Liraglutide

LEADER: cardiovascular outcomes in type 2 diabetes (9,340 people)

LEADER randomized 9,340 people with type 2 diabetes and high cardiovascular risk to liraglutide or placebo on top of standard care, with a median follow-up of 3.8 years. The primary outcome, the first occurrence of cardiovascular death, nonfatal myocardial infarction or nonfatal stroke, occurred in 13.0 percent on liraglutide and 14.9 percent on placebo (hazard ratio 0.87, 95 percent CI 0.78 to 0.97). This result met both the noninferiority test (p less than 0.001) and the superiority test (p = 0.01). Cardiovascular death occurred in 4.7 and 6.0 percent (hazard ratio 0.78) and death from any cause in 8.2 and 9.6 percent (hazard ratio 0.85). Nonfatal myocardial infarction, nonfatal stroke and hospitalization for heart failure were lower with liraglutide, but not significantly. The adverse events that most often led to stopping liraglutide were gastrointestinal (PMID: 27295427).

A prespecified secondary analysis found the composite kidney outcome (new persistent macroalbuminuria, a doubling of serum creatinine, end-stage kidney disease or death from kidney disease) in 268 people on liraglutide and 337 on placebo (hazard ratio 0.78), driven mainly by fewer new cases of persistent macroalbuminuria (PMID: 28854085).

Glucose control in type 2 diabetes

• In LEAD-3, a 52-week double-blind trial in 746 adults with early type 2 diabetes, liraglutide 1.2 mg or 1.8 mg lowered HbA1c by 0.84 and 1.14 percentage points against 0.51 with glimepiride 8 mg, and both differences were significant (p = 0.0014 and p less than 0.0001) (PMID: 18819705).
• LEAD-6 was a 26-week open-label trial in 464 adults on metformin, a sulfonylurea or both. Once-daily liraglutide 1.8 mg lowered HbA1c by 1.12 points against 0.79 with twice-daily exenatide 10 micrograms (p less than 0.0001); weight fell 3.24 and 2.87 kg (PMID: 19515413).
• In a 26-week open-label trial of liraglutide against sitagliptin in 665 adults on metformin, HbA1c fell 1.24 points on liraglutide 1.2 mg and 1.50 on 1.8 mg against 0.90 on sitagliptin 100 mg (p less than 0.0001 for both) (PMID: 20417856).
• DURATION-6 ran open label for 26 weeks in 912 adults. HbA1c fell 1.48 points on liraglutide 1.8 mg daily and 1.28 on exenatide 2 mg weekly, and exenatide did not meet the predefined noninferiority margin (PMID: 23141817).

Children and adolescents with type 2 diabetes: ELLIPSE

ELLIPSE randomized 135 patients aged 10 to under 17 with type 2 diabetes, all taking metformin with or without basal insulin, to liraglutide up to 1.8 mg a day or placebo for 26 weeks, followed by a 26-week open-label extension. HbA1c fell 0.64 points with liraglutide and rose 0.42 with placebo (difference 1.06 points, p less than 0.001), and the difference grew to 1.30 points by week 52. Overall and gastrointestinal adverse event rates were higher with liraglutide (PMID: 31034184).

Weight management in adults: the SCALE program

• In a 20-week phase 2 trial in 564 adults with obesity in Europe, mean weight loss was 4.8, 5.5, 6.3 and 7.2 kg on liraglutide 1.2, 1.8, 2.4 and 3.0 mg against 2.8 kg on placebo and 4.1 kg on open-label orlistat (PMID: 19853906).
• SCALE Obesity and Prediabetes was a 56-week trial in 3,731 adults without diabetes. Weight fell 8.4 kg on liraglutide 3.0 mg and 2.8 kg on placebo in this trial (difference 5.6 kg, p less than 0.001; PMID: 26132939). A loss of at least 5 percent was reached by 63.2 percent against 27.1 percent, and a loss of more than 10 percent by 33.1 percent against 10.6 percent, so all three coprimary end points were met (PMID: 26132939).
• The SCALE 3-year assessment followed 2,254 adults with prediabetes for 160 weeks. Type 2 diabetes was diagnosed during treatment in 2 percent on liraglutide and 6 percent on placebo, and time to onset was 2.7 times longer with liraglutide (hazard ratio 0.21). Weight fell 6.1 percent and 1.9 percent. Half of the participants completed 160 weeks, and those who withdrew were not followed up (PMID: 28237263).
• In SCALE Diabetes, a 56-week trial in 846 adults with type 2 diabetes, weight fell 6.0 percent on liraglutide 3.0 mg, 4.7 percent on 1.8 mg and 2.0 percent on placebo, and no pancreatitis was reported (PMID: 26284720).
• SCALE Maintenance enrolled 422 adults who had first lost at least 5 percent on a low-calorie diet. Over 56 weeks, liraglutide 3.0 mg led to a further 6.2 percent loss against 0.2 percent on placebo, and 81.4 percent against 48.9 percent kept off the run-in loss (PMID: 23812094).
• In SCALE Sleep Apnea, a 32-week trial in 359 adults with obesity and moderate or severe obstructive sleep apnea, the apnea-hypopnea index fell by 12.2 events per hour on liraglutide and 6.1 on placebo (difference 6.1, p = 0.015), and weight fell 5.7 percent against 1.6 percent (PMID: 27005405).
• STEP 8 ran for 68 weeks in 338 adults without diabetes and was open label between the two active drugs. Weight changed by 15.8 percent on once-weekly semaglutide 2.4 mg and 6.4 percent on liraglutide 3.0 mg (difference 9.4 percentage points, p less than 0.001), and 27.6 percent on liraglutide against 13.5 percent on semaglutide stopped treatment for any reason (PMID: 35015037).

Adolescents and children with obesity

• SCALE Teens lasted 56 weeks plus 26 weeks off treatment and enrolled 251 adolescents aged 12 to under 18. The primary end point of SCALE Teens, change in BMI standard-deviation score, favored liraglutide 3.0 mg by 0.22 (p = 0.002; PMID: 32233338). A BMI reduction of at least 5 percent occurred in 43.3 percent against 18.7 percent. After treatment stopped, the BMI standard-deviation score rose more in the liraglutide group (PMID: 32233338).
• SCALE Kids was a 56-week trial in 82 children aged 6 to under 12. BMI fell 5.8 percent with liraglutide and rose 1.6 percent with placebo (difference 7.4 percentage points, p less than 0.001). Body weight rose in both groups, by 1.6 percent on liraglutide and 10.0 percent on placebo. Serious adverse events occurred in 12 percent and 8 percent (PMID: 39258838).

Other conditions studied

• LEAN, a 48-week phase 2 trial, enrolled 52 adults with nonalcoholic steatohepatitis at four UK centres. Steatohepatitis resolved without worsening fibrosis in 9 of 23 patients on liraglutide 1.8 mg (39 percent) against 2 of 22 on placebo (9 percent), and fibrosis progressed in 2 of 23 against 8 of 22 (PMID: 26608256).
• FIGHT followed 300 patients recently hospitalized for heart failure with reduced ejection fraction for 180 days. Liraglutide had no significant effect on the primary rank score built from death, heart failure rehospitalization and change in NT-proBNP (p = 0.31), and the authors concluded the findings do not support liraglutide in that setting (PMID: 27483064).
• ADJUNCT ONE was a 52-week trial in 1,398 adults with type 1 diabetes. Added to treat-to-target insulin, liraglutide 1.8 mg lowered HbA1c by 0.20 points more than placebo and reduced insulin dose and body weight, while symptomatic hypoglycemia and hyperglycemia with ketosis increased (PMID: 27506222).
• ELAD, a 52-week phase 2b trial, enrolled 204 people with mild to moderate Alzheimer's disease and no diabetes. The primary outcome, change in cerebral glucose metabolism on FDG-PET, did not differ from placebo (p = 0.14). An executive function score was better with liraglutide in an unadjusted secondary analysis (p = 0.01), while scores for daily living activities and dementia severity did not differ (PMID: 41326666).

Human pharmacology and absorption of other drugs

• In healthy men given NN2211 for 7 days, the half-life was 12.6 hours, exposure was dose-proportional and the accumulation index was 1.4 to 1.5 (PMID: 11935150).
• A study of oral drugs at liraglutide steady state (1.8 mg) found that single doses of atorvastatin, lisinopril and digoxin reached their peaks up to 2 hours later with 27 to 38 percent lower peak levels, griseofulvin had a 37 percent higher peak, and total exposure (AUC) changed only slightly for all four (PMID: 26426736).
• In 18 adults with type 2 diabetes given paracetamol (acetaminophen), total exposure was equivalent with liraglutide and placebo, with a 31 percent lower peak reached 15 minutes later (PMID: 21792552).
• In 21 postmenopausal women given an oral contraceptive, ethinyl estradiol exposure was equivalent and levonorgestrel exposure 18 percent higher with liraglutide, while peaks were 12 and 13 percent lower and came about 1.5 hours later (PMID: 21228406).

Gastrointestinal effects

Nausea and diarrhea, mostly mild or moderate, were the most frequent adverse events with liraglutide 3.0 mg in the 3,731-person SCALE Obesity and Prediabetes trial, and serious adverse events occurred in 6.2 percent on liraglutide and 5.0 percent on placebo (PMID: 26132939). Against sitagliptin in type 2 diabetes, nausea affected 27 percent on liraglutide 1.8 mg, 21 percent on 1.2 mg and 5 percent on sitagliptin (PMID: 20417856). Against once-weekly exenatide in DURATION-6, nausea affected 21 percent on liraglutide and 9 percent on exenatide, diarrhea 13 and 6 percent, and vomiting 11 and 4 percent (PMID: 23141817). The Victoza label reports severe gastrointestinal adverse reactions in 4.4 percent (1.2 mg) and 4.2 percent (1.8 mg) of trial patients against 1.1 percent on placebo, and does not recommend liraglutide in severe gastroparesis (Victoza label).

In adolescents in SCALE Teens, gastrointestinal adverse events affected 64.8 percent on liraglutide and 36.5 percent on placebo, and 10.4 percent stopped treatment because of adverse events against none on placebo (PMID: 32233338). In children aged 6 to under 12 in SCALE Kids, gastrointestinal adverse events affected 80 percent on liraglutide and 54 percent on placebo (PMID: 39258838).

Pancreas

LEADER recorded adjudicated acute pancreatitis in 18 people (0.4 percent) on liraglutide and 23 (0.5 percent) on placebo. Serum lipase rose 28 percent and amylase 7 percent with liraglutide compared with placebo, and these rises did not predict later pancreatitis (PMID: 28476871). Both US labels warn that acute pancreatitis (including fatal hemorrhagic or necrotizing cases) has been seen with GLP-1 receptor agonists including liraglutide (Victoza label; Saxenda label).

Gallbladder

In the adult Saxenda trials, cholelithiasis was reported in 2.2 percent on liraglutide and 0.8 percent on placebo, and cholecystitis in 0.8 and 0.4 percent (Saxenda label). Most affected patients on liraglutide needed cholecystectomy, and the excess remained after accounting for the amount of weight lost (Saxenda label).

Thyroid

The boxed warning rests on thyroid C-cell tumors in rats and mice; the labels state that the human relevance has not been determined and that postmarketing reports of medullary thyroid carcinoma in liraglutide users are insufficient to establish or exclude a causal link (Victoza label; Saxenda label). A 2024 cohort study across Denmark, Norway and Sweden compared 145,410 people starting a GLP-1 receptor agonist with 291,667 starting a DPP-4 inhibitor and found no increased risk of thyroid cancer over a mean 3.9 years (hazard ratio 0.93, 95 percent CI 0.66 to 1.31; for medullary thyroid cancer 1.19, 95 percent CI 0.37 to 3.86). The Scandinavian study covered GLP-1 receptor agonists as a class rather than liraglutide alone (PMID: 38683947).

Heart rate

Mean resting heart rate rose 2 to 3 beats per minute with Saxenda compared with placebo in adult trials, and more patients on liraglutide had rises of more than 10 beats per minute at two consecutive visits (34 percent against 19 percent) (Saxenda label).

Low blood sugar

Both US labels describe a higher risk of hypoglycemia (including severe hypoglycemia) when liraglutide is combined with insulin or an insulin secretagogue such as a sulfonylurea (Victoza label; Saxenda label). In the adolescent weight-management trial, clinically significant hypoglycemia (below 54 mg/dL) occurred in 1.6 percent on liraglutide and 0.8 percent on placebo, although none of the participants had type 2 diabetes (Saxenda label). In LEAD-6, minor hypoglycemia occurred in 25.5 percent on liraglutide and 33.6 percent on twice-daily exenatide, against a background of metformin, a sulfonylurea or both (PMID: 19515413). Adding liraglutide to insulin in 1,398 adults with type 1 diabetes raised the rate of symptomatic hypoglycemia, and hyperglycemia with ketosis rose with the 1.8 mg dose (rate ratio 2.22) (PMID: 27506222).

Kidneys

Postmarketing reports describe acute kidney injury, sometimes needing dialysis, mostly in patients dehydrated by nausea, diarrhea or vomiting (Victoza label). In LEADER, renal adverse events were similar on liraglutide and placebo (15.1 and 16.5 events per 1,000 patient-years); they included acute kidney injury at 7.1 and 6.2 events per 1,000 patient-years (PMID: 28854085).

Suicidality

In SCALE Teens, one participant on liraglutide died by suicide, an event the investigator judged unlikely to be related to the trial treatment (PMID: 32233338). According to the label's list of recent major changes, the warning on suicidal behavior and ideation was removed from the Saxenda label in February 2026 (Saxenda label).

Other label warnings

Both US labels list serious hypersensitivity reactions (including anaphylaxis and angioedema) from postmarketing reports, and pulmonary aspiration during general anesthesia or deep sedation reported with GLP-1 receptor agonists (Victoza label; Saxenda label).

Pregnancy

In pregnant rats and rabbits, liraglutide caused early embryonic deaths and fetal abnormalities at exposures close to or below those in people (animal data, Victoza label). In a 2026 Danish registry study, liraglutide exposure around conception for diabetes was linked to preterm birth (adjusted odds ratio 1.70), while exposure for weight management was not (adjusted odds ratio 1.01), and the authors suggested the underlying diabetes may explain the difference (PMID: 41852577). A 2026 meta-analysis of seven cohort studies with more than 40,000 exposed pregnancies found no statistically significant increase in congenital malformations with GLP-1 receptor agonists as a class (odds ratio 1.11), reported a signal for urinary malformations (odds ratio 1.24) that rested only on unadjusted data, and rated the certainty of evidence as low (PMID: 42420519).