BACKGROUND
Residency is a critical period in a physician’s training, characterized by significant physical, cognitive, and emotional demands that make residents highly susceptible to stress and associated negative health outcomes. While physiological signals like heart rate have been explored as potential biomarkers of stress, their predictive utility in high-stress environments such as the intensive care unit (ICU) remains inconclusive, especially when factoring in atypical events that can further exacerbate resident stress levels.
OBJECTIVE
This study aimed to investigate the relationship between daily average heart rate (AHR) and perceived stress among ICU residents and to examine the moderating effect of atypical events on this relationship.
METHODS
The TILES-2019 dataset collected longitudinal data from 44 ICU residents who provided daily self-reported stress ratings and wore a Fitbit device to track physiological data over a 3-week period. The main predictor variables were AHR and the occurrence of atypical events (both work/life related and daily hassles). The primary outcome was the level of perceived stress, measured on a 7-point Likert scale. Linear mixed models (LMM) were used to analyze the relationship between AHR and stress, accounting for within-subject and between-subject variance. Interaction effects between AHR and atypical events were also examined.
RESULTS
The analysis revealed a significant positive association between AHR and perceived stress (β = 0.032, P = .04). However, this relationship was attenuated by the presence of atypical events (β = -0.076, P = .20). We further analyzed whether the severity of atypical events had an additional moderating effect but found no statistical significance.
CONCLUSIONS
AHR is a potential physiological marker for perceived stress in ICU residents but its effect is moderated by atypical events. Future research should replicate these findings in more diverse cohorts, assess their generalizability to broader populations, and control for additional confounding variables. Incorporating atypical events in stress assessment could lead to more accurate and context-sensitive interpretations of physiological data.