Does it make sense to study heart rate variability in episodes of paroxysmal atrial fibrillation during the 24-hour electrocardiographic monitoring?

Abstract

Aim. To study the characteristics of heart rate variability (HRV) in episodes of paro xysmal atrial fibrillation (AF) and sinus rhythm (SR) during the 24-hour Holter electro cardiographic (ECG) monitoring and to assess the presence and nature of their relationships.Material and methods. We analyzed 37 records of 24-hour Holter ECG monitoring obtained from patients with paroxysmal AF. For analysis, records were selected from the Long Term Atrial Fibrillation database (LTAFDB) (n=26), as well as from the long-term storage database “Myocard-holter” of the Sverdlovsk Regional Clinical Hospital № 1 (n=11). Each record contained at least 120 min of SR and at least 120 min of paroxysmal AF episodes. In addition, 48 HRV indices were calculated independently for SR and for AF episodes, after which the 96 HRV parameters obtained for each record were subjected to exploratory data analysis.Results. Analysis of HRV in AF paroxysm showed a significant increase in the RMSSD, SD1, pNN50 indices, which confirmed the great importance of parasympathetic influence in the regulation of AF. Despite the fact that single HRV parameters in AF are not directly related to any single parameters in SR, we have shown that the AF is not completely chaotic and independent of the characteristics of baseline HR variability. We found that for some of the HRV indices determined during AF paroxysm (PAF_CVI, PAF_pNN20, PAF_pNN50, PAF_ApEn, PAF_SDNN, PAF_SD2), up to 74% of the variance of these indices can be explained using multivariate linear models, including 4 HRV indices for SR and taken as predictors. It was shown that among all the indices analyzed, only the PAF_HTI, calculated in paroxysmal AF, had a moderate negative correlation with the duration of AF episodes (r=0,60, p<0,01).Conclusion. It is shown that the study of HRV indices on AF paroxysm has a wide clinical and electrophysiological potential. The paper proposes statistical models that demonstrate the relationship between HRV in SR and in the rhythm of AF episode. The PAF_HTI index, assessed on episodes of AF, showed a relationship with the duration of AF episodes, which suggests the expediency of using HRV characteristics on AF to assess the possibility of SR recovery.

Keywords

• heart rate variability
• paroxysmal atrial fibrillation
• 24-hour holter ecg monitoring
• exploratory data analysis