in mg/kg lipid

0.04

0.035

0.03 All participants Subgroup 1 Subgroup 2 Subgroup 3

0.025

0.02 1999

2000

2001

2002 Year

Fig. 3. (Continued)

of the measured values were below the LOQ. There was no great change in these basic trends if one includes all participants. The measured concentrations even at the 95th percentiles were close to the LOQ. Heptachlor epoxide is a persistent metabolite of the insecticide heptachlor (Kielhorn et al., 2006).

Determining factors of POP concentration levels Median concentrations of total PCB, DDT, b-HCH and HCB showed a clear rise with increasing age of mothers. This was especially true for total PCB (Fig. 5). Comparing values of mothers in the age class up to 28 years with mothers in the age class of 34 years and older by means of the U-test (Mann–Whitney) resulted in a

highly significant difference (po0.001) for all four parameters (subgroup 2, participants of the years 2005 and 2006). Concerning the influence of the number of breastfed children on median POP concentrations in human milk samples, a clear decrease of median concentrations with increasing number can be observed (Fig. 6). Especially, in total PCB this effect was less intensive including only participants of the years 2005 and 2006. It should be noted that the median age of the mothers also increased with the number of breastfed children, which may counter this downward trend in parts. There was a remarkable influence of the country of birth and concentrations of total PCB, DDT and b-HCH values. The effect can be seen in Fig. 7. It can be seen that total PCB concentrations in mothers born in

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

0.3 1 child (n=2119) 2 children (n=1360) 3 children (n=375) 4 and more children (n=93)

0.25 mg/kg lipid

Germany were higher than that of mothers born abroad. These foreign countries were especially successor states of the Soviet Union (e.g. Kazakhstan, Russia) and Eastern European countries. Contrary to the PCB values, the median DDT concentrations were much higher in mothers born abroad than in mothers born in Germany. Comparing the group of mothers born in Germany with mothers born in foreign countries (participants of the years 2005 and 2006 breastfeeding their first child) using the U-test (Mann– Whitney), a highly significant difference (po0.001) for

631

0.2 0.15 0.1 0.05 0 PCB

DDT

β-HCH

HCB

Fig. 6. Influence of number of breastfed children on median POP values, only mothers born in Germany (subgroup 1).

0.9 0.8 0.7 0.6 0.5

0.3

0.4

0.25

Germany (n=512) Foreign states (n=83)

mg/kg lipid

0.3 0.2 0.1 0 DDT

PCB

β-HCH

HCB

Fig. 4. Box-and-Whisker plot of total PCB, DDT, b-HCH and HCB including participants from the years 2005 and 2006, only mothers born in Germany and only first breastfed children (n ¼ 512). Units are in mg/kg lipid. Box marks 25th and 75th percentile, the ends of the vertical bar mark the 5th and 95th percentile. Maximum and minimum values are depicted by crosses.

0.2 0.15 0.1 0.05 0 PCB

DDT

β-HCH

HCB

Fig. 7. Country of birth and median concentrations of POPs in participants of the years 2005 and 2006 breastfeeding their first child.

0.4 <24 years (n=27) 24-26 years (n=50)

0.35

27-29 years (n=110) 30-32 years (n=131)

mg/kg lipid

0.3

33-35 years (n=115) 36-38 years (n=62)

0.25

>38 years (n=16)

0.2 0.15 0.1 0.05 0 PCB

DDT

β-HCH

HCB

Fig. 5. Influence of the age of mothers on median POP concentrations in participants of the years 2005 and 2006, born in Germany and breastfeeding their first child (n ¼ 511).

ARTICLE IN PRESS 632

Table 2.

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

Summary of multiple regression analyses of different possible influence factors

Variable

Model all participants – b-coefficients Total PCB

Age of mother Body mass indexa Mother’s place of birth outside Germany Present smoking Former smoking Number of breastfed Two children Three children Four children or more Sampling year 2000 2001 2002 2003 2004 2005 2006 R2

0.017 0.005 0.049

DDT 0.006 0.001 0.263

b-HCH 0.000 0.001 0.095

Model subgroup 3 – b-coefficients HCB 0.002 0.001 0.002

Total PCB

DDT

0.017 0.003 –

–

b-HCH

0.005 0.000

HCB

0.001 0.001 –

0.002 0.001 –

0.015 0.004

0.013 0.005

0.000 0.002

0.002 0.000

0.006 0.006

0.010 0.009

0.002 0.000

children 0.055 0.098 0.162

0.045 0.069 0.096

0.010 0.006 0.016

0.011 0.015 0.023

– – –

– – –

– – –

– – –

0.050 0.011 0.060 0.053 0.087 0.135 0.146 0.388

0.022 0.022 0.019 0.029 0.048 0.063 0.073 0.158

0.003 0.000 0.001 0.007 0.011 0.014 0.013 0.191

0.009 0.010 0.010 0.013 0.022 0.022 0.026 0.146

0.042 0.002 0.042 0.053 0.081 0.135 0.123 0.232

0.020 0.013 0.029 0.029 0.034 0.063 0.055 0.038

0.001 0.000 0.003 0.004 0.007 0.008 0.008 0.180

0.007 0.007 0.010 0.012 0.020 0.022 0.020 0.184

0.000 0.003

Included are b-coefficients and corresponding p-values (two-tailed) using a model covering all 8 years of testing in all participants and also the more homogeneous subgroup 3 (only participants born in Germany with first breastfed children and age of mother only between 27 and 34 years). Values of given sampling years refer to effect versus year 1999 and effect of number of breastfed children versus one breastfed children. Significant results (o0.05) are printed in bold letters. a At the time of sampling.  po0.05.  po0.01. po0.001.

the parameters total PCB, DDT and b-HCH could be detected. No significant difference existed for HCB.

Multivariate analysis Possible associations of the organochlorine compounds with different factors were tested applying a multivariate regression model preceded by a variance analysis for all participants and different subgroups. Results of the model including all participants and also a model including only subgroup 3 can be found in Table 2. For better comparison of the effects, the model uses the original scale. By transforming analytical results of total PCB, DDT, b-HCH and HCB with the natural logarithm, the goodness-of-fit in term of the R2-values can be improved. Significant effects could be seen in many cases for the factors age of mother, body mass index at time of sampling, mother’s place of birth, number of breastfed children and sampling year. Present or former smoking had no influence on POP concentrations. Due to a substantial change of the questionnaire,

factors total length of breastfeeding and school-leaving qualification were only tested in a separate model including subgroup 3 of participants from the year 2006. In this group, total lactation period was an influencing factor only for total PCB (b-coefficient: 0.0002; p ¼ 0.005) and HCB (b-coefficient: 0.00002; p ¼ 0.033) concentrations. These effects were quite low. School-leaving qualification was no significant factor.

Discussion The collected ca. 4300 milk samples of women living in Lower Saxony showed a significant declining time trend of exposure to many contaminants. It has to be taken into account that the collected data of the investigation is not representative for all lactating mothers of that period because it is based on participants, who volunteered themselves. The age distribution for all program participants is shifted to some extent towards a higher age compared to all mothers that gave

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

Table 3.

633

Comparison of different measured concentration levels in human milk investigations of German mothers

Region

Lower Saxony

Source

This Voßmann Raum et Schade and investigation et al. (1995) al. (1998) Heinzow (1998)

Year of sampling

2006

Number of 523 samples Concentration Median values (mg/kg lipid) Total DDT 0.081 b-HCH 0.012 HCB 0.023 Total PCBa 0.183 PCB 138 0.038 PCB 153 0.046 PCB 180 0.024

Lower Saxony

1994

Lower Saxony

Schleswig–Holstein Northern Bavaria

96

July July 1995–June 1992–June 1997 1993 156b 246

Median (mg/kg lipid) 0.19 0.03 0.09 0.49 0.10 0.13 0.07

Median (mg/kg lipid) 0.31 0.04 0.17 0.51 n.a. n.a. n.a.

Median (mg/kg lipid) 0.202 0.036 0.065 0.502 n.a. n.a. n.a.

Former German North Demokratic RhineWestphalia Republic

SaxonyAnhalt

Langstadtler Fu¨rst et al. Horn et al. (1994) Benkwitz (1993) (1994) et al. (2002) 1991 1991 1989/1990 1999/2000

468

113

90

105

Mean (mg/kg lipid) 0.583 0.065 0.274 (0.877)d n.a. n.a. n.a.

Mean (?) (mg/kg lipid) (0.504)c 0.056 0.177 n. a. 0.202 0.254 0.125

Median (mg/kg lipid) (2.02)c 0.15 0.21 0.69 0.14 0.19 0.10

Median (mg/kg lipid) 0.603e 0.040 0.031 0.210 0.040 0.057 0.030

Studies used differing analytical methods and recruitment. Units are in mg/kg lipid; n.a., not available. a Total PCB values calculated by multiplying the sum of the congeners 138, 153 and 180 by the factor of 1.64. b Participants only primiparous, breastfeeding one child, mother’s age 25–35 years. c Value only of metabolite p,p0 -DDE. d Value given as PCB A60. e Sum of p,p0 -DDT, p,p0 -DDE, p,p0 -DDD.

birth to a child in Lower Saxony. It is known from many studies that concentrations of persisting organic pollutants tested in this investigation trend to increase with a higher age of the mothers (Wittsiepe et al., 2007; Schlaud et al., 1995; Dewailly et al., 1996; Tanabe and Kunisue, 2007; UBA, 1999). This can also be seen in our data. Therefore, median concentrations of all participating mothers could be expected to be (moderately) higher than median values of the entirety of breastfeeding mothers in Lower Saxony. There are other factors such as number of breastfed children, time of sampling in the lactation period, and a migration background of the mothers that may differ from the total population of lactating mothers in a particular year, possibly influencing median values. These effects could be controlled at least in parts by the selection of more homogenous subgroups. Each year samples of several hundred mothers have been analyzed. This is much more than in many other published investigations. A higher sample size may diminish the influence of a few outliers on the data. If one assumes that the uptake of PCB and organochlorine pesticides was mainly due to animal food products, which were produced and distributed supraregionally over a long time span in the mothers’ lives (UBA, 1999; La Rocca and Mantovani, 2006), only a limited contamination variability between study groups

from different Western German regions can be expected. This is confirmed by a comparison of available data (cf. Table 3). Because of these facts it can be expected that, although the participants in this investigation were not recruited as a representative group, the detected substance levels would only deviate moderately from a representative collective. Some 8.4% of the mothers stated that they were born outside of Germany. There was no completely matching data for comparison available to us, but at the end of the year 2006 in Lower Saxony 6.3% of all woman had a foreign nationality (NLS, 2007). Other factors influencing body burden with persistent organic pollutants found in this investigation were number of breastfed children (similar to parity), foreign country of origin and body mass index. These factors could be seen in many other studies, too (Wittsiepe et al., 2007; Schade and Heinzow, 1998; Dewailly et al., 1996; LaKind et al., 2001; Tanabe and Kunisue, 2007). The influence of different lifestyle factors especially like dietary habits could be detected in some other investigations (Schade and Heinzow, 1998; Tanabe and Kunisue, 2007). Total lactation period which had a minor influence on total PCB and HCB values in this investigation (multivariate analysis) was also seen as an influencing factor for POP concentrations in some other studies, although results were not always significant or

ARTICLE IN PRESS 634

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

even consistent (LaKind et al., 2001; Wittsiepe et al., 2007; Schade and Heinzow, 1998). In a review of published investigations, LaKind et al. (2001) concluded that the data do not support a specific rate of depuration. The decrease of median concentrations of POPs with increasing parity/number of breastfed children (and to a lesser extend also increasing length of total lactation period) may be interpreted as an elimination process of the mother. It was calculated that during a 3-month breastfeeding period, a mother depurates between 10% and 30% of her body burden of POPs (UBA, 1999). DDT concentrations in human milk have declined in most areas of the world, consistent with restrictions on its use. Nevertheless, levels can be high in areas which used DDT until recent times (Smith, 1999; Wong et al., 2005), such as Mexico (Torres-Sanchez et al., 2007). Population means have declined in much of the world, from 5 to 10 mg DDT/kg milk fat to around or below 1 mg/kg milk fat at the end of the 20th century in many areas (Smith, 1999). The higher median DDT concentrations in mothers born in foreign countries in our investigation probably reflect this different exposure situation of the mothers. Many mothers in this program that were born outside Germany migrated from Eastern European and central Asian states. It is known that pesticides were used in a huge extend in some (former) Eastern bloc countries. For example, Kazakhstan and its neighbor state Uzbekistan have great environmental problems due to agricultural pesticide use and irrigation to produce cotton and rice (Small et al., 2001). In a biomonitoring study of mothers from Kazakhstan sampled in the year 1994, the mean DDT concentration in the 74 samples was 0.3 mg/kg lipid with higher values in specific regions (Hooper et al., 1997). In this program, the median DDT concentration of the 19 mothers from Kazakhstan with their first breastfed child was 0.230 mg/kg lipid (years 2005/2006). Although DDT and other POPs have been banned in Europe for many years, these substances still find their way into the food chain for example due to soil persistence, inappropriate waste disposal or imported products. Therefore, residual sources could be expected to diminish the slope of the decline curve (Smith, 1999). Due to this possible deviation from the main trend, age was included in the multivariate regression model not as a constant variable but in single year steps. Determined concentrations in human milk below the LOQ were included in the statistics as zero, deviating from many other investigations. This was done because many values and in some cases nearly all values of substances and metabolites were below the LOQ. Setting these readings as half of the LOQ would have drawn a wrong characterization. Additionally, the LOQ varies with milk fat content of the samples. Because most of the data described in this evaluation including all time

trends are based on medians and percentiles there is no alteration due to this measure. Of the given data only means and standard deviations were influenced. In the year 1999, reference values for some measured substances were published using data of ca. 1750 samples collected in the year 1994. The reference values were deduced by estimation of the 95th percentiles of a dataset merged from several sources using different analytical methods (UBA, 1999). Personal characteristics and lifestyle details were not considered. The reference values were stated as follows: total PCB 1.2 (sum PCB 138, 153, 180 multiplied by 1.64), total DDT 0.9, b-HCH 0.1, and HCB 0.3, each in mg/kg milk fat. Referring to participants of the year 2006, only a few mothers had measured concentrations exceeding these values, and if only for total PCB, DDT and b-HCH. It should be noted that with the known decreasing time trend of substance concentrations these reference values cannot be expected to fit well anymore.

Results in relation to other human milk studies and monitoring programs in Germany Generally, considering the sampling year there is a remarkable concordance of median or mean concentration levels of PCBs and important organochlorine pesticides in German studies and monitoring programs on breast milk. An overview of key values in different investigations can be found in Table 3. In the following, results of this monitoring program are discussed in relation to basic findings of other important German investigations. Lower Saxony A predecessor of this human milk monitoring program in Lower Saxony was offered by the government laboratory for food investigation (Staatliches Lebensmitteluntersuchungsamt Oldenburg) between the years 1986 and 1994 (Voßmann et al., 1995). They observed a clear downward trend for total PCB, DDT, b-HCH and HCB values. The median values of calculated total PCB including all participants fell from the range of 0.8 to 0.49 mg/kg lipid in the year 1994 (1994: n ¼ 96). That of DDT from the range of 0.5 mg/kg lipid to 0.19, b-HCH from the range of 0.07 to 0.03 mg/kg lipid and HCB from the range of 0.3 to 0.09 mg/kg lipid. Samples were measured by GC–ECD using an extraction method somewhat different to this monitoring program. These results show that the declining trend of the chemicals seen in this investigation dates back many years, with notably higher baselines in the past. A cross-sectional study including 156 primiparous, breastfeeding women from Lower Saxony and 113 from Saxony-Anhalt was conducted from July 1992 to June

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

1993 (Lower Saxony), and from January to June 1995 (Saxony-Anhalt) (Raum et al., 1998; Schlaud et al., 1995). The authors concluded that following group adjustment human milk contamination levels of mothers from former East German Saxony-Anhalt exceeded the contamination in Lower Saxony only for DDT and b-HCH. Heptachlor epoxide and dieldrin concentrations were significantly lower in Saxony-Anhalt (Raum et al., 1998). Schleswig–Holstein Mothers residing in the northern German state of Schleswig–Holstein was offered a human milk monitoring program between the years 1986 and 1997. The median concentration levels of organochlorine pesticides decreased by 80–90% during these 12 years, and the median PCB levels by 60% (Schade and Heinzow, 1998). The recorded values of a neighboring state of Lower Saxony are quite close to the median concentrations detected at the start of our investigation in 1999 (cf. Table 3). Bavaria In the German state of Bavaria, a human milk monitoring program analyzed persistent pesticides and selected halogenated organic compounds (PCBs, DDT, HCB) for more than 20 years (Raab et al., 2007). The levels of these contaminants in human breast milk showed a decreasing trend over this time. Unfortunately, in the last years the number of participants fell below 30 per year. Partial results, only of samples from Northern Bavaria for the years 1984–1991, were published by Langstadtler (1993). Recently, the focus of the monitoring program was switched to the detection of PCDD, PCDF, dioxin-like PCBs (Raab et al., 2007) and perfluorinated compounds (Vo¨lkel et al., 2007). North Rhine-Westphalia A monitoring program in the state of North RhineWestphalia, Germany, measured dioxins, polychlorinated biphenyls and other organohalogen compounds to a different extent in human milk samples. The authors report the results of PCB congeners 138, 153, 180 in 2032 individual human samples analyzed between 1984 and 2003. These congeners were also measured in this study. While the median levels in the 1980s had a quite similar level, the samples collected in the following years showed a clear downward trend. The data for DDT, b-HCH and HCB of more than 1900 individual human milk samples analyzed between 1984 and 2001 were also presented. They summarized that for most of the persistent organohalogen compounds, the levels in human milk at the beginning of this century were 70–95% lower compared to samples from the 1980s. As an exception, the PBDE concentrations in samples of

635

the year 2002 showed higher levels than 10 years ago (Fu¨rst, 2006; Fu¨rst et al., 1994; Wilhelm et al., 2007). Selected toxaphene congeners were also investigated in human milk (Skopp et al., 2002). Blood and associated milk samples of 169 women were compared in the Duisburg birth cohort study for their content of PCDD/F as well as dioxin-like and indicator polychlorinated biphenyls (PCB) in samples collected between September 2000 and January 2003. Contaminant levels in lipid base between both matrices showed good correlations. They observed a dependence of the substance distribution between blood and milk on the molecular weight of the substances and grade of chlorination (Wittsiepe et al., 2007). They suppose from their data that the previously observed continuous decrease of human PCB (and PCDD/F) levels might now have stopped. This is confirmed by our data on PCB levels at least for the years 2005/2006 including all participants. In different subgroups there was still a decline between these years. Hesse A study from Middle Hesse that compared POP concentrations in human milk samples collected in the years 1984/85, 1990/91 and 1995 showed highly significant (po0.001) reductions in levels of b- and g-HCH, HCB, p,p0 -DDE and p,p0 -DDT, between the eldest and the most recent testing group. A weakly significant reduction (po0.05) of a-HCH was also observed. Levels of high-chlorinated PCB congeners fell between the years whereas levels of low-chlorinated PCB congeners showed an increase (Ott et al., 1999). Baden-Wu¨rttemberg As part of a global WHO study of human milk contaminant levels, 39 samples were analyzed from four Baden-Wu¨rttemberg sample pools. Samples were selected according to the WHO criteria (e.g. inclusion only of first and fully breastfed children). Emphasis was placed on dioxin and dioxin-like PCBs. Mean DDT concentration was 0.152 mg/kg lipid, that of b-HCH 0.021 mg/kg lipid and that of HCB 0.042 mg/kg lipid, respectively (Piechotowski et al., 2004). The human milk monitoring program of the state that is in progress since 1980 showed falling mean values of DDT, b-HCH and HCB values until the last reported year of 2002. Total PCB values showed decreasing values in most years since 1984. The number of participants declined constantly in the last reported years (Ministerium fu¨r Umwelt und Verkehr Baden-Wu¨rttemberg, 2003; Sozialministerium Baden-Wu¨rttemberg, 2000). In addition to this monitoring program, POPs were quantified in blood of 10-year-old children between the years 1996 and 2003 (Link et al., 2005). Breastfeeding was associated with about 30% higher median concentrations of PCBs, DDE, and HCB in blood samples.

ARTICLE IN PRESS 636

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

German Demokratic Republic (GDR) and new federal states In the years 1989/1990, in total 90 human milk samples were collected in East Berlin, the district of Go¨rlitz and the district of Strausberg. Compared to studies that measured samples originating from the old territory of the Federal Republic of Germany (FRG), the median values of DDT metabolites were much higher for the corresponding time (Horn et al., 1994). A massive reduction of median pesticide and PCB concentrations was observed for human milk samples collected in the district of Bitterfeld (Saxony-Anhalt) in the years 1990, 1993, 1996 and 1999/2000 (Doering et al., 1999; Benkwitz et al., 2002). Median DDT concentrations were also higher compared to the old states of Germany. In our investigation, the GDR as a place of birth of a mother cannot be assigned completely for the last monitored years due to a substantial change of the questionnaire. But in the years before the change median DDT concentrations of mothers born in the GDR compared to mothers born in Lower Saxony were clearly higher (data not shown). A corresponding long-term decreasing trend over decades could also be monitored in other industrialized countries (Solomon and Weiss, 2002). For example during the course of 20–30 years in Sweden the levels of organochlorine compounds (PCBs, polychlorinated naphthalenes (PCNs), dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), DDT, DDE, HCB, dieldrin) in human milk have all decreased notably. A decrease to half of the original concentration could be seen in the range of 4–17 years periods. In contrast to these substances, the concentrations of polybrominated diphenyl ethers (PBDEs) have increased during the period between 1972 and 1997, indicating a doubling of the levels by 5 years (Nore´n and Meironyte, 2000).

Conclusions A clear downward trend of median total PCB, DDT, b-HCH and HCB values in all participants and also in different selected subgroups could be observed over 8 years. The data indicates a continuation of downward trends in human milk contamination with persistent organic pollutants found in earlier investigations of German nursing mothers. The determined substance concentrations in measured human milk samples reflect the body burden of the mothers, acquired during their lifetime. The ongoing decrease of the PCB, DDT, b-HCH and HCB values confirms a continuing beneficial effect of regulations and other measures taken to minimize the pollution with these POPs, years ago. The associations of personal characteristics and lifestyle factors with contamination levels found in this monitor-

ing program are comparable with results from other studies and monitoring programs. Unlike many other human milk monitoring programs covering a multi-year period, the number of participants remained at a constantly high level (several hundred per year) in this investigation. This was due to regular information and reminders send out to pediatricians, gynecologists, and midwives.

Acknowledgments The authors want to thank Dr. Oliver Hehl, Dr. Ute Jorritsma, Dr. Matthias Pulz (president Landesgesundheitsamt), Prof. Dr. Adolf Windorfer (former president Landesgesundheitsamt), Dr. Klaus-Michael Wollin for their scientific advise and Sedi Asghary, Gabriele Bru¨nner-Ku¨hn, Sandra Heidrich, Jo¨rg Homann, Astrid Nell, Angela Prien, Christa Steffens, Bianca Sterenberg, Birgit Stoeteknuel, Eveline Wieczorek, for their technical support. In the cooperating laboratory Lebensmittelinstitut Oldenburg (Niedersa¨chsisches Landesamt fu¨r Verbraucherschutz und Lebensmittelsicherheit/LAVES), they thank Dr. Ingeborg Block, Dr. Elke Bruns-Weller, Helga Groenhagen, Dr. Annette Knoll and Dr. Ju¨rgen Pfordt.

References AMAP (Arctic Monitoring and Assessment Programme), 2004. AMAP Assessment 2002: Persistent Organic Pollutants in the Arctic. AMAP, Oslo. Angerer, J., Ewers, U., Wilhelm, M., 2007. Human biomonitoring: state of the art. Int. J. Hyg. Environ. Health 210, 201–228. ATSDR (Agency for Toxic Substances and Disease Registry), 2000. Toxicological Profile for Polychlorinated Biphenyls (PCBs). ATSDR, Atlanta. ATSDR (Agency for Toxic Substances and Disease Registry), 2002a. Toxicological Profile for DDT, DDE, and DDD. ATSDR, Atlanta. ATSDR (Agency for Toxic Substances and Disease Registry), 2002b. Toxicological Profile for Hexachlorobenzene. ATSDR, Atlanta. ATSDR (Agency for Toxic Substances and Disease Registry), 2005. Toxicological Profile for Hexachlorocyclohexane. ATSDR, Atlanta. Benkwitz, F., Grossert, H., Neske, P., Seeber, B., 2002. Persistente chlororganische Schadstoffe und MoschusVerbindungen in Muttermilch gehen zuru¨ck. Umweltmed. Informationsdienst 4, 18–20. Dewailly, E., Ayotte, P., Laliberte, C., Weber, J.P., Gingras, S., Nantel, A.J., 1996. Polychlorinated biphenyl (PCB) and dichlorodiphenyl dichloroethylene (DDE) concentrations in the breast milk of women in Quebec. Am. J. Public Health 86, 1241–1246.

ARTICLE IN PRESS B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

Doering, C., Thriene, B., Oppermann, H., Seeber, B., Pfeifer, I., Grossert, H., Benkwitz, F., 1999. Muttermilchuntersuchungen im Landkreis Bitterfeld. Gesundheitswesen 61, 560–566. Fu¨rst, P., 2006. Dioxins, polychlorinated biphenyls and other organohalogen compounds in human milk. Levels, correlations, trends and exposure through breastfeeding. Mol. Nutr. Food Res. 50, 922–933. Fu¨rst, P., Fu¨rst, C., Wilmers, K., 1994. Human milk as a bioindicator for body burden of PCDDs, PCDFs, organochlorine pesticides, and PCBs. Environ. Health Perspect. 102 (Suppl. 1), 187–193. Hooper, K., Hopper, K., Petreas, M.X., She, J., Visita, P., Winkler, J., McKinney, M., Mok, M., Sy, F., Garcha, J., Gill, M., Stephens, R.D., Semenova, G., Sharmanov, T., Chuvakova, T., Hopper, K., 1997. Analysis of breast milk to assess exposure to chlorinated contaminants in Kazakstan: PCBs and organochlorine pesticides in southern Kazakstan. Environ. Health Perspect. 105, 1250–1254. Horn, M., Heinzow, B., Dolk, G., 1994. Belastung der Humanmilch in der ehemaligen DDR mit DDT, HCH, HCB, PCB. Untersuchung und toxikologische Bewertung. Zentralbl. Hyg. Umweltmed. 196, 95–103. Jakubowski, M., Trzcinka-Ochocka, M., 2005. Biological monitoring of exposure: trends and key developments. J. Occup. Health. 47, 22–48. Kielhorn, J., Schmidt, S., Mangelsdorf, I., 2006. Heptachlor. First draft. Concise International Chemical Assessment Documents (CICADS) no. 70. WHO Press, Geneva. LaKind, J.S., Berlin, C.M., Naiman, D.Q., 2001. Infant exposure to chemicals in breast milk in the United States: what we need to learn from a breast milk monitoring program. Environ. Health Perspect. 109, 75–88. Landrigan, P.J., Sonawane, B., Mattison, D., McCally, M., Garg, A., 2002. Chemical contaminants in breast milk and their impacts on children’s health: an overview. Environ. Health Perspect. 110, A313–A315. Langstadtler, M., 1993. Ergebnisse von Frauenmilchuntersuchungen in Nordbayern. Gesundheitswesen 55, 301–307. La Rocca, C., Mantovani, A., 2006. From environment to food: the case of PCB. Ann. Ist. Super. Sanita. 42, 410–416. Link, B., Gabrio, T., Zoellner, I., Piechotowski, I., Paepke, O., Herrmann, T., Felder-Kennel, A., Maisner, V., Schick, K.H., Schrimpf, M., Schwenk, M., Wuthe, J., 2005. Biomonitoring of persistent organochlorine pesticides, PCDD/PCDFs and dioxin-like PCBs in blood of children from South West Germany (Baden-Wuerttemberg) from 1993 to 2003. Chemosphere 58, 1185–1201. Ministerium fu¨r Umwelt und Verkehr Baden-Wu¨rttemberg (ed.), 2003. Umweltdaten 2003. Baden-Wu¨rttemberg. LfU, Mannheim. Available from: /http://www2.lubw.badenwuerttemberg.de/public/abt2/umweltdaten2003_pdf/pdf/ LFU_189_206.pdfS via the INTERNET (accessed 19.03.08). Needham, L.L., Wang, R.Y., 2002. Analytic considerations for measuring environmental chemicals in breast milk. Environ. Health Perspect. 110, A317–A324. NLS (Niedersa¨chsisches Landesamt fu¨r Statistik). Statistische Berichte Niedersachsen, 2006. A II 1-j/05-Eheschließungen,

637

Geborene und Gestorbene 2005. Hannover, NLS–Schriftenvertrieb. Available from: /http://www.nls.niedersachsen. de/Download/StatistischeBerichte/AII1_2005.pdfS via the INTERNET (accessed 19.03.08). NLS (Niedersa¨chsisches Landesamt fu¨r Statistik), 2007. NLS-Online: Tabelle K1000124. Bevo¨lkerung am 31.12.2006 in Niedersachsen. Available from: /http:// www1.nls.niedersachsen.de/statistik/html/mustertabelle.asp? DT=K1000124&LN=DBP&DA=9S via the INTERNET (accessed 19.03.08). Nore´n, K., Meironyte, D., 2000. Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere 40, 1111–1123. Ott, M., Failing, K., Lang, U., Schubring, C., Gent, H.J., Georgii, S., Brunn, H., 1999. Contamination of human milk in Middle Hesse, Germany – a cross-sectional study on the changing levels of chlorinated pesticides, PCB congeners and recent levels of nitro musks. Chemosphere 38, 13–32. Piechotowski, I., Link, B., Malisch, R., Kypke, K., 2004. Muttermilchuntersuchung Baden-Wu¨rttemberg 2001/02 im Rahmen der WHO-Studie zum Gehalt der Muttermilch mit PCBs, PCDDs und PCDFs. Landesgesundheitsamt Baden-Wu¨rttemberg, Stuttgart. Available from: /http:// www.landesgesundheitsamt.de/servlet/PB/show/1194729/ humanmilch%2001-02.pdfS via the INTERNET (accessed 19.03.08). Pronczuk, J., Akre, J., Moy, G., Vallenas, C., 2002. Global perspectives in breast milk contamination: infectious and toxic hazards. Environ. Health Perspect. 110, A349–A351. Pronczuk, J., Moy, G., Vallenas, C., 2004. Breast milk: an optimal food. Environ. Health Perspect. 112, A722–A723. Raab, U., Schwegler, U., Preiss, U., Albrecht, M., Fromme, H., 2007. Bavarian breast milk survey – pilot study and future developments. Int. J. Hyg. Environ. Health 210, 341–344. Raum, E., Seidler, A., Schlaud, M., Knoll, A., Wessling, H., Kurtz, K., Schwartz, F.W., Robra, B.P., 1998. Contamination of human breast milk with organochlorine residues: a comparison between East and West Germany through sentinel practice networks. J. Epidemiol. Community Health. 52 (Suppl. 1), 50S–55S. Schade, G., Heinzow, B., 1998. Organochlorine pesticides and polychlorinated biphenyls in human milk of mothers living in northern Germany: current extent of contamination, time trend from 1986 to 1997 and factors that influence the levels of contamination. Sci. Total Environ. 215, 31–39. Schlaud, M., Seidler, A., Salje, A., Behrendt, W., Schwartz, F.W., Ende, M., Knoll, A., Grugel, C., 1995. Organochlorine residues in human breast milk: analysis through a sentinel practice network. J. Epidemiol. Community Health 49 (Suppl. 1), 17–21. Schulte, E., Malisch, R., 1984. Calculation of the real PCB content in environmental samples. II: Gas chromatographic determination of the PCB concentration in human milk and butter. Fresenius Z. Anal. Chem. 319, 54–59. Skopp, S., Oehme, M., Fu¨rst, P., 2002. Enantiomer ratios, patterns and levels of toxaphene congeners in human milk from Germany. J. Environ. Monit. 4, 389–394.

ARTICLE IN PRESS 638

B.P. Zietz et al. / Int. J. Hyg. Environ. Health 211 (2008) 624–638

Small, I., van der Meer, J., Upshur, R.E., 2001. Acting on an environmental health disaster: the case of the Aral Sea. Environ. Health Perspect. 109, 547–549. Smith, D., 1999. Worldwide trends in DDT levels in human breast milk. Int. J. Epidemiol. 28, 179–188. Solomon, G.M., Weiss, P.M., 2002. Chemical contaminants in breast milk: time trends and regional variability. Environ. Health Perspect. 110, A339–A347. Sozialministerium Baden-Wu¨rttemberg (ed.), 2000. Kindergesundheit in Baden-Wu¨rttemberg. Stuttgart. Available from: /http://www.landesgesundheitsamt.de/servlet/PB/show/ 1142113/kindergesundheit2000.pdfS via the INTERNET (accessed 19.03.08). Steinwandter, H., 1982. Contributions to silica gel application in pesticide residue analysis. III. An On-Line method for extracting and isolating chlorinated hydrocarbon pesticides and polychlorinated biphenyls (PCB’s) from milk and dairy products. Fresenius Z. Anal. Chem. 312, 342–345. Tanabe, S., Kunisue, T., 2007. Persistent organic pollutants in human breast milk from Asian countries. Environ. Pollut. 146, 400–413. Torres-Sanchez, L., Rothenberg, S.J., Schnaas, L., Cebrian, M.E., Osorio, E., Del Carmen Hernandez, M., GarciaHernandez, R.M., Del Rio-Garcia, C., Wolff, M.S., LopezCarrillo, L., 2007. In utero p,p0 -DDE exposure and infant neurodevelopment: a perinatal cohort in Mexico. Environ. Health Perspect. 115, 435–439. UBA (Umweltbundesamt), 1999. Kommission ‘‘Human-Biomonitoring’’ des Umweltbundesamtes. Referenzwerte fu¨r HCB, b-HCH, DDT und PCB in Frauenmilch. Bundesgesundheitsbl 42, 533–539. UNEP Chemicals (United Nations Environment Programme), 2004. Guidance for a Global Monitoring Programme for Persistent Organic Pollutants. UNEP Chemicals, Geneva. Available from: /http://www.chem.unep.ch/gmn/ GuidanceGPM.pdfS via the INTERNET (accessed 19.03.08). Voßmann, U., Bruns-Weller, E., Ende, M., 1995. Bericht u¨ber die in Niedersachsen von 1991-1994 durchgefu¨hrten Mut-

termilchuntersuchungen. Staatliches Lebensmitteluntersuchungsamt Oldenburg. Niedersa¨chsisches Ministerium fu¨r Erna¨hrung, Landwirtschaft und Forsten, Hannover. Vo¨lkel, W., Genzel-Boroviczeny, O., Demmelmair, H., Gebauer, C., Koletzko, B., Twardella, D., Raab, U., Fromme, H., 2007. Perfluorooctane sulphonate (PFOS) and perfluorooctanoic acid (PFOA) in human breast milk: results of a pilot study. Int. J. Hyg Environ. Health (in press), doi:10.1016/j.ijheh.2007.07.024. WHO (World Health Organization), 2007. Fourth WHOCoordinated Survey of Human Milk for Persistent Organic Pollutants in Cooperation with UNEP—Guidelines for Developing a National Protocol. WHO Press, Geneva. Available from: /http://www.who.int/foodsafety/ chem/POPprotocol.pdfS via the INTERNET (accessed 19.03.08). WHO (World Health Organization), 2003a. Polychlorinated Biphenyls: Human Health Aspects. Concise International Chemical Assessment Document No. 55. WHO Press, Geneva. WHO Europe (World Health Organization Europe), 2003b. Health Risks of Persistent Organic Pollutants from LongRange Transboundary Air Pollution. WHO, Geneva. Available from: /http://www.euro.who.int/document/ e78963.pdfS via the INTERNET (accessed 19.03.08). Wilhelm, M., Ewers, U., Wittsiepe, J., Fu¨rst, P., Holzer, J., Eberwein, G., Angerer, J., Marczynski, B., Ranft, U., 2007. Human biomonitoring studies in North RhineWestphalia, Germany. Int. J. Hyg. Environ. Health 210, 307–318. Wittsiepe, J., Fu¨rst, P., Schrey, P., Lemm, F., Kraft, M., Eberwein, G., Winneke, G., Wilhelm, M., 2007. PCDD/F and dioxin-like PCB in human blood and milk from German mothers. Chemosphere 67, S286–S294. Wong, M.H., Leung, A.O., Chan, J.K., Choi, M.P., 2005. A review on the usage of POP pesticides in China, with emphasis on DDT loadings in human milk. Chemosphere 60, 740–752 (Erratum in: Chemosphere 64 (2006) 696).