Safety and Efficacy of Repeated Low-Dose LSD for ADHD Treatment in Adults A Randomized Clinical Trial

Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental condition in adults, characterised by inattention and/or hyperactivity–impulsivity and associated with substantial functional impairment and frequent psychiatric comorbidity. Standard pharmacotherapies—stimulants (eg, methylphenidate, amphetamines) and nonstimulants (eg, atomoxetine)—are effective for many patients but leave approximately 20% to 40% with an inadequate response and face problems with adverse effects and long‑term adherence. In parallel, microdosing of psychedelics such as LSD has gained popular and scientific interest, with surveys and naturalistic studies reporting perceived benefits for mood, cognition, and self‑treatment of disorders including ADHD, but randomised clinical evidence in patient populations is lacking. Mueller and colleagues therefore undertook a multicentre, double‑blind, placebo‑controlled Phase 2A randomised clinical trial to test whether repeated low doses of LSD reduce ADHD symptoms in adults with moderate to severe ADHD. The intervention was a 20 μg LSD dose administered twice weekly for 6 weeks (12 doses total); the primary aim was to compare change in ADHD symptoms (observer‑rated AISRS) from baseline to week 6 versus placebo, while also assessing safety, pharmacokinetics, acute subjective effects, and blinding.

This was a 6‑week, multicentre, double‑blind, parallel‑group Phase 2A randomised clinical trial conducted at two centres (University Hospital Basel, Switzerland, and Maastricht University, the Netherlands). Eligible participants were adults aged 18–65 years with a preexisting DSM‑IV/V diagnosis of ADHD and current moderate to severe symptoms, operationalised as an Adult Investigator Symptom Rating Scale (AISRS) score ≥26 and a Clinical Global Impression‑Severity (CGI‑S) score ≥4. Key exclusions included past or present psychotic disorders in participants or first‑degree relatives, current substance use disorder, and other psychiatric or somatic conditions likely to require hospitalisation or interacting treatments. The extracted text does not provide the full list of prohibited medications in this section but refers to supplements for details. Participants were randomised 1:1 to receive either LSD or placebo, with stratification by site and computer‑generated block randomisation. The sponsor conducted randomisation and provided allocation to the medication manufacturer; study staff and participants remained blinded until study completion. Study drug was manufactured under GMP as a 1 mL drinking solution containing 29 μg LSD tartrate (equivalent to 20 μg LSD base) in 20% v/v alcohol; placebo was the same alcohol solution without active drug. All doses were administered on‑site under supervision twice weekly for 6 weeks (12 doses). After the first dose participants remained on site for 6 hours for monitoring and pharmacokinetic sampling; subsequent dosing visits permitted immediate departure after administration. Participants were asked to guess treatment allocation after the first and last doses to assess blinding. The primary efficacy outcome was the least squares mean (LSM) change in AISRS score from baseline to week 6, analysed using a Mixed Model for Repeated Measures (MMRM) with fixed effects for treatment, visit, treatment-by‑visit interaction and covariates baseline AISRS, sex, and age; an unstructured variance–covariance matrix and the Satterthwaite method for degrees of freedom were used. The trial planned to enrol 52 participants (n = 26 per group) to provide 80% power to detect an effect size of 0.6 with a one‑sided α of .10; analyses followed an intention‑to‑treat principle including all randomised participants who received at least one dose. Secondary outcomes used the same MMRM and imputation procedures but applied two‑sided testing with α = .05; multiple testing correction was reserved for any significant post hoc findings. Missing data were handled by multiple imputation (details referenced to supplements). Safety assessments included pre‑dose pregnancy tests, vital signs, C‑SSRS for suicidal ideation, adverse event monitoring, routine blood tests every 2 weeks, and ECGs at screening and 2 hours after the first and last doses. Pharmacokinetic sampling after the first dose occurred at 0, 0.5, 1, 2, 3, 4, and 6 hours, and acute subjective effects were measured with Visual Analog Scales (VAS), the 5D‑ASC and the MEQ30.

Fifty‑three participants were randomised (LSD n = 27, placebo n = 26). Mean (SD) age was 37 (12) years and 22 participants (42%) were female. Baseline mean (SD) AISRS scores were 37 (6) in the LSD group and 36 (5) in the placebo group. As the primary endpoint, the LSD group showed an LSM AISRS improvement of −7.1 points (95% CI, −10.1 to −4.0) and the placebo group an LSM improvement of −8.9 points (95% CI, −12.0 to −5.8). The between‑group LSM difference in change was reported as 1.8 points with a 95% CI that is not clearly legible in the extracted text (the upper bound could not be reliably determined from the extraction), and the difference did not reach statistical significance. Sensitivity analyses, including a missing‑at‑random imputation approach and stratification by disease severity, likewise did not produce significant between‑group differences. Secondary efficacy analyses also showed no significant differences between groups; both arms exhibited statistically significant within‑group improvements over time across self‑rated and observer‑rated measures, which persisted up to 3 weeks after dosing. Blinding checks indicated that after the last dose a high proportion of respondents believed they had received LSD: the extracted text reports 37 of 46 participants (80%) guessed LSD after the last dose, with 21 of 22 in the LSD group and 16 of 24 in the placebo group; overall, 29 participants (63%) were reported to have correctly guessed their allocation (21 of 22 in LSD and 8 of 24 in placebo). The extraction does not clearly explain why the denominator for the blinding question differs from the full randomised sample. Participants who believed they had received LSD showed nominally larger LSM reductions than those who believed they received placebo. Acute subjective effects were substantially greater with LSD than placebo. The extracted text reports a higher proportion reporting any drug effect in the LSD group versus placebo (statistical comparison P < .001), higher mean 5D‑ASC total scores (13% [SD 11] vs 4.9% [SD 7.8], P = .005) and higher MEQ30 total scores (mean [SD] 22 [20] vs 10 [16], P = .02). Several VAS items and subscales of the 5D‑ASC and MEQ30 also differed significantly in favour of LSD. Pharmacokinetic data indicated rapid absorption with maximal plasma concentrations reached after approximately 1.3 hours; detailed PK parameters were reported in supplements. Safety outcomes showed more adverse events in the LSD group (124 events) than in placebo (64 events). No serious adverse events or deaths occurred. The five most common treatment‑related adverse events were headache, nausea, fatigue, insomnia, and visual alterations; these were more frequent in the LSD arm (23 treatment‑related events) than in placebo (8 events). Two participants in the LSD arm discontinued treatment because of strong or impairing acute effects (one after the first dose, one after five doses). One participant in the placebo group was excluded after a positive pregnancy test. There were no new cases of suicidal ideation during dosing. ECGs, blood pressure, heart rate and routine laboratory values showed no clinically relevant differences between groups that were attributed to the study drug according to the extracted text.

Mueller and colleagues report that this is the first double‑blind, placebo‑controlled randomised clinical trial testing repeated low‑dose LSD (20 μg twice weekly for 6 weeks) in adults with ADHD. The main finding is that the trial did not meet its predefined primary endpoint: LSD did not reduce ADHD symptoms more than placebo on the AISRS or on secondary measures. Both treatment arms showed clinically meaningful within‑group improvements, but no between‑group advantage for LSD was observed. The investigators consider LSD to have been generally well tolerated in the outpatient setting, with treatment‑related adverse reactions mostly mild. The authors note that the chosen 20 μg dose sits at the upper end of commonly described microdosing ranges and may be better characterised as a low dose; it produced measurable acute subjective effects that were substantially greater than placebo. They acknowledge considerable interindividual variability in acute responses and report two discontinuations in the LSD arm due to strong acute effects. While the investigators argue that the selected dose was unlikely to be too low to detect an effect, they do not exclude the possibility that other doses, dosing intervals, or titration strategies could yield different results. A prominent theme in the discussion is the role of expectancy and placebo response. The trial design—with frequent supervised dosing visits, extensive assessments and media attention around psychedelics—may have heightened participant expectations and contributed to the observed improvements in both groups. The majority of participants reported believing they had received LSD at study end, and those who believed so tended to show larger symptom reductions. The authors compare these findings with prior placebo‑susceptible effects reported in ADHD trials and with naturalistic microdosing surveys, concluding that observed benefits in uncontrolled settings may reflect expectancy rather than specific pharmacological effects. Strengths cited include the randomised placebo‑controlled design, supervised on‑site dosing, comprehensive safety and pharmacokinetic monitoring, and use of both observer‑rated and self‑rated efficacy measures. Limitations acknowledged by the investigators are that the trial was powered to detect a relatively large effect (so smaller effects could have been missed), one site contributed the vast majority of participants (95%), expectancy was not systematically quantified, and only a fixed twice‑weekly dosing regimen at a single dose was tested. They also note that other ADHD‑relevant domains (eg, emotion regulation) were not comprehensively assessed. The authors conclude that randomised placebo‑controlled trials are essential to validly assess low‑dose psychedelic interventions, which appear prone to expectancy and placebo effects.

In this Phase 2A randomised clinical trial, repeated low‑dose LSD (20 μg twice weekly for 6 weeks) was safe and generally well tolerated in an outpatient adult ADHD population but did not demonstrate efficacy versus placebo in reducing ADHD symptoms. The findings underscore the importance of placebo‑controlled designs in evaluating microdosing claims and suggest that observed benefits in uncontrolled settings may be driven by expectancy effects rather than pharmacological action.