Heartbeat evokes a dynamically changing cortical theta-synchronized network in the resting state

Abstract

In the resting state, heartbeats evoke cortical responses called heartbeat-evoked responses (HERs). While previous studies reported regional level HERs, researchers have not determined how heartbeat is processed at the cortical network level. Using resting-state magnetoencephalography data from 87 human subjects of both genders provided by the Human Connectome Project, we first showed that heartbeat increases the phase synchronization between cortical regions in the theta frequency, which forms a network structure, and we called this network a heartbeat-evoked network (HEN). The HEN was not an artefactual increase in phase synchronization. The HEN was partitioned into three modules with connector hubs in each module. The first module contained major interoception-related regions and thus was called a visceromotor-interoceptive network (VIN) displaying the strongest synchronization among modules, suggesting a major role for the VIN in processing heartbeat information. Two modules contained regions involved in the default mode network (DMN). The HEN structure was not fixed, but dynamically changed. The most prominent change was observed at approximately 200 ms after R-peak of the electrocardiogram, which was quantified based on the flexibility of the network. Furthermore, the strongest synchronization within VIN was observed before heartbeat stimulated the cortex, which might be related to the prediction of an afferent heartbeat signal, thus supporting an interoceptive coding framework. Based on our results, the heartbeat is processed at the network level, and this result provides a useful framework that may potentially explain previous results of the regional level HER modulation through network-level processing.\n\nSignificance statementThe resting-state network is composed of several networks supporting different functions. However, although the heartbeat is processed in the cortical regions, even in the resting state, the network supporting this function is unknown. Thus, we identified and investigated the heartbeat-evoked network (HEN), a network composed of significantly increased theta-phase synchronization between cortical regions after a heartbeat. The HEN comprised three modules. In particularly, the visceromotor-interoceptive network was likely to play a major role in network-level heartbeat processing and displayed the strongest synchronization immediately before the heartbeat enters the CNS, which supports an interoceptive predictive coding framework. These results provide a novel framework that may improve our understanding of cortical heartbeat processing from a network perspective.