GLP-1 and Muscle Loss: What the Body-Composition Data Shows

Weight lost on a GLP-1 drug is not all fat. Some of it is lean mass — a bucket that includes muscle, but also organs, bone and water. How much, and whether it matters, is a real question with real data behind it, not just a talking point.

The clearest single number: SURMOUNT-1's body-composition substudy

The most directly measured answer comes from a body-composition substudy of SURMOUNT-1, tirzepatide's pivotal obesity trial, using DXA scanning at week 72. Tirzepatide-treated participants lost 21.3% of total body weight, made up of a 33.9% reduction in fat mass and a 10.9% reduction in lean mass — a roughly 75% fat / 25% lean split³. The detail worth sitting with: the placebo arm's split was nearly identical, at roughly the same 75/25 ratio on much smaller total losses³. That is the strongest evidence that this proportion is a property of losing weight itself, at almost any rate, rather than something the drug specifically does to muscle.

Why published lean-mass numbers vary so widely

A 2024 review in Diabetes, Obesity and Metabolism surveyed the GLP-1 and dual-agonist literature and found lean-mass loss reported anywhere from about 15% to 40–60% of total weight lost, depending on the study population, the specific drug, and comorbidities present⁴. That is a wide enough range that no single percentage is "the" answer across all patients. The same review makes a second, easily missed point: DXA-measured "lean mass" is not the same thing as muscle. It includes organs, bone mineral, connective tissue and body water, so a drop in lean mass on a scan is not proof of a proportional drop in muscle you could lift with⁴. MRI-based studies the review cites suggest the muscle-specific change tends to be "adaptive" — roughly what would be expected given age, weight loss magnitude and starting health status, not a distinct drug-driven muscle-wasting process⁴.

The pivotal trials, for scale

Semaglutide's STEP 1 trial produced a mean 14.9% total body weight reduction over 68 weeks (versus 2.4% on placebo)¹; tirzepatide's SURMOUNT-1 produced up to 20.9% at the highest dose over 72 weeks (versus 3.1% on placebo)². Larger total weight loss generally means a larger absolute amount of lean mass lost too, even at a constant proportion — which is the honest reason a bigger number on the scale is not an unambiguous win in every respect.

What is not yet settled

There is no FDA-approved GLP-1 label claim about muscle preservation, and the 2024 review is explicit that purpose-built pharmacological add-ons for muscle retention alongside GLP-1 therapy are still in development rather than available today⁴. Resistance training and adequate protein intake are the interventions with the most independent evidence behind preserving lean mass during any calorie deficit, GLP-1-assisted or not — that is general body-composition science, not a specific finding from these trials, and it is worth raising with your prescriber rather than treating as settled by this article.

Where this fits into choosing a molecule or a provider

Neither pivotal trial nor the review above found a meaningful difference in the fat-to-lean loss ratio between semaglutide and tirzepatide specifically — the SURMOUNT-1 ratio and the review's wide range both cut across drugs. If muscle preservation is a priority, that is a conversation for your prescriber about monitoring and resistance training, not a reason to pick one molecule over the other on this evidence alone. For the rest of the molecule trade-off, see semaglutide vs tirzepatide; for the other widely reported but off-label change, see GLP-1 and hair loss; for the full side-effect and warning picture, GLP-1 side effects; and for the rest of the Desk's sourced evidence, the research library.