Positional Sleep Apnea: When a Wearable May Rival CPAP

CPAP remains the most effective way to reduce breathing events in obstructive sleep apnea. But if your apnea happens mainly when you sleep on your back, a vibrating positional wearable may control daytime sleepiness about as well as CPAP—without a mask.

That conclusion applies to a narrow, testable group: adults with mild-to-moderate positional obstructive sleep apnea (POSA), confirmed on a full sleep study. In a 2025 crossover randomized trial, positional therapy and CPAP each reduced daytime sleepiness by a median of 4 points over eight weeks. CPAP still reduced the apnea-hypopnea index (AHI) more than positional therapy. Symptom relief and complete airway control are different targets, and the distinction matters.

The treatment should match the pattern

POSA means sleep position substantially changes the severity of apnea. In the trial that provides the clearest direct comparison, participants had a supine AHI at least twice their nonsupine AHI. Their total AHI had to fall between 5 and less than 30 events per hour.

Researchers estimate that POSA may affect up to 56% of people with OSA, although that figure should be treated as an upper estimate because the cited source does not identify its original prevalence study. Hernandez et al. also notes that positional therapy has limited utility for non-positional OSA. A person who simply prefers side sleeping is not necessarily a person whose apnea is positional.

CPAP works by holding the airway open regardless of position. It remains the usual standard because it reduces AHI more completely. Its practical weakness is tolerability: people commonly report mask discomfort, leaks, dryness, and claustrophobia. A 2025 randomized crossover study of high-flow nasal cannula versus CPAP described CPAP's long-term effectiveness as limited by poor long-term adherence. Wali et al. did not provide a general-population abandonment rate, and neither should anyone else without a source. The often repeated claim that half of patients abandon CPAP cannot be pinned down from the evidence here.

If position drives the obstruction, avoiding that position can treat the cause of that night's events. It does not make positional therapy interchangeable with CPAP for every form of sleep apnea.

What the head-to-head trial found

In Wilson et al.'s 2025 randomized crossover trial, 52 adults with polysomnography-confirmed mild-to-moderate POSA completed an eight-week vibrotactile positional-therapy period and an eight-week CPAP period in randomized order, separated by washout. A vibrotactile device prompts the sleeper to change position when it detects supine sleep. It is more sophisticated than sewing a tennis ball into a shirt, though the underlying goal is the same.

The primary outcome was the Epworth Sleepiness Scale (ESS), a questionnaire measuring daytime sleepiness. Participants started with a median ESS of 12 (interquartile range, 10 to 14), which indicates meaningful symptoms. Both treatments produced a median 4-point reduction in ESS. Improvement within each treatment period reached P<.001, while the between-treatment comparison was non-significant (P=.782). In practical terms, the trial found no detectable difference in sleepiness improvement between the two approaches over eight weeks.

The crossover design strengthens that result because each participant received both treatments and therefore served as their own comparator. Still, 52 people cannot answer every clinical question. The trial was too small for reliable subgroup findings and too short to establish whether either treatment maintains its effects, or its use, for years.

Positional therapy produced better sleep efficiency than CPAP in this trial, although neither treatment exceeded pretreatment sleep efficiency. The groups did not differ significantly in sleep architecture, arousal index, or device-use hours. The publication did not report mean nightly hours of use, so it cannot settle the question patients usually care about after the honeymoon period: which device will they still use next winter?

The larger evidence agrees on symptoms, not AHI

A 2025 meta-analysis by Gao et al. pooled 19 randomized controlled trials comparing sleep positional therapy with CPAP, oral appliances, or placebo. The authors analyzed changes from baseline rather than relying only on post-treatment group values. They assessed AHI, oxygen desaturation index, total sleep time, arousal index, sleep efficiency, ESS, and device-related complications.

Across the trial pool, positional therapy produced symptomatic outcomes comparable to CPAP, supporting the result seen in Wilson et al. But CPAP achieved greater objective AHI reduction. That is the central tradeoff, not a footnote. ESS captures how sleepy a person feels; AHI counts breathing events per hour of sleep. A treatment can improve one more than the other.

The meta-analysis also has a built-in source of uncertainty: "positional therapy" covered very different interventions, ranging from tennis-ball techniques to modern vibrating devices. Those tools differ in comfort, behavior change, and likely long-term use. Pooling them helps answer whether position avoidance can work, but it cannot tell you that every gadget marketed as a sleep-position solution will perform like the wearable used in the crossover trial.

Why diagnosis needs more than a consumer tracker

Before considering positional therapy, ask for a study report that breaks out supine and nonsupine AHI. Full polysomnography is especially useful because positional dependence can vary by sleep stage. In Ryu et al.'s 2024 observational study, REM and non-REM sleep produced divergent positional-dependency patterns.

That finding creates a real classification problem. A simplified home test, pulse oximeter, or consumer wearable may suggest that you sleep badly on your back, but it may not measure sleep stage and position with the precision used in the treatment studies. Misclassifying non-positional OSA as POSA could steer someone toward a therapy whose evidence does not apply to them.

CPAP side effects are also not imaginary or trivial. In qualitative interviews, CPAP-treated patients described dryness severe enough to wake them and interfere with breathing. Ahonen et al. documented these experiences, and a later qualitative study described xerostomia and oral-health concerns among long-term CPAP users. That research explains why a mask-free option can be appealing; it does not prove that positional therapy produces better long-term adherence.

Who should and should not consider it

Positional therapy is a reasonable conversation for someone who meets all of these conditions:

• A full sleep study confirms positional OSA, with supine AHI at least twice nonsupine AHI.
• Total AHI is mild to moderate, below 30 events per hour.
• Daytime sleepiness or CPAP intolerance is a treatment problem.
• A sleep clinician agrees that residual apnea burden fits the person's cardiovascular and medical risk profile.

It is not a substitute for CPAP in severe OSA, which Wilson et al. excluded, or in non-positional OSA. It also is not an obvious upgrade for a person who tolerates CPAP and wants the greatest AHI reduction. For people who cannot use either CPAP or positional therapy, oral appliance therapy offers another evidence-based route; Pépin et al. found a dose-response relationship between mandibular protrusion and reduced respiratory-effort burden in adults eligible for mandibular advancement therapy.

The cardiovascular question remains open. Positional therapy leaves more residual events than CPAP, and no long-term trial has shown whether that difference changes cardiovascular, metabolic, or cognitive outcomes. In a small 24-week randomized proof-of-concept study, O'Donnell et al. found that CPAP, but not GLP-1-mediated weight loss, improved early cardiovascular disease markers in people with OSA. That result does not directly compare CPAP with positional therapy, but it is a reason not to treat equal ESS scores as proof of equal long-term risk.

The verdict

For confirmed mild-to-moderate positional OSA, vibrotactile positional therapy has randomized evidence for symptom relief comparable to CPAP over eight weeks, backed by a 19-trial meta-analysis. It is evidence-supported, not experimental.

CPAP still produces greater AHI reduction, and no study here shows that positional therapy matches its long-term cardiovascular protection or adherence. Get the positional breakdown from a proper sleep study, then discuss the result with a sleep physician. The useful question is not whether CPAP is outdated. It is whether your airway problem is actually driven by position.