Dose-response relationships of psilocybin-induced subjective experiences in humans

Psilocybin, the primary psychoactive component of Psilocybe mushrooms, produces altered states of consciousness (ASC) through agonism at 5-HT2A receptors. Previous research has linked the phenomenological quality of psilocybin-induced experiences — including mystical-type experiences and ego-dissolution — to positive changes in mood and therapeutic outcomes in several clinical contexts. Despite growing interest in whether the subjective experience itself mediates therapeutic benefits, no meta-analysis had quantified how subjective effects scale with psilocybin dose across studies using standardised psychometric instruments. Hirschfeld and Schmidt set out to estimate dose-response relationships for psilocybin-induced subjective experiences measured with validated questionnaires in controlled, oral-administration studies. The primary aim was to derive linear dose-response estimates across multiple psychometric scales using the Altered States Database (ASDB) and a meta-regression approach suited to dependent effect sizes; a secondary aim was to explore whether including available patient data changed those dose-response patterns. The authors frame these estimates as a reference for experimental and clinical dose selection while noting non-pharmacological influences on subjective effects.

Data were drawn from the Altered States Database (ASDB_v1.1a_2020), a repository of psychometric results from peer-reviewed studies. Inclusion criteria limited the main analysis to studies of orally administered psilocybin in healthy participants that used one or more of the target questionnaires. Studies were excluded if they involved patient populations, intravenous administration, combined interventions (for example meditation), or duplicated previously reported samples; where studies reported subsamples, the larger sample was retained. The authors also conducted an additional analysis that added identified patient datasets published up to 20 December 2020, found by targeted searches and hand-searching reference lists. The meta-analysis pooled outcomes from three standardised instruments: the Altered States of Consciousness Rating Scale (5D-ASC and an 11-factor variant, 11-ASC), the Mystical Experience Questionnaire (MEQ30), and the Hallucinogen Rating Scale (HRS). Where necessary, doses reported as mg per a reference bodyweight were converted to micrograms per kilogram (μg/kg). Psychometric scores given as raw sums were converted to percentage of maximum score to harmonise scales. Statistical analysis used linear meta-regression under a random effects model to estimate intercepts and slopes of dose-response relationships. Although biological dose-responses are typically sigmoidal, the observed dose ranges did not span lower and upper asymptotes, so a linear approximation was adopted. To accommodate within-study dependence from repeated measures, the robust variance estimation (RVE) framework was used with a recommended within-study correlation parameter p = 0.8 and small-sample adjustments; a sensitivity analysis varied p from 0 to 1 to test robustness. Heterogeneity was quantified with Tau² (between-study variance) and I² (percentage inconsistency across studies). Visualisations included spiderplots of predicted percentage-max scores across doses and scatterplots showing individual study effect sizes weighted by their contribution.

Seventeen studies of oral psilocybin in healthy participants were included in the main analysis. Because several studies contributed repeated measurements at different doses, the final datasets comprised, for the 5D-ASC: 14 observations from 7 samples (12 observations from 5 samples for Auditory Alterations and Vigilance Reduction); for the 11-ASC: 8 observations from 6 samples (10 outcomes from 7 samples for some subscales); for the MEQ30: 11 observations from 4 samples; and for the HRS: 8 observations from 3 samples. In the supplementary patient-inclusive analysis, the authors identified data from studies of cancer-related psychiatric distress, major depressive disorder and treatment-resistant depression, with care taken to avoid duplicate subsamples. Across instruments, most scales and factors showed positive dose-response relationships: higher psilocybin doses were associated with larger questionnaire ratings. Exceptions were two 11-ASC subscales — Changed Meaning of Percepts and Impaired Control and Cognition — which did not show clear dose-dependency. The strongest dose-responses on the 5D-ASC were for Visionary Restructuralization (perceptual alterations) and Oceanic Boundlessness (positively experienced ego dissolution, including derealisation and depersonalisation with positive affect). Vigilance Reduction showed a medium dose-response, interpreted as contemplative or dream-like reductions in alertness rather than sedation. Dread of Ego Dissolution (anxiety, loss of self-control) showed only a small dose-response and relatively low mean scores. On the MEQ30, all four factors (Sacredness, Positive Mood, Transcendence of Time/Space, Ineffability) exhibited relatively large and similar dose-related effects. Using the fitted linear estimates, the authors noted that scores exceeding 60% on each MEQ30 factor — sometimes used to denote a complete mystical experience — corresponded to approximately 350 μg/kg and above, though the model had a relatively high intercept and lower-dose effects were less well characterised. HRS results indicated the largest dose responses for Cognition and Perception, while Volition changed little with dose. Sensitivity analyses varying the assumed within-study correlation produced negligible changes in slope estimates and only minor differences in intercepts, indicating robustness of the RVE-based parameters. Including patient data produced meta-regression estimates similar to those from healthy participants. Heterogeneity (I²) and between-study variance (Tau²) were small for most scales, suggesting reliable estimates, but substantial inconsistency (I² > 75%) was observed for Audio-Visual Synesthesia, Changed Meaning of Percepts and Elementary Imagery (11-ASC) and Somaesthesia (HRS), so dose estimates for those scales are less certain. The authors also note limited statistical power given the modest number of observations (8–14) relative to recommendations for meta-regression.

Hirschfeld and colleagues interpret the results as showing that psilocybin intensifies most measured aspects of ASC in a dose-dependent manner, with the largest effects on perceptual alterations and positively experienced ego dissolution. The MEQ30 data suggest that mystical-type experiences, operationalised as high scores across MEQ30 factors, become likely at higher doses (around 350 μg/kg), although the high intercepts and sparse low-dose data limit inference for very small doses. Challenging or aversive aspects of experience (for example anxiety-related scales) were generally less dose-sensitive and had lower mean ratings, but the authors caution that average scores do not rule out that individual participants may undergo highly challenging experiences. The discussion emphasises the role of non-pharmacological factors — often framed as set (personality, expectations, mood, intention) and setting (physical and social environment) — in shaping subjective responses. Reported moderators include current mood, psychological distress, trait absorption, extraversion, openness, neuroticism (inconsistently), prior hallucinogen experience, age, pharmacokinetic differences (plasma psilocin, 5-HT2A receptor occupancy) and brain structural metrics. These variables likely contribute to residual variance and mean that dose alone does not fully predict subjective outcomes. Comparisons with prior reports, intravenous psilocybin studies and combined interventions indicate broadly similar patterns for core perceptual and mystical measures, though earlier single-group analyses sometimes found stronger dose-dependence for Audio-Visual Synesthesia and Changed Meaning of Percepts. The authors attribute some between-study differences to methodological variation, including task demands and imaging environments. They also discuss limitations: reliance on introspective, retrospective questionnaires; the necessity of a linear approximation because available data did not span the full sigmoid dose-response; the RVE framework’s limitation in providing precise heterogeneity parameter estimates; and the selective, highly screened participant samples that limit generalisability to broader or recreational populations. Finally, the authors suggest that these dose-response estimates constitute a practical reference for experimental and clinical work but recommend future studies standardise assessment of non-pharmacological factors to improve predictive models.

The study concludes that oral psilocybin increases almost all questionnaire-measured characteristics of altered states of consciousness, most prominently perceptual alterations and positively experienced ego dissolution, and can occasion mystical-type experiences at higher doses. Despite acknowledged influences of non-pharmacological factors, the authors present robust linear dose-response estimates for most questionnaire factors and scales that can serve as a general reference for relating expected and observed dose-specific subjective effects in controlled research settings. They caution that results may not generalise to recreational use and recommend more standardised measurement of set and setting in future studies to refine dose-response profiles.