Obstructive Sleep Apnea (OSA): How to Stop Treatment Safely

Why "Stopping Treatment" Is Not a Simple Decision

OSA is not a condition that disappears because symptoms feel better. The gold standard for diagnosis and for confirming remission is objective sleep testing, either in-lab polysomnography (PSG) or a validated home sleep apnea test (HSAT). Symptoms like daytime sleepiness improve on CPAP, which can create a false sense that the underlying airway obstruction has resolved. It has not, unless testing confirms otherwise.

What the AHI Threshold Means Clinically

The American Academy of Sleep Medicine (AASM) defines OSA as an AHI ≥5 events per hour in symptomatic patients, or ≥15 events per hour regardless of symptoms [1]. Mild OSA is an AHI of 5 to 14. Moderate OSA is 15 to 29. Severe OSA is ≥30. Before any physician considers treatment discontinuation, repeat testing must show the AHI has dropped below the patient's diagnostic category threshold, ideally below 5.

The Cardiovascular Stakes of Untreated OSA

Stopping treatment prematurely restores the pre-treatment physiological state. The Sleep Heart Health Study (N=6,132) found that severe OSA was associated with a 2.38-fold increase in coronary heart disease risk in adults under 70 [2]. A separate analysis published in the European Heart Journal found that untreated OSA increased the risk of incident atrial fibrillation by 49% compared with no OSA [3]. These are not theoretical risks. They accrue night by night once effective treatment ends.

Conditions Under Which OSA Can Genuinely Resolve

OSA does resolve in some patients. The causal factors that resolve it include substantial weight loss, upper airway surgery with documented anatomical correction, resolution of hypothyroidism or acromegaly when those conditions drove the OSA, and sometimes positional therapy when OSA is strictly positional.

Weight Loss and the AHI Response

Weight is the most modifiable OSA driver in people with obesity. A meta-analysis of 342 patients across randomized controlled trials found that every 1% reduction in body weight was associated with a 3% reduction in AHI [4]. A 10% weight loss reduced AHI by approximately 26% on average. That is meaningful improvement, but a 26% AHI reduction does not equate to remission in a patient who started with an AHI of 40.

The Wisconsin Sleep Cohort study (N=690) demonstrated that a 10 kg/m² reduction in BMI was associated with a 26% decrease in AHI, and that a 10% weight loss predicted a 26% AHI improvement (95% CI: 21 to 31%) [4]. Critically, most patients with moderate-to-severe OSA require far larger weight reductions before AHI normalizes.

GLP-1/GIP Agonists and OSA: The SURMOUNT-OSA Data

Tirzepatide (Zepbound, Eli Lilly) received FDA approval in January 2024 specifically for moderate-to-severe OSA in adults with obesity, making it the first pharmacological agent approved for this indication [5]. The approval was based on the SURMOUNT-OSA program, two phase 3 randomized controlled trials.

In the SURMOUNT-OSA trial 1 (patients not using PAP therapy, N=234) and trial 2 (patients using PAP therapy, N=235), tirzepatide 10 or 15 mg weekly produced a mean AHI reduction of 27.4 events per hour in trial 1 (a 62.8% reduction) versus 4.8 events per hour with placebo (6.4% reduction) at 52 weeks [6]. In trial 1, 42.5% of tirzepatide-treated patients achieved an AHI <5. That means remission by AASM criteria in nearly half of patients who were not on PAP at baseline.

These results are striking. They do not mean that patients on tirzepatide can stop treatment without re-testing. The 42.5% who achieved AHI <5 needed documented testing to confirm that outcome.

Surgical Correction

Upper airway surgery, including maxillomandibular advancement (MMA) and, in children, adenotonsillectomy, can anatomically resolve OSA. A meta-analysis of MMA in adults (N=627 patients across 21 studies) reported a mean post-operative AHI of 9.1 events per hour, down from a pre-operative mean of 62.5, with a surgical success rate (AHI <20 and ≥50% AHI reduction) of 86% [7]. Success in that study still did not mean AHI <5 for all patients, reinforcing the need for post-operative PSG before CPAP discontinuation.

The Step-by-Step Protocol for Safe Treatment Discontinuation

Stopping OSA treatment safely follows a defined sequence. No step should be skipped.

Step 1: Establish That a Remission Trigger Exists

A clinician should identify why remission might have occurred before ordering re-testing. Valid triggers include:

• Documented weight loss of ≥10% body weight (ideally ≥15% in moderate-to-severe OSA)
• Post-surgical anatomy confirmed to have changed (MMA, UPPP with documented airway changes)
• Resolution of a secondary cause (hypothyroidism now euthyroid, acromegaly treated)
• Strictly positional OSA confirmed on original diagnostic study, with sustained weight loss or positional intervention

Without a plausible physiological reason for improvement, re-testing is unlikely to show remission, and stopping treatment places the patient at unnecessary cardiovascular and neurocognitive risk.

Step 2: Order Repeat Objective Sleep Testing

The AASM recommends repeat PSG or HSAT after significant weight loss or other interventions to confirm AHI reduction before discontinuing PAP therapy [1]. The test should be conducted off CPAP (or off the oral appliance) to reflect the patient's untreated status. A single normal HSAT may be sufficient for low-complexity cases, but patients with comorbid cardiovascular disease or a history of severe OSA should have in-lab PSG to capture full data, including oxygen desaturation index and arousal index.

The HealthRX clinical team uses the following discontinuation decision framework. If all four criteria are met, a supervised CPAP trial discontinuation with a 4-week follow-up HSAT is appropriate:

1. AHI <5 on repeat off-therapy sleep study
2. Epworth Sleepiness Scale score <10 at follow-up visit
3. BMI reduction of ≥10% or surgical/anatomical correction documented
4. No history of severe OSA-related arrhythmia or hospitalization

If any criterion is unmet, treatment continues and re-testing is deferred to 6 to 12 months.

Step 3: Supervised Trial Discontinuation

If testing confirms AHI <5, treatment does not end on the day of the result. A supervised trial means the patient stops CPAP for 4 to 8 weeks and returns for a follow-up HSAT or clinical evaluation. The American Thoracic Society notes that OSA can recur, particularly with even modest weight regain [8]. The patient should receive written instructions on symptoms of OSA recurrence (witnessed apneas, re-emerging snoring, morning headaches, daytime hypersomnolence) and a clear instruction to resume therapy and contact their provider if those symptoms appear.

Step 4: Long-Term Monitoring

OSA remission is not permanent. The Wisconsin Sleep Cohort showed that over a 4-year follow-up, incident OSA developed in 7.5% of men and 5.0% of women who had been free of OSA at baseline, driven largely by weight gain and aging [9]. Patients who discontinue CPAP after weight-loss-induced remission should have annual clinical review. Any weight regain of ≥5% should trigger repeat sleep testing before symptoms become severe again.

CPAP-Specific Discontinuation Considerations

What "Non-Adherence" Actually Means

CPAP non-adherence is not the same as treatment discontinuation. Medicare and most commercial insurers define non-adherence as using CPAP fewer than 4 hours per night on fewer than 70% of nights over a 30-day period. Non-adherent patients who stop CPAP because it is uncomfortable still have untreated OSA. They need an alternative effective treatment, not no treatment.

Alternatives with evidence for moderate-to-severe OSA include:

• Mandibular advancement devices (MADs): A Cochrane review (23 trials, N=925) found MADs produced greater sleepiness improvement than no treatment and were subjectively preferred by many patients over CPAP, though CPAP produced greater AHI reduction [10].
• Hypoglossal nerve stimulation (Inspire therapy): FDA-approved for moderate-to-severe OSA in patients intolerant of PAP therapy with AHI 15 to 65; the STAR trial (N=126) reported a 68% responder rate at 12 months [11].
• Tirzepatide in eligible patients (obesity with BMI ≥30): as outlined in the SURMOUNT-OSA data above [6].

Risks of Abrupt Unsupervised CPAP Cessation

Patients who abruptly stop CPAP without medical guidance risk rapid return of OSA within 1 to 2 nights. A crossover study published in JAMA found that 2 nights off CPAP in previously adherent patients restored pre-treatment levels of sympathetic activation and blood pressure [12]. The effect was rapid and measurable. Patients with hypertension or established cardiovascular disease face the highest short-term risk.

Tirzepatide as a Treatment Discontinuation Pathway

Because tirzepatide is now FDA-approved for OSA in adults with obesity, it represents a genuine pharmacological route toward reducing or eliminating PAP dependence for some patients. The SURMOUNT-OSA data show that 42.5% of patients on tirzepatide who were not using PAP therapy achieved AHI <5 at 52 weeks [6]. That outcome makes tirzepatide the only drug with regulatory approval and phase 3 evidence for OSA resolution.

Who Qualifies for Tirzepatide in the OSA Indication

The FDA label approved tirzepatide (Zepbound) for moderate-to-severe OSA (AHI ≥15) in adults with obesity (BMI ≥30 kg/m²). Patients must have a confirmed OSA diagnosis by PSG or HSAT. Tirzepatide is used as an adjunct to, not a replacement for, existing OSA treatment during the treatment period. Discontinuation of PAP therapy requires re-testing as described above, even in patients achieving substantial weight loss on tirzepatide.

Dose and Duration Considerations

The approved tirzepatide dose for OSA follows the same escalation schedule as the obesity indication: starting at 2.5 mg weekly and escalating by 2.5 mg every 4 weeks to a maintenance dose of 10 or 15 mg weekly [5]. The SURMOUNT-OSA trials ran for 52 weeks, and the AHI reduction was assessed at that time point. Whether AHI remission is durable beyond 52 weeks without continued tirzepatide use is not established; weight regain after discontinuation of GLP-1/GIP agonists is well-documented, with patients regaining approximately 60% of lost weight within 1 year of stopping semaglutide 2.4 mg in the STEP-1 extension trial [13].

Special Populations: When Stopping Is Higher Risk

Patients With Cardiovascular Comorbidities

Patients with established coronary artery disease, heart failure, or a history of stroke face compounded risk from OSA recurrence. The ISAACC trial (N=2,717) examined whether CPAP reduced cardiovascular events in patients with moderate-to-severe OSA and established coronary artery disease; CPAP did not significantly reduce the primary composite endpoint, though adherence in that trial averaged only 2.8 hours per night [14]. Low adherence confounds the cardiovascular outcome data, but the physiological harm of OSA in this population remains biologically plausible and supported by epidemiological evidence.

These patients should not discontinue CPAP without repeat PSG showing AHI <5, and should ideally have cardiology involvement in the discontinuation decision.

Patients With Residual Daytime Sleepiness After Documented AHI Normalization

Some patients with confirmed AHI <5 on repeat testing still report excessive daytime sleepiness. The AASM recognizes this as a distinct entity: residual excessive sleepiness (RES) in treated OSA. In these cases, discontinuing CPAP is not the correct next step. Evaluation for narcolepsy, idiopathic hypersomnia, circadian rhythm disorders, or inadequate sleep duration is warranted before any treatment change.

Pediatric OSA

OSA in children is a distinct condition. The American Academy of Pediatrics recommends adenotonsillectomy as first-line treatment for pediatric OSA when adenotonsillar hypertrophy is present [15]. Post-surgical PSG is recommended 6 to 8 weeks after adenotonsillectomy to confirm resolution, particularly in children with obesity, Down syndrome, or craniofacial abnormalities, where resolution rates are lower. Pediatric OSA management is outside the scope of CPAP discontinuation in adults but is noted here because adolescent patients transitioning to adult care often have incomplete re-testing records.

What Guidelines Say About Stopping OSA Treatment

The AASM 2019 clinical practice guideline on PAP therapy states that "clinicians should reassess patients treated with CPAP who have achieved significant weight loss, as OSA may resolve and discontinuation of PAP therapy may be appropriate following objective testing" [1]. The guideline does not specify a weight loss threshold but cites the dose-response data showing AHI improvement with increasing weight loss magnitude.

The American Thoracic Society 2016 update on OSA management states that "periodic re-evaluation with objective sleep testing is recommended after significant anatomical or weight changes" [8]. "Significant" is not defined numerically in that document, but the clinical consensus is ≥10% body weight as a minimum trigger for re-testing.

The Endocrine Society 2021 guideline on obesity pharmacotherapy notes that GLP-1 receptor agonists produce mean weight losses of 12 to 15% at 1 year and recommends re-evaluation of obesity-related comorbidities, including OSA, at 6-month intervals during active treatment [16].

Practical Patient Checklist Before Stopping OSA Treatment

Patients asking about stopping treatment should bring the following to their appointment:

• CPAP download data showing current AHI-on-therapy and usage hours (ideally 90-day average)
• Recent body weight and BMI calculation compared to weight at OSA diagnosis
• Any post-surgical operative notes if upper airway surgery was performed
• Current Epworth Sleepiness Scale score (self-administered before the visit)
• List of current cardiovascular medications and any arrhythmia history

A complete picture at a single visit makes the re-testing decision faster and reduces back-and-forth. Patients with CPAP AHI-on-therapy consistently <1 event per hour and ≥10% weight loss from diagnosis weight are the strongest candidates for proceeding to off-therapy re-testing.