About 15% of weight lost in GLP-1 drug trials is skeletal muscle

How much of GLP-1 weight loss is muscle?

About 15% of the weight lost in incretin drug trials is skeletal muscle measured by CT or MRI, according to a 2026 systematic review ^[4]. Roughly 25% of total weight loss shows up as fat-free mass or lean soft tissue on BIA or DEXA scans ^[4].

The alarming percentages trace back to scans, usually DEXA, and DEXA does not measure muscle. DEXA measures lean mass: muscle fiber, body water, glycogen stores, connective tissue and organ mass, combined into one number. Losing water and glycogen during rapid fat loss moves that number without a single muscle fiber shrinking.

Treating "lean mass lost" and "muscle lost" as the same thing is the first place popular coverage overstates the problem.

Do GLP-1 drugs cause more muscle loss than dieting?

No trial has cleanly tested whether GLP-1 drugs cause disproportionately more muscle loss than diet-only weight loss of the same magnitude. The current trial base, with its small subgroups and mostly secondary endpoints, has not answered that question ^[4].

The benchmark figures, about 25% of weight lost as fat-free mass and about 15% as skeletal muscle, are the expected range for weight loss generally. They are not proof that GLP-1 drugs are unusually harsh on muscle.

How strong is the evidence on GLP-1 drugs and muscle loss?

The evidence on GLP-1 drugs and muscle is weaker than most coverage implies: the typical trial was small and short, and most lacked a body-composition primary endpoint ^[4]. The 2026 systematic review covered 35 randomized trials of liraglutide, semaglutide, tirzepatide or dulaglutide, with a median length of 26 weeks and a median of only 78 participants ^[4].

Only about 43% of those trials were judged at low risk of bias, and just under 29% pre-specified body composition as a primary endpoint rather than an afterthought analysis ^[4]. Those limits describe the trials in the review, not every current clinical trial.

Most of the numbers circulating publicly come from small, secondary, exploratory analyses, not from trials built and powered to answer the muscle question. The weak trial base is no reason to dismiss the concern. It is a reason to hold the "40% is muscle" claim loosely and to doubt anyone who treats it as a fixed property of these drugs.

Does exercise preserve muscle while taking a GLP-1 drug?

Yes, in a Danish randomized trial: exercise combined with liraglutide improved body composition more than either the drug or exercise alone ^[5]. The trial was designed around weight-loss maintenance, not as a GLP-1 muscle trial.

Adults completed an 8-week low-calorie diet and were then assigned for one year to exercise alone, liraglutide alone, the two combined, or placebo. All active strategies outperformed placebo on weight loss, and body-fat percentage was a prespecified secondary outcome ^[5].

The Danish trial is the cleanest available demonstration that pairing a GLP-1 drug with a structured training program improves body composition compared with either intervention alone. It did not test tirzepatide or semaglutide. Liraglutide is an older and less potent GLP-1 drug than those commonly prescribed today, so the size of the effect may not transfer directly.

The trial is still randomized human data on a real countermeasure, which separates it sharply from anything sold as a muscle-preserving peptide add-on.

Do peptide add-ons protect muscle during GLP-1 treatment?

No human trial has tested a peptide add-on, such as a myostatin-inhibiting peptide or a growth hormone secretagogue peptide, alongside a GLP-1 drug for muscle preservation. The argument for these products rests entirely on pathway logic: myostatin regulates muscle growth in animal models, so blocking it should preserve muscle in a person losing weight on a GLP-1 drug.

That argument is a hypothesis, not a finding. A biological pathway existing is not the same as a human outcome being measured. Every product marketed this way lacks the kind of evidence that exists for exercise and protein, and the mechanism is being sold with the confidence of a solved problem. The same gap applies to growth hormone secretagogues paired with GLP-1 drugs.

How much protein preserves lean mass during weight loss?

Protein at 2.4 g/kg of body weight beat 1.2 g/kg in a 2016 trial: lean mass rose 1.2 kg over four weeks, versus a 0.1 kg change ^[6]. The higher-protein group also lost more fat ^[6].

The participants combined intense training with a calorie deficit. Protein, training and pace of loss all have controlled trial support, but not in people taking GLP-1 drugs; that evidence is extrapolated from general weight-loss research.

The protein trial was short and its cohort trained intensively, so treat the exact numbers as directional, not as a universal prescription. The direction is consistent: more protein during a deficit, combined with training, favors lean mass. The muscle section of using GLP-1 medications well puts these levers in the context of the whole treatment.

Does resistance training preserve muscle during weight loss?

Yes: exercise prevented an average of 45.7% of the fat-free mass that calorie restriction would otherwise remove, in a 2026 meta-analysis of calorie-restricted trials ^[3]. Mixed training, strength plus endurance, showed the largest effect, with strength training close behind ^[3].

More volume is not automatically better. In a 2023 trial of trained men, 3 sets and 5 sets per exercise preserved lean mass equally when protein was held constant at 2.8 g/kg of fat-free mass ^[7]. In that trial, a consistent training stimulus mattered more than volume for its own sake.

The broader evidence points the same way: combining structured exercise with a dietary intervention outperforms either approach alone across multiple diet types ^[2]. The practical numbers for preserving muscle during GLP-1 weight loss set out targets for protein, training and deficit size.

Does a slower pace of weight loss protect muscle?

Pace matters in training studies: once an energy deficit reaches roughly 500 kcal/day, resistance training's lean mass gains are essentially prevented, though strength gains hold up ^[1]. That finding comes from a 2022 meta-regression of resistance-training-in-a-deficit trials ^[1].

The meta-regression does not set an ideal deficit for someone on a GLP-1 drug. It supports the practical instinct that a slower, less aggressive rate of loss gives muscle-preserving efforts more room to work.

What is still unknown about GLP-1 drugs and muscle?

The gaps matter as much as the findings:

• Drug versus diet. Whether GLP-1 drugs cause disproportionately more muscle loss than diet-only weight loss of the same magnitude is unanswered by a trial base of small subgroups and mostly secondary endpoints ^[4].
• Tirzepatide and semaglutide with training. None of the trials cited below tests resistance training layered onto tirzepatide or semaglutide as a controlled intervention. The best combined-intervention data use liraglutide ^[5].
• Age, sex and starting muscle. None of the studies cited below addresses how age, starting muscle mass or sex changes these figures.
• Peptide add-ons. No peptide marketed for "muscle preservation" during GLP-1 treatment has human outcome data behind it.

Where the evidence stops, the accurate answer is to say so, not to fill the gap with a mechanism story dressed up as a result.