Should You Use a Sleep Tracker (Economist)

Summary

A Well Informed / Science & Technology column asking two questions of the consumer sleep-tracking boom (nearly half of Americans, ~40% of Britons now track sleep with a watch, ring or app): are the devices accurate, and do users sleep better? Verdict: trackers are reliable at telling sleep from wakefulness, mediocre at sleep staging, modestly useful for changing habits, and harmful for the minority whose data makes them anxious. The piece ends on the low-tech alternative: whether you wake feeling rested is the best single indicator of how good the night was.

Motivating study: the Nature "sleep chart" cohort (Wen et al., Columbia, 13 May 2026) — covered in full on Is Too Much Sleep as Bad as Too Little (Economist) — which found a U-shaped relationship between self-reported sleep duration and 23 biological-ageing clocks in ~500,000 UK Biobank adults, with the slowest ageing at roughly 6.4-7.8 hours a night. Trackers are sold as the way to hit that window.

Key claims

Adoption

• ~50% of US adults and ~40% of UK adults use a smartwatch, ring, phone app or similar to track sleep. (The AASM's 2025 US survey of 2,007 adults puts US tracker use at 48%, up from 35% in 2023.)

How the devices measure

• Phone apps: movement and sound, sometimes inferring breathing rate from audio.
• Watches: accelerometry plus photoplethysmography (low-intensity light reflected off skin to estimate heart and breathing rate).
• Rings: the above plus skin temperature.

Accuracy vs polysomnography (PSG, the electrode-based gold standard)

• Sleep vs wake: a 2024 paper in Sensors tested three popular wearables (Oura Ring Gen3, Fitbit Sense 2, Apple Watch Series 8) against PSG and, per the article, found all three "agreed with polysomnography around 95% of the time". See the flag below on what that number actually is.
• Sleep stages (REM, light, deep): wearables agree with PSG only ~50-80% of the time — "a C+ to B+" per Rebecca Robbins (Harvard Medical School), a co-author of the validation study. Devices nonetheless fold stage estimates into their headline "sleep scores".

Do they change behaviour?

• AASM survey: 55% of adults using a tracker said they changed their behaviour after seeing the data.
• A small 2020 study in the Journal of Clinical Sleep Medicine found a week of wearing a tracker improved self-reported sleep quality (no objective endpoint reported).
• Sleep researchers and therapists mostly view trackers positively: they give an objective sense of time asleep and can reassure people who are in fact sleeping enough.

Downside: Orthosomnia

• Up to 30% of sleep trackers report feeling anxious about their data (the article's figure; the AASM's own 2025 release does not state 30% — see flag).
• Worry about poor sleep is the second most common cause of a sleepless night, behind financial worry.
• Unlike diet or exercise, putting more effort into sleep can backfire.

What is flagged as unproven or weak

• The "95% agreement" claim overstates the Sensors paper. The underlying study (Robbins et al., Sensors 24(20):6532, Oct 2024, doi 10.3390/s24206532) reports sleep-detection sensitivity ≥95% for all three devices but precision of only 72.7-87.8%, and four-stage Cohen's kappa of 0.65 (Oura), 0.60 (Apple Watch) and 0.55 (Fitbit) — "moderate to substantial" agreement. High sensitivity with lower precision means the devices mostly catch sleep but also label some wake as sleep. The study wins: trackers are good, not 95%-good, at sleep/wake.
• The 2020 JCSM behaviour-change study is small and its endpoint is self-report.
• The 30% orthosomnia prevalence is the article's number; it is not in the AASM 2025 release (which reports 76% of adults having lost sleep over worry about sleep, and 55% of tracker users changing behaviour). Treat 30% as unverified.
• The ageing-clock study is observational and uses self-reported sleep duration; it cannot show that moving your sleep into the 6.4-7.8 h window slows ageing.

Closing line

"Sleep data is not worth losing sleep over, because there is an alternative to sleep tracking: whether you wake up feeling rested is the best indicator of how good your night was."

Wikilinks

The Economist — Science & Technology / Well Informed. Sleep Tracking — concept page: what consumer devices measure, how well, and the evidence on behaviour change. Orthosomnia — concept page: data-driven sleep anxiety. Sleep Duration and Mortality — the U-shaped curve the trackers are sold against. Is Too Much Sleep as Bad as Too Little (Economist) — companion piece (2026-07-04) on the same Nature study.

Captured-by trigger

Source file: Economist 2026-06-06 Science---Techn - Should you use a sleep tracker.md. Validation study checked against the Sensors abstract (doi 10.3390/s24206532) and the AASM 2025 survey release (aasm.org).