Decreased directed functional connectivity in the psychedelic state

This MEG study (n=63) found that psilocybin, ketamine, and LSD all decreased directed functional connectivity in the brain. LSD was additionally found to increase undirected functional connectivity.

Abstract

“Neuroimaging studies of the psychedelic state offer a unique window onto the neural basis of conscious perception and selfhood. Despite well understood pharmacological mechanisms of action, the large-scale changes in neural dynamics induced by psychedelic compounds remain poorly understood. Using source-localised, steady-state MEG recordings, we describe changes in functional connectivity following the controlled administration of LSD, psilocybin and low-dose ketamine, as well as, for comparison, the (non-psychedelic) anticonvulsant drug tiagabine. We compare both undirected and directed measures of functional connectivity between placebo and drug conditions. We observe a general decrease in directed functional connectivity for all three psychedelics, as measured by Granger causality, throughout the brain. These data support the view that the psychedelic state involves a breakdown in patterns of functional organisation or information flow in the brain. In the case of LSD, the decrease in directed functional connectivity is coupled with an increase in undirected functional connectivity, which we measure using correlation and coherence. This surprising opposite movement of directed and undirected measures is of more general interest for functional connectivity analyses, which we interpret using analytical modelling. Overall, our results uncover the neural dynamics of information flow in the psychedelic state, and highlight the importance of comparing multiple measures of functional connectivity when analysing time-resolved neuroimaging data.”

Authors: Lionel Barnett, Suresh D. Muthukumaraswamy, Robin L. Carhart-Harris & Anil K. Seth

Summary

Abstract

Neuroimaging studies of the psychedelic state offer a unique window onto the neural basis of conscious perception and selfhood. These studies demonstrate that the brain’s functional connectivity changes during the psychedelic state, with a decrease in directed functional connectivity and an increase in undirected functional connectivity.

Introduction

Psychedelic drugs impact profoundly on conscious experiences of world and self. Their effects are mediated by partial agonist actions at the serotonin 2A receptor (5HT2AR), which trigger different second-messenger pathways and enhance release of the excitatory neurotransmitter glutamate.

Ketamine and the classical psychedelics cause distinct associated subjective phenomena. Many studies now link the subjective effects of psychedelics to altered activity at the relevant receptors, but the large-scale changes in neural dynamics that underlie the dramatic subjective effects remain poorly understood.

Previous FC analyses of the psychedelic state have revealed decreased connectivity within resting-state networks and increased connectivity across such networks, such as the default-mode network.

Despite the insights provided by these studies, fMRI data has inherent limitations. More fine-grained FC analysis is possible using high-time resolution electrophysiological data, such as can be obtained using EEG or MEG.

Several studies have applied fine-grained FC analyses to neuroimaging data obtained in the investigations of the psychedelic state, including increased phase synchronisation under psilocybin, increased transfer entropy in thalamocortical networks under low-dose ketamine, and reduced anterior-to-posterior directed connectivity under sub-anesthetic ketamine.

We analysed source-localised MEG data from healthy volunteers following administration of LSD, psilocybin, or ketamine, and compared these to data following administration of tiagabine, a non-psychedelic GABA reuptake inhibitor.

We measured FC between and within cortical regions using MEG in healthy volunteers in a non-task resting state, and compared FC elicited by psychedelics to placebo and a non-psychedelic drug control.

We used placebo as a control for each drug, and tiagabine as a comparison between psychoactive drugs and non-psychoactive drugs.

We find decreased directed FC between brain regions, but increased undirected FC for all psychedelic compounds, consistent with increasing disorder and functional disorganisation underlying the psychedelic experience.

2. Materials and methods

2.1. Participants and drug delivery

In this study, resting-state pharmaco-MEG recordings of LSD, psilocybin, ketamine and tiagabine were made. Participants were excluded if they were younger than 21, pregnant, had a history of psychiatric disorder, substance dependence, cardiovascular disease, claustrophobia, blood or needle phobia, or had a medically significant condition. LSD, PSI and KET were administered intravenously at fixed single doses over the course of less than 1 min, whereas TGB was administered orally at 15 mg/kg over 3 h. Placebos were saline solution for intravenous drugs (PSI, LSD, KET) and identical inactive capsules for TGB.

2.2. Experimental procedure, data acquisition and preprocessing

All MEG data were collected using the CTF 275-channel axial gradiometer system at the Cardiff University Brain Research Imaging Centre from 2011 to 2014. The data were filtered at 1Hz to suppress slow transients, low-pass filtered at 150Hz to suppress line noise, and downsampled to 300Hz. The MEG data was source-localised to 90 cortical regions using the centroid locations of the standard Automated Anatomical Labelling brain atlas.

We include all 90 AAL regions in our analysis, as not all AAL regions are likely to contribute significantly to functional connectivity changes associated with the various drugs.

Epochs are analysed as stationary multi-trial data in source space, with the assumption that they are realisations of the same underlying stochastic process. This yielded stable autoregressive models, indicating consistent statistical properties across epochs.

2.3. Functional connectivity analysis

In multivariate situations, measures of FC may be confounded by in-direct associations. We used conditional FC measures, which control for indirect associations, and frequency-domain statistics, which integrate across a given frequency band to yield measures of GC in specific frequency ranges.

3.2. Functional connectivity

We next examine changes in per-source undirected (MI) and directed (GC) time-domain FC for drug vs placebo. Results are less clear for undirected measures, with LSD having the strongest effect.

The increase in MI in LSD is strongest in the -band, while psilocybin and ketamine show statistically weak increases. Tiagabine shows a distinct mechanism of action underlying increases in MI as compared with the psychedelics.

We next examined the regional specificity of changes in directed and undirected FC. We found that the decreases in directed connectivity for LSD, psilocybin, ketamine and tiagabine are cortex-wide, both within and between ROIs, with significant intra-ROI decreases in occipital, parietal and sensorimotor regions.

Comparing psychedelics with placebo reveals significant decreases in GC, widespread across cortical regions and across frequency bands, but increases in MI.

We investigated whether changes in power spectra and residuals correlation could account for the observed changes in MI and GC.

3.3. Relation between changes in undirected and directed FC

Our analysis shows that GC and MI relate to each other in a variety of ways, and that GC and MI are reflecting distinct aspects of neural dynamics.

3.3.1. Spectral power and signal-to-noise ratio

Existing functional analyses of the psychedelic state have focused on changes in spectral power, but this explanation is not straightforward because Granger causality is scale invariant, and depends on the full cross-power spectrum.

LSD and tiagabine cause the opposite movement of GC and MI, which is surprising because one might think that increased MI should imply increased mutual information.

Fig. 10a and 10b display heat maps of MI and GC for 0 and c varying within the model parameter space.

Changes in residuals correlation cannot readily explain the pattern of empirical results across all drugs, because for at least half of the entire parameter space, increasing a residuals correlation always leads to increasing MI coupled with decreasing GC, while for the other half, increasing a residuals correlation always leads to decreasing MI coupled with increasing GC.

For LSD, psilocybin and tiagabine, there is a statistically significant increase in residuals correlation in the drug condition, while for KET there is no statistically significant increase in residuals correlation. However, it is unlikely that changes in residuals correlation can explain the relationship between directed and undirected FC.

4. Discussion

In this exploratory study, we examined the effects of three psychedelic drugs on large-scale brain dynamics in terms of directed functional connectivity.

We applied both directed FC measures and undirected FC measures to source-localised MEG data obtained in resting state conditions, contrasting placebo against three different psychedelics as well as against a non-psychedelic control, tiagabine. We observed a decrease in information flow in psychedelic conditions, broadly across cortex.

4.1. Power spectra

We first identified a substantial reduction in broadband spectral power in the psychedelic state, which is consistent with previous studies. We also note a small but significant upward shift in the alpha peak frequency, which is most evident for LSD.

4.2. Functional connectivity and signal variability

By charting patterns of connectivity across brain regions, we can see how neural dynamics are altered in psychedelic states, including a disintegration of communication between and within brain regions.

Our results are most directly comparable to other studies of the psychedelic state using EEG or MEG. However, unlike this previous study, we analysed connectivity of any kind.

fMRI studies of altered FC and signal variability in the psychedelic state have revealed a wide range of effects across a range of different psychedelics. These studies suggest an increased dynamical repertoire in the psychedelic state, which is consistent with our finding of decreased FC in these states. fMRI studies have revealed altered patterns in the psychedelic state, including decreased coupling between “connectivity hub” brain regions and increased FC in high-level association cortices.

Some studies have shown increases in connectivity at sub-anesthetic doses of ketamine, whereas others show decreases. These results are difficult to directly compare with the present results, given the slow temporal dynamics of the fMRI BOLD signal.

4.3. Directed and undirected functional connectivity

The present analysis compares directed and undirected FC on the same data using high-temporal resolution, accurate source-localisation, and steady-state MEG recordings.

Our study shows that directed FC is decreased across all psychedelics (but not tiagabine) and that undirected FC is increased in the psychedelic state. This finding has broader relevance for future studies of functional connectivity.

FC analyses are distinct from EC analyses in that they describe statistical dependencies between variables, whereas EC analyses aim to identify the minimal causal circuit underlying some observed activity pattern.

4.4. Limitations

Our analysis has several limitations, including that muscle artefacts may have affected the MEG signal differently in drug as compared with placebo conditions, and that the MEG recordings represent only a brief snapshot of an 8-hour psychedelic experience.

We note that tiagabine recordings were made with eyes closed, whereas the other recordings were made with eyes open. However, this difference is unlikely to have affected our conclusions.

5. Summary

We measured directed and undirected functional connectivity in human brains while they experienced a psychedelic state, and found that the psychedelic state was associated with a general decrease in directed functional connectivity.

Functional connectivity analyses should include both directed and undirected measures, and future research should probe how these measures behave in other contexts besides psychedelics.

Appendix A.2 Directed measures: Granger causality in time and frequency domain

Directed connectivity measures are based on Granger causality, a 2nd-order statistic for stochastic time-series based on optimal linear prediction.

For multivariate processes, a global-conditional measure (GLO) may be defined as the sum of conditional causalities from X itself to each individual variable Xi.

Appendix A.3 Frequency-band averaging

Appendix B. Correlation and Granger causality (GC) for a minimal VAR model is presented. It consists of the simplest bivariate system, the VAR coefficients matrix, and the residuals covariance matrix. The CPSD of a process can be calculated from the spectral factorisation of the process, where the power spectral density of the process is factored as Sxx z1 4 b czj2 1 2 c2 1 azj2 1 bzj2 on z 1 4 1.

Study details

Compounds studied
Psilocybin LSD Ketamine

Topics studied
Neuroscience

Study characteristics
Bio/Neuro

Participants
63

Authors

Authors associated with this publication with profiles on Blossom

Robin Carhart-Harris
Dr. Robin Carhart-Harris is the Founding Director of the Neuroscape Psychedelics Division at UCSF. Previously he led the Psychedelic group at Imperial College London.

PDF of Decreased directed functional connectivity in the psychedelic state