Retatrutide for Sleep Apnea: What the TRIUMPH-OSA Trials Found
Short answer: yes, retatrutide improved sleep apnea in two phase 3 trials, and by a lot. In results published in the New England Journal of Medicine and featured by Lilly at ObesityWeek on November 15, 2026, the two higher doses cut breathing interruptions by about 25 to 34 events per hour over 80 weeks, versus about 10 on placebo. Roughly 4 to 5 in 10 people on those doses ended with sleep apnea that was gone or mild. The honest part: retatrutide is still not approved for anything, and the lowest dose did not clear the statistical bar.
Key facts
- Published and presented: NEJM online on Oct 21, 2026 (Grunstein et al.), then part of Lilly’s ObesityWeek data in Washington, D.C. on Nov 15, 2026.
- Who was in it: 691 adults with moderate to severe obstructive sleep apnea and a BMI of 27 or higher: 243 without type 2 diabetes (TRIUMPH-1-OSA) and 448 with it (TRIUMPH-2-OSA).
- Without diabetes, at 80 weeks: breathing events fell by 34.3 per hour on 9 mg and 31.7 on 12 mg vs 9.9 on placebo.
- With type 2 diabetes: down 24.7 (9 mg) and 25.4 (12 mg) vs 10.3 on placebo.
- Weight: down 12.5% to 22.2% and 12.1% to 18.5% on retatrutide vs 4.8% and 5.6% on placebo.
- Status, as of Nov 20, 2026: investigational, not approved anywhere. Lilly has said it plans to file with the FDA in Q1 2027.
What are the TRIUMPH-OSA trials?
Obstructive sleep apnea (OSA) is when your airway keeps collapsing during sleep, so breathing stops or goes shallow over and over. The score doctors use is the apnea-hypopnea index (AHI): the average number of those events per hour of sleep.
Lilly tested retatrutide, its once-weekly shot that acts on three hormone receptors (GIP, GLP-1 and glucagon), in two phase 3 trials built inside its larger TRIUMPH obesity trials. Both were randomized, double-blind and placebo-controlled, which is the strongest kind of evidence a drug gets before approval.
- TRIUMPH-1-OSA: 243 adults without type 2 diabetes. Starting AHI averaged 58.6 events an hour and starting weight 119.6 kg (about 264 lb).
- TRIUMPH-2-OSA: 448 adults with type 2 diabetes. Starting AHI 55.9, starting weight 115.9 kg (about 256 lb).
People got 4 mg, 9 mg or 12 mg of retatrutide or a placebo for 80 weeks. Per Healio, everyone started low and stepped up every 4 weeks, and 88% and 89% of participants finished the two trials. Lead author Ronald Grunstein of the Woolcock Institute of Medical Research in Sydney told Healio the group had both severe sleep apnea and severe obesity. On the diabetes trial he said: “No one really looked at that population before.”
New to the drug itself? Start with what retatrutide is.
How much did sleep apnea improve?
The main goal was the AHI change at 80 weeks on the 9 mg and 12 mg doses. Both doses hit it in both trials.
| Result at 80 weeks | 4 mg | 9 mg | 12 mg | Placebo |
|---|---|---|---|---|
| AHI change, no diabetes (events/hour) | -22.9 | -34.3 | -31.7 | -9.9 |
| AHI change, type 2 diabetes (events/hour) | -19.2 | -24.7 | -25.4 | -10.3 |
| Apnea gone or mild, no diabetes | not reported | 51.6% | 42.3% | 12.5% |
| Apnea gone or mild, type 2 diabetes | not reported | 39.5% | 38.9% | 12.6% |
“Gone or mild” is the trial’s own bar, per Healio: an AHI under 5, or 5 to 14 without daytime sleepiness (an Epworth Sleepiness Scale score of 10 or less).
Compared with placebo, that works out to about 24 and 22 fewer events an hour on the two higher doses in people without diabetes, and about 14 and 15 fewer in people with diabetes. When the researchers counted everyone who was randomized (intention to treat), the gaps barely moved: 24.7 and 22.5 without diabetes, 13.9 and 15.0 with it.
Worth noticing: more was not clearly better. Without diabetes, 9 mg landed slightly ahead of 12 mg.
Weight came down too: 12.5% to 22.2% across doses without diabetes and 12.1% to 18.5% with diabetes, versus 4.8% and 5.6% on placebo. Healio also reports larger improvements in blood pressure, cholesterol and hsCRP, a blood marker of inflammation, though the article gives no values for them.
Why Lilly’s own number is bigger
Lilly’s ObesityWeek release (Nov 14, 2026) says retatrutide reduced AHI “by up to an average of 36.1 events per hour (60.6%)” in TRIUMPH-1 and by up to 28.4 events per hour (48.1%) in TRIUMPH-2. Those are the same trials. The release does not say which dose or which counting method produced those figures, and it shows no placebo number next to them. The 36.1 figure goes back to Lilly’s announcement at the American Diabetes Association meeting in June 2026, per Sleep Review.
The NEJM numbers in the table above are the peer-reviewed ones, with the placebo arm shown. Those are the ones worth quoting.
How does it compare with Zepbound and Foundayo?
Nobody has tested these against each other for sleep apnea, so this is context, not a ranking.
- Zepbound (tirzepatide) is the one that is already FDA-approved for moderate to severe OSA with obesity, per Lilly.
- Foundayo (orforglipron), Lilly’s daily pill, cut AHI by 17.8 vs 4.5 events an hour on placebo without CPAP, and 17.3 vs 5.3 with CPAP, over 52 weeks in JAMA (Nov 9, 2026). It is not approved for sleep apnea. Our Foundayo sleep apnea post has the details.
- Retatrutide posted the larger drops, over a longer trial (80 weeks vs 52). Different trials, different people, different lengths.
One thing these trials do not settle is whether a drug can stand in for CPAP, the mask and machine that keeps the airway open. Grunstein told Healio he thinks medicines in this class could potentially replace CPAP for most patients, and that CPAP does not bring weight loss or the blood pressure drop these drugs seem to. That is his read of the field, not something these trials tested. The coverage we read does not report how many participants were using CPAP.
The fine print
Promising and unfinished can both be true:
- Not approved. As of Nov 20, 2026, retatrutide is investigational. Lilly’s own releases say it has not been approved by the FDA or any regulator in the world and cannot be legally sold or marketed for human use.
- The low dose missed. The 4 mg results did not reach statistical significance under the trials’ testing rules in either trial, per Healio. Andrew Namen of Wake Forest University, a spokesperson for the American Academy of Sleep Medicine, wrote in a commentary for Healio that retatrutide “did fulfill its primary endpoint” in both groups and noted that the 9 mg and 12 mg doses were needed to get there.
- Better, not gone, for most. Without diabetes, average AHI went from about 59 to roughly 24 events an hour on 9 mg. About half to 6 in 10 people on the higher doses still did not meet the trial’s “gone or mild” bar.
- Side effects were common. Any adverse event: 88% to 93% on retatrutide vs 76% on placebo without diabetes, and 73% to 84% vs 71% with diabetes. Most were stomach-related, mild to moderate, and showed up while the dose was being raised. Healio gives no separate nausea or vomiting rates.
- Few people quit over them. 9 of 181 people on retatrutide stopped because of side effects vs 4 of 62 on placebo (no diabetes), and 15 of 335 vs 4 of 113 (diabetes).
- Serious events and deaths. Serious adverse events ran 8% on retatrutide vs 10% on placebo in both trials. Four people died across the two trials: three on placebo and one on retatrutide 9 mg in the diabetes trial, a death Healio describes as gastrointestinal in cause.
- Feeling less sleepy is a softer result. On a patient-reported sleep impairment score, only the 9 mg and 12 mg groups in the diabetes trial clearly beat placebo (-10.9 and -12.6 vs -6.9).
- Lilly paid for the trials, and the lead author reports consulting for Lilly.
When could retatrutide be available for sleep apnea?
Not soon, and not on a known date. Lilly said on Sept 29, 2026 that it plans to submit retatrutide to the FDA in Q1 2027. That release does not say which uses the first filing will cover, so a sleep apnea label is not confirmed. Sleep apnea is one of the conditions in the phase 3 program, alongside obesity, type 2 diabetes, knee osteoarthritis pain and others. Our retatrutide availability timeline tracks the filing.
Until then, retatrutide only reaches people through Lilly’s own channels, mainly clinical trials. Lilly’s wording is blunt: anyone purporting to sell retatrutide to consumers is breaking the law. Vials sold on the internet as retatrutide are not the trial drug, and nobody checks what is in them. Our legal status tracker covers where unapproved peptides stand.
What it means for you
If you snore hard, wake up tired or gave up on your CPAP, step one is the same as ever: a sleep study and a real diagnosis. If you have sleep apnea and obesity, an approved option already exists (Zepbound), and a licensed provider can tell you whether it fits, how it works alongside CPAP, and what the side effects look like for you. Retatrutide looks like a strong future option, especially at 9 mg and 12 mg, but it is not available yet, and waiting on it is not a treatment plan. For the joint side of the same trial program, see retatrutide for knee pain. The retatrutide profile has the evidence grade and current FDA status.
Sources
- Grunstein et al. Retatrutide Once Weekly for Treatment of Obstructive Sleep Apnea. N Engl J Med, published online Oct 21, 2026 (PubMed)
- Healio: TRIUMPH-OSA, retatrutide reduces sleep apnea severity in OSA plus obesity (Nov 11, 2026)
- Lilly via PR Newswire: new data across Zepbound, Foundayo, EloraTZP and retatrutide at ObesityWeek 2026 (Nov 14, 2026)
- EMJ: Retatrutide cuts sleep apnea events in phase III trials (Oct 24, 2026)
- Sleep Review: Lilly’s investigational triple agonist retatrutide reduces OSA severity in phase 3 trial (Jun 23, 2026)
- Lilly via PR Newswire: retatrutide TRIUMPH-2 results and plan to submit to the FDA in Q1 2027 (Sept 29, 2026)
- Lilly via PR Newswire: retatrutide vs Ozempic head-to-head release, approval status wording (Oct 28, 2026)
- JAMA: Orforglipron for Obstructive Sleep Apnea, results from 2 randomized clinical trials (ATTAIN-OSA), Nov 9, 2026
Educational content only, not medical advice. Retatrutide is investigational and not FDA-approved for sleep apnea or any other use as of Nov 20, 2026; talk to a licensed provider about diagnosis and treatment options.