Top PubMed Studies on GLP-1 Drugs for Obstructive Sleep Apnea (2026)

SURMOUNT-OSA is the pivotal trial that drove the December 2024 FDA approval of Zepbound for obstructive sleep apnea — the first medication of any class approved for OSA. It enrolled 469 adults with moderate-to-severe OSA and obesity across two parallel phase 3 trials: Trial 1 in patients not using CPAP, Trial 2 in established CPAP users. Tirzepatide reduced apnea-hypopnea index by 25.3-29.3 events per hour versus 5.3-5.5 on placebo at week 52, a roughly 55-60% relative reduction, alongside 17-20% body-weight loss. Daytime sleepiness, hypoxic burden, and patient-reported sleep disturbance all improved.

PMID 38912654 ↗NCT05412004 ↗DOI 10.1056/NEJMoa2404881 ↗

This pre-specified secondary analysis of SURMOUNT-OSA addressed the cardiovascular risk profile that defines clinical OSA management. Tirzepatide produced 7-9 mmHg reductions in systolic blood pressure, large reductions in hypoxic burden (a stronger CV-mortality predictor than AHI alone), and substantial drops in hsCRP, triglycerides, and HOMA-IR. The cardiometabolic benefit tracked weight loss but exceeded what historical CPAP trials have produced, supporting the hypothesis that pharmacologic weight loss may reduce OSA-attributable cardiovascular events — a hypothesis a dedicated phase 3 CV outcomes trial in OSA has not yet tested.

PMID 41540105 ↗NCT05412004 ↗DOI 10.1038/s41591-025-04071-1 ↗

SCALE Sleep Apnea was the first dedicated GLP-1 OSA trial — 359 adults with obesity and moderate-to-severe OSA who were unwilling or unable to use CPAP were randomized to liraglutide 3.0 mg daily or placebo for 32 weeks. Liraglutide reduced AHI by 12.2 events per hour versus 6.1 on placebo (estimated treatment difference -6.1, p=0.015), alongside 5.7% body-weight loss versus 1.6% on placebo. Although the effect size was modest compared to tirzepatide's SURMOUNT-OSA data eight years later, SCALE Sleep Apnea established that pharmacologic weight loss reduces AHI independently of CPAP and seeded the indication-development pathway.

PMID 27005405 ↗NCT01557166 ↗DOI 10.1038/ijo.2016.52 ↗

Sleep AHEAD is the foundational pre-GLP-1 randomized weight-loss-and-OSA trial — a substudy of the Look AHEAD diabetes trial. 264 adults with obesity, type 2 diabetes, and OSA were randomized to intensive lifestyle intervention (low-calorie diet, 175 min/week activity) or diabetes support and education for one year. The intervention group lost 10.8 kg versus 0.6 kg and reduced AHI by 5.4 events per hour versus +4.2 in controls. The trial established that every kilogram lost corresponds to roughly one event per hour reduction in AHI — the dose-response anchor that all later GLP-1 OSA trials are calibrated against.

PMID 19786682 ↗NCT00194259 ↗DOI 10.1001/archinternmed.2009.266 ↗

This four-year follow-up of the Sleep AHEAD cohort showed that even modest sustained weight loss continues to suppress AHI long after the initial lifestyle intervention ends. Among 264 participants, the intensive-lifestyle group maintained a 5.0-kg advantage over controls at year 4 and a 2.5 events/hour AHI advantage. OSA remission was three times more common in the intervention arm. The paper is the durability counterpart to the original Foster 2009 readout and frames a key question for chronic GLP-1 therapy: maintenance of weight loss, not just initial drop, governs long-term respiratory outcomes.

PMID 23633746 ↗NCT00194259 ↗DOI 10.5665/sleep.2618 ↗

This Penn/Schwab MRI study answered the mechanistic question of why weight loss improves OSA. Among 67 adults with obesity and OSA who lost a mean 9.5% body weight, AHI dropped 31%. Multivariable analysis showed reduction in tongue fat — not total body weight, BMI, or neck circumference — was the primary anatomic mediator of AHI improvement. The finding reframed obesity-OSA biology around regional fat depots in the upper airway and is now invoked to explain why GLP-1-class drugs, which preferentially reduce visceral and ectopic fat, produce AHI reductions disproportionate to their total-weight effect.

PMID 31918559 ↗DOI 10.1164/rccm.201903-0692OC ↗

This proof-of-concept RCT randomized 30 adults with newly-diagnosed OSA and obesity to CPAP, liraglutide 3.0 mg, or combined CPAP plus liraglutide for 24 weeks. Liraglutide produced 6.3% weight loss and a 33% reduction in AHI, but only the CPAP arms improved flow-mediated dilatation, a marker of early endothelial dysfunction. The study is small but the only randomized head-to-head between CPAP and GLP-1 monotherapy with a vascular endpoint, and it tempers any premature claim that GLP-1 weight loss can fully substitute for airway-pressure therapy in vascular-risk terms — at least in the first six months.

PMID 38096106 ↗NCT04186494 ↗DOI 10.1513/AnnalsATS.202309-821OC ↗

STEP-HFpEF is included here as adjacent-population context: HFpEF and OSA overlap in roughly 50-80% of obese phenotypes, and many SURMOUNT-OSA enrollees would also meet STEP-HFpEF criteria. The trial randomized 529 patients with HFpEF (LVEF ≥45%) and obesity (BMI ≥30) to semaglutide 2.4 mg weekly or placebo for 52 weeks. Semaglutide improved KCCQ-CSS by 7.8 points more than placebo and reduced body weight by 10.7 percentage points more. The findings imply much of the OSA-related cardiovascular benefit of weight loss is mediated by HFpEF mechanisms that GLP-1s also address.

PMID 37622681 ↗NCT04788511 ↗DOI 10.1056/NEJMoa2306963 ↗

This meta-analysis pooled randomized data across the major GLP-1 OSA trials — SURMOUNT-OSA, SCALE Sleep Apnea, and smaller liraglutide and semaglutide studies — covering roughly 900 patients. Across the pooled cohort, GLP-1-class therapy reduced AHI by a weighted mean of about 11.6 events per hour versus controls and body weight by 11.1%. Subgroup analysis confirmed tirzepatide produced the largest effect and that CPAP-naive patients responded more than CPAP-established patients. The paper is the most-cited class-level summary in the field as of 2026 and the canonical reference for class-effect arguments in coverage policy.

PMID 39626095 ↗DOI 10.1093/sleep/zsae285 ↗