How to Keep Muscle on a GLP-1: Protein, Training & Peptides

Up to 40% of the weight lost on a GLP-1 can be lean mass — but the research shows almost all of it is preventable.

Abstract: DXA substudies of the major GLP-1 trials show 25–40% of weight lost can come from lean mass, yet recent research finds people who train and eat enough protein keep nearly all of it. This guide breaks down the trial data and the research-backed levers — loss rate, protein, resistance training, and GH-axis peptide adjuncts.

How much muscle do you actually lose on a GLP-1?

If you’re trying to figure out how to keep muscle on a GLP-1, start with the real numbers, because they’re both worse and better than the headlines suggest.

In the STEP 1 trial’s DXA substudy, participants on semaglutide lost about 15% of body weight at week 68 — but roughly 40% of the total weight lost was lean soft tissue (Wilding et al., 2021, NEJM). Tirzepatide did better: in the SURMOUNT-1 body-composition substudy, lean mass accounted for about 25% of total loss — roughly 75% of the weight lost was fat (Jastreboff et al., 2022; Look et al., 2025).

Two caveats before you panic:

“Lean mass” on a DXA scan is not all muscle. It includes water, glycogen, organ tissue, and the connective tissue that scales down with any large weight loss — by any method, including diet alone. Some lean-mass reduction during major weight loss is normal and expected.

But none of the trial participants were following a structured muscle-preservation protocol. The trials measured what happens by default — appetite crushed, protein intake collapsing, no resistance training. The interesting research is what happens when you change those defaults.

A 2025 case series tracking people on semaglutide or tirzepatide who resistance-trained 3–5 days per week and ate high protein found lean-soft-tissue changes ranging from −6.9% to +5.8% — some participants gained lean mass — despite total weight losses of 13–33% (Tinsley & Nadolsky, 2025). A randomized controlled trial (LEAN-PREP) testing exactly this combination is now underway.

The conclusion from the data: muscle loss on a GLP-1 is not a drug side effect you have to accept. It’s mostly a behavior gap. Here are the four levers, in order of impact.

Lever 1: Control the rate of loss (≤1 kg / ~2 lb per week)

The fastest way to lose muscle is to lose weight too fast. Physicians working with these compounds make the same observation: in the Huberman Lab episode on peptide therapies, Dr. Craig Koniver describes micro-dosing compounded GLP-1s — starting very low and titrating slowly — specifically so that loss stays at roughly 2 lb (≈1 kg) per week or less, the rate at which his patients hold on to lean mass (clinical observation, not a controlled trial). Rapid 15-lb-in-three-weeks drops are where the “Ozempic face” and rebound stories come from.

In research framing, that means the dose-escalation schedules used in trials are not a target. Slower titration with the lowest effective dose is the most muscle-sparing pattern reported. Our retatrutide vs tirzepatide comparison covers how the major compounds differ in potency and how aggressively each one typically needs to be titrated — relevant because retatrutide, the triple agonist, produces the fastest loss in trials and therefore demands the most deliberate rate control.

Lever 2: Protein — 1.6–2.2 g/kg, front-loaded against a suppressed appetite

GLP-1s work by suppressing appetite. The collateral damage is protein intake: when total food intake drops 30–40%, protein drops with it, and muscle protein synthesis loses its raw material.

The intake range used across muscle-preservation research is 1.2–2.0 g of protein per kg of body weight per day, with the higher end (1.6–2.2 g/kg of target body weight) favored for people training hard in a large deficit. Practical structure matters more than usual because appetite is the constraint:

• 25–40 g protein per meal, protein first on the plate — satiety arrives fast on a GLP-1, so whatever you eat first is what you actually absorb.
• Liquid protein (whey, clear isolates) counts and is often the only realistic way to hit targets in the first months.
• Track it. People on GLP-1s reliably overestimate their protein intake because they remember pre-medication portions.

Lever 3: Resistance training 3–5 days per week (non-negotiable)

This is the lever with the strongest evidence. In the 2025 case series above, training 3–5 days weekly was the common factor among people who kept — or gained — lean mass through 13–33% weight loss. Mechanistically there’s no mystery: muscle that is loaded is muscle the body keeps; in a deficit, unloaded muscle is metabolically expensive tissue the body readily catabolizes.

The research-supported minimum is 3 full-body sessions per week with progressive overload, compound movements first. Cardio is fine and useful for the fat-loss side, but it does not substitute — the signal that preserves muscle is mechanical tension, not energy expenditure. If you’re new to programming this, the principles in our muscle-growth peptide guide apply directly — the training base is identical whether the goal is growth or preservation.

Lever 4: GH-axis peptide adjuncts (the research-grade layer)

This is where peptide research intersects with the GLP-1 muscle question. The growth-hormone axis is anabolic and lipolytic: more GH/IGF-1 signaling means more muscle protein synthesis support and more fat oxidation. Two compounds dominate the research conversation.

Tesamorelin is a GHRH analog and the only one FDA-approved for body composition (visceral fat reduction in HIV-associated lipodystrophy). In controlled trials it reduced visceral adipose tissue while preserving lean mass (Falutz et al., 2007; Stanley et al., 2014). Research protocols use ~1–2 mg SubQ pre-sleep, nothing eaten for several hours prior, often 5 days on / 2 off.

Ipamorelin is the most selective GH secretagogue — it triggers a GH pulse without the cortisol, prolactin, or strong appetite effects of older GHRPs (Raun et al., 1998). Reported protocols cap at ~100 mcg pre-sleep; physician anecdote describes it as the “cleanest” secretagogue, and it is typically paired with a GHRH analog because GHRP monotherapy desensitizes. The combination is common enough that it’s sold as a single blend (ipamorelin + tesamorelin). For a head-to-head on which fits which goal, see tesamorelin vs ipamorelin for fat loss.

Koniver reports compounding a GLP-1 together with a GHRH analog (tirzepatide + sermorelin) explicitly to offset muscle loss — again, clinical observation rather than trial evidence. The honest evidence grade: tesamorelin’s lean-mass preservation is clinical-trial-backed in its approved population; using GH-axis peptides alongside a GLP-1 for muscle preservation is mechanistically coherent and physician-reported, but has no dedicated controlled trial yet.

The key caution: GH-axis compounds reduce insulin sensitivity. Tesamorelin’s label data show small fasting-glucose and A1C increases, worst in people who don’t lose visceral fat, and the effect compounds when stacked with ipamorelin. On a GLP-1 (which improves glycemic control) this is partially offset, but it is exactly the kind of interaction that demands baseline and follow-up labs — fasting glucose, fasting insulin, A1C — as covered in our peptide side-effects and safety guide.

How to keep muscle on a GLP-1: the full protocol

What the combined literature and physician reports describe, framed for research purposes:

Layer	What’s reported in research	Why
GLP-1 dose	Lowest effective dose, slow titration, loss ≤1 kg/week	Rate of loss is the strongest muscle-loss predictor
Protein	1.2–2.0 g/kg/day, 25–40 g per meal, protein first	Appetite suppression collapses protein intake by default
Training	Resistance training 3–5 d/wk, progressive overload	The only lever shown to flip lean-mass change positive
GH-axis adjunct	Tesamorelin ~1–2 mg ± ipamorelin ~100 mcg, pre-sleep, 5 on / 2 off	Lipolytic + lean-mass-sparing signal; trial-backed for tesamorelin alone
Monitoring	Weight-loss rate, DXA or smart-scale lean mass trend, fasting glucose/insulin/A1C	Catches both failure modes: losing too fast, and GH-driven insulin resistance

This protocol is the muscle-preservation core of the broader fat-loss peptide stack (which adds MOTS-c and the behavior base for fat oxidation). If you’re choosing between GLP-1 compounds first, note that tirzepatide’s DXA data (≈25% lean loss untrained) was meaningfully better than semaglutide’s (≈40%), and tirzepatide is the compound with the most published body-composition data of the three.

FAQ

Does everyone lose muscle on Ozempic or tirzepatide? No. Untrained trial participants lost 25–40% of total weight as lean mass, but 2025 data on people combining resistance training with high protein showed lean-mass changes from −6.9% to +5.8% — some gained muscle while losing 13–33% of body weight.

How much protein should I eat on a GLP-1? Research protocols use 1.2–2.0 g per kg of body weight daily, split into 25–40 g servings eaten protein-first. Liquid protein is a practical necessity for many people in the first months of appetite suppression.

Is muscle loss worse on stronger GLP-1s like retatrutide? The risk scales with the rate of weight loss, and stronger agonists produce faster loss at equivalent titration. That makes rate control (micro-dosing, slow escalation, ≤1 kg/week) more important, not less, with retatrutide-class compounds.

Can peptides replace resistance training for keeping muscle? No. No GH-axis peptide has shown muscle preservation in a deficit comparable to resistance training. Tesamorelin preserved lean mass in trials, but the anabolic signal from loading muscle is the foundation; peptides are an adjunct layer in research protocols.

Do GH peptides interfere with GLP-1s? Not directly, but they pull glycemic control in opposite directions: GLP-1s improve insulin sensitivity while GH-axis compounds reduce it. Research protocols monitor fasting glucose, insulin, and A1C at baseline and on-cycle.

References

1. Wilding, J.P.H. et al. (2021). Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1). New England Journal of Medicine, 384(11), 989–1002. doi:10.1056/NEJMoa2032183
2. Jastreboff, A.M. et al. (2022). Tirzepatide Once Weekly for the Treatment of Obesity (SURMOUNT-1). New England Journal of Medicine, 387(3), 205–216. doi:10.1056/NEJMoa2206038
3. Look, M. et al. (2025). Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study. Diabetes, Obesity and Metabolism. doi:10.1111/dom.16275
4. Tinsley, G.M. & Nadolsky, S. (2025). Preservation of lean soft tissue during weight loss induced by GLP-1 and GLP-1/GIP receptor agonists: a case series. SAGE Open Medical Case Reports. doi:10.1177/2050313X251388724
5. Stanley, T.L. et al. (2014). Effects of Tesamorelin on Visceral Fat and Liver Fat in HIV. JAMA, 312(4), 380–389. doi:10.1001/jama.2014.8334
6. Falutz, J. et al. (2007). Metabolic Effects of a Growth Hormone–Releasing Factor in Patients with HIV (Tesamorelin). New England Journal of Medicine, 357(23), 2359–2370. doi:10.1056/NEJMoa072375
7. Raun, K. et al. (1998). Ipamorelin, the First Selective Growth Hormone Secretagogue. European Journal of Endocrinology, 139(5), 552–561. PMID: 9849822
8. Huberman Lab Podcast (2024). Dr. Craig Koniver: Peptide & Hormone Therapies for Health, Performance & Longevity. (Physician clinical observation; not peer-reviewed evidence.)

---

All compounds discussed are for research and educational purposes only. This article is not medical advice, and nothing here is a recommendation to treat, cure, or prevent any condition. Protocols described are summaries of what has been used or reported in published research and physician commentary. Consult a licensed physician before any decisions about your health.