Randomness and changes of heart rate and respiratory frequency during high altitude mountain ascent without acclimatization

The aim of this study was to detect and compare the changes in the time–frequency and fractal scaling behaviors of heart rate (HR) and respiratory frequency (Rf), recorded simultaneously during a high altitude mountain ascent. The time–frequency analysis was performed using the local cosine4 orthonormal bases of Coifman, Malvar and Meyer, whose spectrum is not redundant as those computed with the short Fourier transform. The fractal scaling behavior was obtained using the detrended fluctuation (DFA) and the wavelet leaders (WL) analysis. Results showed that the high altitude mountain ascent differently affected HR and Rf variability. Rf average values increased (p=0.0003) while HR average values did not change. The scaling variability of HR was altered during the mountain ascent, which was detected by the increasing HR short range DFA exponents with altitude (p<0.03). Rf scaling variability remained unchanged. These differences between HR and Rf alterations were also observed for the local cosine4 power law behavior since power law exponents, in absolute values, increased for HR (p<0.003) while those of Rf did not change. Furthermore, the ratio of low over the whole local spectrum energy of Rf decreased with altitude (p=0.04) in contrast to HR. In most of these HR and Rf analyses, one of the two time series was significantly modified but not both. Moreover, the Rf local cosine4 spectrum had higher entropy compared to HR (p<0.01), the difference between the Rf and the HR entropy increased (p=0.04) during the mountain ascent. In consequence, Rf had more randomness than HR and altitude increased this difference.