GLP-1 and body composition: it's not the scale, it's where the fat goes

You're three months in. The scale has more or less parked itself, and you keep stepping off it a little deflated. Then one morning you reach for your usual jeans and they slide on without the familiar fight at the waistband. Same number on the floor. Different body in the mirror.

That gap isn't a measurement error. It's what GLP-1 medications are quietly doing to the body underneath the number you've been trained to watch.

Here's the line that reframes everything. In the body-composition sub-study of STEP 1, semaglutide cut visceral fat by 27.4% — noticeably more than the 19.3% drop in total fat mass. The fat that matters most metabolically came off in the bigger proportion. Your waistband felt that before your scale ever did.

The scale barely moved but the jeans got loose

Plenty of people hit a wall around month three and read it as failure. The math under the hood says otherwise.

Body weight is a single number stacking five different things on top of each other: fat, muscle, bone, water, and whatever's currently in your gut. A GLP-1 can be stripping visceral fat off your liver and your abdominal organs while a Tuesday of extra sodium and a sluggish gut hold a couple of pounds of water steady. The scale averages all of that into one flat reading. It can't tell you that the pound it isn't showing is fat leaving and water arriving on the same day.

So the loose-jeans, flat-scale week isn't a contradiction. It's the most honest signal your body sends you — fat reorganizing faster than a bathroom scale can keep up. The number you'd most want to see is visceral fat, and that's the one number a $30 scale will never report.

Where GLP-1 pulls fat from first: visceral versus subcutaneous

Not all body fat is the same fat, and your body doesn't burn it off in the order you'd guess.

Subcutaneous fat is the kind you can pinch — it sits just under the skin on your thighs, hips, and the backs of your arms. Visceral fat is the deeper, hidden kind. It packs in around your liver, pancreas, and intestines, and you can carry a dangerous amount of it without looking like you do. Two people at the same weight, with the same waist, can hold wildly different amounts of it.

The STEP 1 imaging data is where this gets concrete. Semaglutide dropped total fat mass by 19.3%, which is already a substantial reduction. But visceral fat fell by 27.4% — a meaningfully steeper cut. The medication didn't trim fat evenly across the body. It went after the deep abdominal fat preferentially, which is exactly the fat you'd choose to lose first if you got to pick.

Why does it shake out this way? Visceral fat is more metabolically active than the subcutaneous kind. It mobilizes faster when the body shifts into a calorie deficit, and it's more responsive to the hormonal and insulin changes a GLP-1 sets off. The deep fat goes first because, biologically, it's the easiest to release.

Two people can weigh the same and wear the same jeans, yet one carries far more visceral fat than the other. That hidden difference does more to shape long-term metabolic risk than the number on either of their scales. It's also the number that moves first on a GLP-1 — and the one the scale is structurally incapable of showing you.

This is the quiet good-news headline of the whole drug class. Of all the fat you could lose first, the deep abdominal fat is the prize. And that's exactly where these medications reach first.

Why visceral fat is the kind worth losing: liver, heart, insulin

Calling visceral fat "bad fat" is a shortcut, but it earns the label.

Subcutaneous fat is mostly storage — inert padding that minds its own business. Visceral fat behaves more like a malfunctioning organ. It pumps inflammatory signals into your bloodstream, drives insulin resistance, and sits anatomically close enough to your liver to flood it with free fatty acids. That's the pathway linking a thick midsection to type 2 diabetes, high blood pressure, and heart disease — not the soft fat on your thighs, but the deep stuff crowding your organs.

The liver piece is where a second drug sharpens the picture. In SURPASS-3, a trial run in adults with type 2 diabetes, an MRI sub-study tracked liver fat directly. Tirzepatide produced an absolute reduction in liver fat content of 8.09 percentage points — a substantial drop in the fat marbled into the liver itself. One honest caveat: that was a diabetes trial, not an obesity one, so it isn't a one-to-one read for someone taking these drugs purely for weight. But the direction tells you something real. These medications don't only shrink the fat you can see. They reach the fat wedged inside the organs keeping you alive.

That's the whole case for caring about composition over weight. Losing 10 pounds of subcutaneous fat is fine. Losing visceral and liver fat is the change that moves your blood pressure, your fasting glucose, and your long-term cardiovascular odds. The two can read identically on the scale. Inside your body, they are not the same event.

The numbers behind the body composition shift

Here's the STEP 1 body-composition data, side by side, so the pattern is impossible to miss.

Sit with the bottom two rows for a second, because they hold a subtlety that's easy to mangle.

Lean body mass did drop — by 9.7%. In absolute terms, you lose some muscle. There's no spinning that away. But because fat left in much larger quantities, lean tissue made up a bigger slice of the smaller body left behind. Lean mass as a proportion of total body mass rose by 3.0 percentage points. So the honest two-part summary goes like this: your muscle goes down in absolute pounds while going up as a percentage of you. Both are true at once, and only by holding both do you avoid either panicking or coasting.

How much muscle you actually lose, and how to hold onto it

The muscle question deserves a straight answer instead of reassurance.

You do lose some. STEP 1 put the lean-mass drop at 9.7% — not trivial, not catastrophic. The reason it isn't a five-alarm fire is the ratio: fat left far faster than muscle, which is why your lean-mass percentage climbed even as the raw pounds slipped. The risk isn't that GLP-1s melt your muscle. It's that any rapid weight loss costs some lean tissue, and if you do nothing to defend it, you hand back more than you had to.

The defense is unglamorous and well established. Three levers carry almost all of it.

• Protein. Roughly 1.2 to 1.6 grams per kilogram of body weight per day, spread across your meals rather than crammed into one. Appetite craters on these drugs, so protein is the macro that tends to collapse first — exactly the one you can least afford to drop.
• Resistance training. Two or three sessions a week, hitting the major muscle groups. Lifting is the single clearest signal you can send your body to keep the muscle it still has. Cardio is good for plenty of things, but it doesn't preserve lean mass the way loading your muscles does.
• Pace. Slower, steadier loss protects muscle better than a crash. If the weight is falling faster than a couple of pounds a week for months on end, that's a conversation worth having with your prescriber.

None of this is exotic. A 6-ounce chicken breast runs about 38 grams of protein. A cup of Greek yogurt adds another 20. Two or three honest gym sessions a week, where you genuinely load the muscles, do most of the protective work. The drug handles your appetite. Muscle is the part you own.

Why the scale is a poor instrument for this job

The bathroom scale isn't lying to you. It's answering a question you didn't mean to ask.

It reports one number — total mass — and flatly refuses to break it into parts. On a GLP-1, several things move at once: visceral fat drops fast, subcutaneous fat drops slower, you might be holding a pound or two of water, and your gut content swings day to day. The scale folds all of that into a single figure and hands it back with zero context. A flat week can hide a great composition week underneath it.

There's a psychology cost too. Daily weighing trains you to judge a slow, structural process by a noisy daily signal. Water alone can swing two or three pounds overnight on sodium and hormones — enough to erase a real week of fat loss on the display and leave you discouraged about progress that's genuinely happening. People quit good protocols over scale noise all the time. The number didn't fail them. The instrument was answering the wrong question.

Better ways to track your progress

If the scale is the wrong instrument, what's the right one? You don't need a lab. You need to measure the thing that's actually changing.

The single most useful upgrade is a $5 tape measure. Wrap it around your waist at the navel, same time of day, once a week, and write it down. Waist circumference tracks visceral fat far better than weight does, and it'll often keep dropping during the exact stretches when the scale has stalled. That isn't a coincidence — it's the deep fat leaving while total mass holds steady. If you can get a DEXA scan every few months, that's the clearest window into your real composition. But a tape measure and a phone camera will catch most of what matters, for almost nothing.

What to realistically expect: fat loss versus weight loss

The averages are genuinely strong, and they're also exactly that — averages.

Over two years, a semaglutide study showed a mean body-weight reduction of 15.2%. Tirzepatide, in its obesity program, has gone as high as 22.5%. Those are large, durable numbers. But they sit in the middle of a wide spread. Some people land well above them; a meaningful share land below, and some see relatively little weight change at all. Your mileage varies — your starting point, your dose, your diet, and your activity all push the result around.

Here's the reframe worth carrying, though. Even when the weight-loss number lands lower than you hoped, the composition number — the visceral and liver fat coming off — is often quietly doing its job. The scale measures pounds. Your blood pressure, your fasting glucose, and your waistband are measuring something closer to health. The two don't always move in lockstep, and when they diverge, it's usually the scale that's the lagging, less meaningful indicator.

For brand context in the US: semaglutide for obesity is Wegovy, tirzepatide for obesity is Zepbound, and the same tirzepatide molecule for type 2 diabetes is Mounjaro. Liraglutide for obesity is Saxenda. The same molecules carry different brand names depending on the country and the indication — worth knowing if you're comparing notes with someone abroad.

Questions worth bringing to your next appointment

The visit is short, and the weight number tends to eat the whole conversation. A few questions pull it back toward composition, where the real action is.

Can we track my waist circumference or order a DEXA scan, not just my weight? This reframes the entire visit around composition instead of a single number, and most clinicians will happily order it once you ask.

My weight has stalled but my waist is still shrinking — is that normal? Yes, and naming it out loud usually gets you a clearer read on whether you're losing fat while holding water — a far better question than "why won't the scale move."

What should I be doing to protect my muscle at this dose? This opens the door to a concrete protein target and a resistance-training plan, instead of a vague nudge to "stay active."

Are my metabolic markers improving — blood pressure, fasting glucose, lipids? These often improve alongside visceral-fat loss even when the scale is quiet, and they're a more meaningful scorecard than weight alone.

Walk in with one line written at the top of your notes: "I want to track body composition, not just weight." That single sentence reroutes a 15-minute appointment away from a number that's misleading you and toward the changes that actually move your health.

A smarter checklist for body composition

If you take one thing from all of this, let it be a shift in what you measure and what you let yourself feel good about.

• Retire the daily weigh-in. Once a week at most, same time, same conditions. Daily numbers are mostly water noise dressed up as feedback.
• Measure your waist instead. A tape measure around the navel, weekly, written down. It tracks visceral fat — the fat that fell 27.4% in STEP 1 — far better than your scale ever will.
• Hit your protein. Around 1.2 to 1.6 grams per kilogram a day, spread across meals. It's the lever that holds your muscle while the fat leaves.
• Lift two or three times a week. Resistance training is how you keep the lean mass that's already becoming a bigger share of you.
• Watch the markers that matter. Blood pressure, fasting glucose, how your clothes fit, your energy. These speak to the visceral and liver fat in ways the scale structurally cannot.
• Reframe a flat scale. A stalled number with a shrinking waist isn't a plateau. It's fat reorganizing faster than your scale can report — usually a sign things are working, not stalling.

The number on the floor was never the point. What changed is where your body carries its fat, how much of you is now lean tissue, and what your liver and your arteries are no longer being asked to haul around. The scale can't see any of that. You can — as long as you measure the right things.

The figures here come from published clinical trials and peer-reviewed research, and individual response varies widely. Whether to start, change, or stay on a GLP-1 is a call to make with your doctor, who can read these numbers against your own history.

Start managing your GLP-1 with Blueshot

AI coaching, injection scheduling, and weight tracking in one app