Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The effect of an extra piece of fruit or vegetables at school on weight status in two generations - 14 years follow-up of the Fruit and Vegetables Makes the Marks study

  • Tonje Holte Stea ,

    Roles Conceptualization, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing

    Affiliation University of Agder, Department of Public Health, Sport and Nutrition, Kristiansand, Norway

  • Eline Tønnesson Tveter,

    Roles Data curation, Formal analysis, Writing – original draft

    Affiliation University of Agder, Department of Public Health, Sport and Nutrition, Kristiansand, Norway

  • Saskia J. te Velde,

    Roles Formal analysis, Supervision, Writing – review & editing

    Affiliation University of Agder, Department of Public Health, Sport and Nutrition, Kristiansand, Norway

  • Frøydis Nordgård Vik,

    Roles Conceptualization, Methodology, Supervision, Writing – review & editing

    Affiliation University of Agder, Department of Public Health, Sport and Nutrition, Kristiansand, Norway

  • Knut-Inge Klepp,

    Roles Conceptualization, Funding acquisition, Methodology, Resources, Writing – review & editing

    Affiliation University of Oslo, Department of Nutrition, Oslo, Norway

  • Elling Bere

    Roles Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing

    Affiliation University of Agder, Department of Public Health, Sport and Nutrition, Kristiansand, Norway

The effect of an extra piece of fruit or vegetables at school on weight status in two generations - 14 years follow-up of the Fruit and Vegetables Makes the Marks study

  • Tonje Holte Stea, 
  • Eline Tønnesson Tveter, 
  • Saskia J. te Velde, 
  • Frøydis Nordgård Vik, 
  • Knut-Inge Klepp, 
  • Elling Bere



The obesity epidemic presents a major public health challenge, and a poor diet quality has been identified as one of the most important contributing factors. Whereas a sufficient fruit and vegetable consumption has been associated with several positive health outcomes, the long-term effect on overweight and obesity is unclear. Thus, the aims of this study were to investigate if one year with free school fruit had any effect on weight status 14 years later, and if it affected the birth weight of the participants’ children.


In 2001, 10 -12-year old Norwegian children, received one year of free school fruit in the intervention study “Fruits and Vegetables Make the Marks” (FVMM) and in 2016, a total of 1081 participants of 2049 eligible responded to a follow-up survey. Multilevel logistic regression was used to investigate if one year of free school fruit was associated with weight status and with birthweight status of the offspring. The analyses were adjusted for gender, educational level, and the offspring analysis also for parents’ weight status, and the nested design (child/parent).


The odds ratios of being overweight (OR: 0.93, 95% CI: 0.70, 1.24) or having a child with high or low birth weight (OR: 0.52, 95% CI: 0.21, 1.30) in the intervention group compared to the control group were not statistically significant, 14 years after the intervention period.


One year of free school fruit did not have an effect on weight status on the participants or birth weight of their offspring, 14 years after the intervention period. Although, results from the present study contribute to fill the knowledge gaps concerning long-term effects of public health efforts on weight status, more follow-up studies with larger samples are warranted.


Overweight and obesity are causing challenging public health problems worldwide, and in most countries, rates of overweight and obesity are continuing to rise [13]. Overweight and obesity are associated with several adult chronic illnesses, such as coronary heart disease, type 2 diabetes, and some types of cancer [3,4]. Studies have also shown a strong dose-response association between pre-pregnancy BMI, offsprings`birth weight and childhood BMI [58], and childhood overweight and obesity has shown to increase the risk of adult obesity and chronic disease [3]. Weight-control interventions among overweight and obese pregnant women have shown limited effects on birth weight of the child [911], and it has been concluded that lifestyle changes before pregnancy seem more promising than during pregnancy [11].

Unhealthy dietary habits are closely related to risk of overweight and obesity [3]. Intake of fruit and vegetables (FV), which are low in energy and rich in fibre, has shown to reduce the risk of adiposity [12] and has also been linked to reduced risks for cardiovascular diseases and several cancers [1316]. Therefore, the importance of promoting healthy lifestyle habits, including increased consumption of FV, from an early age using effective surveillance, prevention, and treatment programs has been underlined [17,18].

Systematic review studies have reported that school-based intervention programmes tend to be effective in increasing pupils’ FV intake [19,20], and that school based intervention programmes may effectively contribute in preventing overweight and obesity [21]. Although most of the previously published school-based intervention studies have only reported short-term effects on weight status, few studies have investigated long-term effects [20,21]. Two systematic review studies reported that only eight out of 48 included studies investigated follow-up effects more than a year after the intervention period [19,20] and only one of the reviewed studies, a two-year follow-up study, showed a positive impact on FV consumption and BMI [19]. In addition, the Fruit and Vegetables Make the Marks (FVMM) study, in which Norwegian school-children received one piece of fruit or vegetables for free for one school year (2001/2002), reported a lower prevalence of overweight in the intervention group compared to the control group (15% vs. 25% overweight) seven years after the FVMM intervention period [22]. However, the association was not significant in the final model, and the validity of this finding was hampered by several limitations, including a very low participation rate.

Due to limited scientific evidence from school-based intervention studies and other studies, and a limited number of long-term longitudinal follow up studies, we are not able to conclude that FV consumption has a protective effect on the risk of developing obesity [19,20,23]. Thus, long-term follow-up of school-based intervention studies investigating the association between dietary patterns and weight status during transition from childhood to adolescence, and further into adulthood has been warranted [24,25]. Our study aims to fill this knowledge gap by evaluating the effect of the 15 years follow-up of providing free school fruit for one year in elementary schools on (1) the weight status of the participants and (2) the birth weight of their children.


Study design

In 2001, 48 schools from Hedmark and Telemark counties (24 schools in each county) were randomly selected and invited to participate in the research project FVMM. Nineteen schools from each county agreed to participate [26]. Nine schools within one of the counties were randomly selected as intervention schools and participated in the Norwegian School Fruit Program for free during the school year 2001/2002 [26]. The free subscription program started in October 2001 and lasted throughout the school year (i.e. until June 2002). The pupils in the intervention schools received a piece of fruit or a carrot per day for free each school day for one year, usually in connection with their lunch meal [26]. Apples, pears, bananas, oranges, clementines, kiwis, carrots and nectarines were the most frequently FV given. The remaining 29 schools served as control schools [27]. A more detailed description of the study design has previously been published [28]. The free fruit was given in addition to the normal Norwegian lunch, typically consisting of sandwiches brought from home. There was no focus on decrease of other foods items within the free fruit program. However, we have previously observed a decrease in consumption of unhealthy snacks following the intervention participants compared to the controls, both in the current cohort [29], and also when evaluating the later nationwide implementation of free school fruit lasting from 2007 to 2014 [30].

Some pupils also received an educational program during the intervention year (2001/2002), however a previous publication showed no effect on fruit and vegetable intake of this educational program [31,32]. Therefore, in the current study we only distinguish between those who received free fruit and those who did not.


In September 2001, 6th and 7th graders (age 10–12) completed the baseline survey (before the fruit program started) [26]. Follow-up surveys were conducted in May 2002, May 2003, May 2005 and September 2009 [27], and data collection for a fifth follow-up survey started in January 2016 and ended in November 2016. In the present study, the data collections from 2001, 2005 and 2016 are used. Data from the 2002—and 2003 surveys are not included in this study because the surveys did not include questions regarding weight status. In addition, the 2003 survey only included half of the study sample [28]. Data from the 2009 survey is not included in the present study due to the low participation rate and the fact that results from this follow-up survey have been presented earlier [22,27].

The surveys conducted in the period between 2001 and 2005 were completed by the pupils in the classroom in the presence of a trained project worker [33]. One school lesson (45 min) was used to complete the questionnaires. In 2016, eligible participants were contacted through Facebook or by phone, and were asked to conduct the survey for the fifth follow-up. This strategy was chosen as at that time the participants did not attend schools anymore and Facebook seemed a promising tool to find people.

The FVMM study was conducted according to the guidelines laid down in the Declaration of Helsinki. Ethical approval and research clearance were obtained from The National Committees for Research Ethics in Norway (REK) (file number S-01076) and from The Norwegian Social Science Data Services (NSD) (file number 12395). Written informed consent was obtained from all subjects, i.e. the participants themselves and their parents.

Study sample

Of 2287 eligible schoolchildren, a total of 1950 pupils (85%) completed the survey at baseline. Fifty-nine (3%) children–parents refused to participate, one class (27 children, 1%) was not able to carry out the survey, and 251 (11%) children did not attend this specific school lesson and were not re-contacted for the baseline survey [26]. Those that initially refused to participate (n = 59), and those who resigned throughout the follow-ups (n = 179) were not contacted for the 2016 survey, thus 2049 were eligible for the 2016 survey, and a total of 1081 participants (53%) did participate. The average age in 2016 was 26.5 years. Of the study sample, a total of 190 participants (17.6% of responders, 9.3% of eligible sample) reported to have one or more children. In total, 270 children were included.


Self-reported perceived weight status was assessed by the following question at all surveys: “Are you on (slimming) diet”, with the following response alternatives: “No, my weight is OK”, “No, but I should be on a (slimming) diets”, and “yes”. The two last alternatives were seen to indicate that the responders perceive themselves as too heavy/overweight. A new variable was calculated by combining these two to: 1 = I am too heavy, 0 = my weight is OK [22].

Self-reported height and weight were included in the 2005 and 2016 surveys, but not in the baseline survey conducted in 2001. The reason for not including height and weight in the original survey, was that the main aim of the original study was to evaluate the intervention on fruit and vegetable intake and its behavioral determinants and not changes in BMI [22]. In the present study, BMI was calculated based on self-reported height and weight. For the 2005 survey, age and sex-specific cut off points according to the International Obesity Task Force (IOTF) definitions [34] were used to categorize the adolescents for being overweight or obese (≥25 kg/m2). Furthermore, World Health Organization’s (WHO) definition of adult overweight and obesity (>18 years) was used for the classification of the participants in the 2009 and 2016 survey (≥25 kg/m2) [35]. For the present analyses, the variable was dichotomized into overweight/obese (≥25 kg/m2) vs. not overweight/obese (<25 kg/m2).

In the 2016 survey, questions regarding the participants’ children were included. The first question was formulated “Do you have children”, with the following alternatives: “No” and “Yes”. In total, it was possible to register four children. If the responder answered “Yes”, two new questions were asked: “In what year is it born?” and “What was the birth weight of the new-born (in gram)?”. Birth weight variable was trichotomized into three categories: high birth weight (HBW) (≥4000 gram), low birth weight (LBW) (<2500 gram), and normal birth weight (2500–3999 gram). In the regression analysis, this variable was dichotomized into ‘normal birth weight’ vs. ‘low birth weight’/’high birth weight’ to assess the odds of not being in recommended weight category.

In the surveys, the participants reported their own gender and age. In the 2016 survey, the participants recorded their own educational level as an indicator of socioeconomic status. The educational level variables were dichotomized into lower education (no university or college) or higher education (university or college).

Statistical analyses

Descriptive statistics were used to characterize the current study samples at different time points. Comparisons were made between those participating in the intervention group and the control group. For these comparisons χ2 statistics were used for the categorical variables. Independent Sample t-test was used for the continuous variables. All the continuous variables used in the crude analyses were considered as normally distributed.

Logistic (multilevel) regression analyses was used to assess the association of the intervention condition with weight status (0 = normal weight, 1 = overweight) and with birth weight (0 = normal birth weight, 1 = low or high birth weight). Assumptions regarding multicollinearity were checked by calculating correlation coefficients between all co-variates and these did not indicate multicollinearity problems. Furthermore, the assumption of independence of errors was violated as participants provided more than one assessment. This violation was solved by using multilevel analyses that takes into account the dependency between observations within the same person [36].

Gender and the participant’s education level were included in these analyses as potential confounders. For the analyses with weight status of parents as an outcome, two levels were defined: (1) participants and (2) schools. However, based on changes in the -2 Log Likelihood (-2LL) relative to the changes in degrees of freedom (df), adding school as a level did not significantly improve the model (Δ -2LL = 0.10, Δ df = 1, p = 0.103). Aiming for the most parsimonious model the school level was therefore omitted.

For the analysis with birth weight as an outcome, this model was defined with three levels (1) children, (2), participants (parents) and (3) schools. However, adding school as a level in the model did not improve the model compared to a model not adjusting for the nested design (Δ -2LL = 0, Δ df = 1, p = 1.000). Therefore, we decided not to adjust for the school level and tested the model that was adjusted for the parent level only, which improved the model slightly (-2LL = -139.88, df = 5 vs. -2LL = -138.06, df = 6, Δ = 1.82, df = 1). Despite the non-significant improvement of the model after adding the parental level, we decided to keep this for theoretical reasons. Children from the same parents resemble each other more regarding their birth weight than they resemble other children. Therefore, the model was re-defined with only two levels, (1) children and (2) participants (parents). Besides the study participants’ intervention condition (intervention vs. control), the following variables were included in the models as potential confounders: their gender, their educational level and for the model on birthweight also the participants (i.e. parents of the newborns) BMI (dichotomized into overweight vs. not overweight).

All the descriptive analyses were conducted using IBM SPSS Statistics 24.0. R software, version 3.2.2, was used to run the multilevel binary logistic regression models (lme4 package). All tests were two-sided with p-values <0.05 considered statistically significant.


In the present study (n = 1081), 322 participants were in the intervention group and 759 in the control group. No significant differences were observed between those groups regarding gender. There was a significantly higher prevalence of participants with high educational level in the intervention group compared to the control group (p = 0.035) (Table 1). The total number of children in the present study was 270; 67 of them belonged to a parent from the intervention group, 203 children to a parent from the control group. The first offspring was born in 2000, but most participants had their first child after 2011 (75% of the first children was born in the years from 2012 till 2016). No significant differences were observed between those groups regarding gender, educational level or weight status of their parent (Table 1). Data for characterization of participants at different time points is provided in S1 Table.

Table 1. Characteristics of the present participants (n = 1081) including the offspring of the participants (n = 270).

The prevalence of overweight increased from 11% in 2005 to 34% in 2016 for the intervention group. In the control group, this prevalence increased from 11% in 2005 to 36% in 2016 (Table 2). These differences between the intervention and the control group were not statistically significant. The prevalence of normal birth weight was 80% among children with parents in the intervention group, compared to 72% among children with parents in the control group (p = 0.260). Furthermore, the prevalence of LBW was 3% in the intervention group, compared to 8% in the control group (p = 0.299), and the prevalence of HBW was 17% in the intervention group, and 20% in the control group (p = 0.678) (Table 2). Data on birth weight of offspring and parental weight status 14 years after participation in the intervention group or the control group is provided in S2 Table.

Table 2. Observed means (SD) and prevalence at all time points for the participants (n = 1081) and for the offspring (n = 270).

Multilevel logistic regression analysis (Table 3), showed that the odds of being overweight was 0.91 (95% CI: 0.69, 1.21) for the intervention group compared to the control group (Table 3, model 1). After adjusting for gender and educational level, the odds of being overweight was 0.93 (95% CI: 0.70, 1.24) for the intervention group, compared to the control group (Table 3, model 3).

Table 3. Odds Ratios (OR) and 95% Confidence Intervals (CI) for being overweight in 2016 for the participants or having children with low or high birth weight for the intervention group compared to the control group.

The odds of having a child with high or low birth weight was 0.55 (95% CI: 0.22, 1.38) in the intervention group compared to the control group (Table 3, model 1). Further, the odds of having a child with high or low birth weight was 0.52 (95% CI: 0.21, 1.30) for the intervention group compared to the control group after adjusting for gender, educational level and BMI (Table 3, model 3).


Findings from the present study showed that providing free school fruit for one year in elementary school had no statistically significant effect on weight status of the participants 14 years later. Previously, we have reported a reduced prevalence of overweight in the intervention group compared to the control group after seven-year follow-up, but the effect was not significant after adjusting for gender and parental education [22]. Lack of statistically significant results in the crude model presented in the present study, may be partly explained by the long follow-up period and/or the much larger (and therefore more representative) sample size compared to the seven-year follow-up study.

Previously published systematic review studies have reported a lack of evidence showing an association between fruit and vegetable intake and weight status and that many studies investigating this relationship has been hampered by low study quality [37,38]. However, a ten-year observational follow-up study showed a reduced weight gain among healthy adults with a high intake of FV at the beginning of the study period [39], and a comprehensive observational study among women showed that a diet rich in fruit and vegetables was associated with less weight gain over a 14-year period [40].

A few other intervention studies promoting healthy eating habits among children have shown significant reductions in BMI two and three years after the intervention period [4143]. Based on these studies and other studies showing a dose-response association between pre-pregnancy BMI and offsprings`birth weight (5,6,8), we hypothesized that a one-year intervention promoting fruit and vegetable intake could affect weight status of participants on the long term, and possibly also affect birth weight of their children. Results from the present study, however, showed no statistically significant differences in birth weight status among participants offspring in the intervention group and those participating in the control group.

As few previously published FV intervention studies have evaluated the long-term effect on weight status [44], a major strength of the present study was the extended follow-up period and the effect evaluation on both participants and their children. Another important strength of the FVMM study was the partly randomized intervention design, and that participants were enrolled prior to pregnancy. It is well-known, however, that results may be biased due to differences in background characteristics between the study sample and those lost to follow-up [45]. Among those participating in the last follow-up study (2016) a significantly higher educational level was shown in the intervention group compared to the control group. No other differences in characteristics were observed between the groups.

The present study has also several limitations that should be considered. We did not collect baseline measures of weight and height, and the subsequent measures of participants´ weight and height and birth weight of their children were self-reported. Based on the fact that overweight or obese peoples might tend to understate their weight [46], it is possible that errors have occurred in the process of grouping the BMI of the participants. In addition to the weight and height of the participants, the birth weight of their children was self-reported. The gestational age and prevalence of preterm birth was unknown as well. Furthermore, the intervention schools were only randomized within one of the two participating counties [29]. Hence, the FVMM intervention study was not perfectly randomized, and the control group also included nine schools that from the school year 2001/2002 participated in the Norwegian school fruit subscription program (a parental payment version). Some of the participants also received an FV educational program during the intervention year (2001/2002), but we have previously confirmed that this part of the program did not affect fruit and vegetable intake [32]. Finally, the present study was limited by a low number of participants´ children included in the study, which most likely resulted in low statistical power and reduced possibility to detect possible intervention effects on birth weight of the children. As the present study included only a few children with low birth weight in particular, but also a relatively few children with high birth weight, these categories were merged to compare normal birth weight with non-normal birth weight. Unfortunately, this merging of categories caused loss of relevant information. Thus, future follow-up studies should include a higher number of participants which will most likely result in a sufficient number of participants in each weight category and thereby assuring an adequate power to detect statistically significant associations.


Results from the present study did not show an intervention effect on participants´ weight status 14 years later or on the birth weight of their children. In order to conclude whether free school fruit may have a protective effect on the risk of developing overweight or obesity or on birth outcome, longer term follow-up studies with larger samples are needed. Future studies should also explore whether provision of free school fruit for more than one year may positively effect weight status in adulthood.

Supporting information

S1 Table. Data describing characteristics of the participants at different time points.


S2 Table. Data on birth weight of offspring and parental weight status 14 years after participation in the intervention group or the control group of the free school fruit intervention.



  1. 1. World Health Organization. Report of the Commission on Ending Childhood Obesity. Geneva: World Health Organisation, 2016.
  2. 2. World Health Organization. Preventing Chronic Diseases: a Vital Investment. Geneva: World Health Organisation, 2005.
  3. 3. Ebbeling CB, Pawlak DB, Lugwig DS. Childhood obesity: public health crisis, common sense cure. Lancet. 2002;360:473–82. pmid:12241736
  4. 4. Wang YC, McPherson K, Marsh T, Gortmaker SL, Brown M. Health and economic burden of the projected obesity trends in the USA and the UK. Lancet. 2011;378:815–25. pmid:21872750
  5. 5. Yu Z, Han S, Zhu J, Sun X, Ji C, Guo X. Pre-Pregnancy Body Mass Index in Relation to Infant Birth Weight and Offspring Overweight/Obesity: A Systematic Review and Meta-Analysis. PLoS One. 2013;8.
  6. 6. Adane AA. Maternal preconception weight trajectories are associated with offsprings ‘ childhood obesity. Int J Obes. 2018;42:1265–74.
  7. 7. Castillo-laura H, Santos IS, Quadros LC, Matijasevich A. Maternal obesity and offspring body composition by indirect methods: a systematic review and meta-analysis. Cad Saúde Pública. 2015;31:2073–92. pmid:26735376
  8. 8. Marchi J, Berg M, Dencker A, Olander EK, Begley C. Risks associated with obesity in pregnancy, for the mother and baby: A systematic review of reviews. Obes Rev. 2015;16:621–38. pmid:26016557
  9. 9. Brown J, Na A, West J, Brown S, Cjd M, Farrar D, et al. Lifestyle interventions for the treatment of women with gestational diabetes. Cochrane Database Syst Rev. 2017;5:CD011970. pmid:28472859
  10. 10. Dodd JM, Grivell RM, Crowther CA, Robinson JS. Antenatal interventions for overweight or obese pregnant women: A systematic review of randomised trials. BJOG. 2010;117:1316–26. pmid:20353459
  11. 11. Catalano P, Demouzon SH. Maternal obesity and metabolic risk to the offspring: Why lifestyle interventions may have not achieved the desired outcomes. Int J Obes. 2015;39:642–9.
  12. 12. Schwingshackl L, Hoffmann G, Kalle-Uhlmann T, Arregui M, Buijsse B, Boeing H. Fruit and vegetable consumption and changes in anthropometric variables in adult populations: A systematic review and meta-analysis of prospective cohort studies. PLoS One. 2015;10:1–19.
  13. 13. Boeing H, Bechthold A, Bub A, Ellinger S, Haller D, Kroke A, et al. Critical review: Vegetables and fruit in the prevention of chronic diseases. Eur J Nutr. 2012;51:637–63. pmid:22684631
  14. 14. Wang X, Ouyang Y, Liu J, Zhu M, Zhao G, Bao W, et al. Fruit and vegetable consumption and mortality from all causes, cardiovascular disease, and cancer: systematic review and dose-response meta-analysis of prospective cohort studies. BMJ. 2014;349:g4490. pmid:25073782
  15. 15. World Health Organization. Diet, nutrition and the prevention of chronic diseases. Geneva: World Health Organization, 2003.
  16. 16. World Cancer Research Fund/American Institute for Cancer Research. Diet, Nutrition, Physical Activity and Cancer: a Global Perspective. Continuous Update Project Expert Report 2018. Available at
  17. 17. World Health Organization. The Challenge of Obesity in the WHO European Region and the Strategies for Response. Copenhagen: Word Health Organization, 2007.
  18. 18. Singh AS, Mulder C, Twisk JWR, Van Mechelen W, Chinapaw MJM. Tracking of childhood overweight into adulthood: A systematic review of the literature. Obes Rev. 2008;9:474–88. pmid:18331423
  19. 19. De Sa J, Lock K. Will European agricultural policy for school fruit and vegetables improve public health? A review of school fruit and vegetable programmes. Eur J Public Health. 2008;18:558–68. pmid:18719006
  20. 20. Jaime PC, Lock K. Do school based food and nutrition policies improve diet and reduce obesity? Prev Med. 2009;48:45–53. pmid:19026676
  21. 21. Brown EC, Buchan DS, Baker JS, Wyatt FB, Bocalini DS, Kilgore L. A Systematised Review of Primary School Whole Class Child Obesity Interventions: Effectiveness, Characteristics, and Strategies. Biomed Res Int. 2016;2016:4902714. pmid:27668254
  22. 22. Bere E, Klepp K-I, Overby NC. Free school fruit: can an extra piece of fruit every school day contribute to the prevention of future weight gain? A cluster randomized trial. Food Nutr Res. 2014;58.
  23. 23. Kaiser KA, Brown AW, Bohan Brown MM, Shikany JM, Mattes RD, Allison DB. Increased fruit and vegetable intake has no discernible effect on weight loss: a systematic review and meta-analysis. Am J Clin Nutr. 2014;100:567–76. pmid:24965308
  24. 24. Shaya FT, Flores D, Gbarayor CM, Wang J. School-based obesity interventions: A literature review. J Sch Health. 2008;78:189–96. pmid:18336677
  25. 25. De Vet E, de Ridder DTD, de Wit JBF. Environmental correlates of physical activity and dietary behaviours among young people: A systematic review of reviews. Obes Rev. 2011;12:e130–42. pmid:20630024
  26. 26. Bere E, Klepp K-I. Correlates of fruit and vegetable intake among Norwegian schoolchildren: parental and self-reports. Public Health Nutr. 2004;7:991–8. pmid:15548337
  27. 27. Bere E, Te Velde SJ, Smastuen MC, Twisk J, Klepp KI. One year of free school fruit in Norway—7 years of follow-up. Int J Behav Nutr Phys Act. 2015;12:139. pmid:26556692
  28. 28. Bere E, Veierod MB, Skare O, Klepp K-I. Free School Fruit—sustained effect three years later. Int J Behav Nutr Phys Act. 2007;4:5. pmid:17309800
  29. 29. Bere E, Veierød MB, Klepp KI. The Norwegian School Fruit Programme: Evaluating paid vs. no-cost subscriptions. Prev Med. 2005;41:463–70. pmid:15917042
  30. 30. Øverby NC, Klepp KI, Bere E. Introduction of a school fruit program is associated with reduced frequency of consumption of unhealthy snacks. Am J Clin Nutr. 2012;96:1100–3. pmid:23034961
  31. 31. Bere E, Veierod MB, Bjelland M, Klepp K-I. Free school fruit—sustained effect 1 year later. Health Educ Res. 2006;21:268–75. pmid:16219630
  32. 32. Bere E, Veierød MB, Bjelland M, Klepp KI. Outcome and process evaluation of a Norwegian school-randomized fruit and vegetable intervention: Fruits and Vegetables Make the Marks (FVMM). Health Educ Res. 2006;21:258–67. pmid:16219631
  33. 33. Bere E, Brug J, Klepp KI. Why do boys eat less fruit and vegetables than girls? Public Health Nutr. 2008;11:321–5. pmid:17666125
  34. 34. Cole TJ. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1240–3. pmid:10797032
  35. 35. WHO. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser. 2000;894:1–253.
  36. 36. Hosmer DW, Lemeshow S. Applied Logistic Regression (2nd ed.): New York, John Wiley, 1989.
  37. 37. Newby PK. Plant foods and plant-based diets: Protective against childhood obesity? Am J Clin Nutr. 2009;89:1572S–87S. pmid:19321559
  38. 38. Schwingshackl L, Hoffmann G, Kalle-Uhlmann T, Arregui M, Buijsse B, Boeing H. Fruit and Vegetable Consumption and Changes in Anthropometric Variables in Adult Populations: A Systematic Review and Meta-Analysis of Prospective Cohort Studies. PLoS One. 2015; 10(10):e0140846. pmid:26474158
  39. 39. Vioque J, Weinbrenner T, Castelló A, Asensio L, Garcia De La Hera M. Intake of fruits and vegetables in relation to 10-year weight gain among Spanish adults. Obesity. 2008;16:664–70. pmid:18239583
  40. 40. Boggs DA, Palmer JR, Spiegelman D, Stampfer MJ, Adams-Campbell LL, Rosenberg L. Dietary patterns and 14-y weight gain in African American women. Am J Clin Nutr. 2011;94:86–94. pmid:21593501
  41. 41. Robertson W, Thorogood M, Inglis N, Grainger C, Stewart-Brown S. Two-year follow-up of the “Families for Health” programme for the treatment of childhood obesity. Child Care Health Dev. 2012;38:229–36. pmid:21463350
  42. 42. Taylor RW, McAuley KA, Barbezat W, Farmer VL, Williams SM, Mann JI. Two-year follow-up of an obesity prevention initiative in children: The APPLE project. Am J Clin Nutr. 2008;88:1371–7. pmid:18996874
  43. 43. Reinehr T, Temmesfeld M, Kersting M, de Sousa G, Toschke a M. Four-year follow-up of children and adolescents participating in an obesity intervention program. Int J Obes. 2007;31:1074–7.
  44. 44. Jones R a, Sinn N, Campbell KJ, Hesketh K, Denney-Wilson E, Morgan PJ, et al. The importance of long-term follow-up in child and adolescent obesity prevention interventions. Int J Pediatr Obes. 2011;6:178–81. pmid:21612335
  45. 45. Dumville JC. Reporting attrition in randomised controlled trials. BMJ. 2006;332:969–71. pmid:16627519
  46. 46. Larson MR. Social desirability and self-reported weight and height. Int J Obes. 2000;24:663–5.