Evaluation of the Web-Based Computer-Tailored FATaintPHAT Intervention to Promote Energy Balance Among Adolescents: Results From a School Cluster Randomized Trial | Adolescent Medicine | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
Mar 2012

Evaluation of the Web-Based Computer-Tailored FATaintPHAT Intervention to Promote Energy Balance Among Adolescents: Results From a School Cluster Randomized Trial

Author Affiliations

Author Affiliations: Department of Public Health, Erasmus University Medical Center, Rotterdam, the Netherlands (Drs Ezendam and Oenema); Eindhoven Cancer Registry, Comprehensive Cancer Center South, Eindhoven, the Netherlands (Dr Ezendam); EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, the Netherlands (Dr Brug). Dr Ezendam is now with the Department of Medical Psychology and Neuropsychology, Tilburg University, Tilburg, the Netherlands.

Arch Pediatr Adolesc Med. 2012;166(3):248-255. doi:10.1001/archpediatrics.2011.204

Objective To evaluate the short- and long-term results of FATaintPHAT, a Web-based computer-tailored intervention aiming to increase physical activity, decrease sedentary behavior, and promote healthy eating to contribute to the prevention of excessive weight gain among adolescents.

Design Cluster randomized trial with an intervention group and a no-intervention control group.

Setting Twenty schools in the Netherlands.

Participants A total of 883 students (aged 12-13 years).

Intervention The FATaintPHAT (VETisnietVET in Dutch) Web-based computer-tailored intervention.

Outcome Measures Self-reported behaviors (diet, physical activity, sedentary behavior) and pedometer counts were measured at baseline and at 4-month and 2-year follow-up; body mass index (BMI), waist circumference, and fitness were measured at baseline and at 2-year follow-up. Descriptive and multilevel regression analyses were conducted among the total study population and among students not meeting behavioral recommendations at baseline (students at risk).

Results The complete case analyses showed that FATaintPHAT had no effect on BMI and waist circumference. However, the intervention was associated with lower odds (0.54) of drinking more than 400 mL of sugar-sweetened beverages per day and with lower snack intake (β = −0.81 snacks/d) and higher vegetable intake (β = 19.3 g/d) but also with a lower step count (β = −10 856 steps/wk) at 4-month follow-up. In addition, among students at risk, FATaintPHAT had a positive effect on fruit consumption (β = 0.39 g/d) at 4-month follow-up and on step count (β = 14 228 steps/wk) at 2-year follow-up but an inverse effect on the odds of sports participation (odds ratio, 0.45) at 4-month follow-up. No effects were found for sedentary behavior.

Conclusion The FATaintPHAT intervention was associated with positive short-term effects on diet but with no effects or unfavorable effects on physical activity and sedentary behavior.

Trial Registration Netherlands Trial Registry: ISRCTN15743786.

The high prevalence of overweight and obesity among adolescents is a major public health concern because of its association with various chronic diseases.1 Continuing preventive action is therefore needed. However, the number of effective obesity-prevention interventions for adolescents is limited.2 Computer tailoring has been recognized as a promising health communication technique to promote energy balance–related behaviors.3 Computer tailoring is a technique through which individualized feedback on risk behaviors, cognitions, and perceptions relevant to that behavior can be provided to larger numbers of people.4 Systematic reviews indicate that computer tailoring is likely to be more effective than generic health education for modifying dietary intake5-7 and possibly physical activity8 among adults. At present, only a few studies have evaluated the single effects of computer-tailored interventions among adolescents.9,10 The present evaluation of a stand-alone computer-tailored intervention can add to the evidence on the effectiveness of computer tailoring for adolescents to prevent excessive weight gain and improve dietary behavior, physical activity, and sedentary behavior.

Our study group11 developed an online school-based, computer-tailored intervention called FATaintPHAT (VETisnietVET in Dutch). The present study aims to evaluate the short- and long-term effects of this intervention among adolescents. The predictions were that in the intervention group vs the control group (1) anthropometric outcomes (body mass index [BMI], percentage overweight, waist circumference) and fitness would be more favorable at 2-year follow-up; and (2) the outcomes on the targeted behaviors (consumption of sugar-sweetened beverages, snacks, fruit, vegetables and fiber, screen time, and physical activity behaviors) would be more favorable at 4-month and 2-year follow-up.


Study design

A 2-group cluster randomized trial (n = 883; 20 schools) was conducted with assessments at baseline and 4-month (school year 2006-2007) and 2-year follow-up (school year 2008-2009). Schools were randomized into an intervention group or a no-intervention control group after stratification according to educational level (vocational or preuniversity training) using a random-number generator. The study was approved by the medical ethics committee of the Erasmus Medical Center and registered in the Netherlands Trial Registry (ISRCTN15743786). The methods and intervention have been described in detail previously.11 The study was conducted in collaboration with the Municipal Health Services in the Rotterdam area.

Participants and recruitment

Adolescents aged 12 to 13 years were recruited in a 2-step procedure. First, 88 schools for secondary education in the Rotterdam area were invited to participate. Twenty-three schools were eligible and willing to participate (Figure). Second, adolescents from 1 to 5 first-year classes in each school (depending on the number of first-year classes in the school, maximum of 5) were invited to participate. Students received information and an informed consent form for themselves and their parents for active consent. The completed consent forms were returned through the schools. Three schools withdrew from the study after randomization and before the baseline measurement because they found the informed consent procedure too burdensome. Of the 1494 students, 1156 returned their forms (77%), and 883 students agreed to participate in the study (59%). Students in the intervention group were more likely to participate (33% vs 26%), even though allocation was concealed until the start of the intervention.

Figure. Cluster and participant flow for the FATaintPHAT study.

Figure. Cluster and participant flow for the FATaintPHAT11 study.

The intervention

The objective of the computer-tailored intervention is to help prevent excessive weight gain among adolescents aged 12 to 13 years by improving dietary behaviors (reducing the consumption of sugar-sweetened beverages and high-energy snacks and increasing the intake of fruit, vegetables, and whole-wheat bread), reducing sedentary behavior (reducing screen time), and increasing physical activity (increasing active transport to school, leisure time activities, and sports).

Separate modules (n = 8) addressed the issues of weight management and energy balance–related behaviors. Each module consisted of information about the behavior-health link, an assessment of behavior and determinants, individually tailored feedback on behavior and determinants, and an option to formulate an implementation intention to prompt specific goal setting and action planning. The feedback provided included several elements: behavioral feedback (comparing the student's behavior with guidelines for that behavior [normative feedback] and with behavior of peers [comparative feedback]), prompts for intention formation, decisional balance information to change attitudes, prompts for barrier identification, instructions on how to perform and/or change a behavior to improve self-efficacy, and suggestions on how to organize social support. We used a multiple-theory approach, including the Theory of Planned Behavior,12 the Precaution Adoption Process Model,13 and implementation intentions,14 to inform the intervention. The intervention was accessible through the Internet. The teachers were asked to allocate 15 minutes for each of 8 lessons over 10 weeks to work with the program according to a teacher manual.


The outcome measures were anthropometrics and fitness (at 2-year follow-up), and energy balance–related behaviors (at 4-month and 2-year follow-up). Data on the outcome measures were collected through anthropometric measurements, questionnaires, pedometers, and shuttle-run tests. Fitness was measured with a shuttle-run test administered by the physical activity teacher according to a standard protocol.15 After baseline assessments, the intervention was implemented by the teachers. The control school implemented the regular curriculum.



According to a protocol, a research assistant measured height (average of 2 measurements without shoes using a Seca 225 mobile height rod [Seca, Hanover, Maryland] with an accuracy of 0.1 cm, and the average was calculated) and weight of the students who wore shorts and t-shirt or underwear (using a Seca 888 class III calibrated electronic digital floor scale with an accuracy of 0.2 kg).11 The BMI was calculated as weight in kilograms divided by height in meters squared. Cutoff points were based on the International Obesity Task Force guidelines.16 Waist circumference was measured twice. Third and fourth measurements were taken when the difference between the first 2 measurements was more than 1.0 cm. The average of the last 2 measurements was calculated. Circumference was measured at the waist equidistant from the lowest rib and the hip bone at the end of an expiration.11

Self-Reported Behaviors

Electronic self-administered questionnaires were used to assess behaviors (eTable). Questionnaires were completed within 1 school hour under the supervision of a research assistant. Dietary intake was assessed using a food frequency questionnaire assessing the frequency and quantity of sugar-sweetened beverage consumption in the past week and a self-administered 24-hour recall for snacks and fruit and vegetable consumption.17-19 Physical activity and sedentary behavior were assessed using the Flemish validated questionnaire.20 This questionnaire assessed sports during leisure time, active transportation to school, television viewing, and computer use during leisure time in the past 7 days by asking about the frequency and duration of the activities. In addition, the number of days spent in moderate or vigorous physical activity for at least 60 minutes was assessed. An additional assessment of physical activity was obtained with pedometers (Digiwalker SW200; YAMAX USA Inc, San Antonio, Texas)21 that were worn by a random subsample of 5 students per class for 7 consecutive days after the questionnaire assessment.


Questions on demographic characteristics included sex, age, educational level, country of birth, and parents' country of birth. Ethnicity was defined according to standard procedures of Statistics Netherlands, the Hague, as either Western (both parents born in Europe, North America, Oceania, Indonesia, or Japan) or non-Western (at least 1 parent born elsewhere).

Statistical analysis

Logistic regression analyses were used to identify whether there was selective dropout. Dropout (yes/no) was used as the dependent variable, and sex, education, ethnicity, intervention, BMI at baseline, and compliance with recommendations for each behavior were independent variables.

Baseline group differences were tested using the nonparametric Kolmogorov-Smirnov test or the χ2 test. Multilevel linear and logistic regression analyses were used to establish intervention effects. Each outcome measure was regressed on group, (intervention [1] vs control [0]) and baseline value of the outcome measure. Sex (girls [1] vs boys [0]), education (preuniversity [1] vs vocational [0]), and ethnicity (non-Western [1] vs Western [0]) were included as potential confounders. Intraclass coefficients for the continuous outcomes were calculated as the between-school variance divided by the total variance. Separate analyses were run for the short- and the long-term results. The analyses were conducted for the total study population and then repeated for the students not meeting behavioral recommendations at baseline (at-risk students) because students engaging in risk behavior were expected to benefit more from the intervention. For anthropometric outcomes the risk group included normal, overweight, and obese adolescents because these students received feedback to prevent excessive weight gain.

Complete case analyses and intention-to-treat analyses were performed using baseline observation carried forward (BOCF) and last observation carried forward (LOCF) procedures.22 Multilevel regression analyses were performed in MLwiN 2.02 (University of Bristol, Bristol, England), other analyses in SPSS 15.0 (IBM Corporation, Armonk, New York). The significance level was set at 0.05, and tests were 2 sided.

Loss to follow-up

In the intervention group, 15% of the students were lost to follow-up (left school or had incomplete data at baseline and follow-up); in the control group, 12% were lost (Figure). Loss to follow-up did not differ according to study condition, educational level, ethnicity, or sex.


Student characteristics

The intervention group consisted of more vocational schools and vocational-level students, more boys, and more non-Western students than the control group. At baseline, more students in the intervention group were active for less than 60 minutes per day, and more students in the intervention group engaged less than 2 hours in sedentary behavior (Table 1).

Table 1. School and Student Characteristics at Baseline for the Intervention and Control Groupsa
Table 1. School and Student Characteristics at Baseline for the Intervention and Control Groupsa
Table 1. School and Student Characteristics at Baseline for the Intervention and Control Groupsa

Intervention effects

Table 2 lists the mean values for the outcome measures at baseline and 4-month and 2-year follow-up for the intervention and control groups, the total sample, and for the students at risk. The regression analyses (Table 3) showed no intervention effects on BMI, waist circumference, or percentage of students being overweight or obese in the total sample and among normal-weight, overweight, and/or obese students.

Table 2. Outcome Measuresa for the Intervention and Control Groups and for the Total Sample and the Students at Riskb
Table 2. Outcome Measuresa for the Intervention and Control Groups and for the Total Sample and the Students at Riskb
Table 2. Outcome Measuresa for the Intervention and Control Groups and for the Total Sample and the Students at Riskb
Table 3. Intervention Effects Among All Students and Students at Riska
Table 3. Intervention Effects Among All Students and Students at Riska
Table 3. Intervention Effects Among All Students and Students at Riska

At 4-month follow-up, the intervention group was less likely to report drinking more than 400 mL of sugar-sweetened beverages per day compared with the control group in the total sample but not in the risk group. Mean self-reported snack consumption was lower in the intervention group than in the control group at 4-month follow-up. The difference at 2-year follow-up was not significant. Among the students at risk, those in the intervention group reported eating more pieces of fruit than those in the control group at 4-month follow-up. For vegetable intake, the intervention groups reported consuming more grams per day at 4-month follow-up than the control groups in both the total sample and among the students at risk. There were no differences in self-reported consumption of whole-wheat bread between intervention and control groups.

Among the students at risk, the intervention group was less likely to report participating in sports at 4-month follow-up than the control group. In the intervention group, fewer steps per week were recorded at 4-month follow-up, but more steps at 2-year follow-up (only in the at-risk group), compared with the control group.

Intention-to-treat analyses

Imputation of the 4-month and 2-year follow-up outcomes with BOCF and LOCF procedures resulted in only few differences compared with the complete case analyses. In the at-risk group, imputation led to a significant effect for fruit intake at 2-year follow-up (BOCF β = 0.26; 95% confidence interval [CI], 0.01-0.51), while the group differences for step count at 4-month follow-up (BOCF β = −7413; 95% CI, −15272 to 487) and 2-year follow-up (BOCF β = 2838; 95% CI, −3563 to 9272) (LOCF β = −5017; 95% CI, −16074 to 6098) were nonsignificant.


Main results

The results indicate that the FATaintPHAT intervention11 had no effects on anthropometric outcomes. We found favorable effects on self-reported obesity-related dietary behaviors at 4-month follow-up but not at 2-year follow-up. No effects for sedentary behaviors and some unfavorable effects for physical activity behaviors (sports participation and step counts) were found at 4-month follow-up.

Interpretation of the results

Anthropometrics and fitness were only found at 2-year follow-up, while behavioral effects were only found at 4-month follow-up. The intervention thus appears to be not strong enough to have sustained effects, possibly owing to its short duration (8 sessions of 15 minutes each within 10 weeks). Furthermore, the 4-month changes in behavior were not all in the desired direction, making changes in body composition and fitness less likely.

However, it is promising that we found positive effects on dietary behaviors at 4-month follow-up. These findings compare positively with the effects that have been found in previous comparable studies, both in terms of number of effects and the sizes of the effects. Haerens et al23,24 found a decrease in fat intake at 1-year and 2-year follow-up but not in sugar-sweetened beverage intake or fruit and water consumption. The 1-school-year healthy eating promotion intervention combined environmental changes, computer-tailored feedback, and a parent component. Singh et al25 did not find effects on snack consumption, but they did find a decrease in sugar-sweetened beverage consumption after 8 and 12 months—about a 250-mL difference between groups. They did not find an effect after 20 months.

The 1-school-year intervention to prevent excessive weight gain among adolescents consisted of educational (11 lessons) and environmental components. Martens et al26 did not find a significant effect for fruit intake or breakfast consumption, but they did find a decrease in snack consumption (0.6 pieces) at 3-month follow-up. The dietary intervention consisted of educational (8 lessons) and parental components. Knai et al27 found in their review of studies on fruit and vegetable consumption that of the 4 identified studies, 2 found no significant effects after 3 years; 1 found an increase of 0.3 servings per day among girls after 2 years; and 1 found an increase of 0.9 servings per day at interim evaluation but no effect at 2-year follow-up. Muth et al28 found an increase of 0.9 servings of fruit and vegetables per day directly after the intervention period. It is noteworthy that these other interventions were all high-intensity programs made up of more components than only tailored feedback.

The adverse effects that we found on some physical activity indicators were unexpected. These adverse effects were mainly caused by a larger increase in sports participation and step counts in the control group compared with the intervention group. The observed increase in physical activity in the control group might be owing to seasonal influences over the 4-month period (fall/winter–spring/summer),29 an increase that we at least would have expected to see in the intervention group as well. The findings may indicate that the intervention inhibited the adolescents from increasing their levels of physical activity. This inhibition effect might be the result of unexpected reactions to the feedback messages. Currently, there is limited evidence for the way adolescents respond to personalized feedback.30 More insight in the processing of feedback messages is therefore needed. In addition, we cannot rule out that students might have compensated for their improved dietary behaviors by lowering their physical activity, although additional analyses of our data do not reveal such compensating associations.

Imputation of step count led to nonsignificant effects. As might be expected, BOCF imputation led to smaller effect sizes,22 while LOCF imputation led to effects that were more comparable with outcomes at second follow-up, since any effects on first follow-up are carried forward. Thus, even though we found negative effects for physical activity outcomes, these effects were not consistent. What we did find consistently was no effect on physical activity. The results of our study are in line with those reported in a review by van Sluijs et al,31 which showed that education-only school-based interventions to increase physical activity among adolescents did not result in an increase in physical activity, whereas multicomponent interventions (including, eg, family, community, and/or environmental changes) did have a positive effect. Therefore, to promote physical activity in the school setting, computer-tailored programs might need to be accompanied by family, community, and/or environmental interventions.

Effects of computer-tailored interventions for dietary, physical activity, and sedentary behavior among adolescents have not been studied extensively. Haerens et al9,10 showed that a computer-tailored intervention for increasing physical activity among adolescents was more effective than a no-intervention control group9 but not more effective than generic information.10 De Bourdeaudhuij et al32 concluded, based on a systematic review including multicomponent interventions, that computer-tailored personalized education in the classroom led to better results than a generic classroom curriculum in school-based nutrition and physical activity interventions. Our study indicates adverse effects on physical activity but a favorable impact on dietary behaviors among adolescents compared with a no-intervention control group. This is largely in line with the conclusions of recent reviews that found convincing evidence for the effectiveness of computer-tailored interventions on diet but inconclusive results for physical activity among adults.5,6,8

Strengths and limitations

Important strengths of this study are its randomized design, large size, objective measures for anthropometry and physical activity, and the novelty of addressing the effectiveness of a stand-alone computer-tailored intervention. Limitations to the study are the use of self-reported measures, which might have resulted in less reliable outcomes and might have weakened the effects found in this study. In addition, more students from the intervention group were lost to follow-up, possibly owing to the overrepresentation of vocational schools in this group. Vocational school students choose their specialty after 2 years and then often change to a different school location. Since the pedometers were used in a subsample of students and there was some loss to follow-up, the number of adolescents who used the pedometer at the second follow-up was small compared with number at baseline and at first follow-up. Therefore, results should be interpreted with care. Another limitation is that we did not assess pubertal status to allow correction of the anthropometric data.

In conclusion, our study shows that the computer-tailored intervention FATaintPHAT was not effective in modifying anthropometric outcome measures but that it can have a positive effect on dietary behaviors among adolescents at short-term follow-up. Expanding the intervention with additional components or booster sessions might improve the effectiveness in the short and long term. The results of our study seem to be in line with the findings of recent reviews that have indicated that classroom-based educational programs (not only computer tailored) seem to be effective for promoting dietary intake but not for promoting physical activity. Successful promotion of physical activity might require (additional) environmental changes.

Back to top
Article Information

Correspondence: Nicole P. M. Ezendam, PhD, Department of Medical Psychology and Neuropsychology, Tilburg University, PO Box 90153, 5000 LE Tilburg, the Netherlands (n.p.m.ezendam@uvt.nl).

Accepted for Publication: August 31, 2011.

Published Online: November 7, 2011. doi:10.1001/archpediatrics.2011.204

Author Contributions: Dr Ezendam 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 analyses. Study concept and design: Ezendam, Brug, and Oenema. Acquisition of data: Ezendam. Analysis and interpretation of data: Ezendam, Brug, and Oenema. Drafting of the manuscript: Ezendam, Brug, and Oenema. Critical revision of the manuscript for important intellectual content: Ezendam, Brug, and Oenema. Statistical analysis: Ezendam. Obtained funding: Brug. Administrative, technical, and material support: Ezendam and Oenema. Study supervision: Brug and Oenema.

Financial Disclosure: None reported.

Funding/Support: This study was funded by grant 62200020 from ZonMw, the Netherlands Organization for Health Care Research and Development.

Role of the Sponsors: The funding organization was not involved in any aspect of the analyses or in the preparation of the manuscript.

Reilly JJ. Obesity in childhood and adolescence: evidence based clinical and public health perspectives.  Postgrad Med J. 2006;82(969):429-43716822919PubMedGoogle ScholarCrossref
Gonzalez-Suarez C, Worley A, Grimmer-Somers K, Dones V. School-based interventions on childhood obesity: a meta-analysis.  Am J Prev Med. 2009;37(5):418-42719840696PubMedGoogle ScholarCrossref
Rimer BK, Kreuter MW. Advancing tailored health communication: a persuasion and message effects perspective.  J Commun. 2006;56(s1):S184-S201Google ScholarCrossref
Brug J, Oenema A, Campbell M. Past, present, and future of computer-tailored nutrition education.  Am J Clin Nutr. 2003;77(4):(suppl)  1028S-1034S12663313PubMedGoogle Scholar
Kroeze W, Werkman A, Brug J. A systematic review of randomized trials on the effectiveness of computer-tailored education on physical activity and dietary behaviors.  Ann Behav Med. 2006;31(3):205-22316700634PubMedGoogle ScholarCrossref
Neville LM, O’Hara B, Milat AJ. Computer-tailored dietary behaviour change interventions: a systematic review.  Health Educ Res. 2009;24(4):699-72019286893PubMedGoogle ScholarCrossref
Noar SM, Benac CN, Harris MS. Does tailoring matter? meta-analytic review of tailored print health behavior change interventions.  Psychol Bull. 2007;133(4):673-69317592961PubMedGoogle ScholarCrossref
Neville LM, O’Hara B, Milat A. Computer-tailored physical activity behavior change interventions targeting adults: a systematic review.  Int J Behav Nutr Phys Act. 2009;6:3019490649PubMedGoogle ScholarCrossref
Haerens L, Deforche B, Vandelanotte C, Maes L, De Bourdeaudhuij I. Acceptability, feasibility and effectiveness of a computer-tailored physical activity intervention in adolescents.  Patient Educ Couns. 2007;66(3):303-31017298873PubMedGoogle ScholarCrossref
Haerens L, Maes L, Vereecken C, De Henauw S, Moreno L, De Bourdeaudhuij I. Effectiveness of a computer tailored physical activity intervention in adolescents compared to a generic advice.  Patient Educ Couns. 2009;77(1):38-4119406608PubMedGoogle ScholarCrossref
Ezendam NP, Oenema A, van de Looij-Jansen PM, Brug J. Design and evaluation protocol of “FATaintPHAT”, a computer-tailored intervention to prevent excessive weight gain in adolescents.  BMC Public Health. 2007;7:32417997834PubMedGoogle ScholarCrossref
Ajzen I. The theory of planned behavior.  Organ Behav Hum Decis Process. 1991;50:179-211Google ScholarCrossref
Weinstein ND, Rothman AJ, Sutton SR. Stage theories of health behavior: conceptual and methodological issues.  Health Psychol. 1998;17(3):290-2999619480PubMedGoogle ScholarCrossref
Gollwitzer PM. Implementation intentions: strong effects of simple plans.  Am Psychol. 1999;54(7):493-503Google ScholarCrossref
Léger LA, Mercier D, Gadoury C, Lambert J. The multistage 20 metre shuttle run test for aerobic fitness.  J Sports Sci. 1988;6(2):93-1013184250PubMedGoogle ScholarCrossref
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey.  BMJ. 2000;320(7244):1240-124310797032PubMedGoogle ScholarCrossref
van Assema P, Brug J, Ronda G, Steenhuis 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):377-39011906579PubMedGoogle ScholarCrossref
Van Assema P, Brug J, Ronda G, Steenhuis I, Oenema A. A short dutch questionnaire to measure fruit and vegetable intake: relative validity among adults and adolescents.  Nutr Health. 2002;16(2):85-10612102370PubMedGoogle ScholarCrossref
Bogers RP, Van Assema P, Kester AD, Westerterp KR, Dagnelie PC. Reproducibility, validity, and responsiveness to change of a short questionnaire for measuring fruit and vegetable intake.  Am J Epidemiol. 2004;159(9):900-90915105183PubMedGoogle ScholarCrossref
Philippaerts RM, Matton L, Wijndaele K, Balduck AL, De Bourdeaudhuij I, Lefevre J. Validity of a physical activity computer questionnaire in 12- to 18-year-old boys and girls.  Int J Sports Med. 2006;27(2):131-13616475059PubMedGoogle ScholarCrossref
Schneider PL, Crouter SE, Bassett DR. Pedometer measures of free-living physical activity: comparison of 13 models.  Med Sci Sports Exerc. 2004;36(2):331-33514767259PubMedGoogle ScholarCrossref
Altman DG. Missing outcomes in randomized trials: addressing the dilemma.  Open Med. 2009;3(2):e51-e5319946393PubMedGoogle Scholar
Haerens L, Deforche B, Maes L, Cardon G, Stevens V, De Bourdeaudhuij I. Evaluation of a 2-year physical activity and healthy eating intervention in middle school children.  Health Educ Res. 2006;21(6):911-92117032704PubMedGoogle ScholarCrossref
Haerens L, De Bourdeaudhuij I, Maes L, Vereecken C, Brug J, Deforche B. The effects of a middle-school healthy eating intervention on adolescents' fat and fruit intake and soft drinks consumption.  Public Health Nutr. 2007;10(5):443-44917411463PubMedGoogle ScholarCrossref
Singh AS, Chin A Paw MJ, Brug J, van Mechelen W. Dutch obesity intervention in teenagers: effectiveness of a school-based program on body composition and behavior.  Arch Pediatr Adolesc Med. 2009;163(4):309-31719349559PubMedGoogle ScholarCrossref
Martens MK, Van Assema P, Paulussen TG, Van Breukelen G, Brug J. Krachtvoer: effect evaluation of a Dutch healthful diet promotion curriculum for lower vocational schools.  Public Health Nutr. 2008;11(3):271-27817605839PubMedGoogle ScholarCrossref
Knai C, Pomerleau J, Lock K, McKee M. Getting children to eat more fruit and vegetables: a systematic review.  Prev Med. 2006;42(2):85-9516375956PubMedGoogle ScholarCrossref
Muth ND, Chatterjee A, Williams D, Cross A, Flower K. Making an IMPACT: effect of a school-based pilot intervention.  N C Med J. 2008;69(6):432-44019256179PubMedGoogle Scholar
Tucker P, Gilliland J. The effect of season and weather on physical activity: a systematic review.  Public Health. 2007;121(12):909-92217920646PubMedGoogle ScholarCrossref
Hawkins RP, Kreuter M, Resnicow K, Fishbein M, Dijkstra A. Understanding tailoring in communicating about health.  Health Educ Res. 2008;23(3):454-46618349033PubMedGoogle ScholarCrossref
van Sluijs EM, McMinn AM, Griffin SJ. Effectiveness of interventions to promote physical activity in children and adolescents: systematic review of controlled trials.  BMJ. 2007;335(7622):70317884863PubMedGoogle ScholarCrossref
De Bourdeaudhuij I, Van Cauwenberghe E, Spittaels H,  et al.  School-based interventions promoting both physical activity and healthy eating in Europe: a systematic review within the HOPE project. . 2011;12(3):205-216