Yes, some of the weight lost on GLP-1 drugs is lean mass, but not more than other large weight loss produces. In a 72-week scan substudy of the tirzepatide trial SURMOUNT-1 (n = 160), fat mass fell 33.9% and lean mass 10.9%. About 75% of the weight lost was fat and 25% lean, the same split as on placebo.[3]
When you lose a lot of weight — by any means — some of what you lose is fat and some is lean mass: muscle, bone, organ tissue, and the water they hold. That is not unique to GLP-1 drugs; it is a basic feature of weight loss. But because these medications drive such large reductions so reliably, the absolute amount of lean mass that comes along has become a real clinical question. Here is what the body-composition data actually report — and what they don't.
The fraction, not just the number
The pivotal trials are usually quoted only by total weight: about 14.9% with semaglutide 2.4 mg in STEP 1,[1] and up to roughly 20.9% with tirzepatide in SURMOUNT-1.[2] Those headline figures say nothing about composition. The question that matters physiologically is what share of the lost mass was lean tissue rather than fat — and that requires a body-composition substudy using DXA or MRI, not a scale.
What the tirzepatide substudy showed
A body-composition substudy within SURMOUNT-1 measured the split directly with DXA scans in 160 of the trial’s 2,539 participants. Over 72 weeks, tirzepatide lowered body weight by 21.3%, fat mass by 33.9% and lean mass by 10.9%, against 5.3%, 8.2% and 2.6% on placebo (p < 0.001 for each). About 75% of the weight lost was fat and 25% was lean, in both groups, and that split held across most subgroups by sex, age and amount of weight lost.[3] Because fat fell three times faster than lean mass, participants ended up proportionally leaner, even though some absolute lean mass was lost along with the fat. This is the key nuance: losing some lean mass is expected during large weight loss, but if fat falls faster than lean, overall body composition still moves in a favorable direction. The substudy is the cleanest direct read on the question for these agents.
How big is the lean-mass share, generally?
Across the weight-loss literature, lean tissue accounts for a meaningful minority of total mass lost. The BELIEVE trial authors put it at about 25% to 40% for diet, incretin drugs and bariatric surgery alike,[6] and a nutrition review of the newer anti-obesity drugs gives a wider 20% to 50%, which can mean about 6 kg, or 10% of total lean mass, over 12 to 18 months.[5] The range varies with the scan method, the population, and whether resistance training and adequate protein were in place. A detailed review of how weight loss affects fat-free mass, muscle, and bone, written with the new pharmacotherapies in mind, puts surgery and drugs in the same bracket: over 25% of the weight lost from either typically comes from fat-free mass. It also warns that this matters most for older and pre-frail adults, who face a higher risk of sarcopenia.[4]The open scientific question is whether the lost lean mass is functionally important muscle or largely the expected tissue and water that accompany a shrinking body.
What actually moves the lean-mass needle
The intervention with the strongest support for preserving muscle during weight loss is not a drug at all: it is resistance training plus adequate dietary protein — the same pairing that carries the highest-quality evidence against age-related muscle loss, and the reason a stop-start cycle on these drugs is not composition-neutral. Reviews of muscle health during weight loss on these medications converge on the same practical message — pair the drug with strength work and sufficient protein intake to bias the loss toward fat and protect functional muscle.[5] Much of the alarming framing around “Ozempic muscle loss” collapses two different things: the normal, expected lean-mass component of any large weight loss, versus a drug-specific catabolic effect, for which the trial evidence is far weaker than the headlines suggest.
A pharmacological answer is being pursued, and the two most advanced candidates now have real combination-trial results to report, which change the picture in progress. Bimagrumab, an antibody that blocks the activin type II receptor, is the most direct attempt: in the phase 2 BELIEVE trial (507 adults), published in Nature Medicine in March 2026, the high-dose combination cut body weight by 22.1% at week 72 versus 15.7% on semaglutide 2.4 mg alone, and 92.2% of the weight it removed was fat, versus 75.6% on semaglutide alone. Lean mass fell 2.9% on the combination and 7.4% on semaglutide. SF-36 physical-function scores did not differ from placebo at week 48; by week 72, in the open-label extension, they had improved more on the high-dose combination than in the other groups.[6] That is an early functional signal, not a proven one, and it sits in contrast with bimagrumab’s track record on its own in other muscle-losing populations (inclusion body myositis, sarcopenia, COPD, hip fracture), where across four randomized trials it has not improved physical function. Enobosarm, the SARM better known as ostarine, shows a similar shift: two completed phase 3 trials in cancer muscle wasting raised lean body mass without a functional benefit, but its newer phase 2b trial paired with semaglutide (QUALITY, 168 adults aged 60 or older, 16 weeks) met its primary lean-mass endpoint. On the sponsor’s figures, 17.6% of people on enobosarm 3 mg plus semaglutide lost 10% or more of their stair-climb power versus 44.8% on semaglutide alone. The sponsor’s next step is another phase 2b trial (PLATEAU), planned to start in early 2026, rather than phase 3.[7] Neither result is confirmatory (BELIEVE is one phase 2 trial; QUALITY is phase 2b data reported by the drug’s sponsor, not yet a full peer-reviewed publication), and neither is available outside a trial today. But the lesson is shifting from “mass on a scan and function in a body are separable” toward “the first drugs designed to move both together are starting to show it” — which is exactly why training and adequate protein remain the accessible, proven lever right now, not a placeholder.
“Ozempic muscle loss” specifically
Because Ozempic is the best-known brand, the concern usually gets searched as “Ozempic muscle loss.” Ozempic and Wegovy are both semaglutide, so the semaglutide body-composition evidence is exactly what applies: in the pivotal weight-loss program, participants lost roughly 15% of body weight,[1] and — as with the whole GLP-1 class — a portion of that is lean mass, tracking with the amount of weight lost rather than a drug-specific muscle-wasting effect. There is no evidence that semaglutide melts muscle in a way that differs from other large weight-loss interventions; the lean-mass share and the fat-falls-faster pattern are the same story told above. The one caveat worth naming is that older adults and people losing weight very quickly have more to lose functionally, which is precisely why the protein-and-resistance-training advice matters most for them.[5] In short: “Ozempic muscle loss” is real in the trivial sense that any large weight loss includes some lean mass, and overstated in the sense that implies the drug is uniquely catabolic.
The honest bottom line
Yes, some of the weight lost on GLP-1 medications is lean mass — that is real and measured. But the body-composition data show fat falling faster than lean, an improving fat-to-lean ratio, and a lean-mass share broadly in line with other large weight-loss interventions. What the data do not support is the claim that these drugs uniquely waste muscle. The actionable finding is the boring one: resistance training and adequate protein are what protect lean mass, on or off these medications.