A GLP-1 muscle drug now has real trial data behind it. In the 507-person BELIEVE trial, adding bimagrumab to semaglutide held lean-mass loss to 2.3%, against 6.9% on semaglutide alone (Trial). The same trial measured grip strength. Grip in that top group rose 2.3 kilos, and the placebo group gained 1.7 kilos (Trial).
So the scan changed a lot and the strength test barely moved. That gap is what the rest of the evidence keeps showing. A year ago the answer to muscle loss on these drugs was protein and lifting. Now there is a drug answer too, at least on paper.
What a GLP-1 Muscle Drug Is
Bimagrumab is a lab-made antibody, and it is still experimental. It is not a pen you use at home. It goes in through a drip at a clinic, over about 30 minutes. The trial gave loading doses in weeks 1 and 4, then one dose every 12 weeks (Trial).
A GLP-1 muscle drug of this kind blocks a switch called the activin type II receptor. Think of that receptor as a brake pedal on muscle growth. Block it and the brake comes off, so muscle mass goes up.
A GLP-1 drug works in a completely different way. It acts on appetite, so you eat less and lose weight. Bimagrumab was built for muscle disease instead. It acts on muscle and fat tissue directly (Trial).
The same family includes drugs aimed at myostatin, another brake on muscle growth. One of them, apitegromab, was approved in the United States in September 2026. That approval covers only spinal muscular atrophy, a rare inherited disease. It also applies only to people already on another treatment for it (Announcement). No drug in this class is approved anywhere for keeping muscle during weight loss.
What the BELIEVE Trial Found
BELIEVE ran at 26 centers in the United States, Australia and New Zealand. A coin-flip draw sorted 507 adults with obesity into nine groups. Some got bimagrumab, some semaglutide, some both, some a placebo drip. For the first 48 weeks nobody knew who was getting bimagrumab. Then came a 24-week extension where everyone knew what they were taking. That carried the trial to week 72 (Trial).
The numbers everybody quoted came from the end of that extension. About 22% of body weight gone, roughly 92% of it from fat, and lean mass down about 3%. Those are week-72 figures, and they come from the single highest-dose group. The paper itself labels every week-72 analysis as an after-the-fact look (Trial).
The checkpoint the trial registered in advance was week 48. The goal it registered was plain body weight (Registry). At that point the top combination group had lost 17.8 kilos, and semaglutide alone had lost 14.2 kilos. Lean mass was down 2.3% against 6.9%, and fat mass fell 33.7% against 21.1% (Trial).
The longer follow-up changed less than you might expect. Among the people who stayed on treatment, weight loss deepened from about 20% to about 22%. The share of weight lost as fat barely moved, from 92.3% to 92.2% (Trial).
One result was genuinely odd. People on the higher bimagrumab dose alone gained lean mass while losing weight. Their scans showed lean tissue up about 2.5%, and it held from week 48 through week 72. They lost less weight overall than the semaglutide groups. Every kilo they did lose was fat (Trial).
Mass Is Not Strength
BELIEVE did measure strength, and that part got far less attention. Every participant squeezed a hand dynamometer, a small device that records grip force in kilos (Trial).
The grip results were thin. Of the eight active groups, only one clearly beat placebo, and it was not the group with the best body scans. It was the high bimagrumab dose paired with the lower semaglutide dose. The headline combination improved by 2.3 kilos of grip, against 1.7 kilos on placebo (Trial).
The questionnaires were mixed. On a general health survey, everyday physical function improved about the same in every group at week 48, placebo included. A second questionnaire asked how weight itself affects daily life. That one did favor the higher-dose combinations. By week 72 the top combination scored best of all on both. The extension had no placebo group left, though, so its late scores carry less weight (Trial).
The authors say this plainly in their own limitations. They note the lack of meaningful improvement in grip strength and in what patients reported (Trial). A narrower group of patients might respond better, they suggest. On the one outcome a person can actually feel, this trial found very little.
This is not the first time. Bimagrumab was tested years ago in inclusion body myositis, a muscle-wasting disease. The RESILIENT trial sorted 251 adults into three drug doses or a placebo drip. Doses went in every four weeks for at least a year. Lean body mass rose at the two higher doses, and rose more at the higher one. The main goal was how far people could walk in six minutes, and that did not improve. Thigh strength and grip strength did not improve either (Trial).
Two more years of treatment did not change that. Walking distance kept declining in every group, drug included. The extension was stopped early because the main trial had missed its goal (Trial).
A later review sorted this field into two columns. Across the two trials it pooled, lean mass and thigh muscle volume improved. Strength testing, walking distance and timed mobility did not (Systematic review).
A 2020 editorial on this drug class saw it coming. Blocking these pathways adds mass on a scan, the authors wrote, with no dependable link to strength (Editorial). They added one caveat: the trials so far were short, and most used a single dose.
What the Scan Actually Measures
Part of the problem is the scan itself. DXA is an X-ray body scan that sorts you into fat, bone and lean tissue. A lean reading is not pure muscle. Water, organs and connective tissue land in the same bucket, and all of them shift during weight loss (Review). Fat-free mass is more than double the amount of actual skeletal muscle you carry (Review).
Fat tissue has a wrinkle of its own: it is only about 80% to 85% fat by weight. The rest is water and protein, which a scan files as lean. So losing a lot of fat always costs you some lean reading, even when no muscle goes with it. In one worked example, correcting for that turned an apparent lean loss into a small lean gain (Review).
When researchers used MRI instead of DXA, the picture calmed down. In 246 adults with type 2 diabetes on tirzepatide, thigh muscle loss over a year matched what population data predict for that much weight loss. Fat inside the muscle improved more than predicted (Study). On the top dose, one weight-adjusted muscle score did slip further than expected. This was also an after-the-fact look at scans, rather than a planned goal.
The starting premise is shakier than the headlines suggest. In the SEMALEAN study, 115 adults started semaglutide and 106 finished the year. Grip strength went up, by about 4 kilos. The share of people with low muscle plus high fat fell from 49% to 33% (Study). There was no comparison group, so the gain cannot be pinned on the drug alone.
The Side Effects Are Real
Bimagrumab is not a free add-on. Muscle spasms hit 46% of people on the lower dose and 74% on the higher one, against 5.5% on placebo. In the combination groups spasms ran between 57% and 64% (Trial).
Acne and diarrhea were common too. Acne affected up to 44% of people on bimagrumab alone and up to 55% on the combination, against 3.6% on placebo. Diarrhea ran between 41% and 55% across the drug groups, though semaglutide causes plenty of that by itself (Trial).
People quit over it. Between 14% and 21% of the bimagrumab-only groups stopped because of side effects. On semaglutide alone the figure was under 9%, and on placebo 3.6%. Blood tests moved as well. Liver enzymes rose for a time, and so did a blood marker of muscle turnover (Trial).
This is not a BELIEVE quirk. Four trials of the drug have been pooled, with 268 people between them. Muscle spasms were more than ten times as likely as on placebo, and diarrhea nearly five times as likely. Quitting for side effects was several times more common. Serious problems did not show up more often, though trials this small cannot settle that. LDL cholesterol rose slightly (Meta-analysis).
The two drugs also differ in how you take them. Semaglutide is a weekly injection you give yourself. Bimagrumab is a 30-minute drip at a clinic, with a visit every 12 weeks (Trial).
The Rest of the Pipeline
Three other drugs are chasing the same idea, and all of them sit at the same early stage.
Enobosarm, a pill, was added to semaglutide for 16 weeks in 168 adults over 60. Lean soft tissue fell 1.2%, against 4.1% on semaglutide alone. Fewer people on it had a big drop in stair-climbing power. That is the only hint of a function benefit in the group (Review).
Apitegromab was tested with tirzepatide in 102 people over 24 weeks. The combination kept about 2 kilos more lean mass than tirzepatide alone. Grip and sit-to-stand tests were measured as a side question, and showed no clear difference (Trial).
Trevogrumab, alone or with a second antibody, is being tested with semaglutide in the ongoing COURAGE trial. Over 26 weeks in about 600 people, lean-mass loss was 6.5% on semaglutide alone. With the antibodies added it ran between 2% and 3.8%. The company presented those results at a conference, and they have not been peer reviewed (Announcement).
All three share the same gap. Every one of them is a phase 2 study, the middle stage that tests whether a drug does anything at all. The trials are small, short and built around body scans. The few function results so far come from side measures, which may not reflect real-world strength. None of these drugs has yet shown a benefit to health itself (Review).
In September 2025 Eli Lilly, which owns bimagrumab, withdrew its own planned study of the drug with tirzepatide. That study was meant for people who also had type 2 diabetes. Nobody was ever enrolled, and the registry lists business strategy as the reason (Registry). A parallel study of the same pair in obesity without diabetes is still running (Registry).
What to Do Right Now
None of this is available to you yet. No drug in this class can be prescribed to protect muscle, anywhere. Getting from phase 2 to a pharmacy shelf takes years, and plenty of drugs never get there.
So treat BELIEVE as a reason to watch rather than a reason to wait. What the data does justify is a change in what you track. A scan number is a poor proxy for strength. Strength is the thing that was never shown to improve.
If you are on a GLP-1 and worried about muscle, two questions are worth raising with your clinician. First, how fast are you losing weight, since a gentler pace may be easier on lean tissue. Second, how is your function holding up, measured by something real like grip or standing up from a chair.
Meanwhile the non-drug option is tested and available today. One pooled analysis covered 21 trials where people lost at least a tenth of their body weight. On these drugs, about a third of that loss was fat-free tissue. On diet alone it was about 22%, and on diet plus exercise under 8% (Meta-analysis). Those figures come from separate pools of trials, not a head-to-head test. The protein targets and training details are in our guide to GLP-1 muscle loss, so we will not repeat them here.
Frequently Asked Questions
Can You Get a GLP-1 Muscle Drug on Prescription?
No. Bimagrumab is still experimental, it is owned by Eli Lilly, and its obesity evidence stops at one phase 2 trial (Trial). One related drug did reach approval in September 2026, but only for spinal muscular atrophy (Announcement). Nothing in this class is cleared anywhere for keeping muscle during weight loss.
Does Semaglutide Really Make You Lose Muscle?
Scans say some lean tissue goes, yes. Pooled trials put lean mass at about 30% of the weight lost (Meta-analysis). But lean tissue is not the same as muscle, and strength is rarely measured at all. In one year-long study of people on semaglutide, grip strength went up (Study).
Does More Lean Mass on a Scan Mean More Strength?
Not reliably. Bimagrumab raised lean mass in people with a muscle-wasting disease. Walking distance and strength still did not improve (Trial). Reviewers of this drug class say the link between a scan number and what a muscle can do is not dependable (Editorial).
What Are Bimagrumab’s Side Effects?
Muscle spasms are the big one, affecting roughly half to three quarters of people on the drug. Diarrhea and acne were each common as well, and more people quit bimagrumab than semaglutide (Trial). Pooled across four small trials, the same pattern holds (Meta-analysis).
Do Protein and Lifting Do the Same Job?
They do a different job, and it is the one you can start today. Adding exercise to a diet cuts the lean share of weight lost to under 8%, against about 22% on diet alone (Meta-analysis). Reviewers of GLP-1 trials recommend food and training alongside treatment, partly because the trials themselves never measured strength (Meta-analysis).
Key Takeaways
- The famous numbers are the friendliest ones. The 22% weight loss and 3% lean loss come from week 72, and from the highest-dose group only. The paper itself calls those analyses after-the-fact. The registered checkpoint was week 48, where lean loss was 2.3% against 6.9% (Trial).
- Strength was measured, and it barely moved. Grip in the headline group rose 2.3 kilos, against 1.7 kilos on placebo. Only one of the eight active groups clearly did better (Trial).
- This drug has failed a function test before. In inclusion body myositis it added lean mass at the higher doses and still missed its walking-distance goal. That is the cleanest sign that mass and function are separate things (Trial).
- The side-effect bill is not small. Muscle spasms reached 74% at the higher dose. Up to a fifth of people quit because of side effects. The drug is a clinic drip rather than a home pen (Trial).
- The whole class is stuck at phase 2. Enobosarm, apitegromab and trevogrumab all show better scans over 16 to 26 weeks. None has shown a health benefit yet, and none is approved (Review).
- The available option is still the boring one. When people lose weight by diet plus exercise, under 8% of the loss is fat-free tissue. On diet alone it is about 22%. You can start this week (Meta-analysis).
Watch the Function, Not the Scan
BELIEVE is a genuinely interesting trial. An antibody shifted where the lost weight came from, and the body-composition numbers are striking. That is worth knowing.
It is still a poor reason to wait. The trial measured the thing that matters to a person, and that result came back close to flat. A better scan has not yet been shown to make anyone stronger.
The measure that counts is already in your hands. What you can lift and carry is trackable this week, with no prescription and no clinic visit needed. Track that, and let the pipeline take the time it needs.
The science here is early, and it is worth following closely.
This article is for educational purposes and is not medical advice. Talk to a qualified clinician before changing your health regimen.
Leave a comment