Prenatal exposure to environmental contaminants and body composition at age 7–9 years

Environmental Research 132 (2014) 24–32

Contents lists available at ScienceDirect

Environmental Research journal homepage: www.elsevier.com/locate/envres

Prenatal exposure to environmental contaminants and body composition at age 7–9 years Immle Delvaux a, Jolijn Van Cauwenberghe a, Elly Den Hond b, Greet Schoeters b, Eva Govarts b, Vera Nelen c, Willy Baeyens d, Nicolas Van Larebeke e, Isabelle Sioen a,f,n a

Department of Public Health, Ghent University, UZ 2 Blok A, De Pintelaan 185, 9000 Ghent, Belgium Flemish Institute for Technological Research (VITO), Environmental Risk and Health, Boeretang 200, 2400 Mol, Belgium c Department of Health, Provincial Institute for Hygiene, Kronenburgstraat 45, 2000 Antwerp, Belgium d Department of Analytical and Environmental Chemistry, Free University of Brussels, Pleinlaan 2, 1050 Elsene, Belgium e Department of Radiotherapy and Nuclear Medicine, Ghent University, De Pintelaan 185, 9000 Ghent, Belgium f FWO Research Foundation, Egmontstraat 5, 1000 Brussels, Belgium b

art ic l e i nf o

a b s t r a c t

Article history: Received 17 September 2013 Received in revised form 13 March 2014 Accepted 20 March 2014

The study aim was to investigate the association between prenatal exposure to endocrine disrupting chemicals (EDCs) and the body composition of 7 to 9 year old Flemish children. The subjects were 114 Flemish children (50% boys) that took part in the first Flemish Environment and Health Study (2002– 2006). Cadmium, PCBs, dioxins, p,p0 -DDE and HCB were analysed in cord blood/plasma. When the child reached 7–9 years, height, weight, waist circumference and skinfolds were measured. Significant associations between prenatal exposure to EDCs and indicators of body composition were only found in girls. After adjustment for confounders and covariates, a significant negative association was found in girls between prenatal cadmium exposure and weight, BMI and waist circumference (indicator of abdominal fat) and the sum of four skinfolds (indicator of subcutaneous fat). In contrast, a significant positive association (after adjustment for confounders/covariates) was found between prenatal p,p0 -DDE exposure and waist circumference as well as waist/height ratio in girls (indicators of abdominal fat). No significant associations were found for prenatal PCBs, dioxins and HCB exposure after adjustment for confounders/covariates. This study suggests a positive association between prenatal p,p0 -DDE exposure and indicators of abdominal fat and a negative association between prenatal cadmium exposure and indicators of both abdominal as well as subcutaneous fat in girls between 7 and 9 years old. & 2014 Elsevier Inc. All rights reserved.

Keywords: Children Body composition Endocrine disruptors Growth Prenatal exposure

1. Introduction Obesity in children is an important health problem since most obese and overweight children grow up to be obese adults (Guo et al., 2002). Prevention of overweight and obesity in children is a priority because these diseases are linked to a number of severe health problems such as diabetes mellitus type 2, cardiovascular diseases and certain cancers (Collins, 2005). Besides genetic, behavioural and dietary factors, also environmental factors, e.g. exposure to endocrine disrupting chemicals (EDCs) may be risk factors for developing obesity. EDCs can interfere with the human

Abbreviations: EDCs, endocrine disrupting chemicals; PCBs, polychlorinated biphenyls; p,p0 -DDE, para,para-dichlorodiphenyldichloroethylene; HCB, hexachlorobenzene; BMI, body mass index; FLEHS, Flemish Environment and Health Study; LOD, limit of detection n Corresponding author at: Department of Public Health, Faculty of Medicine and Health Sciences, Ghent University, UZ 2 Blok A, De Pintelaan 185, 9000 Ghent, Belgium. Fax: þ 32 9 332 49 94. E-mail address: [email protected] (I. Sioen). http://dx.doi.org/10.1016/j.envres.2014.03.019 0013-9351/& 2014 Elsevier Inc. All rights reserved.

endocrine system, potentially playing a role in the development of obesity (Newbold, 2010; Tang-Peronard et al., 2011). Polychlorinated biphenyls (PCBs), dioxins (polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans), para,para-dichlorodiphenyldichloroethylene (p,p0 -DDE), hexachlorobenzene (HCB) and heavy metals like cadmium are all considered to act as EDCs. Studies have indicated that dioxins can bind the aryl hydrocarbon receptor, induce the cytochrome P450 1A enzyme and have an antiestrogenic action, while PCBs induce the pregnane X receptor and the constitutive androstane receptor and induce thyroid hormone disruption. HCB and p,p0 -DDE are thought to have an antiandrogenic function (Legler et al., 2011). Cadmium can affect the secretory patterns of pituitary hormones and the synthesis of progesterone (Iavicoli et al., 2009). Recent epidemiological studies suggest that EDCs exposure during the critical period of foetal development is associated with overweight and obesity later in life. Results of these studies are summarized in Table 1. This table is limited to studies with children in the age range between 4.5 and 16 years old.

Table 1 Available data in literature on the effect of prenatal exposure to EDCs and anthropometric parameters in later life. EDC

Reference

Age (years) Number of study Level of exposure participants In cord blood(B)/serum(S)

Cd

Tian et al. (2009)

4.5 4.5

106 106

Dioxins

Su et al. (2010)

5

149

5

149

5

149

6.5

344

PCB

HCB

Gladen et al. (2000) 10–15

594

10–15

594

6.5

343

Valvi et al (2012)

Gladen et al. (2000) 10–16

594

10–16

594

Significant

In maternal blood(B)/serum(S)

In placenta

Median 0.60 mg/L Median 0.60 mg/L

0.15 g/g dry weight 0.15 g/g dry weight

Height Weight

Decrease Decrease

S NS

Median dioxins/PCB Teq 15.15 pg/g lipid Median dioxins/PCB Teq 15.15 pg/g lipid Median dioxins/PCB Teq 15.15 pg/g lipid

Height

Increase

S

Weight

Increase

NS

BMI

Increase

NS

BMI

Increase

Sa

Increase

Sb

Increase

NS

BMI

Increase

Sc

Mean ( 7SD) 0.75 ( 7 1.70) mg/L (S) Median 9.06 ppb ¼ 9.06 mg/L (S) Median o4.27 ppb ¼ o 4.27 mg/L (S) Median o4.27 ppb ¼ o 4.27 Median 9.06 ppb ¼ 9.06 mg/L (S) mg/L (S) Mean ( 7SD) 1.06 ( 7 2.45) ng/mL (S) Median 3.95 ppb ¼ 3.95 mg/L (S) Median 3.95 ppb ¼ 3.95 mg/L (S)

Result

Weight adjusted Median for height o 12 ppb ¼ o 12 mg/L (S) Median o 12 ppb ¼ o 12 Height mg/L (S)

Median 12.60 ppb (S)

Median 6.77 ppb (S)

Weight adjusted for height

Increase

Sd

Median 12.60 (S)

Median 6.77 (S)

Height

Increase

NS

Decrease

S

Ribas-Fito 7 et al. (2002) Gladen et al. (2004) 10–20

1371

Median 24.4 μg/L (S)

Height

304 (only males)

Median 5.7 μg/g lipid (S)

Warner et al. (2013) 7

270

7

270

Mean ( 7 SD) 1.42 ( 7 0.003) mg/g lipid (S) Mean ( 7 SD) 1.42 ( 70.003) mg/g lipid (S)

Height, weight, BMI, skinfolds No association BMI Increase

6.5

405

6.5 6.5

405 405

Smink et al. (2008)

Waist circumference

No association

Median 0.68 mg/L (S)

Height

Median 0.68 mg/L (S) Median 0.68 mg/L (S)

Weight BMI

No association Increase Increase

NS

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

p,p0 -DDE

Valvi et al. (2012)

Median 1.8 mg/L (B) Median 1.8 mg/L (B)

Anthropometric parameter

NS S

Significant for girls only. EDC¼ endocrine disrupting chemical. a

Significant for third tertile, but not for second tertile. Significant for white girls only. c Significant for second tertile, but not third tertile. d Significant for boys only; S¼significant; NS¼ non-significant. b

25

26

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

Several studies investigated the effect of cadmium on anthropometrics at birth (Lin et al., 2011; Loiacono et al., 1992; Nishijo et al., 2002; Odland et al., 1999; Salpietro et al., 2002). Whether cadmium exposure during pregnancy has an effect on the growth of the child later in life has hardly been investigated (Lin et al., 2011; Tian et al., 2009) (Table 1). Concerning prenatal dioxin exposure, a number of studies reported about the relation between prenatal dioxin exposure and anthropometrics at birth with inconsistent results (Eskenazi et al., 2003; Konishi et al., 2009; Nishijo et al., 2002; Pluim et al., 1996; Tawara et al., 2009; Vartiainen et al., 1998). Two studied examined the effects of prenatal dioxin exposure on anthropometrics in children up to 5 years of age (Su et al., 2010; Verhulst et al., 2009) (Table 1). Like for dioxins, the data on the effects of prenatal PCBs exposure on anthropometrics at birth are inconsistent (Govarts et al., 2012; Hertz-Picciotto et al., 2005; Longnecker et al., 2005; Patandin et al., 1998; Ribas-Fito et al., 2002; Rogan et al., 1986; Vartiainen et al., 1998; Verhulst et al., 2009). A number of studies have reported the effect of prenatal PCB exposure on weight or body mass index (BMI) in children (Jacobson et al., 1990; Valvi et al., 2012; Verhulst et al., 2009), on weight and height in adolescents (Gladen et al., 2000; Valvi et al., 2012; Verhulst et al., 2009) and on body composition parameters in adults (Karmaus et al., 2009) (Table 1). Next, different authors reported on the negative association between prenatal exposure to p,p0 -DDE and decreased birth weight (Govarts et al., 2012; Jusko et al., 2006; Longnecker et al., 2001; Rogan et al., 1986; Verhulst et al., 2009; Wojtyniak et al., 2010). There have also been a number of studies that investigated the effect of prenatal p,p0 -DDE exposure on anthropometrics later in life, but the results of these studies did not lead to unambiguous findings (Cupul-Uicab et al., 2010; Garced et al., 2012; Gladen et al., 2004, 2000; Jusko et al., 2006; Karmaus et al., 2009; Ribas-Fito et al., 2006; Valvi et al., 2012; Verhulst et al., 2009; Warner et al., 2013) (Table 1). In contrast to p,p0 -DDE, the effects of prenatal HCB exposure on the developing child have not been extensively investigated. Eggesbo et al. (2009), Schade and Heinzow (1998) and Ribas-Fito et al. (2002) reported no associations between prenatal HCB exposure and anthropometrics at birth (Eggesbo et al., 2009; RibasFito et al., 2002; Schade and Heinzow, 1998). Smink et al. (2008) reported that prenatal exposure to HCB is associated with a significant increase in BMI and a non-significant increase in weight at age 6.5 and this effect is stronger in mothers who smoked during the pregnancy. There was no association found between prenatal exposure to HCB and height at age 6.5 (Smink et al., 2008) or BMI from birth to 3 years of age (Verhulst et al., 2009) (Table 1). Here we test the hypothesis that prenatal exposure to low levels of five EDCs influences body composition of Flemish children between 7 and 9 years old. Earlier results based on analysis of data collected in this Flemish cohort indicated that prenatal exposure to EDCs influences children's birth weight (Govarts et al., 2012) and their body composition on the age of 3 (n¼138 mother-infant pairs) (Verhulst et al., 2009). In this paper we describe the association between prenatal exposure to cadmium, PCBs, dioxins, p,p0 -DDE and HCB on a set of different anthropometric parameters: height, weight, BMI, waist circumference, waist/height ratio and the sum of four skinfolds (triceps, biceps, suprailiac and subscapular). To the authors’ knowledge, this is the first study using both waist circumference, waist/height ratio and skinfolds, besides more common anthropometric measures e.g. weight, height and BMI, in this kind of studies. Moreover, the anthropometric measurements were performed by two trained study nurses. 2. Methods

Within FLEHS I, 1196 mothers and their new-borns were recruited through 25 maternity hospitals between October 2002 and December 2003. These hospitals were spread over eight different areas with characteristic environmental exposures. Inclusion criteria were: living for at least five years in the area of interest, being able to fill out Dutch questionnaires and giving informed consent. About 44% of the mother who gave birth in one of the 25 maternity hospitals participating in the study did not fulfill the inclusion criteria (24% did not live for at least five years in the area of interest and 20% was not able to fill out Dutch questionnaires). Only 3% of the mothers that did fulfill the inclusion criteria, did not give informed consent. Cord blood was sampled at delivery and the exposure to five EDCs i.e. cadmium, PCBs, dioxins, p,p0 -DDE and HCB was measured. More information about the study design of FLEHS I is described by Koppen et al. (2009). In the summer of 2011, the parents of the children that participated as new-borns in 2002–2003 were re-contacted to participate in a new study in order to evaluate the impact of the prenatal exposure to the five EDCs – as measured in FLEHS I – on the body composition of the 7 to 9-year old children. The parents and children received an invitation letter explaining the aims of this study, as well as an informed consent and a postal questionnaire. Additionally, they were asked whether they were willing to be visited by a study nurse to measure length, weight, waist circumference and four skinfolds of the children. In total, 1173 invitation letters for the follow-up study were sent in June 2011, since 23 of the 1196 parents indicated at the baseline study in 2002–2003 that they were not willing to participate in follow-up studies. However, 109 closed envelops came back, because the participants had moved between 2002 and 2011. In total, 281 completed questionnaires came back. Not all families gave their permission for a home visit. Therefore and due to practical reasons, we reached 114 families for a home visit. In this paper, we only used the data from these 114 children. All participating parents provided informed consent for participation. The study was conducted according to the guidelines laid down in the Declaration of Helsinki and the study protocol was approved by the Ethical Committee of the University of Antwerp (Belgium) and the Ghent University (Belgium). 2.2. Prenatal exposure Cord blood was aliquoted and plasma was separated by centrifugation within one day in either the maternity or blood bank laboratories. The aliquoted samples were kept in the refrigerator for maximal one week. Since we only measured persistent chemicals (cadmium and persistent chlorinated compounds) that do not degrade, this method is in line with quality standards. Afterwards they were put at  20 1C until analysis. Due to low volume of the blood samples ( o5 mL blood) and technical reasons in the laboratory not all contaminants could be tested on each sample. In cord blood, cadmium was analysed using High Resolution Inductively Coupled Mass Spectrometry. For more information see Schroijen et al. (2008). The limit of detection (LOD) for cadmium in whole blood samples was 0.09 mg/L. It was possible to measure cadmium concentrations in 106 of the 114 cord blood samples. In 40 of the 106 cases, the concentration was below the LOD and replaced by 0.05 mg/L (LOD/2). Three non-dioxin-like PCB congeners were measured, being PCB 138, 153 and 180, since studies showed that these congeners account for 65 to 80% of the total sum of non-dioxin-like PCBs present in human blood (Apostoli et al., 2005; Needham et al., 2005). It was assumed that the sum of the concentrations of these three congeners give a good reflection of the PCB exposure in general. The PCBs (PCB 138, 153 and 180) and chlorinated pesticides were analysed by gas chromatography - electron capture detection using the method of Gomara et al. (2002). The analyses were performed by two labs. Both laboratories participated in the AMAP proficiency testing scheme (Institute National de Santé Publique, Quebec, Canada). The LOD for all chlorinated compounds was 0.02 μg/L. PCBs and p,p0 -DDE were measured in respectively 108 and 110 of the 114 samples, none of the concentrations were below the LOD. HCB concentration was measured in 108 of the 114 samples, with 32 of the concentrations (29.6%) below the LOD and replaced by 0.01 mg/L (LOD/2). Exposure to dioxin-like compounds was analysed via the CALUXs assay, based on in vitro activation of the aryl-hydrocarbonreceptor of cultured H4IIE rat hepatoma cells by the dioxin-like compounds present in 5 mL cord plasma (BioDetection Systems BV, Amsterdam, The Netherlands). The extraction and clean-up procedures were performed as described in Koppen et al. (2001). The limit of detection was 0.03 pg CALUX-TEQ/mL or 14 pg CALUX-TEQ/g lipids for 5 mL plasma, with a lipid content of 200 mg/dL (Koppen et al., 2009). Dioxins were analysed in 89 of the 114 samples, none of the concentrations were below the LOD. Routinely measured cholesterol and triglycerides were used to calculate total serum lipid content (Covaci et al., 2006), which was used as a confounder in the models studying the associations with prenatal exposure to PCBs, dioxins, p,p0 -DDE and HCB.

2.1. Study population and data collection 2.3. Follow-up questionnaire The study described in this paper was part of the human biomonitoring program in Flanders (Belgium). A cohort of study participants was recruited in the first Flemish Environment and Health Study (FLEHS I 2002–2006) (Schoeters et al., 2012).

In 2011, the parents completed a questionnaire to collect general information. Different aspects were included in the questionnaire: (1) child characteristics: birth

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

27

Table 2 Descriptive and anthropometric data of the study population. Total (n¼ 114)

Age child (y) Birthweight child (kg) Heigth (cm) Weight (kg) BMI child (kg/m²) z-score BMI child Waist circumference (cm) Waist/heigth Sum of 4 skinfoldsa (mm) BMI mother (kg/m²) BMI father (kg/m²) a

Boys (n¼ 57)

Girls (n¼ 57)

N

mean 7SD

Range

N

mean 7 SD

Range

N

mean 7SD

Range

114 114 114 114 114 113 112 112 112 114 114

8.43 7 0.34 3.42 7 0.51 133.07 6.7 29.9 7 6.6 16.8 7 2.5  0.137 1.18 58.79 7 6.96 0.447 0.04 34.137 18.85 24.017 3.99 25.86 7 2.91

(7.65–8.98) (2–5.58) (116.5–149) (19.5–59.5) (13.2–27.2) (  2.86–3.16) (47.00–87.40) (0.36–0.59) (9–120.1) (17.58–36.57) (20.15–34.50)

57 57 57 57 57 56 56 56 56 57 50

8.45 70.35 3.56 70.57 135.0 76.9 31.1 77.5 16.9 72.7 0.02 71.33 59.98 77.60 0.44 70.04 33.15 720.23 24.83 74.11 25.85 72.51

(7.65–8.96) (2–5.58) (121.5–149) (21.6–59.5) (13.2–27.2) (  2.86–3.16) (47–87.40) (0.36–0.59) (9–120.1) (17.58–36.05) (20.15–32.27)

57 57 57 57 57 57 56 56 56 56 52

8.417 033 3.29 7 0.40 131.0 7 5.9 28.6 7 5.3 16.6 7 2.3  0.277 1.00 57.59 7 6.09 0.447 0.04 35.117 17.49 23.197 3.72 25.86 7 3.28

(7.75–8.98) (2.5–4.25) (116.5–149) (19.5–45.1) (13.7–24.7) (  2.02–2.37) (49.40–78.30) (0.39–0.58) (16.4–98.73) (18.17–36.57) (20.20–34.50)

Sum of tricipital, biceps, subscapular and suprailiacal skinfolds.

day, gender, height, weight; (2) social status: number of children and adults in the household, highest educational level of both parents; (3) dietary habits: duration of breastfeeding, short food frequency questionnaire, consumption of locally grown vegetables, eggs and self-caught fish; (4) physical activity; (5) smoking in the neighbourhood of the child: in the first six months after birth, in the first year after birth, and currently (by mother and/or father), smoking habits of the grandparents (before the birth of the mother and father as well as currently); (6) pregnancy information: smoking during pregnancy, use of alcohol during pregnancy, age of the mother at delivery, increase of weight during pregnancy, parity; (7) parents’ characteristics: length and weight of mother and father; (8) health of the child: serious infections since birth and pubertal development. BMI of the parents (kg/m²) was calculated based on the self-reported height (m) and weight (kg). The highest level of education from either mother or father was used as a proxy for socioeconomic status. 2.4. Anthropometric data measured by the study nurses In October and November 2011, the children were visited at home by one of the two study nurses. The study nurses had no access to the prenatal exposure data when performing the measurements. Intra- and inter-observer reliability was enhanced by extensive training. The children were measured barefooted in underwear and /or T-shirt. Weight was measured once with an electronic scale (SECA 815, UK) to the nearest 0.1 kg. Height was measured once with a telescopic height measuring instrument (SECA 213, UK) to the nearest 0.1 cm. The body mass index (BMI) z-score was calculated, adjusting for age and sex using the British 1990 growth reference data (Cole et al., 1998). Skinfold thicknesses (mm) were measured twice on the right side of the body to the nearest 0.2 mm with a skinfold calliper (Holtain, UK, range 0–40 mm) according to the international standards for anthropometric measurement (Marfell-Jones, 2006) and the mean of both measurements was calculated. The tricipital and biceps skinfold were taken halfway between the acromion process and the olecranon process at the back side of the arm for the triceps and at the front side of the arm for the biceps. The subscapular skinfold was measured 20 mm below the tip of the scapula, at an angle of 451 to the lateral side of the body. The suprailiac skinfold measuring point was identified 2 cm above the iliac crest and 2 cm towards the medial line. It was aligned inferiomedially at 451 to the horizontal. The sum of all four skinfold thicknesses was calculated as an indicator of subcutaneous fat. Waist circumference (cm) was measured twice with an inelastic tape (Seca 200, precision 0.1 cm, range 0–150 cm) with the subject in a standing position with arms to the sides. Waist circumference (indicator of abdominal fat) measurement was performed horizontally halfway between the top of the iliac crest and the lower coastal border (10th rib). Waist-to-height ratio was used as a normalized index of body fat distribution. If the first and second measurement of the skinfolds and/or circumferences differed more than 2 mm, a third measurement was performed and the mean of all three measurements was calculated. For circumferences and skinfolds the mean value of the two or three measurements was calculated per child. We performed all statistical analyses with these mean values. 2.5. Statistical analyses Statistical analyses were performed using the SPSS for Windows software program (version 19.0). A two-sided p-value o 0.05 was considered significant. The normality of each outcome variable was examined by performing the Shapiro–Wilk test. Differences in markers of body composition between boys and girls were analysed with an independent samples T-test for the normal distributed markers

and with a Mann–Whitney U-test for the non-normal markers. For the non-responder analysis, an independent samples T-test was used to compare characteristics being continuous variables (age, BMI) between participants and non-participants. A chi-square test was used to compare characteristics being categorical variables (gender of the child, smoking behaviour, level of education) between participants and non-participants. The associations between prenatal exposure and body composition markers at the age of 7 to 9 years were analysed with linear regression analysis. Since the five considered EDCs influence the endocrine system via another mechanism (Legler et al., 2011), a separate regression model was used per EDC. Spearman correlations were calculated to investigate the correlations between the exposures to the different EDCs. The prenatal exposure data were ln-transformed and treated as a continuous variable. Some outcome variables (waist circumference, weight, the waist/height ratio and the sum of all 4 skinfolds) were ln-transformed to fulfill the conditions for linear regression. To facilitate the interpretation of the results, effect estimates were calculated for the effect of a prenatal exposure that would increase from P25 to P75 on the outcome, both expressed on the original scale. When both the exposure and the outcome were modelled on ln-scale, we used the following formula Y2 =Y1 ¼ ðP75=P25ÞB with B being the unstandardised regression coefficient. When the exposure was modelled on ln-scale and the outcome on original scale, the following formula was used: Y2–Y1 ¼ B  ln (P75/P25), with B being the unstandardised regression coefficient. First of all, univariate regression analyses between the five EDCs and the six anthropometric parameters were performed. Only in those cases where the p-value of the univariate regression was o0.200, a multiple regression analysis was performed. Second, effect modification (interaction) by gender was analysed in models including main effects and cross-product terms. A p-value o0.200 for the effect of the cross-product was taken as an indication of interaction and in that case stratified analyses by gender were performed in the next step. Third, multiple regression analyses were performed. A fixed set of confounders (parameters associated with both prenatal exposure and body composition) was applied based on pre-existing knowledge: maternal BMI, age of the mother at birth, smoking during pregnancy (yes/no), highest level of education of both parents (three categories) and total serum lipid content. The following covariates were included since evidence exist that these characteristics influence anthropometric parameters: age of the child and gender of the child (the latter was only included when the analyses were not stratified for gender). Data on food consumption and physical activity were collected via the questionnaire, but did not seem to be correlated with the body composition parameters. Therefore, they were not included in the multiple regression analyses. To avoid loss of subjects in the multiple regression analyses, missing values on maternal BMI (n¼ 1) and the highest level of education of both parents (n¼ 1) were imputed, using the median value for that variable.

3. Results Our study population consisted of 114 children between 7 and 9 years old (50% boys). Their mean age was 8.43 years. Table 2 provides an overview of the characteristics of the study population. The height differed significantly between boys and girls; boys were on average 4 cm (95% CI: 1.6–6.3 cm) taller compared to girls

28

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

Table 3 Characteristics of the participants of the follow-up study (n ¼114) and those who did not participate in the follow-up (n¼ 1082).

Age of the mother at delivery (mean7 SD) BMI of the mother before pregnancy (mean7 SD) Gender of the child: percentage boys Percentage of mothers that ever smoked Percentage of mothers that smoked during pregnancy Highest education level of both parents Percentage in class 1 Percentage in class 2 Percentage in class 3 a b

Participants (n¼ 114)

Non-participants (n¼1082)

p-Value

29.5 7 4.3 23.3 7 4.2 50.0% 30.7% 9.6%

30.6 7 3.8 23.17 4.0 52.2% 36.3% 16.9%

0.008 a 0.532 a 0.659 b 0.236 b 0.047 b o 0.001

20.4% 50.4% 29.2%

46.5% 34.4% 19.2%

b

p-Value of independent samples t-test. p-Value of chi-square test.

Table 4 Contaminants levels in cord blood or plasma. Participants at 7–9 years

PCB (lg/L) c Dioxins (CALUX) (pg CALUX-TEQ/L) c HCB (lg/L) c p,p0 -DDE (lg/L) c Cadmium (lg/L) d

Boys Median (n) [P25; P75] 0.144 (n¼ 53) [0.07; 0.25] 0.05 (n¼ 50) [0.02; 0.08] 0.04 (n¼ 53) [0.01; 0.08] 0.24 (n¼ 54) [0.14; 0.44] 0.17 (n¼ 54) [0.05; 0.59]

Non-participants Girls Median (n) [P25; P75] 0.14 (n¼ 55) [0.10; 0.26] 0.04 (n¼ 39) [0.02; 0.06] 0.05 (n ¼55) [0.01; 0.07] 0.23 (n ¼56) [0.12; 0.44] 0.21 (n ¼52) [0.05; 0.74]

p-Value 0.515 0.122 0.998 0.940 0.404

a

Total Median (n) [P25; P75] 0.14 (n ¼108) [0.09; 0.25] 0.05 (n ¼87) [0.02; 0.07] 0.04 (n ¼108) [0.01; 0.07] 0.24 (n¼ 110) [0.13; 0.44] 0.20 (n ¼52) [0.05; 0.65]

Total Median (n) [P25; P75] 0.14 (n¼ 947) [0.07; 0.22] 0.05 (n ¼770) [0.02; 0.07] 0.04 (n¼939) [0.02; 0.06] 0.22 (n¼ 1004) [0.13; 0.36] 0.24 (n ¼1006) [0.05; 0.66]

p-Value b 0.145 0.908 0.260 0.307 0.428

PCB ¼sum of PCB 138, 153 and 180; HCB ¼ hexachlorobenzene; p,p0 -DDE¼ para,para-dichlorodiphenyldichloroethylene. a

p-Value for difference between boys and girls. p-Value for difference between participants and non-participants. Measured in plasma. d Measured in whole blood. b c

(p ¼0.015). The highest education level amongst two parents was high school in 20% of participants, university college (to obtain a professional bachelor, academic bachelor or master degree) in 50% of the cases and university (to obtain an academic bachelor, master and/or doctoral degree) in another 30%. The frequency of sugar- and/or fat-rich foods varied from 9 to 27 times a week (7 1 to 4 times a day), with a median of 16 times a week (72.5 times a day). An analysis was conducted in order to study the difference between the participants of this follow-up study (n ¼114) and those who did not participate in the follow-up (n ¼1082). The results are presented in Table 3. There was no difference in gender proportion of the children between the participants and non-participants. Concerning the mothers, there was no significant difference in the percentage of mothers who had ever smoked, nor was there a difference in the pre-pregnancy BMI of the mothers between the participants and the non-participants. However, the highest education level of both parents was higher in the group of the participants whereas the percentage of mothers who smoked was lower in the group of the participants. Moreover, the age of the mothers at birth who participated was a bit higher compared to the non-participants. These three parameters (education level, smoking during pregnancy and age of the mother) were therefore included in the fixed set of confounders. The prenatal exposure levels are shown in Table 4. There were no significant differences in prenatal exposure between participating boys and girls, nor between the participants and the nonparticipants. Cadmium exposure was not correlated to any of the other EDCs exposures. Exposure to PCBs was correlated with exposure to dioxins (r ¼0.313; p ¼0.04), HCB (r ¼0.716; p o0.001) and p,p0 -DDE (r ¼0.705; po 0.001). Exposure to dioxins was not

correlated with HCB exposure, but was correlated with p,p0 -DDE exposure (r ¼0.227; p ¼0.036). Exposure to HCB and p,p0 -DDE were correlated with each other (r ¼ 0.682; p o0.001). Table 5 shows the results of the unadjusted regression analyses for the total groups of boys and girls. No significant relations were found between prenatal exposure to PCB and the different anthropometric parameters. Only in those cases where the p-value of the univariate regression was o 0.200, an adjusted regression analyses was performed. Next, the interaction effect of gender was analysed. An indication of interaction (p o0.200) was found for the following associations: HCB versus waist circumference and waist/height ratio, p,p0 -DDE versus waist and waist/height and Cd versus weight, BMI z-score, waist and sum of four skinfolds. The multiple regressions for these associations were stratified by gender. Table 6 shows the adjusted regression coefficients for those association with a p-value o0.200 in the univariate analyses (Table 5) and no indication of interaction effect of gender. After adjustment, none of the associations remained significant. Table 7 shows the adjusted regression coefficients for those association showing an indication of interaction effect of gender. Only for girls, significant associations were found between prenatal EDCs exposure and anthropometric parameters. After adjustment, a negative association was found in girls between prenatal cadmium exposure and weight, BMI z-score, waist circumference and the sum of four skinfolds. A prenatal Cd exposure increase from P25 to P75 (in girls from 0.05 mg/L to 0.74 mg/L) would lead to a BMI z-score being 0.75 lower and would lead to a decrease of the weight with 6%, of the waist circumference with 3% and of the sum of four skinfolds with 27%. This implicates that a higher prenatal cadmium exposure is associated with a decrease in both indicators of subcutaneous as well as

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

29

Table 5 Unadjusted regression coefficient (B and 95%CI) and p-value (p) for the univariate regression to investigate the association between prenatal contaminant exposure and body composition at 7–8 years. PCB (lg/L)

Height (cm) Weightc (kg) BMI z-score Waist circumference (cm) Waist/heightc

c

Sum of 4 skinfoldsc, (mm)

d

Dioxin (pg CALUX-TEQ/L)

a

HCB (lg/L)

a

Estimateb (95%CI) p

Estimateb (95%CI)

0.90. (  0.622; 2.429) 1.011 (  0.912; 1.031) 0.089 (  0.162; 0.341) 1.006 (0.994; 1.018) 1.006 (0.985; 1.028) 1.043 (0.941; 1.157)

0.070 1.710 (  1.238; 4.659) 0.889 1.034 (0.994; 1.075)  0.155 (  0.629; 0.515 0.352 (  0.125; 0.318) 0.827) 1.005 (0.983; 1.028) 0.678 1.020 (0.996; 1.042) 0.989 (0.949; 1.029) 0.570 1.030 (0.988; 1.073) 0.956 (0.778; 1.174) 0.661 1.189 (0.971; 1.453)

p

Estimateb (95%CI) p

0.243 2.514 (  0.207; 5.237) 0.296 1.003 (0.964; 1.042) 0.484 0.320 0.575 0.422

p,p0 -DDE (lg/L) a

a

Estimateb (95%CI)

0.253 1.066 (  0.636; 2.768) 0.092 1.023 (1.001; 1.047) 0.145 0.296 (0.022; 0.568) 0.099 1.014 (1.001; 1.026) 0.168 1.022 (0.999; 1.046) 0.092 1.086 (1.031; 1.218)

Cadmium (lg/L) a p

Estimateb (95%CI)

 2.288 (  4.750; 0.172) 0.036 0.970 (0.938; 1.001) 0.217

p 0.068 0.058

0.034  0.213 (  0.752; 0.187) 0.038 0.982 (0.967; 1.003)

0.066

0.067 0.980 (0.948; 1.013)

0.223

0.150 0.864 (0.733; 1.018)

0.079

0.291

a

Concentration on ln-scale; the regression coefficients with a p-value o 0.200 are indicated in bold. For outcomes modeled on ln-scale: it the exposure increases with IQR, the outcome is multiplied with the estimate. For outcomes on original scale: it the exposure increases with IQR, the outcome increases or decreases with the estimate. c Parameter is ln-transformed. d Sum of biceps, triceps, subscapular and suprailiacal skinfolds. b

Table 6 Adjusted regression coefficient (B and 95%CI) and p-value (p) for the multiple regressions to investigate the association between prenatal contaminant exposure and body composition at 7–8 years in the total group of boys and girls (adjusted for maternal BMI, age of the mother, smoking of mother during pregnancy, highest education level of both parents, total serum lipid content and age and gender of the child) Height (cm)

Dioxin (pg CALUX-TEQ/L)b

Estimatea (95%CI) 2.628 (  0.070; 5.327)

p 0.056 Weight (kg)c

BMI z-score Estimatea (95%CI) 0.142 (  0.352; 0.636)

HCB (lg/L)b

p 0.569

Estimatea (95%CI) 0.991 (  0.062; 0.508)

p 0.494

Weight (kg)c

BMI z-score

p,p0 -DDE (lg/L)b

Estimatea (95%CI) 1.014 (0.975; 1.054)

Skinfolds

p 0.125

Estimatea (95%CI) 1.012 (0.990; 1.036)

Estimatea (95%CI) 1.044 (0.849; 1.283) Skinfolds

p 0.263

c

p 0.983

b

Estimatea (95%CI) 1.031 (0.917; 1.159)

p 0.605

Height (cm)

Cadmium (lg/L)

b

B (95%CI)  1.172 (  3.599; 1.254)

p 0.340

a For outcomes modeled on ln-scale: it the exposure increases with IQR, the outcome is multiplied with the estimate. For outcomes on original scale: it the exposure increases with IQR, the outcome increases or decreases with the estimate. b Concentration on ln-scale. c Parameter is ln-transformed.

abdominal fat. Next, a positive association was found between prenatal p,p0 -DDE exposure and waist circumference as well as waist/height ratio in girls. A prenatal p,p0 -DDE exposure increase from P25 to P75 (in girls from 0.12 mg/L to 0.44 mg/L) would lead to an increase of the waist circumference with 2% and of the waist/ height ratio with 4%.

4. Discussion This study showed a significant association between prenatal EDCs exposure and anthropometric measurements, but only in girls. A negative association was found between prenatal cadmium exposure and both indicators of subcutaneous as well as abdominal fat in girls, whereas a positive association was found in girls for p,p0 -DDE in regard to markers of abdominal fat (waist circumference and waist/ height ratio). This study partly confirms existing hypotheses about the

relationship between environmental toxicants and obesity, a relatively new and not completely understood relation (Grant et al., 2013). 4.1. Effects of prenatal exposure to cadmium Previous studies reported a negative association between prenatal cadmium exposure and height and weight in later life (Lin et al., 2011; Tian et al., 2009), however no association with height was found in this study. The median level of prenatal exposure to cadmium in this study was slightly lower compared to the study performed by Lin et al. (2011) (0.20 mg/L versus 0.31 mg/L) and much smaller compared to the median exposure level to cadmium in the study by Tian et al. (2009) (0.20 mg/L versus 1.8 mg/L). Because this is the first study using circumferences and skinfolds for this hypothesis, comparison with previous results on these parameters is not possible. Comparing the effect of prenatal cadmium exposure on BMI with the effect of maternal smoking during pregnancy was possible using a study in a group of 5 to 11

30

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

Table 7 Adjusted regression coefficient (B and 95%CI) and p-value (p) for the multiple regressions to investigate the association between prenatal contaminant exposure and body composition at 7–8 years in boys and girls separately (adjusted for maternal BMI, age of the mother, smoking of mother during pregnancy, highest education level of both parents and age of the child); only those associations for which an indication for interaction of gender was found are analysed here (the others are indicated in grey). Cadmium (lg/L) Boys

Girls b

Weightc (kg) BMI z-score

Waist circumferencec (cm)

p,p0 -DDE (lg/L)

a

Boys b

Estimate (95%CI)

p

1.025 (0.978; 1.074) 0.405 (  0.222; 1.029) 1.007

0.3

0.937 (0.900; 0.979) 0.2  0.749 (  1.261;  0.237) 0.588 0.973

(0.980; 1.038)

(0.953; 0.997)

Estimate (95%CI)

p

Sum of 4 skinfolds (mm)

1.101 (0.845; 1.430)

0.47

0.728

HCB (lg/L) Girls

b

Estimate (95%CI)

p

a

Boys b

Estimate (95%CI)

p

Girls b

Estimate (95%CI)

p

Estimateb (95%CI)

p

0.003 0.005

0.029 0.99 (0.970; 1.012) 0.978 (0.940; 1.018)

Waist/heightc c

a

0.351 1.018

0.023 0.984

0.305 1.024

0.107

(1.003; 1.034) 0.273 1.037 (1.005; 1.071)

(0.947; 1.019) 0.025 0.953 (0.901; 1.021)

(0.994; 1.054) 0.167 1.050 (0.990; 1.115)

0.098

0.005

(0.587; 0.900)

a

Concentration on ln-scale. For outcomes modeled on ln-scale: it the exposure increases with IQR, the outcome is multiplied with the estimate. For outcomes on original scale: it the exposure increases with IQR, the outcome increases or decreases with the estimate. c Parameter is ln-transformed. b

year old children. In that study, they found a significant difference in BMI z-score, being 0.14 lower in children whose mother did not smoke during pregnancy (Koshy et al., 2011) whereas we found that a prenatal cadmium exposure that would increase from P25 to P75 would lead to a BMI z-score being 0.75 lower. In laboratory studies, cadmium was observed to be able to interact with both estrogen and androgen receptors and to activate the estrogen receptor alpha (Martin et al., 2002; Stoica et al., 2000). In view of the stimulation of lipid mobilization and lipolysis known to be induced by sex hormones (Hackney et al., 2000), this xeno-estrogenic activity might contribute to the negative association with body fat observed for girls, whereas the fact that cadmium is associated with lower testosterone levels in boys might explain the absence of a significant negative association in boys. Moreover, it has to be noted that – besides its endocrine disrupting effect – cadmium is known to affect the kidney and the liver, which can also play a role in the health effects seen in this study (Satarug and Moore, 2004). 4.2. Effects of prenatal exposure to p,p0 -DDE For p,p0 -DDE exposure, Verhulst et al. (2009) did report a significant increase in BMI at the age of 3 in this cohort. Likewise, we found that prenatal exposure to p,p0 -DDE lead to an increase in markers of abdominal fat in 7 to 9 year old girls. In other cohorts, the increasing effect of prenatal exposure to p,p0 -DDE on BMI and/or weight was also reported (Gladen et al., 2000; Karmaus et al., 2009; Valvi et al., 2012; Warner et al., 2013). However, other studies found no association between p,p0 -DDE and anthropometric measures (Cupul-Uicab et al., 2010; Garced et al., 2012; Gladen et al., 2004; Jusko et al., 2006). The exposure to p,p0 -DDE in this study is lower compared to the median exposure in other studies (Valvi et al. (2012) (0.24 mg/L versus 1.06 mg/L) and Gladen et al. (2000) (0.24 mg/L versus 3.95 mg/L)). Most studies measured the concentrations of p,p0 -DDE in maternal blood (Cupul-Uicab et al., 2010; Garced et al., 2012; Gladen et al., 2004; Jusko et al., 2006; Karmaus et al., 2009; Ribas-Fito et al., 2006; Warner et al., 2013), while this study used cord blood. A comparison is therefore not feasible. The mechanisms behind the effect of p,p0 -DDE on body fat are not yet fully understood. The presumed mechanism of action

behind the effects on body fat by p,p0 -DDE is the activation of the peroxisome proliferator-activated receptors (PPAR) which are thought to be involved in numerous pathways affecting the endocrine system (Grant et al., 2013).However, besides disrupting of the endocrine system, also other mechanisms, like oxidation stress or epigenetic mechanisms known to be linked to the exposure to p,p0 -DDE (Mrema et al., 2013) could have plaid a role in the effects seen in this study., 4.3. Effects of prenatal exposure to PCBs, dioxins and HCB Concerning PCBs, Verhulst et al. (2009) reported that PCBs lead to an increase in BMI at the age of 3 for the children of the same cohort. However, at the age of 7 to 9 years old, we did not find any association between prenatal PCB exposure and anthropometric parameters, which is in contrast to studies in other cohorts who did report an association with PCBs and increasing BMI and/or weight (Gladen et al., 2000; Karmaus et al., 2009; Valvi et al., 2012). Prenatal exposure to dioxins was not associated with BMI at the age of 3 for the children in this cohort (Verhulst et al., 2009) nor at the age of 7 to 9 year. In contrast to the positive association between prenatal dioxins exposure and weight reported by Su et al. (2010), we did not find such an association. This difference could be explained by a difference in prenatal exposure, but we are not able to make a comparison since the study by Su et al. (2010) measured the concentration of dioxins in the placenta while this study used cord blood. Another possible explanation may be the lack of statistical power due to the limited sample size in this study. Last, in regard to prenatal HCB exposure, no statistically significant associations with anthropometrics were found in our study after adjustment for confounders and covariates, nor in the previous results at the age of 3 (Verhulst et al., 2009). An earlier study showed that a higher prenatal exposure to HCB leads to a higher BMI in 6.5 year-old children (Smink et al., 2008), however this was not found in our study. This could be explained by the lower prenatal exposure to HCB in our study (0.04 mg/L versus 0.68 mg/L) or by a problem of statistical power due to the limited sample size.

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

4.4. Strengths and limitations of this follow-up study A first strength is the longitudinal design of this study. Prenatal exposure data were collected at birth and the effect of this exposure on birth weight was studied by Govarts et al. (2012). After 3 years, another analysis of the effects of prenatal exposure to EDCs and anthropometric measures was performed by Verhulst et al. (2009). At the age of 7 to 9, the children were revisited to collect a huge battery of anthropometric data and parents were asked to complete a questionnaire. A second strength includes the use of anthropometric data that were collected in a standardized way by two trained study nurses. This kind of data has a higher validity compared to self-reported anthropometric data. Moreover, it was possible to consider a set of different anthropometric parameters, not only height and weight but also waist circumference and skinfolds, which made it possible to distinguish markers of subcutaneous fat from markers of abdominal fat. Weight and height are the basic anthropometric measurements and they can be used to calculate BMI, which is used as an indicator of total body fat. However, when studying the effect of EDCs on children's body composition, it is relevant to differentiate between different types of body fat. Therefore, skinfold thickness and waist circumference measurements were performed in this study. Skinfold thickness assess the subcutaneous fat depots, whereas waist circumference can be used to asses abdominal fat. Waist/height ratio is frequently used in children to standardise the data for different ages and heights (Wells and Fewtrell, 2006). However, the sample size is quite limited, i.e. 114 children. This was due to the fact that the participating families had to agree to be visited at home. It was found that the education of the mothers was higher in the group of the participants (po0.001), the percentage of mothers who smoked was lower (p¼0.047) and the age of the mothers at birth who participated was a bit higher compared to the non-participants (30.5 versus 29.5 years, respectively, p¼ 0.015). To take these differences into consideration, these three parameters (education level, smoking during pregnancy and age of the mother) were included in the statistical analysis as confounders. There were no significant differences between the participants and the non-participants concerning prenatal exposure to cadmium, PCB, dioxins, p,p0 -DDE or HCB. Another limitation of this study is the lack of exposure data for the postnatal period up to the follow-up survey. It is probable that postnatal exposure to some of these contaminants also influences the body composition of the children. Humans are exposed to contaminants (including EDCs) through ingestion of contaminated food or water, inhalation of polluted air or from dermal exposure and these contaminants have a long half live, up to 30 years (Baillie-Hamilton, 2002). Next, it is possible that the five considered EDCs have synergistic or antagonistic effects influencing their actions, e.g. it has been shown that PCBs and PCDDs can act as antagonists (Sanctorum et al., 2007). However, the associations between EDCs and body markers are extremely complex (Grant et al., 2013). As a consequence, a multiple regression in which more than one parameter of internal exposure would be included could shed some light on this question. As a result, the quality of the statistical model would decrease with the addition of more independent variables, certainly when these variables show a certain level of correlation between them. In this paper we made the choice to use only one parameter of internal exposure in a single analysis.

5. Conclusion To the authors’ knowledge, this is the first study investigating the effect of prenatal exposure to EDCs on the body composition at the age of 7 to 9 years old, making use of a variety of anthropometric

31

parameters: weight, height, BMI, waist circumference, waist/height ratio, sum of four skinfolds. Only significant results were found in girls and only with prenatal exposure to p,p0 -DDE and cadmium (not with PCBs, dioxins nor HCB). Prenatal exposure to p,p0 -DDE was associated with an increase in makers of abdominal fat (waist circumference and waist/height ratio) in girls. Prenatal cadmium exposure was associated with a decrease in weight, BMI, waist circumference (indicators of abdominal fat) and with the sum of four skinfolds (indicator of subcutaneous fat) in girls. These results help in creating evidence in the role of prenatal exposure to ECDs in the development of obesity later in life. However, more studies with larger sample sizes using a variety of anthropometric parameters are needed in the future to further investigate this topic.

Funding sources The studies of the Flemish Center of Expertise on Environment and Health were commissioned, financed and steered by the Ministry of the Flemish Community (Department of Economics, Science and Innovation; Flemish Agency for Care and Health; and Department of Environment, Nature and Energy). Isabelle Sioen is financially supported by the Research Foundation—Flanders (Grant no: 1.2.683.11.N.00).

Ethical approval All participating parents provided informed consent for participation. The study was conducted according to the guidelines laid down in the Declaration of Helsinki and the study protocol was approved by the Ethical Committee of the University of Antwerp (Belgium) and the Ghent University (Belgium).

Acknowledgments The studies of the Flemish Center of Expertise on Environment and Health were commissioned, financed and steered by the Ministry of the Flemish Community (Department of Economics, Science and Innovation; Flemish Agency for Care and Health; and Department of Environment, Nature and Energy). Isabelle Sioen is financially supported by the Research Foundation—Flanders (Grant no: 1.2.683.11.N.00). References Apostoli, P., Magoni, M., Bergonzi, R., Carasi, S., Indelicato, A., Scarcella, C., Donato, F., 2005. Assessment of reference values for polychlorinated biphenyl concentration in human blood. Chemosphere 61, 413–421. Baillie-Hamilton, P.F., 2002. Chemical toxins: a hypothesis to explain the global obesity epidemic. J. Altern. Complement. Med. 8, 185–192. Cole, T.J., Freeman, J.V., Preece, M.A., 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Stat. Med. 17, 407–429. Collins, S., 2005. Overview of clinical perspectives and mechanisms of obesity. Birth Defects Res. A Clin. Mol. Teratol 73, 470–471. Covaci, A., Voorspoels, S., Thomsen, C., van, B.B., Neels, H., 2006. Evaluation of total lipids using enzymatic methods for the normalization of persistent organic pollutant levels in serum. Sci. Total Environ. 366, 361–366. Cupul-Uicab, L.A., Hernandez-Avila, M., Terrazas-Medina, E.A., Pennell, M.L., Longnecker, M.P., 2010. Prenatal exposure to the major DDT metabolite 1,1-dichloro2,2-bis(p-chlorophenyl)ethylene (DDE) and growth in boys from Mexico. Environ. Res. 110, 595–603. Eggesbo, M., Stigum, H., Longnecker, M.P., Polder, A., Aldrin, M., Basso, O., Thomsen, C., Skaare, J.U., Becher, G., Magnus, P., 2009. Levels of hexachlorobenzene (HCB) in breast milk in relation to birth weight in a Norwegian cohort. Environ. Res. 109, 559–566. Eskenazi, B., Mocarelli, P., Warner, M., Chee, W.Y., Gerthoux, P.M., Samuels, S., Needham, L.L., Patterson Jr., D.G., 2003. Maternal serum dioxin levels and birth outcomes in women of Seveso, Italy. Environ. Health Perspect. 111, 947–953.

32

I. Delvaux et al. / Environmental Research 132 (2014) 24–32

Garced, S., Torres-Sanchez, L., Cebrian, M.E., Claudio, L., Lopez-Carrillo, L., 2012. Prenatal dichlorodiphenyldichloroethylene (DDE) exposure and child growth during the first year of life. Environ. Res. 113, 58–62. Gladen, B.C., Klebanoff, M.A., Hediger, M.L., Katz, S.H., Barr, D.B., Davis, M.D., Longnecker, M.P., 2004. Prenatal DDT exposure in relation to anthropometric and pubertal measures in adolescent males. Environ. Health Perspect. 112, 1761–1767. Gladen, B.C., Ragan, N.B., Rogan, W.J., 2000. Pubertal growth and development and prenatal and lactational exposure to polychlorinated biphenyls and dichlorodiphenyl dichloroethene. J. Pediatr. 136, 490–496. Gomara, B., Ramos, L., Gonzalez, M.J., 2002. Determination of polychlorinated biphenyls in small-size serum samples by solid-phase extraction followed by gas chromatography with micro-electron-capture detection. J. Chromatogr. B 766, 279–287. Govarts, E., Nieuwenhuijsen, M., Schoeters, G., Ballester, F., Bloemen, K., de, B.M., Chevrier, C., Eggesbo, M., Guxens, M., Kramer, U., Legler, J., Martinez, D., Palkovicova, L., Patelarou, E., Ranft, U., Rautio, A., Petersen, M.S., Slama, R., Stigum, H., Toft, G., Trnovec, T., Vandentorren, S., Weihe, P., Kuperus, N.W., Wilhelm, M., Wittsiepe, J., Bonde, J.P., 2012. Birth weight and prenatal exposure to polychlorinated biphenyls (PCBs) and dichlorodiphenyldichloroethylene (DDE): a meta-analysis within 12 European Birth Cohorts. Environ. Health Perspect. 120, 162–170. Grant, K.L., Carpenter, D.O., Sly, L.J., Sly, P.D., 2013. Environmental contributions to obesity and type 2 diabetes. J. Environ. Immunol. Toxicol. 1, 80–91. Guo, S.S., Wu, W., Chumlea, W.C., Roche, A.F., 2002. Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am. J. Clin. Nutr. 76, 653–658. Hackney, A.C., Muoio, D., Meyer, W.R., 2000. The effect of sex steroid hormones on substrate oxidation during prolonged submaximal exercise in women. Jpn. J. Physiol. 50, 489–494. Hertz-Picciotto, I., Charles, M.J., James, R.A., Keller, J.A., Willman, E., Teplin, S., 2005. In utero polychlorinated biphenyl exposures in relation to fetal and early childhood growth. Epidemiology 16, 648–656. Iavicoli, I., Fontana, L., Bergamaschi, A., 2009. The effects of metals as endocrine disruptors. J. Toxicol. Environ. Health Part B 12, 206–223. Jacobson, J.L., Jacobson, S.W., Humphrey, H.E., 1990. Effects of exposure to PCBs and related compounds on growth and activity in children. Neurotoxicol. Teratol. 12, 319–326. Jusko, T.A., Koepsell, T.D., Baker, R.J., Greenfield, T.A., Willman, E.J., Charles, M.J., Teplin, S.W., Checkoway, H., Hertz-Picciotto, I., 2006. Maternal DDT exposures in relation to fetal and 5-year growth. Epidemiology 17, 692–700. Karmaus, W., Osuch, J.R., Eneli, I., Mudd, L.M., Zhang, J., Mikucki, D., Haan, P., Davis, S., 2009. Maternal levels of dichlorodiphenyl-dichloroethylene (DDE) may increase weight and body mass index in adult female offspring. Occup. Environ. Med. 66, 143–149. Konishi, K., Sasaki, S., Kato, S., Ban, S., Washino, N., Kajiwara, J., Todaka, T., Hirakawa, H., Hori, T., Yasutake, D., Kishi, R., 2009. Prenatal exposure to PCDDs/PCDFs and dioxin-like PCBs in relation to birth weight. Environ. Res. 109, 906–913. Koppen, G., Covaci, A., Van, C.R., Schepens, P., Winneke, G., Nelen, V., Schoeters, G., 2001. Comparison of CALUX-TEQ values with PCB and PCDD/F measurements in human serum of the Flanders Environmental and Health Study (FLEHS). Toxicol. Lett. 123, 59–67. Koppen, G., Den, H.E., Nelen, V., Van de Mieroop, E., Bruckers, L., Bilau, M., Keune, H., Van, L.N., Covaci, A., Van De Weghe, H., Schroijen, C., Desager, K., Stalpaert, M., Baeyens, W., Schoeters, G., 2009. Organochlorine and heavy metals in newborns: results from the Flemish Environment and Health Survey (FLEHS 2002–2006). Environ. Int. 35, 1015–1022. Koshy, G., Delpisheh, A., Brabin, B.J., 2011. Dose response association of pregnancy cigarette smoke exposure, childhood stature, overweight and obesity. Eur. J. Public Health 21, 286–291. Legler, J., Hamers, T., van Eck van der Sluijs-van de Bor, Schoeters, G., van, d., V, Eggesbo, M., Koppe, J., Feinberg, M., Trnovec, T., 2011. The OBELIX project: early life exposure to endocrine disruptors and obesity. Am. J. Clin. Nutr. 94, 1933S–1938S. Lin, C.M., Doyle, P., Wang, D., Hwang, Y.H., Chen, P.C., 2011. Does prenatal cadmium exposure affect fetal and child growth? Occup. Environ. Med 68, 641–646. Loiacono, N.J., Graziano, J.H., Kline, J.K., Popovac, D., Ahmedi, X., Gashi, E., Mehmeti, A., Rajovic, B., 1992. Placental cadmium and birthweight in women living near a lead smelter. Arch. Environ. Health 47, 250–255. Longnecker, M.P., Klebanoff, M.A., Brock, J.W., Guo, X., 2005. Maternal levels of polychlorinated biphenyls in relation to preterm and small-for-gestational-age birth. Epidemiology 16, 641–647. Longnecker, M.P., Klebanoff, M.A., Zhou, H., Brock, J.W., 2001. Association between maternal serum concentration of the DDT metabolite DDE and preterm and small-for-gestational-age babies at birth. Lancet 358, 110–114. Marfell-Jones M. International Standards for Anthropometric Assessment. Olds T, Stewart A Carter L. 2006. South Africa. Ref Type: Generic. Martin, M.B., Voeller, H.J., Gelmann, E.P., Lu, J.M., Stoica, E.G., Hebert, E.J., Reiter, R., Singh, B., Danielsen, M., Pentecost, E., Stoica, A., 2002. Role of cadmium in the regulation of AR gene expression and activity. Endocrinology 143, 263–275. Mrema, E.J., Rubino, F.M., Brambilla, G., Moretto, A., Tsatsakis, A.M., Colosio, C., 2013. Persistent organochlorinated pesticides and mechanisms of their toxicity. Toxicology 307, 74–88. Needham, L.L., Barr, D.B., Caudill, S.P., Pirkle, J.L., Turner, W.E., Osterloh, J., Jones, R.L., Sampson, E.J., 2005. Concentrations of environmental chemicals associated with neurodevelopmental effects in US population. Neurotoxicology 26, 531–545.

Newbold, R.R., 2010. Impact of environmental endocrine disrupting chemicals on the development of obesity. Hormones (Athens) 9, 206–217. Nishijo, M., Nakagawa, H., Honda, R., Tanebe, K., Saito, S., Teranishi, H., Tawara, K., 2002. Effects of maternal exposure to cadmium on pregnancy outcome and breast milk. Occup. Environ. Med. 59, 394–396. Odland, J.O., Nieboer, E., Romanova, N., Thomassen, Y., Lund, E., 1999. Blood lead and cadmium and birth weight among sub-arctic and arctic populations of Norway and Russia. Acta Obstet. Gynecol. Scand. 78, 852–860. Patandin, S., Koopman-Esseboom, C., de Ridder, M.A., Weisglas-Kuperus, N., Sauer, P.J, 1998. Effects of environmental exposure to polychlorinated biphenyls and dioxins on birth size and growth in Dutch children. Pediatr. Res. 44 (538–545). Pluim, H.J., van der Goot, M., Olie, K., van der Slikke, J.W., Koppe, J.G., 1996. Missing effects of background dioxin exposure on development of breast-fed infants during the first half year of life. Chemosphere 33, 1307–1315. Ribas-Fito, N., Gladen, B.C., Brock, J.W., Klebanoff, M.A., Longnecker, M.P., 2006. Prenatal exposure to 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene (p,p0 -DDE) in relation to child growth. Int. J. Epidemiol. 35, 853–858. Ribas-Fito, N., Sala, M., Cardo, E., Mazon, C., De Muga, M.E., Verdu, A., Marco, E., Grimalt, J.O., Sunyer, J., 2002. Association of hexachlorobenzene and other organochlorine compounds with anthropometric measures at birth. Pediatr. Res. 52, 163–167. Rogan, W.J., Gladen, B.C., McKinney, J.D., Carreras, N., Hardy, P., Thullen, J., Tinglestad, J., Tully, M., 1986. Neonatal effects of transplacental exposure to PCBs and DDE. J. Pediatr. 109, 335–341. Salpietro, C.D., Gangemi, S., Minciullo, P.L., Briuglia, S., Merlino, M.V., Stelitano, A., Cristani, M., Trombetta, D., Saija, A., 2002. Cadmium concentration in maternal and cord blood and infant birth weight: a study on healthy non-smoking women. J. Perinat. Med. 30, 395–399. Sanctorum, H., Elskens, M., Baeyens, W., 2007. Bioassay (CALUX) measurements of 2,3,7,8-TCDD and PCB 126: interference effects. Talanta 73, 185–188. Satarug, S., Moore, M.R., 2004. Adverse health effects of chronic exposure to lowlevel cadmium in foodstuffs and cigarette smoke. Environ. Health Perspect. 112, 1099–1103. 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. Schoeters, G., Den, H.E., Colles, A., Loots, I., Morrens, B., Keune, H., Bruckers, L., Nawrot, T., Sioen, I., De, C.S., Van, L.N., Nelen, V., Van de Mieroop, E., Vrijens, J., Croes, K., Goeyens, K., Baeyens, W., 2012. Concept of the Flemish human biomonitoring programme. Int. J. Hyg. Environ. Health 215, 102–108. Schroijen, C., Baeyens, W., Schoeters, G., Den, H.E., Koppen, G., Bruckers, L., Nelen, V., Van de Mieroop, E., Bilau, M., Covaci, A., Keune, H., Loots, I., Kleinjans, J., Dhooge, W., Van, L.N., 2008. Internal exposure to pollutants measured in blood and urine of Flemish adolescents in function of area of residence. Chemosphere 71, 1317–1325. Smink, A., Ribas-Fito, N., Garcia, R., Torrent, M., Mendez, M.A., Grimalt, J.O., Sunyer, J., 2008. Exposure to hexachlorobenzene during pregnancy increases the risk of overweight in children aged 6 years. Acta Paediatr. 97, 1465–1469. Stoica, A., Katzenellenbogen, B.S., Martin, M.B., 2000. Activation of estrogen receptor-alpha by the heavy metal cadmium. Mol. Endocrinol. 14, 545–553. Su, P.H., Chen, J.Y., Chen, J.W., Wang, S.L., 2010. Growth and thyroid function in children with in utero exposure to dioxin: a 5-year follow-up study. Pediatr. Res. 67, 205–210. Tang-Peronard, J.L., Andersen, H.R., Jensen, T.K., Heitmann, B.L., 2011. Endocrinedisrupting chemicals and obesity development in humans: a review. Obes. Rev. 12, 622–636. Tawara, K., Nishijo, M., Honda, R., Maruzeni, S., Seto, T., Kido, T., Saito, S., Nakagawa, H., 2009. Effects of maternal dioxin exposure on newborn size at birth among Japanese mother-infant pairs. Environ. Health Prev. Med 14, 88–95. Tian, L.L., Zhao, Y.C., Wang, X.C., Gu, J.L., Sun, Z.J., Zhang, Y.L., Wang, J.X., 2009. Effects of gestational cadmium exposure on pregnancy outcome and development in the offspring at age 4.5 years. Biol. Trace Elem. Res. 132, 51–59. Valvi, D., Mendez, M.A., Martinez, D., Grimalt, J.O., Torrent, M., Sunyer, J., Vrijheid, M., 2012. Prenatal concentrations of polychlorinated biphenyls, DDE, and DDT and overweight in children: a prospective birth cohort study. Environ. Health Perspect. 120, 451–457. Vartiainen, T., Jaakkola, J.J., Saarikoski, S., Tuomisto, J., 1998. Birth weight and sex of children and the correlation to the body burden of PCDDs/PCDFs and PCBs of the mother. Environ. Health Perspect. 106, 61–66. Verhulst, S.L., Nelen, V., Hond, E.D., Koppen, G., Beunckens, C., Vael, C., Schoeters, G., Desager, K., 2009. Intrauterine exposure to environmental pollutants and body mass index during the first 3 years of life. Environ. Health Perspect. 117, 122–126. Warner, M., Schall, R.A., Harley, K.G., Bradman, A., Barr, D., Eskenazi, B., 2013. In utero DDT and DDE exposure and obesity status of 7-year-old Mexican– American children in the CHAMACOS cohort. Environ. Health Perspect. 121, 631–636. Wells, J.C., Fewtrell, M.S., 2006. Measuring body composition. Arch. Dis. Child 91, 612–617. Wojtyniak, B.J., Rabczenko, D., Jonsson, B.A., Zvezday, V., Pedersen, H.S., Rylander, L., Toft, G., Ludwicki, J.K., Goralczyk, K., Lesovaya, A., Hagmar, L., Bonde, J.P., 2010. Association of maternal serum concentrations of 2,20 ,4,40 5,50 -hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis(p-chlorophenyl)-ethylene (p,p0 -DDE) levels with birth weight, gestational age and preterm births in Inuit and European populations. Environ. Health 9, 56.