Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
April 6, 2009

Dutch Obesity Intervention in TeenagersEffectiveness of a School-Based Program on Body Composition and Behavior

Author Affiliations

Author Affiliations: Departments of Public and Occupational Health (Drs Singh, Chin A Paw, and van Mechelen), VU University Medical Center, EMGO Institute, Amsterdam, the Netherlands (Drs Brug, Singh, Chin A Paw, and van Mechelen).


Copyright 2009 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2009

Arch Pediatr Adolesc Med. 2009;163(4):309-317. doi:10.1001/archpediatrics.2009.2

Objective  To determine whether a multicomponent health promotion intervention for Dutch adolescents (defined as persons between 12 and 14 years of age) would be successful in influencing body composition and dietary and physical activity behavior in both the short and long terms.

Design  Randomized controlled trial.

Setting  Ten intervention and 8 control prevocational secondary schools.

Participants  A total of 1108 adolescents (mean age, 12.7 years).

Intervention  An interdisciplinary program with an adapted curriculum for 11 lessons in biology and physical education and environmental change options.

Main Outcome Measures  Body height and weight, waist circumference, 4 skinfold thickness measurements, and dietary and physical activity behavior data.

Results  Multilevel analyses showed that the intervention remained effective in preventing unfavorable increases in important measures of body composition after 20-month follow-up in girls (biceps skinfold and sum of 4 skinfolds) and boys (triceps, biceps, and subscapular skinfolds). Consumption of sugar-containing beverages was significantly lower in intervention schools both after intervention (boys: −287 mL/d; 95% confidence interval [CI], −527 to −47; girls: −249; −400 to −98) and at 12-month follow-up (boys: −233; −371 to −95; girls: −271; −390 to −153). For boys, screen-viewing behavior was significantly lower in the intervention group after 20 months (−25 min/d; 95% CI, −50 to −0.3). No significant intervention effects on consumption of snacks or active commuting to school were found.

Conclusion  The Dutch Obesity Intervention in Teenagers program resulted in beneficial effects on the sum of skinfold thickness measurements in girls and consumption of sugar-containing beverages in both boys and girls in both the short and long terms.

To curb the obesity epidemic, population-based preventive approaches are required that affect multiple behaviors.1 Related to the reduced opportunities for being physically active and the increased availability and accessibility of high-energy foods—the obesogenic environment2—there is consensus that changes in lifestyle are driving the obesity epidemic.3 Adolescence is an important transition period in the life course in which lifestyle, including energy balance–related behaviors (EBRBs), is subject to important changes4 (eg, levels of physical activity decrease5,6 and food choices become more autonomous7 and less healthy).8 Unfavorable behavioral patterns established during this period of life may be vital in the development of adult health behaviors,9 and health behaviors in youth may track to a certain extent into adulthood.10

The most promising interventions focus on improving dietary and physical activity patterns11 and combine educational and environmental elements.12 Since schools can provide valuable opportunities to influence the lifestyles of adolescents in a collective setting,13,14 recent research within the field of obesity prevention among youth has therefore focused on interventions in the school setting. Most school-based interventions were able to show improvements in dietary or physical activity patterns; however, hardly any study could demonstrate significant effects on indicators of adiposity.15

There is an urgent need for more scientific evidence about intervention programs that effectively contribute to prevention of excessive weight gain in adolescents, especially in the long term. Unless intervention programs and the theoretical models on which they are based are identified and described more accurately, it will be difficult to conclude which programs are effective and whether they have a solid theoretical basis.16,17

We performed a randomized controlled trial, the Dutch Obesity Intervention in Teenagers (DOiT), to evaluate the effectiveness of a school-based multicomponent health promotion intervention aimed at preventing excessive weight gain. The DOiT program was developed, implemented, and evaluated systematically, applying the Intervention Mapping protocol.18 Short-term results of our trial showed that the DOiT intervention was associated with lower skinfold thickness in girls and lower waist circumference in boys.19 In this article, we describe the short-term and long-term effects on the anthropometric and behavioral outcome measures.


A total of 18 prevocational secondary schools participated in the randomized controlled trial. Participating schools were asked to select 3 classes of first-year students (aged 12-14 years). The selection of classes was based on practical reasons (eg, similar timetables for lessons in physical education). No inclusion criteria were set for students to take part in the study. Written informed consent was obtained from all students and their parents. The Medical Ethics Committee of the VU University Medical Center approved the study protocol.

Schools were randomly assigned to either the intervention or control group. Randomization took place at the school level or at the location level (in case 2 schools were located in the same city) and was stratified by degree of urbanization, using SPSS statistical software (SPSS Inc, Chicago, Illinois). We collected data at baseline (at the start of the first school year of secondary school) and after 8 months (immediately after intervention), 12 months (4 months after intervention), and 20 months (12 months after intervention).

Sample size calculation was based on changes in body weight. Assuming an α  of .05, power of 90%, and a 2-sided test, we needed 233 participants per group to show a mean (SD) difference in body weight of 0.5 (1.5) kg between the intervention and control groups. To be able to perform multilevel analyses and at the same time take into account the cluster randomization design, a sample size between 500 and 600 individuals from 16 schools was required. We increased the sample size to allow for dropout.


The aim of DOiT was to increase awareness and to induce behavioral changes concerning energy intake and energy output. Behaviors targeted with regard to energy intake were consumption of sugar-containing beverages and high-energy snacks. Behaviors targeted with regard to energy output were physical activity and screen-viewing behavior. The intervention consisted of an individual component (ie, an educational program covering 11 lessons for the courses of biology and physical education) and an environmental component (ie, encouraging schools to offer additional physical education classes and advice for schools on changes in and around school cafeterias). We developed the DOiT program by applying the Intervention Mapping protocol, which facilitates a systematic process of designing health promotion interventions and is based on theory and empirical evidence.20 The development and content of the DOiT program are described in more detail elsewhere.18 Control schools were asked to maintain their regular curriculum.


Primary outcome measures were changes in body composition (ie, waist circumference, skinfold thickness, and body mass index [BMI]). Secondary outcome measures were changes in dietary and physical activity behavior (ie, EBRBs). All measurements were performed within a 6-week period according to a standardized protocol by a trained research team, which was not blinded to the group assignment.


Body height was measured and recorded with a portable stadiometer (Seca 225; Seca Deutschland, Hamburg, Germany) with an accuracy of 1 mm. We attached a level to the stadiometer to ensure that head posture was positioned correctly for the measurement. Body weight was measured and recorded within 0.1 kg using a calibrated electronic flat scale (Seca 888; Seca Deutschland), which was leveled after each placement. The BMI was calculated as weight in kilograms divided by the height in meters squared; the categories of overweight and obese were defined using the cutoff values as described by Cole et al.21 Skinfolds were measured on the right side of the body to the nearest 0.2 mm, using a Harpenden skinfold caliper (Baty International, West Sussex, England).22 We determined skinfold thickness by averaging 2 measurements. If the 2 measurements differed by more than 1 mm, a third measurement was performed, and skinfold thickness was then taken as the average of the 3 measurements. Waist circumference was measured and recorded with a flexible band (Seca 200; Seca Deutschland) to an accuracy of 0.5 cm. The anatomical landmark for the waist circumference was laterally midway between the lowest portion of the rib cage and the iliac crest. During all measurements, students were dressed in underwear. We ensured privacy during the measurements by measuring 1 adolescent at a time in a separate room.

Before each measurement period, we determined the intrarater and interrater reliability of the skinfold thickness and waist circumference measurements. Values for intrarater reliability varied from 0.94 to 0.99 (skinfold thickness) and 0.82 to 0.97 (waist circumference). Values for interrater reliability varied from 0.90 to 0.99 (skinfold thickness) and 0.88 to 0.97 (waist circumference). All anthropometric measurements were performed by the same research assistant, who was not a co-author of this article.


Demographic and behavioral data were derived from a questionnaire that adolescents completed in the classroom. A teacher and/or a member of the DOiT team supervised the completion of the questionnaires. The following EBRBs were addressed: (1) consumption of sugar-containing beverages (ie, consumption of soft drinks and fruit juices), (2) consumption of high-energy snacks (ie, consumption of savory snacks and sweet snacks), (3) screen-viewing behavior (ie, time spent on television viewing and computer use), and (4) active commuting to school. Because there are no validated Dutch questionnaires addressing the EBRBs selected for our study, we adapted other validated questionnaires on dietary intake,23,24 screen-viewing behavior,25 and physical activity26 in adolescents. The structure and sequence of the questions were equal for all EBRBs (eTable). The questionnaire was pretested for clarity and length among adolescents not participating in the study by means of cognitive interviewing.27 Frequency and quantity of the reported items were multiplied to obtain estimates of mean daily consumption. Behavior reported above the 95th percentile was recoded as the value of the 95th percentile because these unrealistically high values are probably due to overreporting.


We used the nonparametric Kolmogorov-Smirnov test (for all measurements of body composition, physical fitness, and EBRBs) and the Pearson χ2 test (BMI class) to compare groups at baseline. The effects of the intervention on all anthropometric measures were evaluated by multilevel analysis, using a software package (MLwiN, version 2.0; Centre for Multilevel Modeling at the University of Bristol, England). With the use of this technique, regression coefficients could be adjusted for the clustering of observations within 1 school and/or class. We defined 3 levels in our multilevel analysis: (1) student, (2) class, and (3) school. The parameters of interest are the regression coefficients of the linear regression models, indicating the effect of the intervention compared with the control group. In the crude model, the outcome value at 8, 12, or 20 months was adjusted for baseline value. We checked for possible effect modification by sex or ethnicity by including group × sex and group × ethnicity interaction terms in the regression analyses. All analyses were performed according to the intention-to-treat principle. Missing values were not imputed.


In total, we obtained data from 1108 students (ie, anthropometric or behavioral data at baseline and at least 1 follow-up measurement) (Figure). Table 1 presents the baseline characteristics of our sample. Mean baseline data stratified by sex revealed significant differences between intervention and control schools with regard to body weight (boys), body height (girls), and screen-viewing behavior (girls).

Image not available

Flow of the participants through the trial.

Table 1. 
Image not available
Demographic and Anthropometric Characteristics at Baseline in the Intervention and Control Schools

We present the means for all anthropometric outcome measures at baseline, the 3 follow-up measurements, and the results of the multilevel analyses (regression coefficients of the crude model) in Table 2 and Table 3 for girls and boys, respectively. Because no significant group × ethnicity interactions were found, analyses were not stratified for ethnicity. Sex was found to be an effect modifier. Therefore, analyses were conducted separately for boys and girls.

Table 2. 
Image not available
Anthropometric Characteristics in Girls at Baseline and After 8-, 12-, and 20-Month Follow-ups and the Mean Differences in Change Between the Intervention and Control Schools
Table 3. 
Image not available
Anthropometric Characteristics in Boys at Baseline and After 8-, 12-, and 20-Month Follow-ups and the Mean Differences in Change Between the Intervention and Control Schools

At the 20-month follow-up, we found a significant intervention effect on biceps skinfold thickness among girls (−0.7 mm; 95% confidence interval [CI], −1.3 to −0.04 mm). Another significant intervention effect was found with regard to the sum of skinfold thickness: the sum of skinfold thickness was lower in the girls of the intervention schools at both 8-month (−2.3 mm; 95% CI, −4.3 to −0.3 mm) and 20-month follow-ups (−2.0; −3.9 to −0.1 mm).

For boys, waist circumference was significantly lower in the intervention group at 8-month follow-up (−0.6 cm; 95% CI, −1.1 to −0.1 cm), but at 20-month follow-up waist circumference was significantly lower in the control group. For the 20-month follow-up, we found a significant intervention effect on the triceps (−0.7 mm; 95% CI, −1.2 to −0.1 mm), biceps (−0.4 mm; −0.8 to −0.1 mm), and subscapular skinfold thickness (−0.5 mm; −1.0 to −0.1 mm).


The means for all behavioral outcome measures at baseline, the 3 follow-up measurements, and the results of the multilevel analyses (regression coefficients of the crude model) are given in Table 3 for boys and Tables 4 and 5 for girls, respectively. No significant group × ethnicity interactions were found, and the analyses were thus not stratified for ethnicity.

Table 4. 
Image not available
Energy Balance-Related Behaviors in Girls at Baseline and After 8-, 12-, and 20-Month Follow-ups and the Mean Differences in Change Between the Intervention and Control Schools
Table 5. 
Image not available
Energy Balance-Related Behaviors in Boys at Baseline and After 8-, 12-, and 20-Month Follow-ups and the Mean Differences in Change Between the Intervention and Control Schools

Consumption of sugar-containing beverages was significantly lower among boys and girls of the intervention schools both at the 8-month (boys: −287 mL/d; 95% CI, −527 to −47; girls: −249; −400 to −98) and 12-month follow-ups (boys: −233; −371 to −95.1; girls: −271; −390 to −153). We found no significant between-group differences at the 20-month follow-up.

Differences in screen-viewing behavior consistently favored the boys and girls from the intervention schools at all follow-up measurements, with statistically significant differences in favor of the boys of the intervention group after 20 months (−25 min/d; 95% CI, −50 to −0.3 min/d). No significant intervention effect on consumption of snacks or active commuting to school was found.


The DOiT program had significant beneficial effects on body composition in girls and consumption of sugar-containing beverages in boys and girls, compared with the regular curriculum in a sample of prevocational education students. Hence, our results do not show consistently positive findings on all anthropometric and behavioral outcome measures. Our findings are important, especially when considering the need for evidence on the long-term effectiveness of interventions in the field of obesity prevention.


Our intervention was successful in influencing skinfold thickness in the short and long terms. In comparison between BMI and measurement of skinfold thickness, the latter is known to be a more sensitive alternative for determining body fatness, especially in youth, because it discriminates better between lean mass and fat mass.28,29 A recent review30 also recommended the use of skinfold measures when interventions aim to improve physical activity patterns. The fact that we found between-group differences in skinfold thickness but not in BMI might therefore indicate that BMI in children and adolescents represents body build31 rather than body fatness.


Our findings are in line with those of McMurray et al,32 who found significant between-group differences in skinfold thickness but not in BMI. In the study by McMurray et al, the school-based intervention aimed at increasing the aerobic component of the school's physical activity program and at improving knowledge about the relationship of weight control and blood pressure to body fat in early adolescents. This intervention was shorter when compared with the DOiT intervention, although more intensive. Gortmaker et al33 found a decrease in the prevalence of obesity in the girls of the intervention group, as indicated by a composite measure of obesity (ie, BMI and triceps skinfold thickness measurements). However, they assessed the effectiveness of a school-based program of a longer duration and higher frequency (ie, 2 school years with 32 lessons [Planet Health] compared with 1 school year with 11 lessons [DOiT]).

We tried to develop an intervention that is relatively easy to implement (ie, no additional school staff required, low implementation cost, and good fit with existing curriculum, meaning that intervention program lessons could partly replace regular biology lessons). Nevertheless, our process evaluation among teachers indicated that implementation required more time than expected and was perceived as being demanding (A.S.S. unpublished data, presented at the Amsterdam/Cambridge/Ghent Meeting in Ghent, the Netherlands, January 28-30, 2007). Time constraints are a commonly expressed problem in the school setting in general and are related to health promotion projects as indicated by other comparable studies in this field.34,35 This problem suggests that a more intensive intervention would be less feasible in the current Dutch school system.

The DOiT intervention resulted in significantly lower self-reported consumption of sugar-containing beverages after 8 and 12 months (boys and girls). The difference between the intervention and control groups was approximately 250 mL/d. We are aware of 1 other comparable study that reported successfully reducing consumption of sugar-containing beverages. James et al36 studied the effectiveness of an intervention aimed at reducing consumption of carbonated drinks among preschool children. James et al reported changes of 150 mL during 3 days, which indicates that their study aim and methods are similar to ours. Although the magnitude of the effects of the study by James et al was smaller compared with our results, one should bear in mind that the difference in magnitude of the effects might partly be attributed to the difference in the absolute consumption of drinks of all types between preschoolers and adolescents.

With regard to screen-viewing behavior, a statistically significant difference after 20 months was found for boys. This finding is in line with findings of several other studies25,33 that have reported beneficial intervention effects on television viewing and body composition. However, the fact that we found no differences with regard to screen-viewing behavior after 8 and 12 months may be because of the lack of sensitivity of our questionnaire. Bearing in mind that our intervention was aimed at small behavioral changes, a plausible explanation for the lack of effect, besides the absence of effect of the intervention, could be that the questionnaire used was not sensitive enough to detect such small changes as those that occurred between the groups in our study.

We found no significant differences between the intervention and control groups with regard to active commuting to school, neither in the short nor long term. Bicycling is common in the Netherlands, and 95.0% of the adolescents in our baseline sample reported bicycling to school 5 days a week. Thus, our intervention may not have been able to considerably improve this behavior. The intervention program encouraged active commuting not only to school but also during leisure time (eg, the use of a bicycle to visit friends or to go to a sports club). Because our measurements included only active commuting to school, we were not able to measure possible intervention effects on active commuting during leisure time.

Our results indicate no significant intervention effect on consumption of high-energy snacks. A systematic review that reported estimates of the prevalence of being overweight in school-aged youth showed a significant negative relationship between the intake of sweet snacks and the probability of being overweight,37,38 as do our baseline data (Spearman ρ for sum of skinfold thickness measurements and consumption of sweet snacks: boys = −0.15, P = .002; girls = −0.16, P = .001). Although these are cross-sectional data, this finding indicates that weight control behavior may moderate the relationship between EBRBs and body composition.38 This complicates the ability to find the effects of intervention on behaviors that are already well-known weight control strategies. Another explanation for the lack of effect may be a possible lack of sensitivity on the part of the questionnaire in the detection of small behavioral changes.

Body Composition

Because research assistants were involved in arranging to take measurements, conducting the measurements, and delivering the intervention materials, they were not blinded to the group assignment. The remainder of the research team, which consisted of students who were helping with measurements, were not informed about the group assignment. The performance of all measurements according to a standardized protocol minimized the potential for observation bias.

As previously mentioned, the sum of skinfold thickness measurements best reflects body composition, and effects on this measurement are therefore most important. We have presented differences in the sum of skinfold and separate skinfold thickness measurements. Because these separate analyses increase type I error, the results for the separate skinfold thickness measurements should be interpreted with caution.

Energy Balance–Related Behaviors

The use of self-report questionnaires to assess behavioral change is subject to limitations, a drawback our study shares with many studies in the field. Hence, doubt remains as to whether observed changes in self-reported measures reflect behavioral changes. We used an extensive questionnaire that took students approximately 60 minutes to complete. The order in which specific behaviors were assessed was as follows: consumption of sugar-containing beverages, screen-viewing behavior, active commuting to school, and consumption of snacks. We find small but significant between-group differences with regard to the first (consumption of sugar-containing beverages) and second (screen-viewing) questioned behaviors but failed to pick up differences in the third (active commuting to school) and last (snack consumption) questioned behaviors. Confidence in the adequacy of the assessment of behavioral change via questionnaire was somewhat diminished based on observations during the measurements that suggested that some students (especially ones with dyslexia) had difficulties with the length of the questionnaire, which may have caused less adequate completion of the latter parts. On the other hand, it has already been considered unlikely that our intervention would change active commuting behaviors in a country where bicycles are frequently used for transportation.

The fact that we found no differences with regard to screen-viewing behavior after 8 and 12 months contrasts with our findings after 20 months and forces us to question the sensitivity of our questionnaire. As with most questionnaires, we used a tool that was designed to assess actual behavior. Bearing in mind that our intervention was aimed at small behavioral changes, a plausible explanation for the lack of effect could be that the questionnaire used was not sensitive enough to detect such small changes between groups as those that occurred in our study.

The participating schools clearly represented a highly motivated group. This self-selection bias was unavoidable but reflected the reality that only interested schools will participate in evaluation studies such as ours. The compliance rate among adolescents was relatively high and possibly in part due to the motivation of the participating schools and teachers.

Common problems that are likely to have an effect on the intervention effects found on body composition were relatively small but statistically significant. In theory, altering the energy gap by as little as 100 kcal/d (eg, a reduction of intake of approximately 0.2 L of sugar-containing beverages per day or an additional 15 minutes of bicycling [20 km/h] per day) is considered to prevent weight gain.3 On a population level, this would mean that no major changes in body composition and behavior among normal weight youth would be required to prevent excessive weight gain in the long term. Therefore, the DOiT intervention has provided promising evidence that interdisciplinary school-based interventions, developed on a solid theoretical basis, can make a significant contribution to obesity prevention in adolescents on a public health level. To achieve more sustained changes with regard to anthropometry and all targeted behaviors, our intervention might benefit from improvements with regard to parental involvement and more focus on environmental changes. Future interventions might consider the possibility or the need for a booster component (beyond the 11 lessons given) and the maintenance of the environmental changes in the schools to produce greater benefit. Health and population effects can only be expected if interventions are maintained, as has been shown by the implementation of the Planet Health intervention.33


The DOiT intervention resulted in some beneficial effects on the sum of skinfold thickness measurements in girls and consumption of sugar-containing beverages in boys and girls in the short and long terms. Reducing intake of sugar-containing beverages should therefore be considered a good behavioral target for future interventions aimed at the prevention of overweight among adolescents.

Back to top
Article Information

Correspondence: Amika S. Singh, PhD, van der Boechorststraat 7, 1081 BT Amsterdam, the Netherlands (

Accepted for Publication: August 7, 2008.

Author Contributions: Dr Singh had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Singh, Chin A Paw, Brug, and van Mechelen. Acquisition of data: Singh. Analysis and interpretation of data: Singh and Brug. Drafting of the manuscript: Singh, Chin A Paw, and Brug. Critical revision of the manuscript for important intellectual content: Singh, Chin A Paw, Brug, and van Mechelen. Statistical analysis: Singh and Chin A Paw. Obtained funding: Chin A Paw, Brug, and van Mechelen. Administrative, technical, and material support: Singh, Chin A Paw, and van Mechelen. Study supervision: Chin A Paw, Brug, and van Mechelen.

Financial Disclosure: None reported.

Funding/Support: This study is part of the Netherlands Research Programme for Weight Gain Prevention and is funded by grant 2000Z002 from the Netherlands Heart Foundation, the Dutch Ministry of Health, Welfare, and Sports, and the Royal Association of Teachers of Physical Education (KVLO).

Disclaimer: None of the funders had input into protocol development, data collection, or analyses or interpretation.

Additional Contributions: We thank the students and the staff of the schools for their participation in the trial and their support.

World Health Organization, Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation.  Geneva, Switzerland World Health Organization2000;WHO technical report series 894
Egger  GSwinburn  B An “ecological” approach to the obesity pandemic. BMJ 1997;315 (7106) 477- 480
Hill  JOWyatt  HRReed  GWPeters  JC Obesity and the environment: where do we go from here? Science 2003;299 (5608) 853- 855
Kirk  SScott  BJDaniels  SR Pediatric obesity epidemic: treatment options. J Am Diet Assoc 2005;105 (5) ((suppl 1)) S44- S51
Riddoch  CJAndersen  LBWedderkopp  N  et al.  Physical activity levels and patterns of 9- and 15-yr-old European children. Med Sci Sports Exerc 2004;36 (1) 86- 92
Kimm  SYGlynn  NWKriska  AM  et al.  Decline in physical activity in black girls and white girls during adolescence. N Engl J Med 2002;347 (10) 709- 715
Spruijt-Metz  D Personal incentives as determinants of adolescent health behavior: the meaning of behavior. Health Educ Res 1995;10 (3) 355- 364
Lytle  LASeifert  SGreenstein  J McGovern  P How do children's eating patterns and food choices change over time? results from a cohort study. Am J Health Promot 2000;14 (4) 222- 228
Leger  LSNutbeam  D Research into health promoting schools. J Sch Health 2000;70 (6) 257- 259
te Velde  SJTwisk  JWBrug  J Tracking of fruit and vegetable consumption from adolescence into adulthood and its longitudinal association with overweight. Br J Nutr 2007;98 (2) 431- 438
Katz  DLO'Connell  MNjike  VYYeh  MCNawaz  H Strategies for the prevention and control of obesity in the school setting: systematic review and meta-analysis. Int J Obes 2008;32 (12) 1780- 1789Article
Swinburn  BEgger  GRaza  F Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Prev Med 1999;29 (6, pt 1) 563- 570
Summerbell  CDWaters  EEdmunds  LDKelly  SBrown  TCampbell  KJ Interventions for preventing obesity in children. Cochrane Database Syst Rev 2005; (3) CD001871
Stice  EShaw  HMarti  CN A meta-analytic review of obesity prevention programs for children and adolescents: the skinny on interventions that work. Psychol Bull 2006;132 (5) 667- 691
Hardeman  WGriffin  SJohnston  MKinmonth  ALWareham  NJ Interventions to prevent weight gain: a systematic review of psychological models and behaviour change methods. Int J Obes Relat Metab Disord 2000;24 (2) 131- 143
Bartholomew  LKParcel  GSKok  GGottlieb  NH Planning Health Promotion Programs.  San Francisco, CA Jossey-Bass2006;
Kremers  SPVisscher  TLSBrug  J  et al.  Netherlands research programme weight gain prevention (NHF-NRG): rationale, objectives and strategies. Eur J Clin Nutr 2005;59 (4) 498- 507
Singh  ASChin  APaw  JM  et al.  Design of the Dutch Obesity Intervention in Teenagers (NRG-DOiT): systematic development, implementation and evaluation of a school-based intervention aimed at the prevention of excessive weight gain in adolescents. BMC Public Health 2006;6304- 319
Singh  ASChin  APaw  MJBrug  JVan Mechelen  W Short-term effects of school-based weight gain prevention among adolescents. Arch Pediatr Adolesc Med 2007;161 (6) 565- 571
Bartholomew  LKParcel  GSKok  GGottlieb  NHIntervention Mapping: Designing Theory and Evidence-Based Health Promotion Programs. Mountain View, CA Mayfield Publishing2001;
Cole  TJBellizzi  MCFlegal  KMDietz  WH Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320 (7244) 1240- 1243
Durnin  JVRahaman  MM The assessment of the amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 1967;21 (3) 681- 689
Van Assema  PBrug  JRonda  GSteenhuis  IOenema  A A short Dutch questionnaire to measure fruit and vegetable intake: relative validity among adults and adolescents. Nutr Health 2002;16 (2) 85- 106
van Assema  PBrug  JRonda  GSteenhuis  I The relative validity of a short Dutch questionnaire as a means to categorize adults and adolescents to total and saturated fat intake. J Hum Nutr Diet 2001;14 (5) 377- 390
Robinson  TN Reducing children's television viewing to prevent obesity: a randomized controlled trial. JAMA 1999;282 (16) 1561- 1567
Booth  MLOkely  ADChey  TNBauman  A The reliability and validity of the Adolescent Physical Activity Recall Questionnaire. Med Sci Sports Exerc 2002;34 (12) 1986- 1995
van der Horst  KKremers  SFerreira  ISingh  AOenema  ABrug  J Perceived parenting style and practices and the consumption of sugar-sweetened beverages by adolescents. Health Educ Res 2007;22 (2) 295- 304
Nooyens  ACJKoppes  LLJVisscher  TLS  et al.  Adolescent skinfold thickness is a better predictor of high body fatness in adults than is body mass index: the Amsterdam Growth and Health Longitudinal Study. Am J Clin Nutr 2007;85 (6) 1533- 1539
Zimmermann  MBGubeli  CPuntener  CMolinari  L Detection of overweight and obesity in a national sample of 6–12-y-old Swiss children: accuracy and validity of reference values for body mass index from the US Centers for Disease Control and Prevention and the International Obesity Task Force. Am J Clin Nutr 2004;79 (5) 838- 843
Doak  CMVisscher  TLSRenders  CMSeidell  JC The prevention of overweight and obesity in children and adolescents: a review of interventions and programmes. Obes Rev 2006;7 (1) 111- 136
Wright  CMParker  LLamont  DCraft  AW Implications of childhood obesity for adult health: findings from thousand families cohort study. BMJ 2001;323 (7324) 1280- 1284
McMurray  RGHarrell  JSBangdiwala  SIBradley  CBDeng  SLevine  A A school-based intervention can reduce body fat and blood pressure in young adolescents. J Adolesc Health 2002;31 (2) 125- 132
Gortmaker  SLPeterson  KWiecha  J  et al.  Reducing obesity via a school-based interdisciplinary intervention among youth: Planet Health. Arch Pediatr Adolesc Med 1999;153 (4) 409- 418
Neumark-Sztainer  DStory  MHannan  PJRex  J New Moves: a school-based obesity prevention program for adolescent girls. Prev Med 2003;37 (1) 41- 51
Martens  Mvan Assema  PPaulussen  TSchaalma  HBrug  J Krachtvoer: process evaluation of a Dutch programme for lower vocational schools to promote healthful diet. Health Educ Res 2006;21 (5) 695- 704
James  JThomas  PCavan  DKerr  D Preventing childhood obesity by reducing consumption of carbonated drinks: cluster randomised controlled trial. BMJ 2004;328 (7450) 1237
Janssen  IKatzmarzyk  PTBoyce  WF  et al. Health Behaviour in School-Aged Children Obesity Working Group, Comparison of overweight and obesity prevalence in school-aged youth from 34 countries and their relationships with physical activity and dietary patterns. Obes Rev 2005;6 (2) 123- 132
Utter  JScragg  RNi Mhurchu  CSchaaf  D What effect do attempts to lose weight have on the observed relationship between nutrition behaviors and body mass index among adolescents? Int J Behav Nutr Phys Act 2007;450- 57