Early Life Course Risk Factors for Childhood Obesity: The IDEFICS Case-Control Study

Abstract

Background

The early life course is assumed to be a critical phase for childhood obesity; however the significance of single factors and their interplay is not well studied in childhood populations.

Objectives

The investigation of pre-, peri- and postpartum risk factors on the risk of obesity at age 2 to 9.

Methods

A case-control study with 1,024 1∶1-matched case-control pairs was nested in the baseline survey (09/2007–05/2008) of the IDEFICS study, a population-based intervention study on childhood obesity carried out in 8 European countries in pre- and primary school settings. Conditional logistic regression was used for identification of risk factors.

Results

For many of the investigated risk factors, we found a raw effect in our study. In multivariate models, we could establish an effect for gestational weight gain (adjusted OR = 1.02; 95%CI 1.00–1.04), smoking during pregnancy (adjusted OR = 1.48; 95%CI 1.08–2.01), Caesarian section (adjusted OR = 1.38; 95%CI 1.10–1.74), and breastfeeding 4 to 11 months (adjusted OR = 0.77; 95%CI 0.62–0.96). Birth weight was related to lean mass rather than to fat mass, the effect of smoking was found only in boys, but not in girls. After additional adjustment for parental BMI and parental educational status, only gestational weight gain remained statistically significant. Both, maternal as well as paternal BMI were the strongest risk factors in our study, and they confounded several of the investigated associations.

Conclusions

Key risk factors of childhood obesity in our study are parental BMI and gestational weight gain; consequently prevention approaches should target not only children but also adults. The monitoring of gestational weight seems to be of particular importance for early prevention of childhood obesity.

About the Authors:

Karin Bammann

* E-mail: [email protected]

Affiliations Institute for Public Health and Nursing Research (IPP), University of Bremen, Bremen, Germany, BIPS Institute for Prevention Research and Epidemiology, Bremen, Germany

Jenny Peplies

Affiliation: BIPS Institute for Prevention Research and Epidemiology, Bremen, Germany

Stefaan De Henauw

Affiliation: Department of Public Health, Ghent University, Ghent, Belgium

Monica Hunsberger

Affiliation: Department of Public Health and Community Medicine, University of Gothenburg, Gothenburg, Sweden

Denes Molnar

Affiliation: Department of Pediatrics, University of Pécs, Pécs, Hungary

Luis A. Moreno

Affiliation: Growth, Exercise, Nutrition and Development (GENUD) Research Group, University of Zaragoza, Zaragoza, Spain

Michael Tornaritis

Affiliation: Research and Education Foundation of Child Health, Strovolos, Cyprus

Toomas Veidebaum

Affiliation: Department of Chronic Diseases, National Institute for Health Development, Tallinn, Estonia

Wolfgang Ahrens

Affiliation: BIPS Institute for Prevention Research and Epidemiology, Bremen, Germany

Alfonso Siani

Affiliation: Unit of Epidemiology and Population Genetics, Institute of Food Sciences, National Research Council, Avellino, Italy

on behalf of the IDEFICS consortium

Introduction

Excess body fat is one of the major health concerns in childhood populations [1], [2]. Although the energy imbalance resulting from high energy intake and low energy expenditure can contribute to the development of overweight and obesity in children, there is no general consensus about its importance as the main driver of weight gain [3]. In recent years, factors early in the life course emerged as possible determinants of early overweight and obesity [4], [5]. These can lead to epigenetic changes which are associated with the programming of metabolic outcomes in the offspring. Different pre-, peri- and postnatal risk factors for childhood obesity have been investigated and there is an increasing need to disentangle the contribution of factors like maternal weight, gestational weight gain, glycemic control, smoking and alcohol use during pregnancy, birth weight, breast feeding on the adiposity of the offspring, for which partly inconsistent results have been obtained [5]. While high birth weight was consistently shown to be a risk factor for obesity, the situation is less clear with low birth weight that might or might not play a role in the development of subsequent overweight and obesity [6]. Likewise, the impact of breastfeeding on childhood obesity is still a matter of research [7]. Although a protective effect of breastfeeding for subsequent obesity of the offspring has been demonstrated in several studies, this result might at least partly be due to confounding: Amir and Donath show in their studies that maternal smoking and obesity lead to a lesser probability of breastfeeding [8], [9], both, maternal smoking and maternal obesity, being themselves strong risk factors for childhood overweight [5]. Correspondingly, it is suspected by some authors that the association of breastfeeding and childhood overweight is a statistical artifact rather than a true causal association [10].

In summary, there is still a considerable lack of research regarding the effect of early life course factors on the risk of childhood obesity.

The IDEFICS study is a large European multi-center intervention study on childhood obesity. Prior to the intervention phase, a baseline survey was conducted in the control and intervention regions to assess a wealth of factors related to diet- and lifestyle-related diseases in childhood. We used this data to set up a matched case-control study on obesity. The aim of the paper is to investigate the impact of pre-, peri- and postnatal risk factors on the subsequent risk of obesity within this case-control study.

Materials and Methods

Ethics Statement

We certify that all applicable institutional and governmental regulations concerning the ethical use of human volunteers were followed during this research, and that the IDEFICS project passed the Ethics Review process of the Sixth Framework Programme (FP6) of the European Commission. Ethical approval was obtained from the relevant local or national ethics committees by each of the eight study centers, namely from the Ethics Committee of the University Hospital Ghent (Belgium), the National Bioethics Committee of Cyprus (Cyprus), the Tallinn Medical Research Ethics Committee of the National Institutes for Health Development (Estonia), the Ethics Committee of the University Bremen (Germany), the Scientific and Research Ethics Committee of the Medical Research Council Budapest (Hungary), the Ethics Committee of the Health Office Avellino (Italy), the Ethics Committee for Clinical Research of Aragon (Spain), and the Regional Ethical Review Board of Gothenburg (Sweden). All parents or legal guardians of the participating children gave written informed consent to data collection, examinations, collection of samples, subsequent analysis and storage of personal data and collected samples. Additionally, each child gave oral consent after being orally informed about the modules by a study nurse immediately before every examination using a simplified text. This procedure was chosen due to the young age of the children. The oral consenting process was not further documented, but it was subject to central and local training and quality control procedures of the study. Study participants and their parents/legal guardians could consent to single components of the study while abstaining from others. All procedures were approved by the above-mentioned Ethics Committees.

Study Sample

IDEFICS is a multi-center population-based intervention study on childhood obesity that is carried out in selected regions of 8 European countries comprising Belgium, Cyprus, Estonia, Germany, Hungary, Italy, Spain and Sweden. The study was set up in pre- and primary school settings in a control and an intervention region in each of these countries. Two major surveys (baseline and follow-up) were conducted in pre-schools and primary school classes (1st and 2nd grades at baseline). The baseline survey (September 2007–May 2008) reached a response proportion of 51% and included 16 220 children aged 2 to 9 years. The general design of the IDEFICS study has been described elsewhere [11]. A brief description of the study regions can be found in Bammann et al. [12].

Anthropometric measurements were done during a physical examination. Weight to the nearest 0.1 kg and foot-to-foot bioelectrical resistance in Ohm was measured using an electronic scale TANITA BC 420 SMA (TANITA Europe GmbH, Sindelfingen, Germany) with the children being in a fasting status and wearing only underwear. Standing height was measured with the children’s head in a Frankfort plane using a stadiometer SECA 225 (seca GmbH & Co. KG., Hamburg, Germany) to the nearest 0.1 cm. As in the weight measurement, the children were wearing only underwear, all hair ornaments were removed and all braids undone. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m). For obtaining the International Task Force of Obesity (IOTF) category, BMI categories were interpolated for continuous age as proposed [13], [14]. Cubic splines were used for this interpolation. The resistance index (RI) was calculated as squared height (cm2) divided by resistance (Ohm). The RI was shown to be a good predictor for fat free mass in children [15].

Within the baseline survey, a self-administered questionnaire was filled in by the parents to gather information on the children’s behavior, parental attitudes and on the social microenvironment of the children (IDEFICS parental questionnaire). A further questionnaire on health-related and medical information was given to the parents in the course of the physical examination of the child (IDEFICS medical questionnaire). Questionnaires were developed in English, translated to the respective languages and back translated to English to minimize any heterogeneity due to translation problems. Different language versions were available in the centers, and help was offered to those parents who felt they were not able to fill in the questionnaire by themselves. For filling in the IDEFICS medical questionnaire, the help of medical personnel was offered directly at the survey sites.

Eligible for the IDEFICS case-control study on obesity were all children from the baseline survey for whom a basic set of anthropometric indicators and pregnancy information was present (N = 16,113). From these children, all children aged 4 to 8 years that met the IOTF criteria for childhood obesity were identified as cases (N = 1,024). Controls were selected from the group of IOTF normal weighted children (N = 11,193) and matched 1∶1 by study center, sex and age (sliding window +/−0.5 years) to the cases. For all cases, controls with identical sex and study center and a minimal distance with respect to age were selected. The decision for a 1∶1 matched design was based on a sample size calculation. In our study, risk factors with prevalences among controls of at least 10% leading to odds ratios of 1.6 and more can be detected with a power of more than 90% (two-sided, p = 0.05).

Investigated Risk Factors

The investigated risk factors directly related to the child can be grouped into prepartum, peripartum, and postpartum factors (see Table 1). Further family-related risk factors included in the present analysis are familial clustering and parental social background. All information was assessed by the IDEFICS parental questionnaire, except for the occurrence of gestational diabetes that was taken from the IDEFICS medical questionnaire. All investigated risk factors are based on self-report.

[Figure omitted. See PDF.]

Table 1. Risk factors investigated in the study.

https://doi.org/10.1371/journal.pone.0086914.t001

Included prepartum factors were smoking during pregnancy, gestational weight gain and gestational diabetes. The questions relating to smoking during pregnancy and to gestational weight gain were posed to biological mothers, only. Regarding smoking during pregnancy, possible answer categories were “Never”, “Rarely, at maximum once a month”, “Several occasions a week” or “Daily”. We categorized “Rarely, at maximum once a month” or more often as smoking during pregnancy. Validity of maternal recall of smoking during pregnancy and of gestational weight gain is high even after several decades [16], [17].

Included peripartum factors were birth weight and information whether the child was delivered by Caesarian section. The validity of parental long-term recall of birth weight and of Caesarian sections is good to excellent [16]–[18], and is not related to time of recall since birth [19].

Included postpartum factors were duration of breastfeeding and introduction of solid foods. The early infant feeding of the child was assessed by a table indicating different types of feeding where starting and ending age could be given by the parents. The exact wording of the question was “What type of feeding with your child was used prior to being fully integrated into the usual household diet?”. Length of breastfeeding was calculated by subtracting starting age from ending age of breastfeeding, either exclusive or in combination with other types of feeding. The introduction of solid foods was assessed using the question “At what age did you first introduce …?” followed by a table with 5 different foods and the possibility of giving the time of introduction as age of the child in month. Early introduction of solid food was defined as introducing cereals, meat, vegetables or fruits at a child’s age before 4 months. Previous research has shown that the validity of maternal recall for duration of breastfeeding is very high, even after 10 years and more [16], [17], [20], [21]. Repeatability and validity for maternal recall of introduction of solid foods is shown to be higher than for introduction of fluids other than breast milk, but considerably lower than for breastfeeding with no clear direction of bias towards wrong recall of earlier or later dates [20], [22].

The familial clustering of overweight and obesity was assessed using self-reported parental BMI. The parental BMI was calculated as weight (kg)/squared height (m2).

The parental social background was described using the parental education and the age of mother at birth. For the analyses, these were transferred country-by-country to International Standard Classification of Education (ISCED) levels [12], [23] and the maximum ISCED level of both parents was calculated.

Statistical Analyses

To evaluate the impact of a putative risk factor, conditional logistic regression models were fitted to the data. This allows an unbiased estimation of effects in matched case-control studies with regard to the matching variables. Within this approach, each matched set forms a stratum. In the case of 1∶1-matched pairs, the likelihood of being a case for N strata is then given by:

The estimation of the βs was done using Cox’ proportional hazard model with maximum likelihood estimation [24]. 95% confidence intervals (95% CI) were computed.

For all variables, the influence of the child’s sex and age group (< = 6 years, >6 years) on the raw odds ratio (OR) was tested using the Breslow-Day test for homogeneity of the odds ratios [25]. We found all OR to be homogenous regarding age group. OR that were found to be heterogeneous over sexes are reported in the text.

For each of the investigated factors, we built multivariate models adjusting for known or suspected confounders from the literature (cf. Table 1). Since many of the investigated risk factors are correlated, we finally built a multivariate model containing all factors that were shown to be influential in the analyses. We reported the Wald statistics to judge the relative importance of the single factors.

All statistical analyses were done with SAS 9.2 (SAS Institute, Cary (NC), USA).

Raw ORs for parental BMI and parental ISCED level can be found in the Appendix (Table S1).

Results

A basic description of the study sample is displayed in Table 2. Roughly 50% of the case-control-pairs are male, 48% are 4–6 years, and 52% are 7–8 years old. The case-control pairs show large differences with respect to country of origin, ranging from 39.7% from Italy to 3.9 from Belgium and 2.7% from Sweden.

[Figure omitted. See PDF.]

Table 2. Description of the study sample.

https://doi.org/10.1371/journal.pone.0086914.t002

The investigated early life course risk factors are shown in Table 3. Maternal smoking during pregnancy carried a 50% higher risk of the child being a case. Stratification by sex revealed that this elevated risk was present only in boys (OR = 2.15; 95% CI 1.43–3.23), but not in girls (OR = 1.13; 95% CI 0.78–1.62). The elevated obesity risk of maternal smoking during pregnancy remained practically unchanged when adjusting for parental BMI and parental educational level. Gestational weight gain showed a linear dose-response relationship mounting in a twofold risk of obesity if the mother gained 25 kg and more during pregnancy, which was not explained by maternal BMI. Also, gestational diabetes was associated with an excess risk of 32%. After adjustment for maternal BMI the elevated risk disappeared almost completely (OR = 1.05; 95%CI 0.57–1.94).

[Figure omitted. See PDF.]

Table 3. IOTF obesity risk of early life course factors.

https://doi.org/10.1371/journal.pone.0086914.t003

The matched raw odds ratios for birth weight show a clear linear dose-response relationship with childhood obesity (OR = 0.58; 95%CI 0.39–0.86 for a birth weight <2.5 kg up to OR = 1.83; 95%CI 1.26–2.65 for a birth weight of >4 kg). However, when we adjusted for the RI as an indicator of fat free mass of the child, the effect of birth weight on the subsequent obesity risk disappeared completely. Being born via Caesarian section carried a higher obesity risk in our study (OR = 1.43; 95%CI 1.17–1.75). After adjustment for maternal BMI, maternal weight gain during pregnancy and initiation of breastfeeding, this risk was reduced and no longer statistically significant (OR = 1.27; 95%CI 0.99–1.64). The initiation of any breastfeeding carries an unadjusted OR of 0.75 (95%CI 0.61–0.91), and is thus protective against subsequent obesity in childhood. However, parental educational level, maternal BMI and maternal smoking during pregnancy almost completely explained this effect; the adjusted OR is near to unity (OR = 0.91; 95%CI 0.70–1.18). Likewise, the duration of breastfeeding shows unadjusted a u-shaped relationship to obesity with a protective effect only for the two middle categories (4–6 months, 7–11 months of non-exclusive breastfeeding). After adjustment for the afore-mentioned confounders, only 7–11 moths of non-exclusive breast-feeding remained statistically significant protective (OR = 0.70; 95%CI 0.49–0.99). Early introduction of solid foods shows an unadjusted OR of 1.43 (95% CI 1.02–2.01) that is reduced to a non-significant 1.33 (95% CI 0.93–1.90) after adjustment for parental educational level and initiation of breast-feeding.

To assess the combined effect of all investigated factors, we built two multivariate models (Table 4). The first model included gestational weight gain in kg, and design variables for smoking during pregnancy, Caesarian section, early introduction of solid foods and breastfeeding 4 to 11 months. The second model additionally included the confounding factors parental BMI in kg/m2 and parental educational level (unadjusted matched OR of these confounders can be found in Table A of the appendix).

[Figure omitted. See PDF.]

Table 4. Multivariate models for pre-, peri- and postpartum risk factors on IOTF obesity risk.

https://doi.org/10.1371/journal.pone.0086914.t004

In the overall group, statistically significant risk factors for obesity were Caesarian section (adjusted OR = 1.38; 95%CI 1.10–1.74), gestational weight gain in kg (adjusted OR = 1.02; 95%CI 1.00–1.04), maternal smoking during pregnancy (adjusted OR = 1.48; 95%CI 1.08–2.01) and breastfeeding 4 to 11 months (adjusted OR = 0.77; 95%CI 0.62–0.96). After additional adjustment for parental BMI and parental educational level, statistically significant factors were maternal BMI in kg/m2 (adjusted OR = 1.16; 95%CI 1.11–1.20), paternal BMI in kg/m2 (adjusted OR = 1.11; 95%CI 1.07–1.16), and gestational weight gain in kg (adjusted OR = 1.04; 95%CI 1.01–1.07).

Discussion

This paper investigated the impact of early life course risk factors on the risk of subsequent childhood obesity. For many of the investigated risk factors, we found a raw effect in our study. Our results regarding maternal smoking are confirmed by literature. In studies where maternal smoking was assessed, this factor is quite consistently reported to be a marked risk factor [26]–[28]. Similarly, high gestational weight gain was previously shown to be a risk factor not only for overweight and obesity of the child [29], but also of the mother [30]. Sometimes this effect is attributed to the influence of maternal adiposity [5], however in our study, gestational weight gain showed an independent effect, also after control for parental BMI. We were well in accordance with the literature, regarding high birth weight which was also a risk factor in our study [6], [26], [31]. However, when we adjusted for fat-free mass, this effect was completely removed, corroborating the hypothesis that birth weight is mainly influencing lean body mass and not fat mass [32]. Since the BMI is correlated with fat mass and with fat-free mass, one could argue that adjusting for fat-free mass leads generally to over-adjustment. Since the difference of fat-free mass between normal-weight and obese children is much lower than that of fat mass [33], we argue that adjustment for fat-free mass should not be able to mask any true effect on fat mass. Moreover, we checked for each of the other investigated factors whether adjustment for fat-free mass influences the OR. This was, apart from the effect for birth weight, not the case.

Caesarian section is another putative risk factor for childhood obesity that came into focus rather recently [34]. We also did find an elevated risk for obesity in the offspring, also after controlling for gestational gain, maternal BMI and breastfeeding initiation. However, in the final multivariate model Caesarian section was not of particular importance, when judged by the Wald statistic, and no longer statistically significant. The impact breastfeeding exerts on overweight is controversial since it is influenced by several other risk factors, as e.g. socioeconomic status, maternal BMI and maternal smoking [35]. One of the rare RCTs showed no effects of feeding breast milk on the prevalence of obesity until the age of 15 [36], suggesting that the protective effects found in other studies were due to statistical artifacts. Similarly, the RCT of Kramer et al. on breastfeeding promotion in preterm babies in Belarus showed no difference regarding obesity prevalence later in life in the two arms of the study [37]. Many previous studies have investigated either ever versus never-breastfeeding, while fewer have studied the duration of breastfeeding. Moreover, duration of breast-feeding is frequently modeled as months of breastfeeding assuming thus a strictly linear effect without any upper limit. In our study, we categorized the duration of breastfeeding and found a u-shaped effect. Interestingly, a similar pattern was found in a cohort study in Brazil [38] and in a secondary analyses of several cohorts examining the risk of duration of breastfeeding on cardiovascular risk factors [39]. However in our study, the effects were no longer statistically significant when controlling for confounding factors. As many other studies [5], [40]–[42], we also found a strong impact of parental weight status on the risk of childhood obesity. In a full model that included all investigated early life course risk factors for which we could establish an effect, gestational weight gain, smoking during pregnancy, Caesarian section were found to be risk factors for later childhood obesity, and breastfeeding 4 to 11 months was found to be protective. However, after additional adjustment for parental BMI and parental educational status, only gestational weight gain remained statistically significant. Both, maternal as well as paternal BMI were the strongest risk factors in our study, and they confounded several of the investigated associations. The current study has several limitations. It has to be kept in mind that the IDEFICS regions are not representative of the respective countries of Europe, let alone of the European population as a whole. Moreover, due to the high number of Italian obese children in the IDEFICS baseline survey (and the low number in Sweden and Belgium), the case-control sample was highly unbalanced with regard to country. As country was a matching variable this should not bias the results. We did not test for different effects by country, as we did for sex and age group, since the sample size was too small for most countries. For most of the associations that are hypothesized to be of a physiologic nature, like e.g. the association between Caesarian section and obesity, generalizability should not be compromised by the choice of regions. However, the association of social factors, like parental education, with childhood obesity is strongly dependent on context and therefore of limited external validity [12].

The investigated risk factors all rely on parental self-report. The validity of parental recall of most of the investigated factors is quite high, and we did not find much evidence for differential misclassification, which means that recall bias leads to risk attenuation rather than to distorted results. Recall bias might play a role in the lacking effect for introduction of solid food, as this factors has probably lowest validity among all investigated factors. Parental BMI is calculated using self-reported height and weight. Self-reported BMI is known to underestimate the true BMI, especially in the obese [43]. Assuming a familial clustering of obesity this produces a likely differential misclassification for parental BMI in our cases and controls. This misclassification also leads to attenuated Odds Ratios for paternal and maternal BMI, since especially among cases the proportion of overweight and obese parents is underestimated as opposed to the control group.

Our results show that parental BMI is of particular importance for childhood obesity. The parental BMI is a factor or an indicator for very different causal pathways, belonging to categories of shared genetics, a shared obesogenic environment or the process known as early programming. Currently many obesity intervention programs target children. However, reducing the prevalence of parental overweight and obesity would not only help preventing childhood obesity, but would in general lead to an improved health not only of the children, but also of their parents. Beyond this, our results indicate that especially maternal weight gain should be monitored closely during pregnancy. The mechanisms and possible prevention targets of maternal diet [44], [45] and other social and behavioral factors before and during pregnancy as risk factors for parental BMI and high gestational weight gain as well as for obesity in the offspring should be the subject of further studies.

Supporting Information

[Figure omitted. See PDF.]

Table S1.

Distribution of continuous variables in the study population.

https://doi.org/10.1371/journal.pone.0086914.s001

(DOCX)

Author Contributions

Analyzed the data: KB JP. Contributed reagents/materials/analysis tools: KB JP SDH MH DM LAM MT TV WA AS. Wrote the paper: KB JP MH AS. Conceived and designed the study: KB DM LAM WA AS.

Citation: Bammann K, Peplies J, De Henauw S, Hunsberger M, Molnar D, Moreno LA, et al. (2014) Early Life Course Risk Factors for Childhood Obesity: The IDEFICS Case-Control Study. PLoS ONE 9(2): e86914. https://doi.org/10.1371/journal.pone.0086914

References

1. Fiese BH, Bost KK, McBride BA, Donovan SM (2013) Childhood obesity prevention from cell to society. Trends Endocrinol Metab 24: 375–377.

2. Fruhbeck G, Toplak H, Woodward E, Yumuk V, Maislos M, et al. (2013) Obesity: the gateway to ill health - an EASO position statement on a rising public health, clinical and scientific challenge in Europe. Obes Facts 6: 117–120.

3. Bleich SN, Ku R, Wang YC (2011) Relative contribution of energy intake and energy expenditure to childhood obesity: a review of the literature and directions for future research. Int J Obes (Lond) 35: 1–15.

4. Heerwagen MJ, Miller MR, Barbour LA, Friedman JE (2010) Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. Am J Physiol Regul Integr Comp Physiol 299: R711–722.

5. Tounian P (2011) Programming towards childhood obesity. Ann Nutr Metab 58 Suppl 230–41.

6. Yu ZB, Han SP, Zhu GZ, Zhu C, Wang XJ, et al. (2011) Birth weight and subsequent risk of obesity: a systematic review and meta-analysis. Obes Rev 12: 525–542.

7. Beyerlein A, von Kries R (2011) Breastfeeding and body composition in children: will there ever be conclusive empirical evidence for a protective effect against overweight? Am J Clin Nutr 94: 1772S–1775S.

8. Amir LH, Donath SM (2003) Does maternal smoking have a negative physiological effect on breastfeeding? The epidemiological evidence. Breastfeed Rev 11: 19–29.

9. Amir LH, Donath S (2007) A systematic review of maternal obesity and breastfeeding intention, initiation and duration. BMC Pregnancy Childbirth 7: 9.

10. Bartok CJ, Ventura AK (2009) Mechanisms underlying the association between breastfeeding and obesity. Int J Pediatr Obes 4: 196–204.

11. Ahrens W, Bammann K, Siani A, Buchecker K, De Henauw S, et al. (2011) The IDEFICS cohort: design, characteristics and participation in the baseline survey. Int J Obes (Lond) 35 Suppl 1S3–15.

12. Bammann K, Gwozdz W, Lanfer A, Barba G, De Henauw S, et al. (2013) Socioeconomic factors and childhood overweight in Europe: results from the multi-centre IDEFICS study. Pediatr Obes 8: 1–12.

13. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH (2000) Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320: 1240–1243.

14. Cole TJ, Flegal KM, Nicholls D, Jackson AA (2007) Body mass index cut offs to define thinness in children and adolescents: international survey. BMJ 335: 194.

15. Houtkooper LB, Lohman TG, Going SB, Hall MC (1989) Validity of bioelectric impedance for body composition assessment in children. J Appl Physiol 66: 814–821.

16. Tomeo CA, Rich-Edwards JW, Michels KB, Berkey CS, Hunter DJ, et al. (1999) Reproducibility and validity of maternal recall of pregnancy-related events. Epidemiology 10: 774–777.

17. Jaspers M, de Meer G, Verhulst FC, Ormel J, Reijneveld SA (2010) Limited validity of parental recall on pregnancy, birth, and early childhood at child age 10 years. J Clin Epidemiol 63: 185–191.

18. Walton KA, Murray LJ, Gallagher AM, Cran GW, Savage MJ, et al. (2000) Parental recall of birthweight: a good proxy for recorded birthweight? Eur J Epidemiol 16: 793–796.

19. Olson JE, Shu XO, Ross JA, Pendergrass T, Robison LL (1997) Medical record validation of maternally reported birth characteristics and pregnancy-related events: a report from the Children’s Cancer Group. Am J Epidemiol 145: 58–67.

20. Li R, Scanlon KS, Serdula MK (2005) The validity and reliability of maternal recall of breastfeeding practice. Nutr Rev 63: 103–110.

21. Promislow JH, Gladen BC, Sandler DP (2005) Maternal recall of breastfeeding duration by elderly women. Am J Epidemiol 161: 289–296.

22. Launer LJ, Forman MR, Hundt GL, Sarov B, Chang D, et al. (1992) Maternal recall of infant feeding events is accurate. J Epidemiol Community Health 46: 203–206.

23. UNESCO United Nations Educational SaCOIfS (2006) ISCED 1997. International Standard Classification of Education. Montreal.

24. Chen K, Lo SH (1999) Case-cohort and case-control analysis with Cox’s model. Biometrika 86: 755–764.

25. Breslow NE, Day NE (1980) Statistical methods in cancer research Vol. 1 The analysis of case-control studies. Lyon: International Agency for Research on Cancer. 338 S. p.

26. Dubois L, Girard M (2006) Early determinants of overweight at 4.5 years in a population-based longitudinal study. Int J Obes (Lond) 30: 610–617.

27. Chen A, Pennell ML, Klebanoff MA, Rogan WJ, Longnecker MP (2006) Maternal smoking during pregnancy in relation to child overweight: follow-up to age 8 years. Int J Epidemiol 35: 121–130.

28. Howe LD, Matijasevich A, Tilling K, Brion MJ, Leary SD, et al. (2012) Maternal smoking during pregnancy and offspring trajectories of height and adiposity: comparing maternal and paternal associations. Int J Epidemiol 41: 722–732.

29. Poston L (2012) Gestational weight gain: influences on the long-term health of the child. Curr Opin Clin Nutr Metab Care 15: 252–257.

30. Herring SJ, Rose MZ, Skouteris H, Oken E (2012) Optimizing weight gain in pregnancy to prevent obesity in women and children. Diabetes Obes Metab 14: 195–203.

31. Brophy S, Cooksey R, Gravenor MB, Mistry R, Thomas N, et al. (2009) Risk factors for childhood obesity at age 5: analysis of the millennium cohort study. BMC Public Health 9: 467.

32. Labayen I, Moreno LA, Blay MG, Blay VA, Mesana MI, et al. (2006) Early programming of body composition and fat distribution in adolescents. J Nutr 136: 147–152.

33. Wells JC, Fewtrell MS, Williams JE, Haroun D, Lawson MS, et al. (2006) Body composition in normal weight, overweight and obese children: matched case-control analyses of total and regional tissue masses, and body composition trends in relation to relative weight. Int J Obes (Lond) 30: 1506–1513.

34. Blustein J, Attina T, Liu M, Ryan AM, Cox LM, et al. (2013) Association of caesarean delivery with child adiposity from age 6 weeks to 15 years. Int J Obes (Lond) 37: 900–906.

35. Brion MJ, Lawlor DA, Matijasevich A, Horta B, Anselmi L, et al. (2011) What are the causal effects of breastfeeding on IQ, obesity and blood pressure? Evidence from comparing high-income with middle-income cohorts. Int J Epidemiol 40: 670–680.

36. Singhal A, Cole TJ, Fewtrell M, Lucas A (2004) Breastmilk feeding and lipoprotein profile in adolescents born preterm: follow-up of a prospective randomised study. Lancet 363: 1571–1578.

37. Kramer MS, Matush L, Vanilovich I, Platt RW, Bogdanovich N, et al. (2009) A randomized breast-feeding promotion intervention did not reduce child obesity in Belarus. J Nutr 139: 417S–421S.

38. Neutzling MB, Hallal PR, Araujo CL, Horta BL, Vieira Mde F, et al. (2009) Infant feeding and obesity at 11 years: prospective birth cohort study. Int J Pediatr Obes 4: 143–149.

39. Fall CH, Borja JB, Osmond C, Richter L, Bhargava SK, et al. (2011) Infant-feeding patterns and cardiovascular risk factors in young adulthood: data from five cohorts in low- and middle-income countries. Int J Epidemiol 40: 47–62.

40. Fleten C, Nystad W, Stigum H, Skjaerven R, Lawlor DA, et al. (2012) Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child Cohort Study: A Family-based Approach to Studying the Role of the Intrauterine Environment in Childhood Adiposity. Am J Epidemiol 176: 83–92.

41. Dieu HT, Dibley MJ, Sibbritt D, Hanh TT (2007) Prevalence of overweight and obesity in preschool children and associated socio-demographic factors in Ho Chi Minh City, Vietnam. Int J Pediatr Obes 2: 40–50.

42. Durmus B, Arends LR, Ay L, Hokken-Koelega AC, Raat H, et al. (2012) Parental anthropometrics, early growth and the risk of overweight in pre-school children: the Generation R Study. Pediatr Obes 8: 339–350.

43. Connor Gorber S, Tremblay M, Moher D, Gorber B (2007) A comparison of direct vs. self-report measures for assessing height, weight and body mass index: a systematic review. Obes Rev 8: 307–326.

44. Moore VM, Davies MJ, Willson KJ, Worsley A, Robinson JS (2004) Dietary composition of pregnant women is related to size of the baby at birth. J Nutr 134: 1820–1826.

45. Phelan S, Hart C, Phipps M, Abrams B, Schaffner A, et al. (2011) Maternal behaviors during pregnancy impact offspring obesity risk. Exp Diabetes Res 2011: 985139.

46. Oken E, Levitan EB, Gillman MW (2008) Maternal smoking during pregnancy and child overweight: systematic review and meta-analysis. Int J Obes (Lond) 32: 201–210.

47. Durmus B, Mook-Kanamori DO, Holzhauer S, Hofman A, van der Beek EM, et al. (2010) Growth in foetal life and infancy is associated with abdominal adiposity at the age of 2 years: the generation R study. Clin Endocrinol (Oxf) 72: 633–640.

48. Oken E, Rifas-Shiman SL, Field AE, Frazier AL, Gillman MW (2008) Maternal gestational weight gain and offspring weight in adolescence. Obstet Gynecol 112: 999–1006.

49. Oken E, Taveras EM, Kleinman KP, Rich-Edwards JW, Gillman MW (2007) Gestational weight gain and child adiposity at age 3 years. Am J Obstet Gynecol 196: 322 e321–328.

50. Kim SY, Sharma AJ, Callaghan WM (2012) Gestational diabetes and childhood obesity: what is the link? Curr Opin Obstet Gynecol 24: 376–381.

51. Barker DJ (1995) The fetal and infant origins of disease. Eur J Clin Invest 25: 457–463.

52. Singhal A, Lucas A (2004) Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 363: 1642–1645.

53. Verier C, Meirhaeghe A, Bokor S, Breidenassel C, Manios Y, et al. (2010) Breast-feeding modulates the influence of the peroxisome proliferator-activated receptor-gamma (PPARG2) Pro12Ala polymorphism on adiposity in adolescents: The Healthy Lifestyle in Europe by Nutrition in Adolescence (HELENA) cross-sectional study. Diabetes Care 33: 190–196.

54. Poskitt EM, Breda J (2012) Complementary feeding and non communicable diseases: current knowledge and future research needs. Nutr Metab Cardiovasc Dis 22: 819–822.

55. Monasta L, Batty GD, Cattaneo A, Lutje V, Ronfani L, et al. (2010) Early-life determinants of overweight and obesity: a review of systematic reviews. Obes Rev 11: 695–708.

56. Thompson AL (2012) Developmental origins of obesity: early feeding environments, infant growth, and the intestinal microbiome. Am J Hum Biol 24: 350–360.

57. Gibbs BG, Forste R (2013) Socioeconomic status, infant feeding practices and early childhood obesity. Pediatr Obes Apr 2. doi: 10.1111/j.2047–6310.2013.00155.x. [Epub ahead of print].

58. Huh SY, Rifas-Shiman SL, Taveras EM, Oken E, Gillman MW (2011) Timing of solid food introduction and risk of obesity in preschool-aged children. Pediatrics 127: e544–551.

59. Wasser H, Bentley M, Borja J, Davis Goldman B, Thompson A, et al. (2011) Infants perceived as “fussy” are more likely to receive complementary foods before 4 months. Pediatrics 127: 229–237.