Semaglutide in Children Aged 6 to 11: What the STEP Young Phase 3 Trial Actually Found
Novo Nordisk reported that 40.4% of children aged 6 to 11 fell below the obesity threshold on semaglutide against 0% on placebo. What that figure measures, why it is a secondary endpoint, how paediatric BMI thresholds work, and what the announcement did not report.
STEP Young is a Phase 3, randomised, double-blind, placebo-controlled trial of once-weekly semaglutide plus lifestyle intervention in children aged 6 to under 12 with obesity. Novo Nordisk reported on 7 September 2026 that 40.4% of 165 treated children had a BMI below the obesity threshold at week 68 against 0% on placebo, under the trial product estimand, which assumes full adherence. That categorical measure is a registered secondary endpoint; the registered primary endpoint is percent change in BMI from week 0 to week 68, whose value was not reported. No peer-reviewed publication exists yet.
Novo Nordisk reported the first results from STEP Young on 7 September 2026, a Phase 3 trial of once-weekly semaglutide in children aged 6 to under 12 with obesity.1 The headline figure, 40.4% of treated children no longer meeting the BMI threshold for obesity at week 68 against 0% on placebo, is real, and it is also a categorical secondary endpoint reported under the estimand that assumes complete adherence. This article explains what the number measures, how paediatric BMI thresholds work, what the trial registered, and what is still unknown. Condor Research supplies reference materials for laboratory research use only. What follows is commentary on published clinical research in an approved medicine, not medical advice, and contains no dosing guidance for any person.
What did Novo Nordisk actually report?
The company announcement covers 165 children aged 6 to under 12 living with obesity, randomised in a double-blind, placebo-controlled, multinational Phase 3 trial in which both arms received a reduced-calorie diet and increased physical activity.1 Semaglutide was escalated to a maximum dose of either 1.7 mg or 2.4 mg weekly depending on the child’s body weight at baseline. At week 68, 40.4% of children on semaglutide had a BMI below the obesity threshold, against none of the children on placebo. Novo also reported that more than 85% of participants had class II or class III obesity at entry, meaning a BMI at or above 120% or 140% of the 95th percentile.
The trial is registered as NCT05726227.2 The record covers two cohorts, Group Kids aged 6 to under 12 and Group Teens aged 12 to under 18, with a combined estimated enrolment of 210 and 56 sites. It began on 7 July 2023 and reached primary completion on 15 October 2025, with overall study completion estimated for December 2026. The registry does not break enrolment down by cohort, so the 165 figure reported for the younger cohort cannot be reconciled against the record without the study report.
Was 40.4% the primary endpoint?
No. The registered primary endpoint for Group Kids is change in body mass index from week 0 to week 68, expressed as a percentage.2 The categorical measure behind the headline, registered as “improvement in weight category”, is a secondary endpoint, also assessed at week 68 and again at week 104. This is not a criticism of the trial, which registered both prospectively. It is a point about how a result travels: the number that led the announcement, and that will lead most coverage, is not the one the study was designed around, and the announcement did not report the primary endpoint value alongside it.
The threshold result is easier to understand than the primary endpoint, which is exactly why it needs the context the primary endpoint would have supplied.
What does “no longer classified as having obesity” mean in a child?
It means the child’s BMI fell below the 95th percentile for their age and sex on the reference growth chart. Paediatric obesity is not defined against a fixed number the way adult obesity is defined at a BMI of 30. A child’s BMI is compared with a reference distribution that shifts with age, so the same absolute BMI can sit above the threshold at one age and below it at another. Crossing that line is a genuine change in classification, but it is not the same statement as a specific amount of fat loss, and it is not directly comparable to an adult crossing BMI 30.
There is a second mechanism worth naming. Children grow. Over 68 weeks, a child of 7 or 9 gains height, and BMI falls when height rises even if weight is stable. In an untreated population that effect is part of why BMI percentile can drift, and in a treated population it is part of the observed change. Disentangling the contributions of weight change and linear growth requires the full dataset, which is not yet public.
Why was the placebo result exactly zero?
Because of where the children started. With more than 85% of participants at class II or class III obesity, most of the cohort began at 120% to 140% or more of the 95th percentile.1 The distance from there to below the 95th percentile is large. A lifestyle intervention producing a real but modest improvement will move children down within the obesity range without crossing the threshold, and the categorical endpoint records that as no improvement. Zero percent is therefore less surprising than it first appears, and it reflects the severity of the baseline population as much as the ineffectiveness of the comparator. It also makes the contrast visually dramatic in a way a continuous endpoint would not be.
What is a “trial product estimand”, and why does it matter here?
Novo attributes the 40.4% figure to the trial product estimand, described in its own footnote as the treatment effect if all children adhered to treatment.1 An estimand is the precise definition of what a trial estimates, including how it handles events like stopping treatment or starting another one. The two common choices give different answers to different questions. A treatment policy estimand asks what happens to everyone randomised, whether or not they stayed on drug, which is closer to real-world effectiveness. A trial product estimand asks what happens under full adherence, which isolates the pharmacological effect and generally produces a larger number.
Neither is wrong. But a figure quoted without its estimand is incomplete, and the two are not interchangeable. Until the treatment policy result is published, the honest reading of 40.4% is: this is the effect among children who took the drug as intended.
0 peer-reviewed publications of STEP Young existed on 13 September 2026. Detailed results are scheduled for ObesityWeek 2026, 14 to 17 November, in Washington DC.
What was said about safety, growth and puberty?
The company states that overall safety and tolerability were consistent with previous paediatric and adult trials of semaglutide, that no new safety concerns were identified, and that no safety concerns were identified for growth or pubertal development.1 That is a meaningful statement and it is also, at present, a qualitative one. The announcement contains no rates of gastrointestinal adverse events, no discontinuation rates, and no numerical growth or pubertal data.
The reason this cohort raises the question at all is visible in the eligibility criteria. Group Teens requires Tanner stage greater than 1, meaning pubertal development has begun; Group Kids carries no such requirement, so the younger cohort includes prepubertal children.2 Intervening on energy intake during the years preceding the pubertal growth spurt is a different proposition from intervening after it, and the relevant endpoints, height velocity, bone age, pubertal staging over time, are exactly the ones that need numbers rather than reassurance. The registered week-104 assessments will speak to duration in a way week 68 cannot.
How does this compare with the adolescent data?
STEP TEENS, published in the New England Journal of Medicine in 2022, remains the reference point for this drug class in young people.3 In 201 adolescents aged 12 to under 18, mean BMI change at week 68 was -16.1% with semaglutide against +0.6% with placebo, and 73% versus 18% achieved at least 5% weight loss. That trial supported approval in adolescents. What STEP Young adds is not a new mechanism but a younger population, and the questions that come with it are developmental rather than metabolic. For how the wider incretin class is reading out in adults, see our analysis of the TRIUMPH-2 and TRIUMPH-3 results and of what these drugs do to lean mass.
| Item | STEP Young, Group Kids | STEP TEENS |
|---|---|---|
| Age range | 6 to under 12 years | 12 to under 18 years |
| Participants reported | 165 | 201 |
| Duration to primary endpoint | 68 weeks | 68 weeks |
| Primary endpoint | Change in BMI, week 0 to week 68 | Change in BMI, week 0 to week 68 |
| Headline reported | 40.4% below obesity threshold vs 0%, trial product estimand | -16.1% vs +0.6% BMI change |
| Publication status | Company topline only; presentation scheduled November 2026 | Peer reviewed, NEJM 2022 |
| Regulatory position | Outside approved indication in this age group | Supported approval from 12 years |
Figures for STEP Young are from the company announcement of 7 September 2026 and the trial registry; figures for STEP TEENS are from the published paper. The two trials are not designed for head-to-head comparison and the endpoints quoted are not the same measure.
What is the regulatory position for this age group?
Semaglutide for weight management is approved from 12 years of age by the FDA and by the EMA, in both cases in adolescents meeting a BMI percentile criterion.4 Children aged 6 to 11 fall outside the approved indication in both jurisdictions. A positive Phase 3 result is the beginning of a regulatory conversation, not the end of one, and any extension of the indication would require the full dataset, agency review and a paediatric assessment that looks specifically at growth and development. Nothing in the September announcement changes what is currently authorised anywhere.
What is established, and what is not
Established: in a randomised, placebo-controlled Phase 3 trial in 165 children aged 6 to under 12 with predominantly severe obesity, 68 weeks of weekly semaglutide plus lifestyle intervention moved 40.4% of treated children below the obesity threshold under the trial product estimand, against none on placebo plus the same lifestyle intervention.
Not established from the public record: the primary endpoint value, the treatment policy result, adverse event and discontinuation rates, any numerical data on height velocity or pubertal staging, the durability of the effect after treatment stops, and what happens at week 104. Not addressed at all: what this population looks like at 16, or at 30. Weight regain after discontinuation is well described in adults on this drug class, and there is no reason to assume children are exempt; the question in a growing child is not only whether weight returns but what the intervening period did to a developmental trajectory. The evidence that would resolve most of this is the full study report, expected in November 2026.
How this was checked. The company announcement of 7 September 2026 was read in the original, and the trial record NCT05726227 was read directly from the ClinicalTrials.gov API on 13 September 2026, which is the source for the endpoint structure, cohort definitions, Tanner stage criterion, site count and study dates. STEP TEENS figures come from the published NEJM paper. Where the announcement and the registry do not reconcile, in particular the 165 participants reported against a combined estimated enrolment of 210, this page reports the discrepancy rather than resolving it. Version 1.0, first published 13 September 2026; this page will be updated after the detailed results are presented.
Condor Research supplies characterised reference materials for laboratory research use only: not for human or veterinary use, not for diagnostic or therapeutic application, and not for any food or cosmetic purpose. We do not supply semaglutide. This article is scientific commentary on a clinical trial in an approved medicine and is not medical advice, not a recommendation, and not a source of dosing information for any person. Decisions about the treatment of a child belong with that child’s clinician.
Condor Research · Scientific desk
Atrio Sciences s.r.o., IČO 57 669 651, Nitra (SK) · info@condorresearch.com
- Novo Nordisk announced first STEP Young results on 7 September 2026: 40.4% of treated children below the obesity threshold at week 68 versus 0% on placebo.
- 165 children aged 6 to under 12, randomised double-blind with both arms receiving a reduced-calorie diet and increased physical activity; semaglutide escalated to a maximum of 1.7 mg or 2.4 mg weekly by baseline body weight.
- The registered primary endpoint is change in BMI from week 0 to week 68; the threshold-crossing measure is a registered secondary endpoint, assessed at week 68 and week 104.
- The figure is reported under the trial product estimand, described by the company as the effect if all children adhered to treatment, which generally yields a larger number than a treatment policy estimand.
- More than 85% of participants had class II or class III obesity at entry, at or above 120% or 140% of the 95th percentile, which explains why no child on placebo crossed back below the 95th percentile.
- Paediatric obesity is defined against age- and sex-specific reference distributions, so crossing the threshold is not equivalent to an adult crossing BMI 30, and height growth over 68 weeks lowers BMI independently of weight change.
- The trial registry, NCT05726227, covers two cohorts with a combined estimated enrolment of 210 across 56 sites, and does not break enrolment down by cohort.
- The adolescent cohort requires Tanner stage above 1 while the younger cohort does not, so Group Kids includes prepubertal children.
- Safety reporting is qualitative: no new safety concerns and none identified for growth or pubertal development, with no adverse event rates, discontinuation rates or numerical growth data released.
- Semaglutide for weight management is approved from age 12 by both FDA and EMA, so the 6 to 11 age group falls outside the approved indication in both jurisdictions.
What does it mean that 40% of children were no longer classified as having obesity?
It means their BMI fell below the 95th percentile for their age and sex on the reference growth chart. Paediatric obesity is defined against a distribution that shifts with age, not against a fixed value like the adult threshold of BMI 30. Crossing that line is a real change in classification, but it does not translate directly into a stated amount of fat loss, and part of the change in a growing child comes from gains in height rather than reductions in weight.
Why did 0% of children on placebo improve?
Because more than 85% of the cohort entered with class II or class III obesity, meaning at or above 120% or 140% of the 95th percentile. From that starting point, a lifestyle intervention can produce a genuine improvement without moving a child all the way below the 95th percentile, and the categorical endpoint counts that as no improvement. The zero reflects the severity of the baseline population as much as the weakness of the comparator.
What is a trial product estimand?
It is a precise definition of what the trial estimates. The trial product estimand describes the effect under full adherence, isolating the pharmacological effect. A treatment policy estimand describes what happens to everyone randomised regardless of whether they continued treatment, which is closer to real-world effectiveness and generally yields a smaller number. The 40.4% figure is reported under the trial product estimand, and the treatment policy result has not been published.
Are the STEP Young results peer reviewed?
Not yet. The 7 September 2026 announcement is corporate topline data. Novo Nordisk has stated that detailed results will be presented at ObesityWeek 2026, from 14 to 17 November in Washington DC. Until the full dataset is available, the primary endpoint value, safety rates and growth data remain unpublished.
Is semaglutide approved for children under 12?
No. Semaglutide for weight management is authorised from 12 years of age by both the FDA and the EMA, subject to BMI percentile criteria. Children aged 6 to 11 are outside the approved indication in both jurisdictions, and a positive Phase 3 result does not change what is currently authorised.