Enhanced visual contrast suppression during peak psilocybin effects: Psychophysical results from a pilot randomized controlled trial

Surround suppression is a well-established visual phenomenon in which the perceived contrast of a central stimulus is reduced by a higher-contrast surrounding stimulus. Although psychophysical studies have characterised surround suppression extensively, its neurochemical underpinnings are not well understood. Prior pharmacological probes have implicated GABAergic, cholinergic, adenosinergic, dopaminergic and noradrenergic systems, but the role of serotonergic signalling—particularly via 5-HT2A receptors implicated in psychedelic effects—has not been tested in humans. Psilocybin, a serotonergic psychedelic metabolised to psilocin and acting as a 5-HT2A agonist, produces dose-dependent alterations in visual phenomenology, yet psychophysical assessments of centre–surround interactions under psilocybin in people are lacking. This pilot study set out to measure whether peak effects of a high oral dose of psilocybin (25 mg) alter visual contrast surround suppression in healthy adults. Using a contrast-matching (two-alternative forced choice) task with three surround configurations (no surround, orthogonal surround, parallel surround), the investigators compared point-of-subjective-equality (PSE) under psilocybin versus an active placebo (100 mg niacin) in a within-subjects crossover design. The study also assessed subjective psychedelic effects with the 5D-ASC questionnaire and explored correlations between subjective visuals and changes in surround suppression. This work was performed as a small feasibility pilot prior to a larger trial and included safety monitoring for adverse events.

Link and colleagues conducted a triple-blind, randomised, placebo-controlled crossover trial at the University of Minnesota. Eligible participants were adults aged 25–38 years with normal or corrected-to-normal vision, no current or prior mental-health diagnosis, not on prescription medications, and with at least one previous moderate-to-high psilocybin experience; six participants (three male, three female, mean age 33 years) completed the study. The study was approved by institutional and regulatory bodies and registered (NCT04424225). Each participant completed two dosing sessions separated by two weeks: one 25 mg oral psilocybin session and one 100 mg oral niacin (active placebo) session. Drug and placebo were given in identical opaque capsules; participants, experimenters and most study staff were blind to session order, and group assignment counterbalanced who received psilocybin first. Tasks were performed roughly three hours after dosing to coincide with the latter half of peak subjective and plasma effects. Participants rested between dosing and testing. Visual stimuli were annular sinusoidal luminance gratings presented at 3° eccentricity left and right of fixation on a calibrated 23-inch LCD monitor. The target grating had fixed 50% Michelson contrast; a reference grating varied in contrast via adaptive staircases to estimate PSE (the contrast at which target and reference appeared equal). In some conditions the target was surrounded by a 4° outer-diameter annulus at 100% contrast with either parallel (0°) or orthogonal (90°) orientation; a 0.5° gap separated target and surround. There were three stimulus conditions: No Surround (NS), Orthogonal Surround (OS), Parallel Surround (PS). Two staircases per condition (target left/target right) of 40 trials each were interleaved; seven practice trials and 40 easy 'catch' trials (reference fixed at 80%) were included. Stimulus presentation was until response; response time was not limited. Subjective effects were measured using the 5D-ASC (94-item visual-analogue scale) and its 11-subscale factor structure (11D-ASC). Participants completed the questionnaires retrospectively three or four days after each dosing session. For psychophysical measures, PSEs were obtained by fitting logistic functions to staircase data; guess and lapse rates were fixed at 4%. Surround suppression strength was quantified as PSE minus the veridical target contrast (50%). Two complementary statistical analyses were performed. First, after unblinding (double-blind stage), repeated-measures ANOVAs tested effects of drug and stimulus condition on PSE and on 5D-ASC scores, followed by post-hoc t-tests; effect sizes were reported as nG2 (ANOVAs) and Bayes factors (t-tests). Correlations between PSE and subjective scores used repeated-measures correlation (rmcorr), which accounts for within-subject repeated measures, with follow-up robust biweight midcorrelations on difference scores (psilocybin minus placebo). Second, an independent biostatistician performed a triple-blind analysis using linear mixed-effects models comparing PSEs across drug and condition, assuming homogeneous variance; this analysis did not examine catch trials or subjective correlations. Analyses used PsychoPy utilities, Pingouin (Python) and R (version 4.2.2).

Primary psychophysical outcome: In the double-blind repeated-measures ANOVA on PSEs, there was a significant main effect of drug (F1,5 = 7.294, p = 0.043, nG2 = 0.128) and a significant main effect of stimulus condition (F2,10 = 8.822, p = 0.006, nG2 = 0.381), consistent with surround suppression. The drug × condition interaction did not reach conventional significance but trended (F2,10 = 3.0, p = 0.095). Post-hoc paired t-tests comparing PSE across all stimulus conditions confirmed the main drug effect (t5 = 2.701, p = 0.043, Bayes factor = 2.342). When analysed by condition, psilocybin did not significantly affect PSE in the No Surround condition (t5 = 0.665, p = 0.535, Bayes factor = 0.446), while the OS and PS conditions showed trend-level reductions in PSE under psilocybin (OS: t5 = 2.467, p = 0.057, Bayes factor = 1.905; PS: t5 = 2.268, p = 0.073, Bayes factor = 1.594). Mean differences (psilocybin minus placebo) indicated stronger suppression under psilocybin: PS shifted by −7.48% contrast, OS by −5.34% contrast, and NS by −0.63% contrast, signalling enhanced illusory suppression when surrounds were present. Triple-blind analysis: The independent linear mixed-effects model produced a marginal effect of psilocybin on PSE (mean = 0.045, SE = 0.025, t59 = 1.80, p = 0.077). Follow-up contrasts suggested this effect was driven primarily by the PS condition (mean = 0.075, SE = 0.043, t59 = 1.73, p = 0.088); NS and OS contrasts were non-significant. Catch trials: Performance on easy catch trials remained high under both conditions (placebo mean = 99.17%, SD = 1.29%; psilocybin mean = 97.92%, SD = 1.88%). The paired t-test showed a non-significant reduction in accuracy under psilocybin (t5 = 2.24, p = 0.08), suggesting participants were generally able to perform basic contrast discriminations. Subjective effects: Psilocybin produced robust subjective alterations on the 5D-ASC. Repeated-measures ANOVA showed a significant main effect of drug on 5D-ASC scores (F1,5 = 20.109, p = 0.006, nG2 = 0.518), a main effect of subscale (F4,20 = 14.051, p < 0.001, nG2 = 0.364), and a significant drug × subscale interaction (F4,20 = 8.484, p < 0.001, nG2 = 0.342). Correlation between subjective visuals and PSE: Using rmcorr to relate within-subject variations in PSE and 5D-ASC scores across sessions, significant inverse correlations emerged between PSE and the Visionary Restructuralization dimension for OS (r5 = −0.856, p = 0.014) and PS (r5 = −0.836, p = 0.019). That is, higher ratings on Visionary Restructuralization were associated with greater surround suppression (lower PSEs) in conditions with a surround. Analyses using the 11D-ASC subscales indicated that certain lower-level visual subscales (e.g., Audio-Visual Synesthesia) showed strong associations with suppression magnitude; detailed correlation tables and post-hoc robust midcorrelations on difference scores are reported in supplemental materials. The Total 5D-ASC score and several other dimensions (including Oceanic Boundlessness) did not correlate significantly with PSE. Safety: No serious adverse events were reported. All recorded adverse events are tabulated in the supplementary materials; none were judged serious.

Link and colleagues interpret these pilot data as indicating that psilocybin enhanced visual surround suppression during peak subjective effects, rather than weakening it as they initially hypothesised. The psychophysical pattern—reduced PSEs under psilocybin in conditions with a surround but not in the no-surround condition—led the investigators to conclude that psilocybin may selectively amplify processing of visual contextual information while leaving basic contrast discrimination largely intact. The authors note that the effect did not clearly differ between parallel and orthogonal surrounds, which might indicate an orientation-insensitive modulation of surround suppression or could reflect limited statistical power to detect orientation-specific effects. The observed correlations between the Visionary Restructuralization dimension (and some 11D-ASC visual subscales such as Audio-Visual Synesthesia) and increased surround suppression are taken to imply a link between the subjective intensification of visual phenomenology and altered contextual processing in early vision. The authors discuss relevance to clinical questions: surround suppression has been reported as weaker in major depressive disorder and schizophrenia, whereas psilocybin here enhanced suppression, suggesting that psychedelic visual effects may be distinct from the visual abnormalities observed in those disorders. They further relate their findings to neurocomputational frameworks, noting that surround suppression is tied to divisive normalisation and both bottom-up and top-down processes; the results complicate some predictive-processing accounts that posit a simple weakening of top-down feedback under psychedelics. The investigators acknowledge multiple limitations that constrain inference. Chief among these is the small sample size (n = 6) typical of a feasibility pilot, which reduces power and increases the risk that some effects are sample-specific. Blinding was imperfect: participants correctly identified their psilocybin session after both visits, and expectancy effects may have influenced behaviour despite use of niacin as an active placebo and a within-subject crossover design. Statistical analyses included an unblinded (double-blind) set and an independent triple-blind analysis; the latter yielded only marginal effects but was directionally consistent. Other limitations reported include fixed dosing rather than weight-adjusted administration, exclusion of pre-dosing baseline data due to quality problems, potential carryover effects (including concerns about hallucinogen persisting perceptual disorder though deemed unlikely to have driven their findings), lack of receptor-specific pharmacological blockade to isolate 5-HT2A contributions, unlimited stimulus duration without eye-tracking (raising the possibility of fixation instability or longer dwell times under psilocybin), and delayed administration of the 5D-ASC questionnaire more than 24 hours post-session. For future work, the authors propose larger, better-powered studies, potentially using ketanserin to test 5-HT2A specificity, weight-adjusted dosing, baseline pre-dosing measures, tighter control of fixation and response timing, and prospective assessment of participant expectations. They emphasise that psychophysical testing under acute psychedelic effects can be conducted safely and that further study may elucidate how serotonergic neuromodulation shapes contextual visual computations.

In this small pilot crossover trial, psilocybin (25 mg oral) was associated with enhanced visual surround suppression in a contrast-matching task, and greater suppression correlated with the intensity of certain subjective visual experiences. The investigators conclude that serotonergic modulation may play an important role in surround suppression, that psychedelic visual phenomenology may be mechanistically distinct from visual changes reported in depression and schizophrenia, and that existing theoretical accounts of psychedelic effects on hierarchical predictive signalling may require refinement. They present these findings as preliminary and call for larger, more definitive studies to test receptor specificity, dose–response relationships, and clinical relevance.