JMIR Publications

ABSTRACT

Background:

Despite a growing need for digital services to care for geriatric mental health, the development and implementation of digital mental healthcare system for older adults have been hindered by a lack of studies among older adult users and community caregivers.

Objective:

This study aimed to identify whether digital sensing data on heart rate variability (HRV), sleep quality, and physical activity predicts same day or next day depressive symptoms among older adults in their ordinary living environments. In addition, this study tested the feasibility of a digital mental health monitoring platform to inform day-to-day changes in the health status of older adult users and their community caregivers.

Methods:

A living lab pilot study was conducted with community-dwelling older adults (n = 25) and their community caregivers (n = 16) for 6 weeks. Depressive symptoms were assessed daily using PHQ-9 via scripted verbal conversations using a smartphone application chatbot, and digital biomarkers of depression, including HRV, sleep, and physical activity, were assessed using a wearable sensor (Fitbit Sense), which was worn continuously, except charging time, for 6 weeks. Daily individualized feedback on the health status of older adult users was displayed on the applications for the users and their community caregivers. Multilevel modeling (MLM), with within-person changes over time at Level 1 and between-person differences at Level 2, was utilized to examine whether the digital biomarkers predict same day or next day depressive symptoms after adjusting for age, gender, baseline depression, and chronic disease conditions.

Results:

The MLM results showed that fluctuations in daily sleep fragmentation and sleep efficiency were associated with an increased probability of next day depressive symptoms for older adults. The feasibility test results indicated that older adults were able to use one or two functions of the monitoring platform 6 out of 7 days per week. However, the usability levels of older adults decreased from pre to post living lab due to experienced difficulties.

Conclusions:

The findings indicate the feasibility of digital mental health monitoring platforms for socially vulnerable older adults. The results also suggest that wearable sensor assessments of sleep fragmentation and efficiency can be important indicators for passively sensing daily fluctuations of depressive symptoms in older adults.