TL;DR: In December 2024, the FDA approved Zepbound (tirzepatide) as the first drug for moderate-to-severe obstructive sleep apnea in adults with obesity, based on the SURMOUNT-OSA trials. Across two 52-week studies, tirzepatide cut the apnea-hypopnea index by 25.3 to 29.3 events per hour, versus 5.3 to 5.5 events per hour on placebo, alongside 18 to 20% weight loss. It’s a weight-mediated effect, not a direct action on the airway, and that’s worth saying plainly rather than burying it.
The approval: OSA gets its first drug
Obstructive sleep apnea has been a device-and-hardware condition for as long as it’s had a standard treatment. CPAP machines, oral appliances, and in some cases surgery have been the options, and none of them are a pharmaceutical fix for the underlying problem. That changed on December 20, 2024, when the FDA approved Zepbound, Eli Lilly’s tirzepatide, for moderate-to-severe OSA in adults with obesity, on top of its existing weight-management indication. It’s the same molecule already familiar from tirzepatide’s dual GIP/GLP-1 mechanism; what’s new is the label, not the drug.
The approval rested on SURMOUNT-OSA, a pair of phase 3 trials Lilly ran specifically to test tirzepatide against this condition rather than against weight alone. That distinction matters for how you read the data below: this wasn’t a weight trial with a sleep questionnaire tacked on, it was designed around a respiratory endpoint from the start.
What AHI actually measures
The apnea-hypopnea index (AHI) is the standard yardstick for OSA severity. It counts the number of breathing interruptions, apneas (complete pauses) and hypopneas (partial reductions in airflow), per hour of sleep, usually measured overnight in a sleep study. The scale has four widely used bands.
To get into SURMOUNT-OSA, participants needed an AHI of at least 15 events per hour and a BMI of 30 or higher, so this was a moderate-to-severe population by design. Baseline AHI averaged 51.5 events per hour in one trial and 49.5 in the other, well into the severe range, meaning the typical participant was waking up, or shifting out of deep sleep, roughly once a minute all night.
The trials, and the numbers
SURMOUNT-OSA was actually two parallel trials, split by whether participants were already using positive airway pressure (PAP) therapy, the umbrella term that covers CPAP and its variants.
Study 1 enrolled 234 people not using PAP therapy (114 on tirzepatide, 120 on placebo). Study 2 enrolled 235 people who were already on PAP and stayed on it throughout (120 on tirzepatide, 115 on placebo). Both ran 52 weeks with tirzepatide dosed at the maximum tolerated level, 10 mg or 15 mg weekly, the same doses already used in tirzepatide’s standard titration schedule.
At week 52, the AHI reductions were large in both arms:
SURMOUNT-OSA Study 1 (no baseline PAP) and Study 2 (on baseline PAP). Negative values indicate fewer breathing interruptions per hour.
In Study 1, that’s a 50.7% relative drop in AHI on tirzepatide against a 3.0% drop on placebo. In Study 2, tirzepatide produced a 58.7% relative reduction versus 2.5% on placebo, with an estimated treatment difference of 23.8 events per hour (p under 0.001). Roughly 61 to 72% of tirzepatide recipients across the two trials achieved at least a 50% reduction in AHI, and 42.2% (Study 1) to 50.2% (Study 2) reached a threshold researchers defined as disease resolution: an AHI under 5, or an AHI of 5 to 14 with minimal daytime sleepiness. For context on the weight side of these same trials, tirzepatide produced roughly 18 to 20% body weight reduction, compared with about 2% on placebo, numbers in the same neighborhood as the weight-loss results reported for knee osteoarthritis in TRIUMPH-4, a different condition but the same weight-mediated logic.
| Trial | Baseline AHI | Tirzepatide AHI change | Placebo AHI change |
|---|---|---|---|
| Study 1 (no baseline PAP) | 51.5 events/hr | -25.3 events/hr (-50.7%) | -5.3 events/hr (-3.0%) |
| Study 2 (on baseline PAP) | 49.5 events/hr | -29.3 events/hr (-58.7%) | -5.5 events/hr (-2.5%) |
Why weight loss lowers AHI
The mechanism here isn’t mysterious, and it isn’t specific to tirzepatide. Obstructive sleep apnea in people with obesity is driven substantially by two overlapping problems: fat deposited around and within the tissues of the upper airway (the tongue, soft palate, and tissue lining the throat) narrows the passage the air has to move through, and excess abdominal and chest wall fat restricts how much the diaphragm and rib cage can expand, especially lying down. Both effects make the airway more prone to collapsing during sleep, which is the physical event an apnea or hypopnea actually is. Lose a meaningful share of body weight and both mechanical problems ease: less tissue crowding the airway, and more room for the chest to do its job.
That framing also explains why Study 2, the PAP-user group, showed a slightly larger AHI reduction than Study 1. Both groups lost comparable amounts of weight, so the modestly larger drop in the PAP arm is more likely a measurement or population quirk than evidence that PAP and tirzepatide interact directly. Either way, people already using PAP who added tirzepatide got meaningfully better numbers on top of what their machine was already managing.
What this means next to the rest of the tirzepatide picture
For anyone already on tirzepatide for weight or metabolic reasons who also carries an OSA diagnosis, this is one more reason a stalled scale isn’t the whole story. Weight loss on these drugs doesn’t arrive on a straight line, and our GLP-1 plateau guide covers why progress flattens and what actually restarts it, which matters here since AHI improvement in SURMOUNT-OSA tracked the same weight-loss curve.
It’s also worth knowing where tirzepatide sits relative to the newer triple agonist. Retatrutide has posted larger weight-loss numbers in its own trials, and our retatrutide vs tirzepatide comparison lays out the head-to-head picture, though retatrutide doesn’t yet carry an OSA indication or trial data comparable to SURMOUNT-OSA. For now, tirzepatide is the only drug with an FDA label and a purpose-built trial for this specific condition, which is a meaningfully different evidence bar than an indication inferred from a general weight trial.
Frequently asked questions
Is tirzepatide FDA-approved for sleep apnea?
Yes. The FDA approved Zepbound (tirzepatide) in December 2024 for moderate-to-severe obstructive sleep apnea in adults with obesity, making it the first drug approved specifically for this condition. The approval sits alongside tirzepatide’s existing chronic weight management indication and used the same maximum tolerated doses, 10 mg or 15 mg weekly.
How much did tirzepatide lower AHI in the SURMOUNT-OSA trials?
At 52 weeks, tirzepatide reduced AHI by 25.3 events per hour in participants not using PAP therapy and by 29.3 events per hour in participants already on PAP, compared with 5.3 and 5.5 events per hour on placebo. Baseline AHI averaged around 50 events per hour in both trials, so these are reductions of roughly 51 to 59%.
Does tirzepatide treat sleep apnea directly, or through weight loss?
Through weight loss. Tirzepatide drives substantial weight reduction, about 18 to 20% of body weight in the SURMOUNT-OSA trials, and that weight loss eases the fat-related crowding around the upper airway and the mechanical restriction on breathing that obesity causes. It isn’t acting on the airway or respiratory drive directly, which is also why the benefit depends on the weight loss being maintained.
Can I use tirzepatide for sleep apnea instead of CPAP?
SURMOUNT-OSA tested tirzepatide both in people not using PAP and in people who stayed on PAP throughout, and both groups saw large AHI improvements. That trial design doesn’t establish tirzepatide as a standalone replacement for PAP therapy; it establishes that tirzepatide meaningfully lowers AHI on its own and adds benefit on top of PAP for people already using it.
What counts as moderate or severe sleep apnea on the AHI scale?
An AHI under 5 events per hour is considered normal, 5 to 14 is mild OSA, 15 to 29 is moderate, and 30 or more is severe. SURMOUNT-OSA required an AHI of at least 15 to enroll, and average baseline AHI in both trials was around 50, placing the typical participant well into the severe category.
How does the SURMOUNT-OSA weight loss compare to other tirzepatide trials?
The 18 to 20% weight loss seen in SURMOUNT-OSA is consistent with, if slightly below, the roughly 20 to 22% peak weight loss reported in SURMOUNT-1’s obesity population, as covered in our what is tirzepatide overview. That consistency is part of why the AHI results track so closely with the mechanical, weight-loss-driven explanation above.