Skip to content
foodoseGet my numbers
TrialsSeptember 20, 20264 min read

An MRI substudy finds the fat inside the muscle fell too

In a prespecified MRI substudy of the STEP UP trial, semaglutide reduced adipose tissue by 11.1 litres against placebo and cut fat infiltration inside the thigh muscle by 0.92 percentage points, while the change in thigh muscle volume was not statistically significant.

Most body composition studies count kilograms. This one looked inside the muscle, and found something the scales cannot see.

What happened

STEP UP was a phase 3b trial that randomised adults with a BMI of 30 or more to once-weekly semaglutide at 7.2 mg, semaglutide at 2.4 mg, or placebo, for 72 weeks. This prespecified secondary analysis looked at MRI body composition and physical function (PubMed 42764399).

Against placebo, the semaglutide groups pooled:

MeasureDifference against placeboSignificant
Adipose body tissuedown 11.1 litresyes
Visceral adipose tissuedown 1.5 litresyes
Lean body tissuedown 1.7 litresno
Thigh skeletal muscle volumedown 1.1 litresno
Thigh muscle fat infiltrationdown 0.92 percentage pointsyes

On the sit-to-stand test, participants managed a mean of 13 repetitions in 30 seconds at the start and 15 at week 72, in both the semaglutide and placebo groups.

The finding worth understanding

Muscle fat infiltration is the fat marbled through the muscle itself rather than sitting on top of it. It rises with age and with obesity, and it tracks poorly with how big a muscle looks. Two people with identical thigh volume can have very different amounts of it, and the one with less is stronger.

So a fall of 0.92 percentage points, which was statistically significant while the changes in muscle volume were not, says the muscle got leaner as well as slightly smaller. That is a different story from the one usually told about these medicines, and it is consistent with the function result: nobody got worse at standing up out of a chair, and on average both groups got a little better.

Be careful with this one

The limitation is serious and the paper states it. Of 1,407 participants in STEP UP, only 55 had a valid baseline scan, and of those just 6 were on placebo.

Six people is not a comparison group in any robust sense. It is enough to generate a hypothesis and not enough to settle one, which is presumably why the authors close by saying further research into the effect of weight loss on skeletal muscle composition and function should be explored. The lean tissue and muscle volume results not reaching significance may reflect a genuinely small effect, or simply a substudy far too small to detect one.

Read it as a promising signal, not a conclusion. The direction is encouraging and the sample cannot carry much weight.

What it means for you

Two practical things.

Stop treating muscle as a single number. Size, quality and function are three different things, and this study moved in different directions on each. The one that matters day to day is function: whether you can get up off the floor, carry shopping, climb stairs without holding the rail.

Test it rather than guessing. The sit-to-stand test used here is one you can do at home: a dining chair, thirty seconds, arms folded, count how many times you can stand fully and sit again. Do it now, do it in three months, and you will know something about yourself that no scale will tell you.

What to do with this

Whatever the imaging says, protein and load are still the two things under your control. The floor comes from your weight and your dose.

Find your protein floor

People also ask
What did the STEP UP MRI substudy find about muscle?
STEP UP was a phase 3b trial randomising adults with a BMI of 30 or more to once weekly semaglutide at 7.2 mg, semaglutide at 2.4 mg, or placebo, for 72 weeks. In this prespecified MRI analysis, the semaglutide groups pooled against placebo showed adipose body tissue down 11.1 litres and visceral adipose tissue down 1.5 litres, both statistically significant, while lean body tissue was down 1.7 litres and that difference was not statistically significant.
Did semaglutide change the quality of muscle, not just the amount?
That is the interesting part. Fat infiltration inside the thigh muscle fell by 0.92 percentage points, while the change in thigh muscle volume was not statistically significant. The muscle did not measurably shrink, and it got less marbled.
Why does muscle quality matter more than muscle size?
Because muscle quality, meaning how much fat is marbled through it, is not the same thing as muscle size, and it is closer to what strength actually depends on. Almost nothing measures it, which is why most body composition studies count kilograms and miss this entirely.
Read the originalDiabetes, Obesity and Metabolism (STEP UP secondary analysis)Open

From the recipe collection

Recipes that fit

Each with a stated portion and its nutrition calculated from the ingredients.

Go deeper

The news changes weekly. What protects your body doesn't.

Work yours out once: your calorie target, your protein floor, and the an estimate of how much you are likely to manage at your dose. Free, and no sign-up.

More from the Brief