Prolonged Non-metabolic Heart Rate Variability Reduction as a Physiological Marker of Psychological Stress in Daily Life

Abstract

Background: Prolonged cardiac activity that exceeds metabolic needs can be detrimental for somatic health. Psychological stress could result in such "additional cardiac activity."

Purpose: In this study, we examined whether prolonged additional reductions in heart rate variability (AddHRVr) can be measured in daily life with an algorithm that filters out changes in HRV that are purely due to metabolic demand, as indexed by movement, using a brief calibration procedure. We tested whether these AddHRVr periods were related to worry, stress, and negative emotions.

Methods: Movement and the root of the mean square of successive differences (RMSSD) in heart rate were measured during a calibration phase and the subsequent 24 h in 32 participants. Worry, stress, explicit and implicit emotions were assessed hourly using smartphones. The Levels of Emotional Awareness Scale and resting HRV were used to account for individual differences. During calibration, person-specific relations between movement and RMSSD were determined. The 24-h data were used to detect prolonged periods (i.e., 7.5 min) of AddHRVr.

Results: AddHRVr periods were associated with worrying, with decreased explicit positive affect, and with increased tension, but not with the frequency of stressful events or implicit emotions. Only in people high in emotional awareness and high in resting HRV did changes in AddHRVr covary with changes in explicit emotions.

Conclusions: The algorithm can be used to capture prolonged reductions in HRV that are not due to metabolic needs. This enables the real-time assessment of episodes of potentially detrimental cardiac activity and its psychological determinants in daily life.

Keywords: Emotional awareness; Heart rate variability; Metabolic; Perseverative cognition; Physical activity; Worry.

Fig. 1

Calibration and daytime data for one participant. In the upper panel, the relation between acceleration and RMSSD is depicted. The solid black line represents the predicted inverse relation between acceleration and RMSSD, whereas the dashed red line depicts the derived threshold for detecting AddHRVr (i.e., predicted RMSSD levels − 2 × SE). In the lower panel, the data for the daytime assessments (per 30 s epoch) are shown (scatter; all data points depicted in open dots). The red line indicates the continuously calculated threshold, and the open dots below this line are epochs with AddHRVr decreases. For illustrative purposes, acceleration (multiplied by 50) is also included as the blue line. The dashed black lines indicate where hourly AddHRVr periods were detected (i.e., a 7.5 min of AddHRVr precedes this line)

Fig. 2

Scatterplot of the association between the total number of additional HRV reduction (AddHRVr) episodes and the total number of worry episodes

Fig. 3

Estimated marginal means (+/− SEM) of the affect ratings for hours without and with an episode of additional heart rate variability reduction (AddHRVr). PA = positive affect, NA = negative affect

Fig. 4

Estimated marginal means (+/− SEM) of the affect ratings for hours without and with an episode of additional heart rate variability (AddHRVr), split by participants low and high in emotional awareness (median split). PA = positive affect, NA = negative affect, LEAS = Levels of Emotional Awareness Scale