Peptides that activate the glucagon-like peptide-1 receptor have become one of the more studied drug classes in metabolic medicine over the past decade. Most of that research has focused on adults. A 2026 network meta-analysis published in Pharmacological Research set out to fill a gap by comparing how different GLP-1 receptor agonists perform in children and adolescents who carry a diagnosis of overweight or obesity.
The researchers pooled data from 17 randomized controlled trials covering 1,230 participants. They used a Bayesian statistical framework, which is a way of combining indirect comparisons across studies when head-to-head trial data between two specific agents does not yet exist. The goal was to rank the agents on several cardiometabolic measurements: body weight, body mass index, waist circumference, blood sugar control, fasting glucose, systolic blood pressure, and lipid profiles.
The authors were careful to note that most of the comparisons between agents were indirect, meaning the trials did not pit one peptide against another directly. That matters when reading the results. Rankings like these are, in the authors' own words, hypothesis-generating rather than proof of definitive superiority.
What a network meta-analysis actually does
A standard meta-analysis combines results from trials that tested the same treatment against the same comparator, usually a placebo. A network meta-analysis goes a step further. It weaves together trials that share at least one common comparator, even if the active agents being tested differ across those trials. This creates a web, or network, of indirect evidence.
The benefit is a broader picture. The risk is that indirect comparisons carry more uncertainty than direct ones. The Bayesian approach used in this paper produces credible intervals, which are similar to confidence intervals and represent the range of values the true effect likely falls within. When those intervals are wide, as several were in this analysis, the ranking of agents should be treated with caution.
Weight and body composition findings
On body weight, semaglutide at a 2.4 mg subcutaneous dose was associated with the largest reduction among the agents studied. The mean difference versus placebo was approximately 18 kg, with a 95 percent credible interval running from about 11.7 kg to 24.3 kg less than placebo. The analysis rated this finding as high confidence.
The same agent produced the largest point estimate for BMI reduction, at roughly 5.9 kg per square meter, also rated high confidence. Waist circumference fell by an estimated 12.2 cm. These are the kinds of measurements researchers use to track changes in body composition and central adiposity, both of which are linked to longer-term cardiometabolic risk in epidemiological literature.
It is worth noting that semaglutide 2.4 mg is a higher dose than is used in some other contexts, and the trial evidence at this dose in adolescents is still accumulating. The authors did not suggest these findings translate directly into clinical decisions.
Blood sugar control across agents
For glycemic outcomes, a different agent came out ahead. Dulaglutide at 1.5 mg subcutaneous was associated with the largest reduction in HbA1c, a measure of average blood sugar over roughly three months. The estimated reduction was 1.5 percentage points, again rated high confidence, though the authors noted that the network for glycemic endpoints had sparse head-to-head evidence.
Fasting plasma glucose, a snapshot measure of blood sugar taken after an overnight fast, showed the largest reduction with lixisenatide at 20 micrograms subcutaneous, at approximately 3.98 mmol per liter. However, the credible interval for that estimate was wide, ranging from 0.60 to 7.46, which means the true effect could be considerably smaller or larger than the point estimate suggests.
Most of the adolescents enrolled in these trials had obesity with or without type 2 diabetes. GLP-1 receptor agonists work partly by stimulating insulin release in a glucose-dependent manner and partly by slowing gastric emptying, both of which can lower post-meal blood sugar. The specific pharmacokinetic profile of each agent, including how long it stays active and how often it is dosed, likely explains some of the differences in outcomes observed across agents.
Blood pressure and lipid results
Systolic blood pressure, the top number in a blood pressure reading, showed a reduction with exenatide at 20 micrograms subcutaneous. The estimated reduction was about 6.6 mmHg. The authors were explicit that this finding came from limited indirect evidence drawn from early-phase trials, so it should be interpreted cautiously.
No agent produced a statistically significant improvement in lipid profiles, meaning measurements like total cholesterol, LDL, HDL, or triglycerides did not shift meaningfully in either direction across the trials included. Whether longer treatment durations or different dosing strategies might alter that picture is not something the current evidence can answer.
Limitations researchers flagged
The authors were unusually transparent about the constraints of their findings. They pointed out that most active treatment comparisons were indirect, that many of the nodes in the network were informed by a single trial, and that the SUCRA rankings, a statistical tool for ranking treatments by probability of being best, should not be read as evidence of definitive clinical superiority.
With only 1,230 participants spread across 17 trials and multiple agents, individual comparisons can be underpowered. A single large trial can dominate a node and skew rankings. The paper also covered participants with and without type 2 diabetes, which introduces heterogeneity because the baseline metabolic state differs between those groups.
The 17 trials were identified through searches across PubMed, Embase, Web of Science, the Cochrane Library, and ClinicalTrials.gov up to May 2026. Two independent reviewers extracted data and assessed risk of bias following PRISMA guidelines, which is a recognized standard for systematic reviews.
Why the research community is watching this space
Adolescent obesity is associated with early onset of metabolic risk factors that can track into adulthood. Cardiometabolic markers like elevated BMI, high HbA1c, and elevated blood pressure in young people have been linked in epidemiological studies to increased long-term risk of cardiovascular disease and type 2 diabetes. That context explains why researchers are interested in understanding how peptide-based interventions compare across these specific endpoints in younger populations.
GLP-1 receptor agonists as a class act on receptors found not only in the pancreas but also in the brain, gut, and heart. The differential effects seen across agents in this meta-analysis may reflect differences in receptor binding affinity, duration of action, route of administration, or dose, all factors that are actively studied in the pharmacology literature.
For now, the network meta-analysis adds a layer of organized comparison where scattered trial data existed before. Researchers and clinicians working in pediatric metabolic medicine will likely use these rankings as a starting point for designing future head-to-head trials rather than as a final answer. The authors themselves described the findings as hypothesis-generating, which is an honest and accurate framing of what indirect network evidence can and cannot tell us.
