Survodutide vs Tirzepatide for Visceral Fat: What Has Actually Been Measured

That is the honest state of this comparison in late 2026. Tirzepatide has visceral fat data from imaging substudies in two large trials. Survodutide has body-weight percentages, a liver-biopsy trial, and a mechanism that makes a strong case on paper. Those are not the same kind of evidence, and treating them as comparable is how a reasonable hypothesis turns into a confident recommendation it has not earned.

Why visceral fat is the right question to be asking

Total weight loss is the number every trial reports and the number every headline quotes, and it is the less useful one. Visceral adipose tissue — the fat packed around the organs rather than under the skin — is what carries most of the metabolic risk. Two men losing the same 20 kg can end up in very different places depending on where that mass came from.

Visceral fat is metabolically active in a way subcutaneous fat is not. It drains directly into the portal vein, which means the free fatty acids and inflammatory signals it releases arrive at the liver first and at concentration. That is the proposed route from abdominal adiposity to hepatic insulin resistance, dyslipidaemia and the cardiovascular risk that tracks with waist circumference better than it tracks with BMI. Subcutaneous fat on the thighs and hips, by contrast, is close to metabolically inert and in some analyses looks mildly protective.

So "which drug is better for visceral fat" is a better question than "which drug takes off more weight." It is also a harder one to answer, because measuring it requires imaging — DXA or MRI — and imaging substudies are expensive add-ons that most trials skip. A phase 3 obesity programme can run to several thousand participants; scanning even a few hundred of them at two timepoints is a budget line somebody has to defend, and the regulator does not require it. Approval turns on weight and safety, not on body composition.

That is the whole reason this comparison is lopsided. One drug's programme paid for the scans and the other's has not yet. The asymmetry is about trial funding decisions, not about the drugs.

What tirzepatide's visceral fat data actually shows

Tirzepatide has been imaged twice, in two different populations, and the results are strong.

The SURMOUNT-1 DXA substudy enrolled 255 participants with obesity or overweight, of whom 160 completed both baseline and week-72 assessments [1][7]. The headline figures:

The estimated treatment difference on visceral fat was −32.8% [1]. Two things stand out. Visceral fat fell faster than total fat mass — 40.1% against 33.9% — which means the drug is not simply shrinking everything proportionally; it is taking more from the compartment that matters most. And lean mass fell 10.9%, which is roughly a quarter of the total loss and is the figure that tends to get left out of the marketing.

One caveat belongs on this table: the visceral fat results were post hoc [1]. They were not a prespecified endpoint, which lowers their evidential weight even though the effect size is large and the direction is unambiguous.

The second dataset is SURPASS-3 MRI, a substudy of 296 people with type 2 diabetes scanned at baseline and 52 weeks [2][3]. Mean baseline liver fat content was 15.7%. Tirzepatide reduced liver fat by 47.1% at the 15 mg dose against 11.2% for titrated insulin degludec — though that specific pair of figures comes from Lilly's own press release rather than the peer-reviewed paper, and should be read accordingly [2]. On visceral fat the directional result is the more interesting one: VAT fell on tirzepatide and rose on insulin degludec.

That contrast is worth sitting with. Both arms were treating the same disease and both improved glycaemic control, but one did it while adding visceral fat and the other while removing it. It is a reminder that "controls blood sugar" and "improves body composition" are separate questions, and that a drug can do the first while making the second worse.

It also matters that SURPASS-3 was open-label [3]. Neither participants nor investigators were blinded, which is a real limitation for subjective endpoints — but much less of one for an MRI volume measurement, where the scanner does not know which arm it is scanning. Imaging endpoints are among the few places where open-label designs carry close to their full weight, which makes this substudy stronger evidence than its design label suggests at a glance.

The population is the bigger caveat. SURPASS-3 enrolled people with type 2 diabetes who were insulin-naïve, and the MRI substudy skewed toward those at risk of hepatic steatosis. Whether a 52-week VAT reduction in that group transfers cleanly to a non-diabetic man in his forties taking tirzepatide for weight is an assumption, not a finding — a reasonable assumption, given SURMOUNT-1 points the same way in a non-diabetic population, but an assumption.

Anyone acting on this needs their own baseline rather than a trial average, because a trial average tells you what happened to a cohort and nothing about what is happening to you. A fasting glucose and HbA1c, a lipid panel, liver enzymes and a tape measure at the navel, taken before the first dose, gives you something real to compare against at three and six months. Waist circumference is the cheapest visceral fat proxy there is, and SURMOUNT-1's 18.1 cm reduction [1] is a reminder that it moves enough to be worth tracking.

You can compare weight-loss providers on what their intake actually includes — the gap between services that run a full panel and services that run a questionnaire is the main thing separating them. PeterMD and Hims both prescribe tirzepatide where it is appropriate. If you would rather have the monitoring done in person, a local medical weight-loss clinic can handle the follow-up — BODENVY in Winter Park, Florida is one example of a practice that combines the prescribing and the body-composition side. The weight management hub covers what belongs on the ongoing panel.

What survodutide has measured instead

Survodutide's programme is not thin. It is simply pointed somewhere else.

The phase 2 dose-finding trial ran 46 weeks in adults with a BMI of 27 or above without diabetes [4]. Mean body weight change was −6.2% at 0.6 mg, −12.5% at 2.4 mg, −13.2% at 3.6 mg and −14.9% at 4.8 mg, against −2.8% on placebo. Among participants who stayed on 4.8 mg throughout, weight loss reached 18.7%, and 55% of that group lost at least 15% of body weight. Gastrointestinal events were the most common adverse events.

SYNCHRONIZE-1, published in the New England Journal of Medicine, took 725 adults with obesity out to 76 weeks [5]. Mean weight change was −12.2% at 3.6 mg and −13.0% at 6.0 mg against −5.4% on placebo, with 72.6%, 71.9% and 46.3% respectively reaching at least 5% reduction. Funded by Boehringer Ingelheim.

Then the trial that gets cited as visceral-fat evidence but is not: the phase 2 MASH trial, 48 weeks in adults with biopsy-confirmed metabolic dysfunction-associated steatohepatitis and F1–F3 fibrosis [6]. MASH resolution without worsening fibrosis occurred in 47%, 62% and 43% across the 2.4, 4.8 and 6.0 mg groups against 14% on placebo. Nausea ran 66% versus 23%; serious adverse events were 8% versus 7%.

Those are genuinely impressive liver results. Liver fat is not visceral fat. They are correlated, they respond to overlapping mechanisms, and a drug that strips hepatic fat is a reasonable bet to strip visceral fat too — but hepatic steatosis is measured inside the liver and VAT is measured in the abdominal cavity around the organs. A biopsy showing MASH resolution tells you nothing directly about the second. The substitution happens constantly in coverage of this drug, and it is the single biggest reason the comparison looks more settled than it is.

Search survodutide's published record for a DXA or MRI visceral-fat endpoint and you will not find one.

The glucagon argument, and what it is worth

The mechanistic case for survodutide is real and worth stating properly, because dismissing it would be as lazy as over-trusting it.

Tirzepatide is a GIP and GLP-1 receptor agonist. Survodutide is a glucagon and GLP-1 receptor agonist. That glucagon arm is the difference, and glucagon receptor agonism increases energy expenditure and drives hepatic fat mobilisation directly, rather than achieving fat loss purely as a downstream consequence of eating less. On that reasoning, a glucagon-containing agonist should preferentially hit the visceral and hepatic compartments — and the MASH results are exactly what you would predict if that reasoning is correct [6].

So the hypothesis has a mechanism and it has supporting evidence from an adjacent endpoint. That is a respectable position. It is also several steps short of a measurement, and the history of this drug class is full of mechanistic predictions that did not survive contact with imaging. The honest framing is that survodutide is a strong candidate to beat tirzepatide on visceral fat, and that nobody yet knows whether it does.

There is a second-order reason to be careful here. Glucagon receptor agonism raises energy expenditure, which sounds unambiguously good and is the part of the story that gets told. It also raises hepatic glucose output, which is why pairing it with a GLP-1 agonist is necessary rather than merely convenient — the GLP-1 arm is doing work to offset something the glucagon arm is doing. A drug built from two opposing effects can land anywhere depending on the balance between them, and the balance is dose-dependent. That is an argument for measuring rather than predicting.

It is also worth noting what the comparison would look like if the mechanistic logic were applied consistently. Tirzepatide's GIP agonism was widely expected to be the weaker component when it entered trials, on the reasoning that GIP is obesogenic. It turned out to outperform the predictions substantially. The class has a track record of mechanistic forecasts being wrong in both directions, which should temper confidence in this one.

What would settle it is unglamorous: a prespecified MRI or DXA substudy in survodutide's phase 3 programme, or a head-to-head against tirzepatide with imaging endpoints. Neither has reported, and until one does, anyone telling you survodutide is the better choice for visceral fat is extrapolating rather than citing.

What this means if you are choosing today

The practical answer is less interesting than the argument, and that is usually a sign it is right.

Tirzepatide is the only one of the two you can actually get, and it is the only one with visceral fat data. Survodutide is not approved anywhere; its phase 3 programme is still running. A comparison between an approved drug with imaging data and an investigational drug with a mechanism is not a close call for anyone making a decision this year — whatever it might become in 2028.

If you are already on tirzepatide and reading this wondering whether to wait for something better: the 40.1% visceral fat reduction in SURMOUNT-1 is a large effect by any standard [1], and the SURPASS-3 result showing VAT falling where insulin raised it is the kind of finding that holds up [2][3]. There is no evidence-based reason to hold out.

If you are watching survodutide because the glucagon mechanism appeals to you, that is a defensible thing to watch. Watch it for what it is — a promising phase 3 candidate whose visceral fat effect is unmeasured — rather than as an established winner.

One expectation worth setting, because it is where most people get discouraged: these are long timelines. SURMOUNT-1's body composition readout came at 72 weeks and SURPASS-3's at 52 [1][3]. Neither trial is telling you what your visceral fat looks like at week 12, and the scale-shifting that happens in the first few months is not a reliable proxy for what is happening around the organs. Early weight loss on these drugs is a mix of fluid, glycogen, subcutaneous fat, visceral fat and lean tissue in proportions nobody can read from a bathroom scale.

The practical consequence is that a waist measurement every eight weeks is more informative than daily weigh-ins, and a repeat lipid and HbA1c panel at six months is more informative than either. If the waist is coming down faster than total weight, the composition of the loss is going the right way. If weight is dropping and the waist is not following, that is worth raising with whoever is prescribing — it is the pattern that suggests a larger share of the loss is coming from lean tissue, and it is actionable while it is happening rather than after.

And the point that applies regardless of which drug: the lean mass figure deserves more attention than it gets. Losing 10.9% of lean mass alongside 40.1% of visceral fat [1] is a good trade and still a trade, and it is the argument for resistance training and adequate protein running alongside either drug rather than instead of them. We have covered how much protein to eat on these drugs and how tirzepatide compares with semaglutide for men separately, and our earlier look at survodutide against semaglutide covers the weight-loss side of this drug in more depth.

Frequently asked questions

Is survodutide better than tirzepatide for visceral fat?

Unknown. Survodutide has no published visceral fat imaging data — not from its dose-finding trial, SYNCHRONIZE-1, or the MASH trial [4][5][6]. Tirzepatide has it from two substudies [1][2]. The expectation that survodutide wins comes from its glucagon mechanism and its liver results, which is a reasonable prediction rather than a measured outcome.

Can I get survodutide?

No. It is investigational and not approved in any market, with phase 3 trials ongoing [5]. Tirzepatide is approved and available through telehealth and in-person clinics.

Doesn't the MASH trial prove survodutide reduces visceral fat?

It does not. The MASH trial measured hepatic steatosis and fibrosis by biopsy [6]. Liver fat and visceral fat are different compartments measured in different ways. They are correlated and both tend to respond to the same interventions, but a liver result is not a visceral fat result, and the two get conflated constantly in coverage of this drug.

How much visceral fat does tirzepatide remove?

In the SURMOUNT-1 DXA substudy, visceral fat mass fell 40.1% against 7.3% on placebo over 72 weeks, an estimated treatment difference of 32.8% [1]. Visceral fat dropped proportionally more than total fat mass. Those results were post hoc rather than prespecified, which is a genuine limitation.

Will visceral fat come back if I stop?

The imaging substudies did not follow participants off-drug, so there is no direct answer from this data [1][3]. What is well established across the class is that weight regain follows discontinuation in most people without a maintenance strategy, and there is no reason to expect the visceral compartment to be spared. Plan for what happens after, not just what happens during.

Should I worry about lean mass loss?

It is worth managing rather than worrying about. SURMOUNT-1 saw lean mass fall 10.9% against 2.6% on placebo — roughly a quarter of total weight lost [1], which is broadly in line with other drugs in the class and with substantial weight loss generally. Resistance training and adequate protein intake are the standard mitigations, and they matter more at larger losses.

Sources

• [1] Look et al., Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight, Diabetes, Obesity and Metabolism, 2025
• [2] Lilly, Tirzepatide led to greater improvements in liver fat content compared to insulin degludec (SURPASS-3 MRI substudy) — company press release
• [3] NCT03882970 — SURPASS-3: tirzepatide versus insulin degludec in type 2 diabetes
• [4] Glucagon and GLP-1 receptor dual agonist survodutide for obesity: a randomised, double-blind, placebo-controlled, dose-finding phase 2 trial, The Lancet
• [5] Survodutide Once Weekly for the Treatment of Adults with Obesity (SYNCHRONIZE-1), New England Journal of Medicine
• [6] Sanyal et al., A Phase 2 Randomized Trial of Survodutide in MASH and Fibrosis, New England Journal of Medicine, 2024
• [7] NCT04184622 — SURMOUNT-1: tirzepatide in participants with obesity or overweight