Tirzepatide and lean mass: how to read the DXA results

The question

Tirzepatide's SURMOUNT-1 body-composition substudy found reductions in both fat mass and lean mass. The result supports neither “all the weight lost was fat” nor “a quarter of everyone's muscle disappeared.” Those statements confuse different measurements and different denominators.

The first step is to preserve the study's language. DXA, or dual-energy X-ray absorptiometry, estimates body composition. In this report, lean mass means fat-free mass minus bone mineral content: lean soft tissue. It is not a direct measurement of muscle strength. Original DXA substudy.

What the substudy actually compared

The analysis included 160 participants with baseline and end-of-study DXA data: 124 receiving pooled tirzepatide doses and 36 receiving placebo. At 72 weeks, the reported changes with tirzepatide were −21.3% for body weight, −33.9% for fat mass and −10.9% for lean mass. Corresponding placebo changes were −5.3%, −8.2% and −2.6%. Approximately three-quarters of lost weight was fat and one-quarter lean mass in both groups. Lilly funded the work. Substudy results and funding.

These are results from an imaging subset, not measurements from every participant in the parent study. The SURMOUNT-1 trial record supplies the broader protocol context: a randomized study in adults with obesity or overweight, without type 2 diabetes at entry. The subset's pooled result should not be presented as a separate estimate for each dose or as a prediction for one reader.

Two percentages that should never be swapped

“Percentage of weight lost” and “percentage of starting lean mass lost” use different denominators. A hypothetical example shows why this matters without turning a study average into a personal forecast.

Suppose an imaginary person loses 20 kilograms, of which 5 kilograms is classified as lean tissue. Lean tissue represents 25% of the weight lost: 5 divided by 20. If that person began with 50 kilograms of lean tissue, the lean-tissue change is 10% of the starting amount: 5 divided by 50. The same 5 kilograms produces two different percentages because the questions differ.

These are illustrative numbers, not an individual from SURMOUNT-1. They explain why a claim that “25% of the loss was lean mass” cannot be shortened to “25% of muscle was lost.” The latter also substitutes muscle for the broader lean-tissue measurement.

A second trap is treating an improved proportion as evidence of no absolute loss. If fat falls faster than lean tissue, the share of remaining weight represented by lean tissue can rise even while its measured amount falls. Both statements can be true. A useful article or consultation should show the absolute amount and the relevant percentage rather than letting one stand in for the other.

The scan does not answer the whole muscle question

The authors identify small post-hoc subgroups and the absence of an analysis linking nutrient intake and physical activity to DXA changes as limitations. They also call for direct functional outcomes in future work. Study limitations.

That leaves several questions worth keeping separate. How much lean tissue changed? How did strength change? Could the person perform everyday activities more easily? Was the pattern expected for that individual's starting condition and treatment course? An imaging result contributes to that assessment but does not replace it.

Nor does a group-average pattern establish that an individual change is harmless. “Mostly fat loss” is a description of the distribution of weight loss, not a universal safety certificate. The converse is equally unhelpful: seeing any lean-mass decline is not enough to diagnose a specific muscle disorder from a headline.

For readers already receiving medical care, changes in strength, daily function or nutritional intake are useful subjects to raise with their treating clinician. This article does not supply a protein target, exercise prescription or medication adjustment. The substudy did not test a personalized strategy for every possible reader.

Approval belongs to the medicine and indication

US labeling for Zepbound covers defined weight-management and obstructive-sleep-apnea uses; it does not create an indication for preserving muscle. The prescribing information also includes contraindications and warnings, including severe gastrointestinal reactions and a boxed thyroid C-cell tumor warning based on rodent findings, with uncertain human relevance. Zepbound prescribing information.

This is why the most informative question is not whether tirzepatide is simply “good” or “bad” for muscle. Ask which outcome was measured, in whom, over what period, and against what comparison. Our sleep-apnea trial explainer uses the same approach for breathing outcomes. For a separate question about nutrition evidence, see collagen versus whey in muscle research.