Medically reviewed by: [leave blank for reviewer name and credentials]
Someone in a comment section has told you that you’ll lose all your muscle. Someone else has told you that’s a myth. GLP-1 muscle loss is neither of those. The body composition data exists. The numbers are specific. They are also less alarming and more useful than either camp suggests.
Key takeaways
- Some of the weight lost on incretin medicines is lean tissue rather than fat. This is measurable and expected.
- In the STEP 1 body composition substudy, total lean body mass fell 9.7% over 68 weeks while fat mass fell far more.
- That ratio is broadly what happens with any rapid weight loss. The absolute numbers look larger because the total weight loss is larger.
- DXA “lean mass” is not the same thing as skeletal muscle. It includes water, glycogen, organ tissue, and connective tissue.
- Resistance training plus adequate protein is the intervention with the best supporting evidence.
What “lean mass” actually means on a scan
This is the single biggest source of confusion, so it goes first.
Body composition in these trials is measured by DXA, a scan that sorts your body into three buckets: fat mass, bone mineral, and everything else. That third bucket is called lean mass or lean soft tissue.
Everything else includes skeletal muscle. It also includes water, stored glycogen, organ tissue, connective tissue. The blood volume that supports a larger body.
So when a study reports a 9.7% drop in lean mass, that is not a 9.7% drop in muscle. Some of it is water and glycogen, which fall as expected when food intake drops. Some of it is organ and connective tissue adapting to a smaller body. Muscle is part of it, but the scan cannot tell you how much.
This matters because the headline number gets quoted as though it were a muscle number, and it isn’t.
What the body composition trials found
Two large trials included DXA substudies, and they are the main evidence base.
| Trial | Medication | Duration | Total weight change | Fat mass change | Lean mass change |
|---|---|---|---|---|---|
| STEP 1 substudy | Semaglutide 2.4 mg weekly | 68 weeks | About −15% | −19.3% | −9.7% |
| SURMOUNT-1 substudy | Tirzepatide 5, 10, or 15 mg weekly | 72 weeks | About −22% | −33.9% | −10.9% |
Two things stand out.
Fat mass fell much faster than lean mass in both. In STEP 1, people in the DXA subset lost about 10.4 kg of fat against about 6.9 kg of lean soft tissue. The proportion of the body that was lean tissue actually increased, because fat was leaving faster.
The lean fraction of total loss differed between the two. Published analyses put it at roughly 40% for semaglutide in STEP 1 and roughly 25% for tirzepatide in SURMOUNT-1. Estimates vary between analyses depending on which subset and which estimand is used. That is why you will see a range quoted rather than a single figure.
Systematic reviews of randomized trials have mostly concluded that these medicines do not appear to harm skeletal muscle beyond what the size of the weight loss would predict. Put another way, the ratio is normal. The total is large.
Is this different from any other weight loss?
Not really, and that context is the part most articles skip.
Losing weight without resistance training and adequate protein has always cost lean tissue. Dieting does it. Bariatric surgery does it. A meal replacement program does it. Roughly a quarter to a third of weight lost through calorie restriction alone tends to come from lean tissue.
What is different with incretin medicines is the scale of the total loss and the speed. Losing 20% of your body weight in a bit over a year is more weight, faster, than most people have ever lost by other means. A normal ratio applied to a large total produces a large absolute number.
There is also a plausible mechanism specific to appetite suppression. If you eat much less, you may eat much less protein without noticing. That is because protein intake tends to fall in proportion to everything else. That part is fixable.
Does GLP-1 muscle loss matter clinically?
Honestly, the evidence here is less settled than the body composition numbers. It deserves a straight answer rather than a confident one.
Loss of lean tissue does not automatically mean loss of strength or physical function. In the trial substudies, no consistent clinically meaningful decline in physical function was observed. In the prospective SEMALEAN study, DXA-measured lean mass declined early and then stabilized, while average handgrip strength improved over 12 months. That study was not a randomized comparison. So it cannot establish that the medication caused the strength improvement.
The concern that clinicians take seriously is longer term and population specific. Older adults already at risk of sarcopenia, and anyone who was already low on muscle before starting, have less margin. For those groups, monitoring matters more.
What we do not yet have is long-duration data on functional outcomes after treatment ends. That gap is real and should be stated rather than glossed over.
What actually protects lean mass
Three things have evidence behind them, and one trial in particular is worth knowing about.
Resistance training. In a randomized trial published in the New England Journal of Medicine, people who combined a structured exercise program with a GLP-1 medication after an initial diet-induced weight loss lost much less fat-free mass than those on the medication alone, reported at 24.7% versus 38.6% of weight lost. The combination group also roughly doubled the reduction in body fat percentage compared with either strategy alone, and was the only group to show improvements in glycated haemoglobin, insulin sensitivity, and cardiorespiratory fitness.
That is the clearest proof available that the ratio is not fixed. Training changes what comes off.
Adequate protein. Protein intake tends to fall alongside total intake when appetite drops. Keeping it up needs deliberate planning rather than waiting to feel hungry. Specific targets should come from your healthcare provider or a registered dietitian, since they depend on body size, kidney function, and other medical factors.
Gradual titration and not rushing the dose. Our article on whether taking a lower dose actually works covers the related question. Escalating faster than tolerated tends to worsen nausea. This tends to worsen intake. This tends to worsen protein consumption. The titration schedule exists for reasons beyond side effect management.
Practical notes. Two or three resistance sessions a week covering the major movement patterns is the general shape of what was studied, rather than an hour a day. Strength tracked over time is a more useful signal than a DXA scan. That is because a scan cannot separate muscle from water and strength can.
Safety and who should speak to a healthcare provider first
Speak with a licensed healthcare provider before starting or changing a weight management treatment, and above all if you:
- Are over 65, or have been told you have low muscle mass or sarcopenia
- Have reduced kidney function, which affects protein recommendations
- Have a history of an eating disorder
- Have a personal or family history of medullary thyroid carcinoma, or multiple endocrine neoplasia syndrome type 2
- Have had pancreatitis or gallbladder disease
- Have type 1 diabetes, or take insulin or a sulfonylurea
- Are pregnant, planning to become pregnant, or breastfeeding
- Are losing weight much faster than expected, or notice weakness, difficulty climbing stairs, or difficulty rising from a chair
That last one is worth flagging early rather than waiting for a scheduled appointment.
Frequently asked questions
Does semaglutide cause muscle loss?
Some of the weight lost is lean tissue, and DXA substudies measured a 9.7% reduction in total lean body mass over 68 weeks in STEP 1. Lean mass on a scan includes water, glycogen, and connective tissue as well as muscle. So that figure overstates skeletal muscle loss specifically.
How much of GLP-1 weight loss is muscle?
Published analyses put lean tissue at roughly 25% to 40% of total weight lost, varying by medication and by which analysis you read. The muscle-only fraction is smaller than that figure, because DXA lean mass also counts water, glycogen, and connective tissue. No published trial has isolated skeletal muscle on its own.
Is that worse than losing weight by dieting?
By share, no. Losing weight through calorie restriction alone usually costs a similar share of lean tissue, and so does bariatric surgery. The absolute amount is larger with these medicines only because the total weight loss is larger. Apply an ordinary ratio to a very large total and you get a large number.
Will I get my muscle back after stopping?
There is no good long-term data on what happens to lean tissue after treatment ends. That is an honest gap in the evidence rather than a reassuring answer. It is worth knowing before you start. Resistance training and adequate protein during treatment are the more reliable lever. That is because they change the ratio while it is happening.
Do I need a DXA scan?
Most people do not need one. A DXA scan cannot separate skeletal muscle from water, glycogen, and connective tissue. So it answers a narrower question than people expect. Tracking strength on a few basic movements over several months gives you more usable information about muscle, and costs nothing beyond keeping a record.
Should I take creatine or a protein supplement?
Discuss any supplement with your healthcare provider first, since suitability depends on kidney function, other medicines, and your overall clinical picture. As a general principle, total daily protein intake matters more than which source it comes from. A supplement is one way to reach that total, not a requirement for reaching it.
The bottom line
Lean tissue loss on incretin medicines is real, measurable, and roughly proportional to what any large weight loss costs. The scan number is not a muscle number, the ratio is not fixed, and resistance training with adequate protein is the part you control.
To work out what fits your situation, see MyRocky’s weight management treatment options or start an online assessment with a licensed Canadian prescriber.
References
- Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity (STEP 1). New England Journal of Medicine, 2021. https://www.nejm.org/doi/full/10.1056/NEJMoa2032183
- Lundgren JR, Janus C, Jensen SBK, et al. Healthy weight loss maintenance with exercise, liraglutide, or both combined. New England Journal of Medicine, 2021. https://www.nejm.org/doi/full/10.1056/NEJMoa2028198
- Muscle loss and GLP-1 receptor agonist use. National Library of Medicine, PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC12957034/
Disclaimer: This article is intended for informational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always consult your healthcare provider with any questions about a medical condition or treatment.
Editorial Standards: At Rocky Health, we’ve made it our mission to support men and women with trustworthy, easy-to-understand medical and health information online. Read more about our editorial standards here.
