Psilocybin and LSD May Reorganise the Brain's Global Workspace
New research reports that classic psychedelics (LSD, psilocybin and mescaline) increased the number of "gateway" regions in the brain's frontoparietal control and dorsal attention networks, while MDMA and amphetamine did not follow the same pattern.
The authors reanalysed resting-state fMRI scans from two Basel crossover studies using integrated information decomposition, which separates information shared across brain regions from information that emerges only when regions are considered together.
Differences between drugs in these patterns tracked differences in their reported subjective effects, but the data cannot show that the neural changes cause those experiences, and the work has not been peer reviewed.
Classic psychedelics such as LSD, psilocybin and mescaline may reorganise how information is integrated across the brain, according to a new bioRxiv preprint. Reanalysing resting-state brain scans from two studies conducted in Basel, the researchers found that these substances increased the number of “gateway” regions in networks involved in cognitive control and attention. MDMA and amphetamine showed a different pattern, suggesting that these drugs alter brain information processing in distinct ways. The findings remain preliminary, as the preprint has not yet been peer reviewed.
A Theory of Consciousness, Measured With Information Theory
Global workspace theory proposes that the brain's specialised systems, such as visual cortex, process information largely in parallel. Information becomes consciously accessible only when it enters a shared workspace and is broadcast across the brain.
A 2024 paper in eLife, led by Andrea Luppi and colleagues, developed an information-theoretic approach to testing this idea. Integrated information decomposition distinguishes between two forms of information shared by brain regions: redundancy, which can be obtained from either region independently, and synergy, which emerges only when both are considered together. Redundant information can be obtained from either region alone. Synergistic information appears only when both are considered together.
Workspace regions with strongly synergistic links across the brain are labelled "gateways", which integrate information. Those with strongly redundant links are labelled "broadcasters", which redistribute it. The workspace was defined once, from 100 adults in the Human Connectome Project, and comprised 235 of 454 brain regions.
Two Basel Datasets
The scans came from two randomised, double-blind, placebo-controlled crossover studies of healthy volunteers at University Hospital Basel. One (LAM) gave LSD at 100 μg, MDMA at 125 mg and D-amphetamine at 40 mg. The other (LPM) gave LSD at 100 μg, psilocybin at 20 mg and mescaline at 300 mg or 500 mg. All doses were oral, and scans were taken two to three hours after dosing (1.5 to 2.5 hours for MDMA).
After motion screening, the analysis included 26 participants in the first study and 30 to 31 per drug in the second. The studies were funded by the Swiss National Science Foundation; the second also received support from University Hospital Basel and Mind Medicine Inc.
What Changed Under Each Drug
Across the four classic-psychedelic conditions, the number of gateways in the frontoparietal network rose by an average of 4.37 regions. The rise was individually significant under both LSD conditions, psilocybin (+4.52 regions) and mescaline (+4.03). Gateways in the dorsal attention network, which contributes to selecting information for attention, also rose significantly in all four conditions.
Default mode network results were less consistent. Gateway counts rose on average, but significantly only under mescaline and LSD in the second study, and the LSD result lost significance after adjusting for head motion.
MDMA and amphetamine both reduced gateway counts in the visual and somatomotor networks, and MDMA also raised dorsal attention gateways. Taken together, the results point to a reorganisation of brain information processing rather than a uniform increase in integration across all networks.
The distinction extended across both cohorts: patterns of synergy and redundancy consistently separated the classic psychedelics from MDMA and amphetamine.
Links to Subjective Experience
To explore how brain activity related to subjective experience, each drug's neural profile was compared with scores across the 11 subscales of the 5D-ASC questionnaire, a standard measure of altered states of consciousness. Drugs that were more alike in synergy patterns were more alike in reported experience (ρ = .92, q = .003), and redundancy showed a similar but weaker match (ρ = .76, q = .029). Synergy differences were associated with nine of the 11 subscales; redundancy with none.
No link emerged between specific networks and specific experiences. Across the six conditions, no subscale remained significant after adjusting for the mean of the other ten. These associations do not establish that the neural changes cause particular subjective effects.
Limitations and Open Questions
The workspace was defined using an independent sample and then kept fixed, assuming that its constituent regions do not change under psychedelics. Between-drug comparisons may be affected by cohort differences and by shared participants and placebo scans within each study. Only one mescaline fMRI dataset exists, so replication is not yet possible. Resting-state data also cannot reveal which information reaches the workspace, so the model cannot say whether gateway changes relate to, for example, ego dissolution or synaesthesia. Gateways are also a relatively new construct, and their test-retest reliability has yet to be systematically established.
All participants were healthy, so the study says nothing about clinical outcomes. For readers following how psilocybin acts on the brain in depression or the regulatory progress of psilocybin therapy, it offers a mechanistic framework rather than efficacy evidence.
