GLP-1 and muscle loss: how to protect lean mass

Key takeaways

• In the SURMOUNT-1 tirzepatide body-composition substudy (n=160), about 25 percent of the weight lost was lean mass (fat mass fell 33.9 percent, lean mass 10.9 percent), the same lean split as the placebo/diet group .
• In the STEP-1 semaglutide substudy, fat mass dropped about 19 percent and lean mass about 10 percent, and lean mass as a share of body weight actually rose about 3 points .
• DXA lean (fat-free) mass overstates muscle: it bundles in water, glycogen, and organ tissue, and reads systematically higher than direct D3-creatine muscle measurement .
• The two levers with trial evidence are resistance training and higher protein (around 1.6 g/kg/day). Trained men in a 40 percent energy deficit eating 2.4 g/kg gained 1.2 kg of lean mass while losing 4.8 kg of fat .
• Semaglutide cut ad libitum food intake about 35 percent , so protein has to be defended on purpose or it falls with everything else; older adults also need more protein per meal (about 0.40 vs 0.24 g/kg) .

Skip to:

• What the trials measured: about a quarter of the loss was lean
• DXA lean mass is not the same as muscle
• What actually protects lean mass
• Protein falls with your appetite unless you defend it
• Amylin, triple agonists, and muscle-sparing add-ons
• Older adults have less muscle to spare
• What the trials still do not answer
• The honest read

What the trials measured: about a quarter of the loss was lean

The big GLP-1 trials report total weight change as the headline, with smaller DXA substudies measuring fat and lean mass. In the STEP-1 substudy (a 140-person subgroup of the trial where semaglutide 2.4 mg drove about 15 percent weight loss ), total fat mass fell about 19 percent and total lean body mass about 10 percent over 68 weeks, so lean mass as a proportion of body weight rose about 3 points .

The tirzepatide read is cleaner. The SURMOUNT-1 body-composition substudy (n=160) found fat mass down 33.9 percent and lean mass down 10.9 percent at 72 weeks; of the weight lost, about 25 percent was lean tissue on both tirzepatide and placebo . In other words, the drug did not shift the fat-to-lean split of the loss compared with diet alone. SURMOUNT-1 was a 72-week phase 3 obesity trial with weight loss up to about 21 percent at the top dose .

So the fat-to-lean ratio is roughly what any fast weight loss produces in people who are not lifting and not deliberately eating protein. The fraction itself is unremarkable. What makes it worth watching is the size and speed: the total loss is large and quick, so the absolute lean-mass number is bigger and it lands fast.

DXA lean mass is not the same as muscle

DXA fat-free mass is not pure muscle. It includes skeletal muscle, organ tissue, connective tissue, total body water, and glycogen with its bound water (each gram of stored glycogen holds roughly 3 grams of water). A meaningful slice of the early lean-mass drop in any rapid weight loss is glycogen and water leaving, plus shrinking gut and adipose stromal tissue, not contractile muscle.

Direct muscle measurement backs this up. D3-creatine dilution reads systematically lower than DXA lean body mass, meaning DXA overestimates true skeletal muscle mass, and DXA lean mass tracks weakly with strength and fall risk while D3-creatine muscle mass tracks well . That does not erase the concern, but it does mean a scary DXA lean-mass number overstates how much actual muscle you lost.

What actually protects lean mass

• Resistance training, two to three sessions a week, with progressive overload. During an energy deficit lifting mostly protects the muscle you have rather than adds new muscle, because a deficit blunts training-driven lean-mass gains .
• Higher protein, around 1.6 g/kg body weight per day (more is fine). In a controlled trial, trained men in a 40 percent deficit eating 2.4 g/kg gained 1.2 kg of lean mass and lost 4.8 kg of fat, versus roughly flat lean mass in the lower-protein group . Meta-analysis puts the point of diminishing returns for protein plus lifting at about 1.6 g/kg/day .
• A slower rate of loss where that is reasonable. A gentler deficit tends to cost a smaller lean-mass fraction, which trades off against total fat lost.
• Sleep of about 7 hours or more. In a randomized crossover trial, cutting sleep to 5.5 hours raised the share of weight lost as fat-free mass by about 60 percent during the same calorie restriction .
• Keeping daily movement up (steps, general activity). GLP-1 fatigue and low appetite often pull spontaneous activity down, which deepens the deficit in a way that costs lean mass.

Protein falls with your appetite unless you defend it

This is the practical trap on a GLP-1. In a controlled study, semaglutide 2.4 mg cut ad libitum energy intake by about 35 percent versus placebo . If protein just scales down with everything else you eat less of, absolute protein can drop well under the roughly 1.6 g/kg needed to hold lean mass in a deficit.

So protein has to be picked first, before appetite is spent. Whey, casein, lean meat, dairy, eggs, and protein-fortified foods are the usual levers. A workable target is 30 to 40 g of protein at each of three to four meals, which clears the per-meal threshold that maximally stimulates muscle protein synthesis. That threshold is higher for older adults: about 0.40 g/kg per meal versus about 0.24 g/kg for younger adults .

Amylin, triple agonists, and muscle-sparing add-ons

Cagrilintide is a long-acting amylin analog studied on its own and paired with semaglutide as CagriSema. In the REDEFINE-1 phase 3 trial, CagriSema drove about 20.4 percent weight loss versus 3.0 percent on placebo, and more than semaglutide alone . Body-composition data are still accruing, so whether adding amylin changes the lean-to-fat ratio of the loss is an open question.

Tirzepatide (GLP-1 plus GIP) and retatrutide (GLP-1, GIP, plus glucagon) reach lower weight nadirs than semaglutide. Glucagon agonism has been hypothesized to favor fat breakdown specifically, which could in theory shave the lean-mass fraction, but that has not been shown in head-to-head body-composition data.

The newest angle targets muscle directly. In a 2026 phase 2 trial, bimagrumab (an antibody meant to cut fat and add muscle) added to semaglutide produced more weight loss than semaglutide alone (about 17.8 versus 14.2 kg at 48 weeks) and was built to push the loss toward fat rather than lean tissue . It is not approved, and whether protecting DXA lean mass protects real-world strength and function is still unproven.

Older adults have less muscle to spare

Older adults start with less reserve, and sarcopenic obesity (low muscle mass and low muscle function alongside high body fat) is a defined diagnosis with formal criteria from the ESPEN and EASO societies . Losing lean mass fast from a lower baseline is where function starts to matter more than the number on a scan.

The Villareal trial in dieting obese older adults is the closest evidence for what to do about it. Combining aerobic and resistance exercise improved physical function more than either type alone (a 21 percent gain on a performance test versus about 14 percent), and the combined and resistance-only groups lost less lean mass than the aerobic-only group (about 3 and 2 percent versus 5 percent) . Lifting was the part that protected muscle.

What the trials still do not answer

• Whether the lean-mass lost on a GLP-1 translates into long-term function outcomes (falls, frailty, mortality), set against the well-documented metabolic benefits.
• Whether prescribed resistance training plus protein guidance during titration meaningfully shifts the lean-to-fat ratio of the loss.
• Whether multi-receptor or muscle-targeted add-ons (GIP, glucagon, amylin, activin-pathway drugs) preserve real muscle, not just DXA lean mass, at equal total weight loss.
• Whether sex, age, baseline insulin resistance, or starting lean mass predict how much lean mass a given person loses.
• What the best post-nadir maintenance protein and training plan looks like for holding onto muscle once weight stabilizes.

The honest read

Lean-mass loss during fast weight loss is real and well documented, and it happens with dieting and surgery too. What drives it is the energy deficit and the missing training stimulus, the same on a GLP-1 as off one. The levers that change the outcome (lifting, protein, sleep, a loss rate you can hold) have trial and meta-analytic support. Whether or not a prescriber brings any of it up, those are the levers you actually control.

For research and educational purposes only. Not medical advice.

pepSmart has not commissioned independent clinical review of this article.

More on how we write and source these pieces: Editorial process and contributor disclosure and Sourcing posture.

Spot an error? Email corrections via /about.

Sources: 15 entries, all primary canon (PubMed and PubMed Central records from NEJM, Diabetes Obesity and Metabolism, and related peer-reviewed journals), last reviewed 2026-07-08.

Related tools

• Tirzepatide dose calculator - Run tirzepatide-focused vial draw math.
• GLP-1 conversion calculator - Convert a GLP-1 mg dose to U-100 units and ml.
• GLP-1 ramp planner - Preview a linear educational dose-step table.
• Peptide half-life calculator - Estimate single-dose decay from cited half-life constants.
• PK simulator overview - Public overview of the Pro pharmacokinetic simulator.
• Body-recomp simulator overview - Public overview of the Pro body-recomposition simulator.