Guides · Free guide, no sign-up

Which Wearable Numbers Deserve Your Attention

Your ring or watch reports dozens of numbers every morning. A few come close to what a sleep lab or heart monitor would measure. Most are estimates. And the numbers that worry people most are the ones the devices are worst at. This guide sorts them, using independent studies that tested wearables against lab equipment.

Carlyle Jenkins, AHPRA registered chiropractor and lifestyle medicine practitioner

What your device measures well, and what it guesses

The best results published for current devices. Your own device may do worse.

Number	How close to the lab	How to use it
Overnight resting heart rate	Within about 2 to 3% of a heart monitor on the best-tested devices	Watch your own trend
Overnight heart rate variability (HRV)	Within about 6 to 8% on the best-tested ring, wider on others	Trend over weeks only
Total sleep time	About 17 minutes short on the wrist on average and in the same direction, though single nights vary much more	Weekly average
Bedtime and wake time	Falling asleep is picked up within a few minutes on average. Waking is less reliable	How regular your week is
Time awake in the night	Poor on every device tested. Lying still awake often counts as sleep	Ignore the number
Deep sleep and REM minutes	Poor on every device tested. In one study of patients, the best device matched the sleep lab for about half the night	Ignore
Sleep, readiness, recovery and stress scores	No study by anyone has tested them against anything	A prompt to look, not a measurement
Calories burned	More than 10% out in every group studied	Ignore
Blood oxygen	Tends to read high, and can miss low oxygen	Not a reassurance. Reflectance sensors are modelled to be least accurate on darker skin, and almost no validation has included it
VO2max estimate	In the one watch tested against a lab test, it flattered the less fit and squeezed everyone towards the middle. Most makers' versions are untested	Not a fitness grade
Skin temperature, breathing rate	No study has tested them against a reference	No data either way

The four numbers worth watching

Overnight resting heart rate. The best-measured number your device produces. A rise of about three beats or more that holds for a week, against your own baseline, is worth explaining. Illness is a well-documented reason. Alcohol, poor sleep, heat and heavy training are other common explanations. They are not the only ones. See your doctor if the rise comes with chest pain, breathlessness, palpitations, dizziness or fainting, or swollen ankles, or if it has not settled after about two weeks of normal living. Chest pain, severe breathlessness or fainting is a same-day matter, not a trend to watch.
Overnight HRV, as a trend. Compare this week's average with your last one to two months. A move of less than about 10% sits inside the device's error. That is a rule of thumb taken from the published error, not a tested threshold.
Total sleep, as a weekly average. Is your average where you mean it to be? In the one study that measured this, a ring sorted 81 to 93% of nights correctly into under six hours, six to seven, or over seven. That is the level of detail to expect, not 6 hours 40 against 7 hours.
Sleep timing. How much your bedtime and wake time move across the week. Devices are good at telling that you slept, even when they cannot say how well.

How to read any of them

Against your own baseline. Never against someone else's number or a population average.
As a seven-day average set against your last 30 to 60 days.
Never act on one night.
Not across devices. Different devices disagree with each other, so compare a device only with itself.

Why the worrying numbers are the shaky ones

Time awake in the night and deep sleep are the two numbers most likely to worry you in the morning. They are also among the least accurate things your device reports, and they get less accurate on exactly the nights your sleep is broken.

Chasing a better sleep score can make sleep worse. Sleep clinicians have a name for it: orthosomnia.

What a wearable cannot tell you

Whether you have a sleep disorder. The American Academy of Sleep Medicine's position is that these devices cannot be used to diagnose or treat a sleep disorder.
That you do not have sleep apnoea. Watch features that flag it can miss it, moderate cases especially. A flag is worth taking to your doctor. No flag is not an all-clear. Loud snoring, breathing pauses someone else has noticed, waking unrefreshed, or daytime sleepiness, especially at the wheel, are worth assessing whatever your device says.
Whether an irregular rhythm alert is a heart rhythm problem. It is a prompt to get an ECG, not a diagnosis. No alerts is not an all-clear either, because these rhythms come and go. Palpitations, breathlessness, chest discomfort, dizziness or fainting deserve assessment whatever your device reports, and fainting or chest pain means urgent care.
How accurate it is for you. Most testing has been done on younger, healthy adults, often over a single night. Across 186 validation studies the typical share of people over 65 was zero. Treat every figure on this page as the best case.

Let your AI read your export

Most apps let you export your data from their settings. Remove your name and contact details first, check the privacy settings of the AI you use, then attach the file with this prompt.

Prompt

I have attached my exported wearable data. Do not diagnose or give medical advice. Use only these four measures and ignore everything else, including sleep stages, time awake, calories, blood oxygen and any score.

1. Overnight resting heart rate.
2. Overnight heart rate variability (HRV).
3. Total sleep time.
4. Bedtime and wake time.

For each, compare my average for the last 7 days with my average for the 30 to 60 days before that. Show both numbers and the difference.

Mark a measure as changed only if resting heart rate is up or down by more than about three beats per minute, HRV has moved by more than about 10%, average sleep has moved by more than about an hour, or my bedtime or wake time has shifted by more than about an hour. Otherwise say it is within normal variation for the device. Then tell me plainly that these four cut-offs are rules of thumb and not tested thresholds: no study has established how large a change has to be before it is real for any consumer device. The heart rate and HRV figures come from the measurement error of the best-tested devices, and most devices are worse, so on mine a 10% HRV move may still be noise. The sleep and timing figures are judgements about what matters in a routine, not measurements. Add that no change in these four measures does not mean nothing is wrong, and that if I feel unwell I should see my doctor whatever the data shows.

If resting heart rate has risen for a week or more, remind me that illness, alcohol, poor sleep, heat and heavy training are common reasons, and that I should see my doctor if it comes with chest pain, breathlessness, palpitations, dizziness, fainting or swollen ankles, or has not settled after about two weeks. Add that chest pain, severe breathlessness or fainting is a same-day matter, not a trend to watch.

Keep it to a short table and three sentences.

Keep in mind

Every figure here applies to a specific device and software version, and companies change their algorithms faster than researchers test them. Some newer devices have not been tested at all. Checked October 2026.

If you feel unwell, or a number worries you, talk to your doctor. A wearable cannot rule anything in or out.

Nothing here is a diagnosis, a treatment or medical advice, and it does not replace your GP or your specialist. I am a lifestyle medicine practitioner sharing general information, not your treating doctor.

The data is the easy part.

Inside Xone Health your wearable data sits beside your bloods, genetics, glucose and food data in one picture, so we can see which numbers move together and which deserve your attention first.

Where to go from here

1
Take the Energy Check
Five minutes, scored on the spot.
2
Discover more about Xone Health
How I work, who it suits and what it costs.

Sources

Dial MB, Hollander ME, Vatne EA, et al. Validation of nocturnal resting heart rate and heart rate variability in consumer wearables. Physiological Reports. 2025;13(16):e70527. doi.org/10.14814/phy2.70527
Lee YJ, Lee JY, Cho JH, et al. Performance of consumer wrist-worn sleep tracking devices compared to polysomnography: a meta-analysis. Journal of Clinical Sleep Medicine. 2025;21(3):573-582. doi.org/10.5664/jcsm.11460
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
Willoughby AR, Golkashani HA, Ghorbani S, et al. Performance of wearable sleep trackers during nocturnal sleep and periods of simulated real-world smartphone use. Sleep Health. 2024;10(3):356-368. doi.org/10.1016/j.sleh.2024.02.007
Herberger S, Aurnhammer C, Bauerfeind S, et al. Performance of wearable finger ring trackers for diagnostic sleep measurement in the clinical context. Scientific Reports. 2025;15(1):9461. doi.org/10.1038/s41598-025-93774-z
Miller DJ, Sargent C, Roach GD. A validation of six wearable devices for estimating sleep, heart rate and heart rate variability in healthy adults. Sensors. 2022;22(16):6317. doi.org/10.3390/s22166317
Choe JP, Kang M. Apple watch accuracy in monitoring health metrics: a systematic review and meta-analysis. Physiological Measurement. 2025;46(4). doi.org/10.1088/1361-6579/adca82
Jiang Y, Spies C, Roghanizad AR, et al. Performance of wearable pulse oximetry during controlled hypoxia induction: instrument validation study. JMIR Formative Research. 2026;10:e85253. doi.org/10.2196/85253
Al-Halawani R, Qassem M, Kyriacou PA. Monte Carlo simulation of the effect of melanin concentration on light-tissue interactions in reflectance pulse oximetry. Sensors. 2025;25(2):559. doi.org/10.3390/s25020559
Caserman P, Yum S, Göbel S, et al. Assessing the accuracy of smartwatch-based estimation of maximum oxygen uptake using the Apple Watch Series 7: validation study. JMIR Biomedical Engineering. 2024;9:e59459. doi.org/10.2196/59459
de Zambotti M, Rosas L, Colrain IM, et al. The sleep of the ring: comparison of the ŌURA sleep tracker against polysomnography. Behavioral Sleep Medicine. 2019;17(2):124-136. doi.org/10.1080/15402002.2017.1300587
Baron KG, Abbott S, Jao N, et al. Orthosomnia: are some patients taking the quantified self too far? Journal of Clinical Sleep Medicine. 2017;13(2):351-354. doi.org/10.5664/jcsm.6472
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;14(5):877-880. doi.org/10.5664/jcsm.7128
Apple Inc. Estimating Breathing Disturbances and Sleep Apnea Risk from Apple Watch (Sleep Apnea Notifications on Apple Watch). Apple Inc. September 2024. www.apple.com/health/pdf/sleep-apnea/Sleep_Apnea_Notifications_on_Apple_Watch_September_2024.pdf
US Food and Drug Administration. De Novo Classification Order and Decision Summary, DEN230041: Samsung Sleep Apnea Feature (Samsung Electronics Co Ltd). US FDA. Granted 6 February 2024. www.accessdata.fda.gov/cdrh_docs/reviews/DEN230041.pdf
Perez MV, Mahaffey KW, Hedlin H, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. New England Journal of Medicine. 2019;381(20):1909-1917. doi.org/10.1056/NEJMoa1901183
Lubitz SA, Faranesh AZ, Selvaggi C, et al. Detection of atrial fibrillation in a large population using wearable devices: the Fitbit Heart Study. Circulation. 2022;146(19):1415-1424. doi.org/10.1161/CIRCULATIONAHA.122.060291
Schipper RM, Oumar Djibrillah F, Roosink M, et al. The representation of different populations in studies assessing the validity of consumer wearable photoplethysmography-based measurements: scoping review. JMIR mHealth and uHealth. 2026;14:e87324. doi.org/10.2196/87324

Searched on PubMed on 12 October 2026 and found no validation study, by anyone, for readiness, recovery, sleep or stress scores, for wrist or finger temperature against a reference thermometer, or for consumer wearable breathing rate against a reference.