P300-mediated modulations in self-other processing under psychedelic psilocybin are related to connectedness and changed meaning: A window into the self-other overlap

The paper frames ‘‘self’’ as the demarcated subject of experience and highlights neuroscientific evidence that the brain distinguishes self-related from other-related stimuli across sequential processing stages. Early sensory stages (roughly 100–200 ms) show differences in N1/N170 and P2/N2 amplitudes thought to reflect automatic sensory-feedback mechanisms that detect expected, overlearned self-input (for example, one's own voice). Later stages, particularly the P300 component (approximately 300–600 ms), have been linked to elaborative and categorical processing in frontoparietal circuits and typically show a processing bias favouring self-related stimuli. Disruptions of these stages have been implicated in psychiatric symptoms such as schizophrenia, bipolar disorder, and depression, but the neuropharmacology of self–other differentiation and its relationship to subjective experience remain underexplored. To address this gap, T. and colleagues administered a single oral dose of psilocybin (230 μg/kg) versus placebo in a double-blind, within-subject crossover design in healthy volunteers and used an auditory verbal self-monitoring task combined with high-density EEG to probe behavioural, experiential, and neural markers of self–other processing. The study tested whether psilocybin would blur the differentiation between self and other at behavioural and neural levels, with a particular focus on alterations in the N100 and P300 event-related potentials and associations between neural changes and subjective drug effects (dimensions of unity, disembodiment, and changed meaning of percepts).

Design and participants: The experiment employed a double-blind, placebo-controlled, within-subject crossover design with counterbalanced order. Seventeen healthy right-handed participants (9 males; mean age 25.1 ± 4.1 years; mean verbal IQ 104.9 ± 10.7) completed both sessions separated by at least two weeks. The extraction states that recruitment and inclusion/exclusion criteria and a study flow diagram are reported in Supplementary Material, but those details are not present in the provided text. Drug administration and psychometrics: Psilocybin was administered orally at 230 μg/kg body weight; placebo capsules contained lactose. Subjective drug effects were measured using the Altered States of Consciousness rating scale (5D-ASC). From that instrument the authors focused on three self-related dimensions—feeling of unity, disembodiment, and changed meaning of percepts—and one audio-visual synesthesia dimension, given the auditory task modality. Task and behavioural measures: The core behavioural assay was a verbal self-monitoring task initiated 75 minutes post-administration (around the peak plateau of psilocybin effects). On each trial participants vocalised the sound [a:] and received real-time auditory feedback that varied by source (their own voice, a prerecorded other voice matched by sex) and by pitch (unaltered or two semitones down). After each trial participants judged whether the heard voice was their own, someone else's, or they were unsure; they were instructed to respond ‘‘self’’ when they recognised their own voice even if pitch-shifted. Background pink noise was played to attenuate bone conduction; trials included 280 total (70 per condition). Reaction times and responses were recorded; late responses (>5 s) were marked as missing. The extraction does not provide full technical details of voice-processing algorithms beyond noting real-time pitch shifting and prerecorded replacement for the ‘‘other’’ condition. EEG acquisition and preprocessing: EEG was recorded with a 64-channel BioSemi ActiveTwo system, sampling at 2048 Hz. Preprocessing (BrainVision Analyzer 2.1) included down-sampling to 512 Hz, band-pass filtering 0.5–20 Hz, ocular artefact removal via independent component analysis, interpolation of noisy channels, epoching from −200 to +800 ms relative to vocalisation onset, rejection of sweeps exceeding ±80 μV or gradients >30 μV/ms, baseline correction (−200 to 0 ms), and average referencing. The reported mean ± SD of accepted epochs per condition are provided in the extraction (for example, placebo self 67.7 ± 2.0), but the text labels these with ‘‘ms’’ which appears inconsistent; the extraction does not clearly report units for these values. Behavioural analysis: Behavioural responses (correct, incorrect, unsure) were analysed using multinomial logistic regression to estimate response probabilities as a function of treatment (psilocybin/placebo), source (self/other), and pitch (unaltered/pitch-shifted). Two models were used: Model 1 included all trials; Model 2 excluded pitch-shifted trials to focus on the genuine self–other contrast. Missing responses were rare and were not modelled. EEG/statistical approach: The authors used an assumption-free, whole-scalp randomization framework implemented in the RAGU toolbox to assess spatiotemporal ERP differences without selection of channels or time windows a priori. Tests included the topographic consistency test (TCT), topographic analysis of variance (TANOVA) and topographic analysis of covariance (TANCOVA) with 5,000 permutations across the full 800 ms poststimulus window. Global duration statistics were applied to account for temporal autocorrelation and multiple comparisons, accepting effects that exceeded 95% of durations in randomized permutations. TANOVA assessed spatial distribution differences between factors (treatment, source) and constrained permutations to maintain sensitivity. Significant TANOVA effects were followed by post hoc t-maps and multidimensional scaling. TANCOVA tested linear covariation between ERP topography differences and continuous behavioural/subjective variables (ASC differences) using bootstrapping and randomization. Current source density (CSD) inverse analyses were applied to infer likely cortical generators for observed scalp topographies within identified time windows.

Subjective effects: Psilocybin produced robust alterations on the 5D-ASC. A two-way analysis showed a significant drug-by-scale interaction (F10,160 = 16.71, p < .0001) and main effects of scale (F10,160 = 17.29, p < .0001) and treatment (F1,16 = 83.05, p < .0001). Post hoc Tukey tests indicated significant increases after psilocybin for most ASC subscales except anxiety (p = 0.999) and the spiritual experience subscale (p = .991); insightfulness increased at p = .024 and other subscales at p < .0001 according to the extracted text. Behavioural task performance: Response counts and proportions differed between placebo and psilocybin. Under placebo the task yielded 3,823 correct (80.43%), 628 incorrect (13.21%), and 302 unsure (6.35%) responses. Under psilocybin there were 3,442 correct (73.03%), 911 incorrect (19.33%), and 360 unsure (7.64%). Multinomial logistic regression results reported positive coefficients and significant p-values indicating increased probability of inaccurate responses under psilocybin. Specifically, reported coefficients for the treatment × source interaction were β = .373 (p = .004) and β = .676 (p = .003) for Models 1 and 2 respectively, supporting the hypothesis that psilocybin disproportionately reduced correct identification of self voices. Inclusion of pitch produced a significant main drug effect on false recognitions (β = .330, p = .004), but this effect became nonsignificant when pitch-shifted trials were excluded (β = .223, p = .084). A three-way interaction (treatment × source × pitch) was nonsignificant (β = .264, p = .085 for incorrect and unsure responses). EEG topographies and TANOVA: The TCT was used first to assess topographic consistency and led the authors to disregard some early timeframes that lacked consistent scalp configurations. The main TANOVA revealed a drug-by-source interaction in the P300 timeframe (noted in figures as roughly 452–520 ms). Post hoc t-maps showed a significant difference in scalp configuration between self and other under placebo (p = .017; tmax at Cpz = 4.580; tmin at TP7 = −5.728) but not under psilocybin (all other contrasts p > .270, uncorrected). Multidimensional scaling indicated that self and other processing were more distinct under placebo and much more similar under psilocybin, consistent with a loss of topographic distinctiveness in the P300 window. Current source density (inverse) results: Within the identified P300 window, inverse modelling implicated differential contributions from supragenual anterior cingulate cortex and right insular cortex to self–other distinctions that were diminished under psilocybin. The extraction also notes lower CSD in the right auditory cortex under psilocybin, suggesting reduced engagement there. Precise numerical source estimates and statistical details of the inverse solutions are not fully reported in the extracted text. Associations between neural changes and subjective experience: TANCOVA revealed linear covariation between the psilocybin-induced difference in self-related ERPs and ASC subscales in specific time windows: feeling of unity (447–472 ms), changed meaning of percepts (485–520 ms), and disembodiment (94–130 ms). These effects exceeded 20 ms in duration and were reported as significant at p < .05. The initial 5 ms of the unity effect occurred during a period of inconsistent topographies per the TCT and is noted as such. No reliable TANCOVA effects were found for other-related conditions, apart from a disembodiment effect (335–370 ms) that was discarded due to inconsistent topographies. Other observations: The authors report a general psilocybin effect in the N100 timeframe linked to disembodiment, but this early effect did not meet the global duration statistics criterion and was therefore treated cautiously.

T. and colleagues interpret their findings as evidence that psilocybin reduces the neural distinctiveness between self- and other-related auditory stimuli during late, P300-stage processing. Behaviourally, psilocybin disproportionately impaired correct self-attribution in the speech-monitoring task, and neurally the scalp topographies that normally distinguish self from other in the P300 window under placebo were abolished under psilocybin. Multidimensional scaling and waveform inspection supported the view that psilocybin made self- and other-related neural events more similar. The authors link these P300 modulations to inversely modelled changes in supragenual anterior cingulate and right insular cortex engagement, and they note a reduction in right auditory cortical CSD, situating the effects within frontomedial and insular regions rich in 5-HT2A/1A receptor expression—the principal targets of psilocybin. The authors emphasise that the magnitude of the P300 alteration covaried with subjective reports of unity and changed meaning of percepts, suggesting a relationship between altered self–other neural encoding and experiential connectedness or reappraisal of perceptual meaning. They note this aligns with prior imaging work linking psychedelic drugs to altered meaning attribution and with literature showing attenuated P300 self–other differences in meditators and in paradigms that promote self–other overlap. The authors propose that altered meaning attribution and reduced self-focus may be mechanisms relevant to the therapeutic effects of psychedelics, for example by reducing maladaptive rumination or enhancing connectedness in depression. The discussion recognises limitations the authors acknowledge: possible confounding by attentional deficits or task disengagement under psilocybin (though they argue the treatment × condition interaction and the absence of a main drug effect on P300 topography mitigate this), potential carry-over effects inherent to crossover designs and the challenge of effective blinding for conspicuous psychoactive compounds, and domain specificity since the study focused on vocal/auditory self-monitoring. They also note uncertainty about generalisability to other tryptamine psychedelics. The authors recommend future studies examine other perceptual domains and further probe early (N100) mechanisms and sensory-motor corollary discharge processes that relate to bodily aspects of self-experience. Finally, the authors situate their results in broader relevance to social and clinical neuroscience: transient psilocybin-induced reductions in self–other differentiation may underlie phenomenology such as ego-dissolution and feelings of connectedness, phenomena reported to persist and to mediate therapeutic gains in clinical trials of psychedelic-assisted therapy. They suggest that perturbing self-referential processing via serotonergic mechanisms provides an experimentally tractable model for studying the neurobiology of subjectivity and may inform clinical applications targeting excessive self-focus and disconnection.

Using high-temporal-resolution EEG and data-driven whole-scalp analyses, the authors conclude that psilocybin abolishes the distinctiveness of self-related scalp configurations in the P300 timeframe, associated with altered activity in the supragenual anterior cingulate cortex and insula. These findings are presented as advancing mechanistic understanding of self–other processing and the neurobiological basis of subjectivity, and the authors propose that serotonergic psychedelic stimulation offers a valid experimental platform for perturbing and quantifying altered self-referential processing.