A new study reveals that community-based obesity programmes benefit younger children more than adolescents, highlighting the importance of age-specific approaches informed by advanced data analysis techniques.
Researchers have found that community-based obesity prevention programmes do not affect all children in the same way, with younger children appearing to benefit far more than teenagers. In a study published in the International Journal of Data Science and Analytics, a team led by Nu Hoang and Thin Nguyen of Deakin University analysed data from more than 7,000 children enrolled in seven community interventions and concluded that the average result concealed wide variation from one child to the next. The strongest signal was age: it emerged as the clearest factor shaping whether a programme helped, had little effect or may even have worked against some participants.
That finding fits with a broader body of obesity research showing that community efforts can work, but often only modestly and often best when they are paired with activity, dietary change or support in other settings. A systematic review indexed by PubMed found that school-based programmes with both nutrition and exercise components were moderately effective, while community-based interventions also showed moderate benefit, especially when linked with school initiatives. The World Obesity Federation has similarly said results vary widely across studies and that integrated community approaches are usually more promising than isolated campaigns.
What sets the new study apart is its use of causal machine learning, a set of methods designed to identify not only whether an intervention works on average, but for whom it works best. Rather than relying on conventional regression alone, the researchers used a workflow that combined causal discovery, formal identification and double machine learning, then applied a decision tree to sort children into groups with positive, negative or negligible responses. Their analysis pointed to baseline age as the most important upstream variable, with weekly takeaway food consumption also helping to distinguish which children gained the most from the programmes.
The headline estimate for the whole group was small: a reduction of 0.02 in body mass index z-score, with confidence intervals that crossed zero. But the individual estimates ranged much more widely, from a meaningful drop in BMI z-score to a small increase. The researchers found that about 18.7% of children had confidence intervals entirely below zero, suggesting clear benefit, while 8.7% had intervals entirely above zero, implying possible harm. Children under 8.9 years old formed the strongest-benefit subgroup, while adolescents older than 15.2 showed the most negative pattern. Children between those ages who ate takeaway food less than once a week also appeared to benefit.
The results echo other recent evidence that community prevention works best when it changes more than one part of a child’s environment. A 2024 systematic review and meta-analysis found that universal community-engagement programmes had greater impact when they combined environmental and behavioural changes and brought parents into the effort. A randomised clinical trial in JAMA Pediatrics also reported that linking family care with community resources improved child BMI and family-centred outcomes. Another review found that multicomponent community programmes involving several sectors show promise, but that economic evidence remains limited, leaving policymakers with too little information on long-term value for money.
The Deakin-led team is careful not to overstate what the findings mean for any single child. Community programmes are voluntary, which means families and neighbourhoods that join may differ in important ways from those that do not. The authors also note the possibility of spillover effects between communities, and they stress that the individualised estimates are exploratory rather than a basis for personal treatment decisions. Even so, the study offers a practical lesson for public health planners: if community obesity prevention is most effective in younger children and less effective in older adolescents, then age-specific design may matter as much as the intervention itself. In a field where averages often hide important differences, the new analysis suggests that targeting the right children, at the right age, could make the difference between a programme that merely exists and one that actually helps.
Disclaimer: This content is for informational purposes only and is not intended to be a substitute for professional medical judgment, advice, diagnosis, or treatment.
