Using the Hilbert Transform to Evaluate the Effects of Functional Foods on Autonomic Nervous System Activity: A Comparison with the Fast Fourier Transform.

Abstract

Evaluating the autonomic nervous system (ANS) via heart rate variability (HRV) to investigate the effects of food on human health has attracted attention. However, using a conventional HRV analysis via the fast Fourier transform (FFT), it is difficult to remove artifacts such as body movements and/or abnormal physiological responses (unexpected events) from the HRV analysis results. In this study, an analysis combining bandpass filters and the Hilbert transform was applied to HRV data on functional food intake to compare with FFT analysis. HRV data were obtained from six males by recording electrocardiograms on functional food, γ-aminobutyric acid, intake. HRV indices were calculated by both analysis. In the Hilbert analysis, all HRV indices were obtained for the same number of sampling points as the HRV data. The standard errors of all HRV indices tended to be smaller in the Hilbert analysis than in the FFT analysis. In conclusion, the Hilbert analysis was more suitable than FFT analysis for evaluating ANS via HRV on functional foods intake.