Liraglutide: the drug that proved the GLP-1 class

Liraglutide is a synthetic peptide with 97% amino acid homology to native human GLP-1 ^[1], the incretin hormone secreted by intestinal L-cells in response to nutrient ingestion. The native hormone has a half-life of roughly two minutes ^[2], which is why endogenous GLP-1 has never been a viable therapeutic on its own. Novo Nordisk's engineering insight, commercialized as liraglutide, was to attach a C-16 palmitoyl fatty acid side chain that binds reversibly to serum albumin, slowing renal clearance and proteolytic degradation enough to stretch the half-life to about 13 hours ^[3]. That single chemical modification is what made once-daily subcutaneous dosing possible and what launched the GLP-1 class as a real category in medicine.

The receptor, GLP-1R, is a G-protein-coupled receptor expressed across the pancreas, brain, heart, kidneys, and GI tract, which is why a drug aimed at glycemic control turned out to affect appetite, gastric motility, cardiovascular risk, and, in trials still being run, kidney function and cognition. Liraglutide is a full agonist, activating the receptor to maximal effect rather than partially.

The downstream actions are by now familiar to anyone following the category. Glucose-dependent insulin secretion from pancreatic beta cells lowers postprandial glucose without producing the hypoglycemia that destroyed the careers of earlier diabetes drugs. Suppression of alpha-cell glucagon release lowers fasting glucose. Delayed gastric emptying blunts the post-meal glucose spike and produces the early sense of fullness patients describe in the first weeks. And the mechanism that turned a diabetes drug into a weight-loss drug: central activation of pro-opiomelanocortin neurons in the hypothalamic arcuate nucleus, with reciprocal inhibition of the NPY/AgRP hunger-driving neurons through a GABA-dependent circuit (Secher et al., J Clin Invest, 2014) ^[4]. The peptide crosses the blood-brain barrier at the circumventricular organs, where the barrier is naturally fenestrated, and acts directly on the brain's hunger thermostat.

Liraglutide's regulatory trajectory is essentially the regulatory trajectory of GLP-1 medicine itself, compressed into a decade.

Victoza, the 1.2 and 1.8 mg formulation, received initial FDA approval on January 25, 2010 ^[5], as an adjunct to diet and exercise for glycemic control in adults with type 2 diabetes. The pediatric expansion to patients aged 10 and older followed ^[6]. After the LEADER trial readout, more on which below, the FDA added a cardiovascular risk reduction indication to the label for adults with established cardiovascular disease, making Victoza the first GLP-1 RA to carry such a claim.

Saxenda, the 3.0 mg obesity formulation, was approved on December 23, 2014, after a 14-1 advisory committee vote in favor ^[7]. Eligibility was set at BMI ≥30, or ≥27 with at least one weight-related comorbidity such as hypertension, dyslipidemia, or type 2 diabetes. On December 4, 2020, Saxenda became the first GLP-1 RA approved for adolescent obesity, covering ages 12 through 17 ^[8], a milestone that mattered because it established the regulatory and safety framework that semaglutide later inherited for its own pediatric indication.

The approvals stacked deliberately. Glycemic control first, then cardiovascular protection, then chronic weight management in adults, then weight management in adolescents. Each step was earned with a dedicated phase 3 program rather than extrapolated from earlier data. That's unusual in metabolic medicine and is part of why liraglutide's evidence base is, even now, more thoroughly built out than the data supporting some of the drugs that have eclipsed it commercially.

The half-life of roughly 13 hours, with peak plasma concentration at 8 to 12 hours and bioavailability around 55% subcutaneously ^[9], was a large achievement when it was achieved. It's also the principal reason the drug now feels dated. Semaglutide's half-life of approximately one week, engineered through a different fatty acid chain and additional amino acid substitutions, made once-weekly dosing possible and changed what patients are willing to tolerate.

Liraglutide is cleared by general proteolytic degradation rather than through CYP450 metabolism, which means drug-drug interaction concerns are minimal, a real advantage in older patients on polypharmacy. Renal and hepatic dose adjustments are generally not required at standard doses. These are quiet virtues. They don't show up in marketing material because they're the kind of thing you only notice when something goes wrong with the alternatives.

The LEAD (Liraglutide Effect and Action in Diabetes) program comprised six phase 3 trials that brought Victoza to market between 2009 and 2010. Across the LEAD trials, meaningful HbA1c reductions were observed depending on the comparator and background therapy, with weight loss of 1 to 3 kg as a secondary endpoint, a finding that, in hindsight, was the leading edge of the obesity application that followed.

The more clinically useful comparison now is SUSTAIN-10 (Capehorn et al., Diabetes Metab, 2020), which directly pitted semaglutide 1.0 mg once-weekly against liraglutide 1.2 mg once-daily in 577 adults with T2D ^[10]. At 30 weeks, liraglutide reduced HbA1c by 1.0% and body weight by 1.9 kg. Both of those are real, durable effects. Semaglutide did more on both endpoints: an additional 0.69% HbA1c reduction and an additional 3.83 kg of weight loss, both p<0.0001 ^[11]. A subsequent systematic review of head-to-head and indirect comparisons found a pooled HbA1c advantage for semaglutide of 0.56 percentage points (95% CI 0.19 to 0.94), with large between-study heterogeneity (I² = 95.58%) ^[12] that the meta-analysts could not fully resolve.

Two things follow. First, the liraglutide effect is not nothing. A 1% HbA1c drop is what most diabetes drugs aspire to and many fail to deliver. Second, when both drugs are accessible and the patient can tolerate weekly dosing, semaglutide is the better choice on glycemia alone. The question is what happens when both conditions aren't met.

Four trials made up the SCALE program for Saxenda. The headline result, from SCALE Obesity and Pre-diabetes, a large trial in non-diabetic adults with obesity or overweight plus comorbidity, was a notably greater mean weight loss at 56 weeks on liraglutide 3 mg versus placebo, both with diet and exercise counseling. The placebo arm tells you what structured lifestyle counseling delivers on its own, a real but modest effect, and the drug arm tells you what the GLP-1 adds on top.

The same trial's three-year extension in the pre-diabetes subgroup produced what is, to me, the most underappreciated finding in the liraglutide file: a 79% reduction in incident type 2 diabetes (1.8% versus 6.2% incidence, HR 0.21, p<0.0001) ^[13]. That's a prevention signal of a magnitude rarely seen in metabolic medicine, sustained over a clinically meaningful window. SCALE Diabetes, in a cohort of patients with both obesity and T2D, showed meaningfully greater weight loss on the 3 mg dose versus placebo, with about half of treated patients losing 5% or more.

The SCALE Sleep Apnea trial (Blackman et al., 2016) extended the evidence base into a high-risk obesity comorbidity: a cohort of non-diabetic adults with moderate-to-severe OSA who could not or wouldn't use CPAP, with apnea-hypopnea index as the primary endpoint. The trial was designed around a population that genuinely had no good alternative, and the results supported an OSA-specific benefit beyond what weight loss alone would predict.

What none of this addresses is the comparison to semaglutide 2.4 mg and tirzepatide. The most informative real-world data point is the JAMA Network Open cohort from 2024: at one year, mean weight reduction was −5.1% on semaglutide versus −2.2% on liraglutide in routine practice, and among persistent users with an obesity indication, the gap widened further in semaglutide's favor. The network meta-analysis of multiple RCTs ranks semaglutide 2.4 mg first for weight reduction, with liraglutide 3.0 mg producing more modest losses. The honest read of the data: liraglutide produces about 5 to 8% body weight loss in trial conditions and about 2 to 6% in clinical practice, against 10 to 15% for semaglutide 2.4 mg and somewhat more for tirzepatide. That's not a margin of error. It's a generational gap.

The LEADER trial (Marso et al., NEJM, 2016; NCT01179048) enrolled 9,340 adults with type 2 diabetes at high cardiovascular risk, randomized to liraglutide 1.8 mg daily or placebo on top of standard care, with median follow-up of 3.8 years ^[14]. The primary endpoint, first occurrence of cardiovascular death, nonfatal MI, or nonfatal stroke, occurred in 13.0% of the liraglutide arm versus 14.9% on placebo (HR 0.87, 95% CI 0.78 to 0.97, p=0.01 for superiority) ^[15].

The composite was driven largely by cardiovascular mortality: 4.7% versus 6.0%, HR 0.78, p=0.007 ^[16]. All-cause mortality was also lower (8.2% versus 9.6%, HR 0.85, p=0.02) ^[17], and a nephropathy composite favored liraglutide as well (HR 0.78, p=0.003) ^[18]. Nonfatal MI and nonfatal stroke individually were numerically lower but did not reach significance. This was the first GLP-1 RA, and one of the first diabetes drugs of any class, to show cardiovascular superiority over placebo in a properly powered outcomes trial.

The clinical question that has occupied the field since is how this maps onto the newer drugs. Semaglutide's SUSTAIN-6 (in T2D) and SELECT (in non-diabetic patients with established CVD and obesity) both showed cardiovascular benefit, with SELECT in particular extending the indication into non-diabetic primary prevention. Tirzepatide's CVOT, SURPASS-CVOT, has more recently reported its results. The cardiovascular protection associated with this drug class now appears reasonably strong across multiple agents, which means liraglutide's claim to historical primacy matters less for current prescribing decisions. What still matters is that for a 65-year-old with diabetes and established coronary disease, you have a decade of outcomes data on liraglutide. For tirzepatide you have a fraction of that.

The titration schedules are straightforward on paper and harder in life.

For Victoza in T2D, the starting dose is 0.6 mg daily for one week. This dose is explicitly not therapeutic for glycemia. It exists to acclimatize the gut. After a week, escalate to 1.2 mg daily; this is the first dose that does clinical work. If additional glycemic control is needed after at least a week on 1.2 mg, escalate to the maximum of 1.8 mg daily.

For Saxenda in obesity, the climb is longer: 0.6 mg in week 1, 1.2 mg in week 2, 1.8 mg in week 3, 2.4 mg in week 4, and 3.0 mg from week 5 onward as the maintenance dose. The label is explicit that if 3.0 mg is not tolerated, the drug should be discontinued rather than maintained at a lower dose, because the lower doses are not approved for obesity and have not been shown to produce the weight-loss effect. The 16-week reassessment is also explicit: if a patient has not lost at least 4% of body weight by week 16, discontinue ^[19].

That 16-week rule is the most useful single thing in the Saxenda label. It saves patients from being kept on a drug that's not working for them, and it gives the clinician a defined off-ramp before the cost and side-effect burden compounds further. The newer drugs have similar stopping rules with similar logic. Whichever GLP-1 is used, the question to ask at month four is whether the drug is doing what it was supposed to do, and the willingness to act on the answer is what separates good prescribing from inertia.

Administration is subcutaneous in the abdomen, thigh, or upper arm, any time of day, with or without food. Patients should rotate sites. Pens are single-patient devices.

For Victoza: type 2 diabetes in adults or in children aged 10 and older, as add-on to diet and exercise. The cardiovascular indication applies to adults with established cardiovascular disease.

For Saxenda: adults with BMI ≥30, or ≥27 with at least one weight-related comorbidity (hypertension, dyslipidemia, T2D, OSA), and adolescents aged 12 to 17 with body weight above 60 kg ^[20] and BMI corresponding to obesity in adults by standard growth charts.

The absolute contraindication on both labels is personal or family history of medullary thyroid carcinoma or Multiple Endocrine Neoplasia syndrome type 2. This is class-wide for GLP-1 RAs and derives from rodent C-cell tumor findings whose human relevance remains undetermined. The cautious read of the available human data is that the MTC signal has not materialized at the population level, but the contraindication stands and is enforceable.

Relative contraindications and populations to think twice about: a personal history of pancreatitis, severe gastroparesis, active gallbladder disease, pregnancy or active pregnancy planning (animal reproductive toxicity data are concerning and the drug is not recommended), and severe renal impairment in real GI side effects that could precipitate dehydration. Patients on insulin or sulfonylureas should expect dose reductions of those drugs at GLP-1 initiation to avoid hypoglycemia.

The GI burden is the defining feature. Nausea occurs in a substantial share of liraglutide-treated patients across trials, well above the rate seen on placebo. Diarrhea and constipation each affect a notable proportion of users, and vomiting is common. The pattern is dose-dependent and titration-dependent: symptoms are worst in the first weeks at each new dose and typically attenuate over two to four weeks if the patient stays the course. The titration schedule exists to make this tolerable, and patients who escalate faster than the schedule allows pay for it. Worth flagging that in the SUSTAIN-10 head-to-head, GI-related discontinuations actually ran lower on liraglutide than on semaglutide 1.0 mg, even though the broader literature shows similar GI rates across the class; the across-studies and head-to-head pictures don't always line up.

Pancreatitis is the adverse event that most worries prescribers, and the data are imperfect. Victoza trials recorded more cases of pancreatitis in liraglutide arms than in comparators, with some analyses suggesting the rate was several times higher among those on the drug. Saxenda trials produced multiple confirmed cases including some after discontinuation, with serious outcomes reported in some instances. Lipase elevation occurred more frequently in Saxenda patients than in those on placebo. The honest synthesis: pancreatitis risk is real, small in absolute terms, probably elevated by prior pancreatitis or active gallbladder disease, and should prompt discontinuation if it occurs.

Gallbladder disease, cholelithiasis and cholecystitis, runs slightly higher on liraglutide than placebo, with LEADER suggesting roughly one extra case per 100 patients treated for four years ^[21]. Acute kidney injury has been reported postmarketing, almost always in dehydration from severe GI symptoms. A modest heart rate increase is consistently observed and is not generally clinically meaningful. Hypersensitivity reactions including anaphylaxis and angioedema are rare but documented. The regulatory posture on suicidal ideation risk with weight-loss agents, including Saxenda, has shifted over time as postmarketing data have accumulated, and prescribers should consult current labeling for the most recent FDA guidance.

The thyroid C-cell tumor signal sits in its own category: boxed, rodent-derived, human-relevance-undetermined, with a postmarketing MTC registry running for a minimum of 15 years ^[22]. The contraindication for MTC and MEN 2 personal or family history is firm.

The cleanest read: same mechanism, same class, different pharmacokinetics, different commercial generation, meaningfully different efficacy.

On HbA1c, semaglutide is superior by roughly 0.5 to 0.7 percentage points across head-to-head and meta-analytic data. On weight loss, the gap widens. Semaglutide 2.4 mg produces roughly twice the weight reduction of liraglutide 3.0 mg in trials, and the real-world JAMA Network Open data suggest the gap may be larger in clinical practice. Tolerability is broadly similar in kind: same GI symptom profile, similar adverse event categories. Some patients tolerate one better than the other for reasons that are not fully predictable from the pharmacology.

Dosing frequency is the most consequential practical difference. A weekly injection versus a daily one is the difference between something patients incorporate into a routine and something they think about every morning. Adherence data favor the weekly drug substantially. For most patients starting GLP-1 therapy today, that alone is the deciding factor, and the efficacy advantage on top of it is overdetermining.

Tirzepatide is a dual GLP-1 and GIP receptor co-agonist, which is a different pharmacological proposition rather than a more potent version of the same one. The GIP arm of the dual mechanism appears to contribute additional weight loss, and the trial data for tirzepatide place it ahead of semaglutide on weight reduction, which by extension places it well ahead of liraglutide.

The cardiovascular outcomes evidence for tirzepatide is now more developed than it was, but the liraglutide CV file is still the more mature one: a decade of post-LEADER follow-up, real-world registry data, and a consistent benefit pattern across populations. For a patient where cardiovascular risk reduction is the primary motivation and the evidence base needs to be solid, liraglutide remains a defensible choice. For a patient where weight loss is the primary goal and other considerations are secondary, tirzepatide is almost always the better answer.

The December 2020 Saxenda approval for adolescents 12 to 17 was based on a 251-patient randomized trial showing greater BMI reduction on liraglutide than placebo over 56 weeks. The effect was smaller in adolescents than in adults (adolescents physiologically grow into their target weights to some degree, which compresses the visible drug effect), but it was real, and it established that GLP-1 therapy could be deployed in a population where pharmacological options for obesity had been almost nonexistent.

The Victoza pediatric T2D approval (ages ≥10) sits alongside this and was based on adequate glycemic response in pediatric trials. Both pediatric uses carry the same boxed warnings and contraindications as adult use, with additional postmarketing requirements specific to pediatric populations.

For parents weighing adolescent GLP-1 therapy, the honest position is this: the long-term safety data in growing children simply don't exist yet, in the same way they did not exist for statins in pediatric familial hypercholesterolemia for a long time. The decision to start a GLP-1 in a 14-year-old is a decision to accept that the safety evidence will accumulate while the child is taking the drug, and the framework for that decision is the severity of the obesity, the comorbidity burden, the family history, and what has already been tried.

Liraglutide is an FDA-approved branded drug under Novo Nordisk's Victoza and Saxenda labels. It has never been on the FDA shortage list in the way semaglutide and tirzepatide were between 2022 and 2025, and it's not within the scope of the 503A and 503B compounding pathways that drove the compounded GLP-1 market. Compounding of liraglutide for individual patients requires specific clinical justification under state pharmacy law; FDA has taken enforcement action against compounders producing GLP-1 analogues without proper authorization.

Novo Nordisk's primary US patents on liraglutide expired in 2023, opening the door to generic and biosimilar competition. Several biosimilar liraglutide products are in development or have launched outside the US, and the first US generic approvals were issued in late 2024 and 2025. The post-patent pricing dynamics are still settling and will continue to shift through 2026.

The Saxenda postmarketing requirements (the MTC registry running for 15 years minimum, ongoing pediatric studies, and breast cancer risk evaluation in active trials) remain in force. The breast cancer signal is one to watch. Preliminary data have not established a causal link, but the FDA's requirement that the question be studied formally reflects that the question is not yet closed.

A few patient profiles where liraglutide is genuinely the right answer in 2026, rather than the answer of last resort:

Patients with established cardiovascular disease and type 2 diabetes who need a GLP-1 with the longest follow-up data. LEADER's median 3.8-year follow-up and the post-trial registry data give liraglutide an evidence base for CV protection that the newer drugs are still building.

Patients with severe injection anxiety or needle-phobia for whom daily small-volume injections are easier to tolerate than weekly larger-volume ones. This is a small population but it exists, and the pharmacokinetic difference is real.

Adolescents 12 to 17 starting GLP-1 therapy for obesity where the prescriber wants to use the agent with the most established pediatric data. Semaglutide has its own pediatric approval now, but liraglutide's adolescent file is more mature.

Patients with prior pre-diabetes and a strong family history of T2D where diabetes prevention is the primary goal. The 79% incident T2D reduction in the SCALE pre-diabetes extension is the strongest prevention signal in the GLP-1 file and was generated on liraglutide specifically.

Patients who tried semaglutide or tirzepatide and could not tolerate the weekly drug, not because the side effects differ in kind, but because some patients find daily small doses preferable to weekly larger ones, particularly during the titration phase.

For most other patients, the newer drugs are now the better starting point. That's the honest read of the comparative evidence and the honest read of where the field has moved. Liraglutide's place in 2026 is narrower than it was, but it's not nothing. The drug that proved the class still works.