GLP-1 drugs and muscle loss: what the body composition data actually shows

The GLP-1 muscle loss question every new patient asks

Before the first injection, most people want to know two things: how much weight will I lose, and will I lose muscle? The first question has gotten a lot of press. The second one deserves a real answer.

The concern is legitimate. Muscle mass isn't just about aesthetics. It governs resting metabolic rate, insulin sensitivity, physical function as you age, and — critically — what happens if you ever stop the drug. Lose muscle on the way down, and regaining weight becomes much easier because you have a lower metabolic floor.

For years, the honest answer was: we don't have great data. Most GLP-1 trials measured total weight and fat mass. Lean mass was an afterthought. That started changing with a 2025 body composition substudy buried inside SURMOUNT-1.

What the SURMOUNT-1 substudy actually found

SURMOUNT-1 is Eli Lilly's pivotal Phase 3 trial for tirzepatide. The main trial enrolled 2,539 adults. A smaller substudy — 160 participants — ran DXA scans at baseline and week 72. DXA (dual-energy X-ray absorptiometry) is the clinical standard for body composition; it separates fat mass from lean soft tissue with meaningful precision.

The published result: participants on tirzepatide lost an average of 21.3% of body weight over 72 weeks. Of that, 33.9% came from fat mass and 10.9% came from lean mass. Run the math and you get roughly a 75/25 split — three-quarters fat, one-quarter lean. That is meaningfully better than researchers expected going in.

The 10.9% lean mass figure is the one worth sitting with. It represents a real, measurable decline in muscle and other lean tissue. Researchers who analyzed the SURMOUNT-1 data noted that it is roughly equivalent to two decades of age-related sarcopenia — muscle loss that normally unfolds over 20 years — compressed into about 18 months of treatment. For a 40-year-old who is already losing half a pound of muscle per year naturally, that is not a trivial consideration.

Semaglutide versus tirzepatide: the ratio gap

The tirzepatide data looks better than semaglutide — but the comparison needs context.

Meta-analyses of GLP-1 agonist body composition studies put semaglutide's lean-to-fat loss ratio at roughly 39–40% lean. That is substantially worse than tirzepatide's 25%. A 2026 systematic review in the International Journal of Obesity pooling GLP-1 and GLP-1/GIP agonist trials found the dual agonists (tirzepatide) consistently outperformed GLP-1 monotherapy on this metric.

One theory: GIP receptor activation in tirzepatide may directly protect lean tissue. GIP plays a role in anabolic signaling in muscle, and its receptor is expressed in skeletal muscle. That mechanism is biologically plausible but not proven in this context.

Here is the complication: tirzepatide causes significantly more total weight loss. If you lose 21% of body weight on tirzepatide versus 13% on semaglutide, the absolute amount of lean tissue lost can end up comparable — even though the ratio favors tirzepatide. The percentage is not the same as the pound figure on the scale. Anyone comparing the two drugs should think in absolute terms, not just ratios.

Who is most at risk

Not everyone faces the same muscle-loss risk on a GLP-1. The people most likely to lose clinically meaningful lean mass are:

What actually preserves muscle — and what does not

Resistance training is not optional if muscle preservation matters to you. That is the clearest thing the evidence says. A 2025 case series published in SAGE Open Medicine documented individuals who maintained or gained lean mass during GLP-1 therapy — every one of them was doing structured resistance training.

Protein intake matters nearly as much. The current best-evidence target for people in a caloric deficit is roughly 1.6 grams of protein per kilogram of body weight per day. On a GLP-1, appetite suppression can make this hard to hit — nausea and reduced appetite mean many patients eat far less than they normally would. Tracking protein specifically, not just total calories, is useful during the dose-escalation phase.

Creatine monohydrate has the most evidence of any supplement for preserving lean mass and strength during restriction. It is not a replacement for training, but the data for it is cleaner than for almost anything else in the supplement space — 5g daily, the cheapest form, nothing exotic needed.

Things that are often recommended but have weaker evidence in this specific context: HMB, leucine supplementation above normal protein intake, peptide stacks, and BCAAs when protein intake is already adequate. They are not harmful, but the evidence base is thin for GLP-1-specific muscle preservation specifically. We rank the individual peptides people add for this reason, Ipamorelin, CJC-1295, BPC-157, in our guide to peptides and GLP-1s.

The LEAN-PREP trial

There is an ongoing clinical trial — NCT06885736, called the LEAN-PREP study — designed specifically to answer the question of how much resistance training and protein can protect lean mass during semaglutide and tirzepatide therapy. It is estimated to complete in 2029.

We do not have those results yet. What we do have is the general exercise physiology literature, which is consistent: resistance training during caloric restriction preserves muscle across essentially every population studied. There is no reason to expect GLP-1 patients to be an exception. The LEAN-PREP data will give us more precise effect sizes than we currently have.

The weight regain risk is the real reason this matters

There is a larger reason to care about lean mass beyond aesthetics or strength. When people stop GLP-1 drugs — and many do, for cost reasons, side effects, or access — weight typically returns. Studies show 50–60% of lost weight comes back within a year of stopping.

That regained weight tends to come back as fat, not muscle. If you lost 10 pounds of lean mass during treatment, you do not get those 10 pounds back as muscle when you stop — you get them back as fat. That leaves you with a worse body composition than before you started, at the same scale weight. Protecting lean mass during treatment is, in part, protecting yourself against that scenario.

The drugs work. The 21% weight loss in SURMOUNT-1 is real and clinically meaningful. But the muscle question deserves a direct answer, not reassurance. Some lean mass loss is real, it is manageable with training and protein, and it matters more than most patients are told. The same rapid weight loss also triggers temporary hair shedding in a meaningful fraction of users through the same mechanism — both are responses to physiological stress during aggressive caloric restriction.