SURMOUNT-OSA deep dive: tirzepatide for obstructive sleep apnea (Malhotra 2024 NEJM)
Last verified 2026-05-28 · Phase 3 · Completed; primary results published June 2024 · NCT05412004 ↗
SURMOUNT-OSA (NCT05412004) is the pair of phase-3 trials that made obstructive sleep apnea the third FDA-approved indication for tirzepatide. Eli Lilly ran two cohorts under one master protocol: SURMOUNT-OSA-1 randomized 234 adults with moderate-to-severe OSA and obesity not using positive airway pressure (PAP) therapy; SURMOUNT-OSA-2 randomized 235 adults with the same diagnosis already established on PAP. Both cohorts received weekly subcutaneous tirzepatide at maximum tolerated dose (10 or 15 mg) or matching placebo for 52 weeks. The primary endpoint was change from baseline in the apnea-hypopnea index (AHI, events per hour) by overnight polysomnography. Results published by Malhotra and colleagues in the New England Journal of Medicine on June 21, 2024 supported the December 20, 2024 FDA label expansion of Zepbound to moderate-to-severe OSA in adults with obesity.
- Enrollment
- 469
- Duration
- 52 weeks of randomized treatment under a 24-week dose-escalation schedule, plus a 4-week safety follow-up
- Drug
- Tirzepatide (Zepbound)
- Population
- Adults aged 18 or older with a body-mass index of at least 30 kg/m² and a diagnosis of moderate-to-severe obstructive sleep apnea (apnea-hypopnea index ≥15 events/hour on screening polysomnography). Type 2 diabetes was an exclusion. SURMOUNT-OSA-1 enrolled adults unable or unwilling to use positive airway pressure therapy; SURMOUNT-OSA-2 enrolled adults established on PAP therapy for at least 3 months at screening. Across both cohorts the mean baseline AHI was approximately 51 events/hour, mean body weight approximately 116 kg, mean BMI approximately 39 kg/m², and roughly 70% of participants were male — the reverse of the female-skewed obesity trial populations.
Primary endpoint
Change from baseline in apnea-hypopnea index (AHI, events/hour) at week 52 — measured by overnight polysomnography
Treatment arm
Cohort 1 (without PAP): −25.3 events/hour; Cohort 2 (with PAP): −29.3 events/hour
Comparator
Cohort 1: −5.3 events/hour; Cohort 2: −5.5 events/hour
Treatment difference: Estimated treatment difference vs placebo: cohort 1 −20.0 events/hour (95% CI −25.8 to −14.2); cohort 2 −23.8 events/hour (95% CI −29.6 to −17.9); P<0.001 for both
Baseline AHI of approximately 51 events/hour fell by roughly 25-29 events/hour on tirzepatide vs roughly 5 events/hour on placebo — a magnitude that moved most participants from severe OSA (≥30 events/hour) toward mild or remission (<15 events/hour).
Secondary endpoints
| Endpoint | Treatment | Comparator | Difference |
|---|---|---|---|
| Proportion achieving AHI reduction ≥50% from baseline at week 52 More than half of cohort-1 participants and nearly two-thirds of cohort-2 participants achieved a halving of their AHI — the clinical threshold for meaningful OSA response. | Cohort 1: 51.5%; Cohort 2: 65.2% | Cohort 1: 13.6%; Cohort 2: 17.2% | Cohort 1 odds ratio favoring tirzepatide; cohort 2 similar; P<0.001 for both |
| Proportion achieving disease resolution (AHI <5 events/hour or AHI 5-14 with no daytime sleepiness) at week 52 Roughly half of tirzepatide participants met the prespecified definition of OSA disease resolution — a regulatory anchor for the FDA approval. No prior drug has produced this outcome in a randomized trial. | Cohort 1: 43.0%; Cohort 2: 51.5% | Cohort 1: 14.9%; Cohort 2: 13.6% | P<0.001 for both cohorts |
| Percent change in body weight from baseline at week 52 Weight loss in SURMOUNT-OSA was similar in magnitude to SURMOUNT-1 (−20.9% at 72 weeks in adults with obesity without diabetes), confirming that the AHI benefit was layered on top of substantial weight reduction. | Cohort 1: −18.1%; Cohort 2: −20.1% | Cohort 1: −1.3%; Cohort 2: −2.3% | Cohort 1: estimated treatment difference −16.8 percentage points (95% CI −19.3 to −14.4); cohort 2: −17.8 percentage points (95% CI −20.3 to −15.3); P<0.001 for both |
| Change in hypoxic burden at week 52 (% min/hour, area-under-curve of oxygen desaturation) Hypoxic burden integrates the depth and duration of overnight oxygen desaturations and is a stronger predictor of OSA-related cardiovascular mortality than AHI alone. | Cohort 1: −79.3%; Cohort 2: −78.4% (relative reductions) | Cohort 1: −19.0%; Cohort 2: −15.6% | Treatment-difference ratios favored tirzepatide in both cohorts; P<0.001 |
| Change in PROMIS Sleep-Related Impairment T-score at week 52 (patient-reported daytime sleepiness) Patient-reported daytime functional impairment improved on the validated PROMIS Sleep-Related Impairment instrument; the magnitude exceeds the minimal clinically important difference of approximately 2 points. | Cohort 1: −5.0 points; Cohort 2: −5.0 points (improvement) | Cohort 1: −2.3 points; Cohort 2: −1.4 points | Estimated treatment differences −2.6 points (cohort 1) and −3.5 points (cohort 2); P<0.001 for both |
| Change in high-sensitivity C-reactive protein (hsCRP) at week 52 (ratio to baseline) Systemic inflammation, elevated in OSA and obesity, fell substantially on tirzepatide. Consistent with hsCRP reductions seen in STEP-1, SURMOUNT-1, and SELECT. | Cohort 1: 0.63 (37% reduction); Cohort 2: 0.59 (41% reduction) | Cohort 1: 0.97; Cohort 2: 0.96 | Cohort 1: estimated treatment ratio 0.65; cohort 2: 0.61; nominal P<0.001 in both |
| Change in systolic blood pressure at week 52 (mmHg, measured by 24-hour ambulatory monitoring) Magnitude comparable to a low-to-moderate-dose antihypertensive added to standard care. OSA-related sympathetic activation is a major driver of resistant hypertension; the drop here is clinically meaningful. | Cohort 1: −9.7 mmHg; Cohort 2: −7.6 mmHg | Cohort 1: −2.5 mmHg; Cohort 2: −1.0 mmHg | Estimated treatment differences −7.2 mmHg (cohort 1) and −6.6 mmHg (cohort 2); P<0.001 for both |
| Change in HbA1c at week 52 (percentage points) Participants were non-diabetic at entry; HbA1c shifted from upper-normal toward mid-normal range, consistent with the prediabetes-prevention pattern seen across SURMOUNT-1. | Cohort 1: −0.32; Cohort 2: −0.35 | Cohort 1: −0.05; Cohort 2: −0.10 | Estimated treatment differences approximately −0.27 (cohort 1) and −0.25 (cohort 2) percentage points; P<0.001 |
Adverse events
| Event | Treatment rate | Comparator rate |
|---|---|---|
| Diarrhea Most common gastrointestinal AE in both cohorts; clustered during the 24-week dose-escalation phase and generally mild-to-moderate. | Cohort 1: 21.6%; Cohort 2: 21.9% | Cohort 1: 5.9%; Cohort 2: 5.9% |
| Nausea Second-most-common AE; typically transient and concentrated during titration. | Cohort 1: 25.4%; Cohort 2: 18.6% | Cohort 1: 6.8%; Cohort 2: 9.3% |
| Vomiting | Cohort 1: 11.0%; Cohort 2: 7.6% | Cohort 1: 2.5%; Cohort 2: 4.2% |
| Constipation | Cohort 1: 11.9%; Cohort 2: 12.7% | Cohort 1: 2.5%; Cohort 2: 6.8% |
| Injection-site reaction | Cohort 1: 4.2%; Cohort 2: 4.2% | Cohort 1: 0.8%; Cohort 2: 0.0% |
| COVID-19 (any) Background COVID-19 incidence; rates were broadly similar between arms and not attributable to treatment. | Cohort 1: 15.3%; Cohort 2: 16.1% | Cohort 1: 16.1%; Cohort 2: 18.6% |
| Serious adverse events (any) Rates similar between arms; no concentration in any single organ system. | Cohort 1: 6.8%; Cohort 2: 5.1% | Cohort 1: 4.2%; Cohort 2: 5.9% |
| Permanent discontinuation due to adverse events Gastrointestinal events were the leading reason in both cohorts; absolute rates were comparable to other tirzepatide phase-3 trials. | Cohort 1: 5.9%; Cohort 2: 5.1% | Cohort 1: 1.7%; Cohort 2: 1.7% |
Clinical significance
SURMOUNT-OSA is the trial that turned obstructive sleep apnea into a pharmacologically treatable disease for the first time. Tirzepatide cut AHI by 25-29 events/hour across both cohorts — moving most participants from severe OSA toward mild or full resolution — with a 17-18 percentage-point placebo-subtracted weight loss, a 7-9 mmHg drop in systolic blood pressure, a 35-40% reduction in hsCRP, and improvements in patient-reported daytime sleepiness on the PROMIS instrument. The benefit appeared whether or not participants were using PAP therapy, which matters clinically because PAP adherence rates are estimated at roughly 50% in real-world cohorts. The December 20, 2024 FDA approval of Zepbound for moderate-to-severe OSA in adults with obesity made tirzepatide the first drug ever approved for OSA and reframed the condition as part of the cardiometabolic disease spectrum rather than as a standalone sleep disorder.
Frequently asked questions
Does insurance cover Zepbound for OSA in adults without obesity?
No. The December 2024 FDA approval is explicitly for moderate-to-severe obstructive sleep apnea in adults with obesity — defined as a body-mass index of at least 30 kg/m². The SURMOUNT-OSA trials required a BMI ≥30 at screening, so the label and most insurance prior-authorization criteria require documented obesity in addition to the OSA diagnosis. Adults with OSA but BMI under 30 are not covered under the OSA indication. Some commercial payers separately cover Zepbound for chronic weight management at BMI ≥30 or ≥27 with a weight-related comorbidity (including OSA), but Medicare Part D coverage as of 2026 still excludes anti-obesity drugs outside the OSA-specific labeled use.
How big is the AHI improvement with tirzepatide?
Large. Across both SURMOUNT-OSA cohorts, mean baseline AHI was approximately 51 events per hour — well into the severe range. After 52 weeks of tirzepatide at maximum tolerated dose, AHI fell by 25.3 events/hour in the cohort not using PAP and by 29.3 events/hour in the cohort already on PAP, versus drops of about 5 events/hour on placebo in both cohorts. Roughly half to two-thirds of tirzepatide participants achieved a ≥50% AHI reduction, and 43-52% met the prespecified definition of disease resolution (AHI <5 events/hour, or 5-14 with no daytime sleepiness). For comparison, mean AHI reductions with PAP therapy are typically 30-40 events/hour, so the magnitude is in the same range as PAP — but achieved with a weekly injection rather than a nightly mask.
Should I stop my CPAP if I lose weight on Zepbound?
Not without a repeat sleep study. SURMOUNT-OSA showed that a meaningful fraction of treated participants moved from severe OSA into the mild range or full resolution, but individual responses varied and the trial measured AHI by formal in-lab polysomnography rather than home pulse oximetry. The standard of care if a patient reaches substantial weight loss on tirzepatide and wants to discontinue PAP is a repeat overnight sleep study with the patient off PAP for the diagnostic night. Many patients in SURMOUNT-OSA-2 continued PAP through the trial; the additive benefit on PAP-using patients was as large as or larger than the cohort 1 effect, so combining the two modalities is well-supported. Stopping PAP without re-measurement risks untreated nocturnal hypoxia.
Does this apply to central (non-obstructive) sleep apnea?
No. SURMOUNT-OSA enrolled only adults with obstructive sleep apnea — diagnosed by polysomnography showing AHI ≥15 events/hour with the events meeting standard obstructive criteria. Central sleep apnea (CSA), in which the brainstem fails to trigger respiratory drive rather than the upper airway collapsing, was not studied and is not on the Zepbound label. CSA is most commonly seen in heart failure, after stroke, or with opioid use; the mechanism is unrelated to airway anatomy or obesity, so a weight-loss drug would not be expected to help. Complex sleep apnea (mixed obstructive and central events) was also excluded from SURMOUNT-OSA.
How does this compare with STEP-HFpEF for semaglutide in heart failure?
Different organ system, similar logic. STEP-HFpEF (Kosiborod 2023 NEJM) randomized adults with obesity-related heart failure with preserved ejection fraction (HFpEF) to weekly semaglutide 2.4 mg or placebo and showed substantial improvements in patient-reported symptoms (KCCQ score), exercise capacity, and weight. Both trials are part of a broader pattern in which GLP-1 and GLP-1/GIP drugs improve outcomes in obesity-driven downstream diseases — OSA, HFpEF, MACE in obesity, kidney disease in obesity. SURMOUNT-OSA is the OSA-specific trial; STEP-HFpEF is the HFpEF-specific trial. They are not interchangeable, but they argue for the same conceptual reframing: treating obesity treats the diseases obesity drives.
Are there other GLP-1 trials in OSA underway?
Yes, but SURMOUNT-OSA is the only completed phase-3 trial with a positive AHI primary endpoint as of 2026. Novo Nordisk has not yet reported a comparable head-to-head OSA trial for semaglutide; the SELECT cardiovascular-outcomes trial did report subgroup analyses suggesting OSA-related event reductions but did not measure AHI directly. The SURMOUNT-MMO morbidity-and-mortality trial of tirzepatide (NCT05556512, n=15,374) is ongoing and includes OSA-related secondary endpoints. For the foreseeable future, tirzepatide is the only drug with an FDA-approved OSA indication, and SURMOUNT-OSA is the only randomized trial supporting it.
References
- 1.Malhotra A, Grunstein RR, Fietze I, et al.; SURMOUNT-OSA Investigators. Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity. N Engl J Med. 2024. PMID: 38912654.
- 2.Eli Lilly and Company. Obstructive Sleep Apnea Master Protocol GPIF: A Study of Tirzepatide (LY3298176) in Participants With Obstructive Sleep Apnea (SURMOUNT-OSA, NCT05412004). ClinicalTrials.gov. 2024. https://clinicaltrials.gov/study/NCT05412004
- 3.U.S. Food and Drug Administration. FDA Approves First Medication for Obstructive Sleep Apnea (Zepbound, tirzepatide). FDA Press Announcement, December 20, 2024. 2024. https://www.fda.gov/news-events/press-announcements/fda-approves-first-medication-obstructive-sleep-apnea
- 4.Jastreboff AM, Aronne LJ, Ahmad NN, et al.; SURMOUNT-1 Investigators. Tirzepatide Once Weekly for the Treatment of Obesity. N Engl J Med. 2022. PMID: 35658024.