Non-Wearable Sleep Sensors Show Promise in Older Patients with Dementia

The sensors could potentially lead to new treatments as well as early disease detection.

WASHINGTON — The measurement of sleep quality with non-wearable sleep sensors effectively and accurately detected sleep problems in older patients with dementia who resided in skilled nursing facilities, according to new data presented at the Alzheimer’s Association International Conference.

“Non-wearable sleep assessment shows considerable scientific and practical promise,” study researcher Jarod T. Giger, PhD, told Psychiatry Advisor. “Sleep sensors were well tolerated by patients and staff; captured individual and group sleep growth trends; and were strongly associated with caregiver self-report sleep data.”

Giger, of the University of Kentucky College of Social Work, Lexington, KY., and colleagues conducted the pilot study to better understand whether a sensor system that does not utilize wearable technology could detect sleep efficiency in older patients with dementia.

“Most sleep assessment in this population relies on self-report data and wearable technology, such as the wrist actigraphy,” Giger said. “Relying on self-report data, which can be unreliable, and strapping wearable technology on persons with dementia did not seem practical, particularly due to the recent advances in sensor and information technology and statistical computing.

For the study, Giger and colleagues tested the feasibility of unobtrusively and passively capturing sleep assessment in 10 older patients (age, 84.3 years) with moderately severe dementia who resided in a dementia special care unit. They collected 63 days of continuous sleep efficiency data from the system, as well as monthly subjective measures of sleep quality using the Pittsburg Sleep Quality Index (PSQI).

Patients with lower efficiency scores on sensor data (range 0-100) had higher levels of restlessness while in bed.

Results indicated that sleep efficiency had a significant linear effect over time (P=.023), suggesting a temporal decrease. Also significant was the variance of random intercepts (P=.001), indicating a significant variation in baseline efficiency score across the study population.

Furthermore, as sleep efficiency decreased, researchers observed an increase in PSQI scores, signifying poorer sleep quality.

“Non-wearable sleep assessment with older adults with dementia appears possible and could translate into novel interventions/treatments and perhaps empirical early illness detection,” Giger said.

Going forward, family caregivers of persons with dementia, and other neurodegenerative diseases such as Parkinson’s, could be a strong ally in moving this line of research forward, he said.

“Imagine if Jane Does husband is showing early signs and symptoms dementia. Unobtrusive and passive longitudinal measurement of his sleep hygiene could help with care and potentially allow for empirical measurement of disease progression,” Giger said. 


Giger J, et al. Abstract #3172. Non-Wearable Sensors to Detect Sleep Efficiency in Older Adults with Dementia: A Pilot Study.Presented at: Alzheimer’s Association International Conference 2015. July 18-23, 2015; Washington, D.C.