Sleep Tracking

Consumer sleep tracking is the use of a smartwatch, smart ring or phone app to estimate how long and how well a person slept. Adoption is now mass-market: roughly half of US adults (48% in the AASM's 2025 survey, up from 35% in 2023) and ~40% of UK adults track their sleep. The devices are sold as a way to hit the population "sweet spot" of about 6.4-7.8 hours identified by the 2026 Nature ageing-clock study (see Sleep Duration and Mortality).

What the devices measure

Device	Signals	Derived estimates
Phone app	Movement, sound (sometimes breathing rate from audio)	Time asleep, disturbances
Watch / band	Accelerometry + photoplethysmography (reflected light through skin)	Heart rate, breathing rate, sleep/wake, stage
Ring	As watch, plus skin temperature	As watch, plus temperature trends

None measures brain activity. Polysomnography (PSG) — electrodes for eye movement, EEG, muscle tone and heart rate in a sleep lab — is the gold standard everything is validated against.

How accurate they are

Evidence base: Robbins et al., Sensors 24(20):6532, October 2024 (doi 10.3390/s24206532), testing Oura Ring Gen3, Fitbit Sense 2 and Apple Watch Series 8 against PSG.

Sleep vs wake: sensitivity ≥95% for all three devices, but precision 72.7-87.8% — the devices reliably catch sleep and also label some wakefulness as sleep. The Economist's "agree with PSG around 95% of the time" rounds this up; the study wins.
Four-stage classification (wake / light / deep / REM): Cohen's kappa 0.65 Oura, 0.60 Apple Watch, 0.55 Fitbit — "moderate to substantial" agreement. The Economist renders this as ~50-80% agreement and quotes Robbins' grade of "C+ to B+". Devices nonetheless fold stage estimates into headline sleep scores, which are therefore less trustworthy than the duration number.

Reference point: two human PSG scorers typically agree only ~80% of the time, so stage agreement has a ceiling well below 100%.

Do they change behaviour?

AASM 2025 survey (n=2,007 US adults): 55% of tracker users said they changed their behaviour after seeing their data.
A small 2020 study in the Journal of Clinical Sleep Medicine found a week of tracker use improved self-reported sleep quality. No objective-outcome trial is cited.
Clinicians quoted by The Economist view trackers positively on balance: they give an objective time-asleep figure and often reassure people who are in fact sleeping enough.

The downside

A minority of users develop Orthosomnia — anxiety about the data itself — and worry about sleep is itself one of the most common causes of a sleepless night. The Economist's bottom line: the best indicator of a good night is waking up feeling rested, which needs no device.

Open questions

No trial yet links tracker use to an objectively measured improvement in sleep or health outcomes.
The 30% orthosomnia prevalence quoted by The Economist is unverified against the AASM's published figures.
Stage-level metrics (and the "sleep scores" built on them) are not yet accurate enough to act on individually.

Should You Use a Sleep Tracker (Economist) — 2026-06-06 source.
Is Too Much Sleep as Bad as Too Little (Economist) — the duration target trackers are sold against.
Sleep Duration and Mortality · Orthosomnia