Muscle loss on GLP-1 drugs: how much, why, and what actually helps

The STEP 1 DEXA substudy is the most cited number. In a subset of participants on semaglutide 2.4 mg who had body composition scans, about 39 percent of the total weight lost over 68 weeks was lean mass and about 61 percent was fat. Reviews pooling the available body composition data across GLP-1 trials put the fat-free mass share at 30 to 40 percent. The SURMOUNT-1 tirzepatide substudy reported a similar range.

In absolute terms, a person who loses 30 kg with 35 percent of it as fat-free mass has lost about 10 kg of fat-free mass. That sounds alarming until the composition of fat-free mass is unpacked. It includes muscle, but also water, glycogen, bone, organ tissue and the lean fraction of fat tissue itself. A large part of the early fat-free mass drop is water and glycogen that leave as glucose stores shrink, and a further part is organ and connective tissue that scales down with body size.

The number that matters for function is skeletal muscle, and the trials mostly did not measure it directly. DEXA reports lean mass, not muscle mass, and the two are not interchangeable. What the trials did find is that the ratio of lean to fat mass improved in most participants, because fat fell faster than lean. Losing 10 kg of lean mass while losing 20 kg of fat leaves a person with a higher proportion of lean tissue than before.

Rapid weight loss of any kind drops fat-free mass. Bariatric surgery, very low calorie diets and GLP-1 drugs all produce a lean mass share in the 20 to 40 percent range in the studies that measured it. The body in a large energy deficit breaks down some protein for fuel and for gluconeogenesis, muscle that is not loaded tends to be the first source, and a smaller body needs less supporting tissue. GLP-1 drugs reach that deficit faster and sustain it longer than most diets, which is why the absolute loss is larger.

There is a drug-specific contribution too. GLP-1 agonists suppress appetite hard enough that protein intake often falls below what muscle maintenance needs, and the nausea and early fullness of the titration period make large protein meals unpleasant. People eating 1200 calories a day with no attention to protein are in a different situation from people eating the same calories at 1.6 grams of protein per kilogram. The drug does not target muscle; it removes the hunger that would otherwise keep protein intake up.

Age and starting fitness shift the split. Older adults, people with low baseline muscle and people who are sedentary lose a higher fraction as lean mass. The compound pages for semaglutide and tirzepatide both flag lean mass loss as a rate-of-loss effect, and the honest framing is that it is a cost of fast loss rather than a side effect of the molecule.

Resistance training is the one intervention consistently shown to preserve lean mass during weight loss across diet, surgery and drug studies. The mechanism is direct: loaded muscle signals the body to keep it, and the protein synthesis response to training persists even in an energy deficit. Trials of structured lifting during calorie restriction have shown lean mass preserved or even gained while fat fell, which does not happen with cardio alone.

The specific trials of resistance training during GLP-1 treatment are small and recent, but they point the same direction as the broader literature. Two or three full-body sessions a week with progressive load are the pattern most used. The compound pages put it plainly: protein and lifting decide how much of the lost weight is fat. Lifting also counters the fatigue and reduced exercise tolerance that the early weeks of titration bring, though people often need to scale intensity down right after each dose step.

Cardio has its own benefits but is not a muscle-preservation tool. Someone doing only walking or cycling on a GLP-1 drug should expect the trial-average split. Someone lifting with adequate protein should expect a better one, and the better ratio of fat to lean loss compounds over a year of treatment.

Protein is the second evidence-backed counter. Higher protein intake during energy restriction preserves lean mass in trial after trial, with the most studied targets in the range of 1.2 to 1.6 grams per kilogram of body weight per day, and some sports nutrition guidance going higher for people training hard. On a GLP-1 drug the challenge is not knowing the target but hitting it when appetite is suppressed and a chicken breast feels like a full meal.

What people report works: eating protein first at each meal before anything else, splitting intake across more small meals rather than two large ones, and using liquid protein when solid food is unappealing during a titration flare. None of that is a trial result, but it follows from the mechanism, and it is the advice most obesity clinics give alongside the drug. The weight loss slows slightly when calories from protein stay high, which is the point: slower loss with a better composition.

The common additions are growth hormone secretagogues: ipamorelin, usually paired with CJC-1295 without DAC; sermorelin; and tesamorelin. All four raise growth hormone pulses and downstream IGF-1, and growth hormone is anabolic, so the logic is that more of it during a cut should spare muscle. The tirzepatide and ipamorelin stack is the most popular expression of that idea.

The evidence for the logic is thin in the specific case. Sermorelin raised IGF-1 and modestly increased lean body mass over 16 weeks in healthy older adults in a 1997 trial, and tesamorelin reduced visceral fat by roughly 15 to 18 percent over 26 weeks in HIV lipodystrophy. Neither trial was run during GLP-1 treatment, in an energy deficit of that size, or against an outcome of lean mass preserved. Ipamorelin's human data are a single-dose pilot and a failed phase 2 for postoperative ileus. CJC-1295 without DAC has essentially no human efficacy data of its own.

The honest statement is that no human trial shows any of these peptides prevents lean mass loss during GLP-1 treatment. They may do something; nobody has measured it. There is also a practical tension noted on the tirzepatide compound page: ghrelin mimetics like ipamorelin raise hunger, which can partly offset the appetite suppression the GLP-1 drug is there to provide. The ipamorelin vs sermorelin comparison covers how the two differ.

Because lean mass loss tracks the rate of weight loss, the dose that produces the fastest loss also produces the most lean mass loss. Tirzepatide at 15 mg produced about 20.9 percent loss in SURMOUNT-1 against about 15 percent at 5 mg; the higher dose lost more of everything, including fat-free mass. People who care about composition sometimes hold at a lower step and accept slower loss, which the labels allow since every increase is conditional on tolerance.

The same reasoning applies to retatrutide, where the 2026 phase 3 topline reported about 28 percent at 80 weeks. Larger loss means larger absolute lean mass loss unless training and protein scale up with it. The retatrutide body composition data are not yet published. The tirzepatide dosage chart and retatrutide phase 3 results pages cover the dose and trial details.

Measuring is more useful than guessing. A DEXA scan before starting and every few months costs less than a month of peptides and tells you whether the split is on the trial average or better. Strength in the gym is a cheaper proxy: if lifts are holding or climbing during the loss, muscle is being preserved regardless of what the scale says.