The heart rate (HR) is closely coupled with cardiac performance [1]. In response to changes in the oxygenated blood requirements of the body due to exercise, the HR changes with the alteration of cardiac output. The HR is also closely related to myocardial oxygen consumption [2]. Impaired response of heart rate to exercise has recently been demonstrated to be predictive of increased mortality and coronary heart disease incidence [3,4]. Quantitative analyses of the HR response are thus important with respect to cardiac accidents.In high-intensity bicycle ergometer-exercise (120 W), the HR at first increased rapidly followed by a continuous and gradual increase, and this response was fitted to a second-order exponential function [5][6][7][8][9]. In moderate-intensity exercise (50-75 W), the HR increased and reached a plateau level, and this response was fitted to a first-order exponential function [5][6][7][8][9]. In unloaded or low-intensity exercise, the HR increased transiently and then declined, and an adequate equation fitting this response has not yet been obtained [10]. In these previous analyses, it was difficult to determine, in certain cases, whether the first-or second-order exponential equation was appropriate for fitting. It is also reasonable to expect that the central nervous command system regulates smoothly (not suddenly with a switching point) the change of balance between parasympathetic and sympathetic tone. It would be more theoretically reasonable for one single equation to fit all these different HR changes, and such an equation would be very helpful for analyzing the HR changes induced by constant-load exercise Key words: kinetics, regression, least-squares method, electrocardiography.
Abstract:We attempted to fit heart rate (HR) changes induced by constant exercise loads of different intensities to an exponential hyperbolic sine curve by the least-squares method, and we compared the results with the fitting of the changes to exponential curves. Seven healthy male volunteers performed three different intensities of constant-load exercise on a bicycle ergometer. The exponential hyperbolic sine function adequately fitted the HR responses induced by all three different intensities of loads: low (30 W: correlation coefficient, rϭ0.68Ϯ0.13, meanϮSD), moderate (75 W: rϭ0.93Ϯ0.07) and high (125 W: rϭ0.97Ϯ0.02). The first-order exponential curve fitted only the moderate load response. Although the second-order exponential equation fitted the HR response for both the moderate and high loads, the equation did not fit the low-load response (rϭ0.43Ϯ0.26). In lowload exercise, the sum of the power of the residuals for the exponential hyperbolic sine curve fitting was significantly smaller than that for the first-or second-order exponential curve fitting. In conclusion, the exponential hyperbolic sine function is useful for quantitative analyses of the HR response to exercise loads of various intensities.