Sparing the muscle is not yet saving the strength

A myostatin antibody kept more lean mass on the scan during tirzepatide weight loss — but a DXA reading is not a measure of how a patient moves.

Armando Cuesta, MD Tuesday, 9 June 2026 · 2 min read

Analysis. Drafted by The Vital Record’s AI under Armando Cuesta, MD’s direction and reviewed by him before publication. How we label →

The worry behind this trial is real. Drop 15 or 20 percent of body weight on a GLP-1 drug and some of what leaves is muscle, not just fat — and in older or frailer patients that loss is not trivial. So the logic of EMBRAZE is sound: add apitegromab, an antibody that blocks myostatin signaling, and see whether you can shed fat while holding on to lean tissue.

On its own terms, the trial did what it set out to do. EMBRAZE (NCT06445075) was a phase 2, randomized, double-blind, placebo-controlled study of 102 adults with overweight or obesity, all of whom received tirzepatide; participants were randomized 1:1 to add apitegromab or placebo for 24 weeks. The pre-specified primary endpoint was change from baseline in total lean body mass at 24 weeks, measured by dual-energy X-ray absorptiometry — a scan. Per the published report, apitegromab spared roughly 1.9 kg of lean mass versus placebo while total weight loss, a secondary endpoint, held steady. That is a clean readout against the question asked.

But read the endpoint literally. The trial measured lean mass on imaging, not muscle strength, walking speed, grip, or any functional task. More kilograms of lean tissue on a DXA scan is a plausible proxy for preserved function — it is not the same thing.

Preserving the number on the scan is the hypothesis. Preserving what the patient can do with that muscle is the endpoint that still has not been tested.

Three other limits are worth stating plainly. This was a small, 24-week proof-of-concept in adults aged 18 to 65 without diabetes — not a long-term or hard-outcomes study, and not the frail elderly in whom muscle loss matters most. It is a single phase 2 trial, not replicated. And we do not yet know whether sparing lean mass translates into anything a patient or clinician would feel.

The signal is genuine and the mechanism is rational. For now it belongs in the column marked “promising and unproven” — a reason to run the strength-and-function trials, not a reason to prescribe.

AI-drafted under the direction of Armando Cuesta, MD. Not medical advice.