GLP-1 muscle loss: how to preserve lean mass on semaglutide & tirzepatide

Science 10 min read Reviewed

Weight loss from GLP-1 agonists includes lean tissue. Here's what the clinical data shows about the composition of that loss, why it happens, and what strategies are commonly used to mitigate it.

Contents9
  1. 1The composition problem
  2. 2What the clinical data shows
  3. 3Why GLP-1s cause muscle loss
  4. 4Protein intake
  5. 5Resistance training
  6. 6Creatine and other supplements
  7. 7Monitoring body composition
  8. 8Frequently asked questions
  9. 9References

GLP-1 receptor agonists like semaglutide and tirzepatide produce substantial weight loss in trials: a mean -14.9% at 68 weeks for semaglutide 2.4 mg and up to -20.9% at 72 weeks for tirzepatide.12 Part of the weight lost is not fat but fat-free mass, which includes skeletal muscle, organ tissue and water.4

This is not unique to GLP-1 agonists: a 2026 systematic review found muscle-related losses above its benchmark in nearly half of non-drug interventions that produced weight loss, as well as in about two thirds of incretin interventions.4 With weight loss on the scale of these trials, the absolute amount of lean tissue lost can be large, so what shapes the fat-to-lean split matters.

The composition problem

Weight loss is easy to read as fat loss. In practice, weight lost by any method (diet, exercise, surgery or drugs) is a mix of fat mass and fat-free mass, also called lean mass.

Fat-free mass includes skeletal muscle, bone mineral, organ tissue, water and glycogen. Skeletal muscle is the part most people want to keep, for strength and function.

The fat-to-lean split varies widely: in the 2026 review, the share of weight lost from muscle-based measures ranged across studies with an interquartile range of 15.9% to 39.9%.4 Factors usually discussed are the size of the deficit, protein intake, resistance training, starting body composition and age. Appetite suppression strong enough to drive large weight loss can also make it harder to eat enough protein.

Scale weight is misleading

A 15 kg drop on the scale does not mean 15 kg of fat was lost. Across incretin trials, a median 28.3% of weight lost came from muscle-based measures.4 Tracking body composition, not just scale weight, shows more of what is happening.

What the clinical data shows

A 2026 systematic review in Annals of Internal Medicine pooled 35 randomised trials of liraglutide, semaglutide, tirzepatide and dulaglutide that measured body composition by bioimpedance, DXA, CT or MRI.4 Its findings, with the main trials' total weight change for scale:

Trial Compound Total weight loss Measure % Lean of total
STEP 1 Semaglutide 2.4mg -14.9%1 Weight Not in abstract
SURMOUNT-1 Tirzepatide 15mg -20.9%2 Weight Not in abstract
35 incretin RCTs Four incretin drugs Varied All methods Median 28.3% (IQR 15.9-39.9%)4
BIA or DXA studies Incretin drugs Varied Fat-free or lean soft tissue Median ~29%4
CT or MRI studies Incretin drugs Varied Skeletal muscle Median ~25.3%4
Comparator groups Placebo or lifestyle Median -2.5%4 All methods 38% above benchmark

Several points from this review. First, muscle-related losses exceeded the review's benchmarks (about 25% of weight loss for fat-free mass by BIA or DXA, about 15% for muscle by CT or MRI) in about two thirds of incretin interventions.4

Second, methods and reporting differed so much that the authors could not run a meta-analysis, and no study reported objective physical function.4 So a head-to-head lean mass comparison of tirzepatide and semaglutide under identical conditions is not available from this review.

Third, the median study ran 26 weeks with 78 participants.4 Most trial participants were not on structured resistance training or high-protein diets, so these figures describe typical trial conditions rather than an optimised programme.

Why GLP-1s cause muscle loss

Lean mass loss on GLP-1 agonists is not attributed to a single mechanism. Most explanations trace back to the size of the energy deficit.

Caloric deficit magnitude

GLP-1 reduces appetite and food intake and slows gastrointestinal motility.8 A larger deficit and faster loss are generally linked to a higher lean share, though the review above could not pin the relationship down across trials.4

Protein underfeeding

Appetite suppression does not pick between macronutrients. When total food intake drops, protein intake tends to drop with it, which can leave it below what supports muscle.

Reduced physical activity

Less activity and lighter training mean less mechanical stimulus for muscle during an energy deficit. In older adults with obesity, resistance training during caloric restriction prevented 93.5% of the lean mass loss seen with restriction alone.5

Potential direct effects

Whether GLP-1 receptor signalling acts on muscle directly, separate from the energy deficit, is still being studied; no source on it was opened for this page.

Track your GLP-1 protocol

Monitor saturation levels, log doses, and track subjective markers like energy and appetite across your GLP-1 protocol.

Download Milligram

Protein intake

Protein intake is the most discussed dietary factor for lean mass. In a meta-analysis of 49 trials with resistance training, protein supplementation increased gains in fat-free mass.6

The same meta-analysis found no further fat-free mass gain beyond a total protein intake of about 1.62 g per kg of body weight per day.6 For someone weighing 90 kg, 1.62 g/kg/day is about 146 g of protein daily. That trial work was in people training, not specifically in an energy deficit or on GLP-1 drugs.

Practical protein targets

At 1.6 g/kg/day, the arithmetic is: 70 kg = 112g protein, 80 kg = 128g, 90 kg = 144g, 100 kg = 160g. Spread over 3-4 meals, that is roughly 30-50g per meal.

The practical difficulty is that appetite suppression makes eating enough protein hard. At 1,000 calories a day, 120 g of protein is 480 calories, almost half of intake, which takes deliberate food choices such as lean meat, fish, eggs, Greek yogurt and protein shakes.

Protein timing and distribution

Per-meal protein amounts are discussed in muscle physiology research as well as daily totals; no source on a per-meal threshold was opened for this page, so this page gives none.

For people on GLP-1 agonists who manage two meals a day, putting protein first at each meal and adding a protein shake between meals is a common approach.

Resistance training

Resistance training has the clearest trial evidence for keeping lean mass during weight loss. In a meta-analysis of six randomised trials in older adults with obesity, resistance training three times a week for 12 to 24 weeks during caloric restriction prevented 93.5% of the lean mass loss seen with restriction alone.5

What that meta-analysis found:

  • Reduced lean mass loss: resistance training reduced the lean mass lost to caloric restriction by 93.5%, about 0.82 kg, measured by DXA.5
  • Fat and total weight: fat mass and total body mass fell by similar amounts with and without resistance training.5
  • Strength: the change in strength relative to lean mass tended to differ (+20.9% with training vs -7.5% without; p = 0.07).5
  • Population: the trials were in obese older adults on diets of about 15% protein; no trial combining resistance training with GLP-1 therapy was found.

The STEP 3 trial added intensive behavioural therapy to semaglutide and reached -16.0% at 68 weeks, against -14.9% in STEP 1; its abstract does not report body composition.31

1

Frequency: 2-4 sessions per week

The trials in the meta-analysis above used resistance training three times a week.5 Full-body sessions three times a week or upper/lower splits four times a week are common frameworks.

2

Intensity: moderate to heavy loads

Common programming uses moderate to heavy loads, roughly 6-12 repetitions taken close to failure, with load or volume increased gradually over time. No trial in GLP-1 users tested specific loads.

3

Volume: moderate, recovery-aware

Recovery is harder when energy intake is low. A common approach in a deficit is to keep the load per set and trim the number of sets; no trial in GLP-1 users tested this.

4

Compound movements first

Multi-joint exercises (squats, deadlifts, rows, presses, pull-ups) train more muscle per minute than isolation work. A 45-minute session built around 4-5 such movements is a common framework when time and energy are short.

Creatine and other supplements

Creatine monohydrate has a large evidence base. The International Society of Sports Nutrition's position stand says supplementation consistently increases intramuscular creatine, which may explain improvements in high-intensity exercise performance and greater training adaptations.7

No trial of creatine during GLP-1 therapy was found. From the position stand:

  • Training adaptations: improvements in high-intensity exercise performance lead to greater training adaptations.7
  • Recovery and training load: creatine may help athletes tolerate heavy training loads and enhance post-exercise recovery.7
  • Safety: short and long-term supplementation, up to 30 g/day for 5 years, is described as safe and well tolerated in healthy people; habitual intake of about 3 g/day is mentioned for health benefits.7
  • Scale weight: creatine is commonly reported to add some water weight, which can blur a weight trend; no figure was sourced for this page.

Other supplements

HMB (beta-hydroxy beta-methylbutyrate) is a leucine metabolite discussed for lean mass in a deficit. No source on it was opened for this page.

Leucine is discussed as an add-on for lower-protein meals; with enough total protein from food it is usually considered unnecessary. No source on it was opened for this page.

Vitamin D and omega-3 fatty acids are also mentioned in this context; no source on them was opened for this page.

Monitoring body composition

Scale weight cannot tell fat loss from muscle loss, and two people losing the same weight can end up with very different body composition. Tracking body composition shows the difference.

DEXA scan

Dual-energy X-ray absorptiometry (DEXA) gives a regional breakdown of fat mass, lean mass and bone mineral content, and is one of the methods used in the trials reviewed above.4 Repeat scans under the same conditions (time of day, hydration, fasting) compare best.

Bioelectrical impedance

Bioelectrical impedance (BIA) is the other method used in many trials.4 Consumer scales are convenient, and readings move with hydration, so they are better for trends over months than for one-off figures.

Practical proxies

In the absence of DEXA or bioimpedance, several practical measurements provide useful proxy data:

  • Waist circumference: Declining waist measurement alongside stable or slowly declining weight suggests favorable body composition change (more fat loss, less muscle loss)
  • Strength metrics: holding or improving key lifts (squat, bench, deadlift or machine equivalents) during weight loss is a functional signal; no trial in the review measured physical function.
  • Progress photos: Standardized photos (same lighting, angle, time of day) every 2-4 weeks capture visual changes that neither the scale nor the mirror in isolation can convey
  • Subjective markers: Energy levels, recovery quality, and physical performance during daily activities provide qualitative signals about lean mass status
Tracking in Milligram

Milligram's daily check-in system tracks subjective markers like energy, recovery, and appetite alongside your GLP-1 dosing schedule. Over weeks and months, these trends provide context for how your body is responding to the protocol, and whether adjustments to training or nutrition may be worth considering.

Frequently asked questions

How much muscle do you lose on semaglutide?

In STEP 1, mean body weight change with semaglutide 2.4 mg was -14.9% over 68 weeks.1 The STEP 1 abstract does not give body composition. Across 35 incretin trials, a 2026 systematic review found a median 28.3% of weight lost came from muscle-based measures, with an interquartile range of 15.9% to 39.9%.4

Is muscle loss on GLP-1 agonists permanent?

Muscle responds to training and protein, and in trials resistance training kept most lean mass during caloric restriction.5 No study of lean mass regain after stopping a GLP-1 drug was opened for this page.

How much protein should you eat on semaglutide or tirzepatide?

A meta-analysis of resistance training trials found protein supplementation increased fat-free mass gains, with no further gain beyond about 1.62 g/kg/day of total protein.6 Spreading protein over several meals, with shakes or protein-dense foods, is the usual way to reach a target when appetite is low.

Does resistance training prevent muscle loss on GLP-1s?

It reduces it. In six trials of older adults with obesity, resistance training three times a week during caloric restriction prevented 93.5% of the lean mass loss seen with restriction alone.5 No trial combining resistance training with GLP-1 therapy was found.

Does creatine help preserve muscle on semaglutide?

No trial of creatine during GLP-1 therapy was found. The International Society of Sports Nutrition says creatine increases intramuscular creatine and may lead to greater training adaptations, and that supplementation is safe and well tolerated in healthy people.7

References

  1. Wilding JPH et al. Once-weekly semaglutide in adults with overweight or obesity (STEP 1). N Engl J Med. 2021. pubmed.ncbi.nlm.nih.gov
  2. Jastreboff AM et al. Tirzepatide once weekly for the treatment of obesity (SURMOUNT-1). N Engl J Med. 2022. pubmed.ncbi.nlm.nih.gov
  3. Wadden TA et al. Effect of subcutaneous semaglutide vs placebo as an adjunct to intensive behavioral therapy on body weight (STEP 3). JAMA. 2021. pubmed.ncbi.nlm.nih.gov
  4. Batsis JA et al. Effect of incretin-based and nonpharmacologic weight loss on body composition: a systematic review. Ann Intern Med. 2026. pubmed.ncbi.nlm.nih.gov
  5. Sardeli AV et al. Resistance training prevents muscle loss induced by caloric restriction in obese elderly individuals: a systematic review and meta-analysis. Nutrients. 2018. pubmed.ncbi.nlm.nih.gov
  6. Morton RW et al. Protein supplementation and resistance training-induced gains in muscle mass and strength: a systematic review, meta-analysis and meta-regression. Br J Sports Med. 2018. pubmed.ncbi.nlm.nih.gov
  7. Kreider RB et al. International Society of Sports Nutrition position stand: safety and efficacy of creatine supplementation in exercise, sport, and medicine. J Int Soc Sports Nutr. 2017. pubmed.ncbi.nlm.nih.gov
  8. Holst JJ. The physiology of glucagon-like peptide 1. Physiol Rev. 2007. pubmed.ncbi.nlm.nih.gov