What sleep trackers can and cannot tell you

Wearable sleep tracking has become ubiquitous, and it occupies an awkward position: useful enough to be worth having and inaccurate enough to be worth doubting.

How they work

Consumer devices infer sleep from movement, and increasingly from heart rate, heart rate variability, temperature and blood oxygen.

Algorithms combine these to estimate sleep and wake, and then to estimate stages.

The reference standard is polysomnography, which measures brain activity directly along with eye movement and muscle tone — which is what actually defines sleep stages.

No consumer wearable measures brain activity, which is the fundamental limitation.

What they do reasonably well

Detecting whether you are asleep or awake overall, where validation studies generally find good sensitivity for sleep.

Estimating total sleep duration, which is usually within a reasonable margin.

Estimating bedtime and wake time, which is the most useful output and the one people pay least attention to.

Tracking consistency across weeks, which is where the real value lies, since regularity of sleep timing is a strong predictor of outcomes.

And detecting relative changes — a week that was worse than usual — which does not require absolute accuracy.

What they do poorly

Detecting wakefulness while lying still, which is exactly what someone with insomnia does, meaning trackers systematically overestimate their sleep.

Sleep staging, where validation studies find agreement with polysomnography that is well short of reliable, particularly for distinguishing light and deep sleep.

The precise numbers presented for deep and REM sleep should be treated as very rough estimates rather than measurements.

Detecting sleep disorders, which they are not designed or validated to diagnose — though some devices now flag possible breathing disturbance, which can prompt useful investigation.

And single-night interpretation generally, since night-to-night variation is large and normal.

Orthosomnia

A phenomenon described in the sleep literature and increasingly seen in clinics.

It refers to people who develop sleep problems, or worsen existing ones, through preoccupation with achieving perfect tracker scores.

The mechanism is straightforward: sleep is degraded by effort and monitoring, and a device that quantifies performance nightly is an excellent generator of both.

Case reports describe people who slept adequately, were told by a device that their deep sleep was insufficient, and developed genuine insomnia in pursuit of a number produced by an unvalidated algorithm.

Which is the strongest argument for looking at weekly trends rather than nightly scores, and for taking off the device if it has become a source of anxiety.

Using them well

A reasonable approach.

Use them for timing consistency, which is the output most likely to be both accurate and actionable.

Use them to test specific things — alcohol, late meals, exercise timing, room temperature — by comparing periods rather than nights.

Ignore the stage breakdown, or treat it as entertainment.

Ignore the composite score, which is a proprietary construct rather than a clinical measure.

Do not check it first thing in the morning, since the number then shapes how you feel about the day, which is a documented effect — experiments giving participants false sleep quality feedback have found it affects daytime performance and mood independently of actual sleep.

And stop using it if you have insomnia, at least during treatment, since monitoring is part of the problem.

Sleep diaries

Worth mentioning because they outperform devices for the purpose that matters.

Cognitive behavioural therapy for insomnia uses subjective sleep diaries rather than objective measurement, and does so deliberately.

The treatment works on the person's experience of sleep, and the calculations that drive sleep restriction are based on diary data.

A diary also captures what a device cannot: what was going on, what was taken, how it felt.

When measurement actually matters

Where a sleep disorder is suspected, proper testing exists and is accessible.

Home sleep apnoea tests are now the standard route in many systems and measure the relevant variables directly.

Actigraphy — clinical-grade movement monitoring over weeks — is used for circadian rhythm disorders and is validated for that purpose.

Polysomnography remains the reference for complex cases.

A consumer device is a reason to seek assessment, not a substitute for it.

General information only, not medical advice. Consumer sleep trackers do not diagnose sleep disorders — consult a qualified clinician about persistent sleep problems.