Effective connectivity of functionally anticorrelated networks under LSD

This double-blind placebo-controlled study (n=25) assessed whether a change in anticorrelated networks (default mode network (DMN)/salience network (SN)) underlies the peak effects of LSD (100μg) using fMRI. Inhibitory effective connectivity from the SN to DMN became excitatory, and inhibitory effective connectivity from DMN to DAN decreased under the peak effect of LSD suggesting that diminution of the functional anticorrelation between resting state networks that may be a key neural mechanism of LSD and underlie ego dissolution.

Abstract

“Background Classic psychedelic-induced ego dissolution involves a shift in the sense of self and blurring of boundary between the self and the world. A similar phenomenon is identified in psychopathology and is associated to the balance of anticorrelated activity between the default mode network (DMN) – which directs attention inwards – and the salience network (SN) – which recruits the dorsal attention network (DAN) to direct attention outward.

Methods To test whether change in anticorrelated networks underlie the peak effects of LSD, we applied dynamic causal modeling to infer effective connectivity of resting state functional MRI scans from a study of 25 healthy adults who were administered 100μg of LSD, or placebo.

Results We found that inhibitory effective connectivity from the SN to DMN became excitatory, and inhibitory effective connectivity from DMN to DAN decreased under the peak effect of LSD.

Conclusions The effective connectivity changes we identify may reflect diminution of the functional anticorrelation between resting state networks that may be a key neural mechanism of LSD and underlie ego dissolution. Our findings suggest changes to sense of self and subject-object boundaries across different states of consciousness may depend upon the organised balance of effective connectivity of resting state networks.”

Authors: Devon Stoliker, Leonardo Novelli, Franz X. Vollenweider, Gary F. Egan & Katrin H. Preller

Summary

We used resting-state functional MRI to investigate the effective connectivity changes between the amygdala and the default mode network, salience network, and central executive network under 0.215mg/kg psilocybin. We observed a general pattern of decreased top-down effective connectivity and directed connectivity changes associated with altered emotion and meaning under psilocybin.

Psilocybin, the main active compound of so-called magic mushrooms, is a serotonergic psychedelic that can induce profound changes to subjective experience, including visual illusions, hallucinations, and synaesthesia, emotion and cognitive capacities, and transcendence of space and time. Although ego dissolution is rare at the doses used in clinical trials, several studies have associated measurable degrees of ego dissolution to symptom reduction of depression and addiction.

Research has demonstrated that the amygdala and associative connectivity change under psychedelics, which may underlie the therapeutic efficacy of psychedelics. The amygdala is a bilateral subcortical region composed of many subnuclei that integrates cortical and thalamic sensory inputs and their emotional valence. It is also well known for its involvement in fear conditioning and is implicated across internalising disorders for which psychedelics demonstrate efficacy. Damage to the amygdala affects decision-making, memory, and behaviour regulation, and can diminish an individual’s sense of shame. The amygdala is integrated with cognition through top-down projections from midline cortical regions.

Psilocybin and LSD have been shown to influence limbic region emotional processing, and decreased amygdala connectivity has been correlated with positive mood under psilocybin and LSD. Clinical populations measured the day after administration of psilocybin demonstrated an opposite pattern of amygdala activity. This may be due to the fact that the acute effects of psychedelics may disrupt conditioned response patterns and produce enduring connectivity changes that support therapeutic modes of behaviour. Investigation of cortical regions with established connections to the amygdala motivates the creation of resting-state networks.

The RSN regions of particular relevance to this study are abundant in 5107 HT2A receptors and integrate with the 108 amygdala. The default mode network (DMN) is a brain region involved in self-referential mental activity. It is composed of the medial prefrontal cortex (mPFC) and posterior cingulate cortex (PCC). The DMN is related to wellbeing, for example, increased connectivity between the hippocampus and prefrontal cortex is observed in depression, and decreased PCC-mPFC coupling is associated with reduced mind wandering in advanced meditation practitioners.

The dorsolateral prefrontal cortex (dlPFC) and lateral posterior parietal cortex (lPPC) are involved in thinking, planning, and controlling attention, and the dlPFC plays a significant role in the experience of self under psychedelic subjective effects. CEN functions are associated with mental health in various cognitive disorders, and may be task and context-dependent. However, under psychedelics, CEN connectivity deviates from its usual functional pathways and proliferates to regions outside of this network.

Psilocybin renders the phase-locking of blood oxygen level-dependent (BOLD) neurovascular signals flexible, which may be associated with changes in perceptual boundaries and the reduced integrity of the brain’s “rich-club” organisation. The salience network (SN) is a cognitive RSN involved in detecting stimuli associated with biological or goal-directed personal significance. It recruits CEN activity to provide task-relevant information in response to salient features in the environment. The SN detects behaviourally relevant stimuli, receives multimodal sensory input, and includes the amygdala, which may be important for the initial emotional appraisal of salient stimuli. The SN may also act as pre-reflective mechanisms of attention that underlie perception and beliefs.

Serotonergic receptors in the SN and DMN are involved in the selection of channels of attention, and the dACC is associated with changes in self-processing and a reduced sense of social exclusion under psilocybin administration. We applied spectral dynamic causal modelling (spDCM) to investigate the effective connectivity changes between high-level RSNs and the subcortical amygdala under psilocybin. We measured changed meaning of 203 precepts and blissful state collected 360 min following the administration of psilocybin using the five dimensions of altered states of consciousness scale (5D-ASC).

Under psilocybin, the DMN and CEN showed decreased effective connectivity to the amygdala, and the PCC showed decreased effective connectivity to the left amygdala. These changes are hypothesised to be associated with altered meaning of precepts and blissful state.

Fig. 1 shows the change in network connectivity to the amygdala under psilocybin exposure, including changes in effective connectivity, self-inhibition, and connectivity to inhibitory circuits. Several changes in effective connectivity were found between regions of the CEN and amygdala, including increased self-inhibition of the rdlPFC, ldlPFC and decreased self-inhibition of the lLPPC, compared to placebo.

Psilocybin increased inhibitory effective connectivity from cortical regions to the amygdala, and this may relate to subjective and therapeutic effects of psilocybin. The PCC and amygdala showed multiple connectivity changes and behavioural associations under psilocybin, including increased synaptic gain and reduced PCC self-inhibition.

The DLPFC and bilateral lLPPC inhibition to the amygdala is situated within the CEN and is believed to underlie the evaluative aspect of thought, but was not associated with subjective effects. Participants in this study reported nominal anxiety, but the connectivity pattern between regions of the CEN was associated with changed meaning of precepts and ego dissolution. This may suggest that increased intrinsic CEN connectivity is related to resilience.

Psilocybin increases synaptic gain in the ACC and lAI, and decreases inhibition to the right amygdala, which may be therapeutic in internalising disorders characterised by enhanced salience detection. However, psychotic-like behavioural outcomes rarely occur in a controlled clinical setting. Previous research shows that increased functional connectivity between the AI and amygdala is associated with relapse of addiction and behavioural habituation. The inhibition of the amygdala by the AI may be associated with therapeutic behavioural change.

Psilocybin attenuates the recruitment of the amygdala, which contributes to the altered meaning of precepts and subjective emotion. This may explain reduced aversion to emotional stimuli and altered patterns of connectivity that condition emotional, cognitive and self-reflective responses of the brain. The author’s endeavour to investigate the interactions between emotion and cognition in a healthy population under psilocybin has several limitations. This paper is made available under a CC-BY-NC-ND 4.0 International license.

The methods and analysis used in this study have limitations, such as the use of global signal regression and the limited temporal resolution. Future studies should use modalities with temporal specificity capable of resolving insight and emotional change. The results of this study are limited by the small healthy adult sample and the congruence between our results and previous findings. Further research may reveal more precise and nuanced dynamics between the limbic regions and the cortex underlie previously theorised global patterns.

The CEN region shows increased directed excitation, which contrasts with previous analyses, which found a general reduction in associative functional connectivity. We demonstrated altered within-RSN effective connectivity and reduced top-down RSN-418 amygdala effective connectivity associated with emotional and cognitive behavioural changes under psilocybin. These changes may contribute to lasting therapeutic outcomes and enable alternate selection of perceptual hypotheses which underwrite self-beliefs.

Twenty four subjects were recruited through advertisements at universities in Zurich, Switzerland, and were screened for medical history, physical examination, blood analysis, and electrocardiography. The Mini-International Neuropsychiatric Interview, the DSM-IV fourth edition self-rating questionnaire for Axis-II personality disorders, and the Hopkins Symptom Checklist were used. Participants were asked to abstain from prescription and illicit drug use two weeks prior to first testing, and from alcohol use 24 hours prior to testing days. They provided written informed consent statements before participation in the study. A double blind, randomised, placebo-controlled, cross-over study was performed. Participants were orally administered psilocybin or placebo and resting state scans were taken 20, 40 and 70 minutes following administration.

We acquired MRI data on a Philips Achieva 3T whole-body scanner using a whole brain gradient-echo planar imaging (EPI) sequence and two high-resolution anatomical images using 3D magnetization prepared rapid-acquisition gradient echo sequence (MP-RAGE) and a turbo spin-echo sequence, respectively. Three subjects were excluded due to head motion, and one subject did not complete the scan at 70 minutes. We investigated independent models of amygdala effective connectivity with the DMN, CEN and SN.

Neurosynth was used to identify regions of interest in the DMN, SN and CEN, and associations between peak coordinates and cardinal nodes of network regions of interest were determined by expert visual inspection. We used a generalized linear model (GLM) to regress head motion parameters, white matter and cerebrospinal fluid signals from preprocessed data, and used global signal regression in our pre-processing pipeline.

The 11 ROIs for the DMN, SN, CEN, and amygdala were defined, and three independent fully-connected DCM models were specified and inverted for each subject. The effective connectivity that best explains the observed cross-spectral density was inferred for each testing condition. A general linear model was used to decompose individual differences in effective connectivity into hypothesised group-average connection strengths together with unexplained noise. Hypotheses were tested within the parametric empirical Bayes (PEB) framework.