Relationship Between Chronic Stress and Heart Rate Over Time Modulated by Gender in a Cohort of Office Workers: Cross-Sectional Study Using Wearable Technologies

Abstract

Background: Chronic stress is increasing in prevalence and is associated with several physical and mental disorders. Although it is proven that acute stress changes physiology, much less is known about the relationship between physiology and long-term stress. Continuous measurement of vital signs in daily life and chronic stress detection algorithms could serve this purpose. For this, it is paramount to model the effects of chronic stress on human physiology and include other cofounders, such as demographics, enabling the enrichment of a population-wide approach with individual variations.

Objective: The main objectives of this study were to investigate the effect of chronic stress on heart rate (HR) over time while correcting for weekdays versus weekends and to test a possible modulation effect by gender and age in a healthy cohort.

Methods: Throughout 2016 and 2017, healthy employees of technology companies were asked to participate in a 5-day observation stress study. They were required to wear two wearables, of which one included an electrocardiogram sensor. The derived HR was averaged per hour and served as an output for a mixed design model including a trigonometric fit over time with four harmonics (periods of 24, 12, 8, and 6 hours), gender, age, whether it was a workday or weekend day, and a chronic stress score derived from the Perceived Stress Scale (PSS) as predictors.

Results: The study included 328 subjects, of which 142 were female and 186 were male participants, with a mean age of 38.9 (SD 10.2) years and a mean PSS score of 13.7 (SD 6.0). As main effects, gender (χ²₁=24.02, P<.001); the hour of the day (χ²₁=73.22, P<.001); the circadian harmonic (χ²₂=284.4, P<.001); and the harmonic over 12 hours (χ²₂=242.1, P<.001), over 8 hours (χ²₂=23.78, P<.001), and over 6 hours (χ²₂=82.96, P<.001) had a significant effect on HR. Two three-way interaction effects were found. The interaction of age, whether it was a workday or weekend day, and the circadian harmonic over time were significantly correlated with HR (χ²₂=7.13, P=.03), as well as the interaction of gender, PSS score, and the circadian harmonic over time (χ²₂=7.59, P=.02).

Conclusions: The results show a relationship between HR and the three-way interaction of chronic stress, gender, and the circadian harmonic. The modulation by gender might be related to evolution-based energy utilization strategies, as suggested in related literature studies. More research, including daily cortisol assessment, longer recordings, and a wider population, should be performed to confirm this interpretation. This would enable the development of more complete and personalized models of chronic stress.

Keywords: age; chronic stress; circadian rhythm; gender; heart rate; wearable device.

Figure 1

Diagram showing all steps from data upload to model development. ACC: accelerometer; AVGM: average mean acceleration of the x, y, and z coordinates; AVGSD: average standard deviation in acceleration of the x, y, and z coordinates; ECG: electrocardiogram; HR: heart rate; MXYZ: magnitude of acceleration in the x, y, and z directions. See Multimedia Appendix 1 for calculation of the features.

Figure 2

The heart rate (HR) over the time of the day, split for gender and workdays and weekend days. Since the data are summarized per hour, data points are provided in the middle of the hours. The error bars represent the 95% confidence intervals.

Figure 3

The heart rate (HR) over age per hour of the day, split for workdays and weekend days. Since the data are summarized per hour, the labels provided are the start of the hour.

Figure 4

The heart rate (HR) over the Perceived Stress Scale (PSS) score for every hour of the day, split for female and male participants. The time above each subsection provides the starting time of the hour for which the HR is shown.