The main finding
In one sentence: A 2026 update of a systematic review of 38 randomized trials in 25816 adults without diabetes found placebo-subtracted weight loss of up to -19.0% with tirzepatide and -14.8% with injected semaglutide, with common but mostly non-serious side effects.
A pooled result summarizes several studies that differed in who was enrolled, how long they ran and what they compared. It is not a prediction for one person.
On this page
Why this study matters
This review brings together the randomized trials of GLP-1 receptor agonists and related medicines that act on more than one hormone (co-agonists) in adults without diabetes. It updates the same team's earlier review and adds trials published since then. The authors are based in Montreal, at McGill University and the Lady Davis Institute of the Jewish General Hospital, and the review reports no funding source. It gives Canadian readers one place to see what the trials report. It is not a ranking of medicines.
Who was in the study
The review included 38 randomized controlled trials with 25816 participants, all adults with overweight or obesity and without diabetes. This update added 14 new trials with 11000 participants to the earlier review. Each trial had to treat people for at least 16 weeks. For this update, the authors searched MEDLINE, Embase and the Cochrane Central Register of Controlled Trials from 5 October 2024 through 25 March 2026.
What the researchers did
This was a systematic review. A systematic review searches for every study that meets set rules and summarizes them in a planned way. Only randomized controlled trials were included. In these trials, people were assigned by chance to a medicine or a comparison, often placebo, an inactive treatment with no medicine in it. Two reviewers pulled out the data separately, which helps catch errors.
The authors did not pool the trials into a single average for each medicine. They say the trials were too different from each other (heterogeneity) to combine. Instead, they report results from the individual trials.
What they found
Results are given as placebo-subtracted weight loss: the weight change with the medicine minus the weight change with placebo in the same trial. For each medicine, the review gives the largest result reported ("up to"). These are not averages across all trials. A confidence interval, or CI, is the range of values the data are reasonably consistent with.
Among medicines the authors describe as commercially available, placebo-subtracted weight loss reached up to (R33-C14):
| Medicine | Up to (95% CI) |
|---|---|
| Liraglutide (Saxenda, Victoza) | -5.8% (-8.0% to -3.6%) |
| Semaglutide (Wegovy, Ozempic), by injection | -14.8% (-16.2% to -13.4%) |
| Semaglutide, by mouth | -14.3% (-17.2% to -11.4%) |
| Orforglipron | -12.4% (-15.1% to -9.7%) |
| Tirzepatide (Zepbound, Mounjaro) | -19.0% (-21.6% to -16.4%) |
Orforglipron is on this list because the authors describe it as commercially available. As of October 2026, it is not authorized in Canada.
The review reports numerically larger reductions with some medicines still in research: up to -23.9% (95% CI -29.3% to -18.5%) with amycretin and -22.1% (95% CI -24.9% to -19.3%) with retatrutide (R33-C15). "Numerically larger" means the number is bigger, without a formal test showing the difference is real.
Trials that compared medicines directly found more weight loss with semaglutide and with an investigational medicine called JNJ-64565111 than with liraglutide. They also found more weight loss with tirzepatide and with CagriSema (cagrilintide plus semaglutide) than with semaglutide. The abstract does not give these numbers.
Side effects and people who stopped
Across the trials, the review reports these figures for GLP-1 medicines compared with placebo (R33-C16):
| Outcome | GLP-1 medicines | Placebo |
|---|---|---|
| Stomach and bowel side effects | 76.0% | 40.1% |
| Stopping because of side effects | 10.7% | 3.4% |
| Serious side effects | 6.5% | 5.2% |
| Deaths | 0.1% | 0.0% |
The authors describe stopping because of side effects as generally low, but numerically higher with some medicines taken by mouth. They describe serious side effects and deaths as rare and say no new safety signals were found. They also note that safety outcomes were reported inconsistently across trials.
What this study does not tell you
- The trials were not pooled. The "up to" numbers are the largest results for each medicine, not typical results, and they come from trials of different lengths and doses.
- Placing these numbers side by side is not a fair comparison between medicines. Only direct head-to-head trials can show that.
- Side effects were reported differently from trial to trial, which limits what the safety numbers can show.
- The abstract does not say how the overall side effect percentages were calculated across trials.
- It only covers adults without diabetes.
- The review does not cover cost, access or how people do outside trials.
What it means in Canada
This Montreal-led review gives an overview of the trial evidence. It does not describe how any medicine is prescribed or paid for here. As of October 2026, orforglipron is not authorized in Canada; it has been under Health Canada review since January 2026, and the US FDA approved it as Foundayo for weight management on 1 April 2026. Retatrutide is not approved anywhere. Amycretin, CagriSema and JNJ-64565111 are not authorized in Canada. See semaglutide in Canada, tirzepatide in Canada, our medicines page, public drug coverage by province and compare services.
Common questions
How much weight loss did the 2026 Annals review find with GLP-1 medicines?
Placebo-subtracted weight loss reached up to -5.8% with liraglutide, -14.8% with injected semaglutide, -14.3% with oral semaglutide, -12.4% with orforglipron and -19.0% with tirzepatide. These are results reported in the trials, not pooled averages.
How common were side effects with GLP-1 medicines in this review?
Stomach and bowel side effects were reported in 76.0% of people on GLP-1 medicines and 40.1% on placebo. Stopping because of side effects was 10.7% versus 3.4%.
Who did the 2026 Annals review of GLP-1 medicines?
The authors are based at McGill University and the Lady Davis Institute at the Jewish General Hospital in Montreal. The review reports no funding source.
Related reading
Other studies in this library that bear on the same question.
Comparing GLP-1 based weight-loss medicines: a 2026 network analysis
What a Cochrane review says about semaglutide for weight loss
What a Cochrane review says about tirzepatide for weight loss
STEP 8: weekly semaglutide versus daily liraglutide, head to head
ATTAIN-1: the orforglipron pill for weight loss over 72 weeks
Source
Original paper: Areesha Moiz, Updated systematic review of GLP-1 RAs and co-agonists for weight loss without diabetes, Annals of internal medicine, 2026. PubMed: https://pubmed.ncbi.nlm.nih.gov/42673585/. DOI: https://doi.org/10.7326/ANNALS-25-05519. Funding: The review reports no funding source. How we summarized it: abstract and extracted data.
This page explains a published study. It is not medical advice and does not describe whether a medicine is right for you. Talk to your doctor, nurse practitioner or pharmacist about your own situation. Reviewed and signed by two pharmacists registered in British Columbia, 2026-10-04.
Evidence records behind this page
Each record is a row in the clinical evidence file, taken from the published paper named above. The caution column is the limit the researcher recorded for that number.
This table scrolls sideways.
| Record | What the paper reports | Type | Population | Caution | Source |
|---|---|---|---|---|---|
| R33-C14 | An updated systematic review of 38 RCTs (25816 adults without diabetes) reported placebo-subtracted weight loss reaching up to -5.8% (95% CI -8.0 to -3.6) for liraglutide, -14.8% (-16.2 to -13.4) for subcutaneous semaglutide, -14.3% (-17.2 to -11.4) for oral semaglutide, -12.4% (-15.1 to -9.7) for orforglipron and -19.0% (-21.6 to -16.4) for tirzepatide. | cross-trial context | Adults with overweight or obesity without diabetes | No quantitative synthesis (heterogeneity); 'up to' values are best single-trial results, not pooled means, and are not a fair cross-medicine comparison. Orforglipron described as commercially available by authors but not authorized in Canada as of October 2026. Abstract-only. | PubMed 42673585 |
| R33-C15 | The same review reported numerically greater placebo-subtracted weight loss with emerging multiagonists, including up to -23.9% (95% CI -29.3 to -18.5) with amycretin and -22.1% (-24.9 to -19.3) with retatrutide. | cross-trial context | Adults with overweight or obesity without diabetes | 'Numerically greater' is not a formal comparison; single-trial maxima. Amycretin not authorized in Canada; retatrutide not approved anywhere as of October 2026. Abstract-only. | PubMed 42673585 |
| R33-C16 | Across included trials, gastrointestinal adverse events occurred in 76.0% with GLP-1 RAs versus 40.1% with placebo, discontinuation due to adverse events in 10.7% versus 3.4%, serious adverse events in 6.5% versus 5.2%, and deaths in 0.1% versus 0.0%. | cross-trial context | Adults with overweight or obesity without diabetes | Safety outcomes inconsistently reported across trials; method for these overall percentages not described in abstract. Abstract-only. | PubMed 42673585 |