Influence of time-varying mean heart rate in coronary artery disease diagnostic performance of heart rate variability indices from exercise stress testing

In this study, the influence of the time-varying mean heart rate (HR) and respiratory frequency in the ability of HR variability (HRV) indices to diagnose coronary artery disease has been studied. The autonomic nervous system activity has been assessed using a methodology that comprises correction of the HRV signal by the time-varying mean HR and redefinition of the classical high-frequency band to include respiratory frequencies above 0.4 Hz. The obtained clinical indices discriminate patients with coronary artery disease from patients with Framingham risk index lower than 5% with a moderate accuracy of 76%, which is lower than the reported in literature for HRV indices. We claim that time-varying mean HR and respiratory frequency, if not taken into account, introduce apparent improvement of diagnostic performance of HRV indices, adding information nonrelated to the autonomic nervous system activity, which is not what HRV is supposed to measure.