On the trial evidence they work, and the safety picture is thinner than the efficacy one. STEP TEENS randomized 201 participants aged 12 to under 18 with obesity, two to one, semaglutide 2.4 mg against placebo for 68 weeks, both arms receiving lifestyle intervention.
One trial is not the evidence base. A systematic review and meta-analysis pooled every randomized trial in children and adolescents with obesity or type 2 diabetes [2], which is the whole picture rather than the famous study.
On the psychiatric question specifically, an adolescent analysis found that the answer changed with the comparator [3] — two comparators produced two different answers, which is a reason to read any single reassurance carefully.
Prescribing has meanwhile moved below the ages STEP TEENS studied. Prescriptions to children aged 8 to 11 rose 310-fold between 2019 and 2026 [4], and that age band has one randomized trial behind it, of a different drug. SCALE Kids assigned 82 children aged 6 to under 12 to daily liraglutide 3.0 mg or placebo for 56 weeks, both with lifestyle support [5].
BMI fell 5.8 percent on liraglutide and rose 1.6 percent on placebo, a difference of 7.4 percentage points, and 46 percent of the children on the drug cut their BMI by at least 5 percent against 9 percent on placebo. Gastrointestinal events came with it, in 80 percent against 54 percent, and serious adverse events were reported in 12 percent against 8. That trial tested the daily injection sold as Saxenda, and its results do not transfer automatically to the weekly drugs now being written for the same children.
Adolescent prescribing is where this drug class gets least comfortable and least studied, and STEP TEENS is the trial people are arguing about. It randomized 201 participants aged 12 to under 18 with obesity, two to one, to semaglutide 2.4 mg or placebo for 68 weeks, both arms receiving lifestyle intervention [1]. Ninety percent completed treatment. The effect is the largest in the literature for this age group, and the way it is measured is worth reading closely before it reaches a purchasing decision.
What the trial measured
The primary endpoint was the percentage change in BMI, not in body weight. Mean BMI fell 16.1 percent on semaglutide and rose 0.6 percent on placebo, an estimated difference of 16.7 percentage points (95% CI −20.3 to −13.2, P < 0.001). In a population still growing in height, BMI change and weight change are different quantities, and the trial chose the one that accounts for growth.
The confirmatory secondary endpoint was weight loss of at least 5 percent. That was reached by 95 of 131 participants on semaglutide reached it, against 11 of 62 on placebo — 73 percent against 18 percent, an odds ratio of 14.0 (95% CI 6.3 to 31.0). Waist circumference, glycated hemoglobin, lipids other than HDL, and alanine aminotransferase all improved more on the drug.
What it cost
Gastrointestinal adverse events occurred in 62 percent on semaglutide against 42 percent on placebo. Serious adverse events were reported in 15 of 133 on the drug and 6 of 67 on placebo, 11 percent against 9. Five participants on semaglutide developed gallstones and none on placebo did.
Five of 133 is a small number and it points the same way as the much larger gallbladder meta-analysis, whose weight-loss subgroup carried nearly double the pooled risk. In adolescents that finding has a longer horizon attached to it than it does in adults.
What this does not settle
Sixty-eight weeks is not adolescence. The trial says nothing about what happens over the years a 14-year-old has ahead of them, nothing about stopping, and nothing about bone accrual during growth — a question the post-surgical body-composition data raises in adults with far less at stake.
One practical note for anyone reading this as a parent. Adolescent prescribing is a specialist decision, and almost none of the telehealth sellers tracked here publish anything about age criteria, pediatric oversight or who reviews a case — the gap counted in the disclosure census. That absence is the relevant fact here, more than any figure in the trial.