LONELINESS can be continuously and objectively monitored using mobile and wearable technologies, according to a recent study. This innovative approach could transform how mental health issues are tracked and addressed, offering new hope for targeted public health interventions.
Researchers monitored 30 college students over two months, utilising smartphones to track behavioural patterns like location changes, notification types, and incoming and outgoing calls and text messages. Simultaneously, smartwatches and rings collected physiological data, including heart rate, heart rate variability, and sleep duration. Participants also reported their feelings of loneliness multiple times daily through a questionnaire app on their phones.
With this data, the researchers trained a random forest machine learning model to predict loneliness levels. The model obtained an accuracy of 82%, a precision score of 81%, and an area under curve (AUC) of 0.83. Behavioural data from smartphones emerged as the most significant predictor of loneliness across all participants, underscoring the crucial role of daily activities and social interactions. Additionally, physiological data from smartwatches and smartrings provided valuable insights, especially regarding participants’ health and sleep patterns. For example, physiological features were the most predictive of loneliness in Participant 8, whereas Participant 10’s most important predictors were behavioural features.
This study highlights the use of digital approaches to provide real-time data to healthcare professionals, which reduces the burden on patients. As mobile and wearable technologies advance, their application in mental health monitoring could become standard, and lead to timely, personalised interventions in the psychiatric field.
Katrina Thornber, EMJ
Reference
Jafarlou S et al. Objective monitoring of loneliness levels using smart devices: A multi-device approach for mental health applications. PLoS One. 2024;19(6):e0298949.
