I snore heavily but have never been tested. Does the SURMOUNT-OSA result apply to me?

Not directly. Both trials enrolled adults with diagnosed moderate-to-severe obstructive sleep apnoea and obesity, so the AHI reduction of up to 62.8% describes that group and not an untested snorer. Snoring with witnessed pauses, choking at night or heavy daytime sleepiness is a reason for a sleep assessment on its own merits, whatever medication you are on.

My CPAP feels too strong now that I have lost weight. Can I turn it down?

Take it to the sleep service rather than the machine. Pressure settings were derived from a study done at your old weight, and the way that gets corrected is a repeat study, which can also show that severity has changed less than you assumed. Cutting pressure prematurely fragments sleep, and fragmented sleep raises appetite through ghrelin and leptin, which works against the treatment.

If my apnoea resolves, can I come off the drug?

The airway improvement follows the fat that was crowding it, so it is tied to the weight rather than to any lasting change in the airway itself. SURMOUNT-OSA tested the drug being taken, not stopped. Whether treatment continues is a prescriber decision, and any change to apnoea management belongs with the sleep physician after a repeat study.

GLP-1 and Sleep, Including Sleep Apnoea

What this page establishes

SURMOUNT-OSA comprised two phase 3 randomised trials in adults with moderate-to-severe obstructive sleep apnoea and obesity, one in people not using PAP therapy and one in people using it.
Tirzepatide produced a mean apnoea-hypopnoea index reduction of up to 62.8%, roughly 30 fewer events per hour of sleep versus placebo.
43.0% (Study 1) and 51.5% (Study 2) of participants on the highest tirzepatide dose met the trial's criteria for disease resolution.
Tirzepatide holds a US indication for moderate-to-severe obstructive sleep apnoea in adults with obesity under the Zepbound brand; Mounjaro-branded tirzepatide is indicated for type 2 diabetes, not OSA.

Quick answers

I snore heavily but have never been tested. Does the SURMOUNT-OSA result apply to me?
Not directly. Both trials enrolled adults with diagnosed moderate-to-severe obstructive sleep apnoea and obesity, so the AHI reduction of up to 62.8% describes that group and not an untested snorer.
Read the full answer
Snoring with witnessed pauses, choking at night or heavy daytime sleepiness is a reason for a sleep assessment on its own merits, whatever medication you are on.
My CPAP feels too strong now that I have lost weight. Can I turn it down?
Take it to the sleep service rather than the machine. Pressure settings were derived from a study done at your old weight, and the way that gets corrected is a repeat study, which can also show that severity has changed less than you…
Read the full answer
Take it to the sleep service rather than the machine. Pressure settings were derived from a study done at your old weight, and the way that gets corrected is a repeat study, which can also show that severity has changed less than you assumed. Cutting pressure prematurely fragments sleep, and fragmented sleep raises appetite through ghrelin and leptin, which works against the treatment.
If my apnoea resolves, can I come off the drug?
The airway improvement follows the fat that was crowding it, so it is tied to the weight rather than to any lasting change in the airway itself.
Read the full answer
SURMOUNT-OSA tested the drug being taken, not stopped. Whether treatment continues is a prescriber decision, and any change to apnoea management belongs with the sleep physician after a repeat study.

One sleep question on this page has a randomised trial, a published apnoea-hypopnoea index reduction and a US indication behind it. The rest do not. If you have moderate-to-severe obstructive sleep apnoea and obesity, tirzepatide has genuine outcome data: SURMOUNT-OSA found a mean AHI reduction of up to 62.8%, roughly 30 fewer breathing events per hour of sleep than placebo. If you are sleeping badly, having vivid dreams or waking at 3am since starting, the evidence base is anecdotal reports without trial-quantified rates, and this page will keep those two categories apart.

SURMOUNT-OSA, and what tirzepatide actually showed

SURMOUNT-OSA was two phase 3 randomised trials in adults with moderate-to-severe obstructive sleep apnoea and obesity. One enrolled people not using positive airway pressure therapy. The other enrolled people who were using it.

Mean AHI reduction reached up to 62.8%, about 30 fewer events per hour versus placebo. On the trial's own disease-resolution criteria, 43.0% of participants in Study 1 and 51.5% in Study 2 met resolution on the highest tirzepatide dose.

Those are large numbers for a metabolic drug acting on a mechanical airway problem, which is why the indication exists. Read disease resolution as a trial endpoint that a defined proportion of participants met, rather than as a cure rate you can expect.

This is tirzepatide-specific and brand-specific The obstructive sleep apnoea indication in the US attaches to tirzepatide under the Zepbound brand, for moderate-to-severe OSA in adults with obesity. Mounjaro-branded tirzepatide is indicated for type 2 diabetes, not OSA. Semaglutide has no equivalent randomised sleep-apnoea outcome trial and no OSA indication. Weight loss plausibly helps regardless of the molecule, but the evidence and the approval are tirzepatide's.

Why losing weight changes the airway

Obstructive sleep apnoea in people with obesity is driven substantially by fat deposited in the parapharyngeal tissue and the tongue. That tissue narrows the upper airway and makes it collapsible once muscle tone drops during sleep.

Reduce the depot and the airway is wider and less collapsible, so fewer obstructive events occur per hour. The drug is not acting on the airway. It is removing the tissue that was crowding it, which is why an AHI improvement follows a weight-loss mechanism rather than a respiratory one.

It also explains why the effect is not universal. Airway anatomy, jaw structure and muscle tone all contribute to OSA, and only the fat component responds to weight loss.

That is the honest read on why roughly half the participants on the highest dose met resolution criteria and roughly half did not. For someone whose apnoea is largely anatomical, removing the fat depot improves the picture without fixing it. Nobody can tell you in advance which group you sit in, which is the argument for a repeat sleep study rather than an assumption.

Does this replace CPAP?

SURMOUNT-OSA included an arm of people already using PAP therapy and tested tirzepatide alongside it. It did not test withdrawing PAP. There is a real difference between showing a drug works in people on PAP and showing they can stop.

Any decision to reduce or stop PAP is a sleep physician's call, and it is usually made after a repeat sleep study rather than on the basis of a number on the scale. A diagnostic sleep study describes the body you had on the night you slept in it. After substantial weight change, that description may no longer be accurate in either direction.

The practical route people take is to raise it with the sleep service after significant loss and let them decide whether a repeat study is warranted.

There is a second reason not to freelance this. Untreated apnoea fragments sleep, and fragmented sleep raises appetite through ghrelin and leptin, so stopping PAP prematurely works directly against the thing the medication is doing.

Insomnia and vivid dreams: what is actually known

Very little, and this section starts by saying so. Insomnia, vivid dreams and generally disrupted sleep are widely reported by people on GLP-1s, and they are not established adverse reactions with trial-quantified rates in the prescribing information.

That does not mean they are not happening. It means nobody has measured how often, and a page that gives you a percentage here would be inventing one. Weight loss itself changes sleep architecture, GI symptoms disturb nights, and a large energy deficit is its own physiological stressor. Any of those could plausibly sit behind the reports.

There is a plausible non-drug explanation worth ruling out first. A large energy deficit means going to bed less well fed than you used to, and hunger is a documented cause of light, fragmented sleep independent of any medication. If sleep worsened in the same week appetite fell off a cliff, those two may be the same event.

Worth mentioning to a prescriber. Not worth treating as an established drug effect, and specifically not worth treating as a reason to change anything about a dose on your own.

Night-time reflux and lying flat

This one has a labelled mechanism behind it. Delayed gastric emptying means a late meal can still be in the stomach at bedtime, and a full stomach plus a horizontal body is the standard recipe for reflux.

The effect is largest after the first dose and after each dose escalation, so a titration week is when this is most likely to show up. Finishing meals earlier in the evening is the common practical response and it is not a trial-tested intervention, so treat it as a reasonable thing many people try rather than as a recommendation.

Severe night-time reflux or vomiting is a different matter and goes to a clinician.

Overnight hypoglycaemia if you take insulin or a sulfonylurea

This is the highest-stakes item on the page. A GLP-1 on its own rarely drives glucose below normal, because its insulinotropic effect is glucose-dependent. Insulin and sulfonylureas are not glucose-dependent, and adding glucose-lowering on top of them shifts the overnight risk upward unless the doses are adjusted.

The ADA Standards of Care 2026 are explicit that when a GLP-1 RA or dual GIP/GLP-1 RA is added, sulfonylureas should be discontinued or reduced and insulin adjusted, with example figures of a 10 to 20% bolus reduction and about 10% basal reduction if HbA1c is under 7.5%.

Those adjustments are made by the prescriber at the point the GLP-1 is started or escalated. Nothing on this page is an instruction to change an insulin dose. If you take either of these medications and this conversation has not happened, that is the conversation to have.

Sweating, palpitations or confused waking Overnight sweating, palpitations, nightmares or waking confused can be signs of nocturnal hypoglycaemia if you take insulin or a sulfonylurea alongside a GLP-1. Contact your diabetes team. Lows that arrive without warning symptoms (hypoglycaemia unawareness) need urgent attention, because the usual alarm system has stopped working.

Poor sleep works against the drug

Short or fragmented sleep raises ghrelin and lowers leptin, both of which push appetite up. That is general sleep physiology rather than a GLP-1-specific finding, and it holds whether or not you are on medication.

The consequence is that a badly slept week is a week where the drug's main effect is being partially countered by your own hormones. People notice this as appetite breaking through on the days after poor sleep.

It also runs the other way in the case of apnoea. Untreated OSA fragments sleep, fragmented sleep raises appetite, and higher weight worsens the apnoea. Interrupting that loop is part of why the SURMOUNT-OSA result is interesting beyond the AHI number itself.

The practical consequence is unglamorous. Sleep is one of the few inputs here that is free, requires no prescription and pushes in the same direction as the treatment. It is also the one most likely to be sacrificed by someone busy enough to want a pharmacological solution in the first place.

Re-testing your sleep study after significant weight loss

An AHI measured at your starting weight describes airway behaviour at that weight. After a 15% or 20% change, that measurement is a historical document.

Many clinicians re-test after substantial weight change rather than assume the old number holds. That works in both directions: a repeat study is how a reduction in severity gets documented, and it is also how you find out that the severity has not changed as much as you assumed.

The reason to care is that CPAP pressure settings, mandibular device fit and treatment decisions are all built on the original number.

When to contact a clinician

The first two items on this list need attention regardless of what medication you are taking.

Witnessed apnoeas, choking or gasping at night, or heavy daytime sleepiness, which need a sleep assessment whatever else is going on
Falling asleep while driving
Overnight sweating, palpitations or confused waking if you take insulin or a sulfonylurea
Severe night-time reflux or vomiting
Morning headaches with new daytime fatigue

The evidence, one row per claim

Claim	Tier	Source
SURMOUNT-OSA comprised two phase 3 randomised trials in adults with moderate-to-severe obstructive sleep apnoea and obesity, one in people not using PAP therapy and one in people using it.	established	Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity, NEJM
Tirzepatide produced a mean apnoea-hypopnoea index reduction of up to 62.8%, roughly 30 fewer events per hour of sleep versus placebo.	established	Eli Lilly, SURMOUNT-OSA topline results
43.0% (Study 1) and 51.5% (Study 2) of participants on the highest tirzepatide dose met the trial's criteria for disease resolution.	established	Eli Lilly, SURMOUNT-OSA results
Tirzepatide holds a US indication for moderate-to-severe obstructive sleep apnoea in adults with obesity under the Zepbound brand; Mounjaro-branded tirzepatide is indicated for type 2 diabetes, not OSA.	established	SURMOUNT-OSA, NEJM; brand indications per FDA labelling
Semaglutide has no equivalent randomised sleep-apnoea outcome trial; the OSA evidence is specific to tirzepatide.	established	SURMOUNT-OSA, NEJM
Insomnia, vivid dreams and disrupted sleep are widely reported anecdotally on GLP-1s but are not established adverse reactions with trial-quantified rates.	anecdotal	WEGOVY US Prescribing Information (absence of these as labelled reactions)
The ADA Standards of Care advise reducing or stopping sulfonylureas and adjusting insulin when adding a GLP-1 RA, which is directly relevant to overnight hypoglycaemia risk.	established	ADA Standards of Care in Diabetes 2026, Section 9

Sources

Where this comes from

Researched and edited by Thomas Prommer — biohacker, performance athlete, data enthusiast and health coach.

Every number on this page traces to a named source. There are 7 sourced claims below the fold, each with the document it came from.

Sources consulted: Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity, Eli Lilly, SURMOUNT-OSA, WEGOVY US Prescribing Information, ADA Standards of Care in Diabetes 2026.

Not clinically reviewed. This page was researched and written against primary regulatory and trial sources, and no clinician has checked it.

Sources last read 2026-08-23.

Keep reading

Educational content, reviewed 2026-08-23. Not medical advice, not a prescription, and not a substitute for a clinician who knows your history. Doses named here are label schedules or doses used in named trials, never a recommendation to you.