Technology and Dementia Preconference

Area of Science:

• Neuroscience
• Gerontology
• Biomedical Engineering

Background:

• Alzheimer's disease and related dementias (ADRD) significantly impact sleep-wake cycles.
• Commercial wearable devices offer a scalable method for monitoring these changes through passive actigraphy.
• Characterizing rest-activity patterns can provide insights into cognitive decline.

Purpose of the Study:

• To characterize continuous rest-activity patterns using Fitbit™ data.
• To examine differences in these patterns between healthy adults and individuals with ADRD.
• To identify potential biomarkers for cognitive and functional decline.

Main Methods:

• Collected tri-axial actigraphy data (step counts), clinical, cognitive, functional, and mood information from healthy adults, mild cognitive impairment (MCI), Alzheimer's disease dementia (AD), and frontotemporal lobar degeneration (FTLD) cohorts.
• Quantified activity patterns using rest-activity aggregates and minute-level features.
• Applied Principal Component Analysis (PCA) for data reduction and identified five key components explaining over 85% of variance.

Main Results:

• Principal Component 1 (PC1), representing activity variability and amplitude, was negatively associated with cognitive and functional decline scores (CDR®+FTLD-NACC).
• PC1 also showed significant differences between diagnostic groups, with FTLD patients exhibiting reduced PC1 scores compared to healthy and MCI cohorts.
• PC3, related to rest-activity start timing, was negatively associated with the global cognitive score.

Conclusions:

• Activity variability and amplitude derived from Fitbit™ data can distinguish FTLD patients.
• These distinct activity profiles differ from those observed in individuals with mild cognitive impairment or those who are functionally intact.
• Wearable actigraphy shows promise for identifying syndrome-specific rest-activity patterns in neurodegenerative diseases.