News · Heart & Metabolic
Weight-loss drugs worked better with lifestyle, not without
A JAMA Pediatrics network meta-analysis of 42 trials in 3,835 adolescents found every obesity drug performed better paired with lifestyle treatment than on its own.
Based on a peer-reviewed network meta-analysis in JAMA Pediatrics
- A network meta-analysis of 42 randomized trials covering 3,835 participants aged 10-19, published in JAMA Pediatrics.
- It compared lifestyle treatment, four drug classes, and their combinations against control.
- Combinations of drug plus lifestyle treatment were the most effective across every adiposity outcome measured.
- Semaglutide plus counseling produced the largest body mass index reduction, a mean difference of -8.31, though from a limited number of studies.
- Every drug did better when paired with lifestyle treatment than the same drug alone.
- Lifestyle treatment on its own still produced a substantial reduction, a mean difference of -3.85 in body mass index.
- Trials typically ran 6 to 12 months, so nothing here speaks to what happens after that.
- Adolescent bodies are still growing, which is why body mass index z scores are reported alongside raw values.
The arrival of effective weight-loss drugs raised an obvious question in pediatrics: if the medication works, does the hard part still matter? An analysis in JAMA Pediatrics, published by the American Medical Association, pooled 42 trials to answer it.
Treatments combining pharmacotherapy and lifestyle treatment demonstrated the greatest efficacy across all adiposity-related outcomes.
The comparison nobody had assembled
Pediatric obesity is a global health challenge, and the standard of care has long been structured behavior change. Drugs arrived on top of that rather than in place of it.
What was missing until now was a ranking. Although health behavior and lifestyle treatment is foundational, the comparative effectiveness of it, various pharmacotherapies, and their combinations remains unclear. Individual trials each answered one small piece.
What went into the pool
The team searched four databases from inception to June 2025, selecting randomized clinical trials involving children and adolescents aged 10-19 years with obesity.
Interventions covered lifestyle treatment, four drug groups, and combinations of the two. A total of 42 randomized trials with 3,835 participants were included, with a median age of 14.5 years and 59.2% female.
Outcomes ran wider than the scales: body mass index and its z score as primary measures, with waist circumference, fat mass and lean mass alongside.
The ranking, and its asterisk
Semaglutide plus counseling was associated with the largest body mass index reduction, a mean difference of -8.31, and the largest z score reduction at -1.80.
The authors immediately qualify it. That estimate was based on a limited number of studies, which is the difference between a large effect and a reliable one.
The broader pattern is what holds up. All pharmacological treatments were more effective when paired with lifestyle treatment, producing greater reductions than the same medications alone.
The finding that survives the caveats
Strip out the drug rankings and one result remains solid. Lifestyle treatment as monotherapy was associated with substantial decreases in body mass index and in body mass index z score against control, at a mean difference of -3.85.
That is roughly half the effect attributed to the best drug combination, achieved with no prescription. And every medication in the analysis performed better with it than without.
What 6 to 12 months cannot show
The trials are short. The authors describe the effects as short-term, typically 6 to 12 months, which is the window in which almost all of this evidence was generated. Adolescent obesity is not a 12-month condition.
Network meta-analysis also carries its own fragility: comparing treatments never tested against each other assumes the underlying trials were similar enough to link, and 42 trials across a decade strain that.
The conclusion the authors draw is about sequencing rather than substitution: pharmacotherapy was a key component, not solely an adjunct, associated with the greatest improvements, with lifestyle treatment remaining indispensable underneath it.
Both halves of that sentence are load-bearing, and the second half is the one likely to get dropped.
People also ask
What is a network meta-analysis?
A way of ranking treatments that have rarely been tested head to head. If drug A beat control in one trial and drug B beat control in another, shared comparisons let you estimate how A and B compare with each other. It extracts more from existing trials than a simple pooled average, and it rests on the assumption that the trials were similar enough to link, which is a real limitation when they span different ages and settings.
Should adolescents be taking these drugs?
That is a clinical decision for a family and a pediatrician, not something a meta-analysis settles. What this evidence supports is that when medication is used, it works better alongside structured lifestyle support than instead of it. It says nothing about who should be offered medication in the first place, and the trials were short.
Why does the semaglutide number come with a caveat?
Because it rests on few studies. The authors flag that the estimate was based on a limited number of studies, which makes it less stable than it looks. A large effect from thin evidence can move substantially when the next trial reports. The finding that combinations beat monotherapy is supported far more broadly than any single drug's ranking.
Does lifestyle treatment alone still do anything?
Yes, and the paper is emphatic about it. Lifestyle treatment as monotherapy was associated with substantial decreases in body mass index and in body mass index z score against control. The authors describe it as an indispensable component of any effective weight management rather than something medication replaces.
What is a body mass index z score and why report it?
It expresses a child's body mass index relative to others of the same age and sex, because children are still growing and raw values shift with height and development. Reporting both means a change can be read as genuine rather than as an artifact of a growth spurt.