Autonomic Dysregulation in Cardiac Arrhythmias Revealed by Reproducible Long-Term Heart Rate Variability Analysis

This study presents a retrospective computational analysis of heart rate variability (HRV) derived from long-term (24-hour) Holter electrocardiographic recordings obtained from publicly available PhysioNet databases. HRV provides a noninvasive measure of auto-nomic nervous system regulation and cardiovascular complexity, whose alterations are associated with arrhythmic conditions. Although 24-hour Holter monitoring is considered the clinical reference standard for HRV analysis, its large data volume poses significant computational challenges. The objective of this work was to develop and apply a fully reproducible MATLAB-based pipeline for automated HRV analysis and arrhythmic burden quantification. Recordings from approximately 85 subjects with documented rhythm disorders were analyzed and compared with reference recordings from healthy individuals. Signal preprocessing included digital filtering, QRS detection using the Pan–Tompkins algorithm, artifact correction, and NN interval interpolation. HRV metrics were computed in time and frequency domains, as well as through nonlinear methods capturing signal complexity. The results demonstrated a pronounced reduction in HRV indices, decreased spectral power, and increased arrhythmic events in pathological subjects, reflecting impaired autonomic regulation and elevated inter-subject variability. The proposed framework enables standardized, automated, and reproducible HRV analysis, supporting entropy-based characterization of cardiovascular dynamics and future risk stratification studies.