- Across randomized trial substudies, lean mass accounted for roughly 25% to 39% of the total mass lost during 36 to 72 weeks of GLP-1 based weight reduction, according to a 2024 analysis in The Lancet Diabetes & Endocrinology.
- In the SURMOUNT-1 DXA substudy of tirzepatide (160 participants), body weight fell 21.3%, fat mass 33.9% and lean mass 10.9% at week 72. The split was about 74% fat and 26% lean mass, compared with 75% fat and 25% lean mass in the placebo group.
- In the STEP 1 DXA substudy of semaglutide (140 participants), total lean body mass fell by about 9.7% from baseline, yet lean mass increased as a proportion of total body mass because fat mass fell faster.
- DXA lean mass is not muscle. It also contains organs, skin, connective tissue and body water, which is why MRI and strength testing add information that a two-compartment scan cannot provide.
- In the SURPASS-3 MRI analysis, thigh muscle volume declined broadly in line with the amount of weight lost, while muscle fat infiltration decreased, a direction usually considered favourable for muscle quality.
- Evidence on strength and function is thinner and mixed: shorter trials often report preserved grip strength, while longer observational data in older adults point toward declines.
- Muscle-targeted agents such as bimagrumab and apitegromab change scan numbers in phase 2 trials, but none has demonstrated a benefit on strength or physical function, and none is approved for this purpose.
Do GLP-1 medications cause muscle loss?
Short answer: yes, part of the mass lost during GLP-1 based weight reduction is lean tissue, and every published body composition substudy agrees on that point. What the data do not show is that this loss is unique to the drug class, disproportionate for the amount of weight lost, or translated into measurable weakness in the trial populations studied so far.
The most cited synthesis comes from a 2024 commentary in The Lancet Diabetes & Endocrinology by Prado and colleagues, who estimated that lean mass represented roughly 25% to 39% of the total mass lost across GLP-1 receptor agonist trials lasting 36 to 72 weeks. The same authors noted that the absolute annual decline in lean tissue during this kind of rapid weight reduction is considerably larger than the slow loss attributed to normal ageing, which is why the question deserves attention rather than dismissal.
Two framings compete in the literature. The first treats lean mass loss as an unavoidable, largely proportional consequence of losing a large amount of body mass: when total weight falls by 15% to 22%, the body sheds supporting tissue along with fat. The second treats it as a potential pharmacological signal that could matter more in people with little muscle reserve. Current evidence leans toward the first explanation, while acknowledging that the second has not been excluded, particularly in older populations.
This article deals only with lean mass and skeletal muscle. For nausea, gastrointestinal effects and other tolerability questions, see our separate overview of GLP-1 side effects, and for the pharmacology of the class itself, see the GLP-1 guide.
Educational content only. Nothing in this article suggests starting, changing or stopping any medication. Decisions of that kind belong to you and a qualified healthcare professional.
What did the body composition substudies of semaglutide and tirzepatide measure?
Most of the body composition evidence comes from small imaging substudies nested inside much larger weight loss trials. They were not designed as the primary analysis, and their sample sizes are a fraction of the parent trials, which is the single most important caveat to carry through the rest of this page.
STEP 1 (semaglutide). The parent trial randomized 1 961 adults with overweight or obesity to semaglutide 2.4 mg once weekly or placebo for 68 weeks, alongside lifestyle intervention, and was published by Wilding and colleagues in the New England Journal of Medicine in 2021. A DXA substudy of 140 participants measured total fat mass, regional visceral fat and total lean body mass. Exploratory analyses of that subgroup reported a reduction in total fat mass and regional visceral fat mass, a fall in absolute lean body mass of about 9.7% from baseline, and an increase in lean mass expressed as a proportion of total body mass. More on the molecule itself in the semaglutide guide.
SURMOUNT-1 (tirzepatide). Look and colleagues published the DXA substudy in Diabetes, Obesity and Metabolism in 2025. It included 160 participants (124 across pooled tirzepatide groups, 36 on placebo), with a mean baseline weight of 102.5 kg and a mean BMI of 38.0 kg/m², followed to week 72. Details of the molecule are covered in the tirzepatide guide.
SURPASS-3 MRI (tirzepatide, type 2 diabetes). This post-hoc analysis, published in The Lancet Diabetes & Endocrinology in 2025, used magnetic resonance imaging rather than DXA. Of 296 enrolled participants, 246 had valid week-52 scans. MRI allowed the investigators to quantify thigh muscle volume and muscle fat infiltration, a marker of muscle quality that DXA cannot see, and to compare tirzepatide with once-daily insulin degludec.
The methodological gap between these approaches matters. DXA divides the body into fat, bone and a single "lean" compartment. MRI can separate muscle tissue from the fat stored inside and between muscle fibres. Conclusions drawn from one method do not transfer automatically to the other.
How much of the weight lost is lean mass, and is that different from other weight loss?
The SURMOUNT-1 substudy provides the clearest internal comparison, because both the active and placebo groups lost weight and both were scanned. At week 72, participants on tirzepatide lost 21.3% of body weight, 33.9% of fat mass and 10.9% of lean mass. The placebo group lost 5.3% of body weight, 8.2% of fat mass and 2.6% of lean mass. All comparisons versus placebo were reported as statistically significant.
Translated into proportions of the mass lost, tirzepatide produced approximately 74% fat and 26% lean mass, while placebo produced approximately 75% fat and 25% lean mass. The authors reported that these proportions were relatively consistent across clinically relevant subgroups, and concluded that the drug was not associated with a disproportionate loss of lean tissue. This was a post-hoc analysis of a small subgroup, so the finding is indicative rather than definitive.
| Substudy | Duration | Participants scanned | Body weight | Fat mass | Lean mass |
|---|---|---|---|---|---|
| SURMOUNT-1 DXA (tirzepatide) | 72 weeks | 160 (124 active, 36 placebo) | -21.3% | -33.9% | -10.9% |
| SURMOUNT-1 DXA (placebo) | 72 weeks | 36 | -5.3% | -8.2% | -2.6% |
| STEP 1 DXA (semaglutide) | 68 weeks | 140 total substudy | Reduced | Reduced | About -9.7% from baseline |
A 2026 systematic review and meta-analysis in the International Journal of Obesity, pooling seven randomized trials and 821 participants treated at obesity-indicated regimens, found the same two-sided picture: absolute lean mass fell by a pooled 1.74 kg (95% CI -3.04 to -0.45), with very high heterogeneity between trials, while lean mass as a share of total weight rose by about 1.81 percentage points (95% CI 1.1 to 2.52). The semaglutide subgroup showed the largest absolute lean mass reduction. The authors concluded that lean mass loss should not be regarded as a reason to avoid these medicines, while stressing that nutritional and exercise support belongs alongside them.
One analytical trap is worth naming. Several popular articles divide a lean mass figure taken from a small DXA subgroup by a weight loss figure taken from the full trial. Those two numbers come from different populations, and the resulting percentage is not meaningful. If you are comparing the two molecules more broadly, our semaglutide versus tirzepatide comparison sets out the differences in trial design that make head-to-head body composition claims difficult.
Is lean mass the same thing as muscle?
No, and this distinction explains much of the confusion in public discussion. Lean soft tissue as measured by DXA is everything that is neither fat nor bone mineral: skeletal muscle, but also internal organs, skin, connective tissue, blood and, importantly, water. Skeletal muscle typically represents only part of that compartment.
During rapid weight reduction, several components of lean mass change for reasons that have nothing to do with contractile tissue. Glycogen stores fall and carry bound water with them. Plasma volume shifts. The liver and other organs, which are enlarged in obesity, can shrink. Extracellular fluid declines. A DXA scan records all of this as lost "lean mass" without being able to apportion it.
This is precisely why the SURPASS-3 MRI analysis is informative. Over 52 weeks in people with type 2 diabetes, tirzepatide was associated with reduced thigh muscle volume, but the decrease followed the general relationship between body weight change and muscle volume derived from population-based estimates, including longitudinal UK Biobank data used by the authors for context. At the same time, muscle fat infiltration decreased, whereas the expected trajectory with ageing is an increase. Lower intramuscular fat is generally considered a marker of better muscle quality, so the imaging picture is not simply one of deterioration.
Prado and colleagues make the complementary point in their commentary: because DXA conflates tissues and because few trials measure function, a reported kilogram of "lean mass" lost is not interchangeable with a kilogram of muscle lost, and treating the two as equivalent can mislead in either direction. Equally, they argue that dismissing the issue because the compartment is heterogeneous risks ignoring a change that could matter for some patients.
Practically, this means that any strong claim about muscle based only on a DXA percentage, in either direction, is outrunning the measurement.
What happens to strength and physical function?
This is the weakest part of the evidence base, and it is also the part that matters most clinically. Mass is a surrogate. Strength, gait speed and the ability to rise from a chair are outcomes people actually experience.
The SEMALEAN study, a prospective real-world cohort published in Diabetes, Obesity and Metabolism in 2026 by Alissou and colleagues, followed patients with obesity treated with semaglutide 2.4 mg at a French university hospital, with assessments at baseline, 7 months and 12 months. Among the 115 patients enrolled, mean body weight fell by roughly 13% over 12 months, fat mass fell by about 14% at month 7 and 18% at month 12, and lean mass fell by about 3 kg by month 7 and then stabilised. Handgrip strength improved rather than declined between months 7 and 12, resting energy expenditure normalised to lean mass rose, and the reported prevalence of sarcopenic obesity fell from 49% at baseline to 33% at 12 months. The study had no control group, so these changes cannot be attributed to the drug alone.
A 2026 review in the British Journal of Pharmacology by Prokopidis assembled the broader strength literature and reached a nuanced conclusion: lean soft tissue loss is not a reliable predictor of strength change. Shorter trials in adults with obesity, generally six months or less, frequently report preserved handgrip strength despite measurable lean tissue loss. Longer observational and retrospective data in older adults with type 2 diabetes, over about 24 months, have reported reduced handgrip strength and a pattern consistent with accelerated sarcopenia compared with matched controls.
Those two bodies of evidence are not directly comparable. Randomized trials recruit relatively healthy, mostly middle-aged participants and run for a year or two at most. Observational cohorts in older adults carry confounding by indication, frailty and comorbidity. The honest summary is that function has been measured too rarely and for too short a time to settle the question, and that several authors now argue strength testing should be built into obesity trial design rather than added as an afterthought.
Who is most likely to be affected by lean mass loss?
The arithmetic of muscle reserve explains most of the concern. A 35-year-old with abundant skeletal muscle who loses 10% of lean mass remains well above any functional threshold. An 80-year-old close to the diagnostic cut-offs for sarcopenia who loses the same proportion may cross into territory associated with falls, slower recovery from illness and loss of independence.
Older adults are therefore the population most often flagged in the literature. Reviews published in 2025 and 2026 on GLP-1 receptor agonist use in people over 65 consistently note that concomitant lean mass loss is common and that it occurs against a background of age-related muscle decline, while also noting that the direct evidence is thin: dedicated randomized trials with adequate duration in older adults, with strength and function as endpoints, largely do not yet exist.
Other groups discussed as potentially more exposed include people with pre-existing sarcopenia or sarcopenic obesity, people with low habitual protein intake, those who are largely sedentary, and people undergoing very rapid or very large weight reduction. None of these has been tested as a prespecified subgroup in a trial powered for functional outcomes, so these remain plausible risk markers rather than established ones.
A counterweight deserves mention for balance. Obesity itself impairs physical function, loads joints and is associated with intramuscular fat accumulation. Several reviews point out that losing a substantial amount of fat mass can improve mobility and muscle quality even when absolute lean mass declines, and the SURPASS-3 MRI reduction in muscle fat infiltration is consistent with that idea. The net effect on function in any individual is not predictable from a scan.
If you fall into one of these groups, the appropriate step is a conversation with your own clinician about how your muscle health is assessed, not a decision made from an article. This page cannot and does not provide that assessment.
What do trials show about exercise and protein during GLP-1 weight loss?
Mechanistically, resistance exercise and adequate protein intake are the two best established countermeasures against lean mass loss during any form of caloric restriction. The specific question of whether they work during GLP-1 treatment has surprisingly little completed randomized evidence behind it.
The most relevant published trial uses an older molecule in the class. Lundgren and colleagues, in the New England Journal of Medicine in 2021, randomized adults with obesity after a low-calorie diet to exercise, liraglutide, both combined, or placebo for one year. The combination group achieved roughly twice the fat mass loss and twice the waist reduction of either single strategy, while maintaining lean mass, and produced the best overall weight loss maintenance. Liraglutide alone showed a treatment effect of -6.8 kg versus placebo. This is supportive evidence for combining exercise with incretin-based therapy, but liraglutide is not semaglutide or tirzepatide, and the magnitude of weight loss was smaller than in the newer trials.
For the newer agents, the key trial is still recruiting. LEAN-PREP (registered as NCT06885736), whose protocol was published in 2026, plans to randomize 232 adults with obesity treated with semaglutide or tirzepatide to four arms: control, resistance exercise, protein supplementation, or both combined, over six months. Its protein target is 1.6 g per kg per day with at least 25 g per meal. Primary completion is scheduled for 2029, so no outcome data exist yet.
Meanwhile, the pooled meta-analysis evidence and most review authors converge on the same recommendation at the level of research priorities rather than individual advice: pharmacological weight reduction should be studied, and in practice accompanied, by structured nutritional and exercise support. What that support should look like for a given person is a clinical question. Peptides marketed online for muscle building are a separate topic with a very different evidence base, discussed in our review of peptides studied for muscle growth and recovery; none of them has been tested as a countermeasure for lean mass loss during GLP-1 therapy.
Which molecules in development aim to protect muscle, and what have they shown?
Several investigational agents aim to preserve or add lean mass during incretin-based weight loss, mostly by blocking activin or myostatin signalling pathways. All of the results below come from phase 2 trials. None of these agents is approved for muscle preservation during weight loss, and none has demonstrated a benefit on strength or physical function.
Bimagrumab (an anti-activin type II receptor antibody) plus semaglutide was tested in the phase 2 BELIEVE trial, published in Nature Medicine in 2026 with Heymsfield as lead author, in 507 participants. At week 72, the reported changes were a 22.1% reduction in body weight with the high-dose combination versus 15.7% with semaglutide alone and 10.8% with bimagrumab alone; fat mass fell 45.7% with the combination versus 27.8% and 28.5% respectively; and lean mass fell 2.9% with the combination versus 7.4% with semaglutide alone, while rising 2.5% with bimagrumab alone. Reported adverse events in the bimagrumab groups included mild to moderate acne and muscle spasms, and the authors called for further development work to understand them.
Apitegromab (an anti-promyostatin antibody) was tested with tirzepatide in the phase 2 EMBRAZE trial, published by Pratley and colleagues in Nature Medicine in 2026. In 102 adults with overweight or obesity over 24 weeks, lean mass changed by -1.6 kg with apitegromab versus -3.5 kg with placebo, a between-group difference of about 1.9 kg (p = 0.0014), with similar total weight loss in both arms. Critically, strength and chair-stand testing did not differ between groups, which the investigators and independent commentators both highlighted.
That last detail is the central caution for this entire category. A molecule can move a DXA number without moving anything a patient would notice. Preserved mass without demonstrated preserved function is a biomarker result, and the history of body composition pharmacology contains several agents that improved scans without improving outcomes. Longer trials with prespecified functional endpoints are needed before any claim of clinical benefit is justified.
Other programmes in this space, including myostatin pathway antibodies and selective androgen receptor modulators, are at earlier or less transparent stages. Compounds of this type are not approved for human use outside clinical trials.
What remains unknown about GLP-1 therapy and muscle?
Four gaps define the current state of knowledge, and naming them precisely is more useful than a confident summary.
Function is under-measured. Almost all of the data concern mass, measured by DXA, in substudies of 140 to 246 participants. Grip strength appears in a handful of studies, gait speed and chair-stand testing in fewer still, and almost never with the duration needed to detect a slow functional drift. Recent commentary in Diabetes, Obesity and Metabolism has argued explicitly for monitoring strength and function rather than lean mass alone in both trials and practice.
Older and frailer populations are under-represented. The randomized evidence comes largely from middle-aged participants with high baseline muscle reserve. The populations where the concern is most plausible are those least studied, and the available signals there come from observational data with their own well-known confounding.
Reversibility and the maintenance phase are unclear. Whether lean mass recovers after weight stabilises, what happens during treatment interruption or after discontinuation, and whether regained weight returns as fat preferentially are open questions with limited data. The SEMALEAN observation that lean mass fell early then stabilised is consistent with an adaptation pattern, but came from an uncontrolled cohort.
Drug class effects have not been separated from weight loss effects. The Prado commentary frames this as the pivotal question: is lean mass loss simply proportional to the magnitude of weight reduction, which would make it a feature of all effective weight loss, or does incretin signalling contribute independently? The placebo comparison in SURMOUNT-1 and the UK Biobank contextualisation in SURPASS-3 MRI both favour the proportional explanation, but neither constitutes proof.
Medical disclaimer. This article is for educational purposes only and is not medical advice. It does not recommend starting, modifying or discontinuing any medication or supplement. Semaglutide and tirzepatide are prescription medicines whose availability and legal status vary by jurisdiction; bimagrumab and apitegromab are investigational for this indication. Discuss any question about your own treatment, muscle health or monitoring with a qualified healthcare professional. See our medical disclaimer for full terms.
Frequently Asked Questions
How much of the weight lost with GLP-1 medications is lean mass?
Is lean mass loss on semaglutide different from lean mass loss with dieting?
Does a drop in DXA lean mass mean muscle is being lost?
Do GLP-1 medications reduce muscle strength?
Are older adults more affected by lean mass loss during GLP-1 treatment?
Does tirzepatide preserve muscle better than semaglutide?
Can exercise and higher protein intake prevent lean mass loss during GLP-1 therapy?
Are there medicines that preserve muscle during GLP-1 weight loss?
Recommended products
Research peptides selected for quality and purity:
GHK-Cu
Anti-Aging Compound
GHK-Cu Pen
Pre-filled Pen GHK-Cu
GLOW-UP Ritual
3-pen pack: Glow, NAD+, GHK-Cu
Test your knowledge
Quick quiz · 6 questions
Sources
- Look M, et al. (2025). Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight. Diabetes, Obesity and Metabolism. (PubMed)
- Wilding JPH, et al. (2021). Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1). New England Journal of Medicine. (PubMed)
- Sattar N, et al. (2025). Tirzepatide and muscle composition changes in people with type 2 diabetes (SURPASS-3 MRI): a post-hoc analysis of a randomised, open-label, parallel-group, phase 3 trial. The Lancet Diabetes & Endocrinology. (thelancet.com)
- Prado CM, Phillips SM, Gonzalez MC, Heymsfield SB (2024). Muscle matters: the effects of medically induced weight loss on skeletal muscle. The Lancet Diabetes & Endocrinology. (PubMed)
- Alissou M, et al. (2026). Impact of semaglutide on fat mass, lean mass and muscle function in patients with obesity: the SEMALEAN study. Diabetes, Obesity and Metabolism. (PubMed)
- Prokopidis K (2026). Glucagon-like peptide-1 receptor agonists and muscle strength changes in older adults: risks beyond muscle mass reductions. British Journal of Pharmacology. (PubMed)
- Lundgren JR, Janus C, Jensen SBK, et al. (2021). Healthy Weight Loss Maintenance with Exercise, Liraglutide, or Both Combined. New England Journal of Medicine. (nejm.org)
- Pratley RE, Denham DS, Trivedi R, et al. (2026). Apitegromab for lean mass preservation during tirzepatide-induced weight loss: a randomized, double-blind, placebo-controlled phase 2 trial (EMBRAZE). Nature Medicine. (nature.com)
- Heymsfield SB, et al. (2026). Bimagrumab plus semaglutide alone or in combination for the treatment of obesity: a randomized phase 2 trial (BELIEVE). Nature Medicine. (nature.com)
- Systematic review and meta-analysis of 7 randomized controlled trials (821 participants) (2026). Effect of GLP-1 receptor agonists at doses for obesity management on muscle health: systematic review and meta-analysis of randomized controlled trials. International Journal of Obesity. (PubMed)
- LEAN-PREP study investigators (2026). LEAN mass Preservation with Resistance Exercise and Protein during semaglutide and tirzepatide therapy (LEAN-PREP study): a protocol for a randomised controlled trial. Published trial protocol (NCT06885736). (PubMed)