Flashback phenomena after administration of LSD and psilocybin in controlled studies with healthy participants

Earlier research has described episodic recurrences of drug-like experiences—commonly called flashbacks—after the acute effects of hallucinogens such as LSD and psilocybin. These phenomena are typically transient and most often visual, but reported incidence estimates from naturalistic and retrospective studies vary widely and the aetiology remains unclear. When such persistent and distressing perceptual disturbances cause clinically significant impairment they are classed as hallucinogen-persisting perception disorder (HPPD) in the DSM-V; however, most existing evidence on flashbacks and HPPD comes from case reports and uncontrolled naturalistic data, limiting reliable incidence estimates for controlled research settings. Müller and colleagues set out to provide descriptive data on reoccurring drug-like experiences following controlled administration of LSD and psilocybin. Using pooled data from six double-blind, placebo-controlled, randomized crossover trials in healthy volunteers, the study aimed to quantify the frequency, phenomenology, timing, severity, and functional impact of flashback phenomena, and to assess whether any participants met DSM-V criteria for HPPD. Given renewed clinical interest in these compounds, the authors framed this as an opportunity to characterise these events in modern, well-controlled experimental contexts.

The analysis pooled data from six clinical trials conducted in Basel, Switzerland, completed by April 2021. All included trials were double-blind, placebo-controlled, randomised, crossover studies in healthy adult participants; ethical approval and informed consent procedures were reported. The combined sample comprised 142 participants who received at least one dose of LSD or psilocybin during the included studies. LSD was administered a total of 213 times across participants in doses ranging from 0.025 to 0.2 mg (given as capsules or alcohol solution), and psilocybin was administered 103 times in doses of 15 to 30 mg (capsules). Individual study designs varied: two small studies tested single LSD doses (0.1 mg and 0.2 mg), one compared 0.1 mg LSD with d-amphetamine and MDMA, one examined multiple LSD doses and LSD with ketanserin, one studied 25 mg psilocybin after escitalopram or placebo pretreatment, and one compared LSD and psilocybin doses (NCT03604744). Three participants in the last study dropped out after LSD and did not receive psilocybin. Eligibility criteria across studies required adults (age ranges reported per study), German language proficiency, informed consent, contraception when applicable, and abstention from illicit substances during the study; exclusion criteria included seizure, cardiac or neurological disorders, current or past psychotic or major psychiatric disorder, family history of psychosis, substantial prior illicit drug use (>10 occasions, except THC) or recent use within 2 months, pregnancy, and interfering medications. Some trials applied additional, study-specific criteria (for example, fMRI-related restrictions). Adverse events were recorded at each study session and an end-of-study visit asked all participants about any flashback phenomena occurring during the study period; the mean interval from last session to end-of-study visit was 39.8 days (SD 37.2). Participants who reported flashbacks at the end-of-study visit were contacted by e-mail for a follow-up in May 2021 to capture subsequent occurrences, characteristics, triggers, and impairment (mean follow-up interval 31.2 months after end-of-study visit, SD 28.6). Finally, the investigators evaluated reported events against DSM-V criteria for HPPD and conducted a simulation using R (rbinom) to estimate the probability of observing zero HPPD cases in their sample under assumed population incidences of 1%–5%.

The final analysed sample included 142 participants (74 female, 68 male) with mean age 31.7 years (SD 9.0). Prior hallucinogen use was reported by 30.3% of participants (mean 2.2 occasions), MDMA by 40.1% (mean 2.6 occasions), other stimulants by 34.5% (mean 2.0 occasions), and cannabis by 78.9%; use of dissociatives, opioids or sedatives was rare. At the end-of-study visit (mean 39.8 days after the last session), 13 of 142 subjects (9.2%) reported reoccurring drug-like experiences. By substance exposure, seven affected subjects had received LSD (7.8% of LSD recipients), two had received psilocybin (8.3% of psilocybin recipients), and four had received both substances (14.3% of those who received both). Most reports were visual (11 of 13, 84.6%); three of those visual reports included accompanying auditory, cognitive, or feelings of disintegration. Two subjects (15.4% of the 13) reported exclusively emotional alterations. Temporal and phenomenological features were largely brief and transient: 69.2% of episodes lasted seconds, 23.1% minutes, and a single case reported alterations that persisted for hours on one day. In 53.8% of cases the phenomenon occurred only once, while two participants experienced recurrent episodes more than five times (one had about 20 visual flashbacks within ~24 hours after administration; the other reported about 30 visual flashbacks over a 7-month period). Over half of participants reported that events occurred during relaxing periods or around sleep onset/offset. Ten of the 13 reported the phenomena as neutral or pleasant (76.9%), two described them as unpleasant (15.4%), and one case lacked sufficient documentation. Overall, 1.4% of the total sample (2 of 142) reported distressing experiences related to flashbacks; neither of these two participants reported functional impairment in daily life and both episodes resolved spontaneously. Follow-up contact (mean 31.2 months after end-of-study visit) achieved an approximate 92% response rate. One participant (the individual who had reported ~30 flashbacks) described recurrence after the end-of-study visit—approximately 30 brief, benign visual flashbacks over 7 months that had ceased about 10 months before follow-up. All other respondents (12 of the 13 initial reporters; 92.3%) reported no further flashbacks after the end-of-study visit. No participant met DSM-V criteria for HPPD at any point. Simulation results showed the probability of observing zero HPPD cases in a sample of 142 was 24.0% if the true incidence were 1%, 5.7% for 2%, 1.3% for 3%, 0.3% for 4%, and 0.1% for 5%, indicating that observing no cases would be unlikely if true incidence exceeded about 3%.

Müller and colleagues interpret their pooled data as indicating that transient reoccurring drug-like experiences after LSD or psilocybin are relatively common in controlled trials of healthy volunteers, affecting roughly 9% of participants in their sample. Most of these phenomena were visual, very brief, and experienced as neutral to pleasant; severe or persistent events were rare. The authors note that some reported events followed administration of non-hallucinogenic compounds (escitalopram, MDMA, d-amphetamine) in the same trials, which could have influenced occurrence or perception of flashbacks. Compared with higher incidences reported in many naturalistic studies (15%–75%), the observed frequency here was lower; the authors suggest methodological differences, participant selection, and dose/exposure patterns as possible explanations. Incidence was similar for LSD and psilocybin in this pooled sample (7.8% and 8.3% respectively), but the psilocybin subgroup was small and the authors caution interpretation. They also reference a pooled MDMA analysis from their group with comparable reporting rates, proposing that participants in psychoactive substance trials may commonly report flashback-like experiences regardless of specific drug exposure. The authors discuss potential mechanisms without asserting a single cause: hypotheses include substance-induced disinhibition of visual processes, strong encoding of unusual experiences leading to involuntary memory intrusions, and heightened participant awareness of borderline-normal perceptual phenomena. The preponderance of reports around sleep or relaxation may reflect increased awareness or serve as situational cues. Importantly, none of the reported cases met DSM-V criteria for HPPD. The investigators ran simulations to show that observing zero HPPD cases in their sample would be unlikely if true incidence exceeded about 3%, suggesting either a lower true incidence or that HPPD is less likely in carefully screened clinical-trial populations. Limitations acknowledged by the authors include the modest sample size for detecting rare events like HPPD, heterogeneity of doses and substances across studies, potential selection bias favouring participants with positive attitudes toward hallucinogens, substantial prior cannabis use among participants, the fact that only participants who reported flashbacks at end-of-study were also followed up (so late-onset events in others could have been missed), variable follow-up intervals, and exclusion of individuals with psychiatric disorders which limits generalisability to patient populations. The authors therefore constrain their conclusions to healthy volunteer research settings and refrain from extrapolating to clinical or recreational contexts.

In this pooled analysis of six controlled trials in healthy volunteers, drug-like reoccurring perceptual experiences after LSD and psilocybin occurred in about 9% of participants but were predominantly transient, brief, and non‑distressing. Distressing flashback-related experiences were reported by 1.4% of the sample and resolved without treatment; no participant met DSM‑V criteria for HPPD. The investigators conclude that, within the context of well-controlled studies in screened healthy participants, flashback phenomena do not represent a clinically relevant problem.