Wavelet transform shows age-related changes of heart rate variability within independent frequency components

Reduction in overall heart rate variability (HRV) associated with aging is determined by a decreased amplitude of heart rate oscillations at all frequency levels, including high frequency (HF) oscillations attributed to respiratory sinus arrhythmia, low frequency (LF) oscillations attributed to Meyer waves and very low frequency (VLF) oscillations of an uncertain origin, presumably linked among others to thermoregulation. However, no studies were conducted to determine whether heart rate oscillations at independent frequency levels show themselves reduced HRV. Wavelet transform was applied to filter specific frequency components of HRV in a sample of younger (21-34 years old) and older (68-85 years old) healthy subjects. HRV indexes were measured within HF, LF and VLF components. The standard deviation of all RR intervals (SDNN) and the square root of the mean squared differences of successive RR intervals (RMSSD) were used as conventional linear time-domain measures. Sample entropy (SampEn) was used as a measure of nonlinear variability. Aged subjects showed lower SDNN at VLF (p < 0.001), LF (p = 0.007) and HF (p < 0.001). Lower RMSSD was observed in older people at VLF (p < 0.001), LF (p = 0.005) and HF (p < 0.001). SampEn was reduced by aging only at VLF level (p < 0.001). In aged people, linear variability was diminished within all frequency components, while nonlinear variability was lower only at VLF level. Preserved central, nonreflex autonomic modulation over the baroreflex control and the central cardiopulmonary coupling might explain this observation. Potential applications of this method include the study of heart rate regulation during sleep in which a complex interaction between the sympathetic and the parasympathetic activity takes place.