Changes in heart rate variability with respect to exercise intensity and time during treadmill running

BACKGROUND: Heart rate variability (HRV) arises from the complex interplay of sympathetic and parasympathetic autonomic regulation of heart rate. Ultra-low frequency (ULF) and very-low frequency (VLF) components of HRV play a crucial role in automatic HR controllers, but these frequency bands have hitherto largely been neglected in HRV studies. The aim of this work was to investigate changes in ULF and VLF heart rate variability with respect to exercise intensity and time during treadmill running. METHODS: RR intervals were determined by ECG in 21 healthy male participants at rest, and during moderate and vigorous-intensity treadmill running; each of these three tests had a duration of 45 min. Time dependence of HRV was investigated for moderate and vigorous running intensities by dividing the constant-speed stages into three consecutive windows of equal duration ([Formula: see text] 14 min), denoted [Formula: see text] , [Formula: see text] and [Formula: see text] . ULF and VLF power were computed using Lomb-Scargle power spectral density estimates. RESULTS: For both the ULF and VLF frequency bands, mean power was significantly different between the resting, moderate and vigorous intensity levels (overall [Formula: see text] ): mean power was lower for moderate vs. rest ([Formula: see text] ), for vigorous vs. rest ([Formula: see text] ), and for vigorous vs. moderate ([Formula: see text] ). For both ULF and VLF and moderate intensity, mean power was significantly different between the three time windows (overall [Formula: see text] for ULF, overall [Formula: see text] for VLF): for ULF, mean power was lower for [Formula: see text] vs. [Formula: see text] ([Formula: see text] ) and for [Formula: see text] vs. [Formula: see text] ([Formula: see text] ); for VLF, mean power was lower for [Formula: see text] vs. [Formula: see text] ([Formula: see text] ). For ULF and vigorous intensity, there was no significant difference in mean power between the three time windows (overall [Formula: see text] ). For VLF and vigorous intensity, mean power was significantly different between [Formula: see text] , [Formula: see text] and [Formula: see text] (overall [Formula: see text] ): mean power was lower for [Formula: see text] vs. [Formula: see text] ([Formula: see text] ) and for [Formula: see text] vs. [Formula: see text] ([Formula: see text] ). CONCLUSIONS: The degree of HRV in terms of ULF and VLF power was found to decrease with increasing intensity of exercise. HRV was also observed to decrease over time, but it remains to clarify whether these changes are due to time itself or to increases in HR related to cardiovascular drift. For feedback control applications, attention should be focused on meeting performance targets at low intensity and during the early stages of exercise.