Skip to content
GLP Loss
← Research
Evidence

REDEFINE 1 Explained: 3,417 People, and 302 in the Arm That Mattered

REDEFINE 1 put 2,108 participants on cagrilintide-semaglutide and 705 on placebo, and gave each single-drug arm 302. Both coprimary endpoints and all three confirmatory secondaries compared the combination against placebo.

Owen Castellanos9 min read
REDEFINE 1: where the 3,417 were putA 21 to 3 to 3 to 7 allocation, over 68 weeksCombination2,108Placebo705Semaglutide only302Cagrilintide only302Both coprimary endpoints compared one arm to placeboSo did all three confirmatory secondary endpoints.Weight: −20.4% against −3.0%, a gap of 17.3 points.Gastrointestinal events: 79.6% against 39.9% on placebo.Primary completion: 30 October 2024. Results posted: none.256 sites in 23 countries; diabetes was an exclusion.

Seven in ten participants in this trial were assigned to the combination and two in ten to placebo, which made the comparison everybody already expected to win the best-powered comparison in the protocol. The two arms that could answer whether the second drug earns its place — semaglutide alone, cagrilintide alone — got 302 people each.[1] Everything puzzling about how REDEFINE 1 was received follows from that allocation. What the whole CagriSema program has found across its several trials is set out in the CagriSema article; this page is about how this one trial was built.

Four arms, one of them enormous

REDEFINE 1 was a phase 3a, 68-week, multicenter trial, placebo-controlled and active-controlled at once. Adults without diabetes were eligible at a body-mass index of 30 or higher, or 27 or higher with at least one obesity-related complication. A total of 3,417 participants were randomly assigned in a ratio of 21:3:3:7: 2,108 to weekly cagrilintide 2.4 mg with semaglutide 2.4 mg, 302 to semaglutide 2.4 mg alone, 302 to cagrilintide 2.4 mg alone, and 705 to placebo, with lifestyle intervention in every group.[1]

The registry record adds the mechanics. Masking was quadruple, covering participants, care providers, investigators and outcome assessors. The schedule was a 16-week dose escalation followed by a 52-week maintenance period. Enrollment ran across 256 sites in 23 countries. Anyone with a history of type 1 or type 2 diabetes, or a glycated hemoglobin of 6.5% or higher at screening, was excluded.[2]

A four-arm design with two active comparators is unusual and expensive, and running one at all is a real methodological choice — most combination trials never test their own ingredients separately. The question is what those arms were sized to do, and 302 participants is a descriptive reference rather than a powered test.

The 16-week escalation is worth holding on to as well. Roughly a quarter of the trial’s length passed before anyone reached the maintenance dose, so a 68-week result describes 52 weeks at target preceded by four months of climbing toward it. Every weekly injectable in this class carries a schedule of that shape, and what it costs in time before a result appears is set out in the titration article.

The coprimary endpoints, and who they were about

Both coprimary endpoints compared the combination against placebo: the relative change in body weight from baseline to week 68, and the proportion reaching a reduction of 5% or more. Mean body weight changed −20.4% with the combination against −3.0% with placebo, an estimated difference of −17.3 percentage points (95% CI, −18.1 to −16.6; P < 0.001).[1]

The three confirmatory secondary endpoints were reductions of 20% or more, 25% or more and 30% or more — each, again, against placebo, and each reached more often than placebo at P < 0.001. The registry’s own ordering places the comparison of the combination against cagrilintide and against semaglutide below those thresholds in the secondary list.[1][2]

So the hierarchy that carries statistical protection is entirely a set of placebo comparisons, and the arithmetic of combining two drugs — does the pair beat either one alone by enough to justify a second injection, a second set of side effects and a second price — sits outside it. A reader arriving with that question finds a trial of 3,417 people that did not prioritize it.

The abstract also names its analysis explicitly: effect estimates were assessed with the treatment-policy estimand, consistent with the intention-to-treat principle.[1] That is the conservative choice and the right one to publish, and it means the headline figure already absorbs everyone who stopped, switched or had surgery. No on-treatment figure appears in the abstract for comparison.

What the pooled evidence says the second drug bought

Because the within-trial comparison was undersized, the increment has been estimated by pooling across trials instead. A 2026 systematic review of seven randomized trials covering 8,069 participants reported greater weight loss with the combination than with semaglutide alone at a mean difference of −7.58 kg (95% CI, −10.30 to −4.86; P < 0.00001), and than with cagrilintide alone at −9.24 kg (95% CI, −10.46 to −8.02).[3]

A separate 2026 meta-analysis restricted to three randomized trials and 3,545 participants found the combination beating semaglutide by −7.47% (95% CI, −10.58 to −4.36) and −7.60 kg (95% CI, −10.33 to −4.86), and reported that cagrilintide given on its own produced weight loss comparable to semaglutide.[4] A network meta-analysis of six amylin-based trials in 4,642 participants placed high-dose CagriSema at −17.18% against placebo and semaglutide 2.4 mg at −11.45% within the same network.[5]

These are consistent with each other and none of them is the randomized head-to-head that REDEFINE 1 itself could have delivered at scale. What semaglutide alone achieves in its own pivotal trials is set out in the semaglutide weight article, and the arithmetic of applying any of these figures to a personal starting weight is what the expected weight loss tool is for.

Three columns that run against the combination

Pooled against semaglutide alone, overall and serious adverse events were comparable, but three findings went the other way. Low-density lipoprotein cholesterol was higher on the combination, a mean difference of +0.29 mmol/L (95% CI, 0.02 to 0.55; P = 0.03). Administration-site conditions were more frequent, a risk ratio of 3.27 (95% CI, 1.27 to 8.46), as was nausea, at 1.64 (95% CI, 1.01 to 2.66).[4]

The same analysis reports something stranger about the amylin analog on its own: cagrilintide monotherapy carried a higher risk of serious adverse events than semaglutide, a risk ratio of 1.83 (95% CI, 1.03 to 3.24), while producing comparable weight loss.[4] An interval whose lower bound sits at 1.03 has barely cleared no-effect and should be read as a signal rather than a settled difference, but it is the opposite of what a “gentler second ingredient” framing would predict.

On tolerability inside the trial itself, gastrointestinal adverse events affected 79.6% of the combination group and 39.9% of the placebo group, described as mainly transient and mild to moderate in severity.[1] Four in five is the highest such figure in the obesity literature, and the network meta-analysis found high-dose CagriSema the only regimen in its six-trial network that increased discontinuation for adverse events.[5]

The number this trial has not published

How many people stopped the combination in REDEFINE 1 is not stated in the abstract, and the registry record carries no posted results at all — the results-first-posted field is empty despite an actual primary completion date of 30 October 2024 and an overall completion date still listed as estimated.[2] The paper is paywalled with no PubMed Central record, so its own discontinuation table is not readable either.

That leaves the network meta-analysis signal as the only published evidence on the question, and a pooled finding across six trials is not a substitute for one trial’s own column.[5] Anyone comparing this regimen against the drugs already for sale should treat the discontinuation rate as unpublished rather than as low.

More is still to come from the same trial. The registry lists overall completion as estimated October 2026 rather than actual, and describes an extension phase that the placebo group was enrolled into, so the record remains open two years after the primary endpoint was reached.[2] A figure that is missing because a trial is unfinished and a figure that is missing because nobody published it look identical from outside, and here both descriptions apply to different parts of the same record.

What REDEFINE 1 did not establish

It measured weight for 68 weeks and nothing longer. No arm was withdrawn, so it says nothing about what happens when the injections stop. It measured no cardiovascular outcome, no kidney outcome and no mortality endpoint, which means the one class benefit that carried semaglutide beyond weight has not been demonstrated for this combination at all. It ran no comparison against tirzepatide or any other dual agonist. And by excluding diabetes it says nothing about the population where an added amylin analog might behave differently.

The regulatory position closes the loop. No product containing cagrilintide appears in Drugs@FDA or in the DailyMed label archive, alone or in combination.[6] There is nothing to buy, from any seller, at any price. Compounded preparations of the drugs that are approved — the ones listed on the compounded semaglutide board — are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed, and what is and is not inside such a preparation is the subject of the vial contents article.

Any offer of this combination today is therefore an offer of something that was never in this trial, quoting a figure produced over 68 weeks at 256 supervised sites with free investigational product. The distance between that and a monthly shipment is the distance between a protocol and a purchase.

Frequently asked

Why does the size of each REDEFINE 1 arm matter?
Because it decides which question the trial could answer. Of 3,417 participants, 2,108 received the combination and 705 received placebo, while the semaglutide-alone and cagrilintide-alone arms held 302 each. Both coprimary endpoints and all three confirmatory secondary endpoints were comparisons against placebo, so the comparison that decides whether the second drug earns its place sat outside the protected hierarchy.
What did REDEFINE 1 find on weight?
Mean body weight changed −20.4% with cagrilintide-semaglutide against −3.0% with placebo at week 68, an estimated difference of 17.3 percentage points with a 95% confidence interval of −18.1 to −16.6. Reductions of 5%, 20%, 25% and 30% or more were all reached more often than on placebo. Effect estimates used the treatment-policy estimand, which counts everyone as randomized whatever they went on to do.
How much did adding cagrilintide to semaglutide actually gain?
The trial did not test that with adequate power, so the estimates come from pooling. A seven-trial review of 8,069 participants put the advantage over semaglutide alone at 7.58 kg, and a three-trial analysis of 3,545 participants put it at 7.47% and 7.60 kg. The same three-trial analysis reported that cagrilintide given on its own produced weight loss comparable to semaglutide.
Did anything in the pooled data favor semaglutide alone?
Three things. Low-density lipoprotein cholesterol was higher with the combination by 0.29 mmol/L, with a confidence interval of 0.02 to 0.55. Administration-site conditions ran at a risk ratio of 3.27 and nausea at 1.64. And cagrilintide monotherapy carried a higher risk of serious adverse events than semaglutide, at a risk ratio of 1.83 with a lower bound of 1.03.
How many people stopped the drug in REDEFINE 1?
That has not been published. The abstract reports gastrointestinal adverse events in 79.6% of the combination group against 39.9% on placebo but gives no discontinuation figure, the paper is paywalled with no PubMed Central record, and the registry has posted no results despite a primary completion date of 30 October 2024. A network meta-analysis of six amylin-based trials found high-dose CagriSema the only regimen in its network that increased discontinuation for adverse events.
Can this combination be bought anywhere?
No. No product containing cagrilintide appears in Drugs@FDA or in the DailyMed label archive, alone or in combination, so there is no approved version and no lawful compounded version of the full regimen. Compounded preparations of the approved drugs are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed.

Sources

  1. [1] Garvey WT, Blüher M, Osorto Contreras CK, et al. (2025). Coadministered Cagrilintide and Semaglutide in Adults with Overweight or Obesity. N Engl J Med. PMID 40544433
  2. [2] Novo Nordisk A/S (2026). Efficacy and Safety of Cagrilintide s.c. 2.4 mg in Combination With Semaglutide s.c. 2.4 mg Once-weekly in Participants With Overweight or Obesity (REDEFINE 1): study record, NCT05567796. ClinicalTrials.gov. Source
  3. [3] Khan BW, Agha SA, Qamar I, et al. (2026). Maximizing weight loss with cagrisema: a systematic review and GRADE-assessed meta-analysis of randomized controlled trials. Naunyn Schmiedebergs Arch Pharmacol. PMID 42608559
  4. [4] Ahmed M, Hassan M, Tahir M, et al. (2026). Efficacy and Safety of Cagrilintide and Cagrisema Versus Semaglutide as Anti-Obesity Medications: A Systematic Review, Meta-Analysis and Meta-Regression. Diabetes Obes Metab. PMID 41834765
  5. [5] Kamrul-Hasan ABM, Khalil I, Mahajan K, et al. (2026). Novel Amylin-Based Therapies for Weight Management in Adults With Overweight or Obesity Without Diabetes: A Network Meta-Analysis. Endocrinol Diabetes Metab. PMID 42175595
  6. [6] U.S. Food and Drug Administration (2026). Drugs@FDA: FDA-Approved Drugs — no approved product containing cagrilintide, alone or in combination. U.S. Food and Drug Administration. Source

Where to get it

Best GLP-1 injections

Every injectable seller we can verify, with the price each one publishes and an honest read of what the trials measured.

Compare providers →

More in Evidence