Risk Factors Of Childhood Obesity: Lessons From The European IDEFICS Study

ALSO AVAILABLE:

Author(s):

	Wolfgang Ahrens
	Department of Epidemiological Methods and Etiological Research, Leibniz Institute for Prevention Research and Epidemiology – BIPS, Bremen, Germany
	View Author’s Full Biography

	Iris Pigeot
	Department of Biometry and Data Management, Leibniz Institute for Prevention Research and Epidemiology – BIPS, Bremen, Germany  Faculty of Mathematics and Computer Science, University of Bremen, Bremen, Germany
	View Author’s Full Biography

Summary 

The IDEFICS (Identification and Prevention of Dietary- and Lifestyle-Induced Health Effects in Children and Infants) study and its follow-up, the I.Family study, are a large pan-European cohort study investigating childhood obesity, dietary habits, lifestyle factors, and their long-term health effects. 

The IDEFICS study (2006–2012) examined 16,229 children aged 2–9.9 years across eight European countries (Belgium, Cyprus, Estonia, Germany, Hungary, Italy, Spain, and Sweden). It aimed to identify risk factors for childhood obesity and develop intervention strategies. We collected extensive data on diet, physical activity, sleep, socio-environmental factors, anthropometry and biomarkers related to metabolism and inflammation at baseline and repeated these measurements two years later. Building on IDEFICS, the I.Family study (2013–2017) followed up with the original cohort, where an additional survey centre was established in Poland. I.Family investigated 9,617 children now aged 7–17 years, to assess determinants of diet, long-term health impacts of the risk factors measured at baseline and genetic-environmental interactions. A fourth survey wave of the cohort with 5,319 participating children, adolescents and young adults (age range 11 to 26 years) was conducted online with self-completion questionnaires (2020-2021) with a particular focus on digital media exposure and mental health outcomes. Key findings include: 

• Obesity risk factors: Poor diet, insufficient physical activity, short sleep duration, and high screen time were linked to higher obesity risk. 

• Metabolic syndrome: Based on extensive anthropometric measurements and the measured biomarkers such as blood lipids, glucose and inflammatory markers paediatric age- and sex-specific reference values of clinical parameters were newly derived and a new metabolic syndrome (MetS) score was proposed. 

• Parental influence: Children’s dietary habits and body mass index (BMI) were strongly influenced by parental education and socioeconomic status. 

• Sugary drinks and obesity: High consumption of sugar-sweetened beverages was associated with greater weight gain. 

• Healthy diet: Children whose dietary pattern was characterised by high consumption of vegetables and wholemeal had a lower risk of becoming overweight/obese compared to children with the lowest consumption. 

• Sensory taste perception: Children with a taste preference for added fat or for added sugar had significantly higher odds for being overweight/obese. 

• Sleep and weight: Shorter sleep duration correlated with higher BMI and increased obesity risk. 

• Sensitive growth period: A rapid BMI growth from 9 months to <6 years seems to increase later metabolic risk particularly strongly.  

• Digital media exposure: The longitudinal analysis revealed that TV exposure and PC use had a negative impact on children´s well-being, as indicated by increased peer- and emotional problems. Increasing media use trajectories were positively associated with z-scores of MetS. Duration of digital media exposure was positively associated with sweet, fatty, and salty food preference as well as with higher intake of snack foods and energy. Exposure to smartphones and media multitasking were positively associated with impulsivity and cognitive inflexibility and inversely associated with decision-making ability. 

• Physical activity: Green spaces and sports grounds as well as membership in a sports club are associated with higher levels of physical activity (PA) which in turn is linked to a lower risk of excess weight gain and higher bone stiffness (weight-bearing PA). While overweight children tended to reduce their physical activity levels in subsequent years, more active children were less prone to gain excess body weight in the following. 

• Dietary tracking: Children’s dietary habits tend to persist over time, highlighting the importance of early nutrition interventions. 

• Genetics vs. environment: While genetics play a role in obesity, lifestyle factors (diet, exercise, screen time) significantly influence health outcomes. 

• Family and peer influence: While the impact of peer behaviours on food choices and physical activity levels increases during adolescence, the impact of parental behaviours declines. 

• Mental health link: Higher BMI was associated with increased risks of anxiety and depressive symptoms. Emotion-driven impulsiveness and cognitive inflexibility were positively associated with an elevated BMI. 

This research emphasises early intervention in childhood to promote healthy eating, physical activity, and adequate sleep. It highlights the complex interaction between genetics, environment, and lifestyle in shaping long-term health and pave the way for novel participatory and systemic intervention approaches. The findings inform European public health policies on childhood obesity prevention and lifestyle interventions. 

Introduction 

The prevalence of childhood obesity and overweight has increased in most regions of the world1, 2, 3, 4. Several studies indicate that this trend has levelled off in some developed countries like the US, Australia, and some European countries5, 6, 7, 8, 9. This is confirmed by a pooled analysis of 2,416 population-based studies with measured height and weight data10. Based on the body mass index (BMI) of 31.5 million children and adolescents aged 5–19 years it was concluded that the rising trends in the prevalence of childhood overweight and obesity are levelling off in many high-income countries, albeit at a high plateau, while the prevalence is still increasing in other areas, in particular, in parts of Asia (for further details see also a pooled analysis on height and body mass index trajectories of school-aged children and adolescents from 1985 to 201911). As the causal pathway leading to obesity already starts early in life it is important to understand the causes and mechanisms leading to this disorder and to find a way for effective primary prevention interventions in young children. 

The IDEFICS study (Identification and prevention of dietary- and lifestyle-induced health effects in children and infants) (https://www.ideficsstudy.eu/) investigated the aetiology of diet- and lifestyle-related diseases and disorders with a strong focus on overweight and obesity in a large population-based cohort of 16,229 European children aged 2 to 9.9 years who were recruited from eight European countries. According to a standardised protocol, weight status and related health outcomes such as blood pressure and insulin resistance, direct behavioural determinants such as physical activity and diet and indirect determinants such as social/ psychological factors and consumer behaviour as well as environmental drivers of these behaviours like food marketing and the built environment were assessed. In this way, the study tried to disentangle the causal pathways leading to obesity and other health outcomes by analysing the complex interplay of potential risk factors. Details of the objectives, original study design, the proposed measurements and a description of the study sample have been published previously12, 13, 14. Furthermore, the IDEFICS study developed, implemented and evaluated a setting-based community-oriented intervention programme for primary prevention of obesity in a controlled study design15. For this purpose, in each country intervention and control regions were selected with a comparable socio-demographic profile. In the intervention regions, a coherent set of intervention modules were implemented, focusing on diet, physical activity and stress-coping capacity captured in six key messages.  

Building on the IDEFICS cohort, we conducted the I.Family study (https://www.ifamilystudy.eu/) (cohort profile16) immediately afterwards to unravel the factors at play in children’s metabolic health and their complex interplay in even more detail as, e.g., the determinants of dietary behaviour and food choice. Its ultimate aim was to identify targets for effective interventions and to support policy development, enabling more families to make healthier choices. At the third examination wave, i.e. in the framework of I.Family, most IDEFICS children had entered puberty. Even if children had adopted healthy eating and activity patterns, their lives changed considerably as they became teenagers. Healthy routines may get lost and be replaced by unhealthy habits, e.g., because of the influence of marketing or peer pressure. This transition phase had to be carefully studied to get a deeper insight into the effects of changes in lifestyle that may maintain health or may cause ill-health from a life course perspective. We therefore complemented the data on health and nutrition of our IDEFICS children with data on parenting style, family life and peer networks. This enabled us to investigate the dynamic transition from childhood to adolescence, when children´s behaviours are getting influenced by social and environmental factors outside the family environment. 

In this chapter we will present the design of the IDEFICS/I.Family cohort, which provides a unique longitudinal database of children and their families, the evolvement of this cohort over time, and some major results, where we will focus on potential risk factors of childhood obesity.  

Design, subjects and methods 

Study subjects 

A cohort of 16,229 children aged 2 to 9.9 years, referred to as index children in the following, was examined in a population-based baseline survey in eight European countries ranging from North to South and from East to West (Sweden, Germany, Hungary, Italy, Cyprus, Spain, Belgium, Estonia) from autumn 2007 to spring 2008. This baseline survey (T0) was the starting point of the prospective cohort study with the largest European children’s cohort established to date14. At follow-up (T1), two years later, 11,041 (68% of all index children) participated in exactly the same survey modules as at baseline (T0). The mean age (standard deviation) of the children was 6.0 (1.8) years at baseline and 7.9 (1.9) years at T1 where the proportion of boys and girls was nearly balanced in both examination waves. A closer look at the drop-outs revealed that these were more likely to be overweight, to report low well-being scores and to come from low-educated or single parent families17. To offer participation to classmates and to compensate for the attrition, we newly recruited 2,555 children at T1. In between the two examination waves, a primary prevention programme was offered all children living in communities selected as intervention regions in each country. A second community with a similar socio-demographic profile was selected as control region where children were not exposed to the intervention activities. Penetration of the intervention messages was assessed by a mail survey (T2). 

The study was not designed to provide a representative sample for each country. All children in the defined age group who resided in the defined regions and who attended the selected primary schools (grades 1 and 2), pre-schools or kindergartens were eligible for participation. Children were approached via schools and kindergartens to facilitate equal enrolment of all social groups. In addition to the signed informed consent given by parents, each child was asked to give verbal assent immediately before examination. Participants were free to opt out of specific examination modules like blood drawing. Thus, the results presented below are based on different subgroups and varying sample sizes which are described in more detail in the respective original articles. 

In the framework of the I.Family study, Poland joined as ninth survey centre. The third examination wave (T3) took place in 2013/2014, where 6,055 (55%) of the index children and 1,050 (41%) of the children newly recruited at T1 participated. We also invited all siblings who were in the same age range as the index children and newly recruited 2,512 children. The mean age (standard deviation) at T3 was 10.9 (2.9) years. To investigate the influence of familial characteristics, family structure and family life the children’s development we invited at least one parent of each index child to participate. In total, 6167 families with on average 2 children and 4.1 members (including parents) per family participated at T3. 

For a detailed description of all instruments deployed at T0, T1, T3, T4, and T5, we refer to Table A (appendix), the cohort profile paper16 (supplementary Table 1) and a book on instruments for health surveys in children and adolescents18. Some of the examinations and measurements were restricted to subgroups due to time and financial constraints. Table 216 of the cohort profile paper gives a summary of how many subjects provided the respective data. Please note that IDEFICS instruments designed for small children and their proxies were adapted for use in adolescents and adults in I.Family to yield comparable data for longitudinal analyses of repeated measurements. After T3, we invited children from so-called contrasting groups, i.e. children with diverging weight trajectories, for an even more intense examination including for instance functional magnetic resonance imaging (fMRI) of their brains (T4) to assess neural response to food stimuli. 

A fourth survey wave of the cohort was conducted in 2020/2021, i.e. during the corona pandemic, which only allowed a web-based survey with in total 5,319 participating children, adolescents and young adults in the age range from 11 to 26 years and a mean age (standard deviation) of 18.2 (2.4). Figure 1 illustrates the content and the temporal patterns of the IDEFICS/I.Family cohort.  

Figure 1: Overview of examination waves/surveys and overall design of IDEFICS/I.Family cohort 

Questionnaires 

At T0 and T1, parents completed a self-completion questionnaire to assess gestational, behavioural and socio-demographic factors and a children’s eating habits questionnaire (CEHQ) to record food frequency and dietary habits. The latter was complemented by a computer-based 24-hour dietary recall (24-hdr). Parents were offered assistance in filling in the questionnaires. In addition, a face-to-face medical interview was conducted with one parent.  

Educational attainment was classified according to the International Standard Classification of Education (ISCED)19. Further socio-demographic characteristics such as family income (classified by country-specific categories based on the average net equivalence income), employment status, dependence on social welfare and migration background of parents were recorded.  

At T3, parents were asked to complete a kinship questionnaire. Children in the age of 8 years or older were asked to self-assess their maturation stage according to Tanner using pictograms. Medical history was obtained by interview from parents for both their children and for themselves. At T3, the questionnaire on dietary habits and food consumption frequency was combined with the general questionnaire. Teens completed a tailored version of it on a tablet PC and at least one parent completed it also for him/herself in 90% of the families. Since I.Family had a clear focus on dietary habits, the participants were asked to fill a whole series of 24-hdrs: A web-based version of the 24-hdr was offered to all participants with at least 8 years of age. It was recommended to complete the first 24-hdr at the examination centre and two further 24-hdrs on non-consecutive days including one weekend day during the next two weeks. Parents assisted smaller children (<8 years) in completing their 24-hdr. Six months after the T3 examination participants were again asked to fill three further 24-hdrs.  

Examinations 

The examination programme included standard anthropometric measures20, clinical parameters such as blood pressure, collection of urine, saliva and blood for further medical parameters and genetic analyses, and accelerometry to assess physical activity. Additional examinations were only applied in subsamples, either because they were not feasible in small children (e.g., physical fitness tests, sensory taste perception) or because they were expensive (e.g. ,ultrasonography of the calcaneus to assess bone stiffness, analysis of blood fatty acids). Preferably, all examinations of a child took place on the same day but this was not always feasible. As one innovative component of the examination programme sensory taste perception was examined where in total five tastes were included, namely sugar and apple flavour in apple juice (the latter was not tested in Cyprus) as well as monosodiumglutamate, salt, and fat in crackers. Paired comparison tests were used to assess the preference for each taste. That means each child had to choose his/her preferred food sample out of a pair which consisted of a reference sample and a modified version. Each child tasted reference before the modified version and then he/she put the preferred sample on a “smiley” on top of a game board. No preference was not an option. For example, sweet taste was assessed by clear apple juice served in small cups of 30ml at 18 ± 2°C with the reference containing 0.53% added sucrose whereas the high-sugar sample contained 3.11% added sucrose. Sweet was always tested before fat. For fat tasting, crackers were prepared with the reference cracker consisting of water, flour, fat (8%) and salt. The modified cracker contained 18% fat. High-sweet (high-fat) preference was recorded when the child chose the sweetened juice (added fat cracker) over the basic food sample. All food samples were produced centrally and shipped to the survey centres. For further details on these and other examination modules see21, 22, 23. 

At T3, we also conducted neuropsychological tests on decision making, set shifting capacity and inhibitory capacity. 

Blood collection: We aimed to obtain fasting blood from all children via either venipuncture or capillary sampling. It was anticipated that a sizeable number of children would refuse the venipuncture even with local anaesthesia with EMLA patches provided. To ensure that basic data on metabolic disturbances was available for as many children as possible a point-of-care analyser was used to assess blood glucose, HDL and LDL cholesterol and triglycerides in one drop of capillary blood from the finger tip on the spot. All blood, serum, urine and saliva samples were transferred to a central bio-repository to coordinate the laboratory analyses and to ensure standardised storage and handling of samples24. 

Physical activity: To monitor physical activity children wore a uni-axial accelerometer (ActiGraph® or ActiTrainer®) on a hip belt over three consecutive days including one weekend day. In school children the accelerometer was combined with a Polar® heart rate monitor using a chest belt. Resting heart rate was assessed in conjunction with the physical fitness tests. Accelerometry was complemented by an activity diary that was completed by parents over the measurement period.  

Physical fitness: Components of the physical fitness tests were adopted from the European battery of cardiorespiratory and motor tests (Eurofit battery) (flamingo balance test, backsaver sit and reach, handgrip strength, standing broad jump, 50 m sprint, shuttle-run test)25 that were restricted to school children. 

Bone stiffness: Heel ultrasonometry which had shown good correlations with bone mineral density assessed by dual-energy-x-ray absorptiometry (DEXA) in adults26 and children27, as well as a high prognostic value of bone fractures in adults28 was included as an optional component to assess bone stiffness of the calcaneus of the left and right foot. 

Brain activation: In the framework of the I.Family study, functional magnetic resonance imaging of the brain (fMRI) was performed in three countries to assess brain activation by visual food cues in a smaller subgroup of normal- and overweight children and their parents. 

Quality management 

All measurements followed detailed standard operating procedures (SOPs) that were laid down in the general survey manual and finalised after the pre-test of all survey modules22. Field personnel from each study centre participated in central training and organised local training sessions thereafter. The coordinating centre conducted site visits to each study location during field surveys to check adherence of field staff to the SOPs. Questionnaires were developed in English, translated to local languages, and then back-translated to check for translation errors. All study centres used the same technical equipment that was purchased centrally to maximise comparability of data.  

Databases and computer-assisted questionnaires included automated plausibility checks. All numerical variables were entered twice independently. Inconsistencies identified by additional plausibility checks were rectified by the study centres. In addition, a panel of statisticians supported state-of-the-art data analyses. 

To further check for the quality of data, sub-samples of study subjects were examined repeatedly to calculate the inter- and intra-observer reliability of anthropometric measurements23. In addition, the reliability of questionnaires was checked by re-administering the CEHQ and selected questions of the parental questionnaire to a convenience sample of study participants29,30. Food consumption assessed by the CEHQ was validated against selected nutrients measured in blood and urine31. The new method to analyse the fatty acid profile in a dried drop of blood was compared to the standard analysis of serum and erythrocytes from venous blood. A validation study was carried out to compare uni-axial and tri-axial accelerometers in children and to validate them using doubly labelled water as the gold standard, and to also validate body composition measures using 3- and 4-compartment models32. Ultrasonometry was compared to DXA to assess bone mineral density in a sample of children from Sweden and Belgium33.  

Major findings 

This section will describe some major findings related to (1) the intervention programme, (2) the first cohort phase (IDEFICS study), and (3) the second cohort phase (I.Family study). We will highlight some prominent results without any claim to completeness. 

The IDEFICS intervention programme 

The intervention mapping protocol34 was applied to develop the components of the IDEFICS intervention35. Based on the major suspected risk factors for the development of obesity, i.e. physical activity, dietary and stress-related behaviours, the IDEFICS intervention focussed on three main intervention areas formulated as six key messages35: (1) increase daily physical activity levels, (2) decrease daily television (TV) viewing time, (3) increase the consumption of fruit and vegetables, (4) increase the consumption of water, (5) strengthen parent-child relationships and (6) establish adequate sleep duration patterns (see Figure 2). 

Figure 2: The six key messages of the IDEFICS intervention, illustrations were used in the corresponding leaflets for parents and children36 

We searched the literature for established national or international recommendations regarding the health-related behaviours listed above with respect to the prevention of childhood obesity (for details see37) and determined the percentage of children who “spontaneously” complied with these recommendations at baseline. Figure 3 shows the overall percentage of children adhering to these recommendations. A more detailed figure is given in37. 

Figure 3: Percentage of children adhering to recommendations based on the six key messages of the IDEFICS intervention37 

The results of the IDEFICS primary prevention programme are presented in detail in a supplement of Obesity Reviews38. With respect to indicators of body fatness, we used mixed models that were adjusted for age and socioeconomic status allowing for a random effect for country to account for the clustered study design to compare the prevalence of overweight/obesity and mean values of body mass index z-score, per cent body fat and waist-to-height ratio between baseline examination and first follow-up examinations two years later39. Overall, no statistically significant differences between the intervention and control groups were observed. However, we detected changes in favour of the intervention in some indicators but these were counterbalanced by changes in favour of the control group in other indicators. In total, the prevalence of overweight and obesity increased from 18.0% at baseline to 22.9% at follow-up in the control group and from 19.0% to 23.6% in the intervention group. 

Based on comparable models, we also investigated potential effects on the behaviours of the IDEFICS children targeted by the intervention. No overall significant time by condition interaction effects, neither for boys nor girls, could be shown, but a few favourable intervention effects were observed on specific behaviours in some countries40.  

Interestingly, our analyses provided some evidence of a beneficial effect of the IDEFICS intervention on overweight children. Although these children were not specifically targeted, we observed a significantly greater probability of a normalised weight status after two years among those children with prevalent overweight or obesity at baseline41. 

Furthermore, we were in the lucky situation to investigate behavioural effects of the IDEFICS intervention in the framework of the I.Family study, i.e., six years after the intervention phase. In particular, we observed that families who participated in the IDEFICS intervention had a significantly lower propensity of sugar consumption and a higher propensity of water consumption compared to control families42. 

Key results of the first cohort phase (IDEFICS study) 

Metabolic health: The combined prevalence of overweight/obesity in 2- to 9.9-year olds ranged from more than 40% in southern Europe (highest percentage in Italy) to less than 10% in northern Europe (lowest percentage in Belgium). The prevalence was higher in girls (21.1%) than in boys (18.6%), in particular in pre-school children. The prevalence increased with increasing age and was negatively associated with education and income43.  

Fractional polynomial multilevel models were used to derive individual body mass index (BMI) trajectories and to estimate – among others – age at infancy BMI peak (IP) and age at adiposity rebound (AR)44. It turned out that age and BMI at IP and AR could not be assessed in 5.4% and 7.8% of the children, respectively. These children showed a significantly higher BMI growth during infancy and childhood.  BMI values at ages 1 and 5 years could be used instead as potentially better predictors for later weight status. 

Based on the extensive anthropometric measurements and the measured biomarkers such as blood lipids, glucose and inflammatory markers as well as blood pressure, we were able to calculate for the first time age- and sex-specific reference values of clinical parameters in a large sample of healthy children45. These were published in a supplement volume of the International Journal of Obesity. Of particular importance in this context were the newly derived definition of the metabolic syndrome46 and the proposed novel metabolic syndrome (MetS) score for children which have received major attention over the years (see among others47) and has been widely used worldwide in recent years.  

We also identified various determinants of the metabolic syndrome in children, where, e.g., children with a low socioeconomic background carried an elevated risk to develop a metabolic syndrome independently of diet, physical activity, sedentary behaviours and well-being48. In addition, the transition a metabolically unhealthy state was associated with higher levels of hsCRP at follow-up, independent of their weight status at baseline49.  

Taking advantage of the longitudinal database, we used linear-spline mixed-effects models to investigate the association between BMI trajectories during childhood and later metabolic health50. We observed that all exposures, i.e., BMI at birth, rates of BMI change during infancy (0 to <9 months), early childhood (9 months to <6 years) and later childhood (≥6 years) as well as current BMI z-score were strongly associated with the MetS score at follow-up. It turned out that the most sensitive period of growth affecting health is a rapid BMI growth starting from birth, especially in the time window of 9 months to <6 years, leading to an increased later metabolic risk.  

Diet: Numerous sources of measurement error have been encountered when operating with dietary data. In young children, dietary data are commonly assessed through parental proxies which may result in additional measurement error. Meals or snacks not observed by the proxies may lead to underreporting of certain foods and total energy intake whereas difficulties in estimation of portion sizes/consumption frequencies as well as social desirability may lead to both, over- or underreporting51. In IDEFICS, the prevalence of underreporting and overreporting estimated based on age- and sex-specific Goldberg cut-offs were 8.0% and 3.4% (24-hdr data), respectively52. The prevalence of underreporting increased with BMI z-score, household size and was higher in low income groups. Especially social desirability and the parental perception of their child’s weight status seemed to affect the reporting accuracy. When analysing diet-obesity associations, the associations were strongly affected or even masked by measurement errors where Börnhorst et al.53 found that consideration of the reporting status and inclusion of a propensity score for misreporting were useful tools to counteract attenuation of effect estimates.  

Due to these problems to correctly assess children´s dietary behaviour it is difficult to reveal individual associations between diet and overweight/obesity: Using a principal component analysis, Pala et al.54 identified four major dietary patterns: snacking, sweet/fat, vegetables/wholemeal, and protein/water in children´s dietary behaviour assessed by food frequency questionnaires. In a multilevel mixed regression analysis of the longitudinal data with change in BMI category55 from thin/normal weight at T0 to overweight/obese at T1 as outcome, adjusted for baseline BMI, age, sex, physical activity and family income, they observed a lower risk of becoming overweight/obese for children in the highest tertile of the vegetables/wholemeal pattern compared to children in the lowest tertile. 

In addition, based on a cluster analysis that distinguished three dietary patterns, we observed that children with a higher socioeconomic background were more likely to be allocated to the healthy cluster at baseline and follow-up whereas children with a lower socioeconomic background showed unhealthier dietary profiles over a two-year period56. 

Further effects of dietary behaviour can be observed taking TV viewing additionally into account as potential risk factor. 

TV viewing: Lissner et al.57 investigated the association between daily TV time and the presence of a TV/video/DVD in the child´s bedroom and overweight/obesity by estimating odds ratios adjusted for sex, age and parental education. Both, having a TV in the child´s bedroom and consumption daily TV time of more than 60 minutes showed a positive association with the weight status of children in all countries57. It could also be shown that, independent of taste preferences, children who watched more TV had a higher propensity to consume foods high in fat and/or sugar57. 

Moreover, associations between screen habits and sweetened beverage consumption were observed which could also be seen longitudinally: children who were exposed to commercial TV at baseline (T0) had a higher risk of consuming sweetened beverages at T158. A further longitudinal analysis revealed a substantial impact of TV viewing and other screen habits on the consumption of sugary drinks and on increase in BMI59. 

Longitudinally, we could also show that TV exposure and PC use had a negative impact on children´s well-being, as indicated by increased peer- and emotional problems particularly in girls and impaired family life in both sexes60. 

Physical activity: A ‘moveability index’ was developed as a tool for urban planners to reflect opportunities for physical activity in the urban environment of children. Based on geographical data, the index integrated different urban measures such as the availability of destinations, i.e. playgrounds, green spaces and sport facilities, as well as the street connectivity considering intersections, foot paths and cycle lanes that were both assessed using a so-called kernel density approach (Figure 4). Additionally, residential density and land use mix were included in the index. In a pilot study that was conducted in the intervention region in Germany, it was shown that opportunities for physical activity in the urban neighbourhood of school children, i.e. short routes and particularly the availability of destinations, were positively associated with physical activity levels61. 

Figure 4: Availability of playgrounds within the German intervention community, Delmenhorst, estimated via kernel density  

The analysis of physical activity concentrated on its effect on bone stiffness and weight status. The duration of moderate-to-vigorous physical activity (MVPA) showed huge variations across Europe62 and had a protective effect against overweight/obesity, in particular in school-age children. The prevalence of obesity was elevated in children exercising less than the recommended 60 minutes moderate-to-vigorous physical activity per day63 (Figure 5).  

Figure 5: Duration of MVPA (60 sec. interval, Evenson64) by age and weight 

We also investigated the impact on physical activity on children´s bone stiffness. This analysis revealed that an additional 10 minutes per day of MVPA or sports activities with weight-bearing exercises were related to an increased bone stiffness where muscle strength and sedentary behaviour were taken into account65.  

Family life: Based on the parental questionnaires, a “health-related quality of life score (QoL)” adapted from the Questionnaire for Measuring Health-Related Quality of Life in Children and Adolescents KINDL66, a “difficulties score” and a “pro-social behaviour score”, where the latter two were adapted from the Strengths and Difficulties Questionnaire67 and two family lifestyle measures were constructed. The health-related QoL score showed a substantial variation among all countries. Based on a generalised mixed model with country as random effect, adjusted for sex and age group a negative association between the QoL and overweight/obesity was observed regardless of the socioeconomic status of the families. In additional analyses, we especially considered the question whether parents and their children share family meals at least one per day (Do you sit down with your child when he/she eats meals?) as a proxy for family life during meals. Here, we observed a clear gradient of an increasing prevalence of overweight/obese children ranging from a prevalence of 17.1% among those sitting always together to a prevalence of 36.2% among those who reported to never/rarely sit together during meals. 

Sleep: Sleeping behaviour was investigated with respect to the factors influencing sleep duration, the association of sleep duration and obesity, and the physiological changes involved in this association. Sleep duration showed marked variation across Europe, but exhibited an ecological correlation with the prevalence of overweight/obesity68. This correlation was confirmed by individual level analysis as sleep duration was negatively associated with weight status, particularly in school-age children (see Table 1)69. 

*adjusted for age (continuous), ambient temperature (continuous), European region (north versus south) 

Table 1: Odds ratios (OR) and 95% confidence intervals (CI) for the association between sleep duration and overweight/obesity (reference > 11 hours)69 

Multivariate linear regression and quantile regression models confirmed an inverse relationship between sleep duration and measures of overweight/obesity. The estimate for the association of sleep duration and body mass index (BMI) was approximately halved after adjustment for fat mass (FM), but remained statistically significant. The strength of this association was also markedly attenuated when adjusting for insulin mainly for the upper BMI quantiles. This means that the inverse relationship between sleep duration and BMI is mainly explained by the association between sleep duration and FM. Insulin may explain part of this association, in particular at the upper tail of the BMI distribution70.  

Bone stiffness: Sioen et al.71 investigated the association between various markers of body fat and the bone status that was assessed as calcaneal bone stiffness by ultrasonography. Partial correlation analyses, linear regression analyses, and ANCOVA stratified by sex and age groups showed that pre-school children with higher BMI had a lower calcaneal stiffness index (SI), while primary school children with higher BMI had a higher calcaneal stiffness index. After adjusting for fat-free mass, both pre-school and primary school children showed an inverse association between BMI and calcaneal stiffness. Thus, fat-free mass seems to be a confounder in the association between SI and weight status in primary school children but not in pre-school children. We concluded that muscle mass is an important determinant of bone stiffness. 

Genetic susceptibility (for abbreviations please see Table 2): The analysis of the FTO gene (Ref.-SNP 9939609) showed that the odds ratio for overweight/obesity was elevated by 40% among children carrying the AA-allele as compared to the TT-allele. Similar positive associations were found for waist circumference, waist-to-height ratio and the sum of skinfold thicknesses. These associations were confirmed in the longitudinal analysis even after adjustment for age, sex, country, intervention group and BMI at T072. Based on a structural equation model (SEM) with the latent constructs obesity, dietary intakes, physical activity and fitness habits, and parental socioeconomic status (SES), we investigated a potential interaction of the FTO gene with SES. The SEM revealed that children carrying the protective FTO genotype TT seemed to be more protected by a favourable social environment regarding the development of obesity than children carrying the AT or AA genotype73. 

It was also investigated74 whether NMU single nucleotide polymorphisms (SNPs) and haplotypes, as well as functional ADRB2 SNPs, are associated with bone stiffness in children. An additional question was whether NMU and ADRB2 interact with each other. The reasoning behind this question is that energy metabolism and bone mass are both regulated by the neuromedin U, encoded by the NMU gene, which is a hypothalamic neuropeptide, while effects of catecholamine hormones and neurotransmitters in bone are mediated by the beta-2 adrenergic receptor, encoded by the ADRB2 gene. After adjusting for multiple testing, the stiffness index (SI) was significantly associated with all NMU SNPs. A non-significant decrease in SI was observed in in ADRB2 rs1042713 GG homozygotes, while children carrying SI-lowering genotypes at both SNPs (frequency = 8.4%) showed much lower SI than non-carriers. Thus, it was for the first time shown that the NMU gene impacts on the regulation of bone strength through an interaction with the ADRB2 gene. 

Table 2: List of genetic markers and corresponding abbreviations 

In a subsample of children, we performed micro-RNA analyses in peripheral blood cells (PBCs) to assess the effect of diet on gene expression. These showed that children with a low frequency consumption of sugary foods displayed higher TAS1R3 expression levels compared to those with intermediate or high frequency. In turn, children with high frequency consumption of fatty foods showed lower UCN2 expression levels compared to those with low or intermediate frequency. Thus, transcripts of TAS1R3 and UCN2 in PBCs may serve as potential biomarkers of consumption of sugary and fatty food75.  

Biomarkers: The analysis of transcriptional biomarkers in peripheral blood showed the following: high expression levels of CPT1A, SLC27A2, INSR, FASN, or PPARα were indicative of a lower risk for the insulin-resistant or dyslipidemic state associated with obesity, whereas low LEPR mRNA levels appeared as a marker of high low-density lipoprotein cholesterol, independently of body mass index76. 

Sensory taste perception: A unique feature of the IDEFICS study was the assessment of taste thresholds and taste preferences in order to reveal possible associations with overweight/obesity in a population-based approach with a large number of subjects. The cross-sectional analysis77 of the baseline survey showed that both fat and sweet taste preference were independently associated with weight status. Children with a taste preference for added fat and those with a taste preference for added sugar had significantly higher odds for being overweight/obese after adjusting for possible confounders. The positive associations with overweight/obesity were seen in all age groups and both sexes, but most pronounced in girls. 

Key results of the second cohort phase (I.Family study) 

Metabolic health: A range of determinants of the metabolic syndrome in children and adolescents was identified. These determinants include genetic78, socioeconomic48, and lifestyle factors, e.g., exposure to digital media79, as well as a chronic inflammatory state80. Estimated probabilities for transitioning into high-risk metabolic phenotypes early in life revealed that children who showed several metabolic syndrome components already in early childhood have almost no chance of reaching a healthy metabolic status during later childhood or adolescence. We showed latent transition analysis to be a powerful tool for estimating changes in metabolic status from childhood to adolescence81. While lipid disturbances can be quickly reversed in childhood, abdominal obesity is a likely precursor of subsequent metabolic disturbances. In this context, we studied the role of factors influencing the development of metabolic disturbances from childhood to adolescence. While media devices in the bedroom seemed to have a detrimental influence, membership in a sports club, and positive well-being seemed to strengthen metabolic health82. 

In addition, we tried to account for the complexity of childhood obesity by using a causal discovery algorithm to estimate a cohort causal graph (CCG) over the life course from childhood to adolescence83. Considering only modifiable risk factors, this analysis revealed some possible risk factors acting only on indirect causal paths leading to obesity (measured by age- and sex-adjusted BMI z-scores) six years later. 

Cognitive functioning: Mixed-effect regression analyses were conducted for markers of cognitive functioning (emotion-driven impulsiveness, cognitive inflexibility, decision-making ability as predictor with standardised BMI (zBMI) as dependent variable to investigate their association in adolescents aged 12-18 years at T384. We showed that emotion-driven impulsiveness and cognitive inflexibility were positively associated with an elevated zBMI, while no statistically significant association with zBMI for decision-making ability was observed. 

To further investigate the effect of emotion-driven impulsiveness, we considered its association with snack food consumption in adolescents aged 12-18 years at T3. Mixed-effect regression analyses were conducted for each snacking behaviour with emotion-driven impulsiveness as exposure85. After controlling for zBMI, age, sex, country and socioeconomic status, emotion-driven impulsiveness was positively associated with daily consumption frequency of snacks and consumption frequency of energy-dense snacks. Importantly, adolescents with stronger emotion-driven impulsiveness reported a higher snacking frequency (specifically more energy-dense snacks).  

Digital media: We investigated the increasing digital media exposure of children and adolescents as well as the consequences of increased media time on overweight, obesity, and the metabolic syndrome. We observed that digital media exposure increased with increasing age, most strongly in Estonian children and least strongly in Spanish children. In particular, increasing media use trajectories were positively associated with z-scores of MetS and its components after adjustment for puberty, diet and other confounders86. In addition, duration of digital media exposure was positively associated with sweet, fatty, and salty food preference as well as with higher intake of snack foods and energy87. This may be explained by media-related stress and response to negative emotions88, which was corroborated by a hypothetical intervention analysis on improving well-being and reducing impulsivity89. A detailed analysis of the association of cognitive functioning and digital media use revealed that exposure to smartphones and media multitasking were positively associated with impulsivity and cognitive inflexibility and inversely associated with decision-making ability. Extensive smartphone/internet exposure combined with low computer/medium TV exposure was associated with higher impulsivity and cognitive inflexibility scores, especially in girls.  

Physical activity: Based on objectively measured physical activity data of three subsequent examination waves we revealed a bi-directional relationship between moderate-to-vigorous physical activity (MVPA) and overweight or obesity by means of multilevel regression models90. This means that overweight children tended to reduce their physical activity levels in subsequent years while more active children were less prone to gain excess body weight.  

Family life and peer network: Based on IDEFICS/I.Family cohort, we showed that high physical activity and energy intake seem to explain differences in the body mass index between siblings and twin pairs living in the same home91. We also observed that while sibling resemblance in fast food consumption and screen time decreased with age, while peer resemblance increased92. For instance, for fast food consumption, the peer resemblance was more than 6-fold higher than the sibling resemblance and the peer resemblance surpassed the sibling resemblance by the age of 9-10 years. In particular, longitudinal results showed that children’s changes in fast food consumption were more strongly associated with that of their peer group than that of their siblings, in particular when the age gap between siblings was large. In line with this, behaviours changed around the age of 10 years when, e.g., the consumption of snacks, fast foods, and sweetened drinks increased. Unfavourable media use and unhealthy eating patterns were most frequent in children from families with a low or medium educational level and in non-traditional families. Known risk factors for overweight or obesity seemed to exert their effects almost independently of parental education93,94. 

Well-being and sleep: We found that higher levels of well-being were associated with lower levels of insulin resistance, blood pressure, triglycerides, and higher concentrations of high-density lipoprotein cholesterol, mediated through behavioural factors (diet, physical activity, sleep, media use) and waist circumference95. A better well-being was also associated with longer sleep duration and less sleep disturbances96. Longitudinally, longer sleep duration seemed to exert beneficial effects on insulin resistance through reduction of abdominal obesity97. Lower emotion-driven impulsiveness, which was associated with fewer adverse life events, also showed positive associations with well-being98. 

Bone health: Linear mixed-effects models were used to estimate the cross-sectional and longitudinal associations between physical activity, sedentary behaviour and percentiles of the bone stiffness index, stratified by weight status. It was shown that sufficient MVPA is associated with improved bone strength, particularly in children with overweight or obesity99.  

We also investigated the effect of serum 25-hydroxyvitamin D (25(OH)D) and the physical activity level on bone health. Using linear mixed-effects models it could be shown that serum 25(OH)D and serum C-terminal telopeptides of type I collagen were inversely associated. In addition, it turned out that only in participants with at least 60 minutes MVPA daily sufficient serum vitamin D levels were associated with greater bone stiffness100.  

Biomarkers: Using a linear mixed-effect model we investigated the association of serum 25(OH)D as independent variable and z-scores of inflammatory markers (CRP, cytokines, adipokines, combined inflammation score) as dependent variables. It was shown that the vitamin D status was positively associated with adiponectin and inversely associated with the inflammatory score. After stratification by weight status, these associations were only apparent in normal weight children while in children with overweight/obesity, only a positive association between 25(OH)D and the inflammatory marker IP-10 (interferon gamma inducible protein) was observed101.  

In addition, reference percentiles for two important micronutrients in children and adolescents, vitamin D and iron, were derived. The analyses revealed that the majority of European children had a deficient vitamin D status102, while only a small number of children suffered from insufficient iron status103. A better iron status was observed in children consuming more iron from animal sources. 

fMRI: As part of the contrasting groups (T4), 27 children aged 10-12 years and 32 adults aged 32-52 years participated in an additional experiment, where fMRI data were acquired while pictures of unhealthy and healthy food were presented to the participants104. In general, it seemed that unhealthy foods elicited more attention both in children and in adults, where it turned out that children had stronger brain activation while viewing unhealthy compared to healthy foods in areas involved in reward, motivation, and memory. Furthermore, children activated a motivation and reward area located in the motor cortex more strongly than did adults in response to unhealthy foods. In particular, children with a higher BMI had less activation in inhibitory areas in response to unhealthy foods, which might imply they are more susceptible to tempting food cues. 

Conclusion and future perspectives 

The above summary of some of the results obtained from the IDEFICS/I.Family cohort confirms that childhood obesity results from a complex interplay of a variety of health-related lifestyle factors. The living environment, social conditions, economic pressures and family lives have drastically changed over recent decades. Often both parents are working and the time spent together with their children is limited. Self-prepared meals from local ingredients have been replaced by fast and ready-made foods. Concerns about safety on the streets, limited availability of play spaces, exposure to TV and social media have pushed physical activity out of the daily lives of young people. These changes profoundly impact children’s health, particularly those in the most vulnerable groups. Obesity preventions programmes have therefore to follow a systemic approach addressing the overall obesogenic environment.  

In a follow-up project, called GrowH! (https://www.growh.eu/), we used an interdisciplinary strategy to further develop novel participatory prevention approaches, paying special attention to vulnerable groups most in need of effective interventions. For this purpose, we searched for successful obesity prevention programmes for children and identified the Guelph Family Health Study105 and the Amsterdam Youth-led Participatory Action Research (YPAR) approach106 that both have shown promising effects. These approaches were successfully adapted to reach families and children from socially disadvantaged backgrounds living in deprived city districts in Bremen (Germany) and Zaragoza (Spain). Taking advantage of longitudinal data from the European IDEFICS/I.Family and the Dutch ABCD107 cohorts, we were able to derive a life course perspective on the developmental trajectories of risk factors and health outcomes in children, adolescents and young adults. Using the wealth of data provided by the IDEFICS/I.Family and the ABCD cohorts, we used causal inference methods as the parametric g-formula to estimate the hypothetical effects of different interventions on unhealthy weight gain and explored, whether the impact of such interventions differs by age, sex or social position. This approach allowed us to identify critical time windows and the most promising intervention targets as well as the simulation of complex interventions108. For instance, the 13-year risk of developing overweight/obesity was 30.7% under no intervention and 25.4% when adhering to six behavioural intervention messages (see above). Here, following recommendations to limit screen time was the most effective approach, followed by enforcing physical activity. For some behaviours, we observed that already small behavioural changes showed a similar effect as compared to strictly meeting the related recommendations where membership in a sports club seemed to be a particularly promising intervention target for children of vulnerable groups. Eventually, GrowH! went beyond local prevention projects by putting them in a systems perspective and by paying special attention to barriers that might impede their successful implementation109. 

In a most recent project, the so-called Biomarkers4Pediatrics Collaboration (https://www.bips-institut.de/en/biomarkers4pediatrics.html), we are inviting principal investigators of epidemiological studies in children and adolescents with measurements of biomarkers to join our collaboration and to contribute data for pooled analyses. Our first aim is to derive new reference curves of clinical biomarkers for the paediatric practice. At the beginning of 2025, the data set already comprised more than 150,000 children and adolescents from 14 population-based studies covering 22 countries worldwide. We expect this data set to further grow substantially over the next couple of years. 

References

1. Flodmark CE, Lissau I, Moreno LA, Pietrobelli A, Widhalm K. New insights into the field of children and adolescents’ obesity: the European perspective. Int J Obes (Lond) 2004; 28: 1189-96. 
2. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA 2004; 291: 2847-50. 
3. McArthur LH, Holbert D, Peña M. Prevalence of overweight among adolescents from six Latin American cities: a multivariable analysis. Nutr Res 2003; 23: 1391-402. 
4. Moreno LA, Pigeot I, Ahrens W, eds. Epidemiology of obesity in children and adolescents – prevalence and etiology. New York, USA: Springer Publisher, 2011. 
5. Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children and adolescents, 2003-2006. JAMA 2008; 299: 2401-5.  
6. Olds TS, Tomkinson GR, Ferrar KE, Maher CA. Trends in the prevalence of childhood overweight and obesity in Australia between 1985 and 2008. Int J Obes (Lond) 2009; 34: 57-66. 
7. Péneau S, Salanave B, Maillard-Teyssier L et al. Prevalence of overweight in 6- to 15-year-old children in central/western France from 1996 to 2006: trends toward stabilization. Int J Obes (Lond) 2009; 33: 401-7.  
8. Stamatakis E, Wardle J, Cole TJ. Childhood obesity and overweight prevalence trends in England: evidence for growing socioeconomic disparities. Int J Obes (Lond) 2009; 34: 41-7. 
9. Lissner L, Sohlström A, Sundblom E, Sjöberg A. Trends in overweight and obesity in Swedish schoolchildren 1999-2005: has the epidemic reached a plateau? Obes Rev 2010; 11: 553-9. 
10. NCD Risk Factor Collaboration (NCD-RisC) Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128·9 million children, adolescents, and adults. Lancet 2017; 390: 2627-42. 
11. NCD Risk Factor Collaboration (NCD-RisC) Height and body-mass index trajectories of school-aged children and adolescents from 1985 to 2019 in 200 countries and territories: a pooled analysis of 2181 population-based studies with 65 million participants. Lancet 2020; 396: 1511-24.  
12. Ahrens W, Bammann K, De Henauw S et al., European Consortium of the IDEFICS Project. Understanding and preventing childhood obesity and related disorders – IDEFICS: A European multilevel epidemiological approach. Nutr Metab Cardiovas Dis 2006; 16: 302-8. 
13. Bammann K, Peplies J, Sjöström M et al., on behalf of the IDEFICS consortium. Assessment of diet, physical activity biological, social and environmental factors in a multi-centre European project on diet- and lifestyle-related disorders in children (IDEFICS). J Public Health 2006; 14: 279-89. 
14. Ahrens W, Bammann K, Siani A et al., on behalf of the IDEFICS consortium. The IDEFICS cohort: Design, characteristics and participation in the baseline survey. Int J Obes (Lond) 2011; 35(Suppl. 1): S3-15. 
15. De Henauw S, Verbestel V, Marild S et al., on behalf of the IDEFICS consortium. The IDEFICS Community-Oriented Intervention Program. A new model for childhood obesity prevention in Europe? Int J Obes (Lond) 2011; 35(Suppl. 1): S16-23. 
16. Ahrens W, Siani A, Adan R et al.; I.Family consortium. Cohort Profile: The transition from childhood to adolescence in European children-how I.Family extends the IDEFICS cohort. Int J Epidemiol 2017; 46: 1394-1395j. 
17. Hense S, Pohlabeln H, Michels N et al. Determinants of attrition to follow-up in a multicentre cohort study in children — Results from the IDEFICS study. Epidemiol Res Int 2013; vol 2013: Article ID 936365, 9 pages. 
18. Bammann K, Lissner L, Pigeot I, Ahrens W. Instruments for health surveys in children and adolescents. Springer International Publishing, Cham, 2019. 
19. International Standard Classification of Education (ISCED). UNESCO 1997. (Accessed April 2025 at https://uis.unesco.org/sites/default/files/documents/international-standard-classification-of-education-1997-en_0.pdf)  
20. Marfell-Jones M, Olds T, Stewart A, Carter L. International standards for anthropometric assessment. Potchefstroom, South Africa: International Society for the Advancement of Kinanthropometry, 2006. 
21. Knof K, Lanfer A, Bildstein MO, Buchecker K, Hilz H, on behalf of the IDEFICS consortium. Development of a method to measure sensory perception in children at the European level. Int J Obes (Lond). 2011; 35(Suppl 1): S131-6. 
22. Suling M, Hebestreit A, Peplies J et al., on behalf of the IDEFICS consortium. Design and results of the pretest of the IDEFICS study. Int J Obes (Lond) 2011; 35(Suppl 1): S30-44. 
23. Stomfai S, Bammann K, De Henauw S et al., on behalf of the IDEFICS consortium. Intra- and inter-observer reliability in anthropometric measurements in children. Int J Obes (Lond) 2011; 35(Suppl 1): S45-51. 
24. Peplies J, Guenther K, Bammann K et al., on behalf of the IDEFICS consortium. Influence of sample collection and preanalytical sample processing on the analyses of biological markers in the European multicentre study IDEFICS. Int J Obes (Lond) 2011; 35(Suppl 1): S104-12. 
25. Committee of Experts on Sports Research EUROFIT. Handbook for the EUROFIT tests of physical fitness, 2nd ed. Strasburg, France: Council of Europe, 1993. 
26. Trimpou P, Bosaeus I, Bengtsson BA, Landin-Wilhelmsen K. High correlation between quantitative ultrasound and DXA during 7 years of follow-up. Eur J Radiol 2010; 73: 360-4. 
27. Ahuja SP, Greenspan SL, Lin Y, Bowen A, Bartels D, Goyal RK. A pilot study of heel ultrasound to screen for low bone mass in children with leukemia. J Pediatr Hematol Oncol 2006; 28: 427-32. 
28. Khaw KT, Reeve J, Luben R et al. Prediction of total and hip fracture risk in men and women by quantitative ultrasound of the calcaneus: EPIC-Norfolk prospective population study. Lancet 2004; 363(9404): 197-202. 
29. Lanfer A, Hebestreit A, Ahrens W et al., on behalf of the IDEFICS consortium. Reproducibility of food consumption frequencies derived from the Children’s Eating Habits Questionnaire used in the IDEFICS study. Int J Obes (Lond) 2011; 35(Suppl. 1): S61-8. 
30. Herrmann D, Suling M, Reisch L et al., on behalf of the IDEFICS consortium. Repeatability of maternal report on prenatal, perinatal and early postnatal factors: findings from the IDEFICS parental questionnaire. Int J Obes (Lond) 2011; 35(Suppl 1): S52-60. 
31. Huybrechts I, Börnhorst C, Pala V et al., on behalf of the IDEFICS consortium. Evaluation of the Children’s Eating Habits Questionnaire used in the IDEFICS study by relating urinary calcium and potassium to milk consumption frequencies among European children. Int J Obes (Lond) 2011; 35(Suppl 1): S69-78. 
32. Bammann K, Sioen I, Huybrechts I et al., on behalf of the IDEFICS consortium. The IDEFICS validation study on field methods for assessing physical activity and body composition in children: design and data collection. Int J Obes (Lond) 2011; 35(Suppl 1): S79-87.  
33. Sioen I, Goemare S, Ahrens W et al., on behalf of the IDEFICS consortium. The relationship between paediatric calcaneal quantitative ultrasound measurements and dual energy X-ray absorptiometry (DXA) and DXA with laser (DXL) as well as body composition. Int J Obes 2011; 35(Suppl 1): S125-30. 
34. Bartholomew L K, Parcel GS, Kok G, Gottlieb NH, Fernandez ME. Planning health promotion programs: An intervention mapping approach. 3rd ed. San Francisco, California: Jossey-Bass, 2011.  
35. Verbestel V, De Henauw S, Maes L et al. Using the intervention mapping protocol to develop a community-based intervention for the prevention of childhood obesity in a multi-centre European project: the IDEFICS intervention. Int J Behav Nutr Phys Act 2011; 8: 82.  
36. Pigeot I, Baranowski T, De Henauw S; IDEFICS Intervention Study Group; IDEFICS consortium. The IDEFICS intervention trial to prevent childhood obesity: design and study methods. Obes Rev 2015; 16(Suppl 2): 4-15. 
37. Kovacs E, Siani A, Konstabel K, et al., on behalf of the IDEFICS consortium. Adherence to the obesity-related lifestyle intervention targets in the IDEFICS study. Int J Obes (Lond) 2014; 38(Suppl 2): S144-51. 
38. De Henauw S, Baranowski T, Pigeot I (eds.) Prevention of childhood obesity: Results from the IDEFICS study. Obes Rev 16(Suppl 2), 2015. 
39. De Henauw S, Huybrechts I, De Bourdeaudhuij I et al.; IDEFICS consortium. Effects of a community-oriented obesity prevention programme on indicators of body fatness in preschool and primary school children. Main results from the IDEFICS study. Obes Rev 2015; 16(Suppl 2):16-29. 
40. De Bourdeaudhuij I, Verbestel V, De Henauw S et al.; IDEFICS consortium.  Behavioural effects of a community-oriented setting-based intervention for prevention of childhood obesity in eight European countries. Main results from the IDEFICS study. Obes Rev 2015; 16(Suppl 2): 30-40. 
41. Lissner L, De Bourdeaudhuij I, Konstabel K et al.; IDEFICS consortium. Differential outcome of the IDEFICS intervention in overweight versus non-overweight children: did we achieve ‘primary’ or ‘secondary’ prevention? Obes Rev 2015; 16(Suppl 2): 119-26. 
42. Arvidsson L, Bogl L, Eiben G, et al.; IDEFICS Consortium. Fat, sugar and water intakes among families from the IDEFICS intervention and control groups: first observations from I.Family. Obes Rev 2015; 16(Suppl 2): 127-37. 
43. Ahrens W, Pigeot I, Pohlabeln H, et al.; IDEFICS Consortium. Prevalence of overweight and obesity in European children below the age of 10. Int J Obes (Lond) 2014; 38(Suppl 2): S99-107. 
44. Börnhorst C, Siani A, Tornaritis M et al. Potential selection effects when estimating associations between the infancy peak or adiposity rebound and later body mass index in children. Int J Obes (Lond) 2017; 41: 518-26. 
45. Ahrens W, Moreno LA, Pigeot I (eds.) Filling the gap: international reference values for health care in children. Int J Obes (Lond) 38(Suppl 2), 2014. 
46. Ahrens W, Moreno LA, Mårild S et al.; IDEFICS consortium (2014) Metabolic syndrome in young children – definitions and results of the IDEFICS study. Int J Obes (Lond) 2014; 38(Suppl 2): S4-14.  
47. Chiarelli F, Mohn A. Early diagnosis of metabolic syndrome in children. Lancet Child Adolesc Health 2017; 1: 86-8. 
48. Iguacel I, Michels N, Ahrens W et al.; IDEFICS consortium. Prospective associations between socioeconomically disadvantaged groups and metabolic syndrome risk in European children. Results from the IDEFICS study. Int J Cardiol 2018; 272:333-40. 
49. González-Gil EM, Moreno LA, Nappo A et al.; on behalf the IDEFICS consortium. Impaired metabolic health over-time and high abdominal fat are prospectively associated with high-sensitivity C-reactive protein in children: The IDEFICS study. Pediatr Obes 2021; 16: e12817. 
50. Börnhorst C, Tilling K, Russo P, et al. Associations between early body mass index trajectories and later metabolic risk factors in European children: the IDEFICS study. Eur J Epidemiol 2016; 31: 513-25. 
51. Börnhorst C, Bel-Serrat S, Pigeot I et al., on behalf of the IDEFICS consortium. Validity of 24-h recalls in (pre-)school aged children: Comparison of proxy-reported energy intakes with measured energy expenditure. Clin Nutr 2013; 33: 79-84. 
52. Börnhorst C, Huybrechts I, Ahrens W et al., on behalf of the IDEFICS consortium. Prevalence and determinants of misreporting among European children in proxy-reported 24 h dietary recalls. Br J Nutr 2013; 109: 1257-65. 
53. Börnhorst C, Huybrechts I, Hebestreit A et al., on behalf of the IDEFICS consortium. Diet-obesity associations in children: approaches to counteract attenuation caused by misreporting. Public Health Nutr 2013;16: 256-66. 
54. Pala V, Lissner L, Hebestreit A et al. Dietary patterns and longitudinal change in body mass in European children: a follow-up study on the IDEFICS multicenter cohort. Eur J Clin Nutr 2013; 67: 1042-9. 
55. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1240-3. 
56. Fernández-Alvira JM, Börnhorst C, Bammann K, et al. Prospective associations between socio-economic status and dietary patterns in European children: the Identification and Prevention of Dietary- and Lifestyle-induced Health Effects in Children and Infants (IDEFICS) Study. Br J Nutr 2015; 113: 517-25. 
57. Lissner L, Lanfer A, Gwozdz W et al. Television habits in relation to overweight, diet and taste preferences in European children – the IDEFICS study. Eur J Epidemiol 2012; 27: 705-15. 
58. Olafsdottir S, Eiben G, Prell H et al. Young children’s screen habits are associated with consumption of sweetened beverages independently of parental norms. Int J Public Health 2014; 59: 67-75. 
59. Olafsdottir S, Berg C, Eiben G et al. Young children’s screen activities, sweet drink consumption and anthropometry: results from a prospective European study. Eur J Clin Nutr 2014; 68: 223-8. 
60. Hinkley T, Verbestel V, Ahrens W,et al.; IDEFICS Consortium. Early childhood electronic media use as a predictor of poorer well-being: a prospective cohort study. JAMA Pediatr 2014; 168: 485-92. 
61. Buck C, Pohlabeln H, Huybrechts I et al. Development and application of a moveability index to quantify possibilities for physical activity in the built environment of children. Health Place 2011, 17: 1191-201. 
62. Konstabel K, Veidebaum T, Verbestel V et al., on behalf of IDEFICS consortium. Objectively measured physical activity in European children: the IDEFICS study. Int J Obes (Lond) 2014; 38(Suppl 2): S135-43.  
63. Publishable final activity report: IDEFICS –Identification and prevention of dietary- and lifestyle-induced health effects in children and infants. Brussels 2012. https://www.ideficsstudy.eu/fileadmin/idefics/downloads/publications/c_Final_Publishable_Activity_Report.pdf; last access 08 Apr 2025 
64. Evenson KR, Cattellier D, Gill K, Ondrak K, McMurray RG. Calibration of two objective measures of physical activity for children. J Sports Sci 2008; 26: 1557-65. 
65. Herrmann D, Buck C, Sioen I et al.; IDEFICS Consortium. Impact of physical activity, sedentary behaviour and muscle strength on bone stiffness in 2–10-year-old children-cross-sectional results from the IDEFICS study. Int J Behav Nutr Phys Act 2015; 12: 112. 
66. Ravens-Sieberer U, Bullinger M. Assessing health-related quality of life in chronically ill children with the German KINDL: first psychometric and content analytical results. Qual Life Res 1998; 7: 399-407. 
67. Goodman R, Meltzer H, Bailey V. The Strengths and Difficulties Questionnaire: a pilot study on the validity of the self-report version. Eur Child Adolesc Psychiatry 1998; 7: 125-30. 
68. Hense S, Barba G, Pohlabeln H, et al. Factors that influence weekday sleep duration in European children. Sleep 2011; 34: 633-9. 
69. Hense S, Pohlabeln H, De Henauw S et al. Sleep duration and overweight in European children: is the association modified by geographic region. Sleep 2011; 34: 885-90. 
70. Börnhorst C, Hense S, Ahrens W, et al.; IDEFICS Consortium. From sleep duration to childhood obesity–what are the pathways? Eur J Pediatr 2012; 171: 1029-38. 
71. Sioen I, Mouratidou T, Herrmann D et al. Relationship between markers of body fat and calcaneal bone stiffness differs between preschool and primary school children: results from the IDEFICS baseline survey. Calcif Tissue Int 2012; 91: 276-85. 
72. Lauria F, Siani A, Bammann K et al., on behalf of the IDEFICS consortium. Prospective analysis of the association of a common variant of FTO (rs9939609) with adiposity in children: Results of the IDEFICS study. PLoS One 2012; 7: e48876.  
73. Foraita R, Günther F, Gwozdz W et al.; IDEFICS consortium. Does the FTO gene interact with the socioeconomic status on the obesity development among young European children? Results from the IDEFICS study. Int J Obes (Lond) 2015; 39: 1-6. 
74. Gianfagna F, Cugino D, Ahrens W et al., on behalf of the IDEFICS consortium. Understanding the links among neuromedin U gene, beta2-adrenoceptor gene and bone health: an observational study in European children. PLoS One 2013; 8: e70632.  
75. Priego T, Sánchez J, Picó C, et al.; IDEFICS and I.Family consortia. TAS1R3 and UCN2 transcript levels in blood cells are associated with sugary and fatty food consumption in children. J Clin Endocrinol Metab 2015; 100: 3556-64. 
76. Sánchez J, Priego T, Picó C et al., on behalf of the IDEFICS consortium. Blood cells as a source of transcriptional biomarkers of childhood obesity and its related metabolic alterations: results of the IDEFICS study. J Clin Endocrinol Metab 2012; 97: E648-52.  
77. Lanfer A, Knof K, Barba G et al. Taste preferences in association with dietary habits and weight status in European children: results from the IDEFICS study. Int J Obes (Lond) 2012; 36: 27-34. 
78. Nagrani R, Foraita R, Gianfagna F et al. Common genetic variation in obesity, lipid transfer genes and risk of metabolic syndrome: Results from IDEFICS/I.Family study and meta-analysis. Sci Rep 2020; 10: 7189. 
79. Sina E, Buck C, Veidebaum T et al.; IDEFICS, I.Family consortia. Media use trajectories and risk of metabolic syndrome in European children and adolescents: the IDEFICS/I.Family cohort. Int J Behav Nutr Phys Act 2021; 18: 134. 
80. González-Gil EM, Moreno LA, Nappo A et al.; on behalf the IDEFICS consortium. Impaired metabolic health over-time and high abdominal fat are prospectively associated with high-sensitivity C-reactive protein in children: The IDEFICS study. Pediatr Obes 2021; 16: e12817. 
81. Börnhorst C, Russo P, Veidebaum T et al. Metabolic status in children and its transitions during childhood and adolescence-the IDEFICS/I.Family study. Int J Epidemiol 2019; 48: 1673-83. 
82. Börnhorst C, Russo P, Veidebaum T et al.. The role of lifestyle and non-modifiable risk factors in the development of metabolic disturbances from childhood to adolescence. Int J Obes (Lond). 2020; 44: 2236-45. 
83. Foraita R, Witte J, Börnhorst C et al. A longitudinal causal graph analysis investigating modifiable risk factors and obesity in a European cohort of children and adolescents. Sci Rep 2024; 14: 6822. 
84. Coumans JMJ, Danner UN, Ahrens W et al. The association of emotion-driven impulsiveness, cognitive inflexibility and decision-making with weight status in European adolescents. Int J Obes (Lond) 2018; 42: 655-61. 
85. Coumans JMJ, Danner UN, Intemann T et al.; I.Family Consortium. Emotion-driven impulsiveness and snack food consumption of European adolescents: Results from the I.Family study. Appetite 2018; 123: 152-9. 
86. Sina E, Buck C, Veidebaum T et al.; IDEFICS, I.Family consortia. Media use trajectories and risk of metabolic syndrome in European children and adolescents: the IDEFICS/I.Family cohort. Int J Behav Nutr Phys Act 2021; 18: 134. 
87. Sina E, Buck C, Ahrens W et al., on behalf of the I.Family consortium. Digital media use in association with sensory taste preferences in European children and adolescents – Results from the I.Family study. Foods 2021;10: 377. 
88. Sina E, Boakye D, Christianson L, Ahrens W, Hebestreit A. Social media and children’s and adolescents’ diets: A systematic review of the underlying social and physiological mechanisms. Advances in Nutrition 2022; 13: 913-37. 
89. Do S, Didelez V, Börnhorst C et al.; IDEFICS/I.Family consortia. The role of psychosocial well-being and emotion-driven impulsiveness in food choices of European adolescents. Int J Behav Nutr Phys Act 2024; 21: 1. 
90. Sprengeler O, Pohlabeln H, Bammann K et al. Trajectories of objectively measured physical activity and childhood overweight: longitudinal analysis of the IDEFICS/I.Family cohort. Int J Behav Nutr Phys Act 2021;18: 103. 
91. Bogl LH, Mehlig K, Intemann T et al.; I.Family Consortium. A within-sibling pair analysis of lifestyle behaviours and BMI z-score in the multi-centre I.Family study. Nutr Metab Cardiovasc Dis 2019; 29: 580-9. 
92. Bogl LH, Mehlig K, Ahrens W et al.; IDEFICS and I. Family Consortia. Like me, like you – relative importance of peers and siblings on children’s fast food consumption and screen time but not sports club participation depends on age. Int J Behav Nutr Phys Act 2020;17: 50. 
93. Börnhorst C, Ahrens W, De Henauw S et al. Age-specific quantification of overweight/obesity risk factors from infancy to adolescence and differences by educational level of parents. Int J Public Health 2023;68: 1605798. 
94. Stahlmann K, Hebestreit A, De Henauw S et al. A cross-sectional study of obesogenic behaviours and family rules according to family structure in European children. Int J Behav Nutr Phys Act 2020; 17: 32. 
95. Thumann BF, Börnhorst C, Ahrens W et al.; IDEFICS and I.Family consortia. Cross-sectional and longitudinal associations between psychosocial well-being and cardiometabolic markers in European children and adolescents. Psychosom Med 2020; 82: 764-73. 
96. Thumann BF, Börnhorst C, Michels N et al.; IDEFICS and I.Family consortia. Cross-sectional and longitudinal associations between psychosocial well-being and sleep in European children and adolescents. J Sleep Res 2019; 28: e12783. 
97. Thumann BF, Michels N, Felső R et al.; IDEFICS and I. Family Consortia. Associations between sleep duration and insulin resistance in European children and adolescents considering the mediating role of abdominal obesity. PLoS One 2020; 15: e0235049. 
98. Do S, Coumans JMJ, Börnhorst C et al. Associations Between Psychosocial Well-Being, Stressful Life Events and Emotion-Driven Impulsiveness in European Adolescents. J Youth Adolesc 2022; 51: 1106-17. 
99. Cheng L, Pohlabeln H, Ahrens W et al.; IDEFICS and I. Family Consortia. Cross-sectional and longitudinal associations between physical activity, sedentary behaviour and bone stiffness index across weight status in European children and adolescents. Int J Behav Nutr Phys Act 2020; 17: 54. 
100. Cheng L, Pohlabeln H, Wolters M et al., on behalf of the IDEFICS and I.Family consortia. Moderate-to-vigorous physical activity modulates the association between serum 25-hydroxyvitamin D and bone stiffness in European children and adolescents. Nutrition Research 2025; 133: 64-77. 
101. Wolters M, Foraita R, Moreno LA et al.; IDEFICS and I.Family consortia. Longitudinal associations between vitamin D status and biomarkers of inflammation in a pan-European cohort of children and adolescents. Eur J Nutr 2024; 63: 3047-60. 
102. Wolters M, Intemann T, Russo P et al. 25-Hydroxyvitamin D reference percentiles and the role of their determinants among European children and adolescents Eur J Clin Nutr 2022; 76: 564-73. 
103. Floegel A, Intemann T, Siani A et al., on behalf of the IDEFICS and I.Family consortia. Cohort-based reference values for serum ferritin and transferrin and longitudinal determinants of iron status in European children aged 3 – 15 y. J Nutr 2024; 154: 658-69. 
104. van Meer F, van der Laan LN, Charbonnier L et al.; I.Family Consortium. Developmental differences in the brain response to unhealthy food cues: an fMRI study of children and adults. Am J Clin Nutr 2016,104:1515-22. Erratum in: Am J Clin Nutr 2017; 105: 772. 
105. Haines J, Douglas S, Mirotta JA et al. Guelph Family Health Study: pilot study of a home-based obesity prevention intervention. Can J Public Health 2018; 109: 549-60. 
106. Anselma M, Altenburg T, Chinapaw M. Kids in Action: the protocol of a Youth Participatory Action Research project to promote physical activity and dietary behaviour. BMJ Open 2019; 9: e025584. 
107. van Eijsden M, Vrijkotte TG, Gemke RJ, van der Wal MF. Cohort profile: the Amsterdam Born Children and their Development (ABCD) study. Int J Epidemiol 2011; 40: 1176-86. 
108. Börnhorst C, Pigeot I, De Henauw S et al.; GrowH! consortium. The effects of hypothetical behavioral interventions on the 13-year incidence of overweight/obesity in children and adolescents. Int J Behav Nutr Phys Act 2023; 20: 100. 
109. De Henauw S, Altenburg T, Sotomayor Sainz A et al.; GrowH! consortium. Striving for a systems approach in childhood obesity prevention. A White Paper based on the GrowH! Project. 2024; https://www.growh.eu/publications.html; last access 09 Apr 2025.