Do GLP-1s Cause Muscle Loss? How Much of the Loss Is Muscle

By Eric Parker, founder of TrueCal. Published October 1, 2026. Sources checked October 1, 2026.

Written with AI assistance from the cited studies. Every claim was checked against its source. Editorial policy.

Key takeaways

In SURMOUNT-1, about 75 percent of the weight lost on tirzepatide was fat and about 25 percent was lean mass, the same split as placebo.¹
Lean mass is not the same as muscle. About 74 percent of fat-free mass is water.⁵
A 2026 review of 35 trials of incretin drugs found a median 28.3 percent of weight lost was in measures the review called muscle-related.⁷
In BELIEVE, grip strength went up in every group, including semaglutide alone.³
In 2,292 older adults, leg and grip strength predicted death over about 5 years, while muscle size was not strongly related.¹²

Do GLP-1s cause muscle loss?

In the SURMOUNT-1 body scan substudy, about 75 percent of the weight people lost on tirzepatide, the drug in Mounjaro and Zepbound, was fat and about 25 percent was lean mass. The placebo group, who lost less weight without the drug, had the same 75 to 25 split. Lean mass is not the same as muscle. It also counts water, glycogen, organs and the non-fat part of fat tissue, so the share that is actual skeletal muscle is smaller.¹⁵¹⁸¹⁹

Ozempic and Wegovy contain semaglutide. Mounjaro and Zepbound contain tirzepatide. This article covers what the trials measured. It is not advice on starting, stopping or dosing any medicine.¹⁶¹⁷¹⁸¹⁹

How much of the weight lost on GLP-1 drugs is lean mass?

Body composition in these trials was measured with DXA scans in small subgroups. Here is what the main studies reported.¹³

So most of the weight lost on these drugs is fat. Roughly a quarter to a third shows up as lean mass on a scan.¹³⁶⁷

Study	Who	What was lost
SURMOUNT-1 substudy	160 adults with obesity, 72 weeks	On tirzepatide, weight fell 21.3 percent, fat 33.9 percent and lean 10.9 percent. On placebo, weight fell 5.3 percent, fat 8.2 percent and lean 2.6 percent. About 25 percent of the loss was lean in both groups.¹
BELIEVE	507 adults with obesity, 48 weeks	On semaglutide alone, in the higher dose arm, people lost 9.6 kilograms of fat and 3.9 kilograms of lean, so 71 percent of the loss was fat.³
A 2026 meta-analysis	36 trials of GLP-1 and SGLT2 drugs	Lean mass was 28 percent of weight lost, with a range of likely values of 22 to 34 percent.⁶
A 2026 systematic review	35 trials of incretin drugs	A median 28.3 percent of weight lost was in measures the review called muscle-related. The middle half of studies fell between 15.9 and 39.9 percent.⁷

Is lean mass the same as muscle?

No. A DXA scan or a home body fat scale splits the body into fat and everything else. Everything else includes muscle, but also organs, skin, blood, bone, water and food in the gut.⁵

At the molecular level, fat-free mass is about 74 percent water, 19 percent protein and 6.5 percent mineral. A few other things inflate the lean number during weight loss.⁵

A 2026 study of 374 adults on a 6 month diet found that once the lean part of fat tissue was accounted for, arm and leg lean tissue made up less than 10 percent of the weight lost.¹⁰

Glycogen and its water. The body stores glycogen with three to four parts water. When stores drop early in a diet, that water leaves too, and scans count it as lean loss.⁸
The lean part of fat tissue. Fat tissue is only about 80 to 85 percent fat. The rest is roughly 15 percent water and 5 percent protein. When fat tissue shrinks, that water and protein go with it and get counted as lean.⁵
Organs. In 45 women on a low calorie diet, MRI showed the liver, heart and kidneys shrank by 4 to 6 percent, more in relative terms than muscle, which shrank 3 percent.⁹

So how much of the loss is actual muscle?

MRI can measure muscle directly. A 2024 model built from MRI scans of 897 adults predicted that people dieting without a structured exercise program lose about 2 to 2.5 kilograms of skeletal muscle per 10 kilograms of weight lost in men, and 1 to 1.5 kilograms in women. Two follow-up studies of real weight loss came out close to those numbers.⁴

For a GLP-1 drug specifically, the SURPASS-3 MRI substudy followed 246 adults with type 2 diabetes on tirzepatide for a year. Thigh muscle volume went down 0.64 liters. That drop matched what a large UK population dataset predicts for the same amount of weight loss. Fat inside the muscle fell more than predicted.²

Some real muscle is lost. In BELIEVE, lean mass in the arms and legs, which is mostly muscle, fell 7.9 percent on semaglutide alone at the higher dose.³

Do GLP-1s cause more muscle loss than dieting?

The trials so far don't show that. They show more weight loss, with a similar mix.¹²⁷

Some lean loss comes with losing weight at all. A classic study of 104 women found that extra body weight is itself about 22 to 30 percent lean tissue.¹¹

These numbers have limits. They come from small scan substudies, 160 of 2,539 people in SURMOUNT-1, mostly analyzed after the fact.¹

In SURMOUNT-1, the drug and placebo groups lost the same 75 percent fat and 25 percent lean share. The drug group simply lost more weight in total.¹
In SURPASS-3, thigh muscle loss was in line with the expected amount for the weight lost.²
In the Batsis review, about two thirds of drug trials lost more lean than the review's benchmark, which was about 25 percent of weight lost on DXA or home scales. So did nearly half of the diet or placebo groups that lost weight.⁷

Does losing lean mass hurt strength or function?

In the BELIEVE trial, strength held up. Grip strength went up in every group, including semaglutide alone, up 2.2 kilograms against 1.7 kilograms on placebo. An experimental drug added to semaglutide cut lean loss sharply, so 92 percent of the weight lost was fat. Even so, self-reported physical function at 48 weeks was not different from placebo in any group. The Batsis review noted that none of its 35 trials reported objective physical function tests.³⁷

Outside drug trials, strength tracks health better than muscle size. In 2,292 older adults, leg and grip strength predicted death over about 5 years, while muscle size by CT or DXA was, in the authors' words, not strongly related. The European sarcopenia definition now uses low strength as its main sign, with muscle amount used to confirm.¹²¹³

Does exercise change the picture?

Exercise has been tested next to a GLP-1 drug. In a 1 year trial of 195 adults who had already lost weight on a low calorie diet, exercise plus liraglutide lowered body fat percentage by 3.9 points, about twice as much as either one alone.¹⁴

In dieting older adults without drugs, lean mass fell 2 to 3 percent in groups doing resistance or combined training, and 5 percent in the group doing only aerobic training. Our article on building muscle in a calorie deficit covers lifting and protein in more detail. Talk with your clinician before starting a new exercise plan.¹⁵

How to track it

A home body fat scale reports lean mass, the same mix of water, organs and muscle described above. It can't separate muscle from the rest, and its lean number moves with glycogen and water, so a single reading says little.⁵⁸

What you can track at home.

The trials above measured body composition with DXA scans. If you want a measurement, your doctor can tell you whether a scan makes sense for you.¹³

Weight as a weekly average. Daily weight swings with water and glycogen. A weekly average shows the trend.⁸
Protein. Logging meals shows how much protein you actually eat on days when appetite is low. In TrueCal you type what you ate and protein is counted with the rest.
Strength. Write down what you lift or a simple test you repeat. Strength is the measure tied to health outcomes in the studies above.¹²¹³

Common myths

Myth, up to 40 percent of the weight lost on GLP-1s is muscle. That figure is lean mass, not muscle, and it sits near the top of the range. The median across trials was about 28 percent, and a large part of lean mass is water, glycogen and organ tissue.⁷⁵
Myth, GLP-1 drugs attack muscle. In SURMOUNT-1 the fat to lean split was the same on drug and placebo. The drug changed how much weight was lost, not the mix.¹
Myth, if more than 25 percent of my loss is lean, something is wrong. The 25 percent is an average across a group of 160 people. It isn't a cutoff for one person, and a home scale measures lean mass by a different method than the DXA scans behind that number.¹

A note on medical advice

This article is for general information and isn't medical advice. Talk with your clinician about your own situation.

Worked example

Take a man and a woman who each lose 10 kilograms without a structured exercise program. The 2024 MRI model predicts about 2 to 2.5 kilograms of skeletal muscle lost for him, which is 20 to 25 percent of the 10 kilograms, and about 1 to 1.5 kilograms for her, which is 10 to 15 percent. Compare that with SURMOUNT-1, where lean mass was about 25 percent of the loss. For 10 kilograms, 25 percent is 2.5 kilograms of lean mass, and lean mass also counts water, glycogen and organs, not only muscle.⁴¹⁵

A closer look

Part of the lean mass lost during weight loss is organ tissue. In 45 women on a low calorie diet, MRI showed the liver, heart and kidneys shrank by 4 to 6 percent, more in relative terms than muscle, which shrank 3 percent.⁹

Questions and answers

How much of the weight lost on Ozempic or Wegovy is muscle?

In the BELIEVE trial, about 71 percent of the weight lost on semaglutide alone was fat and the rest was lean mass. Lean mass includes water, glycogen and organs, so the share that is muscle is lower.³⁵

How much of the weight lost on Mounjaro or Zepbound is muscle?

In the SURMOUNT-1 scan substudy, about 25 percent of the weight lost on tirzepatide was lean mass, the same share as placebo. MRI in another tirzepatide trial found thigh muscle shrank about as much as expected for the weight lost.¹²

What is the difference between lean mass and muscle?

Lean mass is everything that isn't fat, including water, organs, skin, bone and gut contents. Muscle is one part of it. About 74 percent of fat-free mass is water.⁵

Will I get weaker on a GLP-1?

In BELIEVE, grip strength went up in every group, including semaglutide alone. If you notice weakness, talk with your clinician.³

Does protein stop muscle loss on GLP-1s?

Protein and resistance training are the two approaches researchers name to limit lean loss during weight loss. Ask your doctor or a dietitian what amount fits you, especially if you have kidney disease.¹⁵²⁰

Keep a food log with TrueCal

Type a meal in plain words and TrueCal adds it to your log, so your intake and your weight trend sit side by side.

Get TrueCal on the App Store Get TrueCal on Google Play

Sources

Diabetes Obes Metab. Look. 2025.
Lancet Diabetes Endocrinol. Sattar. 2025.
Nat Med. Heymsfield. 2026.
Obesity. Heymsfield. 2024.
J Endocr Soc. Tinsley, Heymsfield. 2024.
Diabetes Metab Res Rev. Jobanputra. 2026.
Ann Intern Med. Batsis. 2026.
Am J Clin Nutr. Kreitzman. 1992.
Am J Clin Nutr. Bosy-Westphal. 2009.
Am J Clin Nutr. Roberts. 2026.
Hum Nutr Clin Nutr. Webster. 1984.
J Gerontol A Biol Sci Med Sci. Newman. 2006.
Age Ageing. Cruz-Jentoft. 2019.
N Engl J Med. Lundgren. 2021.
N Engl J Med. Villareal. 2017.
Wegovy label. DailyMed. 2026.
Ozempic label. DailyMed. 2026.
Mounjaro label. DailyMed. 2026.
Zepbound label. DailyMed. 2026.
American Journal of Clinical Nutrition. Wycherley. 2012.