A procedure to correct the effect of heart rate on heart rate variability indices: description and assessment

Abstract Objective: To develop a method to correct the nonlinear effect of the heart rate (HR) on different heart rate variability (HRV) indices of heart rate variability. Methods: The study included 265 healthy participants (17–69 years old), a group of 36 type 1 diabetes mellitus patients, including 15 patients with positive diagnosis of cardiovascular autonomic neuropathy (CAN), and a group of 24 CAN positive type-2 spinocerebellar ataxia patients. HR and HRV indices were calculated for 5-min resting ECG recordings. The proposed correction method (CM) included the joint application of multiple regression analysis and Z-transformations of HR and HRV indices. To assess the effect of the CM, correlation analysis, multivariate factor analysis, and the ANOVA test were applied to both groups before and after corrections. Results: The CM was able to remove the effect of HR on HRV indices, and at the same time, were preserved the expected differences between HR and HRV indices between controls and patients. Sample size was not a factor. Conclusion: Our method may be considered a novel approach, and may represent an alternative to the use of currently developed procedures. Significance: Studies of HRV without an appropriately HR correction should not be considered in the future.