THE EVIDENCE REVIEW

Retatrutide and lean mass: the body-composition evidence

Retatrutide body-composition research indicates that fat mass decreases substantially, while some lean-mass reduction may also occur. The proportion should not be exaggerated into 'muscle destruction,' because DXA lean mass also includes water and non-muscle lean tissue. This page assembles the verified record behind that sentence.

The anchor: the retatrutide DXA substudy

Coskun and colleagues, Lancet Diabetes & Endocrinology 2025 ↗ is the only retatrutide DXA body-composition dataset published to date. In people with type 2 diabetes, DXA-measured total fat mass fell by up to 26.1% at week 36 (8 mg pooled), with the proportion of lean-mass loss reported by the authors as similar to other obesity treatments. What it supports: retatrutide weight loss is predominantly fat loss in this population and timeframe. What it does not prove: muscle-function or bone outcomes, composition in people without diabetes, or completeness — 103 of 189 substudy participants completed both DXA scans. The parent trial (Rosenstock 2023, Lancet ↗) reported up to −16.9% bodyweight at 36 weeks with glycaemic endpoints only.

The phase 2 obesity trial (Jastreboff 2023, NEJM ↗) established dose-dependent weight loss up to −24.2% at 48 weeks — total body weight only, with no fat-versus-lean breakdown. The first phase 3 result, TRANSCEND-T2D-1 ↗, reported −11.5% to −15.3% bodyweight at 40 weeks in type 2 diabetes — again without body-composition endpoints. The composition question is open by design of the programme, not by absence of interest.

The benchmarks: tirzepatide and semaglutide

The SURMOUNT-1 DXA substudy ↗ (n=160 of 2539) measured −33.9% fat mass and −10.9% lean mass at week 72 on tirzepatide — roughly 75% of weight lost as fat and ~25% as lean mass in both active and placebo arms. What it supports: a consistent fat-to-lean loss ratio under incretin weight reduction. What it does not prove: a direct skeletal-muscle measurement (DXA lean mass includes water and non-muscle tissue) or any muscle-function outcome.

Earlier, the SUSTAIN 8 DXA substudy ↗ showed semaglutide 1.0 mg reducing fat mass by 3.4 kg and lean mass by 2.3 kg over 52 weeks — with no placebo arm, a caveat its own authors attach. STEP 1 ↗ and SURMOUNT-1 ↗ supply the class weight-loss frame (−14.9% and −15.0% to −20.9% respectively) without DXA analysis in the primary papers.

Aggregating across the class, the Batsis systematic review of 35 RCTs ↗ found a median ~28.3% of weight lost on incretin therapies attributable to muscle-based indices, with two-thirds of studies exceeding the ~25% benchmark — and no included study reporting objective physical-function outcomes.

Established versus open

What is established

Retatrutide produces substantial, predominantly fat-mass weight reduction (phase 2 DXA substudy, type 2 diabetes).
Across incretin therapies, roughly a quarter of weight lost comes from lean or muscle-based indices.
DXA lean mass includes water and non-muscle lean tissue; it is not a direct skeletal-muscle measurement.
No retatrutide study to date reports objective muscle-function or bone endpoints.

What remains open

The tissue composition of lean-mass change under retatrutide (water versus contractile tissue).
Functional consequences — strength, performance, falls — of composition change during rapid weight loss.
Whether nutritional architecture (protein, resistance exercise, micronutrient status) alters the fat-to-lean loss ratio.
Whether a weekly thiamine component can meaningfully support thiamine status during appetite-suppressed weight loss.