Heart Rate Asymmetry (HRA) is a phenomenon of the imbalance between the accelerations and decelerations in human Heart Rate (HR). The majority of young healthy population exhibit this type of behavior in their HR, however, the scale of asymmetry varies depending on the assessment method. The standard HRA methods are based on the Poincaré plots analysis, i.e. interpretation of the distribution of points referring to accelerations and decelerations in the HR. The essential step before the analysis is data filtering.
We have examined 413 heartbeat interval series (5 min) of young, healthy people (232 females; age: 22.1 ± 1.8; BMI: 22.4 ± 2.9) to evaluate the most commonly used HRA measures: Porta Index (PI), Guzik Index (GI), Ehlers Index (EI) as well as those less popular: Slope Index (SI) and Deceleration Input (DI). We used three methods of automatized detection of possibly incorrect heartbeat intervals: square (300 ms < normal < 2000 ms), quotient (incorrect if change between pairs of consecutive intervals is greater than 20%) and indicated by an algorithm of the intervals extraction. We compared the results with the values obtained from the shuffled interval series.
In three (GI, EI and SI) out of five HRA measures the differences between various filtering methods were greater than the differences between results obtained for shuffled heartbeat series and filtered by those methods. Moreover, in 10 series (2.4%) removing a single point (out of ~340) on Poincaré plot determined the occurrence of asymmetry by GI.
When it is impossible to manually indicate incorrect HR intervals it is necessary to provide detailed information about the automated method used. Otherwise, the correct comparison of the results between different studies is not feasible. The issue of differences resulting from filtration made by hand by various specialists remains open.