SURMOUNT-OSA: tirzepatide cut sleep apnoea events by 20 to 24 an hour more than placebo

SURMOUNT-OSA tested tirzepatide in adults who have both obesity and moderate-to-severe obstructive sleep apnoea. After 52 weeks, breathing interruptions per hour of sleep fell by 25.3 on tirzepatide against 5.3 on placebo in one trial, and by 29.3 against 5.5 in the other [1]. The estimated differences were 20.0 and 23.8 events an hour [1].

Why this study matters

Obstructive sleep apnoea is repeated narrowing or closure of the airway during sleep. Excess body fat is a leading risk factor, and a 2026 review calls obesity the dominant reversible cause [2]. The standard treatment, a mask that holds the airway open with air pressure (positive airway pressure, or PAP), keeps the airway open but does not change body weight [2].

That leaves an obvious question. If obesity narrows the airway, does a large drop in weight on an injectable such as tirzepatide reduce the number of breathing interruptions? SURMOUNT-OSA was built to test exactly that. The review reports that the US Food and Drug Administration approved tirzepatide in December 2024 for moderate-to-severe obstructive sleep apnoea in adults with obesity [2].

What they did

The two trials differed in one respect. Trial 1 enrolled people who were not using PAP at baseline. Trial 2 enrolled people who were already using it [1]. Within each trial, participants were assigned 1:1 to tirzepatide or placebo for 52 weeks, and the review puts the combined total at 469 adults [2].

The primary end point was the change in the apnoea-hypopnoea index (AHI). The AHI counts the apnoeas (breathing pauses) and hypopnoeas (shallow breaths) in an average hour of sleep, so a lower number means fewer interruptions [1].

What they found

At the start, the mean AHI was 51.5 events an hour in trial 1 and 49.5 in trial 2. Mean body-mass index was 39.1 and 38.7 [1].

On our own arithmetic, a fall of 25.3 from a starting average of 51.5 leaves the tirzepatide group near 26 events an hour in trial 1. That is a large drop and still well above zero. The abstract reports averages for each group, and it does not report how many individual people stopped having apnoea, so we do not quote a remission figure.

Every key secondary end point the authors prespecified also favoured tirzepatide: percent change in AHI, percent change in body weight, hypoxic burden (a measure of how far oxygen levels dipped during sleep), patient-reported sleep impairment and disturbance, high-sensitivity C-reactive protein (a blood marker of inflammation), and systolic blood pressure [1].

The most frequent side effects with tirzepatide were gastrointestinal, and the authors describe most as mild to moderate [1].

How big is a 20-event drop?

The AHI is also used to grade severity. An AHI of 15 or more is the threshold the review gives for entry into the trials [2], and both groups started near 50 events an hour, which is roughly one interruption every minute and a quarter. On placebo, the average fell by about five events an hour, and the abstract does not explain why, though people in a trial are also followed closely and may change habits. The tirzepatide groups fell by five times as much or more. The trial authors report significant improvements on every prespecified key secondary end point, so the AHI result did not stand alone [1].

Caveats

• Sponsor-run. Eli Lilly, which makes tirzepatide, funded both trials, and several authors work for the company [1].
• Weight and drug cannot be separated. Tirzepatide lowers body weight substantially, and the trials were not designed to show how much of the AHI change comes from the weight loss itself. The review explains the likely route as less fat around the tongue and the sides of the throat [2]. That is the reviewers' account of mechanism, not something these trials measured.
• Maximum tolerated dose. Doses were set by tolerance, so the result describes the drug as people could take it, not at one fixed dose.
• Intermediate end points. The trials measured breathing events and risk markers. They were not built to count heart attacks, strokes or deaths [1].
• One population. Everyone had obesity, with an average BMI near 39. The result says nothing about sleep apnoea in people without obesity.

What it means in practice

Two separate randomised trials, one in people without PAP and one in people already using it, moved in the same direction. That is stronger than either alone.

The second trial deserves a careful reading. Its participants were already on PAP at baseline, but the abstract does not say what happened to their mask use over the year [1]. So the result is evidence that tirzepatide lowers AHI, not evidence that it replaces a mask.

Tirzepatide is a prescription medicine. Whether it is licensed for sleep apnoea depends on the country, and the decision sits with a prescriber and a sleep clinic. For the wider evidence on the drug, see our tirzepatide guide, the SURMOUNT-1 weight-loss trial and the head-to-head with semaglutide.

What this means

We read SURMOUNT-OSA as the clearest evidence so far that, when obesity is driving sleep apnoea, a large weight drop on tirzepatide produces a large drop in breathing interruptions. We are confident in the direction, because two double-blind trials agree, and less sure of the size outside a sponsor-run trial and beyond one year. Take from it that weight is a real lever on sleep apnoea. Do not take from it that a mask or a sleep-clinic assessment is optional.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.