Comparing GLP-1 based weight-loss medicines: a 2026 network analysis

Why this study matters

There are now many medicines that work through the GLP-1 hormone pathway, and few trials compare them directly. A network meta-analysis tries to fill that gap. It combines trials that compared each medicine with placebo, or with another medicine, and uses those shared links to estimate how the medicines compare. Canadian readers may see news stories based on studies like this one. It is useful context, but it is not a ranking and it does not replace head-to-head trials. Some of the medicines in it are not authorized in Canada.

Who was in the study

The review included 58 randomized controlled trials with 24214 participants. All were adults with overweight or obesity. Trials that enrolled people with diabetes were left out, as were trials where diabetes status was not clear. Each trial had to last at least 12 weeks. The authors searched three research databases, Embase, PubMed and Web of Science, for trials published from 1 January 2000 to 6 March 2026.

What the researchers did

The authors included trials comparing GLP-1 receptor agonists, or related medicines that act on more than one hormone (co-agonists), with placebo or with another medicine. Placebo is an inactive treatment with no medicine in it. Randomized means people in each trial were assigned to a group by chance.

They then ran a network meta-analysis. A regular meta-analysis pools trials of one medicine against one comparison. A network meta-analysis links many trials together. For example, if medicine A was tested against placebo in one trial, and medicine B against placebo in another, the method estimates how A and B compare, even though no trial tested them side by side. These are called indirect comparisons. The authors also rated how certain each result was.

The full paper says the study plan was registered with PROSPERO, a public register of reviews, after the literature search was finished rather than before.

What they found

Compared with placebo, the estimated average change in body weight was as follows (R33-C12). A confidence interval, or CI, is the range of values the data are reasonably consistent with. The table is in alphabetical order, not in order of results.

CagriSema combines cagrilintide with semaglutide, the medicine in Wegovy and Ozempic. The authors write that older GLP-1 receptor agonists, which they call conventional, showed more modest effects. The abstract does not give the numbers for those medicines. Waist size and blood fat (lipid) results followed a similar pattern.

The authors also calculated a ranking of the medicines. We do not show it. They note that the confidence intervals overlapped for several comparisons, and a ranking can make small or uncertain differences look certain. The authors concluded that newer medicines acting on more than one hormone led to more weight loss than conventional GLP-1 receptor agonists, and that differences in tolerability, the small number of head-to-head trials and remaining uncertainty should be considered.

Side effects and people who stopped

The abstract reports, based on low certainty evidence, that more people stopped treatment with danuglipron and retatrutide, while mazdutide showed better tolerability (R33-C13). It does not give the numbers. The full paper notes that the results for stopping because of side effects were not consistent across the network, which lowers confidence in those comparisons. The published summary we reviewed does not report serious side effects for each medicine.

What this study does not tell you

• Most comparisons are indirect. They are estimates built from separate trials with different people, doses and lengths, not results from one trial.
• The confidence intervals overlapped for several comparisons, so the order of the results is not certain.
• The full paper says some medicines were represented by only one trial, and that long-term safety data for newer medicines were limited.
• It only covers adults without diabetes, and it does not cover cost, access or how people do outside trials.
• The study plan was registered after the search was done, which makes it harder to check that the methods were set in advance.
• The summary we reviewed does not state the funding source or the authors' conflicts of interest.

What it means in Canada

This analysis gives an overview of trial results, not a guide to choosing a medicine. As of October 2026, retatrutide is not approved in Canada or anywhere else, and CagriSema, danuglipron and mazdutide are not authorized in Canada. For medicines available in Canada, see semaglutide in Canada, tirzepatide in Canada, our medicines page, public drug coverage by province and compare services.

Related reading

Other studies in this library that bear on the same question.

Source