Factors Associated with Wearable-Derived Physiological Stress Indicators Among Clinical Nurses: A Longitudinal Observational Study Using Continuous Monitoring.

Background: Work-related stress affects both nursing care quality and workforce wellbeing, yet conventional self-report assessment is prone to recall bias and is unable to capture real-time fluctuations. This study used a consumer-grade wearable device to continuously monitor physiological stress and identify associated factors among clinical nurses. Methods: In this longitudinal observational study, 52 clinical nurses at a tertiary medical center located in Taiwan wore a Garmin Vivosmart 4 over a period of 90 consecutive days. The wearable-derived stress score (0-100), computed from heart rate variability, was then averaged to a daily mean per nurse. A multivariable generalized estimating equation (GEE) model identified independent predictors of daily mean stress. Results: Older age was associated with lower physiological stress (β = -0.55 per year of age; p = 0.001). Relative to day shift, night-shift work was associated with substantially lower stress (β = -11.97; p < 0.001), whereas evening shift work did not differ significantly from day shift work. Two or more shift changes monthly were associated with lower stress than a consistent schedule (β = -14.74; p = 0.002). A higher daily step count was associated with higher stress (β = 1.06 per 1000 steps/day; p < 0.001). Education, position, and department showed no significant association. Conclusions: Younger age, day shift work, a consistent schedule, and a higher daily step count were associated with higher scores on the wearable-derived physiological stress indicator. Wearable-based monitoring may offer a feasible, objective complement to traditional self-report assessment. Our findings require confirmation against validated self-report measures before informing occupational-health decisions.