Not for weight loss. STEP 8 is the only trial to put the two drugs head to head [1]. Adults without diabetes lost 15.8% of their body weight on weekly semaglutide 2.4 mg and 6.4% on daily liraglutide 3.0 mg, against 1.9% on placebo. Liraglutide clearly beat placebo, and semaglutide more than doubled liraglutide.
That gap is the whole comparison, and most other numbers follow from it. A reader weighing how much weight a GLP-1 takes off should read the class average as a semaglutide-and-newer figure. Liraglutide is the oldest drug in the group, and it is still the one some weight programs offer first.
What STEP 8 measured
The trial ran at 19 US sites and randomized 338 adults without diabetes. Each had a body mass index of 30 or more, or 27 or more with a weight-related condition. Both drugs came with counseling on diet and physical activity.
The difference in weight change was 9.4 percentage points (95% CI −12.0 to −6.8), and the thresholds widen it. Weight loss of 10% or more was reached by 70.9% on semaglutide against 25.6% on liraglutide. The figures were 55.6% against 12.0% for 15% or more, and 38.5% against 6.0% for 20% or more.
One design point limits the result. The comparison between the two drugs was open-label, so participants knew which one they were taking. Each drug was blinded only against its own matched placebo.
Side effects and stopping
Gastrointestinal adverse events were reported by 84.1% on semaglutide and 82.7% on liraglutide, so on that measure the two drugs are the same.
Discontinuation for any reason was 13.5% on semaglutide and 27.6% on liraglutide. The trial abstract does not give the reasons behind that gap.
A network meta-analysis of 27 trials and 15,584 patients looked at seven drugs in the class [2]. It found that none of them raised serious adverse events, and no significant difference in withdrawal for adverse events in head-to-head comparisons. So liraglutide is not the harsher drug of the two. It is the weaker one.
The same analysis put retatrutide 12 mg, retatrutide 8 mg and tirzepatide 15 mg at the top for weight loss against placebo: 22.10%, 20.70% and 16.53%. Neither liraglutide nor semaglutide made that top three. For the next step up, the tirzepatide premium is its own question.
Is liraglutide cheaper per result?
Per course in one model, and per result it costs more. A Greek analysis priced both drugs over 68 weeks from the national payer’s side, using STEP 8’s efficacy [3]. Semaglutide cost more to supply, at €3,285.55 against €2,742.47.
It still cost less per patient who reached a target. At 5% weight loss that was €3,768.72 on semaglutide and €4,718.66 on liraglutide. The gap grew with the target, to €6,064.20 at 10%, €17,005.23 at 15% and €37,296.00 at 20%. These are Greek list prices, not a US bill.
Where liraglutide does win on paper
Against an older diabetes tablet, not against semaglutide. A Markov model of China’s health system compared liraglutide with sitagliptin, each added to metformin, over 30 years [4]. The ratio came out at 267,985.96 RMB per quality-adjusted life-year gained.
That sat below the threshold the authors used, three times China’s 2025 GDP per capita. In the probabilistic analysis it was cost-effective in 63.6% of iterations and not in the other 36.4%. The authors call it “likely to be cost-effective”, which is the accurate form for a result that close to the line.
The scenario analysis found the advantage grew as liraglutide’s price fell. Cost-effectiveness is a property of a price rather than of a molecule. That is why the broader economic studies change their answers when the comparator or the price changes.
Who might still take liraglutide
The trial answers efficacy, not every choice a buyer faces. Someone already stable on liraglutide, or offered it through a plan that covers nothing newer, faces a different decision. And in one cohort of adolescents on liraglutide, the spread of results was bigger than the average result.
What the evidence does settle is the ranking. On weight, head to head, semaglutide took off more than twice as much, and stopping early was twice as common on liraglutide.