Survodutide

An experimental peptide that switches on both the glucagon receptor and the GLP-1 receptor. It is being developed by Boehringer Ingelheim and Zealand Pharma and is not approved as a medicine.

Sold as

Injection under the skin (only within clinical trials).

Legal status in the EU

No marketing authorisation in the EU or the US. Available only within clinical trials.

It acts on two receptors at once. The GLP-1 side gives fullness and blood sugar control, while the glucagon side is being studied particularly for effects on the liver; a separate phase 3 programme is running in fatty liver disease. The sequence has not been verified against a citable source and is left empty.

Part of the bodyWhat happensHow it is knownKind of evidence

The liver

This is the part that separates survodutide from the drugs it is related to. Fat builds up inside liver cells in people with weight and blood-sugar problems; when the fat is joined by inflammation and swollen, damaged liver cells, the condition has a name — metabolic dysfunction-associated steatohepatitis, shortened to MASH. Without inflammation it is called metabolic dysfunction-associated steatotic liver disease, or MASLD. On survodutide the fat drains out of the liver fast and far, the two blood tests that go up when liver cells are being damaged come back down, and on a second biopsy the inflammation has often settled. The scarring — the part that actually decides whether someone goes on to liver failure — moves much less.

Measured in 293 adults whose MASH was confirmed by biopsy, over 48 weeks. On the highest amount, the fat in the liver fell 62.0% against 5.7% on dummy injections. The blood test called ALT fell 38.5 units per litre against 5.7, and AST fell 32.2 against 2.4. Measured again in 216 adults with fatty liver disease diagnosed mostly by scan rather than biopsy, over 48 weeks: 84.2% of people on survodutide lost at least 30% of their liver fat against 24.3% on dummy injections, and 61.0% ended with a liver that was less than 5% fat — the normal range — against 5.7%.

Measured in people

Source [04]Source [05]Source [06]

Appetite, and how much people eat

People eat less and lose weight, and the weight loss is large. What is odd is that when the largest trial asked people directly about their eating — how much they wanted to eat, how well they could resist — the answers on the drug were barely different from the answers on dummy injections. Something is reducing what people eat that the questionnaire did not catch. Loss of appetite was reported as a side effect by roughly one person in six.

Weight measured in 725 adults over 76 weeks. On the same people, the eating-behaviour questionnaire fell 8.1 points on the higher amount against 7.3 points on dummy injections — a difference of 0.9 points, with a range running from 2.6 points better to 0.9 points worse, which answers nothing. The threshold the questionnaire's authors call meaningful is 8 points. Loss of appetite was reported by 11.6% and 17.8% on the two amounts against 7.9% on dummy injections.

Measured in people

Source [07]

Stomach and gut

The most common thing that happens on this compound, by a very wide margin, is that people feel sick. Vomiting, loose stools and constipation follow, and constipation and loose stools can arrive in the same person in different weeks. It clusters in the months when the amount is being stepped up, and it is the single biggest reason people leave the trials — about one person in five stopped taking survodutide because of gut trouble, against roughly one in thirty-five on dummy injections.

Measured in 483 adults on survodutide and 242 on dummy injections over 76 weeks: some gut reaction in 80.9% and 89.7% on the two amounts against 47.9%. Nausea 61.0% and 64.9% against 17.4%. Vomiting 40.7% and 44.6% against 6.2%. Loose stools 32.0% and 39.7% against 18.6%. Constipation 31.1% and 32.6% against 12.4%. Gut trouble ended treatment for 17.8% and 20.2% against 2.9%. The same pattern in the 48-week liver trial in 219 adults: nausea 66% against 23%, loose stools 49% against 23%, vomiting 41% against 4%.

Measured in people

Source [04]Source [07]

Blood sugar

There is a reason to worry here and the worry has not been borne out. Glucagon is the hormone that pushes blood sugar up — it is the one the body releases when sugar runs low — so a compound that switches on the glucagon receptor could in principle make diabetes worse. It did the opposite. In people with type 2 diabetes, long-term blood sugar fell as far in 16 weeks as it did on semaglutide given in the same trial. In people without diabetes it fell slightly, and high blood sugar was reported less often on the drug than on dummy injections.

Measured in 411 adults with type 2 diabetes over 16 weeks, all of them already on metformin. On 1.8 mg once a week, long-term blood sugar fell 1.71 percentage points; on open-label semaglutide in the same trial it fell 1.47 points. Measured again in 725 adults without diabetes over 76 weeks: long-term blood sugar fell 0.2 percentage points against no change on dummy injections, and fasting blood sugar fell 7.4 and 6.0 mg/dl against 1.2. High blood sugar was reported as a side effect by 0.8% and 1.7% on the drug against 5.8% on dummy injections.

Measured in people

Source [07]Source [08]

Heart and blood vessels

The resting heart runs a little faster. This is the effect that killed off two earlier compounds of the same kind, so it is watched closely. Blood pressure fell, but it fell about as much on dummy injections, so the trial could not credit the drug with it. Feeling faint, and blood pressure dropping when standing, were reported more often on the higher amount. Nobody yet knows what any of this adds up to over years: the trial built to answer that question finished in June 2026 and has published nothing.

Measured in 483 adults over 76 weeks: resting heart rate up 3.2 to 3.5 beats a minute against down 0.4 on dummy injections. The top blood pressure number fell 6.7 and 5.4 mmHg against 5.9 — a difference of under one point either way, which answers nothing. Low blood pressure including fainting spells was reported by 2.1% and 7.4% against 1.7%. In the 48-week liver trial the heart-rate rise was 2.4 to 4.2 beats a minute against 1.6 on dummy injections. No electrical disturbance of the heart was seen in either trial. The heart-outcomes trial in 5,531 adults is recorded as completed on 30 June 2026 with no results posted.

Measured in people

Source [04]Source [07]Source [09]

Fat and muscle, and where the fat sits

The fat that goes first is the fat packed around the organs inside the belly — the kind most closely tied to diabetes and heart disease. It goes faster than the fat under the skin, and much faster than the weight on the scales would suggest. Muscle and other lean tissue come off too. Whether it comes off in a better proportion than on other weight drugs is the company's claim and is not something this substudy can settle, because it was never compared against another drug.

Scanned by MRI in a planned substudy of 25 people per group inside the 725-person trial, at the start and at 76 weeks. On the higher amount: total body fat down 27.8% against 9.5% on dummy injections, the fat inside the belly down 34.0% against 11.8%, the fat under the skin down 28.1% against 9.8%, liver fat down 63.1% against 24.5%. Lean tissue fell 9.8% against 2.9%. Of every unit of tissue lost, lean tissue made up 10.8% on the drug and 12.8% on dummy injections. Seventy-five people is a small number and this is the only body-composition data that exists.

One small study

Source [07]

How much energy the body burns

This is the part everyone assumes and nobody has shown. The argument for switching on the glucagon receptor is that it makes the body burn more fuel and pushes the liver to burn its own fat, so that weight comes off through two routes rather than one. In mice, that is what happens: the GLP-1 half does the eating, the glucagon half does the burning. In people it has never been measured. The receptor that would do the burning is barely present in the parts of the human brain that control appetite, which is the strongest human evidence that the two halves do different jobs — but it is evidence about where a receptor sits, not about what a person's body does.

Established in lean and fat mice, with human tissue used only to check which cells carry the receptors. The authors state plainly that the functional testing was done exclusively in mouse models and not in human tissue. Nine of the paper's authors are employees of the company developing survodutide. The study designed to measure this in people — 64 adults with obesity, comparing survodutide against semaglutide, with metabolic rate measured overnight in a sealed chamber — is recorded as active with results expected in January 2027.

Only seen in animals

Source [10]Source [11]

The phase 2 study (Lancet Diabetes & Endocrinology 2024, 386 people treated at 43 centres in 12 countries, 46 weeks) gave between −6.2% in the lowest dose group and −14.9% in the highest, against −2.8% for placebo. Only 233 of 386 participants (60.4%) completed the treatment period. The phase 3 programme SYNCHRONIZE-1, SYNCHRONIZE-2, SYNCHRONIZE-MASLD and a heart outcomes study are ongoing; baseline data for SYNCHRONIZE-1 (725 participants) has been published, but that paper reports no results on effect.

Slow to get on, and nothing at all is known about getting off. Every trial spent months stepping the amount up — 20 weeks in the first obesity trial, 24 weeks in both liver trials, and up to 32 weeks in the 76-week phase 3 trial. The gut trouble is concentrated in exactly that stretch: the phase 3 paper places the nausea, vomiting and loose stools mainly in the step-up period, and the liver trial's authors note that most of the people who quit did so during the fast step-up, and suggest going slower. Weight comes off through the first year and then flattens. The heart-rate rise was measured at 48 weeks and again at 76 weeks and was about the same at both, 2.4 to 4.2 beats a minute and then 3.2 to 3.5, which suggests it arrives early and stays. The liver results were read at 48 weeks in both liver trials, and no published measurement exists earlier than that. Beyond 76 weeks nothing has been published on anything. And there is no published follow-up whatsoever after stopping — not one trial has reported what happens to weight, liver fat or heart rate in the weeks or months after the last injection.

What was measuredWhat happenedIn whomHow it is known

Inflammation in the liver settling without the scarring getting worse

This is the compound's strongest and most distinctive result, so it is worth unpacking. Everyone had a piece of liver taken at the start and again at 48 weeks, and a pathologist who did not know who had received what compared the two. The target was a real improvement in the fat, the inflammation and the swollen damaged cells, with the scarring no worse than before. It was reached by 47% on the lowest amount, 62% on the middle amount (95% CI 51 to 73) and 43% on the highest, against 14% on dummy injections (95% CI 8 to 23). The middle amount beat the highest, which is not what a dose-response is supposed to look like and is why the trial's own statistics describe a curve rather than a straight line. Anyone who did not come back for the second biopsy was counted as a failure.

293 adults aged 18 to 80 with MASH confirmed by biopsy and scarring at stage 1, 2 or 3 of 4, 48 weeks; 41% were at stage 2 and 35% at stage 3, and 39% also had type 2 diabetes

Measured in people

Source [04]

The scarring in the liver itself getting better

Much weaker, and the paper says so. Scarring improved by at least one stage in 34%, 36% and 34% across the three amounts against 22% on dummy injections — a gap of about twelve points that the authors describe as suggesting a possibility of benefit rather than showing one. On the highest amount, 32% had their scarring improve with no worsening of the inflammation, against 18%. The confidence ranges around these were not adjusted for the number of comparisons made, and the paper states outright that they should not be used to infer a definite treatment effect.

The same 293 adults, all with paired biopsies at the start and at 48 weeks

Measured in people

Source [04]

Fat draining out of the liver, measured by scan

The largest effect anywhere in this record. At least 30% of the liver fat gone: 84.2% on survodutide against 24.3% on dummy injections. At least 70% gone: 55.5% against 2.9%. A liver back under 5% fat, which is the normal range: 61.0% against 5.7%. Counted the harder way, which includes everyone who stopped early, the first of those figures becomes 68.5% against 28.6%. This is a scan, not a biopsy: it measures how much fat is in the liver and says nothing directly about the inflammation or the scarring.

216 adults with obesity or overweight and fatty liver disease with signs of inflammation or scarring, 146 on survodutide and 70 on dummy injections, 48 weeks. 90.3% were diagnosed by scan and blood tests rather than biopsy; the other 9.7% had a biopsy within the previous three years. People with cirrhosis were excluded.

Measured in people

Source [05]Source [06]

Body weight in adults with obesity and no diabetes, over 76 weeks

The same trial produced two headline numbers and both are true. Counting everyone as they were assigned, including the quarter who stopped taking it and the people who went and got another weight drug, weight fell 12.2% on the lower amount and 13.0% on the higher (95% CI 14.4 to 11.6), against 5.4% on dummy injections (95% CI 6.9 to 4.0) — a difference of 7.5 percentage points. Counting only what happened while people were actually taking what they were given, weight fell 15.3% and 16.6% against 3.2%. The gap between 13.0% and 16.6% is not spin; it is the cost of the people who could not stay on it. The gap between 5.4% and 3.2% has its own explanation: 16.5% of the dummy group obtained another GLP-1 medicine, approved or made up by a compounding pharmacy, while still in the trial.

725 adults with a BMI of 30 or more, or 27 or more with an obesity-related problem, and no diabetes; mean age 47.1, 40.6% men, mean weight 108.8 kg; 116 sites in 14 countries

Measured in people

Source [07]Source [12]

How many people lost a lot of weight rather than a little

Counting everyone as assigned: at least 5% of body weight gone in 72.6% and 71.9% on the two amounts against 46.3% on dummy injections. At least 10% in 55.2% and 56.6% against 26.0%. At least 15% in 35.7% and 45.9% against 12.0%. At least 20% in 24.9% and 28.5% against 6.6%. The dummy figures are unusually high for a weight trial, for the reason above.

The same 725 adults, 76 weeks

Measured in people

Source [07]

Body weight in the earlier dose-finding trial

Weight fell 6.2% on the smallest amount, 12.5%, 13.2% and 14.9% as the amount went up, against 2.8% on dummy injections. But only 233 of 386 people — 60.4% — finished the 46 weeks, and that was true in both groups: 61% of those on survodutide and 60% of those on dummy injections. A trial that loses two people in five leaves the question of how well it is tolerated genuinely open.

386 adults with a BMI of 27 or more and no diabetes, treated at 43 centres in 12 countries, 46 weeks, with 20 of those weeks spent stepping the amount up

Measured in people

Source [01]

Body weight in people who had fatty liver disease

Down 12.2% against 1.0% on dummy injections over 48 weeks, counting what happened while people were taking what they were given; down 8.7% against 1.4% counting everyone. At least 5% of body weight gone in 79.1% against 14.3%.

The same 216 adults with obesity or overweight and fatty liver disease, 48 weeks

Measured in people

Source [05]Source [06]

Blood sugar and weight in people with type 2 diabetes, against semaglutide

The only head-to-head comparison against another drug that exists for survodutide, and it is short and small. Over 16 weeks, long-term blood sugar fell 1.71 percentage points on 1.8 mg of survodutide a week and 1.47 points on up to 1.0 mg of semaglutide a week. Weight fell up to 8.7% on survodutide against 5.3% on semaglutide. Side effects were reported by 77.8% of people on survodutide, 52.0% on semaglutide and 52.5% on dummy injections. The semaglutide arm was open-label — everyone knew who was on it — and the amount used was below the 2.4 mg licensed for weight loss.

411 adults with type 2 diabetes not well controlled on metformin, BMI 25 to 50, 16 weeks

Measured in people

Source [08]

Waist, blood fats and uric acid

Waist down 10.5 cm and 11.5 cm on the two amounts against 5.4 cm on dummy injections. Triglycerides down 22.7% and 27.7% against 8.8%. The lightest, most harmful kind of cholesterol down 24.6% and 30.1% against 10.6%. LDL cholesterol down 8.8% and 9.0% against 5.7%. Uric acid, which causes gout, down 53.0 and 73.8 units against 8.6. Fasting insulin down about 29% against 8.6%.

The same 725 adults, 76 weeks

Measured in people

Source [07]

How people described their own eating

A null result, printed because it is surprising. On a twelve-question survey about wanting to eat and being able to resist, scores fell 7.8 and 8.1 points on the two amounts against 7.3 on dummy injections. The differences from dummy injections were 0.5 and 0.9 points, with ranges that crossed zero. The survey's own threshold for a meaningful change is 8 points on the total — which both the drug groups and the dummy group roughly met. Whatever is making people eat less, this questionnaire did not detect a difference.

The same 725 adults, 76 weeks

Measured in people

Source [07]

Blood pressure

Nothing that could be credited to the drug. The top blood pressure number fell 6.7 mmHg and 5.4 mmHg on the two amounts and 5.9 mmHg on dummy injections. The bottom number fell 4.1 and 3.6 against 3.4. This is what it looks like when a change is real but is not the drug's doing. The 48-week liver trial did report a fall against dummy injections in both numbers, in a different and smaller group of people.

725 adults over 76 weeks, and separately 216 adults with fatty liver disease over 48 weeks

Studies in people disagree

Source [06]Source [07]

Heart attacks, strokes and heart deaths

Nothing has been measured. In the 725-person trial an independent committee confirmed exactly one such event over 76 weeks — a stroke, in a person on dummy injections — and none in either survodutide group. One event against nothing is far too little to mean anything in either direction. The trial designed to answer this enrolled 5,531 adults with heart or kidney disease or risk factors for it, across 524 sites in 34 countries, and is recorded as completed on 30 June 2026 with no result published.

725 adults over 76 weeks; separately, 5,531 adults in a completed but unreported trial

One small study

Source [07]Source [09]

In phase 2, side effects were reported in 281 of 309 people on survodutide (91%) against 58 of 77 on placebo (75%). They were mainly stomach and bowel effects, affecting 232 of 309 (75%) in the survodutide groups. That nearly 40% of participants did not complete the study leaves tolerability an open question. Long-term safety is unknown.

EffectWhat it feels likeHow often

Gallstones and an inflamed gallbladderSerious

Losing weight quickly causes gallstones by itself, and no trial of this compound has been able to separate the two. Stones can lead to hospital admission and to the gallbladder being taken out. The signs are pain under the right ribs, fever and yellowing of the eyes.

Uncommon. Gallstones in 1.2% and 2.1% of people on the two amounts against 0.4% on dummy injections; an inflamed gallbladder in 0.8% and 0.4% against none. Measured in 725 adults over 76 weeks.

Source [07]

Blood sugar dropping too lowSerious

Shaking, sweating, confusion and sudden hunger. The pattern in the trials is clear: on its own this compound did not drive blood sugar dangerously low, and it lowered it less than a drug for diabetes normally would in people who did not have diabetes. The risk was carried entirely by people already on medicines that push insulin out.

In people without diabetes: rare, and never severe. Reported by 1.2% and 0.8% on the two amounts against 0.4% on dummy injections in 725 adults, with no severe episode at all. In the liver trial, where 39% of people had type 2 diabetes: 5% against 3%, and every case was in someone who had diabetes or high insulin and was taking other blood-sugar medicines.

Source [04]Source [07]

Kidney damageSerious

Two events is not a rate and cannot be turned into one. It is listed here because the mechanism is well known for this whole class of drugs — days of vomiting or loose stools dry the body out and the kidneys fail because of it — and because both of the two side-effect reports ever filed for survodutide with the US regulator name kidney injury.

Two people out of 483 on the drug over 76 weeks — one in each dose group, 0.4% — against none of the 242 on dummy injections. Both were the mildest of the four stages.

Source [07]Source [13]

A suicide attemptSerious

The trial records that this happened in the context of family, relationship and academic difficulties. One event in 483 people establishes nothing either way, and it is here because leaving it out would be the dishonest choice. People whose mood disorder was not well controlled were kept out of the trial in the first place, so the group most likely to be affected was never tested.

One person out of 483 on the drug over 76 weeks, on the lower amount. Nobody on dummy injections. Set against that, depression was reported by 3.3% and 2.5% on the two amounts against 5.0% on dummy injections.

Source [07]

Feeling sick

The defining side effect of this compound, and the one that ends treatment. It arrives while the amount is being stepped up and is worst then. Across the trials it is the biggest single reason people leave: gut trouble ended treatment for about one person in five on survodutide against about one in thirty-five on dummy injections.

Very common — 61.0% and 64.9% of people on the two amounts against 17.4% on dummy injections, in 725 adults over 76 weeks. 66% against 23% in the 48-week liver trial.

Source [04]Source [07]

Vomiting

Seven times as common as on a dummy injection, which is a wider gap than for any other effect in the tables. It clusters in the weeks when the amount is being increased. It is also the route by which people get dried out, which is what the kidney warnings rest on.

Very common — 40.7% and 44.6% against 6.2% on dummy injections. In the liver trial, 41% against 4%.

Source [04]Source [07]

Loose stools

Notably common on dummy injections too, so a good part of it is not the drug. What is the drug is the extra twenty points on the higher amount. People who moved to survodutide from tirzepatide describe this as the direction the bowel trouble runs on this compound.

Very common — 32.0% and 39.7% against 18.6% on dummy injections. In the liver trial, 49% against 23%.

Source [07]

Constipation

The mirror image of the loose stools, and it can happen to the same person in a different week. About one person in three had it, which is almost exactly the proportion who had the opposite problem.

Very common — 31.1% and 32.6% against 12.4% on dummy injections.

Source [07]

Tiredness

About twice as common as on a dummy injection in both trials, which makes it one of the more consistent non-gut effects. Nothing published says how long it lasts or whether it settles.

Common — 18.7% and 14.9% against 8.7% on dummy injections over 76 weeks. In the liver trial, 17% against 8%.

Source [04]Source [07]

Heartburn, indigestion and belching

Belching is ten times as common as on a dummy injection on the lower amount, and it is the kind of thing nobody warns people about. Indigestion was also more common in the liver trial, at 14% against 4%.

Common — acid coming back up the throat in 11.6% and 14.5% against 3.7% on dummy injections; indigestion in 12.0% and 11.6% against 4.1%; belching in 12.4% and 7.4% against 1.2%.

Source [04]Source [07]

A faster resting heart

This is the effect that matters most for this particular compound, because it is the one plausibly coming from the glucagon half rather than the GLP-1 half, and because two earlier compounds of the same kind were abandoned partly over it. Most people will not notice a few beats a minute. Neither trial found any electrical disturbance of the heart. What a small permanent rise does over years is not established for any drug in this family, and the trial that could answer it for this one has not reported.

An average rise of 3.2 to 3.5 beats a minute at 76 weeks, against a fall of 0.4 beats a minute on dummy injections. In the 48-week liver trial, a rise of 2.4 to 4.2 beats a minute against 1.6.

Source [04]Source [07]

Dizziness, and blood pressure dropping when standing

This runs strongly with the higher amount: dizziness was less common than on a dummy injection at 3.6 mg and nearly twice as common at 6.0 mg. The trial recorded the fainting episodes as mostly mild to moderate and not of the serious kind. Fainting still means falling over, which is how people break bones.

Dizziness in 5.4% and 13.2% on the two amounts against 7.0% on dummy injections. Low blood pressure including fainting spells in 2.1% and 7.4% against 1.7%. Fainting itself in 2.1% and 1.7% against 0.8%.

Source [07]

Odd skin sensations — burning, tingling or numbness

Three to four times as common as on a dummy injection, from a small base. It is unexplained, it is not a familiar effect of the GLP-1 drugs, and nothing published says whether it goes away.

Uncommon — 3.7% and 2.5% on the two amounts against 0.8% on dummy injections. The trial records all of them as not serious and mostly mild to moderate.

Source [07]

Pancreas enzymes rising with no symptoms at all

A blood test comes back at three times the normal level or higher, and the person feels nothing. It matters because the same enzymes rise when the pancreas is genuinely inflamed, which is the dangerous version. In 725 people over 76 weeks, and again in 293 people over 48 weeks, not one confirmed case of an inflamed pancreas was found on survodutide.

Uncommon — confirmed by independent review in 2.5% and 1.7% on the two amounts against 1.2% on dummy injections. Also more common on the drug than on dummy injections in the 48-week liver trial.

Source [04]Source [07]

These are the amounts the studies below gave their participants, and they are here as facts about those studies. They are not a recommendation, not a starting point and not a range to pick from. What a person should take, if anything, is a question for a doctor who knows them.

SYNCHRONIZE-1, weight loss in 725 adults with obesity or overweight and no diabetes

Once a week under the skin, stepped up to either 3.6 mg or 6.0 mg. By the end, 75.5% of the lower-amount group and 62.4% of the higher-amount group were still on the amount they had been assigned; the rest finished on less

76 weeks

Source [07]

The phase 2 trial in 293 adults with inflamed fatty liver disease confirmed by biopsy

2.4 mg, 4.8 mg or 6.0 mg once a week under the skin, reached over a 24-week step-up period

48 weeks — 24 weeks going up, then 24 weeks holding the amount steady

Source [04]

SYNCHRONIZE-MASLD, 216 adults with obesity and fatty liver disease with signs of inflammation or scarring

Once a week under the skin, stepped up over 24 weeks to a maximum of 6.0 mg, with 2.4 mg as the lowest amount anyone was held at

48 weeks

Source [05]Source [06]

The phase 2 dose-finding trial in 386 adults with obesity and no diabetes

0.6 mg, 2.4 mg, 3.6 mg or 4.8 mg once a week under the skin, reached over a 20-week step-up period

46 weeks — 20 weeks going up, then 26 weeks holding the amount steady

Source [01]

The phase 2 trial in 411 adults with type 2 diabetes, which also ran an open-label semaglutide arm

Up to 0.3 mg, 0.9 mg, 1.8 mg or 2.7 mg once a week, or up to 1.2 mg or 1.8 mg twice a week, all under the skin and all on top of metformin

16 weeks

Source [08]

What people report

• Whether the liver improvement stops anyone becoming ill — no trial has yet measured liver failure, transplant or death, and the two that will do not finish until 2029 and 2031
• How much of the effect comes from the glucagon half and how much from the GLP-1 half — the split has only been established in mice
• Whether it makes a person burn more energy at all — the study measuring it in 64 adults reports in January 2027
• What happens after stopping — not one trial has published weight, liver fat or heart rate in the weeks or months after the last injection
• Whether it works in people who have both obesity and type 2 diabetes — that trial finished in March 2026 and has published nothing
• Whether it causes or prevents heart attacks and strokes — the 5,531-person trial built to answer that finished in June 2026 and has published nothing
• What a permanent rise of three to four heartbeats a minute does over years
• Why the higher amount worked less well than the middle one on the liver biopsy in the phase 2 trial
• Whether the scarring in the liver actually improves, as opposed to merely not getting worse — the one measurement of that was described by its own authors as not usable to infer a treatment effect
• Whether the fat that drains out of the liver on a scan corresponds to anything happening to the inflammation or the scarring in the phase 3 liver trial, which took no routine biopsies
• Why people report craving sugar in the first weeks, when the questionnaire in the largest trial found no difference from dummy injections in how people described their eating
• How much muscle is lost over a full course — the only body-composition scans cover 25 people per group in a single substudy
• Whether the odd burning and tingling skin sensations, reported three to four times as often as on dummy injections, mean anything or go away
• What it does in anyone under 18, in pregnancy, while breastfeeding, or over 80 — none of them has been studied
• What it does in people with uncontrolled high blood pressure, unstable depression, a recent heart event or a history of an inflamed pancreas — all of them were excluded from the trials
• Whether taking it beyond 76 weeks is safe or keeps working — nothing published goes further
• What is actually in the vials being sold online to people outside the trials, and what it does to them

Can I get this?

Not legally, and not from a doctor. Survodutide has no marketing authorisation in the European Union, the United States or anywhere else, and no application for one has been announced. The only lawful way to receive it is to be enrolled in one of the trials. It has been given a fast-track status in three places — accepted into the European regulator's PRIME scheme in November 2023, granted Breakthrough Therapy designation by the US regulator in October 2024 for inflamed fatty liver disease with moderate or advanced scarring, and a similar designation in China in June 2024 — but every one of those is a promise of faster review, not a decision that it works or is safe. It is also banned in sport at all times, under the anti-doping rule that covers any substance still in clinical development.

Source [03]Source [14]

What is MASH, and what does "the inflammation improved without the scarring getting worse" actually mean?

Fat collects inside the cells of the liver in people with weight and blood-sugar problems. On its own that is called metabolic dysfunction-associated steatotic liver disease, or MASLD, and for most people it sits there quietly. In some people the fat is joined by inflammation and by liver cells that swell up and die, and that is metabolic dysfunction-associated steatohepatitis, or MASH. The dying and repairing lays down scar tissue, and it is the scar tissue — not the fat and not the inflammation — that eventually stops the liver working. So the trial's target had two halves. The first half is that the fat, the inflammation and the swollen cells all measurably improved when a pathologist compared a piece of liver taken at the start with a piece taken 48 weeks later. The second half is that the scarring had not got any worse in the meantime. For a person, the honest translation is: the process that makes the scarring calmed down, and the scarring already there mostly stayed where it was. Whether that means fewer people go on to liver failure is exactly what nobody knows, and the two trials that could answer it run to 2029 and 2031.

Source [04]Source [15]

Why do I see 16.6% in one place and 13.0% in another for the same trial?

Because the same 725 people were counted two ways, and both counts were published. The first way counts only what happened while people were actually taking what they were given: 16.6% lost on the higher amount, against 3.2% on dummy injections. The second way counts everyone as they were assigned, including the quarter who stopped taking it and anyone who went and got a different weight drug: 13.0%, against 5.4%. The trial's own primary analysis was the second one — the stricter one — because that is what regulators ask for and because it is closer to what happens to a real group of people, some of whom will not manage to stay on it. The company's press release led with 16.6%. Both numbers are in the paper. Reading either without the other gives a false picture in a different direction.

Source [07]Source [16]

What does the glucagon half actually add?

In theory: the GLP-1 half makes you eat less, and the glucagon half makes your body burn more and pushes the liver to burn its own fat, so the two work by different routes. In mice, that is what has been shown — food intake is driven by the GLP-1 receptor, energy burning by the glucagon receptor acting on the liver. In people, the split has never been measured. The strongest human evidence is indirect: the glucagon receptor is barely detectable in the parts of the human brain that control appetite, which fits the idea that it is not doing the appetite work. The study that would settle it is running now — 64 adults with obesity, metabolic rate measured overnight in a sealed chamber, survodutide against semaglutide — and reports in January 2027. Until then, how much of survodutide's effect comes from its second receptor is an argument, not a fact.

Source [10]Source [11]

Is it better than semaglutide or tirzepatide?

Nobody has run the trial that would answer that for weight. There is exactly one head-to-head comparison anywhere in survodutide's record: 16 weeks, 411 people with type 2 diabetes, against semaglutide given openly at up to 1.0 mg a week — below the 2.4 mg licensed for weight loss. Survodutide took off up to 8.7% against 5.3%, and lowered long-term blood sugar 1.71 points against 1.47. Sixteen weeks is short and 1.0 mg is not the weight-loss amount, so that comparison settles very little. Comparing across separate trials is not evidence, but for orientation: the 13.0% and 16.6% from survodutide's 76-week trial sit near semaglutide's published weight figures and below tirzepatide's. Where survodutide has something the others do not is the liver, and even there it has never been compared against the one drug approved for that disease.

Source [07]Source [08]

How bad is the sickness, really?

Worse than the class it belongs to, on the published numbers. Some gut reaction was recorded in 80.9% and 89.7% of people on the two amounts, against 47.9% on dummy injections. Nausea reached 64.9% and vomiting 44.6%. The number that matters most is how many people it drove off the drug: gut trouble ended treatment for 17.8% and 20.2%, against 2.9% on dummy injections. In the earlier 46-week trial only 60.4% of people finished at all — and that was true in the dummy group too, at 60%. Both sets of trial authors say the same thing about it: the trouble clusters in the months when the amount is being stepped up, and going up more slowly may reduce it. The phase 3 trial changed its rules part-way through to allow more flexibility, and its authors state that whether doing that earlier would have helped is unknown.

Source [01]Source [04]Source [07]

If glucagon raises blood sugar, does this make diabetes worse?

It did the opposite in every trial that looked. In 411 adults with type 2 diabetes, long-term blood sugar fell as far in 16 weeks as it did on semaglutide given alongside. In 725 adults without diabetes, it fell slightly against no change on dummy injections, fasting blood sugar fell, and high blood sugar was reported as a side effect four times less often on the drug than on dummy injections. The phase 3 authors flag this as the thing that has been hardest to get right in this class of compound — getting the balance between the two receptors such that the glucagon side does not spoil blood sugar control. In survodutide the glucagon receptor is switched on about one eighth as strongly as the GLP-1 receptor, and on the published evidence that balance holds.

Source [04]Source [07]Source [08]

What still has to happen before it could be approved?

Two of the four big trials in obesity have reported and two have not. SYNCHRONIZE-2, in 755 adults who have both obesity and type 2 diabetes, is recorded as completed on 27 March 2026 with no published result. The heart-safety trial in 5,531 adults is recorded as completed on 30 June 2026, also with nothing published — and that is the one regulators require before approving a weight drug. In liver disease the position is further back: everything published so far measured either a biopsy over 48 weeks in 293 people, or fat on a scan. The trials that measure whether people actually get less ill — 1,800 adults with moderate or advanced scarring, and 1,590 adults who already have cirrhosis — are still recruiting and run to 2031 and 2029.

Source [09]Source [15]Source [17]Source [18]

1. [01]le Roux CW m.fl. Glucagon and GLP-1 receptor dual agonist survodutide for obesity: a randomised, double-blind, placebo-controlled, dose-finding phase 2 trial. Lancet Diabetes Endocrinol (2024)
2. [02]Survodutide for treatment of obesity: Baseline characteristics of participants in a randomized, double-blind, placebo-controlled, phase 3 trial (SYNCHRONIZE-1). Diabetes Obes Metab (2026)
3. [03]WADA Prohibited List 2026, S0 non-approved substances: substances in clinical development without regulatory approval are prohibited at all times (official publication, Austrian BGBl. III no. 219/2025) (2025)
4. [04]Sanyal AJ et al. A Phase 2 Randomized Trial of Survodutide in MASH and Fibrosis. N Engl J Med (293 adults, 48 weeks) (2024)
5. [05]Kaplan LM et al. Survodutide in adults with obesity and metabolic dysfunction-associated steatotic liver disease: SYNCHRONIZE-MASLD, a randomized, double-blind, placebo-controlled phase 3 trial. Nat Med (2026)
6. [06]SYNCHRONIZE-MASLD supports survodutide for liver fat content and bodyweight reductions — conference report of the ADA 2026 presentation, Springer Medicine (2026)
7. [07]le Roux CW et al. Survodutide Once Weekly for the Treatment of Adults with Obesity (SYNCHRONIZE-1). N Engl J Med, Tables 1 and 3 (2026)
8. [08]Blüher M et al. Dose-response effects on HbA1c and bodyweight reduction of survodutide, a dual glucagon/GLP-1 receptor agonist, compared with placebo and open-label semaglutide in people with type 2 diabetes: a randomised clinical trial. Diabetologia (2024)
9. [09]ClinicalTrials.gov NCT06077864 — SYNCHRONIZE-CVOT, 5,531 participants, status completed, no results posted (2026)
10. [10]Zimmermann T et al. Survodutide acts through circumventricular organs in the brain and activates neuronal regions associated with appetite regulation. Molecular Metabolism (2026)
11. [11]ClinicalTrials.gov NCT06745284 — effect of survodutide on energy expenditure and fatty acid oxidation in subjects with obesity, 64 participants, active, no results posted (2026)
12. [12]Survodutide substantially reduces weight and liver fat in adults with obesity, phase 3 SYNCHRONIZE studies show. Cleveland Clinic Journal of Medicine, ADA 2026 report (2026)
13. [13]openFDA drug adverse event counts for survodutide, reaction terms by frequency (data last updated 28 April 2026) (2026)
14. [14]Boehringer receives U.S. FDA Breakthrough Therapy designation and initiates two phase III trials in MASH for survodutide (company press release) (2024)
15. [15]ClinicalTrials.gov NCT06632444 — LIVERAGE, 1,800 adults with MASH and moderate or advanced scarring, completion December 2031 (2026)
16. [16]Zealand Pharma announces Boehringer Ingelheim's survodutide Phase III trial showed targeted 34% visceral and 63% liver fat reduction (company press release, 7 June 2026) (2026)
17. [17]ClinicalTrials.gov NCT06066528 — SYNCHRONIZE-2, 755 participants, completed 27 March 2026, no results posted (2026)
18. [18]ClinicalTrials.gov NCT06632457 — LIVERAGE-Cirrhosis, 1,590 participants, recruiting, completion June 2029 (2026)