A patient presents with severe obstructive sleep apnea — apnea-hypopnea index (AHI) of 64 events per hour — and a BMI of 39 kg/m². Two CPAP trials have failed: one due to claustrophobia, one due to pressure intolerance. A mandibular advancement device produced partial relief but no resolution. Until late 2024, this clinical picture had no pharmacological pathway. The SURMOUNT-OSA trial (NCT05412004) tested a fundamentally different approach: target the adiposity driving the obstruction rather than the obstruction itself.
SURMOUNT-OSA Trial Design and Enrollment Criteria
SURMOUNT-OSA enrolled adults with moderate-to-severe obstructive sleep apnea (AHI ≥15 events/hour confirmed by polysomnography) and obesity (BMI ≥30 kg/m²) across two parallel Phase 3 randomized controlled trials. Trial 1 enrolled participants not currently using positive airway pressure (PAP) devices. Trial 2 enrolled participants actively receiving PAP therapy, with PAP suspended before polysomnographic assessments so that underlying OSA severity could be objectively measured.
Participants in both trials were randomized to subcutaneous tirzepatide (Mounjaro/Zepbound; Eli Lilly) or matching placebo, administered once weekly and dose-escalated using the standard protocol: 2.5 mg for four weeks, increasing in 2.5 mg increments every four weeks, targeting a maximum tolerated dose of 10 mg or 15 mg weekly. Treatment duration was 52 weeks. The primary endpoint in both trials was change from baseline in AHI — measured via polysomnography at week 52.
Key exclusion criteria included predominant central sleep apnea, a major cardiovascular event within 90 days prior to enrollment, severe renal impairment (eGFR <30 mL/min/1.73 m²), active malignancy, or prior bariatric surgery. These exclusions meaningfully limit direct generalizability — particularly for patients with complex sleep-disordered breathing phenotypes, advanced CKD, or post-surgical anatomy — and clinicians should factor these boundaries when evaluating individual eligibility.
Primary Endpoint: AHI Reduction in the Non-PAP Cohort (Trial 1)
Trial 1 enrolled 469 participants (mean age approximately 52 years, 66% male, mean baseline body weight 115 kg, mean baseline AHI approximately 51.5 events/hour). Tirzepatide reduced AHI by a mean of 55.0 events/hour from baseline versus 5.3 events/hour in the placebo group — a between-group difference of −49.2 events/hour (95% CI: −57.3 to −41.1; p<0.001). This effect size is large by any benchmark applied in OSA pharmacology research, a field that had not previously produced an approved systemic agent.
The proportion of participants achieving AHI below 5 events/hour — the standard polysomnographic threshold for OSA resolution — was 51.5% in the tirzepatide arm versus 13.6% in the placebo arm. An additional cohort achieved AHI in the mild range (5–14 events/hour). In aggregate, roughly 70% of tirzepatide-treated participants in Trial 1 shifted out of the moderate-to-severe AHI category by week 52 — a clinically meaningful shift for a population that entered the trial with an average of 51 obstructive respiratory events per hour of sleep.
Mean body weight declined 20.1% in the tirzepatide arm (approximately −23 kg from baseline) versus 2.3% in the placebo arm. The correlation between degree of weight loss and magnitude of AHI reduction was strong across individual participants, consistent with the mechanistic hypothesis that pharyngeal adiposity deposition and thoracic wall mass loading are primary biomechanical drivers of obesity-associated OSA. The trial was not designed or powered to isolate any direct receptor-mediated airway effect independent of weight loss — that mechanistic question remains open in the literature.
PAP-Concurrent Cohort: Trial 2 Findings
Trial 2 (n=389) addressed a distinct but equally relevant clinical question: among patients already using PAP therapy — where airway patency is mechanically maintained nightly — does substantial pharmacological weight loss produce measurable reduction in the underlying structural OSA severity? The answer was affirmative, though the absolute effect magnitude was smaller than Trial 1.
Tirzepatide reduced AHI by a mean of 29.3 events/hour versus 5.5 events/hour for placebo, yielding a between-group difference of −23.8 events/hour (95% CI: −31.4 to −16.2; p<0.001). Body weight decreased by a mean of 18.1% in the tirzepatide arm versus 1.0% in the placebo arm. The smaller absolute AHI reduction relative to Trial 1 is mechanistically plausible: PAP-using patients may represent a cohort with longer disease duration, greater degree of fixed upper airway narrowing, or neuromuscular compensatory adaptations associated with chronic intermittent hypoxia — factors that would attenuate AHI response independently of adiposity change.
For clinicians managing patients in this cohort, the data support reassessing AHI via polysomnography — or at minimum, reviewing autoPAP pressure trending — after a patient achieves clinically significant weight reduction (a reasonable threshold is ≥10–15% body weight loss). Pressure requirements may decrease meaningfully, and in a subset of patients, supervised PAP modification or discontinuation evaluation becomes appropriate. There is currently no published consensus protocol specifying the optimal timing or frequency of such reassessment.
Secondary Endpoints: Hypoxic Burden, Blood Pressure, and Patient-Reported Outcomes
Raw AHI reduction, while the regulatory primary endpoint, does not fully capture the physiological burden of sleep-disordered breathing. Hypoxic burden — quantified as the area under the oxygen desaturation curve expressed as %minutes per hour of sleep — carries stronger associations with cardiovascular mortality and neurocognitive consequence than AHI alone in the OSA outcomes literature. Tirzepatide produced statistically significant reductions in hypoxic burden in both SURMOUNT-OSA trials, directionally consistent with the degree of AHI reduction observed in each cohort.
Systolic blood pressure declined significantly in the tirzepatide arm relative to placebo in Trial 1. This finding carries clinical weight because OSA independently contributes to resistant hypertension — a condition frequently co-occurring in this patient population. Whether the observed BP reduction is attributable primarily to relief of nocturnal hypoxia, direct GLP-1 receptor activity on vascular smooth muscle tone, or systemic weight-loss effects cannot be disaggregated from the SURMOUNT-OSA dataset alone; the trial was not designed to isolate these mechanisms.
Patient-reported outcomes, assessed via the Epworth Sleepiness Scale (ESS) and the PROMIS Sleep Disturbance instrument, showed statistically significant improvement with tirzepatide versus placebo in Trial 1. This is an important confirmatory signal: objective AHI reduction is translating, at least partially, into subjective symptom relief. However, daytime sleepiness in OSA is multifactorial, and patients with residual hypersomnia after achieving objective AHI normalization warrant evaluation for comorbid sleep disorders — including idiopathic hypersomnia, circadian rhythm disruption, or mood-related fatigue — before attributing residual symptoms to incomplete OSA treatment.
Cardiometabolic secondary markers — including triglycerides, C-reactive protein, and glycemic parameters — showed favorable changes in tirzepatide-treated participants consistent with the metabolic profile documented in SURMOUNT-1 (PMID: 35916895) and SURMOUNT-2 (PMID: 36652616). These improvements occur in parallel with AHI reduction, reinforcing the compound's potential role in addressing the cardiometabolic cluster that frequently co-presents with obesity-associated OSA: hypertension, dyslipidemia, insulin resistance, and elevated inflammatory markers.
Adverse Event Profile and Tolerability in the SURMOUNT-OSA Population
The adverse event profile observed in SURMOUNT-OSA was consistent with tirzepatide's established safety signal across the SURMOUNT and SURPASS trial programs. Gastrointestinal events were the most common treatment-emergent adverse events: nausea in 13.4% of tirzepatide participants versus 2.6% for placebo; diarrhea in 10.5% versus 4.3%; constipation in 10.1% versus 4.3%; and vomiting in 6.4% versus 1.7%. The majority of GI events were mild-to-moderate in severity, most frequent during the dose-escalation phase, and largely transient.
Serious adverse events occurred in 6.4% of tirzepatide participants and 7.6% of placebo participants in Trial 1 — a non-significant difference aligned with the overall SURMOUNT safety profile. Discontinuation due to adverse events occurred in 5.8% of tirzepatide-treated participants versus 1.7% of placebo-treated participants, driven predominantly by GI intolerance during titration. No cases of confirmed acute pancreatitis were reported in the SURMOUNT-OSA publication.
Cholelithiasis warrants specific attention in this context. Rapid weight loss — regardless of mechanism — is associated with increased biliary cholesterol supersaturation and elevated gallstone formation risk. In SURMOUNT-1, cholelithiasis was reported in approximately 1.6% of participants receiving tirzepatide 15 mg versus approximately 0.4% of placebo-treated participants. SURMOUNT-OSA did not report cholelithiasis as a primary safety endpoint, but clinicians initiating tirzepatide in patients with prior gallbladder disease, cholecystectomy history, or trajectories of rapid weight loss exceeding 1–2% body weight per week should incorporate this risk into monitoring discussions.
Resting heart rate elevation — a known pharmacodynamic effect of GLP-1 receptor agonism on sinoatrial node automaticity — was observed in the tirzepatide arm, consistent with the compound class. In patients with pre-existing cardiac arrhythmias or baseline sinus tachycardia, this signal merits individual risk-benefit consideration. The FDA boxed warning for GLP-1/GIP receptor agonists regarding thyroid C-cell tumor risk — based on rodent carcinogenicity data — applies to tirzepatide; patients with a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia type 2 (MEN2) syndrome are contraindicated.
FDA Approval and Regulatory Landscape
On December 20, 2024, the U.S. Food and Drug Administration approved tirzepatide injection (Zepbound; Eli Lilly) for the treatment of moderate-to-severe obstructive sleep apnea in adults with obesity — the first drug approval ever issued specifically for OSA treatment. For decades, the condition was managed exclusively through device-based interventions (PAP therapy, mandibular advancement devices) or surgical approaches. The SURMOUNT-OSA Phase 3 data formed the primary basis for the supplemental indication.
The approved labeling specifies use "in conjunction with a reduced-calorie diet and increased physical activity," mirroring the obesity indication language. The label does not position tirzepatide as a standalone replacement for PAP therapy in patients who are tolerating it. Clinicians considering supervised PAP modification or discontinuation in patients who achieve significant AHI normalization should base those decisions on serial polysomnographic data — not on weight loss percentage alone as a surrogate.
Semaglutide 2.4 mg weekly (Wegovy; Novo Nordisk) reported positive Phase 3 data for OSA via the SCALE Sleep Apnea trial (NCT05035095), with an FDA supplemental indication under review as of early 2025. No head-to-head trial comparing tirzepatide and semaglutide in an OSA population has been conducted. Cross-trial comparisons between SURMOUNT-OSA and SCALE Sleep Apnea carry the limitations of non-equivalent patient populations, baseline AHI differences, and protocol design variation — comparative efficacy inferences remain indirect and should be treated as hypothesis-generating rather than definitive.
Clinical Implications and Unresolved Questions
The SURMOUNT-OSA findings reframe obesity-associated OSA as a condition potentially addressable through pharmacological weight loss targeting the upstream adiposity driver rather than the downstream airway obstruction. For patients with high PAP non-adherence rates — estimated at 30–60% in long-term observational data — this represents a pathway that did not previously exist at the Phase 3 evidence tier. The question is no longer whether tirzepatide affects AHI; the data answer that affirmatively. The clinical questions that remain are more nuanced.
Durability is the most clinically important unresolved issue. SURMOUNT-OSA was a 52-week trial. The SURMOUNT-4 trial (PMID: 38381878) demonstrated that participants who transitioned from tirzepatide to placebo after an initial treatment period regained approximately two-thirds of prior weight loss within 52 weeks of withdrawal. If AHI reduction in SURMOUNT-OSA is predominantly weight-mediated — which the correlation data strongly suggest — a comparable AHI re-elevation after tirzepatide discontinuation would be mechanistically expected. This trajectory has not been directly measured in an OSA-specific withdrawal cohort, and its absence is a meaningful gap for long-term management planning.
Patient selection criteria require careful application. SURMOUNT-OSA enrolled participants with mean BMI approximately 39 kg/m² whose OSA was presumed to be predominantly obesity-driven. Normal-weight OSA — estimated to represent 10–20% of the OSA patient population — involves anatomically fixed upper airway narrowing, retrognathia, adenotonsillar hypertrophy, and craniofacial skeletal factors. These phenotypes are not represented in the SURMOUNT-OSA dataset, and these patients are unlikely to achieve meaningful AHI reduction through weight-loss pharmacotherapy at any dose level.
Cardiovascular outcome data in an OSA-specific context remain absent. SURMOUNT-OSA was not powered for MACE endpoints. Whether pharmacological AHI reduction translates to reduction in hard cardiovascular event rates — a relationship that PAP therapy RCTs including SAVE (NCT00738179) have failed to demonstrate definitively despite robust AHI control — is an open research question requiring dedicated long-term outcome trial designs with adequate event rates.
For clinicians building a practical monitoring framework, the available evidence supports: baseline polysomnography before initiating tirzepatide; reassessment after achieving ≥10–15% body weight reduction or at the 52-week mark (whichever comes first); documentation of PAP pressure changes via autoPAP trending throughout treatment; and coordination with sleep medicine specialists before any supervised PAP modification is implemented. No published consensus protocol yet establishes optimal reassessment intervals in the post-approval setting — this is an area where clinical practice is outpacing formalized guidance.
This article summarizes research and does not constitute medical advice. Consult a licensed clinician for diagnosis, treatment, or any decisions about medications or supplements.