Physical activity for the management of obesity in children up to the age of 9 years.

rationaleChildhood obesity is a major global public health concern. Although physical activity is recognised as an effective non-pharmacological intervention, most existing evidence synthesis has primarily focused on the role of physical activity in preventing obesity, with limited attention given to its effects on managing children with obesity.

objectivesTo synthesise evidence on the benefits and harms of physical activity for the management of obesity in children up to 9 years of age. SEARCH

methodsWe searched CENTRAL, MEDLINE, Embase, and two trial registries from 2012 to 2 June 2023, with an update on 4 December 2025. We also used Google Scholar to identify additional studies. We did not impose language or publication status restrictions. ELIGIBILITY CRITERIA: We included randomised controlled trials (RCTs) that investigated the effects of physical activity interventions of any frequency, mode (e.g. aerobic, resistance), or intensity, with a minimum intervention duration of 12 weeks, in children up to 9 years of age with obesity (as defined by the trialists) at baseline. Eligible comparators were standard care, waiting-list, no physical activity, or any other active interventions, such as comparisons of different physical activity parameters (e.g. duration, frequency, intensity). OUTCOMES: Critical outcomes: body mass index (BMI), BMI z-score, body weight, health-related quality of life, adiposity and fat distribution, glycaemia, and adverse events (minor and serious). Important outcomes: physical well-being, mental well-being, physical activity levels, blood pressure, hyperinsulinemia, resistance to insulin, alterations in lipid metabolism, presence of obesity-related comorbidities or any non-communicable diseases, obesity-associated disability, lipid hormones, alterations in hunger or satiety, disability, mortality, prevalence of obesity in adulthood, and access to health services. RISK OF BIAS: Pairs of review authors independently assessed the risk of bias in included studies using the original version of the Cochrane tool (RoB 1). SYNTHESIS

methodsWe synthesised results using meta-analysis when appropriate. Given the clinical and statistical heterogeneity, we predominantly used a random-effects model alongside sensitivity analyses. When meta-analysis was not feasible, we employed synthesis without meta-analysis (SWiM) methods. We assessed the certainty of the evidence using the GRADE approach. INCLUDED STUDIES: We included four studies (five references, 517 children; female 46%; mean age ranged from 8.9 to 9.9) conducted in three countries (the USA, Brazil, and Iran). Three studies (75%) delivered the physical activity interventions in school settings. Three studies (75%) compared physical activity interventions with non-exercise controls, while one compared physical activity with a behaviour-changing intervention. The duration of physical activity interventions ranged from 12 to 32 weeks. We assessed three studies as having an overall high risk of bias. SYNTHESIS OF

resultsPhysical activity interventions compared to control The evidence is very uncertain about the effect of physical activity interventions on BMI (mean difference (MD) -1.52 kg/m², 95% confidence interval (CI) -2.74 to -0.29; I² = 0%; 2 studies, 118 children; very low-certainty evidence), BMI z-scores (MD -0.10 z-score units, 95% CI -0.22 to 0.02; I² = 29%; 1 study, 222 children; very low-certainty evidence), body weight (MD -0.86 kg, 95% CI -3.17 to 1.46; I² = 0%; 2 studies, 118 children; very low-certainty evidence), health-related quality of life (MD -0.60 points, 95% CI -4.22 to 3.02; 1 study, 175 children; very low-certainty evidence), adiposity and fat distribution, assessed as body fat percentage (MD -1.23%, 95% CI -2.50 to 0.03; I² = 0%; 3 studies, 456 children; very low-certainty evidence), minor adverse events (risk ratio (RR) 3.58, 95% CI 1.95 to 6.55; 1 study, 222 children; very low-certainty evidence), and serious adverse events (RR 1.08, 95% CI 0.10 to 11.76; 1 study, 222 children; very low-certainty evidence). No studies assessed glycaemia. Combined training (aquatic exercises) versus combined training (video game exercises) The evidence is very uncertain about the effect of combined training in an aquatic setting compared to video-game-based combined training on BMI (MD -0.90 kg/m², 95% CI -3.21 to 1.41; 1 study, 39 children; very low-certainty evidence), BMI at 4-week follow-up (MD 0.03 kg/m², 95% CI -2.54 to 2.60; 1 study, 39 children; very low-certainty evidence), body weight (MD -1.30 kg, 95% CI -5.66 to 3.06; 1 study, 39 children; very low-certainty evidence), and body weight at 4-week follow-up (MD -0.70 kg, 95% CI -5.70 to 4.30; 1 study, 39 children; very low-certainty evidence). This study did not assess the other critical outcomes. Low-dose versus high-dose combined training The evidence is very uncertain about the effect of low-dose versus high-dose combined training on BMI z-scores (MD 0.12 z-score units, 95% CI 0.10 to 0.14; 1 study, 144 children; very low-certainty evidence), adiposity and fat distribution, assessed as body fat percentage (MD 1.07%, 95% CI -0.74 to 2.88; 1 study, 144 children; very low-certainty evidence), glycaemia, assessed as fasting glucose level (MD -0.50 mg/dL, 95% CI -2.77 to 1.77; 1 study, 144 children; very low-certainty evidence), minor adverse events (RR 0.91, 95% CI 0.64 to 1.30; 1 study, 144 children; very low-certainty evidence), and serious adverse events (RR 3.08, 95% CI 0.13 to 74.45; 1 study, 144 children; very low-certainty evidence). This study did not assess the other critical outcomes. AUTHORS'

conclusionsThe evidence from four randomised studies on the effects of physical activity interventions in children aged 0 to 9 with obesity is of very low certainty. Serious methodological limitations, clinical heterogeneity, small-study effects, and imprecise results constrained this evidence base. Important knowledge gaps remain because none of the included RCTs enroled children with disabilities. The RCTs provided little information on contextual factors. Future high-quality and better-reported RCTs will likely change our findings.

fundingThe Department of Nutrition and Food Safety at the World Health Organization (WHO) commissioned and provided financial support for this work. WHO acknowledges financial support from the Norwegian Agency for Development Cooperation, the Swedish International Development Cooperation Agency, the Government of the Grand Duchy of Luxembourg, and the Government of Germany to the Department of Nutrition and Food Safety. REGISTRATION: Protocol available via https://doi.org/10.17605/OSF.IO/DSHUP.