Exercise & Training

A One-Year Study Using Digital Biomarkers From Sensing Technologies to Assess Changes in Physical Activity Levels and Sleep Quality in Nursing Home Residents With Dementia: Observational Study.

TL;DR

Sensing technologies (smartwatch and radar-based system) demonstrated high adherence and acceptability in nursing home residents with dementia, detecting statistically significant changes in nighttime physical activity and four sleep biomarkers over one year, though long-term group-level reliability was poor, warranting cautious use for clinical decision-making.

Key Findings

Statistically significant differences were found in nighttime Euclidean norm minus one (ENMO) over the one-year observation period.

  • ENMO is a measure of physical activity level derived from accelerometer data
  • Nighttime ENMO showed a statistically significant difference across time points (P=.01)
  • 9 participants were included in the final analysis, recruited from 2 dementia care units in Norway
  • Data were collected at baseline, 6 months, and 1 year using a Garmin smartwatch (Vivoactive 5 or Venu 3)

Four sleep biomarkers showed statistically significant differences over the one-year study period.

  • Total sleep time (TST) differed significantly across time points (P=.02)
  • Sleep efficiency (SE) differed significantly (P=.02)
  • Wake after sleep onset (WASO) differed significantly (P=.01)
  • Sleep regulatory index (SRI) differed significantly (P=.01)
  • Sleep data were collected using a radar-based system (Vital Things, Somnofy) over 6 nights at each time point

Long-term group-level reliability of ENMO was poor, while reliability between individuals at each time point was moderate to strong.

  • Intraclass correlation coefficient (ICC) for group ENMO across the study was 0.00–0.02, indicating poor long-term reliability
  • ICC between individuals at each individual time point was moderate to strong at 0.58–0.79
  • This discrepancy suggests that ENMO can reliably differentiate individuals at a given moment but may not reliably track changes within groups over time
  • The authors concluded this finding requires recommendation for 'cautious, well-designed use of digital biomarkers for clinical decision-making'

Adherence and acceptability of the sensing technologies was high, with no adverse events reported.

  • Adherence and acceptability rates ranged from 88% to 96%
  • No adverse events were reported among participants
  • Participants ranged in age from 79 to 93 years
  • The application of devices was described as 'well tolerated by the participants'
  • 11 participants were initially recruited, with 9 included in the final analysis

The study used a multimodal sensing approach combining wrist-worn accelerometry and radar-based sleep monitoring to generate digital biomarkers in a real-world nursing home environment.

  • Smartwatches used were Garmin Vivoactive 5 or Garmin Venu 3 for physical activity data
  • A radar-based system (Vital Things, Somnofy) was used for sleep monitoring
  • Digital biomarkers assessed included ENMO, sleep efficiency (SE), wake after sleep onset (WASO), sleep regulatory index (SRI), sleep fragmentation index (SFI), total sleep time (TST), and time out of bed (no presence)
  • Observations were designed to be 'aligned with real-world conditions in which such sensing technologies would be applied within a nursing home environment'
  • The Personal Self-Maintenance Score and Neuropsychiatric Inventory-Nursing Home Version (nighttime behaviors section K) were also administered as proxy-rated measures

Proxy-rated questionnaires remain the current standard for assessing activity and sleep in people with dementia in nursing homes, and sensing technologies offer a complementary objective approach.

  • The study background notes that 'proxy-rated questionnaires remain the standard for the assessment of activity and sleep for people with dementia living in nursing homes'
  • Sensing technologies 'can generate continuous data that provide quantitative insights into daily activities and behavioral and psychological symptoms, such as sleep disturbance'
  • The study aimed to explore 'long-term capabilities of multimodal sensing technologies for assessing physical activity levels and sleep quality using selected digital biomarkers'
  • The authors suggest the use of sensing technologies 'could enable more objective, data-driven future care models for people with dementia residing in nursing homes'

What This Means

This research suggests that wearable smartwatches and radar-based sleep monitors can successfully detect meaningful changes in physical activity and sleep patterns in elderly nursing home residents with dementia over the course of one year. The study, conducted in Norway with 9 participants aged 79–93, found significant changes in nighttime movement and four measures of sleep quality — including total sleep time, sleep efficiency, and how often residents woke up during the night — when comparing data collected at the start of the study, at six months, and at one year. Importantly, the devices were well accepted by participants, with adherence rates between 88% and 96% and no adverse events, suggesting this kind of continuous, passive monitoring is feasible even in a vulnerable, older population with dementia. However, the study also found important limitations in how reliably the devices tracked changes across the group over time. While the technology was reasonably good at distinguishing differences between individuals at any given moment, it was poor at consistently tracking the same individuals' activity levels across the full year at the group level. This means that while the sensors show promise, results should be interpreted carefully and the technology is not yet ready to replace clinical judgment or serve as a standalone tool for making care decisions. This research matters because people with dementia often cannot self-report how well they are sleeping or how active they are, and current methods rely on caregivers filling out questionnaires, which can be subjective and time-consuming. Continuous, objective monitoring using everyday wearable technology could one day help nursing home staff detect early signs of decline, sleep disturbance, or changes in behavior more quickly and accurately — potentially enabling more personalized and responsive care. The authors emphasize that more carefully designed studies are needed before digital biomarkers from such devices can be used reliably in clinical practice.

Check Your Own Numbers

Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.

Upload Your Labs

Have a question about this study?

Citation

Boyle L, Patrascu M, Husebo B, Haugarvoll K, Steihaug O, Marty B. (2026). A One-Year Study Using Digital Biomarkers From Sensing Technologies to Assess Changes in Physical Activity Levels and Sleep Quality in Nursing Home Residents With Dementia: Observational Study.. JMIR nursing. https://doi.org/10.2196/95194