Attia called GLP-1s a muscle-wasting trap. The data and the lived experience of physique-focused users tell a more nuanced story when resistance training and protein are programmed correctly.
Peter Attia spent most of 2023 warning that GLP-1 agonists cause "extreme muscle loss" and shouldn't be used by people trying to lose the last 10 pounds. A year later, that take is aging poorly. The headline DEXA numbers from STEP and SURMOUNT trials were generated in sedentary, often elderly subjects eating whatever protein their appetite allowed - which on semaglutide is not much. Strip those two variables out and the picture changes substantially. The looksmaxxing-lean question isn't whether GLP-1s can cause muscle loss; it's how much of that loss is actually muscle versus other lean tissue, and how completely resistance training plus aggressive protein offsets it.
The widely cited "40% of weight lost is lean mass" figure from STEP-1 is a DEXA estimate of fat-free mass (FFM), not a biopsy of skeletal muscle. FFM includes:
When people drop 15% of bodyweight rapidly, organ mass shrinks, gut volume contracts, glycogen and intracellular water fall, and skin slowly remodels. All of that lands in the "lean mass loss" column on a DEXA. The MRI substudy of SURMOUNT-1 found that thigh muscle volume loss tracked roughly proportional to total weight loss - meaning the muscle-specific fraction of lean loss was much smaller than the FFM number implies. This is the nuance Attia's original framing flattened.
This doesn't mean muscle is untouched. It means the panic number is inflated, and the levers that protect skeletal muscle in any deficit - protein and mechanical tension - protect it on tirzepatide and semaglutide too.
GLP-1 trial subjects famously eat 1,200-1,500 kcal and around 0.6-0.8 g/kg of protein because the drug crushes appetite and protein is the most satiating macro. Physique-focused users who pre-commit to protein get a completely different body-composition outcome. The protocol that the bodybuilding and looksmaxxing community has converged on:
| Lever | Sedentary trial subject | Physique-focused protocol |
|---|---|---|
| Protein | ~0.6 g/kg | 1.6-2.2 g/kg lean mass, front-loaded |
| Resistance training | None or walking | 3-5x/week, progressive overload |
| Deficit size | ~25-30% via appetite crash | 15-20%, controlled |
| Dose escalation | Per label, fast titration | Slow titration, lowest effective dose |
The slow-titration piece matters more than people give it credit for. Jumping straight to 2.4 mg semaglutide or 10-15 mg tirzepatide collapses appetite so hard that hitting 180 g of protein becomes a daily grind through nausea. Staying at 0.25-0.5 mg semaglutide or 2.5-5 mg tirzepatide for an extended block keeps the appetite suppression in a useful range - enough to enforce the deficit, not enough to make food aversive.
Volume management is where most users get this wrong. The instinct is to train harder to "save" muscle. In a real deficit on a drug that's blunting recovery via reduced food intake and lower insulin spikes, that backfires. The literature on resistance training in a deficit (Longland 2016, Helms et al. reviews) is consistent: heavy load and proximity to failure preserve muscle better than high-volume metabolic work.
A defensible template:
The single best biofeedback signal is whether working-set loads are holding. DEXA noise will tell you nothing useful week to week. A flat or rising top set on incline press while bodyweight drops 0.5-0.7% per week is the actual proof that the protocol is working.
The practical bottleneck on a GLP-1 isn't knowing protein matters - it's getting it down when nothing sounds good. Strategies the community has settled on:
Creatine monohydrate at 5 g/day is non-negotiable on this protocol. It's the cheapest, best-evidenced muscle-preservation tool in any deficit and the case for it gets stronger when food intake is compressed.
GLP-1s are most useful in the 15-25% body fat range, where appetite is the rate limiter and the deficit is large in absolute terms. Below ~10%, the drugs become a worse tool: appetite is already fragile, training output is already compromised, and the marginal fat loss is small enough that the protein-intake friction outweighs the appetite benefit. Most experienced users taper off the GLP-1 around 12-13% body fat and finish the cut on diet alone, often with a low-dose beta-2 agonist or yohimbine for the stubborn-fat phase.
"GLP-1 antagonists don't cause muscle loss if you lift and eat protein" is the emerging community consensus pushing back on Attia's original framing - and the n-of-many DEXA scans posted across physique forums broadly support it.
The cleanest use case: a lean-bulk overshoot or an offseason that drifted to 18-20%, brought back to 12% over 12-16 weeks on 0.5-1 mg semaglutide or 2.5-5 mg tirzepatide, with protein at 2 g/kg lean and four hard lifting sessions a week. Strength holds, top sets hold, and the DEXA "lean loss" that shows up is mostly water and gut mass that comes back the moment maintenance calories return.
The muscle-wasting narrative around GLP-1s was built on trials of sedentary subjects eating sub-threshold protein. Physique-focused protocols - slow titration, 1.6-2.2 g/kg protein, heavy resistance training 3-5x/week, creatine, controlled 15-20% deficit - close most of that gap. Skeletal muscle is protected by the same levers that protect it in any deficit; the drug just makes the deficit easier to hold. Used in the 15-25% body fat range and tapered off before single digits, GLP-1s are one of the cleanest cutting tools available to the looksmaxxing toolkit.
Powered by BTST