Why age changes how well a tracker reads your sleep
A sleep tracker doesn't watch your brain. It infers sleep stages from body movement, heart rate, and similar signals, then applies rules learned from people it was tested on. If those people were mostly young, the rules fit young sleep patterns well and older patterns less well. One evaluation of consumer wearables and nearables in healthy young and older adults found the devices were much less accurate in older adults, and the authors tie this to the fact that most sleep-tracking development has been done for young adults [1].
The mismatch is not just about age in the abstract; it is about what older sleep looks like on those sensors. Older adults spend more of the night awake after falling asleep and get less deep sleep [3]. A device that expects a young person's rhythm has to interpret a sparser, more fragmented signal, which makes stage labels harder to get right.
The disagreement is biggest for deep sleep
When trackers are compared with polysomnography, the broad sleep-versus-wake call often holds up better than the finer stage labels. In a lab comparison of several trackers, the best-performing consumer device reached a two-stage (sleep/wake) agreement score of 0.64, while four-stage agreement was lower [3]. That pattern matters for older adults because the stage most affected by age is deep sleep, the N3 stage, which older adults get less of [3][4].
So the practical weak spot is not "did I sleep?" but "how much deep sleep did I get?" A tracker can be roughly right about whether you were asleep and still be off on the stage breakdown, and that error is likely to be larger in an older person whose deep sleep is already scarce.
What this means if you use a tracker
Treat stage-level numbers from a consumer device as a rough trend rather than a measurement, especially if you are older. The evidence points to devices being less accurate in older adults overall [1], and the stage where they are least reliable is the one older adults have least of [3]. Total sleep time and sleep-versus-wake patterns are the safer things to watch; the deep-sleep figure is the shakiest.
It is also worth knowing that the comparison studies are mostly lab validations in healthy people, so they describe how the devices perform under controlled conditions rather than in every home and every health situation. The age gap is a consistent signal across these evaluations, but the exact size of the error will vary by device and by person.
Why the age comparison is still thinly studied
Direct young-versus-old comparisons of consumer trackers against polysomnography are few. Much of the validation work has been done in adolescents or young adults, and some of those studies explicitly call for including older people in future work [2]. That leaves the older-adult side of the picture supported by a smaller set of studies, which is why the finding is best stated as a consistent tendency rather than a precise, device-by-device number.
The broader sleep-architecture data do confirm the underlying age difference: older adults show shorter sleep, more wake after sleep onset, and less N3 and REM sleep [4]. That is the biological backdrop against which a tracker's errors play out, and it is why the same device can look good in a 25-year-old and mediocre in a 70-year-old.
Sources used in this answer
Performance evaluation of consumer sleep-tracking wearables and nearables in healthy young and older adults
Consumer sleep-tracking wearables and nearables were much less accurate against polysomnography in older adults than in young adults, with development largely focused on young adults.
Validation of a consumer sleep wearable device with actigraphy and polysomnography in adolescents across sleep opportunity manipulations
A consumer sleep wearable was validated against polysomnography and actigraphy in adolescents across sleep-opportunity manipulations, but the study did not include older adults.
Selecting a sleep tracker from EEG-based, iteratively improved, low-cost multisensor, and actigraphy-only devices
Among several trackers compared with polysomnography, the best consumer device reached a two-stage sleep/wake agreement score of 0.64, with lower four-stage agreement, and older adults had more wake after sleep onset and less deep sleep.
The Human Sleep Project: A multi-center clinical polysomnography dataset across the human lifespan
A large multi-center polysomnography dataset showed older adults had shorter sleep, more wake after sleep onset, and less N3 and REM sleep, but it did not involve consumer trackers.
Altered sleep-stage dynamics and reduced EEG power in insomnia-related sleep-state misperception
In insomnia-related sleep-state misperception, polysomnography and a consumer sleep tracker were used to examine sleep stages, but the study did not compare age groups.