Studies on EEG correlates of non-ordinary states of consciousness and slow-paced breathing

Abstract

Studies on the non-ordinary states of consciousness (NSCs) induced by meditation, hypnosis, trance, and slow-paced breathing are gaining visibility due to their potential efficacy in treating various clinical conditions. Slow-paced breathing at six cycles per minute (cpm) has been labeled as coherent or resonant breathing since it has been suggested to induce synchronous resonance frequency in various physiological signals. These self-regulatory or guided processes are practiced primarily to reduce stress and manage emotions and mental health. However, the underlying physiological mechanisms causing the health benefits of these practices still need to be better understood. Electroencephalography (EEG), a non-invasive neurophysiological tool to investigate brain's electrical activity, is deployed to study the changes in brain dynamics during different NSCs and slow-paced breathing. All the analyses are carried out at the sensor level. Different methods are explored to study the neural correlates during the practice of different NSCs and slow-paced breathing. In the first part of the work, the results of our studies on the bivariate functional connectivity (FC) using pairwise phase consistency (PPC) as the metric are presented. We have investigated the phase coupling within the anterior, posterior, left, and right hemisphere clusters to study the short, long-range, and across-hemisphere interactions during eyes-open Rajyoga meditation (RM) and music-listening session by the controls. Twenty-seven long-term Rajyoga meditators and thirty controls are recruited for the study. Both similar and distinct patterns are observed in distinct frequency bands in meditators and control groups. Node degree strength (NDS), a graph-based measure, is used to study the contributions of different cortical regions for integrating connectivity (IC). NDS is observed to be consistently higher in meditators than controls in higher theta and alpha bands both within and across hemispheres. Controls with no knowledge of meditation show no change in theta band during the music-listening session. Higher IC is observed in frontal electrodes in meditators during meditation implying increased self-awareness. The occipital electrodes are observed to have lower PPC in meditators than controls which may be the trait effect due to long-term practice of meditation with eyes open. Overall, meditators show increased functional integration during meditation as compared to controls during music-listening session supporting the hypothesis of cortical-integration theory. In the second part of the work, we have investigated the functional interdependencies during RM, hypnosis and self-induced cognitive trance (SICT) using two higher-order measures. This is a multicentric study involving datasets recorded at different laboratories. Synergistic and redundant information measures are used to compare and contrast the higher-order interactions among the different scalp EEG channels during the three NSCs. 22 long-term meditators, nine volunteers undergoing hypnosis, and 21 practitioners of SICT are considered for the study. Common and distinct patterns are observed in synergy and redundancy among the NSCs. Synergistic interdependencies in delta and theta bands increase during RM. On the other hand, synergy decreases in delta and beta2 bands in selected regions during hypnosis and in the delta, theta, and alpha bands during SICT. Redundant interdependencies in delta, beta1, and beta2 bands decrease during RM, whereas no significant changes occur in redundancy during hypnosis or SICT. The changes in the patterns of synergy and redundancy are related to the commonalities and differences in the phenomenology of NSCs including changes in environmental awareness, sense of self, selective attention, and sensory perception. In the third part of the work, we have studied the entrainment of brain oscillations by respiration and heart rate variability (HRV) during slow-paced breathing at different breathing rates (BRs), using coherence as a metric. EEG, electrocardiogram, and breathing data have been analyzed from 63 young, healthy adults while they synchronize their breathing to a visual metronome at ten, six, and four cpm. The analyzed segments of protocol involve initial baseline (IBL), symmetric breathing (SB), and symmetric breathing with breath-hold (SBH) after inspiration and expiration. Coherence between respiration and EEG significantly increases during slow-paced respiration at all BRs, with no common localization across subjects. The coherence is the highest during slow-paced breathing at six cpm during both SB and SBH conditions. The synchronization is observed at the whole brain level during SB and localized to the left frontal and central regions during SBH. Phase-amplitude coupling between respiration and EEG shows distinct patterns during normal and slow-paced breathing at six cpm in specific EEG frequency bands. Higher coupling is observed in frontal, centroparietal, and occipital cortices at the peak of inhalation in gamma bands. The modulation index increases during slow-paced breathing compared to normal respiration, supporting the link between respiration and brain activity and providing possible insight into the benefits of therapeutic breathing exercises like pranayama. In the last part of the work, we have studied changes in entropy and distance between covariance matrices of successive epochs of EEG during RM. Minimum variance modified fuzzy entropy is used to quantify the changes in the complexity of EEG during RM and baseline conditions. 14 long-term Rajyoga meditators are considered for the study. Increased entropy is observed during RM compared to the baseline condition in all lobes and maximum in the frontal lobe. Study of individual frequency bands in the frontal lobe electrodes has shown that the entropy changes are higher in the gamma band implying the role of cognitive processes such as attention and memory retrieval during RM. Another study using distance between covariance matrices of one-second successive epochs has revealed reduced variance in the Frobenius norms of the difference matrices in the frontal region. Another study is carried out on EEG-based biometrics using meditation datasets with FC measures such as phase lag index, phase lag value and correlation. It is observed that in spite of the long-term practice of meditation, certain EEG signatures are specific to individuals. This study has used multi-class support vector machines for identifying the subjects.