Can Smartwatches Really Detect Sleep Apnoea?

Why it sounds plausible

A pause or reduction in breathing can change movement, oxygen levels and pulse patterns. Sensors on a wrist can detect some of those signals, and an algorithm can look for patterns associated with obstructive sleep apnoea. That gives the idea a credible basis. The important test is comparison with an appropriate clinical reference, not whether a watch produces an impressive sleep score. [1][2]

What Apple's validation actually found

Apple's published clinical validation reported 66.3% sensitivity and 98.5% specificity for its notification feature against the study's moderate-to-severe threshold. Those figures describe a particular algorithm, protocol and reference test. They are manufacturer-reported evidence, not an independent verdict on every smartwatch. [1]

Sensitivity describes how often people meeting the disease threshold are flagged. Roughly two thirds were identified in that validation, leaving a substantial minority without a notification. Specificity describes how often those below the threshold avoid an alert. It does not mean an alerted person has a 98.5% chance of having the condition; that probability also depends on how common it is in the population being screened. [1]

Other validated watches are a separate question

The FDA's review of Samsung's sleep-apnoea feature reported sensitivity of 82.7% in its study. Its hardware, oxygen-based approach, eligibility and study protocol differ from Apple's movement-based feature. Comparing headline percentages cannot establish which is best for an individual or prove that every similarly named app works. [2]

Regulatory authorisation is specific to the stated feature and intended use. It is not permission to reinterpret an ordinary sleep-stage estimate as a diagnostic test. Availability and supported hardware also vary by country and software version. Check the actual instructions for the feature you own. [1][2]

The catch: screening is not diagnosis

These systems focus on possible moderate-to-severe obstructive sleep apnoea. Mild disease, other sleep disorders and some affected users may be missed. Sensor contact, usable recording nights and eligibility restrictions matter. A reassuring score cannot explain persistent sleepiness or witnessed breathing pauses. [1][2]

NICE recommends a clinical pathway beginning with home respiratory polygraphy for suspected obstructive sleep apnoea, with further testing when indicated. A sleep laboratory is not necessary for every initial assessment, but an appropriate clinical test is still different from an unsolicited wrist alert. [3]

What this means for you

Treat a notification as a reason to discuss assessment with a clinician. If someone notices breathing pauses or you have persistent excessive daytime sleepiness, seek advice even if the watch has never alerted you. Do not change prescribed sleep-apnoea treatment because a watch score improves. [3]

Before relying on a feature, ask what condition it screens for, who it was validated in and what its instructions say a negative result means. A useful early warning can help someone access care without replacing that care.

Research & Sources

Check the evidence yourself. Access limitations and indirect evidence are identified below.

Evidence reviewed: October 2026

Targeted editorial review of sources available on 24 October 2026, not an exhaustive systematic review. Evidence vs Hype provides education, not individual medical diagnosis or treatment. Original AI-generated artwork is illustrative and does not show a study result. No commercial product was selected or endorsed.