Sleep trackers count your hours well, but their sleep stages are rough guesses

Key takeaways

• In a Navy lab study of seven consumer sleep trackers, every device caught at least 93% of the time people slept, but correctly flagged only 18% to 54% of the time they were awake.
• Labels such as “deep” and “REM” sleep match lab results only about 50% to 70% of the time, according to a 2024 review for the Sleep Research Society.
• The FDA has authorized sleep apnea alerts on some Samsung and Apple watches, but Apple’s alert caught only about two out of three people with moderate to severe apnea in testing.
• In a 2026 national survey, about 1 in 3 U.S. adults tracked their sleep, and nearly a third of those trackers screened positive for worry about their sleep data.
• If you have trouble sleeping at least three nights a week for three months, a clinician and a proven talk-based therapy will help more than a better score.

In May 2026, researchers from Harvard Medical School and Brigham and Women’s Hospital reported on a national survey of 1,280 American adults¹. About 1 in 3 said they tracked their sleep with a digital tool. Most of them, 70%, said they did it to sleep better.

The survey also asked trackers how often they felt nervous about their sleep data or couldn’t stop worrying about it. Nearly 31% scored high enough to be flagged at risk of orthosomnia¹.

Orthosomnia is an unhealthy preoccupation with getting perfect sleep. Sleep researchers coined the word in 2017². The survey used a quick screening tool, not a diagnosis, and a gift from the ring maker Oura paid for it¹. Still, it points to an odd problem: a device meant to improve sleep can make some people sleep worse.

That doesn’t make trackers useless. The skill is knowing which numbers to trust, which to take lightly, and when a number means you should see a clinician.

Trackers are good at spotting sleep and poor at spotting wakefulness

The gold standard for measuring sleep is polysomnography, an overnight lab test that records brain waves, breathing, heart rhythm and more. Researchers test trackers by having people wear them during that test.

In one of the most careful tests, researchers at the Naval Health Research Center studied seven devices⁴. Four were wearables from Fatigue Science, Fitbit and Garmin, and three were bedside devices. Thirty-four healthy young adults spent three nights in the lab, including one night when researchers disrupted their sleep on purpose.

Every device caught at least 93% of the time people were truly asleep⁴. But they correctly labeled only 18% to 54% of the time people were awake. In plain terms, when you lie quietly awake, many trackers assume you’re asleep.

The devices also did worse on disrupted nights, the ones you’re most likely to check⁴. Most matched or outperformed actigraphy, the wrist motion sensors long used in sleep research. Both Garmin models fell short of it.

A 2025 analysis in the Journal of Clinical Sleep Medicine pooled 24 studies of wrist trackers with 798 people⁵. The trackers’ numbers differed significantly from lab results for total sleep, time to fall asleep and time awake. The authors called them less reliable than lab testing but useful for following general patterns.

Deep sleep and REM numbers are educated guesses

Many apps split your night into light sleep, deep sleep and REM, the stage when most dreaming happens. The lab sorts stages using brain waves, which wrist and finger devices can’t record, so they infer stages from movement and heart signals.

In 2024, a team commissioned by the Sleep Research Society reviewed wearable sleep tracking⁶. Staging formulas have improved, they found, but identify stages correctly only about 50% to 70% of the time. Total sleep time was the most consistent measure across devices, and time awake at night was a main weakness.

The Navy study found the same pattern⁴. Five of the six devices it checked for stages overestimated light sleep. Three overestimated deep sleep, and three underestimated REM.

The 2024 review flagged other limits⁶:

• Most validation studies test young, healthy adults in controlled labs. Results may not hold for older adults, people with sleep disorders or chronic illness, or ordinary nights at home.
• Skin color and body weight can affect sensor readings, which could widen existing health gaps.
• Companies rarely share raw data or explain their formulas, so outside scientists can’t fully check them.
• Users can’t control software updates, and newer versions have scored sleep differently with no apparent change in hardware.

So when your app reports 47 minutes of deep sleep, read it as a rough estimate. A small nightly swing may reflect the formula, not your body.

The one FDA-reviewed sleep feature: apnea alerts

Most sleep tracking never goes before the Food and Drug Administration. In its January 2026 guidance, the FDA lists claims “to track sleep trends” as an example of general wellness, a low-risk category the agency says it does not intend to examine⁷. Sleep scores and stage charts fall in that zone.

Sleep apnea alerts are different. In obstructive sleep apnea, the airway repeatedly narrows or closes during sleep. The FDA has authorized at least two watch features that flag signs of moderate to severe apnea.

Samsung’s feature received FDA De Novo authorization, a pathway for new types of lower-risk devices, on February 6, 2024⁸. It’s for adults 22 and older who haven’t been diagnosed with apnea. The watch’s light sensors estimate blood oxygen and look for repeated drops during sleep. It needs two qualifying nights within 10 days.

In Samsung’s study, the feature caught 82.7% of people whose overnight sleep study showed moderate to severe apnea⁸. It correctly cleared 87.7% of people without it, which fell short of the study’s own preset goal. A recalculation the company did afterward, counting some false alarms differently, reached 91.1%. About 1 in 6 nights produced too little data to judge.

The Samsung feature isn’t meant for people with certain conditions, such as atrial fibrillation, heart failure or COPD, or during pregnancy⁸. It can’t detect central sleep apnea, a type in which the brain doesn’t signal the body to breathe.

Apple’s feature was cleared through the FDA’s 510(k) route on September 13, 2024⁹. It’s for adults 18 and older without an apnea diagnosis. It uses the watch’s motion sensor to spot small breathing disturbances and reviews them in 30-day windows.

Apple’s study enrolled 1,499 people and compared the watch with a home sleep apnea test⁹. The feature caught 66.3% of people with moderate to severe apnea, about two out of three. It rarely raised false alarms: 98.5% of people with no apnea or mild apnea got no alert. The FDA summary states that the lack of an alert does not mean you don’t have sleep apnea.

Neither feature diagnoses apnea⁸⁹. If you get an alert, take it to a clinician, who can order a proper sleep test.

Sleep doctors’ view: a conversation starter, not a diagnosis

In 2018, the American Academy of Sleep Medicine (AASM) issued a position statement on consumer sleep technology¹⁰. Because most devices lacked validation and FDA clearance, it said they could not be used to diagnose or treat sleep disorders. It added that the data may improve conversations between patients and clinicians, as part of a proper evaluation.

The AASM also called for more validation, access to raw data and formulas, and FDA oversight¹⁰. The apnea alerts show progress on oversight. The 2024 review shows the rest is unfinished⁶.

That framing fits values many natural-health readers hold. You know your body better than any app. A tracker can sharpen that self-knowledge, for instance by showing that your shortest nights follow late work or evening drinks. But it can’t feel rested for you, and it shouldn’t overrule what you feel.

When chasing a good score backfires

The term orthosomnia comes from a 2017 paper in the Journal of Clinical Sleep Medicine by sleep researcher Kelly Glazer Baron, now at the University of Utah, and colleagues. They described three patients who were preoccupied with tracking their sleep and still felt unrested during the day². According to a 2024 study that summarized the paper, the patients’ sleep problems worsened as they strictly followed tracker data that often clashed with objective sleep tests³.

How common is it? No one knows, because there’s no agreed way to measure it. A 2024 study of 523 mostly young women put the rate between 3% and 14%, depending on the cutoff, using an unvalidated method³.

The 2026 national survey found nearly 31% of trackers at risk using a two-question screen¹. Its authors said that was roughly twice earlier estimates.

In 2025, Baron and Norwegian researchers published a 12-item Bergen Orthosomnia Scale². It measures how much tracking interferes with daily life and how rigid people are about sleep. Rigidity went along with unhelpful beliefs about sleep and obsessive-compulsive symptoms.

The irony: the leading treatment for chronic insomnia often uses sleep restriction, which limits your time in bed at first and adds time back as sleep improves¹⁴. If you lie in bed longer just to raise a score, you work against that approach.

Trackers can also help. In two small trials led by Baron’s team and published in 2026, adults in coaching programs wore Fitbits to help them sleep longer¹¹. In one trial, 21 of the 29 people who answered (about 72%) said they often or always enjoyed wearing the device, and only 7 (24%) checked their sleep data daily.

No one reported more anxiety from tracking, though some felt discouraged when they missed goals¹¹. The trials’ main sleep results were still being analyzed.

How to read a week of sleep data in 10 minutes

Daily scores invite worry. Pick one morning a week and work through these steps.

1. Fix the start and end of each night. Mark bedtime and wake time by hand when your app allows it. A device can’t know when you started trying to sleep⁶.
2. Average your total sleep. Look at the seven-day average, not last night. Total sleep is the most dependable number⁶.
3. Compare with the targets. The CDC lists at least 7 hours for adults 18 to 60, 7 to 9 hours for ages 61 to 64, and 7 to 8 hours for 65 and older¹².
4. Check your rhythm. Compare bedtimes and wake times on workdays and days off. Clinicians ask about both¹³.
5. Treat “time awake” as a minimum. Trackers tend to miss quiet wakefulness⁴.
6. Glance at stages, then move on. Don’t change your habits because of one week of deep sleep or REM numbers⁶.
7. Add your own notes. Write one line a day on how rested you felt. If data and notes disagree, trust your notes.
8. Act on the signals that matter. An apnea alert, or a partner who hears loud snoring or gasping, deserves a clinician’s attention⁹¹³.

Use this table to weigh each number.

Number on your app	How much to trust it	What to do with it
Total sleep time	Fairly good for trends	Compare the weekly average with your age target
Bedtime and wake time	Good if you mark them yourself	Aim for a steady schedule
Time awake at night	Low; usually undercounted	Trust your own memory of restless nights
Light, deep and REM sleep	Rough; right about half to two-thirds of the time	Ignore night-to-night swings
Sleep or readiness score	Unknown; company formulas aren’t public	Treat it as a nudge, not a verdict
FDA-authorized apnea alert	Reviewed by FDA; misses some cases	Book a visit with a clinician

The evidence at a glance

Claim	Evidence level	What the research shows
Trackers detect when you’re asleep	Strong evidence	At least 93% of sleep detected in lab testing46
Trackers follow total sleep well enough for trends	Strong evidence	Most consistent measure across devices, though less reliable than lab testing56
Trackers measure time awake accurately	Evidence says no	Correctly flagged only 18% to 54% of awake time4
Stage labels are accurate enough to act on	Evidence says no	Stages right about 50% to 70% of the time6
FDA-authorized watch alerts flag moderate to severe apnea	Promising but early	One maker-run study per feature; caught 66% to 83% of cases, and no alert doesn’t rule apnea out89
Tracking can feed sleep anxiety in some people	Promising but early	Case reports and surveys; no agreed measure yet123
Tracking plus coaching helps people sleep longer	Not enough evidence	People used and liked the devices; main trial results pending11
CBT-I treats chronic insomnia	Strong evidence	Strong guideline recommendation based on moderate-quality evidence15

When to see a clinician, and what helps insomnia first

The National Heart, Lung, and Blood Institute (NHLBI) calls insomnia chronic when trouble falling or staying asleep happens at least three nights a week for three months or longer¹³.

For chronic insomnia, the American College of Physicians strongly recommends cognitive behavioral therapy for insomnia, or CBT-I, as the first treatment for all adults, based on moderate-quality evidence¹⁵. Medication comes in only if CBT-I alone doesn’t work, after a shared decision with your clinician.

CBT-I is a structured talk-based program, usually six to eight weeks long, delivered by a doctor, nurse or therapist in person, by phone or online¹⁴. It combines easing anxiety about sleep, relaxation, sleep education, sleep restriction and stimulus control, which means using the bed only for sleep and sex. For readers who prefer drug-free care, it is the first-line option.

Before an appointment, NHLBI suggests keeping a sleep diary for one to two weeks, noting bedtime, wake time, naps, daytime sleepiness, caffeine, alcohol and exercise¹³. Your tracker’s weekly totals can sit beside the diary, but they don’t replace it. Your clinician may add actigraphy, a small wrist motion sensor worn for 3 to 14 days, or a sleep study.

See a clinician if you get an apnea alert, if you snore loudly or wake up gasping whatever your watch says, if poor sleep has lasted three months, or if worry about your data keeps you up⁸⁹¹³.

Questions readers ask

The bottom line

Sleep trackers do one job well: they tell you roughly how long you slept, and over weeks they show real patterns. They miss much of the time you lie awake, and their deep sleep and REM figures are estimates. The FDA-authorized apnea alerts are useful, but they miss some cases.

Use your tracker as a mirror, not a judge. Check it weekly, trust how you feel, and bring lasting sleep trouble, loud snoring or an apnea alert to a clinician. If the numbers start to cost you sleep, set the device aside.

Related reading

• “FDA registered” isn’t “FDA approved”: a buyer’s map to wellness devices
• What a medical test can and cannot tell you
• What to bring, write down and ask before a doctor visit
• What protects your data when HIPAA doesn’t cover most health apps

Sources

1. Goel R, Ejikeme C, Quan S, Weaver MD, Sablan J, Czeisler C, Robbins R. “Exploring the Prevalence of Sleep Tracking and Orthosomnia in a National Survey of Adults in the US” (abstract 0553). Sleep, 2026, Supplement 1. doi.org
2. Guldbrandsen BV, Baron K, Vedaa Ø, Bjorvatn B, Pallesen S. “Development of a scale for measuring orthosomnia: the Bergen Orthosomnia Scale (BOS).” Frontiers in Sleep, 2025. frontiersin.org
3. Jahrami H, Trabelsi K, Husain W, et al. “Prevalence of Orthosomnia in a General Population Sample: A Cross-Sectional Study.” Brain Sciences, 2024. pubmed.ncbi.nlm.nih.gov
4. Chinoy ED, Cuellar JA, Huwa KE, et al. “Performance of seven consumer sleep-tracking devices compared with polysomnography.” Sleep, 2021. pubmed.ncbi.nlm.nih.gov
5. Lee YJ, Lee JY, Cho JH, Kang YJ, Choi JH. “Performance of consumer wrist-worn sleep tracking devices compared to polysomnography: a meta-analysis.” Journal of Clinical Sleep Medicine, 2025 (abstract via TheMedNet). themednet.org
6. de Zambotti M, Goldstein C, Cook J, et al. “State of the science and recommendations for using wearable technology in sleep and circadian research.” Sleep, 2024. pubmed.ncbi.nlm.nih.gov
7. U.S. Food and Drug Administration. “General Wellness: Policy for Low Risk Devices.” Final guidance, January 2026. fda.gov
8. U.S. Food and Drug Administration. “De Novo Classification Request for Sleep Apnea Feature” (DEN230041, Samsung Electronics), decision summary, 2024. accessdata.fda.gov
9. U.S. Food and Drug Administration. “510(k) Summary: Sleep Apnea Notification Feature” (K240929, Apple Inc.), 2024. accessdata.fda.gov
10. Khosla S, Deak MC, Gault D, et al. “Consumer Sleep Technology: An American Academy of Sleep Medicine Position Statement.” Journal of Clinical Sleep Medicine, 2018. link.springer.com
11. Taylor E, Takgbajouah M, Duffecy J, et al., Baron KG. “Using consumer sleep trackers for sleep extension interventions: retention, adherence, and user experiences.” SLEEP Advances, 2026. pubmed.ncbi.nlm.nih.gov
12. Centers for Disease Control and Prevention. “About Sleep.” CDC, updated 2024. cdc.gov
13. National Heart, Lung, and Blood Institute. “Insomnia: Diagnosis.” NIH, updated March 2022. nhlbi.nih.gov
14. National Heart, Lung, and Blood Institute. “Insomnia: Treatment.” NIH, updated March 2022. nhlbi.nih.gov
15. Qaseem A, Kansagara D, Forciea MA, Cooke M, Denberg TD. “Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians.” Annals of Internal Medicine, 2016. pubmed.ncbi.nlm.nih.gov