Elife
March 2, 2021
Balázs Szigeti, Laura Kärtner, Allan Blemings et al.
192 citations
A self-blinding citizen science trial with 191 participants tested whether microdosing psychedelics produces psychological benefits beyond a placebo. All psychological outcomes improved from baseline to after the four-week dose period in the microdose group, but the placebo group also improved, and no significant between-group differences were observed. Small, significant differences in acute measures (emotional state, drug intensity, mood, energy, creativity) and post-acute anxiety appeared, but these could be explained by participants breaking blind. The findings suggest that anecdotal benefits of microdosing can be explained by the placebo effect.
Psychological medicine
June 1, 2024
Balázs Szigeti, Brandon Weiss, Fernando E Rosas et al.
74 citations
In a double-blind trial comparing escitalopram and COMP360 psilocybin for major depressive disorder, patients held higher expectations for psilocybin than for escitalopram. Higher pre-trial expectancy for escitalopram predicted better outcomes with escitalopram, but expectancy for psilocybin did not predict response to psilocybin. Pre-treatment trait suggestibility was linked to therapeutic response in the psilocybin arm but not the escitalopram arm. These findings suggest that psychedelic therapy may be less influenced by expectancy biases than previously thought, and that highly suggestible individuals may be especially responsive to psilocybin treatment.
Biological psychiatry. Cognitive neuroscience and neuroimaging
May 1, 2024
Balázs Szigeti, Boris D Heifets
62 citations
Clinical trials of psychedelics such as psilocybin, LSD, and DMT have challenged how nondrug factors like participant expectations are measured and controlled in mental health research. Higher doses of these psychoactive substances make it harder to conceal treatment conditions in double-blind, placebo-controlled designs. Growing public enthusiasm for psychedelic therapy raises questions about whether trial results are biased by positive expectancy. This review covers key concepts of expectancy and its measurement, examines expectancy effects reported in modern microdose and macrodose trials, and considers expectancy as a physiological process that can be independent of or interact with drug effects. Expectancy can be harnessed to improve outcomes and managed to enhance trial rigor.
Sci Rep
July 26, 2023
Balázs Szigeti, David Nutt, Robin Carhart-Harris et al.
33 citations
Blinding in medical trials aims to evenly distribute expectancy effects between treatment groups, but it often fails. Using computational modeling, this work shows that weak blinding combined with positive treatment expectancy creates an 'activated expectancy bias' (AEB), which can inflate treatment effect estimates and produce false positive results. The authors introduce the Correct Guess Rate Curve (CGRC) to estimate outcomes of a perfectly blinded trial from imperfectly blinded data. Re-analyzing a self-blinding psychedelic microdose trial dataset, they find that observed placebo-microdose differences are susceptible to AEB and may be false positives, suggesting microdosing acts as an active placebo. The findings highlight the distinction between trials with a placebo control group and genuinely placebo-controlled trials.
Nature medicine
June 3, 2025
Chloé Pronovost-Morgan, Kyle T. Greenway, Leor Roseman et al.
26 citations
A Delphi consensus study with 89 experts from 17 countries identified 30 extra-pharmacological variables that are important or very important for reporting the setting in psychedelic clinical trials. These variables, forming the ReSPCT guidelines, cover physical environment, dosing session procedure, therapeutic framework and protocol, and subjective experiences. The findings reveal significant ambiguities in current conceptualizations of set and setting. The guidelines provide a new standard for designing and documenting contextual factors in psychedelic research.
Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology
April 9, 2025
Ellen R Bradley, Kimberly Sakai, Gisele Fernandes-Osterhold et al.
23 citations
In an open-label pilot trial, 12 people with mild to moderate Parkinson's disease plus depression or anxiety received psilocybin (10 mg then 25 mg) with psychotherapy. No serious adverse events occurred, and no worsening of Parkinson's symptoms was observed. Non-motor and motor symptoms improved, and gains in some cognitive domains were sustained one month later. Depression and anxiety scores improved to a clinically meaningful degree and remained improved three months after dosing. These first results in any neurodegenerative disease suggest psilocybin therapy for Parkinson's disease warrants further study.
Journal of Psychopharmacology
May 7, 2018
Balázs Szigeti, Adam Winstock, David Erritzøe et al.
17 citations
Neuroimaging studies of people who regularly use ecstasy have focused on unusually heavy users, not typical users. A systematic review of 10 imaging studies that measured serotonin transporter levels in recreational ecstasy users found that the average number of pills taken per session and sessions per month corresponded to the top 5-10% of users in the Global Drug Survey, a large international self-report database. Imaging participants consumed, on average, 720% more pills per year than the Global Drug Survey participants. This suggests that conclusions from these brain imaging studies likely overestimate the extent of serotonergic alterations experienced by the majority of people who use ecstasy.
British Journal of Clinical Pharmacology
July 17, 2023
Balázs Szigeti, Lawrence D. Phillips, David Nutt
13 citations
Randomized controlled trials (RCTs) are often considered the gold standard in medical research, but they have limitations including reliance on null hypothesis significance testing and poor generalizability. Bayesian analysis of real-world evidence (RWE) offers a complementary approach. In a case series of 20 children with epilepsy treated with medical cannabis, all experienced reduced seizures; Bayesian analysis with a flat prior gives a 95% probability that the next patient will improve (95% credible interval 87%–100%). For treatment-resistant depression treated with psilocybin, the probability of a favorable response ranges from 62% (QIDS-16) to 82% (MADRS). These analyses require fewer patients than traditional RCTs and provide directly actionable probabilities for clinicians and patients.
JAMA Psychiatry
March 18, 2026
Zachary J. Williams, Hannah Barnett, Balázs Szigeti
10 citations
In trials for major depression, psychedelic-assisted therapy (PAT) was no more effective than open-label traditional antidepressants (TADs) such as SSRIs and SNRIs. Across 8 PAT trials (249 patients) and 16 open-label TAD trials (7921 patients), the estimated difference in symptom improvement was 0.3 favoring TADs, a statistically non-significant result. Open-label TADs outperformed blinded TAD treatment, but PAT showed no such difference, confirming that PAT trials are effectively always open label. These findings challenge overly optimistic claims about PAT and underscore the importance of maintaining blinding integrity in clinical research.
Alcohol and alcoholism (Oxford, Oxfordshire)
May 14, 2025
Hannah Thurgur, Ben Sessa, Laurie Higbed et al.
3 citations
In an open-label feasibility study, 14 adults with alcohol use disorder who had recently completed detoxification underwent an eight-week course of ten psychotherapy sessions, including two sessions with MDMA. Bayesian analysis estimated a 55%–63% probability of a two-level reduction in World Health Organization drinking risk three months after treatment. Preliminary findings also indicated reductions in alcohol craving and improvements in sleep and aspects of psychosocial functioning at the three-month follow-up compared to baseline. The results provide initial insights into MDMA-assisted psychotherapy's potential to improve quality of life and well-being beyond reducing drinking.
PsyArXiv
May 24, 2022
Balázs Szigeti, David Nutt, Robin Carhart-Harris et al.
3 citations
preprint
Blinding in medical trials is meant to distribute expectancy effects evenly across treatment arms, but it often fails. Using computational modeling, this work shows that weak blinding combined with positive treatment expectancy can create an 'activated expectancy bias' (AEB), which inflates treatment effect estimates and can produce false positive findings. To address this, the authors introduce the Correct Guess Rate Curve (CGRC), a statistical tool that estimates what a perfectly blinded trial would have found from imperfectly blinded data. Re-analyzing a self-blinding psychedelic microdose trial, they find that observed placebo-microdose differences are susceptible to AEB and may be false positives, suggesting microdosing acts as an active placebo. The results underscore the difference between trials with a placebo control group and genuinely placebo-controlled trials.
October 19, 2022
Stefan Baumann, Robin Carhart-Harris, David Nutt et al.
2 citations
preprint
In a placebo-controlled citizen science trial with 240 participants, microdosing tolerance was assessed by tracking whether correct guesses of receiving a microdose decreased with more doses taken. Correct guess probability declined overall, indicating tolerance developed. This tolerance was specific to LSD and LSD-analogue microdoses, not psilocybin microdoses. The findings suggest that microdosers may need to periodically suspend their routine to avoid tolerance and that psilocybin may be better suited for long-term protocols.
December 11, 2020
Balázs Szigeti, Laura Kärtner, Allan Blemings et al.
1 citation
A self-blinding citizen science study tested whether psychedelic microdosing improves well-being and cognition beyond placebo. 191 participants who already planned to microdose were randomly assigned to receive four weeks of microdoses, placebos, or a mix. All psychological outcomes—including well-being, mindfulness, and life satisfaction—improved from baseline in the microdose group, but the placebo group also improved, and no significant differences emerged between groups. Small acute differences in mood, energy, and creativity were observed, but these could be explained by participants correctly guessing whether they took a microdose. The findings suggest that the anecdotal benefits of microdosing are likely due to the placebo effect.
Balázs Szigeti
1 citation
preprint
The FDA rejected MDMA-assisted therapy for PTSD partly due to concerns about functional unblinding, where participants or clinicians guess treatment assignment because of the drug's noticeable effects, potentially biasing trial results. The authors define unmasking bias and calculate its magnitude for two other drugs, ketamine and escitalopram, using published data. They find that unmasking bias for these two drugs exceeds the treatment-versus-control effect size observed in MDMA trials. This indicates that the effect sizes for MDMA-assisted therapy are not too large to be explained by unmasking bias, though the findings do not prove that the therapy's effects are entirely or partially due to this bias.
July 7, 2025
Balázs Szigeti
preprint
Psilocybin therapy under controlled conditions appears safe and potentially effective for reducing depressive symptoms and improving quality of life in individuals with bipolar II disorder experiencing moderate-to-severe depression. In an open-label pilot trial, 14 participants received 10 mg of psilocybin, followed by 25 mg if symptoms persisted, alongside psychotherapy. No serious adverse events occurred; common effects included mild-to-moderate anxiety, nausea, and headache. Three participants experienced suicidal ideation or hypomania, which resolved with support. Depression scores improved at all timepoints, with a 21-day reduction of 12.7 points after 10 mg and 18.6 after 25 mg. Quality of life also improved at 90 days. Feared outcomes like mania, psychosis, and suicidality were not elevated relative to other clinical populations treated with psilocybin.
Balázs Szigeti, Ellen Bradley, Joshua Woolley
preprint
Unmasking bias—where participants or researchers can guess who received a treatment due to its noticeable effects—may account for the reported benefits of MDMA-assisted therapy for PTSD. Analyzing data from trials of ketamine and escitalopram, the authors found that the magnitude of unmasking bias is larger than the treatment-versus-control effect size observed for MDMA. This indicates that MDMA-AT's effect sizes are not too large to be explained by unmasking alone, though the findings do not prove that the effects are entirely or partially due to this bias.
The lancet. Psychiatry
May 1, 2026
Joseph J Taylor, Balázs Szigeti, Noah D Silverberg et al.
Placebo effects remain a major paradox in medical research, with more data available from placebo-controlled trials than on any treatment, yet little deep analysis of how they are measured, appraised, and interpreted. This review shifts focus from clinical practice to clinical trials, examining established and emerging approaches for managing placebo effects in randomized controlled trials. It highlights three key challenges in contemporary psychiatric research: blinding and expectancy in psychedelic trials, large placebo responses in interventional psychiatry trials (device or procedure-based treatments), and the implications of overlapping neurobiological mechanisms between placebo effects and psychiatric treatments.