Association Between Ultra-Short-Term Heart Rate Variability and Atrial Fibrillation in Heart Failure Population: A Retrospective Cohort Study

BackgroundPatients with heart failure are at high risk for atrial fibrillation (AF) due to impaired heart function and sympathetic excitation.Ultra-short-term heart rate variability (usHRV), a parameter that reflects autonomic nervous system activation, has been proposed.However, the association between usHRV and AF has not been investigated in the heart failure with reduced ejection fraction (HFrEF) population. ObjectiveThis study explored the association between usHRV and AF risk in the HFrEF population. MethodWe conducted a retrospective cohort study involving 2,122 participants from the Critical Care Medical Information Mark-IV (MIMIC-IV) database with 2 years' observation.We used Cox proportional hazards models, subgroup analysis, interaction effect evaluation and E-value to exclude confounding factors.Furthermore, the Kaplan-Meier curve was used to describe the survival probability in different usHRV quantiles. ResultsDuring an average follow-up of 1.67 years, 475 participants (22.4%) developed AF.A negative association between usHRV and the risk of AF was evident after adjusting for multiple variables.In the fully adjusted Cox proportional hazard models, for every 1 unit increase, the hazard ratio (HR) for incident AF was 0.84 for log(standard deviation of normal-to-normal RR intervals) (95% confidence intervals [CI] 0.66-1.07),1.05 for log(standard deviation of successive differences) (95% CI 0.84-1.32),1.04 for log(root mean square of successive differences) (95% CI 0.83-1.31),0.88 for log(low frequency [LF]) (95% CI 0.80-0.96),0.93 for log(high frequency [HF]) (95% CI 0.84-1.04),0.82 for log(LF/HF) (95% CI 0.70-0.95),0.79 for log(LF normalised units) (95% CI 0.65-0.97),2.48 for log(HF normalised units) (95% CI 1.42-4.35),and 0.90 for log(total power) (95% CI 0.80-0.98).The Kaplan-Meier curve shows that the lower the frequency-domain index of usHRV, the higher the risk of AF occurrence in the HFrEF population.Subsequent subgroup analysis and E-value implied the current associations tended to be more stable. ConclusionsOur study findings suggest that decreased usHRV frequency domain metrics levels may be linked to an increased risk of AF in the HFrEF population.