What percentage of weight loss on GLP-1 is muscle?

In the trials that scanned people's bodies, about 25–40% of the weight lost on GLP-1 medications was lean mass, and a 2025 pooled analysis of placebo-controlled trials put the average at about 31%. But lean mass is not all muscle: it also counts water, organs and other tissue. Skeletal muscle is roughly half of lean mass, so the muscle share is likely closer to 10–20% of the weight lost, with fat making up most of the rest.

Below: the numbers from each major trial in one table, what "lean mass" really measures, how the share compares with dieting alone, what it means in pounds for someone losing 40 lb, who loses more, what happens if the weight comes back, and how to keep your own share at the low end.

The short answer, in numbers

So the honest answer is a range. A common figure in headlines, "up to 40% of the weight is muscle", takes the top of the lean-mass range and calls all of it muscle. Both steps overstate it.

Lean mass is not the same as muscle

Almost every number you'll read comes from DXA scans, which split the body into three parts: fat, bone, and "lean soft tissue". Lean soft tissue is everything that isn't fat or bone:

• skeletal muscle (the muscle you move with), roughly half of it;
• water, inside and outside cells;
• organs, such as the liver, which shrinks when it loses stored fat and sugar;
• glycogen, the sugar stored in muscle and liver, which holds about 3 g of water per gram. When you eat much less, glycogen and its water drop early, and DXA counts that as lean loss.

That is why a "30% lean" result does not mean 30% of the weight was muscle. Some of that lean loss is water and stored sugar that come back as soon as you eat more, and some is a smaller liver and other organs, which is part of normal weight loss.

MRI scans, which can measure muscle itself, add a second point. In a post-hoc MRI analysis of a tirzepatide trial in type 2 diabetes (SURPASS-3), thigh muscle volume fell, as expected with weight loss, but the fat inside the muscle fell too. Muscle with less fat in it tends to work better. Researchers are still arguing about how much this offsets the loss; it is a reason the raw "lean mass" number doesn't tell the whole story.

From percentages to pounds: a worked example

Say you lose 40 lb on a GLP-1 medication. Using the trial ranges:

The "about half" is a rough rule from body composition research, not a measured number for any one person, so read the muscle column as a ballpark. Even so, it shows the scale: for most people the medication takes off far more fat than muscle. What matters is whether you can keep the muscle number closer to the top row than the bottom one, which the second half of this guide covers.

Is that more muscle than losing weight by diet?

Not clearly. Any large weight loss takes some lean mass with it. A long-standing rule from diet studies is that about a quarter of the weight lost by eating less is fat-free mass. A 2014 critical review by Heymsfield and colleagues found the "quarter rule" holds roughly, with a lot of variation between people: it is higher in leaner people and with very fast loss, lower in people who start heavier.

Against that baseline:

• In SURMOUNT-1, the lean share on tirzepatide was the same as on placebo, around 25%.
• In the ENDO 2025 study, the share on semaglutide was similar to a diet and lifestyle program, though the medication group lost more weight.
• Some analyses comparing GLP-1 trials with diet-and-exercise trials report a larger lean share on the medications. Those comparisons mix very different people and methods.

A 2024 viewpoint in JAMA by Conte, Hall and Klein made the same point: the lean loss in GLP-1 trials is in line with what other ways of losing weight produce, and part of it is an adaptation to carrying less body weight. Where everyone agrees: the bigger and faster the loss, the more lean mass goes in absolute pounds. These medications produce big losses, so the pounds add up, even if the percentage is ordinary.

Why the trial numbers differ so much

• How much and how fast people lost. More loss means more lean mass lost in total, and in some studies a higher share.
• Who was in the trial. Age, sex, starting weight and diabetes all change the share. Women and older adults tend to lose a larger share of lean mass.
• What the scan counts. DXA can't tell muscle from water. A scan taken on a dehydrated morning shows more "lean loss" than one taken a week later.
• Time point. Lean loss is often front-loaded in the first months, when glycogen and water drop; the share can change by the end of a trial.
• Small numbers. The body-scan groups are a few hundred people at most. One small study can swing the headline.
• What people ate and did. Most trials did not set a protein target or a strength-training program. The share you see is for people doing whatever they did, which for many was little protein and no lifting.

Who loses a bigger share of muscle?

The ENDO 2025 study found three things linked to more muscle loss on semaglutide, after allowing for how much weight people lost: older age, being female, and eating less protein. Reviews add a few more:

• Adults over 60. Muscle already falls about 3–8% per decade after 30, faster after 60. Losing more with weight loss matters more when there is less to lose. Our guide to symptoms of muscle loss on GLP-1 covers the signs to watch.
• People who start leaner. With less fat to lose, a bigger share of each pound comes from lean tissue.
• Very low intake. People who eat very little for weeks, which these medications make easy, lose more lean mass.
• No strength training. Muscle that is not used is not kept, especially in a body that now weighs less.
• Long illness or bed rest during treatment. Days in bed speed up muscle loss at any weight.

Does it matter if some of the weight is muscle?

It depends on how much and who you are. For a 45-year-old with plenty of muscle who loses 5 lb of it along with 30 lb of fat, the trade is usually fine: lighter bodies need less muscle to move, and many people feel stronger for their size. For a 68-year-old woman with little muscle to start with, the same loss can push her closer to weakness, falls and frailty. Reviews flag older adults as the group to watch, though the long-term data on GLP-1 users are still thin.

Muscle also matters for what comes after. It burns more energy at rest than fat does, it holds your blood sugar steady after meals, and it is what lets you stay active. People who keep more of it tend to have an easier time staying active once the medication stops.

What happens to muscle if the weight comes back?

Weight often returns after stopping. In the STEP 1 extension, people regained about two-thirds of the weight they had lost within a year of stopping semaglutide. Our guide on weight gain after coming off GLP-1 has the details.

The concern is what that regained weight is made of. A study of older women after a diet (Beavers et al., 2011, not on GLP-1 medication) found that for every kilogram of fat lost they lost about a quarter of a kilogram of lean mass, but for every kilogram of fat regained they got back only about an eighth of a kilogram of lean mass. If the same holds after a GLP-1, a cycle of loss and regain could leave someone at the same weight with less muscle than they started with. That has not been measured directly in GLP-1 users yet; it is a reason to keep strength training going through and after the medication.

How to keep your share at the low end

Two habits carry most of the weight, and the advisory and US activity guidelines agree on both:

1. Eat 1.2–1.6 g of protein per kg of body weight a day (often using an adjusted weight if you carry a lot of extra), spread over 3–4 meals of about 25–30 g. Our guide on how much protein to eat on GLP-1 has a table by height and weight, and the guide to the best protein sources on GLP-1 lists the foods that pack the most into a small appetite.
2. Strength train on 2 or more days a week, working all major muscle groups, and make it a little harder over time. In a review of diet trials in older adults, adding resistance training to a calorie cut prevented about 93% of the lean loss seen with diet alone. In a 2017 NEJM trial of dieting adults around 70, those doing resistance work lost about 2% of lean mass against about 5% with aerobic exercise alone. These are diet trials; trials of training on GLP-1 medications are still running. Our guide on how to prevent muscle loss on GLP-1 has a 12-week home routine.

Then the smaller ones: keep walking most days, sleep 7 or more hours, drink enough (dehydration makes everything harder and muddies a scan), and don't let your intake fall to almost nothing for weeks. If you have no appetite, a protein shake or a yogurt counts.

Not sure how far below your protein number you are today? The free 2-minute protein check gives your daily target in grams and the gap to close.

How to know what you are losing

The simplest plan: do the chair-stand test before you start or as soon as you can, then once a month. If the number holds or rises while the scale drops, you are mostly losing fat. Our symptoms guide has the test steps and the numbers to compare against. If you want a scan, ask your prescriber whether one makes sense for you, especially if you are over 65 or already frail.

When to talk to your doctor

• You are losing strength you can feel: needing your hands to stand from a chair, trouble with stairs, a fall or near fall.
• You are losing weight very fast or eating very little for weeks.
• You are over 65, have had a fall, or have osteoporosis or a condition that affects muscle.
• You have kidney disease and want to raise your protein: your target should come from your care team.

Questions people ask

How much muscle do you lose on Ozempic or Wegovy?

In the STEP 1 body-scan group on semaglutide 2.4 mg, close to 40% of the weight lost was lean mass. Since roughly half of lean mass is muscle, that suggests about 20% of the weight lost was muscle, with a lot of variation between people.

Does Zepbound or Mounjaro cause less muscle loss than Ozempic?

The tirzepatide trial reported a lower lean share (about 25%) than the semaglutide trial (close to 40%), but the trials differed in people, length and method, and no head-to-head trial has compared body composition directly. Don't pick a drug on this number; that is your prescriber's call.

What is the average muscle loss on GLP-1?

The best pooled figure, from a 2025 analysis of placebo-controlled trials, is about 31% of the weight lost as lean mass, or about 2.5 kg of lean mass for an 8 kg loss. Muscle itself is a smaller share than that.

Is losing some muscle on GLP-1 normal?

Yes. Any large weight loss takes some lean mass, and diets alone take roughly a quarter. The aim is to keep the share low, not to reach zero.

Can you build muscle back after GLP-1?

Yes, with strength training and enough protein, at any age, though it is slower when you are older or eating less. It is easier to keep muscle during the loss than to rebuild it after.

Does muscle loss slow your metabolism on GLP-1?

A little. Muscle burns more energy at rest than fat, so losing some lowers your daily burn slightly. Most of the drop in energy use with weight loss comes from simply weighing less.

To find your protein number and see where your meals fall short, take the free 2-minute protein check. More on keeping muscle on weight-loss medication: all our muscle plan guides.

Sources: Wilding J.P.H. et al., STEP 1, "Once-weekly semaglutide in adults with overweight or obesity", N Engl J Med, 2021 (DXA substudy). Jastreboff A.M. et al., SURMOUNT-1, "Tirzepatide once weekly for the treatment of obesity", N Engl J Med, 2022 (DXA substudy). Beavers K.M. et al., GLP-1 receptor agonist therapy and DXA-measured lean mass: a meta-analysis of placebo-controlled trials, Obesity, 2025. Haines M. et al., semaglutide and muscle loss, ENDO 2025 (Endocrine Society, July 2025). Neeland I.J., Linge J., Birkenfeld A.L., "Changes in lean body mass with GLP-1-based therapies and mitigation strategies", Diabetes Obes Metab, 2024. Sattar N. et al., tirzepatide and muscle composition (SURPASS-3 MRI), Lancet Diabetes Endocrinol, 2025. Conte C., Hall K.D., Klein S., "Is weight loss-induced muscle mass loss clinically relevant?", JAMA, 2024. Heymsfield S.B. et al., "Weight loss composition is one-fourth fat-free mass", Obes Rev, 2014. Wilding J.P.H. et al., STEP 1 extension, Diabetes Obes Metab, 2022. Beavers K.M. et al., lean mass after weight regain in postmenopausal women, Am J Clin Nutr, 2011. Weinheimer E.M. et al., Nutr Rev, 2010. Villareal D.T. et al., N Engl J Med, 2017. Mozaffarian D. et al., joint nutrition advisory on GLP-1 therapy, Am J Clin Nutr, 2025. US Physical Activity Guidelines for Americans, 2nd edition. This guide is about food and training and is not medical advice; keep to your prescriber's advice.