A framework for quantifying the coupling between brain connectivity and heartbeat dynamics: Insights into the disrupted network physiology in Parkinson's disease

Abstract

Parkinson's disease (PD) often shows disrupted brain connectivity and autonomic dysfunctions, progressing alongside with motor and cognitive decline. Recently, PD has been linked to a reduced sensitivity to cardiac inputs, that is, cardiac interoception. Altogether, those signs suggest that PD causes an altered brain–heart connection whose mechanisms remain unclear. Our study aimed to explore the large‐scale network disruptions and the neurophysiology of disrupted interoceptive mechanisms in PD. We focused on examining the alterations in brain–heart coupling in PD and their potential connection to motor symptoms. We developed a proof‐of‐concept method to quantify relationships between the co‐fluctuations of brain connectivity and cardiac sympathetic and parasympathetic activities. We quantified the brain–heart couplings from electroencephalogram and electrocardiogram recordings from PD patients on and off dopaminergic medication, as well as in healthy individuals at rest. Our results show that the couplings of fluctuating alpha and gamma connectivity with cardiac sympathetic dynamics are reduced in PD patients, as compared to healthy individuals. Furthermore, we show that PD patients under dopamine medication recover part of the brain–heart coupling, in proportion with the reduced motor symptoms. Our proposal offers a promising approach to unveil the physiopathology of PD and promoting the development of new evaluation methods for the early stages of the disease.

Abstract

We propose a new framework to measure brain–heart interactions. The methodological pipeline consists in: (a) Computation of time‐varying electroencephalogram (EEG) power at different frequency bands (α, β, γ) and (b) the estimation of time‐varying connectivity between two EEG channels. (c) Computation of the heart rate variability series from electrocardiogram and the estimation of cardiac sympathetic‐parasympathetic activity. (d) Brain connectivity‐cardiac coupling estimation by computing the Maximal Information Coefficient (MIC). The coupling quantification is achieved by assessing the similarities between two time series, regardless of the curvature of the signals. The MIC method evaluates similarities between distinct segments individually, using an adjusted grid as depicted in the figure. The overall measure combines the similarities observed throughout the entire time‐course. A network cluster permutation pipeline is applied to the connections coupled with cardiac activity. These connections are grouped based on their neighboring definition. Then, the identified clusters undergo a permutation test. This process ultimately defines the networks whose coupling with cardiac activity changes under the experimental conditions being tested.