Guide · Updated November 2026

GLP-1 Body Composition: What the Scale Doesn’t Tell You

GLP-1 body composition data shows the scale hides what matters. In a SURMOUNT-1 DEXA substudy of tirzepatide (Mounjaro, Zepbound), about 75% of the weight lost was fat and about 25% was lean mass. In a STEP 1 DEXA analysis of semaglutide (Ozempic, Wegovy), fat mass fell 19.3% and lean mass fell 9.7%. Scale weight alone cannot tell fat loss from muscle loss.

My scale said I lost 41 pounds in nine months on Mounjaro. My first DEXA scan said something different. Fat loss: 34 pounds. Lean mass loss: 7 pounds. Same 41 pounds on paper, two completely different stories underneath.

This is the part nobody prepares you for. The scale is a liar of omission. It tells you a real number, but it hides the mix. And the mix matters for strength, health, and how you feel after you stop.

This guide covers what the scale misses, what the clinical trials found on tirzepatide and semaglutide, what my DEXA scans showed across three data points, and how to track what matters.

Why the scale hides the real story

A pound is a pound, but not all pounds leave the body equally. When you lose weight on any intervention, diet, surgery, or medication, some percentage comes from fat and some from lean mass. Lean mass includes muscle, bone, organs, and water. Muscle is the part you want to keep.

Muscle drives strength, helps control blood sugar, and affects how you look and move at a given body weight. Two people at 180 pounds can look completely different depending on their fat-to-muscle ratio. The scale cannot see any of this.

On GLP-1 medications, appetite drops so much that many people undereat protein, and many do not lift weights. If you do not measure body composition, you will not know how much lean mass you are losing.

What SURMOUNT-1 actually measured

SURMOUNT-1 (NEJM, 2022, 2,539 adults) is the main tirzepatide obesity trial. Mean weight change at 72 weeks was -20.9% on the 15 mg dose. A DEXA substudy of 160 participants measured body composition. On tirzepatide, body weight fell 21.3%, fat mass fell 33.9%, and lean mass fell 10.9%.

That works out to about 75% of weight lost as fat and 25% as lean mass. One detail most people miss: the placebo group showed the same 75/25 split. So tirzepatide did not make the mix worse. It made the total loss much bigger. The split also held fairly steady across sex, age, and amount of weight lost.

For context, a 2024 review of GLP-1 trials found that lean mass made up anywhere from about 15% to 60% of weight lost, depending on the study. It also notes that lean mass is not the same as muscle, since it includes organs, bone, and water.

What STEP 1 found on semaglutide

STEP 1 (NEJM, 2021, 1,961 adults) is the semaglutide 2.4 mg trial. Mean weight change at 68 weeks was -14.9%. A body composition analysis of 140 participants found fat mass fell 19.3% and lean mass fell 9.7%. Lean mass as a share of total body mass actually rose by 3.0 percentage points, so the body as a whole got leaner.

Can you compare that with tirzepatide? Only loosely. The two substudies used different people and methods, and there is no head-to-head body composition trial. The bigger difference between the drugs is the total amount of weight lost. In the SURMOUNT-5 head-to-head trial, mean weight change at 72 weeks was -20.2% with tirzepatide and -13.7% with semaglutide.

If you are choosing between Ozempic and Wegovy versus Mounjaro and Zepbound, base it on total efficacy, side effects, and cost, not on a claimed muscle advantage. I covered this in more detail in my fat loss vs weight loss guide.

Tirzepatide vs semaglutide: body composition at a glance

The numbers are not directly comparable since the substudies had different protocols. The safest reading is that both drugs cause mostly fat loss, with some lean mass loss, as diet-driven weight loss also does.

My DEXA data: three scans, one stubborn trend

I started tracking body composition three months into Mounjaro, not at baseline. I wish I had done a DEXA before starting. The pattern across my three scans at months 3, 6, and 9:

Between months 3 and 9, I lost 41 pounds total. 32 pounds of that was fat. 9 pounds was lean mass. That is 22% of my loss coming from lean tissue, a bit lower than the roughly 25% in the SURMOUNT-1 substudy. I credit protein and resistance training, but one person’s scans cannot prove that.

The months I stayed on top of both protein (180g+) and two weekly lifting sessions, lean mass loss slowed to near zero. The months I slipped on either one, the scan punished me. Protein intake on a GLP-1 matters if you want the body composition math to work.

I wrote a deeper breakdown in my DEXA scan after 6 months on Mounjaro post, which has the raw scan images.

Why lean mass matters after you stop

Most people regain weight after stopping a GLP-1. In the STEP 1 extension, people who stopped semaglutide regained about two-thirds of their prior weight loss within a year. In STEP 4, people switched to placebo after 20 weeks on semaglutide regained 6.9% of body weight over the next 48 weeks, while those who stayed on it lost another 7.9%.

Appetite coming back is the main driver. Keeping muscle will not prevent regain on its own, but it keeps you stronger and better able to stay active, which helps whatever happens next.

This is why GLP-1 exit strategy planning matters even before you start.

The four things that protect body composition

Not a 10-item listicle. Four things I focus on:

1. Enough protein. Many people on GLP-1s eat less protein than they think. Set a daily target with your prescriber or a dietitian, and spread it across meals. If you have kidney disease, ask before raising protein.
2. Resistance training 2 to 3 times per week. Compound lifts: squat, deadlift, bench, row, overhead press. Cardio helps your heart, but lifting is what tells your body to keep muscle.
3. A titration pace you can tolerate. The Mounjaro label calls for at least 4 weeks on a dose before each increase. Your prescriber decides when to move up. Faster weight loss did not change the fat-to-lean split in the SURMOUNT-1 substudy, but severe side effects can make it hard to eat enough protein.
4. Creatine, if your prescriber agrees. Some people add creatine to support strength. Ask first, especially if you have kidney disease. I covered the data in creatine on Ozempic and Mounjaro.

How to actually measure body composition

Five methods, roughly ranked by accuracy:

DEXA is what I use. Prices vary by city and provider, so call ahead. A sensible rhythm is a baseline scan before you start, then one every few months during the loss phase.

InBody machines at gyms trend reasonably well scan-to-scan on the same machine. They are not accurate in absolute terms, but they are useful for tracking direction. Avoid switching between different units if you use this approach. Our body composition calculator gives a rough estimate from tape measurements if you want a free starting point.

The scale still has a role

I am not saying to throw away the scale. I weigh myself daily. It is fast, free, and captures water and glycogen swings. The rule I follow: scale weight is a leading indicator, DEXA is the scorecard.

The scale still tells you something. It just cannot tell you what the weight was made of.

Bottom line

The scale is a summary statistic. GLP-1 body composition data is the full report. In the SURMOUNT-1 substudy, about 25% of weight lost on tirzepatide was lean mass, the same split as placebo. In the STEP 1 analysis, semaglutide cut fat mass by 19.3% and lean mass by 9.7%. Protect muscle with enough protein and regular resistance training, and track with DEXA every few months rather than relying on the scale alone.

If you care about keeping the loss after stopping the medication, plan for it with your prescriber from the start. Compare GLP-1 providers to find one that supports the pace and follow-up you want.

FAQ

Is muscle loss on GLP-1s permanent?

Not necessarily. Muscle responds to training. With adequate protein and resistance training, lost muscle can be rebuilt. The harder part is rebuilding while appetite returns to baseline. See the GLP-1 maintenance guide for protocols.

How often should I get a DEXA scan on GLP-1s?

Every few months during the loss phase is usually enough to see real change. More frequent scanning adds noise without new information. Prices vary, so check with local providers.

Does tirzepatide preserve muscle better than semaglutide?

There is no head-to-head body composition trial. In separate substudies, about 25% of weight lost on tirzepatide was lean mass, and semaglutide reduced fat mass by 19.3% and lean mass by 9.7%. The methods differ, so the numbers cannot be compared directly.

Can I use my scale at home to track body composition?

Home smart scales with bioimpedance are directionally useful but not accurate in absolute terms. They read hydration and recent meals as much as body fat. Use them for trends on the same scale, not for absolute numbers.

Should I stop my GLP-1 if I am losing too much muscle?

Do not stop on your own. Add protein and resistance training first, then talk to your prescriber about your dose. See our guide on what to do if your doctor will not adjust your dose.

What we checked

This AI-assisted public-source desk check reviewed the SURMOUNT-1, STEP 1, SURMOUNT-5, STEP 1 extension, and STEP 4 abstracts on PubMed, the SURMOUNT-1 and STEP 1 body composition analyses, and a 2024 review of lean mass changes on GLP-1 drugs on November 4, 2026. This is not a clinical review and it is not medical advice. Prices and policies change, so confirm them with the provider or your prescriber before you act.

Related

Guides:

• GLP-1 Muscle Preservation Protocol · Fat Loss vs Weight Loss on GLP-1s · Protein Intake on a GLP-1

Provider Reviews: Ro · Hims · MEDVi

Compare: All Providers · Best GLP-1 Programs · All Guides