Description of heart rate parameters changes during sleep stages by means of regression equations

Abstract

The functional state of the autonomic nervous system changes during the transition of night sleep stages. These changes are reflected in the results of heart rate (HR) autonomic control analysed by the method of power spectrum. The results of our study demonstrate similar rules for HR parameters changes during transition of sleep stages and cycles for both healthy subjects (HSs) and ischemic heart disease patient (IHD pts) groups. However, HSs and IHD pts differ in HR parameters at an initial level and the level of the amplitude of their changes during shifts in sleep stages, depending on the functional status of the person investi-gated. These rules for changes have been described by multiple linear regression equations, based on a relationship between HR parameters during individual sleep stages and HR parameters at the baseline level just before sleep or HR parameters evaluated during all night sleep. A coefficient of determination was evaluated for every regression equation. The estimates of HR parame-ters for any person investigated have been evaluated in individual sleep stages by means of dependent regression variables; the accuracy of the model HR parameters has been evaluated. The examination has shown the following results: the average relative error of the model was 1-3% - for the average of RR interval, 13-18% - for the square deviation of RR intervals, 10-24% - for the absolute parameters of HR power spectrum and 12-37% - for the relative parameters of HR power spectrum. This study demonstrates that the regression equations are accurate for modeling changes in HR parameters over transition of sleep stages if based on all night sleep HR values, although they are less accurate if the model is based on HR parameters just before sleep. Both situations, all night sleep and wakefulness before sleep, can be used for modeling of the rules the HR changes during indi-vidual sleep stages in practice.