Abstract
Emerging efforts toward prevention of stress-related mental disorders have created a need for unobtrusive real-life monitoring of stress-related symptoms. We used ecological momentary assessments (EMA) combined with wearable biosensors to investigate whether these can be used to detect periods of prolonged stress. During stressful high-stake exam (versus control) weeks, participants reported increased negative affect and decreased positive affect. Intriguingly, physiological arousal was decreased on average during the exam week. Time-resolved analyses revealed peaks in physiological arousal associated with both self-reported stress and self-reported positive affect, while the overall decrease in physiological arousal was mediated by lower positive affect during the stress period. We then used machine learning to show that a combination of EMA and physiology yields optimal classification of week types. Our findings highlight the potential of wearable biosensors in stress-related mental-health monitoring, but critically show that psychological context is essential for interpreting physiological arousal detected using these devices.
Teaser Smartwatches combined with daily diaries of mood can detect stress periods using individualized machine learning models.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
Only cosmetic changes were performed in this revision. Minor textual corrections were applied. Figures were reformatted, and Figure 5 has been changed for easier interpretation.
