Evaluation of consistency of HRV indices change among different emotions

Recent studies have indicated that there are significant changes in physiological signals among multiple emotions. This work aimed to study the differences of heart rate variability (HRV) indices among multiple emotions, and thus to explore the consistency of the differences within and without individuals. Sixty healthy students were involved in this study. Electrocardiography (ECG) from each subject was recorded among six different emotions: neutral, fear, sadness, happiness, anger and disgust. Variability series (RR interval time series) were abstracted from the ECG signals. Six HRV indices of the RR interval time series were calculated, including SDNN and RMSSD in time-domain, LF, HF, and LF/HF in frequency-domain, and SampEn in nonlinear domain. The results showed that there were significant differences of the above HRV indices among different emotions. No notable consistency of these differences was acquired, as no similar reference group orders were found, and the inconsistency measure (IM) of each single index was approximately normal distribution.