Zepbound for Sleep Apnea: What SURMOUNT-OSA Showed and What Is Approved

Yes, Zepbound is FDA-approved for sleep apnea: since December 20, 2024, its label has covered moderate to severe obstructive sleep apnea in adults with obesity, alongside a reduced-calorie diet and more physical activity^[1,6]. The FDA called it the first drug treatment option for certain patients with the condition^[6]. The approval rests on SURMOUNT-OSA, where after 52 weeks the number of breathing interruptions per hour of sleep fell by 25.3 on tirzepatide vs. 5.3 on placebo in people not using CPAP, and by 29.3 vs. 5.5 in people using it^[4].

This is general information, not a treatment guide. Diagnosing and treating sleep apnea is up to your doctor, usually together with a sleep medicine specialist.

What obstructive sleep apnea is

Sleep apnea is a common condition in which breathing stops and restarts many times during sleep, which can keep the body from getting enough oxygen^[7]. There are two types. In obstructive sleep apnea (OSA), the most common, the upper airway gets blocked again and again during sleep. In central sleep apnea, the brain doesn’t send the signals needed to breathe^[7]. Everything here is about the obstructive kind: SURMOUNT-OSA excluded people with central or mixed sleep apnea^[5].

The usual clues are snoring or gasping during sleep, often noticed by a bed partner, and excessive daytime sleepiness^[7]. Untreated sleep apnea raises the risk of stroke, heart attack and other serious problems^[7].

How it’s measured: the apnea-hypopnea index (AHI)

Diagnosis takes a sleep study^[7,10]. The main number is the AHI: the count of apneas (pauses in breathing) and hypopneas (partial drops in airflow) per hour of sleep^[4].

The moderate and severe rows are the definitions in the footnotes to Table 8 of the Zepbound label, section 14.2^[1]. The first two rows are my reading of an endpoint in its Table 9, “remission or mild non-symptomatic OSA,” which the label defines as an AHI under 5, or an AHI of 5 to 14 with an Epworth Sleepiness Scale score of 10 or less^[1]. The Epworth scale is a daytime sleepiness questionnaire scored from 0 to 24^[5].

The AHI isn’t the whole picture. An international consensus document published by the Spanish Society of Pulmonology and Thoracic Surgery (SEPAR), drawn up with 17 scientific societies, considers a severity rating based on the AHI alone limited^[10]. It suggests also weighing the time spent with oxygen saturation below 90%, daytime sleepiness, body mass index (BMI) and related conditions such as high blood pressure or type 2 diabetes^[10]. It defines OSA as an AHI of 15 or more, or 5 or more with excessive daytime sleepiness, unrefreshing sleep, fatigue or poorer quality of life not explained by anything else^[10].

How obesity is linked to sleep apnea

Obesity is one of the main risk factors. According to the NHLBI, people with obesity can have more fat deposits in the neck, which can block the upper airway^[8]. A MedlinePlus Genetics page from the National Library of Medicine, last updated in 2018, says 60% to 70% of people with OSA have obesity^[9]. The FDA puts it this way: OSA can affect anyone, but it’s more common in people with overweight or obesity^[6].

It isn’t the only cause. Age, large tonsils or tongue, hormone changes, family history, alcohol and smoking also raise the risk^[8], which is why some people with sleep apnea have a normal weight.

The SEPAR consensus recommends treating overweight or obesity in everyone with OSA, starting with an intensive lifestyle program, and having a specialized unit consider anti-obesity drugs or surgery in severe obesity^[10]. It stresses that treatment options, CPAP included, are complementary, not mutually exclusive^[10]. If you want to place your own BMI, there’s a BMI calculator; the link between sleep and appetite is in sleep, stress and appetite.

SURMOUNT-OSA: what was tested

SURMOUNT-OSA was really two parallel trials, funded by Eli Lilly, at 60 sites in nine countries^[4,5]. They enrolled 469 adults with moderate to severe obstructive sleep apnea (AHI of 15 or more) and obesity (BMI of 30 or more; 27 or more in Japan), without diabetes^[4,5]:

• Trial 1: 234 people not using positive airway pressure (PAP, such as CPAP), because they couldn’t or wouldn’t^[1,4].
• Trial 2: 235 people who were using it. They were asked to pause it for 7 days before each assessment, so the machine wouldn’t mask the result^[1,5].

In each trial, half got tirzepatide at the maximum tolerated dose (10 or 15 mg once weekly) and half got placebo, for 52 weeks. Everyone was counseled on a reduced-calorie diet and physical activity^[1,4]. At baseline, mean AHI was 51.5 and 49.5 events per hour and mean BMI was 39.1 and 38.7^[4]. Most had severe sleep apnea: 63% in trial 1 and 68% in trial 2^[1].

One detail before the numbers. The paper’s main analysis is the treatment-regimen estimand: it includes everyone who started, whether or not they kept taking the drug^[5]. The efficacy estimand estimates what would have happened if everyone had taken it as planned, and in this trial gave somewhat bigger numbers^[5]. The table uses the first one, whose numbers match the Zepbound label^[1,5].

Sources: Table 2 of the paper^[5] and Table 9 of the Zepbound label, section 14.2^[1].

• Difference vs. placebo. For the AHI, −20.0 events per hour in trial 1 (95% CI −25.8 to −14.2) and −23.8 in trial 2 (95% CI −29.6 to −17.9)^[4]. For weight, −16.1 and −17.3 percentage points^[4]. The 95% CI (confidence interval) is the range where the true effect probably lies.
• Other measures. Compared with placebo, tirzepatide also improved hypoxic burden (how often, how deeply and for how long oxygen drops with breathing events), C-reactive protein, systolic blood pressure and participants’ own reports of sleep-related impairment^[4,5]. The label says the AHI fell regardless of age, sex, ethnicity, baseline BMI or baseline sleep apnea severity^[1].
• Who finished. 91.5% of people on tirzepatide and 74.4% on placebo completed the trial^[5].

There’s a summary with the trial’s limitations in SURMOUNT-OSA.

Side effects in the trial

They were mostly digestive, mostly mild to moderate, and most common while the dose was being raised^[5]. These are the numbers from Table 4 of the paper^[5]:

There were no deaths. There were two confirmed cases of acute pancreatitis, both on tirzepatide in trial 2^[5]. The label says side effects on 10 mg or 15 mg were similar to those in the weight-management trials, and gives a pancreatitis rate in the sleep apnea studies of 0.84 patients per 100 years of exposure on Zepbound and 0 on placebo^[1]. The full list of side effects and warnings is in the tirzepatide guide and in Mounjaro and Zepbound side effects by dose.

What the Zepbound label says

Section 1 of the Zepbound label lists two uses, weight management and sleep apnea, and the FDA added the second on December 20, 2024, based on SURMOUNT-OSA^[1,6]. The schedule starts the same way for both and differs at maintenance^[1]. This isn’t a dosing plan for anyone: the prescriber decides the dose.

• Start. 2.5 mg once weekly for 4 weeks; 2.5 mg is for starting treatment and isn’t approved as a maintenance dose^[1].
• Increases. In 2.5 mg steps, after at least 4 weeks on the current dose^[1].
• Maintenance for sleep apnea. 10 mg or 15 mg once weekly, vs. 5, 10 or 15 mg for weight management^[1].
• Maximum. 15 mg once weekly^[1].

The label also says coadministration with other tirzepatide-containing products or with any GLP-1 receptor agonist is not recommended^[1]. Mounjaro has the same active ingredient, but its US label covers type 2 diabetes and lowering the risk of major cardiovascular events in adults with type 2 diabetes at high risk, not sleep apnea^[2]. Why one molecule has two brands, and how insurance and Medicare treat each one, is in Mounjaro vs. Zepbound.

What the trial doesn’t show

• That it replaces CPAP. SURMOUNT-OSA compared tirzepatide with placebo, not with CPAP^[4]. In the CPAP trial, participants paused their machine for 7 days before measurements, and the label states that the clinical studies for OSA did not evaluate the timing or appropriateness of PAP discontinuation in patients who were previously compliant with PAP therapy^[1]. The EU label goes further: Study 2 does not allow any conclusion about a potentially added benefit of tirzepatide on top of PAP therapy, because PAP was paused before the measurement^[3].
• That it changes CPAP use. Trial 2 wasn’t designed to find out whether the drug changes how consistently people use their machine^[5].
• What happens long term. At 52 weeks, it can’t assess heart attacks, strokes or other cardiovascular outcomes, and it didn’t measure what happens to sleep apnea after stopping the drug^[5]. What’s known about weight after stopping is in what happens when you stop a GLP-1 medication.
• What happens without obesity or with diabetes. It only included people with obesity and without diabetes, and excluded central or mixed sleep apnea^[5].

Add to that: Lilly funded the trial, and the statistical analyses were done by its employees^[5].

Outside the US: the EU and Mexico

• European Union. Section 4.1 of the EU label approves Mounjaro for type 2 diabetes and for weight management in adults with an initial BMI of 30 kg/m² or more, or 27 to under 30 kg/m² with at least one weight-related condition, such as hypertension, dyslipidaemia, obstructive sleep apnoea, cardiovascular disease, prediabetes or type 2 diabetes^[3]. For sleep apnea it points to the trial results in section 5.1^[3]. So the approved use is weight management, and Lilly’s medical information for Spain says plainly that tirzepatide isn’t indicated to treat obstructive sleep apnea^[11].
• Mexico. Lilly México announced COFEPRIS approval of tirzepatide for type 2 diabetes^[12]. In May 2025, a Lilly de México press release carried by enFarma announced its authorization in Mexico for weight management^[13]. These are company announcements: I couldn’t check the full COFEPRIS registration, and neither mentions sleep apnea. Until an official source says otherwise, I don’t treat sleep apnea as an approved use there.

The EU label reports different numbers for the same trial: an AHI change of −27.4 vs. −4.8 events per hour in Study 1 (difference −22.5) and −30.4 vs. −6.0 in Study 2 (difference −24.4), and weight changes of −18.1% vs. −1.3% and −20.1% vs. −2.3%^[3]. For the AHI, these match the efficacy estimand in the paper^[5]. The EU label doesn’t name the analysis, so if two sources disagree, check which one each uses.

Who decides, and what to ask

Sleep apnea is diagnosed and followed by a doctor, usually with a sleep medicine specialist. The SEPAR consensus asks for the diagnosis to be confirmed with a validated sleep study before any treatment, and considers CPAP still the treatment of choice for many patients^[10]. In SURMOUNT-OSA’s discussion, the authors note that the American Academy of Sleep Medicine’s guidelines prioritize PAP for people with symptoms^[5].

If you already use CPAP, any change is something to discuss with your sleep doctor: the Zepbound label itself says the studies didn’t evaluate when to stop it^[1]. A few questions worth bringing:

• How severe is my sleep apnea, looking at the AHI and also at oxygen levels and sleepiness?
• Could treating my weight help in my case, and with what options?
• How would a drug like Zepbound fit with my CPAP or other treatments?
• When should the sleep study be repeated to see if anything has changed?

You can put them together and print them with the doctor visit checklist.

Frequently asked questions

Is Zepbound FDA-approved for sleep apnea?

Yes. Since December 20, 2024, Zepbound has been approved for moderate to severe obstructive sleep apnea in adults with obesity, with a reduced-calorie diet and more physical activity^[1,6]. It isn’t approved for sleep apnea in people without obesity^[1].

Does Zepbound cure sleep apnea?

Neither the label nor the trial talks about a cure. In SURMOUNT-OSA, after 52 weeks, 42.2% and 50.2% of people on tirzepatide reached remission or mild sleep apnea without sleepiness, vs. 15.9% and 14.3% on placebo^[1]. The rest didn’t get there, and the trial didn’t measure what happens after stopping^[5].

Can Zepbound replace a CPAP machine?

The trial doesn’t show that: it compared tirzepatide with placebo, not with CPAP^[4]. The label says the studies didn’t evaluate when or whether to stop PAP, and the EU label says the CPAP trial allows no conclusion about added benefit on top of PAP^[1,3]. That’s a decision for your sleep doctor.

What dose of Zepbound is used for sleep apnea?

The label’s maintenance dose is 10 mg or 15 mg once weekly, after starting at 2.5 mg for 4 weeks and going up in 2.5 mg steps^[1]. The trial used whichever of those two was the highest dose each person tolerated^[4]. The actual dose is up to the prescriber.

Is Mounjaro approved for sleep apnea?

Not in the US: its label covers type 2 diabetes and cardiovascular risk in type 2 diabetes^[2]. Not in the EU either, where Mounjaro covers diabetes and weight management, with sleep apnea listed as an example of a weight-related condition and the trial results in section 5.1^[3].

My CPAP feels different since I lost weight. Why?

Weight loss can change what your airway needs at night. The American Academy of Sleep Medicine’s guideline on the long-term care of obstructive sleep apnea says that after substantial weight loss, 10% or more of body weight, a follow-up sleep study is routinely indicated to check whether PAP is still needed or whether its pressure needs adjusting^[15]. The guideline dates from 2009, before these medicines, and doesn’t describe how a mismatch feels. In SURMOUNT-OSA, average weight loss on tirzepatide was well above that 10% mark^[3]. If your CPAP feels different or your symptoms come back, tell your sleep doctor: changing the pressure or stopping is their call, as what the trial doesn’t show explains.