Agitation Monitoring in Alzheimer Disease: Passive Sensing Shows Early but Limited Promise
Key Clinical Summary
- Passive sensing technologies (smartphones, wearables, in-home sensors) may help detect agitation, loss of pleasure, and other subjective symptoms in Alzheimer disease and related dementias (AD/ADRD).
- Evidence remains preliminary; most studies are proof-of-concept with significant challenges in validation, data standards, usability, and ethical implementation.
- Improved harmonization and user-centered design could support future clinical translation, particularly for long-term care settings managing behavioral symptoms.
A recent study highlights the expanding role of passive sensing technologies to assess subjective experiences—including agitation, sadness, and anhedonia—in persons living with AD/ADRD. As traditional self-report becomes less reliable with cognitive decline, investigators emphasize the need for innovative measures that can capture behavioral symptoms central to caregiver burden and long-term care utilization.
Passing Sensing Methods and Evidence to Date
Passive sensing refers to the automatic collection of behavioral or physiological data without active patient participation. The commentary outlines three primary device categories currently explored in dementia research: smartphones, wearables, and in-home sensors. Smartphones can capture GPS location, accelerometry, Bluetooth connections, typing patterns, and speech features. Wearable devices provide heart rate variability, sleep parameters, movement data, and physiologic signals such as galvanic skin response. In-home sensors—including infrared motion detectors, pressure sensors, microphones, and 3-D cameras—monitor mobility, ambulation speed, room-to-room transitions, and social or vocal behavior.
Pilot studies suggest these tools may reflect within-person dynamics of agitation, apathy, loneliness, or early cognitive change. For example, home-based mobility sensors have been used to examine fluctuations in agitation and apathy, while other systems have detected changes in time spent in specific rooms that correspond with functional decline. Additional exploratory work links Bluetooth-based social proximity, actigraphy, and GPS patterns to mood states in non-dementia samples, with early efforts underway in AD/ADRD populations.
Despite enthusiasm, the authors underscore that passive sensing remains largely in the exploratory stage. Validation is hindered by the lack of reliable “gold standards,” since self-report, caregiver ratings, and clinician assessments all carry biases—especially as cognitive impairment progresses. Replication is limited, device capabilities vary widely, and the computational procedures required to convert raw sensor data into clinically meaningful features lack standardization. Initiatives such as findability, accessibility, interoperability, and reusability (FAIR) data principles and harmonization platforms like the Collaborative Aging Research Using Technology (CART) Initiative are proposed as pathways to improve reproducibility and cross-study comparison.
Ethical considerations are substantial. Informed consent complexities, privacy risks, device usability challenges in frail older adults, and caregiver workload must be carefully addressed. The authors emphasize the need for user-centered design, community engagement, and clearer guidance for consent when cognitive impairment limits decisional capacity.
Clinical Implications
For long-term care providers, the potential value of passive sensing lies in its ability to continuously monitor agitation and related behavioral symptoms—key drivers of caregiver distress, acute transfers, and institutionalization. Such tools could eventually support earlier intervention, better tracking of treatment response, and individualized care plans. Continuous sensing may also help identify environmental or contextual triggers for behavioral symptoms, enabling more precise nonpharmacologic strategies.
However, current evidence is insufficient for routine clinical adoption. Challenges include cost, data management, alert fatigue, and the need for clear clinical pathways when sensors detect concerning behavior. Future integration into long-term care will depend on validated algorithms, standardized data practices, caregiver-friendly interfaces, and alignment with regulatory and privacy frameworks.
Conclusion
Passive sensing offers a promising yet unproven approach to monitoring agitation and other subjective symptoms in AD/ADRD. With coordinated efforts in validation, ethics, and design, these technologies may eventually support more personalized and responsive care models in long-term care settings, addressing some of the field’s most persistent challenges.
Reference
Depp C, Holden J, Granholm E. Digital measurement of subjective experiences in Alzheimer disease and related dementias (AD/ADRD). JMIR Aging. 2025 Nov 18;8:e71920. doi:10.2196/71920.
