Learn · Sleep

Sleep stages: deep, REM and core sleep explained

In short: You sleep in cycles of about 90 minutes in which core sleep, deep sleep and REM take turns. Watches estimate the stages only roughly: they detect that you are asleep well, but often not which stage it is. Rely more on the length and regularity of your nights than on single percentages.

Four colours for one night

Tap a stage to highlight it across the night.

Example night from 11 pm to 7 am, not a measurement. Tap a stage to highlight it across the night. The numbers on top show cycles of about 90 minutes each.

In a sleep lab, five stages are told apart: awake, REM and three levels of non-REM sleep, N1 to N3. N1 is light, transitional sleep, N2 makes up most of non-REM sleep, and N3 is deep sleep1. The stages follow each other in cycles: a night has about four to six of them, and on average one lasts about 90 minutes2.

Apple Watch and Salvor combine N1 and N2 as “core”. Other watches call this “light sleep”. Core sleep is still important and is often more than half of the night1. Deep corresponds to N3, and REM and awake are named as in the lab.

What is normal, and what does age change?

There is no ideal value for the shares. In an analysis of 65 lab studies in healthy people, the stages changed with age: in adults the share of deep sleep and REM fell, while light sleep (N1 and N2) and the time awake after falling asleep rose3.

The study also shows how uncertain such norms are: how much the stages differed with age depended on how strictly participants were screened for illnesses and drug or alcohol use3. So compare your stages mainly with your own nights, not with a table.

What affects the stages

  • lowers: alcohol in the evening: REM later and less of it4
  • lowers: caffeine up to six hours before bed5
  • lowers: older age: less deep sleep and REM3

A review of all studies in healthy volunteers found that alcohol shortens the time to fall asleep and delays the first REM period at every dose, and at moderate and high doses people get less REM sleep overall. In the second half of the night, sleep is more disrupted4.

Caffeine: in one study, a dose of 400 mg still disturbed sleep when taken six hours before the usual bedtime. What was measured was mainly how much sleep was lost, not a single stage5.

How well does your watch recognise the stages?

The gold standard is polysomnography in a sleep lab, which measures brain activity among other things. A watch on the wrist estimates the stages from sensor signals instead. Apple Watch uses the movements of your wrist1, 7.

In a comparison with a sleep lab, fitness trackers detected sleep very reliably (sensitivity at least 93%) but wake much worse (specificity only 18 to 54%). The stages agreed with the lab only partly, and it was worse on disrupted nights6. For Apple Watch, Apple reports an average agreement with the lab of 0.63 for Apple Watch (kappa, where 1 would be perfect). The most common error is scoring deep sleep as core sleep1.

What Salvor does with it

Salvor reads the stages from Apple Health (awake, core, deep, REM) and shows them as a timeline of your night. Because the estimate is uncertain, it counts little in the sleep score: duration weighs 45%, quality (efficiency, time awake, awakenings) 25%, regularity 20% and the share of deep sleep plus REM only 10%.

Regularity weighs this much because of a study of more than 60,000 people: those who slept more regularly had a lower risk of death, and regularity predicted it more strongly than sleep duration8. The share of deep sleep and REM is also compared with your own average, not only with a fixed range.

When to see a doctor

The stages on your watch are not a diagnosis. Talk to a doctor if you are very tired during the day for weeks despite enough sleep, or if someone notices that you snore loudly and pause in your breathing. This applies whatever a watch shows. In an emergency, call 112.

Sources

  1. Apple. Estimating Sleep Stages from Apple Watch. September 2023. www.apple.com/ca/healthcare/docs/site/Estimating_Sleep_Stages_from_Apple_Watch_Sept_2023.pdf
  2. Patel AK, Reddy V, Shumway KR, Araujo JF. Physiology, Sleep Stages. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2024. www.ncbi.nlm.nih.gov/books/NBK526132/
  3. Ohayon MM, Carskadon MA, Guilleminault C, Vitiello MV. Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. Sleep. 2004;27(7):1255–1273. doi.org/10.1093/sleep/27.7.1255
  4. Ebrahim IO, Shapiro CM, Williams AJ, Fenwick PB. Alcohol and sleep I: effects on normal sleep. Alcohol Clin Exp Res. 2013;37(4):539–549. doi.org/10.1111/acer.12006
  5. Drake C, Roehrs T, Shambroom J, Roth T. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. J Clin Sleep Med. 2013;9(11):1195–1200. doi.org/10.5664/jcsm.3170
  6. Chinoy ED, Cuellar JA, Huwa KE, et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep. 2021;44(5):zsaa291. doi.org/10.1093/sleep/zsaa291
  7. de Zambotti M, Cellini N, Goldstone A, Colrain IM, Baker FC. Wearable sleep technology in clinical and research settings. Med Sci Sports Exerc. 2019;51(7):1538–1557. doi.org/10.1249/mss.0000000000001947
  8. Windred DP, Burns AC, Lane JM, et al. Sleep regularity is a stronger predictor of mortality risk than sleep duration: a prospective cohort study. Sleep. 2024;47(1):zsad253. doi.org/10.1093/sleep/zsad253

Last updated: 16 October 2026. General information on wellbeing and fitness. Not a substitute for medical advice, not a diagnosis.