Muscle Loss on a GLP-1: What the Evidence Actually Shows

Ordered by how far the published human evidence goes — strongest first. Availability is a separate question from evidence, so it is labeled separately.

1. 1.Tirzepatide (Zepbound, Mounjaro)

   FDA-approved
   Randomized trials

   Evidence strength: Randomized trials. Randomized controlled trials in humans, at scale.

   What it is

   A weekly injection that activates two gut-hormone receptors, GIP and GLP-1. Approved for weight management; the question here is what the weight is made of.

   What the evidence shows

   A body-composition substudy of SURMOUNT-1 scanned 160 participants by DXA at baseline and week 72. Body weight fell 21.3%, fat mass 33.9% and lean mass 10.9% on tirzepatide, against 5.3%, 8.2% and 2.6% on placebo. About 75% of the weight lost was fat and 25% lean — the same split in both arms, and consistent across sex, age and weight-loss tertile. A 2025 network meta-analysis of 22 randomized trials in 2,258 people ranked tirzepatide 15 mg among the most effective for weight and fat reduction and among the least effective at preserving lean mass.^1,3

   The catch

   The 25% share is ordinary; the absolute loss is not, because the total loss is not. DXA measures mass, not what you can lift — the substudy did not test strength, so nothing in it says whether those kilograms mattered functionally.

   Approved for this use and available by prescription.

2. 2.Semaglutide (Wegovy, Ozempic)

   FDA-approved
   Randomized trials

   Evidence strength: Randomized trials. Randomized controlled trials in humans, at scale.

   What it is

   A weekly GLP-1 receptor agonist, and the incretin drug with the largest published body-composition literature.

   What the evidence shows

   In the SEMALEAN study, 106 adults with obesity took semaglutide 2.4 mg for a year: weight fell 13%, fat mass 18%, and lean mass dropped about 3 kg by month 7 and then held. Handgrip strength rose 4.5 kg and the share meeting criteria for sarcopenic obesity fell from 49% to 33%. The 2025 network meta-analysis is less flattering — lean-mass loss at about a quarter of total weight loss across its 22 trials, with semaglutide 2.4 mg among the least effective at sparing it. A 2024 review put class-wide lean-mass loss near 10%, or roughly 6 kg, and compared that to a decade or more of ageing.^2,3,7

   The catch

   SEMALEAN was a prospective cohort at one hospital, not a randomized comparison, so it cannot separate the drug from the weight loss. Its rising grip strength and the meta-analysis's falling lean mass are not contradictory: mass and function are different things, and grip improving while mass falls is what tends to happen when someone stops carrying 15 extra kilograms.

   Approved for this use and available by prescription.

3. 3.Liraglutide (Saxenda)

   FDA-approved
   Randomized trials

   Evidence strength: Randomized trials. Randomized controlled trials in humans, at scale.

   What it is

   A daily GLP-1 injection, the oldest approved incretin drug for weight management and the weakest for total weight loss.

   What the evidence shows

   In the 2025 network meta-analysis, liraglutide at 3.0 mg daily or 1.8 mg daily was the GLP-1 receptor agonist in that 22-trial set that reduced weight significantly without a significant reduction in lean mass. A separate randomized trial took 195 adults who had lost a mean 13.1 kg on an 8-week low-calorie diet and assigned them for one year to exercise, liraglutide 3.0 mg, both, or placebo. The combination reduced body-fat percentage by 3.9 points — about twice the fall in the exercise-alone (1.7 points) and liraglutide-alone (1.9 points) groups — and it was the group where HbA1c, insulin sensitivity and cardiorespiratory fitness all improved.^3,4

   The catch

   Liraglutide also produced much less weight loss than the newer drugs, which is the likeliest reason lean mass held. Reading that as muscle-sparing rather than as weaker is the mistake to avoid. The more useful result is the combination arm, and the ingredient doing the work there may be the training.

   Approved for this use and available by prescription.

4. 4.Bimagrumab

   Research use only
   Early human

   Evidence strength: Early human. Small, open-label, or phase 1/2 trials. Promising, unproven.

   What it is

   An investigational monoclonal antibody that blocks activin type II receptors. It is not a peptide, and it has the most direct evidence on this page for the goal readers arrive with.

   What the evidence shows

   A 48-week phase 2 trial in 75 adults with type 2 diabetes and obesity reported fat mass down 20.5% (−7.5 kg) on bimagrumab against 0.5% (−0.18 kg) on placebo, while lean mass rose 3.6% (+1.70 kg) against −0.8% (−0.4 kg), alongside a 9.0 cm fall in waist circumference. A 2026 phase 2 trial randomized 507 adults with obesity across nine groups for 48 weeks: least-squares mean weight change was −9.3 kg on bimagrumab 30 mg/kg, −14.2 kg on semaglutide 2.4 mg and −17.8 kg on the two together, against −3.3 kg on placebo. Muscle spasms, diarrhoea and acne were the common bimagrumab adverse events.^5,6

   The catch

   Two phase 2 trials totalling 582 participants is not an approval, and both administered the drug as an intravenous infusion in a clinic — not something a person gives themselves. Nothing marketed as a muscle-preserving peptide is this molecule, and no result here carries across to one.

   Sold for laboratory use. Not a legal medicine for people.

5. 5.CJC-1295, ipamorelin and sermorelin

   Compounding restricted
   Animal only

   Evidence strength: Animal only. Rodent or cell studies. Nothing published in humans.

   What it is

   Growth-hormone-releasing peptides, usually sold as blends, on the reasoning that raising growth hormone protects muscle during a calorie deficit.

   What the evidence shows

   A 2026 orthopaedic review groups these with tesamorelin and AOD-9604 as secretagogues that activate IGF-1 signalling and satellite-cell repair, and states that clinical trials are currently lacking. A 2026 sports medicine review reaches the same conclusion about the unapproved peptides it examined: favourable animal results, scarce rigorous human data, and potential for serious harm.^8,9

   The catch

   As of August 2026, a ClinicalTrials.gov intervention search returned three studies for ipamorelin — post-operative ileus, gastrointestinal dysmotility, and one listing PTSD, cognitive dysfunction and brain injury — and one for CJC-1295, a phase 2 in HIV-associated visceral obesity terminated in 2006 with no results posted. None of those four lists lean mass, muscle, obesity or weight loss as what it measured. Peptide compounding is separately under active FDA restriction.

   Under active FDA restriction — compliant pharmacies decline to sell it.

6. 6.AOD-9604

   Research use only
   Anecdotal

   Evidence strength: Anecdotal. User reports and forum consensus. No controlled data.

   What it is

   A fragment of human growth hormone, sold on a body-composition promise — the name is short for 'anti-obesity drug 9604'.

   What the evidence shows

   It appears in the 2026 orthopaedic peptide review among the growth-hormone secretagogues and in the 2026 sports medicine review of unapproved peptides. In both, the supporting work described is preclinical.^8,9

   The catch

   As of August 2026, a ClinicalTrials.gov intervention search for AOD-9604 returned zero studies. That is a compound named after an outcome, with no registered trial measuring that outcome in anyone — let alone measuring lean mass in someone taking an incretin drug.

   Sold for laboratory use. Not a legal medicine for people.