Tirzepatide for Sleep Apnea: The Zepbound OSA Approval

Two parallel trials, 469 adults across nine countries, all with moderate-to-severe OSA and obesity, followed for a year. The first enrolled people who could not or would not use a CPAP machine; the second enrolled people already using one who intended to continue. Each group was randomized to weekly tirzepatide or placebo.

The measure that matters in sleep apnea is the apnea-hypopnea index — the number of times per hour breathing stops or shallows. On tirzepatide, that number fell by about 25 events per hour in the non-CPAP trial and 29 in the CPAP trial, against 5 and 6 on placebo. Put differently, severity fell by roughly half to two-thirds.

After a year, 42% of the non-CPAP group and 50% of the CPAP group had either mild OSA or none at all, against 16% and 14% on placebo. Weight fell by 18–20%.

Side effects were the class's usual gastrointestinal ones. The label carries the same warnings as for any use of tirzepatide.

The indication is specific: adults with moderate-to-severe OSA and obesity. Read both halves.

It does not cover mild OSA, where the evidence is thinner and the trade-off different. It does not cover sleep apnea in people of normal weight, where the airway problem is usually anatomical rather than adipose and weight loss does little. And it is not an indication for tirzepatide instead of diagnosis — the approval assumes the diagnosis has been made properly, which means a sleep study.

That last point is the one that matters most to how we work. We screen for sleep-disordered breathing with a validated questionnaire and refer for a sleep study. We do not diagnose OSA and we will not prescribe tirzepatide on the assumption that someone has it. If you snore, wake unrefreshed and carry weight, the sequence is: study, diagnosis, then a conversation about what to treat it with.

The CPAP trial is the interesting one. Participants on tirzepatide kept using their machines and still saw a larger drop in AHI than the non-CPAP group. Two conclusions follow.

First, the medication does not replace CPAP on day one. The weight comes off over months, and the airway improves behind it. Someone with severe OSA who stops their machine the week they start tirzepatide has traded a treatment that works tonight for one that will work next year.

Second, for the many people who own a CPAP and cannot tolerate it — the first trial's population, and in our experience the larger one — the medication offers a route to treating the condition that does not depend on a mask they will not wear. That is the population the approval most changes things for.

A sleep physician decides when, and whether, CPAP can be stepped down. We do not.

Semaglutide produces weight loss and will improve sleep apnea for the same reason. It has not been trialled for OSA as an endpoint and does not carry the indication. For a patient whose sleep apnea is the main driver of the decision, the trial evidence and the label both favor tirzepatide; for a patient whose cardiovascular history is the main driver, the SELECT trial points the other way. That is what a clinician is for.