The GLP-1 Drugs: Ozempic, Wegovy, Mounjaro

TL;DR. Your gut has been talking to your pancreas and your brain about food for as long as you have been eating. The messenger is a hormone called GLP-1, released by cells in the intestinal wall the moment food arrives, which tells the pancreas to release insulin, tells the stomach to empty more slowly, and tells the brain that you have eaten. Natural GLP-1 is destroyed within about two minutes, which made it useless as a medicine for decades. The breakthrough was engineering versions that survive for a week. That is what Ozempic, Wegovy, Mounjaro, and Zepbound are: a normal gut hormone, rebuilt to last. They work, they reduce heart attacks and kidney decline in the people studied, and they carry real dangers that are widely under-explained, including one that has killed people: they are not, and cannot be, a replacement for insulin.

Key takeaways

• GLP-1 means "glucagon-like peptide-1." It is a short protein made in the gut whose sequence resembles the hormone glucagon, and it was the first of two such peptides found in that gene. The name describes its chemistry, not its job.
• It works on five things at once: insulin release, glucagon suppression, stomach emptying, brain appetite centres, and food reward. The weight loss comes mostly from the last three.
• It cannot replace insulin. Its insulin effect requires working beta cells. In type 1 diabetes there are none, so a person who reduces or stops insulin because a GLP-1 drug is "handling the sugar" can be in life-threatening ketoacidosis within a day.
• The vomiting is not a nuisance to push through. Severe or persistent vomiting causes dehydration, and dehydration in someone with existing heart or artery disease can precipitate a cardiac event or acute kidney injury. Persistent vomiting means stop and seek advice, not persevere.
• Around 10 deaths and 100 hospitalisations in the US have been linked to compounded semaglutide, mostly from dosing errors of ten-fold or more when people drew doses from vials themselves.
• At the same time, the outcome trials are strongly positive: semaglutide cut major cardiovascular events by about 20 percent in people with obesity and established heart disease. Both things are true.

What GLP-1 actually is, and why it is called that

In short: A gut hormone named for resembling glucagon, destroyed in two minutes by an enzyme, and turned into a weekly drug by borrowing a trick from a venomous lizard.

Decoding the name

GLP-1 = Glucagon-Like Peptide-1.

• Peptide: a short chain of amino acids. Proteins are long chains; peptides are short ones. GLP-1 is 30 or 31 amino acids long. This matters practically, because peptides are digested by the stomach like any other protein, which is why almost all of these drugs are injected.
• Glucagon-like: its amino acid sequence closely resembles glucagon, the hormone that raises blood sugar. They resemble each other because they are cut from the same original protein. A single gene, proglucagon, produces a long precursor that different tissues chop up in different places using different enzymes. Pancreatic alpha cells cut it to release glucagon. Intestinal L-cells cut the same precursor at different points and release GLP-1 and GLP-2 instead.
• -1: because two glucagon-like peptides came out of that sequence, and this was the first. GLP-2 exists and does something else entirely (it promotes growth of the intestinal lining, and a GLP-2 analogue is used to treat short bowel syndrome).

So the name is a statement about molecular ancestry, not function. GLP-1 and glucagon are molecular siblings with nearly opposite jobs: glucagon raises blood glucose, GLP-1 helps lower it.

The observation that started it: the incretin effect

Swallow a measured amount of glucose, and your pancreas releases a certain amount of insulin. Now inject exactly enough glucose into a vein to produce the identical blood glucose curve. The insulin response is much smaller, by roughly half to two-thirds.

The blood glucose is the same in both cases, so glucose alone cannot explain the difference. Something else is telling the pancreas to release insulin, and that something is triggered by food passing through the gut rather than by sugar in the blood. This gap is called the incretin effect, and the hormones responsible are incretins: substances from the intestine that increase insulin secretion.

Two of them matter:

The system exists because it is a warning shot. Food in the gut has not been absorbed yet, so signalling the pancreas early means insulin is already circulating by the time the glucose arrives, which blunts the peak. It is anticipatory control rather than reactive control.

In type 2 diabetes, the incretin effect is measurably reduced, which was the observation that made these hormones a drug target in the first place.

Why it took forty years to turn into a medicine

GLP-1 was identified in the early 1980s and its glucose-lowering effect confirmed soon after. The problem was its half-life of roughly one to two minutes.

An enzyme called DPP-4 (dipeptidyl peptidase-4, meaning it clips off amino acids two at a time from the end of a peptide) circulates in blood and on cell surfaces and destroys GLP-1 almost immediately. Physiologically this is sensible: a meal signal should not persist for hours. Pharmacologically it is fatal to the idea of a drug. An infusion of natural GLP-1 works beautifully and stops working the moment you switch off the pump.

Two routes around it were taken, and both produced drug classes still in use:

1. Stop the enzyme. DPP-4 inhibitors (sitagliptin, linagliptin, and the other "gliptins") block the destroyer, so your own GLP-1 lasts a few times longer. This raises natural GLP-1 to roughly twice its normal level, which lowers HbA1c by about half a point. Modest, because you are limited to what your gut produces.
2. Build a version the enzyme cannot cut. This is the route that produced Ozempic, and it started with a lizard.

The Gila monster

In the early 1990s, John Eng, a researcher at a Veterans Affairs hospital in New York, was following up an old observation that the venom of certain reptiles caused pancreatic enlargement. Working on venom from the Gila monster (Heloderma suspectum), a venomous lizard of the southwestern United States that eats a few large meals a year, he isolated a peptide he named exendin-4.

Exendin-4 turned out to be about 53 percent identical to human GLP-1, enough to activate the human GLP-1 receptor, and different in exactly the right place: it lacks the amino acid pair that DPP-4 cuts. Its half-life in humans is around 2.4 hours rather than 2 minutes.

Synthetic exendin-4 became exenatide, approved in 2005, the first drug of the class. It had to be injected twice a day, which limited it, and it proved the concept.

How semaglutide was engineered to last a week

Semaglutide is human GLP-1 with three deliberate modifications, and each solves a specific problem:

The result is a half-life of about one week, which is why it is a weekly injection. That is the whole trick: not a new mechanism, but the same natural hormone made to survive.

Oral semaglutide (Rybelsus) solves a different problem. Peptides are digested in the stomach, so it is co-formulated with an absorption enhancer (SNAC) that locally raises the pH and helps a small fraction of the dose cross the stomach lining intact. Only about 1 percent is absorbed, which is why the oral dose is far larger than the injected one and why it must be taken on an empty stomach with a small sip of water and nothing else for 30 minutes.

Which drug is which

In short: Ozempic and Wegovy are the same molecule at different doses, and tirzepatide is a genuinely different drug that hits two receptors.

The naming causes genuine confusion, because the same molecule is sold under different brand names for different conditions at different doses.

Ozempic and Wegovy are the same drug. Semaglutide, approved at up to 2.0 mg weekly for type 2 diabetes as Ozempic and at 2.4 mg weekly for obesity as Wegovy. The difference in the box is the licensed indication and the maximum dose, not the molecule. Much of the confusion in public discussion, including about who is "taking a diabetes drug for weight loss," comes from this one fact not being stated.

Tirzepatide is a different thing. It activates the GIP receptor as well as the GLP-1 receptor, which is why it is called a dual agonist or twincretin. GIP's contribution is still being worked out, and it appears to add appetite and metabolic effects beyond GLP-1 alone, plus possibly less nausea per unit of weight loss. It produces the largest weight loss of any approved drug.

How it actually works: five mechanisms

In short: Insulin release, glucagon suppression, a slowed stomach, brain appetite circuits, and food reward, and the weight loss comes mostly from the last three.

This is the part usually compressed into "it makes you feel full," which is a third of the story.

1. It tells the pancreas to release insulin, but only when glucose is high

GLP-1 binds its receptor on pancreatic beta cells and amplifies the insulin response to glucose. The critical word is glucose-dependent. The signalling pathway it activates only produces insulin release when glucose is already entering the cell and raising its metabolic activity. At normal or low blood glucose, the effect switches off.

This is why GLP-1 drugs alone almost never cause hypoglycaemia, unlike sulfonylureas, which force insulin release regardless of glucose level and can drive it dangerously low. It is a genuinely elegant property: a drug with a built-in off switch.

It is also the property that creates the danger described later in this chapter, because amplifying a beta cell's response requires beta cells to exist.

2. It suppresses glucagon

Glucagon tells the liver to pour glucose into the blood. In type 2 diabetes, glucagon is inappropriately high after meals, which adds to the post-meal glucose rise. GLP-1 suppresses it, also in a glucose-dependent way, so the suppression lifts if blood sugar falls.

3. It slows the stomach

GLP-1 slows gastric emptying, the rate at which the stomach releases its contents into the intestine. Three consequences follow, and all three matter:

• Glucose enters the bloodstream more gradually, so the post-meal peak is lower.
• The stomach stays fuller for longer, which produces satiety and reduces how much is eaten at the next meal.
• Everything that goes wrong with these drugs in the gut follows from this. Nausea, vomiting, bloating, reflux, constipation, the rare cases of severe gastroparesis, and the risk during anaesthesia are all the intended mechanism, operating more strongly than intended.

4. It acts directly on the brain's appetite circuits

This is the mechanism that produces most of the weight loss, and it is why these drugs work where diets fail.

GLP-1 receptors are present in the hypothalamus (the arcuate nucleus, which houses the appetite-regulating neurons described in Chapter 19) and in the hindbrain, particularly the area postrema, a region deliberately left outside the blood-brain barrier so it can sample circulating signals. Long-acting GLP-1 drugs reach these sites and activate the satiety-promoting neurons while suppressing the hunger-promoting ones.

The subjective result is what patients describe as the disappearance of food noise: the continuous background negotiation about what and when to eat simply quietens. That is a description of a regulatory set point being lowered, which is why the weight loss is not experienced as deprivation and why it is not the same as willpower.

It is also why the weight comes back when the drug stops. Nothing has been permanently reset; the signal has been supplied from outside, and when it is withdrawn the original regulation resumes (Chapter 19 explains why the body defends its highest weight).

5. It reduces the reward value of food

Imaging and behavioural studies show reduced activation in reward-related brain regions in response to food cues. This is the most likely explanation for the reports of reduced craving for alcohol and nicotine in people taking these drugs, which is now being formally tested (Chapter 43).

What it achieves

In short: Substantial weight loss, plus proven reductions in cardiovascular events, kidney decline, and sleep apnoea severity.

This is a genuinely impressive list, and it is the reason the safety section that follows should be read as a description of trade-offs rather than as a case against the drugs.

The dangers, in detail

In short: The insulin-replacement error, the vomiting and dehydration chain, unregulated compounded products, and the anaesthesia risk, in that order of importance.

The one that kills people: this is not a replacement for insulin

This is the most important paragraph in the chapter.

A GLP-1 drug lowers blood glucose partly by amplifying the beta cell's insulin response. In type 1 diabetes, the beta cells have been destroyed (Chapter 18). There is nothing to amplify. The drug's other effects (slowed stomach, reduced appetite, lower glucagon) will still lower glucose readings somewhat, which is exactly what makes the situation dangerous: the numbers look acceptable while the fundamental problem is untreated.

Without insulin, cells cannot take up glucose regardless of how much is in the blood. The body reads this as starvation, breaks down fat for fuel, and produces acidic ketones. Blood turns acidic. This is diabetic ketoacidosis, it develops over hours to a day, and it is fatal without emergency treatment.

The same logic applies more subtly in long-standing type 2 diabetes, where beta cell function has declined far enough that insulin is required. Someone in that position who eats much less because of appetite suppression, and who therefore reduces their insulin, can also develop ketoacidosis.

The specific chain that has caused harm:

1. The drug suppresses appetite, so the person eats far less.
2. Eating less makes blood glucose readings fall, and they reduce or skip insulin, sometimes on their own initiative, sometimes because nobody explained the distinction.
3. Insulin does more than dispose of glucose: it is the signal that suppresses fat breakdown and ketone production. Without it, ketones rise even when glucose is not very high.
4. Nausea and vomiting from the drug look like an expected side effect, and they are also the early symptoms of ketoacidosis. The two are indistinguishable without a ketone test.
5. Vomiting means less fluid intake and more fluid loss, accelerating everything.

The practical rules that follow: GLP-1 drugs are used in type 1 diabetes only as an addition to insulin, under specialist supervision, never as a substitute. Reassuringly, large recent studies of GLP-1 drugs used as add-on therapy alongside insulin in type 1 diabetes have not found an increase in ketoacidosis, which supports the supervised add-on use and says nothing about replacement. Anyone on insulin who starts one of these drugs needs an explicit insulin adjustment plan and a way to test ketones at home, and should be told that persistent vomiting requires a ketone check rather than a wait.

The vomiting, and why it is not simply an inconvenience

Nausea affects a large minority to a majority of users. In the STEP 1 trial of semaglutide 2.4 mg, roughly 44 percent reported nausea and about 25 percent vomiting, against 17 and 7 percent on placebo. It is dose-related, worst after each dose increase, and usually improves over weeks.

For most people it is unpleasant and manageable. The danger is in the tail of the distribution, and the mechanism is worth spelling out because it is rarely explained:

Severe or persistent vomiting causes volume depletion. You lose water, sodium, potassium, and chloride, and you stop drinking because you cannot keep fluids down. The consequences:

• Blood volume falls, so blood pressure falls and heart rate rises to compensate.
• In a person with narrowed coronary arteries, the combination of a faster heart (which needs more oxygen) and lower perfusion pressure (which delivers less) can push a stable narrowing into ischaemia. This is the mechanism by which severe dehydration from any cause, gastroenteritis included, can precipitate a heart attack in someone with existing coronary disease. It is not unique to these drugs and it is a real pathway.
• Blood becomes more concentrated and more prone to clotting, adding to that risk.
• The kidneys are underperfused, causing acute kidney injury, and this is worse in anyone taking ACE inhibitors, ARBs, diuretics, NSAIDs, or SGLT2 inhibitors, which is a very large proportion of the people prescribed these drugs.
• Potassium and magnesium losses can provoke cardiac arrhythmias.

So the correct advice is the opposite of "push through it." Persistent vomiting on a GLP-1 drug is a reason to stop, rehydrate, hold the interacting medicines under advice, and be assessed. It is also a reason for the dose to be escalated slowly rather than quickly, which is the single most effective way to avoid the whole problem.

Don't be confused: expected nausea and dangerous vomiting are different situations. Feeling queasy for a day or two after a dose increase, eating smaller meals, and finding fatty food unappealing is the drug working as designed. Vomiting repeatedly, being unable to keep fluids down for more than a few hours, severe abdominal pain, not passing urine, or feeling faint on standing is not a side effect to tolerate. The first needs patience. The second needs a clinician the same day.

Deaths and injuries from compounded and counterfeit products

This is where documented deaths cluster, and the cause is not the molecule.

During the shortages of 2023 and 2024, US regulations permitted pharmacies to compound their own semaglutide, and a large market of online clinics, medical spas, and unregulated sellers appeared. Novo Nordisk reported roughly 10 deaths and about 100 hospitalisations linked to compounded semaglutide in FDA adverse event data, and the FDA issued repeated warnings about dosing errors.

The mechanism of harm was mostly arithmetic:

• Branded pens deliver a fixed, pre-set dose. Compounded product often came as a vial plus a syringe, requiring the patient to calculate and draw the dose themselves.
• Doses were sometimes expressed in units (as on an insulin syringe) and sometimes in milligrams, and the two were confused, producing overdoses of five to twenty times the intended amount.
• Some products used semaglutide salts (sodium or acetate forms) that are not the approved active ingredient and have not been tested for safety or potency.
• Some counterfeit pens seized internationally contained insulin instead of semaglutide, which is a direct route to severe hypoglycaemia.

A ten-fold semaglutide overdose produces intractable vomiting for days, severe dehydration, and in some reported cases multi-organ failure. There is no antidote and the drug takes weeks to clear, so treatment is supportive.

The FDA declared the shortages resolved in December 2024 (tirzepatide) and February 2025 (semaglutide) and has moved to prohibit compounding of these drugs. The practical rule for anyone reading this: obtain these drugs through a regulated prescription and pharmacy, in manufacturer-labelled pens, and never draw a dose from a vial you were told to measure yourself.

Gastroparesis, ileus, and surgery

Slowed gastric emptying is the mechanism, and in a small number of people it becomes severe and persistent: gastroparesis, in which the stomach empties so slowly that food is retained for many hours, causing vomiting of undigested food, bloating, and weight loss beyond what is wanted. Pharmacovigilance analyses found a clear disproportionate reporting signal, with an absolute incidence on the order of 1 case per 1,000 person-years. Ileus (the intestine stopping altogether) was added to the labels in 2023.

The related and more immediately dangerous problem is anaesthesia. Patients are fasted before surgery so the stomach is empty, because an anaesthetised person loses the reflexes that protect the airway, and stomach contents entering the lungs (aspiration) causes severe pneumonitis and can be fatal. Endoscopy studies found substantial residual stomach contents in patients on these drugs despite standard fasting, and pulmonary aspiration during general anaesthesia was added to the Ozempic label in November 2024.

Practical rule: tell every anaesthetist, surgeon, endoscopist, and dentist that you are on one of these drugs. Current professional guidance generally involves holding the drug before elective procedures (typically the weekly dose skipped the week before), extending clear fluid fasting, and in some cases ultrasound assessment of stomach contents or treating the patient as having a full stomach regardless of fasting time.

The rest of the risk profile

Stopping, and what happens next

In the STEP 1 trial extension, participants who stopped semaglutide regained about two-thirds of the weight they had lost within a year, and their cardiometabolic improvements reversed in proportion.

This is not a failure of the drug and it is not usually explained properly before starting. The drug lowers the body's defended weight while it is present. Withdraw it and the defence resumes. Framed correctly, these are long-term treatments for a chronic condition, in the same category as antihypertensives, not a course of treatment with an end date. Anyone starting one should decide with that in mind, because it has cost, supply, and commitment implications over years.

Who should not take them

• Personal or family history of medullary thyroid carcinoma or MEN 2 syndrome.
• Previous pancreatitis (relative, and requires a considered discussion).
• Pregnancy, breastfeeding, or planning pregnancy in the near term.
• Severe gastroparesis or significant existing gastrointestinal motility disease.
• Type 1 diabetes as a substitute for insulin, under any circumstances.
• Active severe eating disorder, where appetite suppression can be dangerous.

And a category rather than a contraindication: anyone who has not been told what to do when they vomit, what to do before surgery, how their insulin or sulfonylurea dose changes, and that stopping means regaining. Those four conversations are where most of the avoidable harm sits.

What the person taking one can do

In short: Escalate slowly, drink deliberately, have sick-day rules, tell every clinician before any procedure, and buy only from a regulated pharmacy.

• Escalate the dose slowly. Most of the nausea comes from going up too fast. If a step is badly tolerated, staying at the lower dose longer is usually better than pushing on.
• Eat smaller, slower, less fatty meals, and stop at the first sense of fullness. The stomach genuinely holds more than it can process.
• Drink deliberately, since reduced appetite reduces fluid intake too, and dehydration is the pathway to most of the serious complications.
• Have sick-day rules. If you cannot keep fluids down for more than a few hours: stop the drug, seek advice, and if you are on insulin, check ketones. Ask in advance which of your other medicines (ACE inhibitors, ARBs, diuretics, NSAIDs, SGLT2 inhibitors, metformin) should be paused during vomiting.
• Protect muscle: resistance training two or three times a week and roughly 1.2 to 1.6 g of protein per kg of body weight per day, adjusted for kidney function.
• Tell every clinician, especially before any procedure involving sedation.
• If you have diabetes, agree an insulin or sulfonylurea reduction plan before the first dose, not after the first hypoglycaemic episode.
• Buy only from a regulated pharmacy with a manufacturer-labelled pen, and never self-calculate a dose from a vial.
• Report sudden vision loss, severe persistent abdominal pain, or inability to keep fluids down as emergencies rather than side effects.

The honest summary

In short: Almost all documented deaths trace to a dosing error from an unregulated supply or to someone taking this instead of insulin, and both are preventable with information.

These drugs do something that nothing previously did: they lower the body's defended weight rather than fighting it, and in doing so they improve outcomes that matter, not just numbers on a chart. A 20 percent reduction in cardiovascular events is a serious result.

They are also powerful drugs acting on the gut, the pancreas, and the brain simultaneously, prescribed at enormous scale, frequently through channels with minimal follow-up, to people who are often not told the four things that matter: what to do when they vomit, what happens before surgery, how insulin doses change, and what happens when they stop.

Almost all of the documented deaths trace to one of two failures, and neither is intrinsic to the molecule: a dosing error from an unregulated supply, or a person on insulin taking this instead of it. Both are preventable with information, which is the entire argument for explaining a drug rather than just naming it.

Sources and notes

Incretin effect and the physiology of GLP-1 and GIP: standard endocrinology references and Drucker's reviews in Cell Metabolism and Nature Reviews Endocrinology. Exendin-4 isolation: Eng et al., Journal of Biological Chemistry, 1992. Semaglutide molecular design (Aib8 substitution, C18 diacid albumin binding): Lau et al., Journal of Medicinal Chemistry, 2015. STEP 1 weight loss and adverse event rates: Wilding et al., NEJM, 2021. STEP 1 extension weight regain: Wilding et al., Diabetes, Obesity and Metabolism, 2022 (roughly two-thirds of lost weight regained one year after withdrawal). SURMOUNT-1: Jastreboff et al., NEJM, 2022. SELECT: Lincoff et al., NEJM, 2023 (17,604 participants, about 20 percent MACE reduction). FLOW: NEJM, 2024. Gastroparesis and ileus signals: FDA label updates (Ozempic, September 2023) and pharmacovigilance analyses; absolute incidence approximately 1 per 1,000 person-years. Pulmonary aspiration label update: November 2024; residual gastric content studies and American Society of Anesthesiologists guidance, 2023 onward. Severe pancreatitis label update: January 2025; randomised trial meta-analyses do not show a significant excess. NAION: cohort analyses in type 2 diabetes reporting 0.04 percent versus 0.02 percent, plus the 2024 Harvard case series; American Academy of Ophthalmology and NANOS advisory. Compounded semaglutide deaths and hospitalisations: FDA adverse event reporting data as reported by Novo Nordisk, 2024, and FDA compounding statements; shortage resolutions December 2024 and February 2025. Suicidality review: European Medicines Agency PRAC conclusion, 2024. GLP-1 as add-on in type 1 diabetes: target trial emulation and meta-analyses published 2025 to 2026 finding no increase in ketoacidosis with supervised add-on use.

Open questions. Long-term safety across decades of continuous use is unknown, since the drugs have been used at scale for only a few years. Whether the cardiovascular benefit is mediated by weight loss, by direct vascular effects, or by both is unresolved. Whether the lean-mass loss has long-term consequences, particularly in older patients, is being studied. Whether these drugs genuinely reduce addictive behaviour is not established.

Next: how to read a drug name, a lab result, or a medical term you have never seen before. 👉