Tirzepatide and Obstructive
Sleep Apnea: What the Evidence Shows

The approval is specific: branded Zepbound, adults, obesity, and moderate-to-severe OSA

FDA approved branded Zepbound for moderate-to-severe OSA in adults with obesity on December 20, 2024, alongside reduced-calorie eating and increased physical activity. The label lists 10 mg or 15 mg weekly as OSA maintenance doses; 2.5 mg is an initiation dose. Dosing and escalation belong to the prescriber—not this trial summary.^{[1] [2]}

This is not an approval for every product containing tirzepatide. It does not transfer automatically to a different brand, a different dose or presentation, or an unapproved compounded preparation. FDA also says Zepbound should not be coadministered with another tirzepatide-containing product or any GLP-1 receptor agonist.^{[1] [8]}

What AHI measures—and why a sleep evaluation still matters

OSA occurs when the upper airway becomes blocked repeatedly during sleep. AHI is the number of apneas (breathing pauses) and hypopneas (shallow-breathing episodes) per hour of sleep. In the Zepbound OSA trials, moderate-to-severe disease meant an AHI of at least 15 events per hour. AHI is an outcome from sleep testing; it is not something a person can reliably establish from snoring, a wearable score, or daytime tiredness alone.^{[1] [3]}

NHLBI says a clinician considers symptoms, risk factors, and family history, and may refer a person for a sleep study to identify the type and seriousness of sleep apnea. This matters because central sleep apnea is not OSA, and SURMOUNT-OSA excluded central or mixed sleep apnea. A medicine discussion should follow an appropriate evaluation rather than substitute for one.^{[4] [5]}

What SURMOUNT-OSA actually tested

SURMOUNT-OSA comprised two 52-week randomized, double-blind, placebo-controlled trials involving 469 adults with obesity and moderate-to-severe OSA. Participants received weekly tirzepatide, escalated to a maximum tolerated 10 mg or 15 mg, or placebo. Everyone received diet and activity counseling. Type 2 diabetes was excluded, limiting direct generalization to that population.^{[1] [3]}

Study 5 enrolled 234 people unable or unwilling to use PAP. Study 6 enrolled 235 people already using PAP and intending to continue. The primary endpoint was laboratory-measured AHI change at week 52. Secondary outcomes included weight, oxygen-related hypoxic burden, defined AHI responses and sleep-related symptoms.^{[1] [3]}

The principal results, with groups and denominators

The FDA label reports results in the modified intention-to-treat population: randomized participants who received at least one study dose. These are group averages and percentages after 52 weeks, not predictions for an individual. The placebo groups also received diet and activity counseling, and they had smaller average AHI reductions.^[1]

The AHI-change row compares a change in breathing events per hour; it is not a percentage of people “cured.” The combined AHI-and-sleepiness category is also not a guarantee of absence of OSA. It describes the trial’s defined endpoint: AHI below 5, or AHI from 5 to 14 with an Epworth Sleepiness Scale score of 10 or less. The label says results used multiple imputation for missing week-52 data, and categorical percentages were calculated from imputed datasets.^[1]

The publication also reported improvements in hypoxic burden and pooled sleep-related patient-reported outcomes. These group findings do not establish an individual’s alertness, driving safety, cardiovascular outcome or need for another OSA treatment.^[3]

PAP was not replaced by the trial—and should not be stopped without clinical direction

PAP uses pressurized air during sleep to help keep the airway open. NHLBI calls PAP the most common treatment for sleep apnea and describes CPAP, BPAP, and APAP as different forms. The AASM PAP guideline recommends PAP rather than no therapy for adults with OSA and excessive sleepiness, and calls for objective testing, follow-up, troubleshooting, and monitoring of efficacy and use data.^{[6] [9]}

The PAP cohort is important evidence that tirzepatide was studied alongside ongoing PAP care. Participants in Study 6 temporarily suspended PAP for seven days before designated study assessments so researchers could measure sleep-disordered breathing without PAP’s immediate effect. That protocol procedure is not a self-management instruction. The current FDA label explicitly says the OSA studies did not evaluate the timing or appropriateness of PAP discontinuation in previously compliant patients. Do not stop, pause, or change PAP based on symptom improvement or a medication response without the sleep-care team’s direction and appropriate reassessment.^{[1] [3] [7] [9]}

What the evidence does not show

The 52-week trials did not establish long-term cardiovascular outcomes or study PAP adherence. Participants were mostly male and White, with mean baseline BMI near 39 and AHI near 50 events per hour. They excluded people without obesity, limiting generalization to all adults with OSA.^[3]

The trials also cannot distinguish exactly how much improvement came from weight reduction, other effects of treatment, or the full trial context. FDA’s approval announcement says the OSA improvement is likely related to body-weight reduction. It is more accurate to say the trials found improvement with the defined regimen than to promise airway “cure,” direct anatomical repair, or a replacement for comprehensive sleep-apnea care.^{[2] [3]}

Safety screening and follow-up remain part of the evidence

Zepbound has a boxed warning about thyroid C-cell tumors observed in rats; the human relevance is unknown. It is contraindicated for a personal or family history of medullary thyroid carcinoma or for multiple endocrine neoplasia syndrome type 2, and for known serious hypersensitivity to tirzepatide or its excipients. The label also addresses severe gastrointestinal reactions, volume-depletion-related acute kidney injury, gallbladder disease, pancreatitis, hypersensitivity, hypoglycemia with insulin or an insulin secretagogue, diabetic-retinopathy complications in people with type 2 diabetes, and pulmonary aspiration risk during anesthesia or deep sedation.^[1]

Gastrointestinal effects were most frequent during dose escalation. In Study 5, diarrhea occurred in 26.3% of tirzepatide participants versus 12.5% with placebo. Two adjudication-confirmed pancreatitis cases occurred in Study 6’s tirzepatide group. Trial safety findings do not replace individual risk assessment and label-directed follow-up.^[3]

Branded-product evidence is not compounded-product evidence

The approval, label, and SURMOUNT-OSA data concern FDA-approved Zepbound. FDA says unapproved versions of GLP-1 drugs, including tirzepatide, do not undergo FDA review for safety, effectiveness, or quality before marketing. FDA says compounded drugs should be used only when a patient’s medical needs cannot be met by an FDA-approved drug. A compounded preparation should not be called “the same as” Zepbound, and these trials do not establish its quality, bioequivalence, dosing behavior, safety, or OSA outcomes.^[8]

A safe conversation starts with verified diagnosis and coordinated monitoring

Discuss the sleep-study diagnosis, PAP use, symptoms, medicines, contraindications, pregnancy considerations and planned procedures with the care team. Follow-up should assess tolerability and whether repeat testing is appropriate. AASM guidance allows follow-up sleep testing after clinically significant weight change; it supports clinician-led reassessment, not unadvised device changes.^{[5] [7]}

For related reading, see what tirzepatide is, GLP-1 real-world effectiveness, and Luma’s treatment information.

Frequently asked questions