Tirzepatide Cuts Sleep Apnea Severity in Obesity—With Limits

Tirzepatide has changed the treatment conversation for adults with obesity and moderate-to-severe obstructive sleep apnea (OSA). In two 52-week phase 3 trials, the weekly drug reduced the apnea-hypopnea index (AHI) by roughly 22 events per hour more than placebo. That is the largest pharmacologic effect reported for this population, and it came with weight loss, lower hypoxic burden, lower high-sensitivity C-reactive protein (hsCRP), and lower systolic blood pressure.

But an AHI reduction is not a permission slip to abandon CPAP. A small randomized study found that CPAP improved early cardiovascular disease markers whereas GLP-1-mediated weight loss did not. Tirzepatide may be a major option for people with obesity who cannot use positive airway pressure, but no trial has shown that it provides cardiovascular protection equivalent to CPAP.

The phase 3 result is large, but applies to a specific group

Malhotra et al.'s 2024 NEJM trial, known as SURMOUNT-OSA, tested tirzepatide in two parallel, multicenter, double-blind, placebo-controlled trials lasting 52 weeks. Participants had obesity, defined as a body-mass index of at least 30 kg/m², and moderate-to-severe OSA, defined as an AHI of at least 15 events per hour. Each trial had roughly 230 participants per arm.

Study 1 enrolled people who had not used positive airway pressure (PAP) or could not tolerate it. Study 2 enrolled established PAP users, who stopped PAP for seven days before sleep-study assessments so investigators could measure the drug's effect without PAP masking it. Participants received tirzepatide titrated to their maximum tolerated dose, either 10 or 15 mg weekly, or placebo.

The primary endpoint was change in AHI. Tirzepatide reduced AHI by approximately 22 events per hour versus placebo and significantly improved the proportions reaching an AHI below 5 events per hour and at least a 50% AHI reduction. Those are meaningful sleep-breathing outcomes, not merely a flattering change in a blood test.

The trial also found significant reductions in body weight, hypoxic burden, hsCRP, and systolic blood pressure. These outcomes moved together, but the study does not prove which change caused which. Weight loss is the leading explanation for the airway benefit; whether dual GIP/GLP-1 signaling adds an independent anti-inflammatory or vascular effect remains unanswered.

People also reported better sleep-related and daytime outcomes. In a 2025 secondary analysis, Kanu et al. found statistically significant improvements at weeks 20 and 52 across PROMIS sleep-related impairment and sleep disturbance measures, the Epworth Sleepiness Scale, FOSQ-10, SF-36v2, EQ-5D-5L, and global impression scales. The seven-day PAP withdrawal in Study 2 helps separate those reports from the immediate effects of wearing PAP. Still, statistical significance does not settle clinical importance: validated minimum clinically important difference thresholds for the PROMIS sleep measures remain incomplete in OSA populations.

Among diabetes drugs, tirzepatide ranks first—indirectly

A 2026 systematic review and network meta-analysis by Yang et al. combined 8 randomized trials and 2 nonrandomized studies, totaling 1,280 participants. It estimated that tirzepatide reduced AHI by 21.85 events per hour versus control (95% CI, −27.52 to −16.34). Its direct pairwise estimate was similar: −21.89 events per hour (95% CI, −26.01 to −17.77).

The same analysis estimated smaller AHI reductions for SGLT-2 inhibitors, −7.73 events per hour (95% CI, −9.83 to −5.64), and GLP-1 receptor agonists such as liraglutide and semaglutide, −5.19 events per hour (95% CI, −6.87 to −3.51). Tirzepatide also ranked highest for weight loss at −19.41 kg (95% CI, −21.00 to −17.82), along with reductions in systolic and diastolic blood pressure.

The estimates make tirzepatide the leading drug option in this evidence base. They do not establish a final league table. No head-to-head OSA trial has compared tirzepatide with an SGLT-2 inhibitor or a GLP-1 receptor agonist, so the class ranking rests on indirect comparisons across different studies and populations. SGLT-2 inhibitors appeared to reduce AHI more than GLP-1 receptor agonists despite less weight loss. Fluid handling and reduced overnight fluid shift are plausible explanations, but they remain hypotheses rather than demonstrated OSA mechanisms.

Why a month of keto did not improve AHI beyond CPAP

Rapid weight loss did not automatically produce extra sleep-apnea benefit in a small dietary trial. Schiavo et al. randomized 70 people with severe OSA, extreme obesity, and planned bariatric surgery to four weeks of CPAP plus a low-calorie ketogenic diet (LCKD) or CPAP alone. The LCKD group fell from 143.6 to 129.7 kg, about a 9.7% reduction, with p=0.0052. The CPAP-only group changed from 132.7 to 131.6 kg, which was not statistically significant (p=0.816).

Despite that difference in weight, the primary endpoint—an incremental AHI reduction beyond CPAP—was not statistically significant. A trend toward lower C-reactive protein in the diet group likewise did not translate into an additional AHI benefit.

This pilot cannot show that a ketogenic diet never helps OSA. It was short, small, conducted in a presurgical population with extreme obesity, and used CPAP in both groups, which may have left little room for an additional AHI improvement. The more modest conclusion is the useful one: four weeks of a low-calorie ketogenic diet has not shown an OSA-specific benefit beyond CPAP. Early losses on such diets can include glycogen and water; whether sustained calorie restriction over months reduces pharyngeal fat enough to further improve OSA needs direct trials.

CPAP still has a job the drug has not proved it can do

CPAP prevents airway collapse while it is worn. Tirzepatide appears to reduce the conditions that make collapse more likely, but the two strategies do not yet have interchangeable evidence. In a randomized proof-of-concept trial, O'Donnell et al. found that CPAP improved early cardiovascular disease markers while GLP-1-mediated weight loss did not. The trial does not directly test tirzepatide, and it does not provide cardiovascular outcomes data. It does show why a lower AHI alone cannot be assumed to deliver the same cardiovascular result as PAP.

A 2025 meta-analysis of CPAP, oral appliances, and positional therapy is useful context for established OSA treatments, but it does not justify calling tirzepatide "CPAP-level" therapy here: a comparable numerical CPAP effect estimate for moderate-to-severe OSA is not available from the reported evidence.

Who should consider tirzepatide?

The best-supported use is in adults with obesity and moderate-to-severe OSA who cannot tolerate or adhere to PAP, under clinician supervision. For them, tirzepatide has phase 3 evidence of substantial AHI reduction. It may also complement PAP, but treatment changes should follow reassessment of OSA severity and cardiovascular risk—not a scale reading or a wearable's congratulatory fireworks.

• Do not generalize these results to non-obese OSA. The trials reviewed enrolled people with BMI of at least 30 kg/m²; craniofacial anatomy, airway muscle response, and arousal threshold can drive OSA in leaner people.
• Expect tradeoffs. Dose-dependent nausea, vomiting, diarrhea, and constipation occurred across the SURMOUNT program, and gastrointestinal intolerance led some participants to stop treatment. The available reports do not provide exact SURMOUNT-OSA rates here.
• Account for access. Cost, insurance coverage, prior authorization, and local availability can determine whether treatment is realistic. No OSA-specific cost-effectiveness analysis is available in this evidence.
• Treat 52 weeks as one year of evidence, not a lifetime guarantee. The durability of AHI improvement, the need for indefinite dosing, and the effect of stopping tirzepatide remain unknown.

Tirzepatide has strong evidence for reducing OSA severity in adults with obesity. It has weaker evidence for ranking against other drug classes and no evidence yet that it replaces CPAP's cardiovascular effects.

SURMOUNT-OSA was designed, funded, and statistically analyzed by Eli Lilly, with sponsor-employed medical writing support. It also excluded people with type 1 or type 2 diabetes and those with recent weight change greater than 5 kg, limiting how neatly its results map onto everyday sleep-clinic populations. Independent replication and cardiovascular outcomes trials would answer the questions that matter next.