Wearable and app-based resilience modelling in employees: exploring the possibilities to model psychological resilience using wearable-measured heart rate variability and sleep

Stress has a major impact on both an individual and a societal level. Early recognition of the negative impact of stress or reduced resilience can be used in personalized interventions that enable the user to break the identified pattern through timely feedback, and thus limit the emergence of stress-related problems. The emergence of wearable sensor technology makes it possible to continuously monitor relevant behavioral and physical parameters such as sleep and heart rate variability (HRV). Sleep and HRV have been linked to stress and resilience in population studies, but knowledge on whether these relationships also apply within individuals, which is necessary for the aforementioned personalization, is lacking. This thesis introduces a cyclical conceptual model for resilience and four observational studies that test relationships between sleep, HRV and subjective resilience-related outcomes within participants using different types of data analysis at different timeframes. The relationships from the conceptual model and the related hypotheses are broadly confirmed in these studies. Participants tended to have more favorable subjective stress- and resilience-related outcomes on days with a relatively high resting HRV or long total sleep duration. Also, having a resting HRV that fluctuates relatively little from day to day was related to less stress and somatization. However, the strength of the relationships found was modest. The current findings can therefore not yet be directly implemented to initiate meaningful feedback, but they do provide starting points for future research and take a relevant step towards the possible future development of automated resilience interventions.