Associations of psychological self-assessments and HRV in long-term measurements at home

Psychological self-assessments and heart rate variability analysis are popular methods of estimating subject's mental load and his/her load on the autonomic nervous system, thus the level of mental stress. The aim of the study was to test a recent heart rate variability analysis method against self-assessments. Long-term measurements carried out in everyday life ensured a sufficient amount data (12 subjects, 35 variables and approximately 300 measurements per subject/variable). Significant (p<0.05) correlations between data from the two sources were found (?<0.359). Associations within self-assessment variables were studied with Principal Factor Analysis and the most powerful of the resulting variables (feature vectors) were retained. The distributions of these new variables in space were studied with a clustering algorithm which concluded in a division into two groups. Based on the characteristics of the values assigned to the groups, they were named as bad (40% of grouped points) and good (60%) days/evenings. The same division was applied to heart rate variability-related variables which showed significantly higher relative measurement time spent in stress on days marked as bad (mean 60%, std 19%) when compared to good (mean 52%, std 17%).