Trial summary

Liraglutide vs semaglutide: what STEP 8 found

By David Chen·Editor·Reviewed September 2026

Participants

338 randomized

Duration

68 weeks

Comparator

Liraglutide 3.0 mg daily (and pooled placebo)

Published

JAMA, 2022

Registration

NCT04074161

PubMed

PMID 35015037

Two molecules, four brands³

Liraglutide and semaglutide are both GLP-1 receptor agonists made by Novo Nordisk, and each is sold under two names — one approved for weight, one for type 2 diabetes. The brand pairs people search for are really the same two molecules at different doses.

The starting schedules are the other practical difference. Saxenda starts at 0.6 mg a day and rises every week, reaching 3 mg in week 5.⁴ Wegovy starts at 0.25 mg a week and rises every four weeks, reaching maintenance in week 17 — the full ladder is on the dosage chart. The Wegovy label does not recommend using it with any other GLP-1 receptor agonist, which includes liraglutide. (Wegovy now also comes as a once-daily 25 mg tablet, and adults using the injection for weight can go up to 7.2 mg; STEP 8 tested the 2.4 mg injection.)

Who was in the trial¹

338 adults at 19 sites in the United States, all with a BMI of 30 or more — or 27 or more with at least one weight-related condition — and none with diabetes. Their mean age was 49, 78.4% were women, and they started at a mean weight of 104.5 kg. That makes it a Saxenda-vs-Wegovy trial, not a Victoza-vs-Ozempic one: both arms used the weight-loss dose.

How it was run¹

Participants were randomized 3:1:3:1 to semaglutide 2.4 mg weekly (126 people), its matching placebo, liraglutide 3.0 mg daily (127), or its matching placebo; the two placebo groups were pooled (85). All four arms had counseling on diet and physical activity.

The comparison between the two drugs was open-label — one is injected daily and the other weekly, so participants knew which drug they might be on — while each drug was blinded against its own placebo. Semaglutide climbed to 2.4 mg over 16 weeks, and anyone who could not tolerate it could stay at 1.7 mg. Liraglutide climbed to 3.0 mg over 4 weeks, and, in line with its label, anyone who could not tolerate 3.0 mg stopped treatment and could restart the titration.

What it found¹

The difference carries a 95% confidence interval of -12.0 to -6.8 percentage points (P<.001). Pooled placebo lost 1.9%. In kilograms the two drugs came to -15.3 kg and -6.8 kg.

These are observed shares among the 117 people in each arm weighed at week 68. The 10%, 15% and 20% rows were confirmatory end points; the 5% row was exploratory and was not tested statistically. For scale, the semaglutide arm of STEP 1 lost 14.9% over the same 68 weeks, and the authors describe STEP 8’s results as generally consistent with earlier trials of each drug.

Who stopped, and why¹

On placebo, 17.6% stopped, 55.3% reported gastrointestinal events and 7.1% a serious adverse event. No one died. Gastrointestinal events were most common during and shortly after dose escalation, and half of the adverse-event discontinuations on liraglutide were gastrointestinal.

The authors offer possible reasons for the gap in discontinuations, and present them as possibilities rather than findings. Among them: liraglutide’s shorter half-life (13 to 15 hours, against 165 for semaglutide) could make hunger return more noticeably when treatment is paused; its label required stopping, rather than stepping down, when 3.0 mg was not tolerated; and it is injected every day. Treatment satisfaction was not measured, so which of these mattered is unknown.

What limits the comparison¹

The authors name three limits. The rule for intolerance differed between the arms, so liraglutide users may have spent less time on treatment and lost less for that reason. Participants knew which active drug they could be on. And missing weights were imputed, although 92% of each group was weighed at week 68. It was also a modest trial — 338 people, all in the United States, none with diabetes.

One difference sits in the labels rather than the trial. Saxenda’s label tells prescribers to check weight 16 weeks in:⁴

Evaluate the change in body weight 16 weeks after initiating SAXENDA® and discontinue SAXENDA® if the patient has not lost at least 4% of baseline body weight, since it is unlikely that the patient will achieve and sustain clinically meaningful weight loss with continued treatment.

The Wegovy label has no equivalent cut-off. What it says instead is to consider treatment response and tolerability when choosing the maintenance dose. How that plays out for someone who is not losing weight is on not losing weight on semaglutide.³

Victoza vs Ozempic: SUSTAIN 10²

The diabetes brands have their own head-to-head. SUSTAIN 10 randomized 577 adults with type 2 diabetes, already on one to three oral diabetes medicines, to semaglutide 1.0 mg weekly or liraglutide 1.2 mg daily for 30 weeks, open-label.

Semaglutide was superior on both blood sugar and weight. Unlike STEP 8, it was the arm with more gastrointestinal side effects and more discontinuations. Read the doses before the headline, though: 1.2 mg is not Victoza’s ceiling — its label allows up to 1.8 mg⁶ — and Ozempic’s label goes to 2 mg.⁵ The trial chose 1.2 mg as the liraglutide dose most prescribed in Europe, so it is a comparison of common doses, not of top doses. For the semaglutide trial against another weekly drug in diabetes, see SUSTAIN 7.

What this means for a compounded vial

This page reports what the trials and labels recorded. It does not suggest which drug anyone should take or switch to — that is a decision for a prescriber. See the medical disclaimer.