GLP-1 Weight-Loss Drugs: Trial Review

A 2026 systematic review combined 24 phase 3 randomized trials of liraglutide, injectable semaglutide and tirzepatide in adults with overweight or obesity. Twenty-three trials entered its network meta-analysis. Every active treatment reduced body weight more than placebo, while tirzepatide 15 mg ranked first for percentage weight change. That ranking is not the same as a 23-trial head-to-head contest. Most estimates joined separate trials through a shared comparator, and efficacy results varied substantially across studies^[2].

What the review actually compared

The authors searched for phase 3 randomized controlled trials lasting at least 52 weeks. Eligible participants were adults with a body mass index of 25 kg/m² or higher. The review examined approved weight-management doses against placebo or another eligible treatment. Four liraglutide trials, 13 semaglutide trials and five tirzepatide trials were included, plus two direct-comparison trials. The network analysis used random-effects models and reported treatment rankings with P-scores^[2].

One direct trial compared semaglutide 2.4 mg with liraglutide 3.0 mg. Another compared tirzepatide with semaglutide, but it was excluded from the quantitative network because the dose reporting did not support connection to the model. Its findings were discussed descriptively instead. The remaining comparisons often relied on indirect evidence, so the network estimates should not be read as if every medicine and dose had been randomized within one protocol^[2].

How the medicines ranked for weight change

All assessed doses produced a significantly greater percentage reduction in body weight than placebo. Tirzepatide 15 mg ranked first by P-score in the main analysis and every subgroup analysis. In the main network, it produced a significantly greater reduction than the other interventions except semaglutide 7.2 mg. Among participants with diabetes, the comparison with semaglutide 7.2 mg also favored tirzepatide 15 mg significantly^[2]. The result describes trial groups and does not predict an individual's response.

Semaglutide 7.2 mg ranked second for percentage weight reduction in the overall network. It ranked above 2.4 mg across efficacy endpoints, but the difference in percentage weight change was statistically significant only in the diabetes subgroup. The authors described the average incremental difference as roughly 2 to 3 percentage points and questioned whether that would be clinically meaningful for most patients. Liraglutide 3.0 mg produced the smallest weight reduction among the active treatments evaluated^[2]. Rankings alone do not show how much two options differ or whether that difference is precise.

What the safety analysis found

The safety networks included 18 trials for any adverse event, 23 for serious adverse events, 23 for discontinuation because of adverse events and 12 for injection-site reactions. The analysis found no significant differences between treatments for serious adverse events. Liraglutide 3.0 mg had a higher risk of any adverse event than tirzepatide 15 mg and placebo, and a higher discontinuation risk than tirzepatide 5 mg, semaglutide 2.4 mg and placebo^[2].

Injection-site reactions were more frequent with all three tirzepatide doses than with either semaglutide dose, liraglutide 3.0 mg or placebo. The network found no significant safety difference between semaglutide 7.2 mg and 2.4 mg for the assessed outcomes, although the point estimates suggested more adverse events and discontinuations at the higher dose. A nonsignificant trend is not proof that the doses are equally safe, particularly when long-term evidence for the newer dose remains limited^[2].

Why the ranking needs restraint

Efficacy heterogeneity was considerable. I² was 92% for percentage weight change, 86% for waist circumference and 92.4% for BMI. The trials enrolled different populations, including people with or without diabetes and groups selected for cardiovascular disease, heart failure or knee osteoarthritis. Follow-up, background care and participant characteristics also differed. The authors noted that diabetes status did not explain all the remaining variation^[2].

Most included trials were judged at low risk of bias, while four raised some concerns. The network showed limited evidence of broad inconsistency between direct and indirect estimates, but that does not remove the heterogeneity or create missing head-to-head evidence. A P-score expresses a ranking probability within the assembled network; it does not by itself establish a clinically important advantage. The authors called for further direct randomized trials and real-world follow-up^[2].

What remains unknown

This analysis does not settle which treatment is best for a particular patient. It cannot fully separate medicine effects from differences among trial populations, and it offers less evidence for long-term safety than for weight change over one to two years. It also does not establish that a higher-ranked dose provides enough added benefit to outweigh tolerability, contraindications or monitoring needs. The review reported no author conflicts of interest, but its evidence base consisted of previously published trials with their own sponsorship and protocol constraints^[2].

The dose names above report research arms, not instructions. Liraglutide, semaglutide and tirzepatide have different indications, warnings and eligibility requirements. A qualified clinician must assess whether any approved treatment is appropriate, then use current product information and local guidance rather than a network ranking.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.