Liraglutide

A modified version of the human gut hormone GLP-1 that lasts about a day. In the EU it is sold as Victoza for type 2 diabetes and as Saxenda for weight management, and it was the first GLP-1 medicine approved for weight management in the EU.

Sold as

Injection under the skin.

Legal status in the EU

Victoza was approved by the European regulator on 30 June 2009 and Saxenda on 23 March 2015 (Novo Nordisk A/S). Prescription only. Victoza is approved for type 2 diabetes from age 10; Saxenda for weight management in adults, adolescents from 12 and children from 6. Liraglutide does not appear on the 2026 WADA list.

Liraglutide switches on the GLP-1 receptor, which increases insulin release, reduces the hormone that raises blood sugar, and increases the feeling of fullness. One amino acid is swapped at position 34 and a fatty acid is attached at position 26, which makes it stick to a blood protein and last far longer than natural GLP-1, which is broken down within minutes.

Part of the bodyWhat happensHow it is knownKind of evidence

Appetite, in the brain

This is the part people notice first and the part the whole drug rests on. Hunger drops, the feeling of being full arrives earlier and lasts longer, and the constant background thought about the next meal quietens. When people were then given a free buffet and told to eat as much as they wanted, they ate about 16% less than on a dummy injection. The body does not burn more. The European product information states plainly that liraglutide does not raise the amount of energy the body burns. In the buffet study, 24-hour energy burn on the higher amount was about 5% lower, in line with the smaller body.

Measured in 49 adults with obesity and no diabetes, of whom 44 finished. Each person took two of the three treatments — the drug at one of two amounts, or a dummy injection — for 5 weeks each. The appetite ratings and the buffet meal were done under supervision. That there is no rise in energy burn is stated in the European product information for Saxenda.

Measured in people

Source [07]Source [08]

Stomach and gut

The stomach empties more slowly, and for most people that shows up as feeling sick. Nausea is the single most common effect by a wide margin, followed by loose stools, vomiting and constipation — and constipation and diarrhoea can both happen in the same person at different points. It starts in the first weeks and mostly settles within days or weeks of carrying on. It is also what most often ends treatment: 9.8% of people stopped because of a side effect against 4.3% on dummy injections, with nausea and vomiting the top two reasons.

Some gut reaction was recorded in 67.9% of the 5,813 adults across the five trials behind the European approval. In the pooled US safety analysis of 3,384 adults on liraglutide against 1,941 on dummy injections: nausea 39.3% against 13.8%, loose stools 20.9% against 9.9%, constipation 19.4% against 8.5%, vomiting 15.7% against 3.9%. Stomach emptying measured directly in 49 adults was 23% slower one hour after a meal on the higher amount, but no different over five hours.

Measured in people

Source [07]Source [08]Source [09]

Body fat, and where it sits

Weight comes off mainly as fat, and the fat around the organs inside the belly goes faster than the fat under the skin. That shows up as a waist that shrinks more than the scales alone would predict. It is also the part of the effect that reverses fastest once the injections stop.

Waist measured in 3,662 adults over 56 weeks: down 8.2 cm on liraglutide against 4.0 cm on dummy injections, a difference of 4.2 cm. That weight loss is mainly fat, and that the fat inside the belly falls faster than the fat under the skin, is stated in the European product information for Saxenda.

Measured in people

Source [08]

Muscle and bone

Losing weight quickly takes some muscle and some bone with it, and liraglutide on its own does not protect against that. In the one trial designed to look, bone density in the hip and the lower spine fell more in the group given liraglutide alone than in the group given supervised exercise instead. The exercise group had lost less weight over the year — 11.2 kg against 13.7 kg — and still kept more bone. Adding exercise to the drug kept bone density where the dummy group's was. In a much smaller study, the absolute amount of lean tissue in the trunk and limbs fell, while lean tissue as a share of body weight held steady.

Bone density scanned in 195 adults with obesity, randomised to exercise, liraglutide, both or dummy injections, over 52 weeks after an 8-week low-calorie diet. Hip bone density fell 0.026 g/cm² on liraglutide alone (95% CI 0.033 to 0.018) against 0.012 on dummy injections (95% CI 0.020 to 0.004); lower-spine density fell 0.020 against 0.001. Lean tissue measured by scan in 20 adults with obesity over 35 days.

Measured in people

Source [10]Source [11]

Blood sugar, and the pancreas

The drug tells the pancreas to release insulin only when blood sugar is already high, and to release less of glucagon, the hormone that pushes blood sugar up. Because the insulin release is tied to the sugar level, liraglutide on its own almost never drives blood sugar dangerously low. It does when it is combined with insulin or with the older sulfonylurea tablets, which push insulin out regardless. The pancreas can also become inflamed, which is uncommon but is the reason the drug is stopped and never restarted if it happens.

Long-term blood sugar measured in 3,662 adults over 56 weeks fell 0.3 percentage points against 0.1 on dummy injections; in 623 adults who already had type 2 diabetes it fell 1.3 points against 0.4. Low-blood-sugar episodes needing another person's help were reported by 0.7% of adults with type 2 diabetes on the drug, and only in those also taking a sulfonylurea. Inflamed pancreas confirmed by an independent review in 9 of 3,291 people on the drug (0.3%) against 2 of 1,843 on dummy injections (0.1%); in the 9,340-person diabetes trial the rates were 0.4% and 0.5%.

Measured in people

Source [08]Source [09]Source [12]

Heart and blood vessels

Two things happen at once and they point in opposite directions. Blood pressure falls. The resting heart rate rises, by a small average amount, and peaks about six weeks in; it goes back to where it was when the drug is stopped. Set against that, in people with type 2 diabetes and existing heart disease, fewer had a heart attack, a stroke or died of a heart problem over nearly four years.

The top blood pressure number fell 4.3 mmHg against 1.5 on dummy injections, measured in 3,662 adults over 56 weeks. Across the phase 3 trials the average heart rate rise was 2.5 beats a minute, ranging from 1.6 to 3.6. When heart rate was tracked continuously for 24 hours it ran 4 to 9 beats a minute above the dummy group. Heart outcomes were measured in 9,340 adults with type 2 diabetes over a median of 3.8 years.

Measured in people

Source [08]Source [09]Source [13]

SCALE Obesity and Prediabetes (NEJM 2015, 3,731 participants, 56 weeks) gave −8.4 kg versus −2.8 kg for placebo (difference −5.6 kg; 95% CI −6.0 to −5.1); 63.2% versus 27.1% lost at least 5% of their weight, and 33.1% versus 10.6% lost more than 10%. The European regulator's assessment of Saxenda rests on five studies in more than 5,800 adults and reports 7.5% weight loss versus 2.3% for placebo. The approvals for children rest on one study in 251 adolescents and one in 82 children.

Fast on, fast off, with one long tail. Appetite changes within the first weeks: in the 49-adult study the effect on fullness, hunger and how much people ate was already there at five weeks. The gut reactions arrive in the same first weeks and usually fade within days or weeks of carrying on. Sleeplessness and dizziness cluster in the first three months. Resting heart rate climbs to its peak at about six weeks and then holds. Twelve weeks is the point the trials treat as decisive. In trial 1, 67.5% had lost at least 5% of their weight by then. Of the people who had not, 93.4% never reached 10% by the end of the year. Weight loss itself happens mostly in the first year and then flattens: the 160-week figure, down 6.2%, is smaller than the 56-week figure, down 8.0%, not larger. Coming off is quicker than going on. Across six trials of this class in 3,236 people, about 60% of the lost weight was back within a year of stopping. Regain levelled off at about three quarters of what had been lost, at a half-life of 23 weeks. Blood pressure came back faster: 70 to 80% of the fall in the top number returned within 12 weeks, and about half the improvement in long-term blood sugar within 8 to 12 weeks. In the adolescent trial, 26 weeks off the drug undid most of the BMI change built up over the previous year. The heart rate rise, at least, reverses when the drug is stopped.

What was measuredWhat happenedIn whomHow it is known

Body weight in adults with obesity and no diabetes

Down 8.4 kg against 2.8 kg on dummy injections, a difference of 5.6 kg (95% CI 6.0 to 5.1). 63.2% lost at least 5% of their weight against 27.1%, and 33.1% lost more than 10% against 10.6%. The regulator's own table, which analyses 2,437 people on the drug and 1,225 on dummy rather than everyone randomised, gives 63.5% against 26.6% and 32.8% against 10.1% — the same trial, a slightly different set of people counted.

3,731 adults with obesity, or overweight plus a blood-fat or blood-pressure problem, randomised two to one, 56 weeks; everyone also got diet and exercise counselling

Measured in people

Source [01]Source [08]

Whether the weight loss held over three years

It shrank but did not vanish. At week 160 the drug group was 6.2% lighter than at the start against 1.8% for the dummy group, a difference of 4.3 percentage points. 49.6% were still at least 5% down against 23.4%. Most of the loss happened in the first year and then held roughly flat.

2,254 adults who had prediabetes at screening, of whom 1,472 on the drug and 738 on dummy injections were analysed at week 160; 1,128 completed

Measured in people

Source [08]

Going on to develop type 2 diabetes

3% of people on the drug were diagnosed with type 2 diabetes by week 160 against 11% on dummy injections. Time to diagnosis was 2.7 times longer (95% CI 1.9 to 3.9). Among those who had prediabetes at the start, 69.2% no longer did at 56 weeks against 32.7%.

2,254 adults with prediabetes, 160 weeks, inside the same trial

Measured in people

Source [08]

Heart attacks, strokes and death in people with type 2 diabetes

A heart attack, stroke or death from a heart cause happened to 13.0% of people on liraglutide against 14.9% on dummy injections (hazard ratio 0.87, 95% CI 0.78 to 0.97). Death from a heart cause: 4.7% against 6.0% (hazard ratio 0.78, 95% CI 0.66 to 0.93). Death from any cause: 8.2% against 9.6% (hazard ratio 0.85, 95% CI 0.74 to 0.97).

9,340 adults with type 2 diabetes that was not well controlled, most of them with existing heart disease, median follow-up 3.8 years

Measured in people

Source [08]Source [13]

Kidneys and eyes in that same heart trial

Kidney damage came less often on the drug (hazard ratio 0.78, 95% CI 0.67 to 0.92). Damage to the back of the eye came slightly more often, but the range crossed no difference and so answered nothing (hazard ratio 1.15, 95% CI 0.87 to 1.52).

The same 9,340 adults with type 2 diabetes

Measured in people

Source [12]

Body weight in adults who already had type 2 diabetes

Down 5.9% against 2.0% on dummy injections. 49.8% lost at least 5% against 13.5%. Long-term blood sugar fell 1.3 percentage points against 0.4. The weight loss is smaller than in people without diabetes, and that gap is consistent across this class.

846 adults with obesity or overweight and type 2 diabetes that was not well controlled, 412 on the drug and 211 on dummy injections analysed, 56 weeks

Measured in people

Source [08]

Breathing interruptions during sleep

Episodes of stopped or shallow breathing fell by 12.2 an hour against 6.1 an hour on dummy injections, a difference of 6.1 (95% CI 11.0 to 1.2). Weight fell 5.7% against 1.6%. People started the trial with about 49 episodes an hour, so both groups were still well inside the severe range at the end.

359 adults with obesity and moderate or severe obstructive sleep apnoea, 32 weeks; 276 completed

Measured in people

Source [08]

Keeping off weight already lost by dieting

After everyone had first lost at least 5% on a low-calorie diet, the drug group lost a further 6.3% against 0.2% on dummy injections. 81.4% held on to the weight they had already lost, against 48.9%.

422 adults with obesity or overweight plus high blood pressure or a blood-fat problem, 207 on the drug and 206 on dummy injections, 56 weeks; 305 completed

Measured in people

Source [08]

Head to head against semaglutide

Liraglutide lost 6.4% of body weight (95% CI 8.2 to 4.6) against 15.8% for weekly semaglutide (95% CI 17.6 to 13.9). At least 10% lost: 25.6% against 70.9%. At least 15%: 12.0% against 55.6%. At least 20%: 6.0% against 38.5%. Gut side effects were near identical (82.7% against 84.1%), but twice as many people abandoned liraglutide — 27.6% against 13.5%.

338 adults with obesity and no diabetes, randomised to weekly semaglutide, daily liraglutide or a dummy, 68 weeks

Measured in people

Source [14]

What happened in the year after people stopped

The group who had been on liraglutide alone regained 9.6 kg in the twelve months after stopping. That is 6.0 kg more than the group who had been doing supervised exercise instead (95% CI 2.1 to 10.0). Two years after the trial began they were 8.7 kg heavier than at the point the injections started, which was itself after an 8-week low-calorie diet.

109 of the original 195 adults with obesity came back for the follow-up one year after all treatment stopped

Measured in people

Source [15]

How much weight comes back across this whole class of drugs

About 60% of the weight lost during treatment was back one year after stopping. Regain levelled off at about 75.3% of what had been lost (95% CI 68.9 to 81.6), so roughly a quarter of the loss looks like it stays. The pace works out at a half-life of 23 weeks (95% CI 17.3 to 34.3). Blood pressure went back faster than weight: 70 to 80% of the fall in the top number returned within 12 weeks.

6 randomised trials pooled, 3,236 participants, covering liraglutide, semaglutide and tirzepatide; the liraglutide trial included was SCALE Obesity

Measured in people

Source [16]

Body weight in adolescents

Body weight fell 2.65% on the drug and rose 2.37% on dummy injections, a difference of 5.01 percentage points (95% CI 7.63 to 2.39). 43.25% cut their BMI by at least 5% against 18.73%. In the 26 weeks after the injections stopped the drug group's BMI score climbed back by 0.22 against 0.07 — most of what had been gained in the year before.

251 adolescents aged 12 and over with obesity, 125 on the drug and 126 on dummy injections, 56 weeks, then 26 weeks off it

Measured in people

Source [08]

Body weight in children aged 6 to 11

BMI fell 5.80% on the drug and rose 1.60% on dummy injections, a difference of 7.40 percentage points (95% CI 11.56 to 3.24). 46.2% cut their BMI by at least 5% against 8.7%. The regulator records more gut trouble in this age group than in adolescents or adults, with vomiting about twice as common as in adolescents.

82 children aged 6 to under 12 with obesity, 56 on the drug and 26 on dummy injections, randomised two to one, 56 weeks

Measured in people

Source [08]

Body weight in children and teenagers with Prader-Willi syndrome

No difference worth the name. In the older group the BMI score fell 0.20 on the drug against 0.13 on dummy injections at 16 weeks; in the younger group it fell 0.50 against 0.44. At 52 weeks the falls were 0.31 against 0.17, and 0.73 against 0.67. The regulator records these as similar with the drug and with dummy injections.

56 patients with Prader-Willi syndrome and obesity — 32 aged 12 to under 18 and 24 aged 6 to under 12 — randomised two to one, 16 weeks blinded followed by 36 weeks open

Measured in people

Source [08]

How long people actually stay on it outside a trial

16.6% were still taking liraglutide 12 months after starting, against 24.0% for phentermine/topiramate tablets and 28.8% for orlistat tablets. Average time on it was 114.6 days. Weight at six months was down 4.24%, against 4.29% for orlistat and 5.40% for phentermine/topiramate in the same hospital. That is about half the drop the trials reported at a year.

1,910 courses of weight medicine at one large hospital in Seoul between 2018 and 2025, of which 628 were liraglutide

Measured in people

Source [17]

Nausea, vomiting, diarrhoea and constipation affect more than 1 in 10 and usually fade after a few weeks. In SCALE, serious side effects were reported in 6.2% versus 5.0% for placebo. Gallbladder disease and pancreatitis are known risks across the whole GLP-1 class.

EffectWhat it feels likeHow often

Inflammation of the pancreasSerious

Severe pain high in the belly that bores through to the back, often with vomiting. The European product information instructs that the drug is stopped if it is suspected and never restarted if it is confirmed. The rarer form in which pancreas tissue dies has been reported since the drug went on sale.

Uncommon — fewer than 1 in 100. Confirmed by independent review in 9 of 3,291 people on the drug (0.3%) against 2 of 1,843 on dummy injections (0.1%). In the 9,340-person diabetes trial, 0.4% against 0.5%.

Source [09]Source [12]

Gallstones and an inflamed gallbladderSerious

Losing a lot of weight quickly causes gallstones by itself, and the European regulator's position is that this only partly explains the higher rate on the drug. Stones can lead to hospital admission and to the gallbladder being removed. The signs are pain under the right ribs, fever and yellowing of the eyes.

Gallstones in 2.2% of people on the drug against 0.8% on dummy injections; inflamed gallbladder in 0.8% against 0.4%. Across 55 randomised trials of this whole class covering 106,395 people, the risk of gallstones was 1.46 times higher (95% CI 1.09 to 1.97) — about 2 extra cases per 1,000 people.

Source [08]Source [09]Source [18]

Kidneys failing after being dried out by vomiting or diarrhoeaSerious

This is a knock-on effect, not a direct one: days of vomiting or loose stools empty the body of fluid, and the kidneys fail because of it. Most reported cases followed exactly that sequence. Acute kidney injury is the 23rd most reported term in the US regulator's public side-effect file for liraglutide, with 884 reports.

Rare in the trials — fewer than 1 in 1,000. Reported since the drug went on sale without a rate, some cases needing dialysis.

Source [08]Source [19]

A whole-body allergic reactionSerious

Blood pressure drops, the heart races, breathing becomes hard and the face or throat swells. The European product information calls this potentially life-threatening and says the drug must not be restarted afterwards.

Rare — fewer than 1 in 1,000. Reported with marketed use rather than counted in trials.

Source [08]

A blocked bowelSerious

A belly that swells, pain, vomiting and no bowel movement. It sits in the same column as cutaneous amyloidosis, meaning the regulator has reports but no denominator to turn them into a rate.

Frequency not known — it appears in the regulator's table only as something reported after the drug went on sale.

Source [08]

Breathing stomach contents into the lungs during an operationSerious

The drug slows the stomach down, so food can still be sitting there when someone is put under general anaesthetic or deep sedation, and it can go the wrong way. The warning is directed at the anaesthetist rather than the patient.

No rate published. The European regulator added it to the warnings after cases were reported.

Source [08]

Blood sugar dropping too lowSerious

Shaking, sweating, confusion and hunger. On its own liraglutide releases insulin only when blood sugar is already up, which is why it barely causes this in people without diabetes. Combined with insulin or a sulfonylurea, which push insulin out regardless, the risk becomes real. Clinically significant episodes were also reported in adolescents on the drug — 1.6% against 0.8%.

Depends entirely on what else is being taken. In adults with type 2 diabetes on the drug, episodes needing another person's help were reported by 0.7%, and only among those also on a sulfonylurea tablet. Documented low readings with symptoms: 43.6% against 27.3% on dummy injections. In adults without diabetes, symptoms were reported by 1.6% against 1.1%, and none were severe.

Source [08]

Feeling sick

The defining side effect. It appears in the first weeks and mostly fades within days or weeks of continuing. Nausea was the single most common reason people quit, at 2.9% against 0.2%. Nausea is also the most reported term in the US regulator's public side-effect file for liraglutide, with 7,250 reports.

Very common — 39.3% of 3,384 people on the drug against 13.8% of 1,941 on dummy injections.

Source [09]Source [19]

Loose stools

Comes on with the nausea in the first weeks and usually settles with it. It matters beyond the discomfort: it is the route by which people get dried out, and severe diarrhoea can stop other tablets being absorbed properly.

Very common — 20.9% against 9.9% on dummy injections.

Source [08]Source [09]

Constipation

The mirror image of the loose stools, and it can happen to the same person on a different week. It comes from the same slowing of the gut.

Very common — 19.4% against 8.5% on dummy injections.

Source [09]

Vomiting

Usually early and usually short-lived, but it is the second most common reason people quit, at 1.7% against under 0.1%.

Very common — 15.7% against 3.9% on dummy injections. Twice as common in adolescents as in adults, and twice as common again in children aged 6 to 11.

Source [08]Source [09]

A faster resting heart

It peaks about six weeks in and goes back to where it started when the drug is stopped. Most people never notice it; some feel a thumping heartbeat while sitting still. The European regulator asks for heart rate to be checked at regular intervals, and for the drug to be stopped if a meaningful rise persists. It also states that the long-term consequences of the average rise are not established.

An average rise of 2.5 beats a minute across the trials, ranging from 1.6 to 3.6. A racing heart reported as a side effect in 0.6% against 0.1%. On continuous 24-hour monitoring the difference was 4 to 9 beats a minute; in children 3 to 7.

Source [08]Source [09]

Reactions where the needle goes in, and lumps under the skin

Usually mild redness or itching that fades while treatment carries on. The lumps are a separate problem. Injecting the same patch of skin over and over can build a deposit of protein there, and the product information says injection sites should always be rotated to reduce that risk.

13.9% against 10.5% on dummy injections — most of these reactions happen with a dummy injection too. Lumps of amyloid protein under the skin have no published rate.

Source [08]Source [09]

Headache, tiredness and trouble sleeping

The headache figure is worth reading twice. It is very common on the drug and very nearly as common on a dummy injection. That is what it looks like when most of a side effect is not the drug. Tiredness and sleeplessness show a real gap.

Headache 13.6% against 12.6% on dummy injections — almost the same. Tiredness 7.5% against 4.6%. Sleeplessness is listed as common, mainly in the first three months.

Source [08]Source [09]

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.

SCALE Obesity and Prediabetes (trial 1), weight loss in adults with obesity or overweight and no diabetes, 3,731 adults

3.0 mg once a day, injected under the skin

56 weeks for everyone; 160 weeks for the 2,254 who had prediabetes, each followed by 12 weeks of observation off the drug

Source [01]Source [08]

SCALE Diabetes (trial 2), weight loss in adults who also had type 2 diabetes, 846 adults

3.0 mg once a day, injected under the skin

56 weeks

Source [08]

SCALE Sleep Apnoea (trial 3), breathing interruptions during sleep, 359 adults

3.0 mg once a day, injected under the skin

32 weeks

Source [08]

SCALE Maintenance (trial 4), holding on to weight already lost by dieting, 422 adults

3.0 mg once a day, injected under the skin

56 weeks

Source [08]

LEADER, heart attacks and strokes in adults with type 2 diabetes, 9,340 adults

Up to 1.8 mg once a day, injected under the skin, on top of usual diabetes care

Between 3.5 and 5 years, median 3.8 years

Source [08]Source [13]

STEP 8, the head-to-head trial against weekly semaglutide, 127 adults in the liraglutide group

3.0 mg once a day, injected under the skin

68 weeks

Source [14]

The Danish trial of exercise against liraglutide against both, 195 adults, which is where the bone and post-treatment data come from

3.0 mg once a day, injected under the skin

52 weeks, after an 8-week low-calorie diet, then a year with no treatment at all

Source [10]Source [15]

Trial 4180 in 251 adolescents aged 12 and over with obesity; 82.4% of them ended the trial on the full amount

3.0 mg once a day, injected under the skin. 103 of the 125 reached that amount and stayed on it; the other 22 finished the trial on less, depending on what they could tolerate

56 weeks, then 26 weeks off the drug

Source [08]

The mechanism study in 49 adults with obesity and no diabetes that measured appetite, stomach emptying and how much they ate at a free buffet

1.8 mg or 3.0 mg once a day, injected under the skin; each person took two of the three treatments, one in each period

5 weeks per period

Source [07]

What people report

• Whether taking it for more than about four years is safe — the longest weight trial ran 160 weeks, the longest diabetes trial a median of 3.8 years, and nothing published goes further
• Whether it changes anything for people who stop and stay off — every published figure for what happens after stopping runs out at one year
• How much muscle is lost, in kilograms, over a full course — the only body-composition scans in liraglutide-specific trials cover 20 adults over 35 days and one bone-focused analysis in 195 adults
• Whether the bone density lost during treatment recovers afterwards, or leads to more fractures later — the trial that measured it stopped measuring at 52 weeks
• What the small rise in resting heart rate does over years — the European regulator states the long-term clinical impact has not been established
• Whether the thyroid tumours seen in rats and mice mean anything for people — no human study has been long enough to find out, and the two regulators disagree about how to act in the meantime
• Whether the slightly higher counts of breast cancer in the trials — 17 of 2,379 women against 3 of 1,300 — are the drug or chance, along with the same question about growths in the bowel
• How often it causes lumps of amyloid protein under the skin at injection sites — the regulator lists it with no rate at all
• Whether it is safe in pregnancy — the European product information says the human data are too limited and the risk unknown, and instructs that it be stopped
• How it behaves in people over 75, in severe kidney failure or in severe liver failure — not studied in any of them, which is why it is not recommended there
• Whether it does anything for children with Prader-Willi syndrome — the one trial in 56 of them found the same BMI change on the drug as on dummy injections
• Why nearly everyone stops taking it in ordinary practice — 16.6% were still on it at 12 months in one hospital's records, and no study has asked those people why they left
• Whether combining exercise with it changes anything beyond weight and bone — the one trial that tested the combination measured body composition, bone and blood sugar, not illness or death

How much weight will it actually take off?

About 8 kg over a year in the largest trial, against about 3 kg for people injecting a dummy — so about 5 kg of it is the drug. Put another way, 63% lost at least 5% of their body weight against 27% on the dummy, and about a third lost more than 10% against about one in ten. In people who already had type 2 diabetes the figure was smaller: 5.9% against 2.0%. And outside a trial it is smaller again — in one hospital's records the average at six months was 4.2%.

Source [01]Source [08]Source [17]

What happens if I stop?

Most of it comes back, and faster than it went. In the Danish trial, the group on liraglutide alone regained 9.6 kg in the twelve months after stopping. That is 6.0 kg more than the group that had been doing supervised exercise instead, and it left them two years on heavier than when the injections started. Pooled across six trials of this class in 3,236 people, about 60% of the lost weight was back within a year, levelling off at about three quarters. Blood pressure came back faster still: 70 to 80% of the improvement was gone within 12 weeks. In adolescents, 26 weeks off the drug wiped out most of the BMI change from the previous year.

Source [08]Source [15]Source [16]

Does it take muscle as well as fat?

Some, and the drug on its own does nothing to stop it. The clearest evidence is about bone rather than muscle. In 195 adults over a year, hip bone density fell more than twice as far in the liraglutide-alone group as in the dummy group, and lower-spine density fell where the dummy group's did not move. The exercise group lost less weight than the liraglutide group and kept more bone. Adding supervised exercise to the drug brought bone density back in line with the dummy group. For muscle the data are much thinner: in 20 adults over 35 days, the absolute amount of lean tissue in the trunk and limbs fell while lean tissue as a share of body weight held steady.

Source [10]Source [11]

How does it compare with semaglutide?

Badly, on weight. The two were tested head to head in 338 adults over 68 weeks: liraglutide took off 6.4% of body weight, weekly semaglutide 15.8%. At least 15% lost: 12.0% against 55.6%. Gut side effects were about the same in both groups, 82.7% against 84.1%. But twice as many people gave up on liraglutide, 27.6% against 13.5%, which is what a daily injection costs against a weekly one. Liraglutide's advantage is age: it has been on the market since 2009, so its long-term record is longer.

Source [14]

Does it cause thyroid cancer?

It caused thyroid tumours in rats and mice, and the two regulators disagree about what that means for people. The US label carries a boxed warning and bars anyone with a personal or family history of medullary thyroid cancer, or the inherited condition multiple endocrine neoplasia type 2. The European regulator looked at the same animal data. Its reading is that the tumours came from a mechanism rodents are unusually sensitive to, that the relevance to humans is likely to be low but cannot be completely ruled out, and that no other treatment-related tumours were found. No human trial has been long enough or large enough to settle it either way.

Source [09]Source [12]

Is it true it causes suicidal thoughts?

The European regulator looked and said no. Its safety committee reviewed animal work, trial data and the reports sent in since the drugs went on sale. It also read a large study of electronic health records comparing semaglutide against other medicines, and ran a second health-records study of its own in people with type 2 diabetes. Its conclusion, published after the meeting of 8 to 11 April 2024, was that the available evidence does not support a causal link between this group of drugs and suicidal or self-harming thoughts and actions. It made no change to the product information.

Source [20]

Does it do anything besides shift weight?

Yes, and this is the strongest part of its record. 9,340 adults with type 2 diabetes, most of them with existing heart disease, were followed for a median of 3.8 years. Heart attacks, strokes and heart deaths happened to 13.0% on liraglutide against 14.9% on dummy injections, and death from any cause to 8.2% against 9.6%. In people with prediabetes, 3% had developed type 2 diabetes by week 160 against 11%. In people with moderate or severe sleep apnoea, breathing interruptions fell by 12.2 an hour against 6.1. Kidney damage in the heart trial came less often; damage to the back of the eye came slightly more often, but that comparison answered nothing either way.

Source [08]Source [12]Source [13]

Will the sickness ever stop?

For most people, within days or weeks. The European product information states that gut reactions usually appear in the first weeks and diminish while treatment continues, and that most of them were mild or moderate and did not lead to stopping. For a substantial minority it does not settle: 9.8% of people on the drug stopped because of a side effect against 4.3% on dummy injections, and nausea and vomiting were the top two reasons. People aged 65 and over, and people with mild or moderate kidney impairment, get more of it.

Source [08]Source [09]

Each letter represents one amino acid.

1. [01]Pi-Sunyer X m.fl. A Randomized, Controlled Trial of 3.0 mg of Liraglutide in Weight Management (SCALE). N Engl J Med (2015)
2. [02]EMA – Saxenda (liraglutid), EPAR
3. [03]EMA – Victoza (liraglutid), EPAR
4. [04]PubChem CID 16134956 – Liraglutide (formula C172H265N43O51, full chemical name; the sequence was read residue by residue from it)
5. [05]NCATS Inxight Drugs – Liraglutide (UNII 839I73S42A), systematic peptide name, residue by residue
6. [06]WADA Prohibited List 2026 (official publication, Austrian BGBl. III no. 219/2025) – liraglutide does not appear in any class (2025)
7. [07]Effects of the once-daily GLP-1 analog liraglutide on gastric emptying, glycemic parameters, appetite and energy metabolism in obese, non-diabetic adults (49 adults, 5 weeks) (2014)
8. [08]European Medicines Agency — Saxenda (liraglutide) product information, section 5.1 Pharmacodynamic effects
9. [09]DailyMed — SAXENDA (liraglutide) injection, US prescribing information, Adverse Reactions 6.1
10. [10]Bone Health After Exercise Alone, GLP-1 Receptor Agonist Treatment, or Combination Treatment: A Secondary Analysis of a Randomized Clinical Trial (195 adults, 52 weeks) (2024)
11. [11]Effects of liraglutide treatment for 35 days on total and regional fat free, lean, and bone mass (20 adults, randomised placebo-controlled crossover). Diabetes Research and Clinical Practice (2026)
12. [12]European Medicines Agency — Victoza (liraglutide) product information, section 4.8 Pancreatitis
13. [13]Marso SP et al. Liraglutide and Cardiovascular Outcomes in Type 2 Diabetes (LEADER). N Engl J Med (2016)
14. [14]Rubino DM et al. Effect of Weekly Subcutaneous Semaglutide vs Daily Liraglutide on Body Weight in Adults With Overweight or Obesity Without Diabetes (STEP 8). JAMA (2022)
15. [15]Jensen SBK et al. Healthy weight loss maintenance with exercise, GLP-1 receptor agonist, or both combined followed by one year without treatment: a post-treatment analysis of a randomised placebo-controlled trial. eClinicalMedicine (2024)
16. [16]Trajectory of weight regain after cessation of GLP-1 receptor agonists: a systematic review and nonlinear meta-regression. eClinicalMedicine (2026)
17. [17]Comparative effectiveness of long-term anti-obesity medications in routine clinical practice: a retrospective cohort study using real-world data. Translational and Clinical Pharmacology (2025)
18. [18]Glucagon-Like Peptide-1 Receptor Agonists and Gastrointestinal Adverse Events: A Systematic Review and Meta-Analysis (55 trials, 106,395 participants). Gastroenterology (2025)
19. [19]openFDA drug adverse event counts for liraglutide, reaction terms by frequency (data last updated 28 April 2026) (2026)
20. [20]European Medicines Agency — Meeting highlights from the Pharmacovigilance Risk Assessment Committee (PRAC) 8-11 April 2024 (2024)