mind & bodyanalysisThe Conversation
Does your sleep tracker get it right? A Pittsburgh neuroscientist weighs in
Consumer wearables estimate sleep from heart rate, its variability, movement and skin temperature, not from brain activity. Against polysomnography, the clinical gold standard that records brain waves, eye movements and muscle tone, they are fairly accurate on a basic question: a 2025 meta-analysis of 24 studies found they overestimate total sleep by about 17 minutes a night. Their weak point is quiet wakefulness, when a person lies still and awake but gets scored as asleep. Sleep-stage tracking, such as how much deep or REM sleep you had, is much less reliable, and studies of six wrist devices found only modest agreement. Trackers do capture total sleep duration well, which matters: a Fitbit study of over 600 students found that those sleeping six hours or less a night early in term had lower grades later.
Wearables use sensors on the wrist or finger to record heartbeats, movement and temperature. Algorithms look for patterns that differ between wake, light, deep and REM sleep, such as the heart's rhythm in REM. Lying still while awake looks similar to sleep to these sensors, which is why they overcount sleep.
Trackers are useful for spotting trends in how long you sleep, but nightly stage numbers are noisy. Because the devices are often replaced before studies catch up, their readings should be read as rough guides.
Illustrative photo · PamD, CC BY-SA 4.0, via Wikimedia Commons
