Retatrutide sleep apnea trials report large drops in breathing pauses

Two phase 3 trials of retatrutide sleep apnea treatment report that breathing stopped far less often after eighty weeks on the drug than on placebo. The trials enrolled 691 adults in total, all with moderate to severe obstructive sleep apnea and a body mass index of at least 27, one trial in people without type 2 diabetes and one in people with it. Results appeared in The New England Journal of Medicine on 21 October 2026. Retatrutide is still investigational and is not approved anywhere for any use.

How the two trials were built

Both studies sat inside larger programmes testing the same drug for weight, which is why the sleep groups are smaller than the headline figures for retatrutide usually quoted. TRIUMPH 1 OSA randomised 243 adults without diabetes and TRIUMPH 2 OSA randomised 448 adults with type 2 diabetes. Each assigned participants to placebo or to one of three weekly amounts of retatrutide, given by injection, with the two higher amounts carrying the main comparison. The primary measure was the change after eighty weeks in the apnea hypopnea index, the average number of breathing pauses and shallow breaths per hour of sleep, scored from overnight sleep studies rather than from how rested people said they felt.

The participants were severely affected at the start. Mean index values were 58.6 and 55.9 events per hour in the two trials, which is well above the threshold for the severe category. The analysis used what the authors call a hybrid estimand, meaning that people whose course suggested treatment failure were handled as if they had stopped the drug, while other interruptions were handled as they occurred. Results using the plainer intention to treat approach are reported alongside and point the same way. Retatrutide acts on three receptors at once, for glucose dependent insulinotropic polypeptide, for GLP 1 and for glucagon, which is what the triple agonist label in the literature refers to.

What happened to the apnea hypopnea index

In the trial without diabetes, the index fell by 34.3 events per hour on the middle amount and 31.7 on the highest, against 9.9 on placebo. In the diabetes trial the falls were 24.7 and 25.4 against 10.3. The differences from placebo are the numbers that matter, and they were larger in the group without diabetes, as the table shows. Gastrointestinal complaints were the most commonly reported side effects, which is the familiar pattern for this class of drug. Placebo groups improved too, by roughly ten events per hour, a reminder that sleep study scores move on their own across a long trial.

In the trial without diabetes, the index fell by 34.3 events per hour on the middle amount and 31.7 on the highest, against 9.9 on placebo.

Read against the starting severity, the better performing groups moved from severe disease towards the mild end of the scale, while most placebo groups stayed severe. That is a large effect by the standards of drug treatment for this condition, where the usual alternative is a machine that holds the airway open all night and a diet programme that rarely shifts much weight. The index is only a count of events, though. It says nothing directly about daytime sleepiness, blood pressure, accidents or heart outcomes, which are the reasons the condition is treated in the first place. Sleep studies also vary from night to night in the same person, which is part of why placebo groups drift.

Why weight loss makes the result hard to read

Body weight fell steeply in the same participants. The report describes reductions of roughly twelve to twenty two percent across the retatrutide groups without diabetes and about twelve to nineteen percent with diabetes, against roughly five percent on placebo. Fat around the neck and abdomen narrows the airway and loads the chest, so losing a fifth of body weight would be expected to improve breathing during sleep on its own. The authors say the size of the weight loss suggests it contributes to the effect. The trials were not designed to separate a weight independent action of the drug from the consequences of being lighter, and no analysis reported here settles that question.

This matters for what the result means in practice. If the benefit travels with weight, then any treatment producing similar loss should produce similar improvement, and the comparison that clinicians need is against other options rather than against placebo. Tirzepatide is already approved for this condition in people with obesity, so the open question is relative benefit, not whether a drug can work at all. A trial that pitted retatrutide sleep apnea treatment against an existing agent, or that matched groups for weight change, would answer it. Neither of these trials did that, and placebo is a low bar after eighty weeks.

What the trials do not answer

Duration is the first gap. Eighty weeks is long for a sleep study programme but short against a chronic condition that people live with for decades, and nothing here shows what happens when treatment stops. Reviews of how fast weight returns once people stop these drugs suggest the question is not idle. The trials also enrolled a narrow group, severely affected adults with a high body mass index, so the findings do not transfer to people with mild disease, to those with a normal weight or to anyone whose airway collapses for anatomical reasons. Safety across a longer horizon remains to be established, and retatrutide has no approval anywhere.

There is also the matter of who ran the work. The trials were sponsored by the company developing the drug, employees of that company appear among the authors, and the results were released alongside a wider programme of trials in obesity. None of that makes the numbers wrong, and the design, with randomisation, blinding, placebo control and an objective overnight measure, is the strongest available. It does mean the independent replication and the comparisons against existing treatments still have to come from somewhere else. Regulators will see the full dataset before any of this reaches a prescription pad, and the published summary of the retatrutide sleep apnea work is what readers have for now.

Conclusion

Retatrutide sleep apnea trials show a clear and large reduction in breathing events over eighty weeks in two randomised comparisons against placebo, with gastrointestinal side effects as the main complaint. The catch is that participants also lost a great deal of weight, the trials cannot separate that from a direct effect, and no comparison against an approved treatment was made. The drug remains investigational. What it does for sleepiness, blood pressure and heart risk over years is still unknown.

Source Grunstein RR and colleagues. Retatrutide Once Weekly for Treatment of Obstructive Sleep Apnea. The New England Journal of Medicine, published 21 October 2026. DOI 10.1056/NEJMoa2611477

Amino Weekly reports research. It is not medical advice and never gives doses.