Does retatrutide cause muscle loss? What the DXA scans show, and the big gap in the data

The short answer: retatrutide lowers lean mass, as every large weight loss does. Whether it costs more muscle than other drugs is unknown, because the only published scan data is one small phase 2 study in people with type 2 diabetes.

That study's authors said the share of weight lost as lean mass was "similar to other obesity treatments." But those people lost less weight, over less time, than people in the phase 3 trials, and I can't find body composition results from any TRIUMPH trial.

Status as of November 6, 2026: Retatrutide is investigational and not approved anywhere. Lilly plans to file for U.S. approval in Q1 2027. Tirzepatide and semaglutide are FDA-approved. Nothing here is medical advice or a dosing guide.

Key takeaways

Want the short version of what's actually approved, legal and tested? Get the free BHN cheat sheet.

Fat mass fell sharply at the higher doses. Lean mass fell too, in a proportion the authors called in line with other treatments. The paper is a substudy of a phase 2 trial using DXA, the X-ray scan that splits the body into fat, lean tissue and bone (Coskun et al., 2025). Fat mass change at 36 weeks:

Doses are the trial arms, shown only to identify each group. The limits:

Probably, but only in the full paper, which is paywalled, and I couldn't read it. The public abstract gives no lean mass figure. So I'm not printing one, including the percentages that circulate online.

The authors wrote that the data "could provide reassurance that a greater proportion of lean mass is not lost with retatrutide." But "other obesity treatments" spans about 25% to 40% in the two trials below, so "similar" doesn't say where reta landed.

A 2026 review citing the full paper reads it less warmly. It places the higher-dose groups "above the conventional one-quarter estimate," with lean share rising with dose, and calls the dose pattern hypothesis-generating (Diabetes & Metabolism Journal, 2026). That's a reviewer's reading, not a new measurement.

Nobody can rank them yet. No trial has compared them on lean mass head to head, and these studies differ in almost every way that matters.

Sources: Coskun et al., 2025, Look et al., 2025, Wilding et al., 2021. The STEP 1 share comes from two later papers that read it slightly differently: 38% in the joint advisory and 40% in Look et al. These are cross-trial numbers. Read them as context, not a ranking.

The tirzepatide split looks like weight loss, not a drug effect. About 75% fat and 25% lean showed up on both tirzepatide and placebo. Lean mass fell 5.6 kg on the drug and 1.2 kg on placebo: far more, in about the same proportion (Look et al., 2025).

A 25% vs 38% gap doesn't prove tirzepatide spares more, either. These are small substudies with different people at different sites. The weight numbers behind each drug are in reta vs tirz.

No. DXA lean mass is everything that isn't fat or bone mineral, and muscle is only part of that.

The joint advisory's rough rule: "muscle mass is about half of lean body mass." By that math, it says about 20% of STEP 1's weight reduction was muscle (joint advisory, 2025). That's an estimate, not a measurement.

So a lean mass drop overstates muscle loss. It doesn't clear anyone, though. The advisory notes these tools track "muscle mass but not muscle health, quality, or function." And in calorie restriction studies of nonexercising adults, muscle really did fall: about 2 to 2.5 kg per 10 kg lost in men, 1 to 1.5 kg in women (Heymsfield et al., 2024). That was diet, not a drug.

Not that I can find. The TRIUMPH-1 abstract covers weight, knee pain and sleep apnea, with no lean mass (NEJM, 2026). Lilly's June and September phase 3 releases don't mention it either (Lilly, Sept 29, 2026).

On November 6 I checked all 32 Lilly retatrutide records on ClinicalTrials.gov. None names lean mass, fat mass, DXA or muscle as a primary or secondary outcome. That doesn't rule out a substudy: the phase 2 record doesn't list body composition either, and its substudy got published.

This is the real gap. TRIUMPH-1's top dose averaged -25.0% at 80 weeks, and nobody has published what that weight was made of. A 2026 review says the net effect on skeletal muscle "remains an open question," partly because of glucagon, the receptor reta adds (Int J Mol Sci, 2026).

Want the reta scan data the day it's published? Get BHN's updates. If Lilly reports TRIUMPH body composition, I'll update this post.

Strength training first, with enough protein behind it. None of this was tested on retatrutide. It comes from the 2025 joint advisory by the American College of Lifestyle Medicine, American Society for Nutrition, Obesity Medicine Association and The Obesity Society, which covers GLP-1 drugs and never mentions reta (joint advisory, 2025).

Strength training. The advisory calls structured resistance or mixed training "well established" for preserving lean mass during weight reduction, with a smaller effect from aerobic work alone. It aims for strength training at least 3 times a week plus at least 150 minutes of moderate aerobic exercise. But the GLP-1-specific support it cites for fat-free mass is retrospective, and the randomized trial it cites showed bone, abdominal fat and inflammation results, not a muscle result.

Protein. The advisory sets no single number:

It's also unclear which body weight to use, and actual weight can overestimate need. Since appetite drops, it suggests eating protein first and leaning on lower-volume foods like fish, eggs and Greek yogurt. The numbers are laid out in how much protein on reta.

The catch. The advisory says more protein alone is "likely inadequate" without structured strength training.

The advisory doesn't mention creatine. The basics are in creatine is boring because it works. What's known about taking it with reta is in creatine and reta.

Possibly, by a small amount, in a narrow setting. A 2024 meta-analysis of 8 studies found cold water immersion right after resistance training "may modestly attenuate" muscle growth compared with training alone (Piñero et al., 2024). The limits:

So nothing shows that plunging changes muscle loss on reta.

Retatrutide also isn't approved, the legal routes to it are narrow, and the wider side effect picture is in is reta safe?

If you've had DXA scans before and during a GLP-1 drug, reply and tell me what changed: fat, lean, and whether you were lifting. When TRIUMPH scan data lands, I'll put the trial numbers next to what readers actually saw.

Both, mostly fat. In the phase 2 DXA study, fat mass fell 26.1% at 8 mg and 23.2% at 12 mg over 36 weeks. Lean mass fell too, and the authors said the proportion was similar to other obesity treatments.

No verified figure is public in the study's abstract, and DXA doesn't measure muscle directly anyway. For scale, about 25% of weight lost on tirzepatide was lean mass, and about 38% to 40% on semaglutide, in separate trials. Muscle is roughly half of lean mass.

Unknown. No trial has compared them on lean mass, and their studies differ in population, length and dose.

No trial has tested that. The joint advisory for GLP-1 drugs mentions proposed targets of 1.2 to 1.6 g/kg a day during weight reduction, or a flat 80 to 120 g a day. It says protein alone likely isn't enough without strength training.

No trial or guideline says so for reta. The joint advisory says DXA could be considered yearly or every 2 years for people on GLP-1s, and it lists cheaper options like bioimpedance.

Want the short version of what's actually approved, legal and tested? Get the free BHN cheat sheet.