Obesity Medications Work Best for Young People When Combined With Lifestyle Support

Key Takeaways: 

• Among young people with obesity, pairing medication with structured behavioural and lifestyle support produced the largest short-term reductions in BMI.
  
• Metformin combined with behaviour and lifestyle treatment lowered BMI by 4.95, whereas metformin used on its own showed no significant change.
  
• Semaglutide plus counselling was linked to the biggest BMI reduction of any approach, although this estimate rested on a single trial.
  
  
  

Support matters as much as the medicine

Children and adolescents with obesity who received a combination of medication and structured lifestyle treatments achieved the greatest short-term reductions in BMI, according to a new evidence synthesis. The findings point away from prescribing in isolation and towards a model in which medication is layered on top of behavioural and family support rather than used as a stand-alone fix.

How the study was conducted

Researchers carried out a systematic review and network meta-analysis, searching the literature databases through June 2025, to work out which obesity treatments perform best for young people. A network meta-analysis allows multiple interventions to be compared against one another even where they have not all been tested head-to-head in the same trial.

The final analysis brought together 42 randomised clinical trials involving 3835 participants aged 10–19 years with obesity. The median age was 14.5 years, and 59.2% of participants were female individuals. Most of the included studies followed up participants over 6–12 months, placing the emphasis firmly on short-term outcomes.

The interventions assessed fell into several categories: structured behavioural and lifestyle treatments, in both standard and intensive forms; counselling; medications, including GLP-1 receptor agonists, metformin, orlistat, and phentermine–topiramate; and combinations of medication with lifestyle treatment.

The primary outcomes were changes in BMI and BMI z-score, while the secondary outcomes were changes in waist circumference, fat mass, and lean mass. The interventions were then ranked in order of effectiveness.

On study quality, the risk for bias was judged low in 21.4% of trials and high in 26.2%, with the remaining 52.4% raising some concerns. The overall certainty of the evidence ranged from very low to high, so the strength of the findings varies considerably from one comparison to another.

What the analysis found

Across the 35 trials that reported BMI and the 19 that reported BMI z-score, medications produced larger reductions when paired with lifestyle treatments than when used alone. Metformin illustrates the pattern clearly: combined with behaviour and lifestyle treatment it was associated with a reduction of 4.95 in BMI, whereas metformin used on its own showed no significant change in BMI.

Semaglutide plus counselling was associated with the largest reduction in BMI (mean difference [MD], −8.31) and in BMI z-score (MD, −1.80). This estimate, however, came from a single trial, so it should be read with caution.

Behavioural and lifestyle treatment on its own was associated with reductions in BMI (MD, −3.85; five studies) and in BMI z-score (MD, −0.89; one study) – results that matched or exceeded the effect of certain medications used alone. Combination treatments were linked to the largest reductions in fat mass, drawing on 21 studies.

What it means in practice

The authors framed the combined approach as consistently outperforming medication given without support. “[The] finding suggests that even combining medication with basic counselling was still superior to giving medication without any lifestyle support,” the researchers wrote. “Medications should never be prescribed in isolation; a person-centered, family-centered approach matching treatment intensity to medical need is essential,” they added.

Where the research came from

The study was led by Ke-wen Wan, MSc, of Hong Kong Baptist University in Hong Kong SAR, China. It was published online on 22 June in JAMA Pediatrics.

Limitations to consider

Several caveats temper the results. The findings for newer medications were based on only a few small trials, which limits confidence in those specific estimates. The wide age range may have obscured differences by age or stage of puberty, since a 10-year-old and a 19-year-old can respond very differently to the same intervention. Most of the trials also did not report data on race, ethnicity, or income, leaving open questions about how the findings apply across different populations.

Funding and disclosures

The study received funding from grants from Hong Kong Baptist University. One author reported serving on professional boards related to childhood obesity and receiving travel grants or reimbursements, and another author reported receiving consulting fees from pharmaceutical companies. Detailed disclosures are available in the original article.