BELIEVE is the first properly published trial of a drug given specifically to protect muscle while a GLP-1 removes weight. It ran for 72 weeks, randomised 507 adults with obesity across nine groups, and reported in Nature Medicine in March 2026.
The headline is genuinely striking. The detail is more interesting than the headline.
What bimagrumab is
An antibody that blocks activin type II receptors — part of the signalling system that tells muscle to stop growing. Block it and muscle grows, or at least stops shrinking, largely independently of whether you are training.
It is given by infusion, not injection: in this trial, intravenously every 12 weeks. That matters for any picture you have of how it would be used.
The three results that matter
| Arm | Weight lost | Share of it from fat | Lean mass |
|---|---|---|---|
| Bimagrumab alone | 10.8% | 100% | up 2.5% |
| Semaglutide alone | 15.7% | 71.8% | down |
| Both together | 22.1% | 92.8% | largely preserved |
Read the middle row first, because it is the one that describes almost everyone taking these medications today. On semaglutide alone, 71.8% of the weight lost was fat — so 28.2% of it was not.
That figure is a direct body-composition measurement over 72 weeks, and it sits almost exactly on the 30% baseline our own muscle check has been using. Our formula was built from earlier trials; this is a newer and more direct measurement arriving at the same place, which is about as good a corroboration as an estimate can get.
What the combination actually achieved
Two things at once, and it is worth separating them.
It produced more weight loss — 22.1% against 15.7%. And it changed what the weight was made of, from 71.8% fat to 92.8% fat.
The second is the point. The first is a bonus that, on its own, would be unremarkable — plenty of things produce more weight loss. Very few change the composition of it.
The trial also reported large falls in inflammatory markers, with hsCRP down by up to 83% in combination groups, and among prediabetic participants some combination groups saw complete reversion to normal glucose regulation.
The parts usually left out
It is a Phase 2 trial. 507 people across nine arms means each individual group is small. Phase 2 results are a reason to run Phase 3, not a conclusion.
Side effects tracked the mechanisms. Muscle spasms and acne were prominent with bimagrumab; gastrointestinal problems with semaglutide. Discontinuations were highest in the bimagrumab monotherapy group — the drug is not free of cost to the people taking it.
Bimagrumab has a history. It has been studied in other conditions before now and did not become a treatment in them. A promising body-composition result does not by itself make a medicine.
Lean mass is not the same as function. The trial measured tissue. It did not show that people were stronger, fitter, or healthier for longer as a result. That is the study that has not been done yet, and it is the one that would actually settle whether this matters.
What it changes for you today
Practically, nothing. It is not approved for weight management, it is not prescribable for this, and it is an infusion in a clinical trial.
What it changes is the argument. The composition of GLP-1 weight loss is now something a major trial has measured directly and a pharmaceutical company has spent a great deal of money trying to fix. If anyone tells you the muscle concern was invented by the internet, this is the answer.
The thing worth noticing
Bimagrumab on its own increased lean mass by 2.5%.
Resistance training does that too, in people who start doing it, at no cost and with no infusion. The drug’s real contribution in this trial is that it worked without requiring anyone to train — which is a genuine achievement for people who cannot, and a smaller one for people who simply have not started.
If you are in the second group, the lever is already available to you. The muscle check will tell you what it is currently costing you not to pull it.