Heart rate variability in mental stress aloud

Abstract 

Previous investigations on arithmetic stress with verbalization showed that spectral measures of heart rate variability (HRV) did not assess changes in autonomic modulation, although the heart rate (HR) increased. In this study non-linear measures of HRV are determined and linear measures are re-examined in order to understand this apparent discrepancy between HR and HRV changes. In 23 healthy subjects 5-min electrocardiograms (ECGs) were recorded at rest and during arithmetic stress aloud. We determined non-linear (short-term scaling exponent, sensitivity to the initial condition and signal complexity) and linear (low-frequency and high-frequency spectral powers) measures. Our results showed that averaging concealed out an opposite effect of mental stress aloud on spectral measures and that this could be the main reason why the effect was not quantified. We found that increase of HR upon mental stress aloud could be achieved through the decreased as well as increased modulation in high-frequency band (HF). We also showed that non-linear measures distinguished this opposite effect of mental stress aloud on linear measures. Decreased HF power is associated with increase in short-term scaling exponent and decrease in signal complexity, while increased HF power increased sensitivity to the initial conditions. Apart from their opposite response to the mental stress, the two groups differed in baseline in sensitivity to the initial conditions. We suggest that variety of changes in HR dynamics upon different perturbation could be due to some differences in intrinsic properties of the system.

Keywords: Heart rate variability, Nonlinear time series analysis, DFA, Sample entropy, Largest Lyapunov exponent, Mental stress