Article Text

Download PDFPDF

Impact of child summertime obesity interventions on body mass index, and weight-related behaviours: a systematic review and meta-analysis protocol
  1. Jennette P Moreno1,
  2. Lydi-Anne Vézina-Im1,
  3. Elizabeth M Vaughan2,
  4. Tom Baranowski1
  1. 1 Department of Pediatrics-Nutrition, USDA/ARS Children’s Nutrition Research Center, Baylor College of Medicine, Houston, Texas, USA
  2. 2 Department of Medicine, Baylor College of Medicine, Houston, Texas, USA
  1. Correspondence to Dr Jennette P Moreno; palcic{at}


Introduction In previous studies, it has been found that on average, children consistently gained weight during the summer months at an increased rate compared with the 9-month school year. This contributed to an increased prevalence of overweight and obesity in children. Several obesity-related interventions have occurred during or targeting the summer months. We propose to conduct a systematic review and meta-analysis of the impact of obesity prevention and treatment interventions for school-age children conducted during the summer or targeting the summer months when children are not in school on their body mass index (BMI), or weight-related behaviours.

Methods and analyses A literature search will be conducted by the first author (JPM) using MEDLINE/PubMed, Cochrane Library, Scopus, CINAHL, PsycINFO, EMBASE and Proquest Dissertations and Theses databases from the date of inception to present. Studies must examine interventions that address the modification or promotion of weight-related behaviours (eg, dietary patterns, eating behaviours, physical activity (PA), sedentary behaviour or sleep) and target school-age children (ages 5–18). The primary outcomes will be changes from baseline to postintervention and/or the last available follow-up measurement in weight, BMI, BMI percentile, standardised BMI or per cent body fat. Secondary outcomes will include changes in dietary intake, PA, sedentary behaviour or sleep. Risk of bias will be assessed using the Cochrane risk of bias tool for randomised and non-randomised studies, as appropriate.

Ethics and dissemination Because this is a protocol for a systematic review, ethics approval will not be required. The findings will be disseminated via presentations at scientific conferences and published in a peer-reviewed journal. All amendments to the protocol will be documented and dated and reported in the PROSPERO trial registry.

PROSPERO registration number CRD42016041750

  • summer
  • childhood obesity prevention
  • dietary intake
  • physical activity
  • sedentary behaviour
  • sleep

This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • This will be the first review to systematically examine the impact of behavioural obesity prevention and treatment interventions conducted during or targeting the summer months when children are not in school on changes in children’s weight status and/or weight-related behaviours during summer. Results will help to inform future behavioural obesity prevention and treatment interventions for summertime.

  • Since a limited number of studies may have delivered obesity prevention interventions during summer months, the scope of the review will also include obesity treatment interventions occurring during the summer break from school.

  • In addition to objectively measured body composition outcomes, we will include studies assessing diet, physical activity, sedentary behaviour and sleep outcomes using self-reported or parent-reported measures. These measures are known to be subject to substantial error.


Obesity is a critical public health problem in the USA with 17% of children and 38% of adults classified as obese or extremely obese.1 2 Globally, 73 countries have experienced a doubling in the prevalence of obesity among children and adults since 1980.3 Obesity increases the risk for diabetes,4 heart disease4 and cancer.5 It is associated with all leading causes of preventable death6 and has a US estimated annual medical cost of US$149.4 billion.7 In 2015, 7% of all deaths globally (4.0 million) were associated with having a high body mass index (BMI).3 Cardiovascular disease and diabetes were the leading causes of obesity-related deaths worldwide.3 In the USA, rates of obesity appeared to be higher among school-age children with 17.5% of those aged 6–11 years and 20.5% of those aged 12–19 years being classified as obese compared with only 8.9% of those aged 2–5 years,8 suggesting this is a time when increased focus on the prevention of weight gain is needed.

Findings from studies of two large nationally representative samples of children in the USA,9 10 and one each in Texas,11 12 Minnesota,13 Wyoming14 and Osaka Prefecture and Tokyo, Japan15 16 consistently demonstrated that school-age children gained weight at an accelerated rate during the summer months. Summertime increases in standardised BMI contribute to increased risk for obesity,10 17 18 especially among children at risk for developing chronic health conditions associated with obesity.18 A large 5-year longitudinal study that examined children’s standardised BMI biannually for 5 years throughout elementary school showed that about 19% of children began a trajectory towards overweight or obesity during early elementary school when they were 5–8 years.17 The greatest increases in risk for overweight and obesity occurred during summer vacation.17 18

Less is known about summer weight gain among preschoolers, with one study in Japan failing to find evidence of summer weight gain until 5 years old15 and another demonstrating summer weight gain among overweight children enrolled in HeadStart (ie, early childhood education for children from low-income families) and healthy-weight children receiving Medicaid (ie, government-assisted health insurance for children from low-income families) in Michigan.19 Studies involving middle school students have yielded inconsistent findings. Primarily Caucasian middle schoolers in the US Pacific Northwest demonstrated improvements in their BMI during summer months.20 No significant change in BMI during summer was observed among American Indian middle schoolers in Wyoming.14 However, results from a randomised controlled trial (RCT) of an intensive school-based obesity intervention in Texas reported significant increases in standardised BMI among Hispanic adolescents randomised to both the treatment and control conditions, though summertime increases in standardised BMI were significantly less in the treatment condition.21

Summer weight gain seems to affect school-age children regardless of race, ethnicity, weight status classification or socioeconomic status10 11; however, some studies suggest that overweight children and adolescents, African Americans, Hispanics and children of low socioeconomic status may display greater rates of accelerated summer weight gain.12 18 22 Further, there is evidence that improvements achieved by school-based obesity and fitness interventions are lost over the summer.23–26 While less is known about the influence of accelerated weight gain during out-of-school periods globally, summer weight gain in school-age children may be significantly impeding progress towards addressing the global obesity pandemic in children, representing an important opportunity to address the problem.27

Previous investigations into the causes of summer weight gain have focused primarily on physical activity (PA) and to a lesser extent diet, screen media use or sleep.28 Several studies have demonstrated that children increase their activity levels during summer29–31 while others showed a decrease13 or no differences in activity levels or energy expenditure compared with the school year.32 Three studies examined changes in diet during the school year and summer. One study found that children aged 9–11 years in Minnesota demonstrated no seasonal differences in kilocalories and macronutrient intake13 while another showed a summertime increase in total calories (+82±31 kcal) among Greek children aged 3–18 years.33 Children in the US National Health and Nutrition Examination Survey 2003–2008 aged 6–19 decreased their vegetable intake by 0.2 servings per day, increased their added sugar intake by 2.1 teaspoons per day during summer break compared with the school year and increased their screen media use by 18 min per day during summer.31 Finally, sleep duration among elementary school-age children has been demonstrated to be 40.5 min shorter during summer compared with winter, 31.1 min shorter compared with fall and 14.8 min compared with spring.34 Overall, relatively few studies have examined seasonal differences in weight-related behaviours and their association with seasonal weight change. Studies that have examined these associations have produced inconsistent findings. As a result, the weight-related behaviours that underlie summer weight gain and their relationship to weight change during the school year and summer are poorly understood.

Typically, observational studies are used to study seasonal differences in weight-related behaviours and their association with differences in seasonal weight gain. Obesity prevention and treatment studies experimentally manipulate weight-related behaviours, offering another opportunity to study the impact of changes in these behaviours on weight outcomes. Such studies may help identify important behavioural targets for the prevention and treatment of summer weight gain, behaviours amenable to change, effective intervention components and the level of BMI changed.

While few studies have examined determinants of summer weight gain, several obesity interventions have been conducted during summer.35–76 Summer obesity interventions are typically delivered via day camps35 36 56 57 62–64 71–73 or residential camps,37–39 44 49 50 52–54 56 65–69 but also include school-based interventions41 43 and multilevel community-based interventions.61 74–76 We are not aware of any systematic reviews or meta-analyses focusing specifically on the impact of summertime childhood obesity prevention and treatment interventions on BMI, dietary intake, PA, screen media use or sleep. We propose to conduct a systematic review and meta-analysis of interventions for obesity prevention and treatment among school-age children conducted during or targeting the summer months when children are not in school, and assessing their impact on BMI, diet, PA, screen media use and sleep. This systematic review will examine (1) the impact of behavioural obesity prevention interventions on children’s weight status and/or weight-related behaviours during summer, (2) whether the effects of the interventions varied by the behaviours targeted or behavioural intervention techniques employed77 and (3) biases in included studies.



This protocol has been registered with PROSPERO (registration number CRD42016041750) and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA),78 PRISMA-Protocol79 and PRISMA-Protocol for Children Extension80 guidelines.

Study selection criteria

Interventions and population

Studies reporting results of behavioural obesity prevention and treatment interventions occurring during or targeting the summer season and/or summer break from school will be included. Behavioural obesity prevention interventions will be defined as those targeting weight-related behaviours (eg, dietary patterns, eating behaviours, PA, screen media use, sedentary behaviour and sleep) among school-age children (ages 5–18). The summer season or summer break from school for studies conducted within the USA will include the following months: June, July and August. For studies conducted in the southern hemisphere, summer will be considered December, January and February, though for all studies occurring outside of the USA, we will research when the typical break from school occurs in that particular country.


Outcomes must include one or more of the following: changes in objective or self-reported measurement of weight, BMI, BMI percentile, standardised BMI, per cent body fat, waist circumference, dietary intake, PA, screen media use, sedentary behaviour or sleep from baseline to postintervention and the last available follow-up measurement (table 1). Studies will be grouped for meta-analysis according to the outcome(s) provided.

Table 1

Outcomes and operational definitions

Study designs

RCTs, quasi-experimental RCTs, and controlled or uncontrolled before-and-after comparisons will be included to allow for the inclusion of pilot studies. Studies using a waitlist, usual care, no-treatment or alternative treatment controls will be included.


Studies must be published in English, French or Spanish, with no restriction on the year of publication.

Exclusion criteria 

Studies will be excluded for the following reasons: (1) participants were diagnosed with pre-existing medical conditions (eg, diabetes, non-alcoholic fatty liver disease, cancer); (2) active intervention phase of less than 1 day; (3) letters to the editor, meta-analyses, case reports and reviews; or (4) abstracts and conference proceedings.

Search strategy

A literature search will be conducted by the first author (JPM) using MEDLINE/PubMed, Cochrane Library, Scopus, CINAHL, PsycINFO and EMBASE databases from the date of inception to present. Proquest Dissertations and Theses will also be explored for unpublished trials (a form of grey literature). Structured search strings will be developed using the following key words: (1) summer or seasonal; (2) children or adolescents or kid or boy or girl; (3) overweight, obese, obesity, or weight gain, weight, body mass index, BMI, dietary intake, diet, PA, sleep, sedentary behavior, screen media use, television, TV, video game use, computer use, mobile phone, body fat, adiposity or weight status; (4) intervention, program, prevention, or treatment. Additional studies will be gleaned from the reference lists of studies chosen for inclusion in the systematic review. Search strategy results will be reported using a PRISMA flow diagram.81 A draft of the search strategy is provided in online supplementary file 1.

Supplementary file 1

Study selection

Initial title and abstract review will be conducted by the first author (JPM). Duplicates and articles clearly not meeting selection criteria will be removed. The remaining list will be retrieved and two reviewers (JPM and EMV) will independently review the retrieved articles for eligibility. Overall agreement rate will be calculated using Cohen’s κ. A third reviewer (LAVI) will be consulted in situations when agreement is not achieved. In the event that multiple studies reporting results of the same intervention with the same sample are identified, reviewers will attempt to verify if this is the case by contacting the authors, and preference for inclusion will be given to studies reporting changes in outcomes from baseline to postintervention and follow-up and studies using more rigorous methods.

Data extraction

Codebooks adapted from the Cochrane data collection form for intervention reviews will be developed in Excel and piloted by two reviewers using five manuscripts. Two reviewers (JPM and EMV) will meet to discuss discrepancies and propose any needed changes to the codebooks.

Using the refined codebooks, data will be abstracted independently by two reviewers (JPM and EMV). Information will include setting, location of intervention delivery (eg, day camp, school, community, residential camp), study design, country, participant recruitment, participant recruitment sources, inclusion and exclusion criteria, participant allocation, sample size, demographics of the sample, intervention components,77 intervention duration, theoretical framework, dose of intervention delivered (duration in minutes, length of the intervention in days or weeks and frequency of contact), treatment fidelity, treatment compliance, outcome measurements, time points measured, means and SD for continuous outcomes measured, and handling of missing data. Codebooks created independently by reviewers will be merged and reviewers will meet to review for agreement. Discrepancies will be resolved by discussion and a third party (LAVI) will be consulted as needed.

Outcomes and prioritisation

The primary outcome will include changes in weight status-related measures including weight, BMI, BMI percentile,82 83 standardised BMI,84 85 per cent body fat and waist circumference.86 Preference will be given for standardised BMI and if not available, the authors will attempt to calculate standardised BMI by using available data or data made available to the authors. In the event that we are unable to convert data to standardised BMI, outcomes will be grouped according to the outcome measure collected. Secondary outcomes will include changes in targeted behaviour such as dietary intake, PA, screen media use, sedentary behaviour or sleep.

Risk of bias and quality assessments

Risk of bias will be assessed independently by JPM and EMV using the Cochrane risk of bias tool for randomised and non-randomised studies, as appropriate.87 The Cochrane risk of bias tool for RCTs considers the following criteria: (1) randomisation sequence generation: was the allocation sequence adequately generated?; (2) treatment allocation concealment: was the allocated treatment adequately concealed from study participants and clinicians and other healthcare or research staff at the enrolment stage?; (3) blinding: were participants and research staff blind to treatment allocation?; (4) blinding: were the personnel assessing outcomes and analysing data sufficiently blinded to the intervention allocation throughout the trial?; (5) completeness of outcome data: were participant exclusions, attrition and incomplete outcome data adequately addressed in the published report?; (6) selective outcome reporting: is there evidence of selective outcome reporting and might this have affected the study results?; (7) other sources of bias: was the trial apparently free of any other problems that could produce a high risk of bias?.87 Questions are rated as having a high, low or unclear level of bias across the seven domains.

The Cochrane risk of bias tool for non-randomised studies of interventions considers risk of bias across three domains: (1) preintervention (ie, potential confounders and selection of participants), (2) intervention (ie, classification of interventions) and (3) postintervention (ie, protocol deviations, missing data, measurement of outcomes and selective reporting of results). Studies are rated as having a low, moderate, serious and critical risk for bias. Discrepancies in ratings will be discussed until consensus is achieved or a third party (LAVI) will be consulted. Questions are rated as having a critical, moderate, low or unclear level of bias across the three domains.

The Grading Quality of Evidence and Strength of Recommendations (GRADE) guidelines will be used to assess the quality of the evidence.88 GRADE addresses methodological flaws, consistency of results across studies, generalisability of results to the target population and effect size. Each quality outcome is rated as high, moderate, low or very low.

Data management and statistical analyses

All articles retrieved from the search will be imported into EndNote X7.89 Data obtained through data extraction will be entered into Excel. Outcomes will be imported into Comprehensive Meta-analysis software90 for meta-analysis.

We will create a table describing study characteristics and major outcomes. We will provide a narrative summary of the results from the included studies, organised by type of intervention (ie, prevention or treatment), treatment modality (ie, day or residential camp, community based, school based or clinic based), targeted behaviours (ie, dietary patterns and intake, PA, sleep, sedentary behaviour), target population characteristics (eg, ethnicity, sex, age, culture), geographic location of the study, type of outcome and intervention content. Intervention effects for each study will be calculated using standardised mean differences (for continuous outcomes). We will stratify studies according to whether they are obesity treatment or prevention interventions. Studies involving the same type of intervention, treatment modality, comparator and outcome measure will be grouped.

Results will be pooled using a random-effects meta-analysis, with standardised mean differences for continuous outcomes, and we will calculate 95% CIs and two-sided p values for each outcome. For studies that have not accounted for the effects of clustering, SD will be adjusted for the design effect. Heterogeneity between the studies in effect measures will be assessed using both the χ2 test and the I2 statistic.91 We will consider an I2 value greater than 75% is indicative of substantial heterogeneity.92 When a meta-analysis of results is not appropriate, a narrative summary of results will be provided. We will conduct sensitivity analyses to explore heterogeneity in effect estimates according to participant characteristics (eg, gender, age, weight status), location, theoretical basis of the intervention (eg, theory vs non-theory based), behaviours targeted (eg, diet only, PA only, diet and PA, screen media use, sleep), intervention components,77 risk of bias and study quality. If our meta-analysis contains greater than 10 studies, a funnel plot analysis will be performed to assess for publication bias and a visual inspection for asymmetry will be performed93 and by using Egger’s94 regression test for continuous variables.


This systematic review and meta-analysis will be the first to examine the efficacy and effectiveness of behavioural summer obesity prevention and treatment interventions to reduce weight outcomes among school-age children during the summer months when children are not in school and the impact on targeted behavioural outcomes (eg, diet, PA, screen media use, sedentary behaviour and sleep). Secondary analyses will examine differences in outcomes according to participant characteristics such as gender, age, weight status, location, theoretical basis of the intervention, intervention components77 and quality of studies. By evaluating the impact of obesity interventions on weight outcomes, we hope to identify the overall efficacy and effectiveness of interventions conducted during the summer or targeting summer weight gain.

A previous review examined non-curricular approaches to increasing PA, including summer day camps and active travel to school, extracurricular activities and interventions delivered during school break periods (eg, recess). Their review of summer day camps found that day camps were effective in increasing PA, though further research was needed regarding the promotion of long-term habitual physical activity.59 A similar review examined the impact of obesity prevention and treatment interventions conducted during out-of-school times (eg, after school and summertime) among African American youth and identified only one summertime obesity prevention intervention, which demonstrated a trend towards reduced BMI.58 While we are primarily interested in the impact of obesity prevention interventions on summer weight gain, there are likely too few such studies to warrant a review. Thus, the focus of this review was expanded to examine the prevention of obesity in adulthood and, therefore, obesity treatment interventions. A review of immersion programmes, many of which occurred during the summer break from school, found immersion programmes produced significantly greater weight loss compared with outpatient interventions. These findings suggested immersion programmes were a promising method of intervening with youth.60 The current review will build on the previous findings, focusing specifically on summer as a time to prevent or treat obesity in children, and will examine the impact of the interventions on weight outcomes and weight-related behaviours.

Examination of the impact of the interventions on weight-related behaviours will identify important behavioural targets for the prevention of summer weight gain and effective strategies for inducing behaviour change. Results will help to inform factors associated with summer weight gain in children as well as refinement and future directions for development of behavioural obesity prevention interventions for summertime. Prevention of summer weight gain is an important opportunity to address the worldwide childhood obesity epidemic.

We would like to acknowledge some of the likely limitations of this review. First, while objective measures of outcomes are preferred, we will include studies assessing secondary outcomes using self-reported or parent-reported behavioural weight-related outcomes. Self-report or parent reports of diet, PA and sleep are known to overestimate sleep95 96 and PA97 and underestimate dietary intake.98 Second, it is challenging to group interventions together even by type and modality of intervention and to determine the impact of the intervention as methods and outcome measures may vary greatly across studies. As such, we will stratify studies according to whether they are obesity treatment or prevention interventions and group them according to treatment modality, comparator and outcome measure. Finally, this review may be limited by the number of summertime interventions fitting into these categories. In the event that an insufficient number of studies are identified for performing a meta-analysis, we will provide a narrative review of the studies and their outcomes.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.


  • Contributors JPM, the guarantor of the protocol, drafted the protocol and registered it in PROSPERO. TB reviewed and commented on the protocol in PROSPERO. JPM, EMV, LAVI and TB all reviewed and commented on this protocol.

  • Funding This work is a publication of the USDA (USDA/ARS) Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine (Houston, TX) and has had been funded, in part, with federal funds from the USDA/ARS under Cooperative Agreement Number 58-6250-6001. JPM is the recipient of grants from the Boyd and Evelyn Mullen Foundation and the Texas Children’s Hospital Pediatric Pilot Research Fund, which played no role in the development or review of the protocol.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.