Psilocybin desynchronizes the human brain

The study investigates how psilocybin alters human brain network organization acutely and over subsequent weeks, and how these changes relate to subjective psychedelic experiences. Psychedelics act primarily via 5-HT2A receptor agonism and can induce profound alterations in perception of self, time, and space, with clinical trials showing rapid and sustained symptom relief for depression, addiction, and end-of-life anxiety after a single high dose. Rodent studies show transient 5-HT2A activation induces persistent plasticity, including synaptogenesis in medial frontal cortex and anterior hippocampus. However, species differences and limitations of animal models constrain translation to humans. Prior human imaging work reports increased glutamate signaling and glucose metabolism during psilocybin, reduced electrophysiological power, decreased hemodynamic fluctuations, and reduced segregation between functional networks, but the drivers—especially subcortical effects—and persistence of changes are not fully characterized. The default mode network (DMN), functionally connected with ventromedial prefrontal cortex and anterior/middle hippocampus and implicated in self-referential processing and depression, is a candidate circuit mediating therapeutic effects. Using precision functional mapping with dense longitudinal fMRI sampling, the authors aimed to track individual-specific acute and persistent brain changes before, during, and after psilocybin, compare with methylphenidate to control for arousal, relate neural changes to mystical experience ratings, assess task state effects, and test for weeks-long alterations, particularly in hippocampal–DMN connectivity.

The paper situates its work within evidence that psychedelics cause acute alterations in human brain metabolism and electrophysiology and reduce functional network segregation. Clinical trials demonstrate rapid and persistent symptom relief in depression, addiction, and anxiety following high-dose psilocybin. Animal studies show 5-HT2A-mediated neuronal communication changes and plasticity (e.g., synaptogenesis in medial frontal cortex and anterior hippocampus) potentially underlying antidepressant-like effects. However, rodent-to-human translation is limited by receptor homology differences. Previous human resting-state studies report diverse acute changes (psilocybin, LSD, ayahuasca) converging on increased global integration and reduced network segregation; persistent post-psilocybin effects have been mixed. Increased hippocampal–DMN FC has been linked to depressive symptoms, with reductions following treatment. The DMN and anterior hippocampus are rich in 5-HT2A receptors and implicated in self and time perception, suggesting their involvement in psychedelic effects. The study also references work showing that dissociatives like nitrous oxide and ketamine similarly reduce network segregation, and grounding or external stimulation can modulate psychedelic neurodynamics.

Design: Precision functional mapping with dense longitudinal fMRI sampling tracked individual-specific brain changes across conditions. Healthy young adults (n=6) each completed roughly 18 MRI visits, including dense pre-drug baseline sessions, drug sessions, and post-drug follow-up for 3 weeks; participants returned for an additional psilocybin dose 6–12 months later. Drugs: Single high-dose psilocybin (25 mg) and methylphenidate (MTP; 40 mg), administered 1–2 weeks apart, with MTP serving as an arousal-matched control. Imaging: Resting-state fMRI scans (15 min per scan) acquired across visits; quality control metrics reported for 129 total MRI visits. Individualized parcellations and network definitions were used to account for inter-individual variability. Metrics and analyses: Functional connectivity (FC) change quantified as Euclidean distance from each participant’s same-participant baseline across cortex and subcortical structures. Linear mixed-effects (LME) models tested drug effects and associations with behavior, with permutation testing (TFCE) for spatial maps; rotation-based null models (spin tests) assessed network-level effects. Multidimensional scaling (MDS) on parcellated FC matrices (blind to session labels) identified latent dimensions capturing variance across scans. Normalized global spatial complexity (NGSC) quantified spatial entropy of brain activity (global and parcel-level), reflecting desynchronization (higher NGSC indicates more independent spatial time courses). Relationship to subjective experience: Session-by-session correlations between whole-brain FC change/NGSC and Mystical Experience Questionnaire (MEQ30) scores; nuisance variables (e.g., head motion) assessed. Task modulation: During some drug sessions, participants performed a simple auditory-visual matching task (>80% accuracy), enabling LME tests for task × drug interactions on FC change and NGSC; regression of evoked responses and scan order controls tested robustness. Replication and generalization: Dimension-1 scores computed in extant datasets with intravenous psilocybin (n=9) and LSD (n=16); comparison to stimulant effects in the ABCD Study (n=487) for generalizability; analysis of LSD dataset for NGSC replication. Neurochemical correspondence: Spatial correlations between psilocybin/LSD NGSC maps and 5-HT2A receptor density (C-11 Cimbi-36 PET). Persistent effects: FC changes 1–21 days post-psilocybin compared to pre-psilocybin; targeted analysis of anterior hippocampus and hippocampal–DMN FC time courses; comparison to post-MTP to test specificity. Statistics: LME models with estimates and confidence intervals; permutation/TFCE for imaging maps; spin tests; paired t-tests in external datasets; equivalence testing for post-MTP persistent effects. Data resources: Extended Data and Supplementary materials include unthresholded maps, full FC distance matrices, preprocessing robustness checks, task performance, and evoked response analyses.

• Psilocybin induces widespread acute FC disruption across cortex and subcortex, strongest in association networks and DMN-connected subcortical regions (thalamus, basal ganglia, cerebellum, hippocampus; anterior hippocampus foci at MNI −24, −22, −16 and 24, −18, −16). - Compared to methylphenidate (MTP), psilocybin produced >3× larger FC changes (post hoc two-sided t-test, P = 3.6×10^-12, uncorrected). LME for whole-brain FC change: 10 psilocybin doses (275 observations), estimate (95% CI) = 15.83 (13.50, 18.15), P = 1.36×10^-10, uncorrected. - Relative effect sizes (normalized): day-to-day 1.00; task 1.22; MTP 1.10; high head motion 1.29; psilocybin 3.52; between-person 3.53. - Network specificity: FC change largest in DMN (spin test, 1,000 permutations, one-sided P_spin < 0.001; other networks P_spin > 0.05). MTP effects largest in motor/action networks (P_spin = 0.002). - Data-driven MDS revealed a “psychedelic dimension” (dimension 1) that separated psilocybin from non-drug and MTP scans; top-weighted edges indicated reduced segregation between typically anticorrelated networks (DMN vs fronto-parietal, dorsal attention, salience, action-mode). Similar dimension-1 increases observed in independent datasets with i.v. psilocybin (n = 9; paired t = 2.97, P = 0.0177) and LSD (n = 16; paired t = 4.58, P < 0.001), suggesting generalization across psychedelics. - Network-level FC changes under psilocybin showed decreased within-network integration and decreased between-network segregation (loss of anticorrelations), indicating dissolution of functional brain organization. - Subjective experience linkage: Whole-brain FC change correlated strongly with MEQ30 mystical experience scores across drug sessions (r = 0.81; LME effect of FC change: t(13) = 7.68; P = 3.5×10^-5, uncorrected); head motion was not significantly related. Association cortex primarily drove this relationship; the MEQ30 “transcendence of time and space” subscale correlated most strongly (r^2 = 0.86). - Desynchronization (entropy) explains FC change: Psilocybin significantly increased NGSC (global spatial entropy), returning to baseline by the next session. Global NGSC LME estimate (95% CI) = 0.0510 (0.0343, 0.0676); P = 2.0×10^-6. Parcel-level NGSC: estimate = 0.0149 (0.0071, 0.0228); P = 2.30×10^-4; largest in association cortex, minimal in primary cortex. NGSC correlated with MEQ30 (r = 0.80, P = 3.52×10^-7, uncorrected, after single outlier removal). - Neurochemical correspondence: Spatial distribution of psilocybin- and LSD-associated NGSC increases correlated with 5-HT2A receptor density (Cimbi-36 binding): r = 0.39, P = 1.9×10^-13 (psilocybin); r = 0.32, P = 4.5×10^-5 (LSD), uncorrected. - Task engagement attenuates effects: Performing a simple perceptual task reduced psilocybin-associated FC change and NGSC (LME interaction task × psilocybin on FC change: estimate = −6.48 (95% CI −9.59, −3.37), P = 5.49×10^-5; on NGSC: estimate = −0.042 (95% CI −0.056, −0.027), P = 4.82×10^-4). Task-evoked response amplitude in primary visual cortex was reduced by psilocybin; other task-related responses were not significantly changed. - Persistent effects: Whole-brain FC largely returned to baseline 1–21 days post-psilocybin, but a targeted region in anterior hippocampus showed significant persistent FC change in the 3-week post-drug period (LME mean change = 0.095, P_pre–post-psilocybin = 0.0033; uncorrected). No persistent FC differences post-MTP (90% CI −0.056, 0.080; equivalence δ = ±0.086; P_pre–post-MTP = 0.77). - Anterior hippocampus–DMN connectivity decreased post-psilocybin for ~3 weeks: mean (95% CI) pre = 0.180 (0.169, 0.192) vs post = 0.163 (0.150, 0.176); values returned to baseline by 6–12 months (replication underpowered, n = 4).

Findings demonstrate that psilocybin acutely desynchronizes human brain activity at global and regional scales, dissolving typical network organization by reducing within-network integration and between-network segregation (loss of anticorrelations), especially in association cortex and DMN-connected subcortical structures. This neurodynamic shift aligns with a latent “psychedelic dimension” that generalizes across datasets and psychedelic compounds (psilocybin, LSD), and with prior observations of increased global integration under psychedelics. Strong correlations between neural desynchronization/FC change and mystical experience (particularly transcendence of time and space) tie neurophysiology to subjective phenomenology. The spatial correspondence with 5-HT2A receptor density supports a receptor-level mechanism underpinning these macro-scale effects. Task engagement attenuated desynchronization and FC disruption, consistent with clinical grounding strategies that direct attention externally to mitigate overwhelming experiences, highlighting state- and context-dependence of psychedelic effects. Although whole-brain FC largely normalized within weeks, a persistent decrease in anterior hippocampus–DMN connectivity suggests a circuit-specific subacute modification aligning with neuroplastic processes observed in animal models and with depression-relevant circuitry. The authors propose that acute desynchronization (increased entropy) perturbs typical network activity patterns, engaging glutamate-dependent and homeostatic plasticity mechanisms (e.g., BDNF, mTOR, EEF2), potentially contributing to pro-plastic and therapeutic outcomes. The pronounced DMN and anterior hippocampal involvement integrates micro- and macro-scale findings and offers a mechanistic framework linking receptor pharmacology, network dynamics, and clinical effects.

Using precision functional mapping with dense longitudinal fMRI, the study shows that a single high dose of psilocybin causes large-scale desynchronization of brain activity and widespread FC reorganization, far exceeding effects of an arousal-matched stimulant. These acute changes track the intensity and qualities of the mystical experience, generalize across psychedelic datasets, and are modulated by task engagement. Importantly, psilocybin induces a persistent reduction in anterior hippocampus–DMN connectivity for weeks, suggesting a candidate circuit-level correlate of pro-plasticity and therapeutic effects. The work bridges micro-scale 5-HT2A pharmacology with macro-scale network dynamics and subjective experience, advancing understanding of psychedelic mechanisms. Future research should test these mechanisms in clinical populations using precision individual-level mapping, establish causal links to symptom change, and develop in vivo markers of neurotrophic plasticity to integrate cellular, circuit, and psychological levels.

• Small sample size (n = 6) for primary longitudinal dataset limits generalizability and statistical power, especially for detecting subtle persistent effects; replication of long-term effects had n = 4 pre-psilocybin visits. - Many statistical tests reported as uncorrected for multiple comparisons; findings, particularly spatial maps and correlations, may be susceptible to Type I error. - Healthy, non-depressed volunteers were studied; clinical relevance and mechanisms in patient populations remain to be validated. - Reliance on resting-state fMRI and derived metrics (FC Euclidean distance, NGSC) may be influenced by preprocessing and physiological noise, though robustness checks were performed. - The i.v. psilocybin and LSD dataset analyses were re-analyses and not part of a unified experimental protocol; differences in acquisition and dosing could confound generalization. - Persistent whole-brain FC largely returned to baseline, suggesting circuit-specific rather than global subacute effects; the functional significance of anterior hippocampal changes requires further study. - MEQ30 associations are correlational and depend on self-report; causality between neural changes and subjective experience cannot be inferred.