Is the weight you're losing fat? What the trials actually measured

The short version: in the trials that measured body composition, weight lost on GLP-1 medicines was mostly fat — but a meaningful share was lean mass. The scale can't tell you which is which for your own body; body-composition measurement can.

What the evidence says

Most big GLP-1 trials measured weight, not composition. The exceptions are informative. In the STEP 1 trial of semaglutide, a sub-group measured with DEXA scans lost both fat mass and lean mass, with the proportion of lean mass to total mass improving overall — because fat fell faster than lean. Published analyses of these substudies commonly report lean mass making up very roughly a quarter to a third of total weight lost, varying widely between people. Tirzepatide's SURMOUNT-1 trial reported a similar pattern in its composition sub-study.

Two honest caveats. First, "lean mass" on a DEXA scan is everything that isn't fat or bone — water, organs and muscle together — so a falling lean-mass number is not automatically shrinking muscle, and short-term shifts in water show up in it. Second, some lean-mass loss happens in *any* substantial weight loss, by any method; it is not unique to these medicines.

Why the split matters

Two people can lose the same 10 kg with different things inside it. Fat loss drives most of the health benefit people are seeking; keeping lean mass is associated with maintaining strength and day-to-day function. That's why researchers measure the split — and why a bathroom scale alone can't tell you how your own weight loss is going beyond the total.

What you can actually do about it

The evidence-backed levers during weight loss are not exotic: adequate protein and resistance training are the two with the most consistent support for retaining lean mass while losing fat, in weight-loss research generally. (Our protein guide covers the protein evidence in detail; a guide to strength training during weight loss follows.) Neither is a guarantee — responses vary — but both are under your control.

How to see your own split

Three tiers of measurement, in rising accuracy and cost: a smart scale's body-fat estimate (convenient, but an estimate with real error margins), repeated under consistent conditions; a DEXA scan (the reference method used in the trials, available privately in NZ's main centres); or simply tracking strength and tape measurements alongside weight. Whatever you use, the trend over months matters more than any single reading, and scan figures and scale estimates shouldn't be blended into one number — they measure differently.

What we still don't know

How lean-mass changes on these medicines translate into long-term strength and function; whether the newer agents differ meaningfully on composition; and what happens to composition after stopping. The full trial publications report composition only in sub-groups, so every number above comes with that limit attached.