Development of integrated heart rate variability indexes using bootstrapping factor analysis

Heart rate variability (HRV) has a great potential as a crucial indicator of psychophysiological conditions of participants in behavioral, physiological, and neuroscience research. Despite its potential utility, HRV has not been widely used due to the availability of numerous and seemingly interchangeable HRV measures. The present study used a bootstrapping Factor Analysis approach to analyze the HRV data from 5001 healthy adults. We found that a majority of variance in the different HRV measures could be explained by six orthogonal factors. They include: 1) RR interval statistics, 2) relative spectral powers, 3) absolute spectral powers, 4) large RR interval differences, 5) pulse fractal structure, and 6) heart rate. Our results suggest that these six underlying dimensions are sufficient to characterize an individual’s HRV. Further, the six dimensional indexes can be synthesized into one single integrated index of HRV. We further developed a web application to showcase computation of the six underlying dimensions and integrated HRV, each as indicators of stress. Potential contributions to the growing effort of HRV as a physiological indicator of stress within research and clinical understandings are discussed.