Wrist actigraphy provided acceptable estimates of average total sleep time for epidemiological monitoring but showed substantial errors for sleep onset latency and wake after sleep onset when validated against in-home portable EEG in community-dwelling older adults.
Key Findings
Results
Actigraphy slightly overestimated total sleep time and sleep efficiency compared to portable EEG in community-dwelling older adults.
49 community-dwelling older adults contributed 217 nights of concurrent actigraphy and portable EEG recordings
Monitoring was conducted for five consecutive nights when possible, extended to up to seven nights when feasible
Actigraphy overestimated TST and sleep efficiency relative to EEG reference measurements
Actigraphy underestimated SOL and WASO compared with EEG
Results
Intraclass correlation coefficients for TST were acceptable at the single-night level and improved substantially when averaged across nights.
Single-measure ICC for TST was 0.73
Average-across-nights ICC for TST was 0.84
Single-measure ICC for WASO was only 0.38
Average-across-nights ICC for WASO was 0.55
Agreement was assessed using Bland-Altman plots and multilevel modeling
Results
Mean absolute percentage error for TST was 11%, indicating acceptable accuracy, while MAPE for SOL and WASO exceeded 50%, indicating poor accuracy.
MAPE for TST was 11%, classified as acceptable for epidemiological use
MAPE for SOL exceeded 50%
MAPE for WASO exceeded 50%
High MAPE values for SOL and WASO indicate substantial measurement error for these indices
Results
Wrist actigraphy showed poor reliability and accuracy for sleep onset latency and wake after sleep onset in older adults.
Single-measure ICC for WASO was 0.38, well below conventional thresholds for acceptable reliability
MAPE exceeded 50% for both SOL and WASO
The flexible sleep-wake schedules common in older adults may contribute to actigraphy's reduced validity for these parameters
Authors note these findings are specific to relatively healthy community-dwelling older adults and should not be generalized to those with clinically significant sleep disorders
Results
Averaging actigraphy measurements across multiple nights substantially improved reliability for both TST and WASO.
TST ICC improved from 0.73 (single night) to 0.84 (averaged across nights)
WASO ICC improved from 0.38 (single night) to 0.55 (averaged across nights)
Multi-night averaging is recommended to improve the reliability of actigraphy-derived sleep estimates
Even with averaging, WASO reliability remained below levels typically considered acceptable for individual-level assessment
Conclusions
The study concluded that wrist actigraphy is suitable for estimating average TST at a group or population level but is not sufficiently accurate for individual-level assessment of SOL or WASO.
Authors state actigraphy 'provided acceptable estimates of average TST for epidemiological monitoring'
Authors note 'substantial errors for SOL and WASO'
Findings are limited to relatively healthy community-dwelling older adults
Authors explicitly caution that results 'should not be generalized to older adults with clinically significant sleep disorders or broader clinical populations'
What This Means
This research suggests that wrist-worn actigraphy devices — which measure sleep by tracking wrist movement — provide reasonably accurate estimates of how long older adults sleep overall (total sleep time), but are much less reliable for measuring how long it takes to fall asleep or how much time is spent awake during the night. The study had 49 healthy older adults living in their own homes wear both an actigraphy device and a portable brainwave (EEG) monitor simultaneously for up to seven nights, allowing researchers to compare the two measurements directly in a real-world setting. The EEG served as the more accurate reference standard.
For total sleep time, actigraphy performed reasonably well, with an error rate of about 11% and good agreement with EEG when measurements were averaged across multiple nights. However, for sleep onset latency (how long it takes to fall asleep) and wake after sleep onset (time spent awake during the night), errors exceeded 50%, meaning actigraphy readings could be very misleading for these measures. Actigraphy also tended to overestimate how much people slept and underestimate how long they lay awake.
This research suggests that consumer and research-grade wrist actigraphy devices are useful tools for tracking general sleep duration trends in groups of older adults — such as in large epidemiological studies — but may not be reliable enough to assess individual sleep quality, particularly for measuring disrupted or fragmented sleep. People and researchers using actigraphy to evaluate insomnia symptoms or nighttime wakefulness in older adults should be aware of these limitations, and the findings may not apply to older adults with diagnosed sleep disorders.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Deguchi N, Hatanaka S, Daimaru K, Maruo K, Sasai H. (2026). In-home validation of wrist actigraphy against portable electroencephalography for sleep assessment in older adults.. Sleep & breathing = Schlaf & Atmung. https://doi.org/10.1007/s11325-026-03790-0