AI-generated research brief — verify at source
Dietary Behaviour, Not Physical Activity, Mediates Body Composition Gains in Paediatric Obesity Trial
A 9-month cluster-randomised trial in 1,180 school-age children found that the DECIDE-Children intervention reduced whole-body and trunk fat mass percentage by 1.003% and decreased a composite adiposity factor (PCA_1) by 0.942 standard deviations, while concurrently increasing fat-free and muscle mass. Dietary behaviour accounted for 16.1% of the mediated effect on the composite adiposity score, whereas neither screen time nor moderate-to-vigorous physical activity showed statistically significant mediation.
What Was Studied
This study investigated whether changes in dietary behaviour and physical activity mediated the effects of the DECIDE-Children intervention on detailed body composition outcomes in school-age children. Rather than relying solely on BMI, the researchers derived a composite adiposity factor using principal component analysis (PCA) to capture the shared variance across whole-body and trunk fat mass, fat-free mass, and muscle mass indices — providing a more comprehensive picture of how intervention-related body changes unfold and through which behavioural pathways.
How It Was Studied
The analysis drew on data from a cluster-randomised controlled trial conducted across multiple regions, enrolling 1,180 children. The DECIDE-Children intervention was multifaceted and digitally assisted, combining health education, physical activity reinforcement, and app-based BMI monitoring with feedback for children, alongside school food policy changes and structured family engagement — all delivered over a 9-month period. A control group was assigned at the cluster level, as is standard in school-based trials. Body composition was measured using bioelectrical impedance analysis, and potential behavioural mediators — dietary behaviour, screen time, and moderate-to-vigorous physical activity (MVPA) — were assessed using validated questionnaires. Mediation analysis was performed within linear mixed-effects models adjusted for age, sex, and region, with PCA used to reduce the dimensionality of the body composition outcome variables.
What Was Observed
- The intervention produced meaningful reductions in adiposity across multiple body composition metrics. Whole-body and trunk fat mass percentage fell by 1.003%, and the composite adiposity factor PCA_1 — where higher scores indicate greater fat accumulation and lower muscle mass — decreased by 0.942 standard deviations. These changes were accompanied by increases in both fat-free mass percentage and muscle mass percentage at the whole-body and trunk levels, indicating a favourable shift in body composition rather than simple weight loss.
- Dietary behaviour was the only statistically significant mediator of the intervention’s effect on body composition. It mediated 16.1% of the total intervention effect on PCA_1, and mediation was consistent across all individual body composition outcomes tested. This finding held up across sensitivity analyses, lending stability to the conclusion.
- Physical activity — whether measured as screen time or as moderate-to-vigorous physical activity (MVPA) — did not significantly mediate any of the body composition outcomes. Despite physical activity reinforcement being a core component of the intervention, changes in MVPA and screen time did not explain the observed improvements in adiposity or muscle mass through the mediation pathway tested.
Why This Matters
Paediatric obesity trials have often treated dietary and physical activity components as roughly equivalent contributors to body composition change, yet this analysis offers formal statistical evidence that dietary behaviour carries the dominant mediating role in this intervention context. The use of PCA to derive a composite adiposity outcome addresses a recognised methodological challenge in this field — the presence of multiple correlated body composition indicators — and may serve as a model for future trial design. The null mediation finding for physical activity is particularly notable given the prominence of activity-based components in the DECIDE-Children programme, suggesting that body composition gains are being driven through the dietary rather than the movement pathway.
How to Read This Result
This is a well-powered, multi-region cluster-RCT with a rigorous analytical approach, but the reliance on validated self-report questionnaires for mediator assessment introduces the possibility of measurement error, and the absence of a stated age range for participants limits how broadly the findings can be generalised.
Limitations
The abstract does not explicitly report study limitations.