steady

Sleep · The rhythm

The tracker isn't the doctor

Evidence: worth trying

Do this

For one week, don't open the sleep score in the morning. Judge the night by how you feel mid-morning.

How to do it

  1. Don't open the score in the morning for a week. What you're told about your night changed how people felt.
  2. Judge the night by mid-morning. By then the grogginess of waking has usually cleared.
  3. Keep wearing it if you like. Treat the totals as rough; stages and time awake are the shakiest part.
  4. Worried by the numbers? Take it to a doctor, not the app.

How the score gets in

  1. Trackermisses time awake

    Against a sleep lab, seven consumer devices caught at least 93% of sleep but only 18 to 54% of time awake. Sleep stages were hit and miss.

  2. Morning scoresets how you feel

    People with insomnia who were told, falsely, that they'd slept badly felt sleepier and less sharp that day than those told they'd slept well.

  3. Worrycan feed bad nights

    Three patients sought treatment over what their trackers showed, chasing the ideal night. That's three case stories, not a study.

  4. Youjudge by the day

    Whether a week without the score helps hasn't been tested. It's my suggestion, built on the two findings above.

Why it works

Sleep doctors have a name for people whose worry about sleep grew while they tried to perfect a tracker score: orthosomnia, roughly "correct sleep". The name comes from a report on three patients. In each, worry about the numbers had become part of the problem. Evidence: worth trying

Three people is a story, not a study. But what you're told about your night can change how the day feels, whatever the night was really like. Evidence: decent And the numbers are weakest at time spent awake. Evidence: decent If the morning score decides your mood for the day, try a week without it.

The long readEvery step of the chain, each study with its numbers, and what's still unproven. 3 min read.

What happens in your brain

The tracker misses time awake. Seven consumer sleep trackers, four worn and three by the bed, were tested against a full sleep-lab recording in 34 healthy young adults over three nights. They caught 93% or more of the time asleep, but only 18 to 54% of the time awake, so time awake often gets counted as sleep. Their sleep-stage results were inconsistent, and they did worse on nights when sleep was broken. Evidence: decent

The morning score sets how you feel. 63 people with insomnia wore a watch that gave them a sleep score on waking. The score was fake: half were told they'd slept well, half badly. By the evening, the "bad night" group felt sleepier, more tired and less mentally sharp. The "good night" group's mood and alertness rose more through the morning. Their reaction-time test didn't differ. Evidence: decent

Worry can feed bad nights. The three patients in the orthosomnia report went to a sleep clinic because of what their trackers showed, and found the tracker's numbers fitted their experience better than validated lab tests. The authors describe a perfectionist chase for the ideal night. Evidence: worth trying That this worry makes sleep itself worse is their reading of three cases, not something measured. Evidence: worth trying

You judge by the day. Nobody has tested a week without the score. It's my suggestion, built on the two findings above. Evidence: worth trying

The proof

  • 63 people with insomnia were given a fake sleep score on waking, good or bad, at random. Those told they'd slept badly reported less alertness and more sleepiness and fatigue by evening. Reaction times and attention tests didn't differ. One day, one sample. Evidence: decent
  • 34 healthy young adults slept three nights in a lab wearing or using seven consumer sleep trackers. The devices caught at least 93% of sleep but only 18 to 54% of time awake; sleep stages were inconsistent; they did worse on disrupted nights. Evidence: decent
  • Three patients came to a sleep clinic worried by their tracker data. The doctors named it orthosomnia and noted that tracker data often felt more true to patients than validated tests. Evidence: worth trying
  • Nobody has tested skipping the morning score. Evidence: worth trying

How to do it

  • For one week, don't open the score in the morning. Evidence: worth trying
  • Judge the night by how you feel mid-morning. In a brain-scan study, full alertness took 20 to 30 minutes to come back after waking. Evidence: decent
  • Keep wearing it if you like. Treat the totals as rough, and don't lean on stages or time awake. Evidence: decent
  • If the data worries you, take the worry to a doctor, not to the app.

Watch out

Nobody has tested skipping the morning score. Evidence: worth trying The orthosomnia report is three patients: a story, not a study. Evidence: worth trying The fake-score experiment was one day, in people with insomnia, and it changed how they felt, not their reaction times. Evidence: decent The device test was in healthy young adults; how trackers do in older or poorer sleepers wasn't tested there. Evidence: decent

If the data worries you, take the worry to a doctor, not to the app.

Arrows to move, Enter to open, Esc to close.