A randomized, double-blind, placebo-controlled single-ascending-dose study to identify a non-hallucinogenic dose of psilocybin in healthy adults

Psilocybin is an orally active prodrug that is converted to psilocin, which acts mainly through serotonergic mechanisms and, at conventional doses, produces intense perceptual and cognitive changes. Earlier research has suggested therapeutic potential for depression, anxiety and other psychiatric conditions, but most clinical studies have used moderate to high doses that require prolonged in-clinic supervision because of the altered states they produce. The paper notes that this creates cost, scalability and feasibility problems, and that it remains unresolved whether any potentially beneficial pharmacology of psilocybin can be separated from its hallucinogenic effects. Levy-Cooperman and colleagues therefore set out to characterise the safety, pharmacokinetics and pharmacodynamics of single low doses of psilocybin in healthy adults, with the specific aim of identifying a dose that is perceptible and biologically active but not hallucinogenic and not cognitively impairing. The broader motivation was to define a candidate low-dose window that could support later outpatient studies, including trials in generalised anxiety disorder and other neuropsychiatric disorders.

This was a Phase I, randomised, double-blind, placebo-controlled, single-ascending-dose study conducted at a single clinical research site in Toronto. Healthy adults aged 18 to 55 years were screened and enrolled after written informed consent. Eligibility required a body mass index of 18.0 to 34.0 kg/m2, minimum weight of 50 kg, protocol-defined blood pressure limits and non-smoking status for at least 6 months. The main exclusions were clinically significant medical or psychiatric illness, personal or family history of psychotic or bipolar disorders, recent or current major depression or anxiety disorders, suicidal ideation or behaviour, and recent or extensive non-therapeutic use of perception-altering substances. Participants were assigned within cohorts of eight in a 3:1 ratio to psilocybin or matching placebo, with six active and two placebo participants per cohort and at least one participant of each sex in the active arm. Randomisation was handled by an unblinded statistician and pharmacist, while participants, investigators, site staff and the bioanalytical laboratory remained blinded. To reduce expectancy effects, the consent process used active-drug masking language, telling participants they might receive placebo or one of several psychoactive agents, although only psilocybin or placebo were actually administered. Psilocybin was given orally as an aqueous solution with 0.2% sucralose; placebo contained the sucralose vehicle only. The dose-escalation sequence was 0.5, 1.0, 1.5, 2.5, 3.5 and 4.0 mg, with the planned 2.0 and 3.0 mg levels not used. A Drug Safety Review Committee reviewed blinded safety and pharmacodynamic data after each cohort and decided whether to escalate. The protocol allowed escalation up to 5 mg and defined the threshold dose as the dose immediately before escalation stopped. Safety, pharmacokinetic and pharmacodynamic assessments were collected across three visits: screening, a 3-day inpatient treatment phase and an outpatient follow-up 7 ± 2 days later. Psilocin concentrations were sampled repeatedly up to 24 hours after dosing and analysed with non-compartmental methods for Cmax, Tmax, AUC and terminal half-life, with dose proportionality assessed using a power model. Pharmacodynamic measures included visual analogue scales for subjective effects, the 5-Dimensional Altered States of Consciousness scale, state and trait anxiety, CANTAB tasks for attention, working memory and impulse control, and pupillometry as an objective physiological readout. Safety assessments included adverse events, vital signs, ECGs, laboratory tests, physical examination and suicidality screening. The statistical analyses were mainly descriptive; no inferential between-group testing was prespecified for the primary programme-level analyses, although several exploratory correlations and tests were applied to subjective, cognitive, anxiety and pupil measures.

Of 144 people screened, 56 were randomised and dosed: 42 received psilocybin and 14 received placebo. The psilocybin groups comprised six participants each at 0.5, 1.0, 1.5, 2.5 and 3.5 mg, and 12 participants at 4.0 mg because that dose was repeated in an additional cohort. All participants completed the study, and all were included in the safety and pharmacodynamic analyses. Psilocin appeared rapidly in plasma after oral dosing, was usually detectable by 0.25 to 0.5 hours, and reached peak levels at about 1 hour. Peak concentrations rose in a dose-ordered fashion from 0.5 mg to 4.0 mg, and most participants fell below the assay limit by 10 to 18 hours. The terminal half-life was short, roughly 1.6 to 2.5 hours. Overall exposure increased with dose in a dose-proportional manner, while Cmax rose slightly less than proportionally. Subjective drug effects increased with dose. On the Any Drug Effects scale and the Bowdle “I felt high” item, peak ratings rose most clearly at doses of 2.5 mg and above, with the strongest responses at 3.5 mg and 4.0 mg. By contrast, hallucination ratings and items reflecting perceptual distortion stayed low and did not show dose-related increases; the highest mean hallucination score actually occurred in the placebo group. Scores on the 5D-ASC remained low across all doses, indicating minimal alteration of consciousness, and the pattern was dominated by Reduction of Vigilance, which the authors interpret mainly as mild drowsiness rather than psychedelic change. When the authors analysed peak subjective effects across the full dose range, both Any Drug Effects and Bowdle “High” ratings showed modest monotonic dose relationships. The reported Spearman correlations were 0.34 for Any Drug Effects and 0.30 for “High”, with the clearest rise between 2.5 mg and 3.5 mg. However, the time-averaged effect exposure over 24 hours showed only weak, non-significant dose relationships. The fitted Emax model did not plateau within the administered doses, so the authors treat it as a descriptive monotonic dose-response rather than a firm estimate of potency. Cognitive performance was preserved. Attention, vigilance, working memory and impulse control showed no dose-dependent impairment across the tested range. Some changes appeared to reflect practice effects rather than drug effects, and a few small dose-related signals clustered around the 2-hour point near peak exposure, but these were not consistent with meaningful cognitive disruption. By 4 hours, performance had largely returned towards baseline. Psilocybin produced a clear dose-dependent pupil dilation, beginning within 0.5 to 1 hour after dosing and peaking most strongly at 2.5 mg, 3.5 mg and 4.0 mg. Peak dilation within the first 4 hours was correlated with dose, and psilocybin doses of 2.5 mg and above produced greater peak dilation than placebo. This objective autonomic signal was interpreted as evidence of physiological engagement in the absence of hallucinations or cognitive harm. State anxiety did not increase with psilocybin and was slightly reduced at the higher doses. Baseline anxiety scores were in the mild-to-moderate range for this healthy-volunteer sample. Change-from-baseline scores at 6 hours were small, but several individual items moved in an anxiolytic direction: participants reported feeling more content and secure, and less frightened, as dose increased. Trait anxiety was unchanged. The masking procedure appeared reasonably effective, especially at lower doses. Most participants said they did not know which treatment they had received, and only three of 14 placebo recipients believed they had received an active psychoactive drug. At higher psilocybin doses, more participants guessed they had received an active drug. Safety was favourable. There were no deaths, no serious adverse events and no discontinuations due to treatment-emergent adverse events. All adverse events were mild and judged related to study drug. Somnolence was the most frequent event and occurred across both psilocybin and placebo conditions. Euphoric mood, recorded verbatim as “feeling high”, was reported only at 3.5 mg and 4.0 mg. One participant at 4.0 mg had an event coded as auditory hallucination, but the source description suggested “auditory disturbances” and thought disturbance rather than a clear hallucination. Laboratory tests, vital signs and ECGs remained within normal limits, and there was no suicidal ideation or behaviour.

The authors argue that this Phase I study shows single oral doses of psilocybin from 0.5 mg to 4.0 mg were safe and well tolerated in healthy adults. They emphasise the absence of serious adverse events, discontinuations and clinically significant changes in laboratory values, vital signs, ECGs or suicidality. In their view, the data also show that low oral doses in the 2.5 mg to 4.0 mg range produce measurable pharmacodynamic effects, including objective pupil dilation and small changes in state anxiety, but without the hallucinatory or cognitive effects that characterise conventional doses. Levy-Cooperman and colleagues interpret the subjective findings as evidence of a monotonic, dose-related increase in perceptible drug effect that becomes distinguishable from placebo at 2.5 mg and above, while remaining largely separate from overt perceptual distortion. They stress that the increase in “high” ratings was not accompanied by meaningful hallucinations or altered states, and therefore conclude that the perceptible pharmacology of psilocybin can be dissociated from its hallucinogenic effects at low doses. They present this as an important translational finding because it addresses whether psilocybin can be biologically active without inducing major alterations in consciousness. The authors also highlight the preserved cognitive performance as a major safety outcome for any outpatient or at-home dosing model. They report no dose-dependent decrements in attention, working memory, reaction time or impulse control, and interpret the small changes that did occur as mostly practice-related and also present in placebo. They suggest that the low doses engaged cognitive-affective systems without impairing function. The pupillometry results are presented as an objective biomarker of autonomic engagement. The authors note that pupil dilation tracked dose and plasma psilocin concentrations, and they argue that this supports the idea that the doses studied were pharmacologically active even though they did not produce overt impairment. They view the slight reduction in state anxiety and the dose-related rises in feelings of contentment and security as encouraging, but they are cautious about over-interpreting this signal. They explicitly note that the healthy-volunteer sample had relatively low baseline anxiety, so the findings may not predict treatment effects in a generalised anxiety disorder population. In positioning the study against earlier work, the authors say it is consistent with prior reports of favourable safety at sub-hallucinogenic doses and with broader naturalistic interest in microdosing. They also refer to emerging clinical observations in other populations where daily low-dose psilocybin was reportedly safe and associated with symptom improvement. However, they stress that their trial provides controlled pharmacological evidence that earlier survey and open-label findings could not. The main limitations they acknowledge are that this was a small, exploratory first-in-program study in healthy volunteers, with descriptive analyses and limited sensitivity to subtle effects. Cognitive testing did not coincide with the approximate 1-hour peak effect, and healthy participants were near ceiling on several tasks. They also note that the fitted Emax model did not reach a plateau and therefore does not define a true potency estimate. Because the study did not include patients with anxiety disorders, they say additional trials are needed to test repeated dosing, safety and efficacy in clinical populations. The authors conclude that the study supports a low-dose window in which psilocybin is perceptibly active and physiologically engaging but not meaningfully hallucinogenic or cognitively impairing, and they present this as a rationale for advancing to outpatient Phase 2 studies.