Schmitter-Edgecombe M, Luna C, Dai S, Cook DJ. Predicting daily cognition and lifestyle behaviors for older adults using smart home data and ecological momentary assessment. Clin Neuropsychol. Mar 19 2024:1-25. doi:10.1080/13854046.2024.2330143
This study assessed whether digital markers extracted from passive home sensor data could predict cognitive fluctuations, lifestyle behaviors, and risk and protective factors for cognitive decline. Here, 44 midlife and older adults (70% female ranging from 48-89 years old) were monitored in their homes using smart sensors for 3-4 months. The sensor data was categorized into eight digital markers: time of day, day of the week, most recent location when survey prompts were sent, total sensors activated, and time spent in specific locations of the home. These categories allowed for the exploitation of which theoretical digital marker best predicted real-time cognitive ability. In addition to the passive data, the researchers collected active responses to ecological momentary assessment prompts. These prompts were sent semi-randomly four times a day for two weeks. Prompts included behavioral surveys as well as cognitive assessments. Aspects of the sensor data, such as overall activity level and time spent outside successfully predicted fluctuations in cognitive ability as measured in the iPad prompts. Smart home digital markers may be valuable predictors of cognitive health and lifestyle behaviors in aging populations. This method of sensor data collection might be particularly advantageous in populations where remembering to wear sensors, such as watches or rings, may pose a challenge.