Information-Theoretic Analysis of EEG Wave Amplitude and Heart Rate Variability Reveals the Time Scale-Dependent Nature of Brain–Heart Interactions

Goal: Brain-heart interactions have been linked to physiological and pathological states and are typically studied through the use of electroencephalographic (EEG) signal and heart rate variability (HRV) time series. However, there are still major challenges to overcome, particularly in establishing a robust methodology to assess these complex multi-scale interactions and to extract meaningful information. To this end, we explore the time scale-dependent nature of brain-heart interactions by exploiting information-theoretic measures. Methods: We analyze cardiac vagal activity and EEG brain wave amplitudes at two time scales—heart rhythm ($\sim$1 s) and longer ($\sim$1 min)—in two groups of healthy subjects monitored during wakefulness and sleep, respectively. Different entropy-based measures are then employed to evaluate the regularity of each system's dynamics, as well as their static and dynamic coupling. Results: Different time-scales are involved in different physiological coupling mechanisms. While overall coupling strength values are low, longer time-scales show a stronger presence of coupling in terms of statistically validated brain-heart connections compared to shorter time-scales. Conclusions: This study shows that the presence and the strength of brain-heart interactions are highly dependent on the time-scale, which in turn is affected by the underlying physiological processes.