"You lose muscle on GLP-1s" is one of the most repeated claims in the category and one of the most poorly sourced. The underlying data comes from a small number of body-composition substudies using dual-energy X-ray absorptiometry. This is what they measured and what the measurement can and cannot tell you.
Key Takeaways
- DEXA substudies in STEP 1 and SURMOUNT-1 both found that a meaningful share of total weight lost was lean mass — a finding consistent with weight loss generally, not unique to GLP-1s.
- In both substudies, lean mass as a proportion of total body mass increased, because fat mass fell faster than lean mass.
- DEXA 'lean mass' is not synonymous with skeletal muscle: it includes organ tissue, connective tissue and body water.
- No published GLP-1 trial has used the intervention most likely to change the outcome — supervised progressive resistance training — as a randomized arm.
- Comparison against surgical and caloric-restriction literature suggests the lean-mass fraction lost on GLP-1s is broadly in the expected range for the magnitude of loss.
What DEXA actually measures
Dual-energy X-ray absorptiometry passes two X-ray energies through the body and infers tissue composition from differential attenuation. It resolves the body into three compartments: bone mineral content, fat mass, and lean soft tissue mass.
That third compartment is where interpretation goes wrong in popular coverage. Lean soft tissue includes skeletal muscle, but also organs, skin, connective tissue and — importantly — water. A person losing weight rapidly experiences glycogen depletion, and glycogen is stored with water. Some of the early "lean mass" decline on any weight-loss intervention is water leaving with glycogen rather than contractile tissue disappearing.
DEXA also cannot distinguish muscle quality, cross-sectional area of specific muscle groups, or functional strength. A study can report a lean-mass decline while measured strength is unchanged, and both results can be correct.
The STEP 1 body-composition substudy
A subset of STEP 1 participants underwent DEXA scanning at baseline and at week 68. The substudy found that total fat mass fell substantially and that lean mass also fell in absolute terms. Critically, the ratio shifted in the favorable direction: lean mass as a percentage of total body mass increased, because adipose tissue was lost disproportionately.
Regional analysis showed the largest absolute reductions in visceral fat, which is the depot most strongly associated with cardiometabolic risk.
The SURMOUNT-1 substudy
SURMOUNT-1's body-composition substudy used the same approach on tirzepatide. The pattern reported was similar in shape: absolute reductions in both compartments, with fat mass falling faster and the lean-to-total ratio improving. Given tirzepatide's larger total weight reduction in the parent trial, the absolute lean-mass change was correspondingly larger.
Placing the finding in context
The number that circulates online — some fraction of weight lost being lean tissue — is only interpretable against a comparator. Three comparators are available in the literature.
- Caloric restriction without pharmacotherapy. Decades of intervention studies report a lean-mass fraction in a similar range for comparable weight loss. This is the baseline expectation for losing weight by any means.
- Bariatric surgery. The surgical literature reports lean-mass losses at least as large, often larger, reflecting the greater magnitude and speed of loss.
- Loss with concurrent resistance training and adequate protein. This is where the literature consistently shows preserved or improved lean mass — and it is the arm the GLP-1 trials did not run.
The reasonable conclusion is that the lean-mass changes observed are a property of substantial, relatively rapid weight loss rather than a pharmacologic property of GLP-1 receptor agonism specifically. That is a materially different claim from "these drugs waste muscle," and the distinction is not rhetorical: it determines whether the appropriate response is to avoid the drug or to add a training and protein intervention alongside it.
| Compartment | What DEXA sees | What it includes | Common misreading |
|---|---|---|---|
| Fat mass | Adipose tissue by attenuation | Subcutaneous and visceral depots | Regional estimates are less precise than whole-body |
| Lean soft tissue | Non-fat, non-bone soft tissue | Skeletal muscle, organs, connective tissue, water | Reported as if it were skeletal muscle alone |
| Bone mineral content | Mineralized bone | Skeletal mineral | Usually omitted entirely from popular summaries |
The evidence gap that matters
The most useful trial in this space has not been published: a randomized comparison of GLP-1 therapy alone versus GLP-1 therapy plus supervised progressive resistance training and a protein intake target, with DEXA and functional strength testing as co-primary outcomes. Until that exists, recommendations to train while on these medications rest on extrapolation from the general weight-loss literature. That extrapolation is well-founded, but it is extrapolation.
Smaller trials combining GLP-1 therapy with exercise have been conducted and generally point in the expected direction. Their sample sizes limit how firmly conclusions can be drawn about body-composition endpoints specifically.
Frequently Asked Questions
Do GLP-1s cause more muscle loss than other weight-loss methods?
The published DEXA data does not support that framing. The lean-mass fraction reported in GLP-1 substudies falls broadly within the range reported for comparable weight loss achieved by caloric restriction or surgery.
Does the lean mass come back if weight is regained?
Regain after weight loss is generally fat-predominant in the broader obesity literature. The GLP-1 discontinuation studies did not include serial DEXA in their off-treatment extensions, so this has not been directly measured in this drug class.
Is grip strength or functional testing available from these trials?
Functional strength was not a primary or key secondary endpoint in the major registration trials. This is one of the clearest holes in the evidence base.
How reliable is DEXA compared with MRI?
MRI and CT resolve specific muscle groups and can distinguish intramuscular fat, which DEXA cannot. DEXA is used in large trials because it is faster, cheaper and lower-dose, not because it is the most precise available method.
References
- Wilding JPH et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. New England Journal of Medicine, 2021 (STEP 1, including body-composition substudy reporting).
- Jastreboff AM et al. Tirzepatide Once Weekly for the Treatment of Obesity. New England Journal of Medicine, 2022 (SURMOUNT-1).
- Heymsfield SB et al. Reviews of body-composition methodology and the interpretation of lean soft tissue by DEXA.
- Weiss EP et al. Effects of weight loss on lean mass, strength and bone — comparative intervention literature.
Comparing providers, not just trials?
Our homepage keeps a maintained side-by-side grid of GLP-1 telehealth providers with current pricing, formats and disclosures. That is where the provider comparisons live — this research desk stays citation-first.
Open the provider comparison grid