ISSN: 1223-1533

< Return to the list of articles


Authors: Jerzy Sacha, Szymon Barabach, Gabriela Statkiewicz-Barabach, Krysztof Sacha, Alexander Müller, Jaroslaw Piskorski, Petra Barthel, Georg Schmidt


Background: Previously, we have demonstrated that the dependence between heart rate variability (HRV) and heart rate (HR) can be mathematically modified. Using a cohort of 1455 post-infarction patients, we have shown that if HRV becomes more dependent on HR its prediction power increases for cardiac death and decreases for non-cardiac death. However, this has been observed only in men - in women the HRV prognostic ability decreases for all outcomes if HRV gets more dependent on HR.


The aim of the present study was to validate these findings.


Methods: A separate group of 946 post-infarction patients (followed up for 5 years) took part in this study. In the analyzed group there were 763 males. Seven classes of spectral HRV indices with increasing dependence on HR were calculated, i.e.: hrv1, hrv2, hrv3, hrv4, hrv5, hrv6, hrv7. Their prognostic power was tested by calculating the areas under receiver-operator characteristic curves (AUC).




Figure 71.1. Prognostic powers (AUC) of very low frequency component of HRV spectrum in men (A) and women (B) are shown.



Results: As HRV was getting more dependent on HR (i.e. from hrv1 to hrv7 class), its predictive power progressively increased for cardiac but decreased for non-cardiac death in men. This phenomenon almost exactly corresponded to that observed in our previous male cohort - Figure 1A (in the Figure, all AUC above dashed lines are significantly different from 0.5, those below are not). However, such validation was not possible in women, since only 2 of them died from cardiac and 4 from non-cardiac causes and consequently all AUCs were not significantly different from 0,5 - Figure 1B only shows the results of our previous study.


Conclusion: HR plays a consistent and pivotal role in the HRV prognostic power in men, i.e. its role is positive in predicting cardiac death but negative in non-cardiac death. The role of HR in the HRV predictive ability in women is less certain, however, our previous study suggests the HR impact to be negative for the prediction of both cardiac and non-cardiac death. To summarize, by calculating HRV with different dependence on HR, one may get specific predictors for cardiac and non-cardiac mortality in men after myocardial infarction.