Abnormal visual experiences in individuals with histories of hallucinogen use: A web-based questionnaire

The study received institutional review board approval and recruited participants via the drug information website Erowid over an 80-day period. A link inviting hallucinogen-experienced visitors to complete a "visual experiences survey" was posted; respondents who clicked saw an information sheet that did not explicitly state a focus on chronic visual changes. Anonymity was preserved through encrypted connections and storage of encrypted IP addresses. The questionnaire collected self-reported drug-use histories, past and current psychiatric and neurological diagnoses, and detailed information about visual experiences. Drug-use questions asked participants to estimate the number of exposures to 14 specific psychoactive substances drawn from six pharmacological classes, including classical serotonergic hallucinogens (examples: LSD, psilocybin, DMT, LSA, and several phenethylamines/tryptamines), NMDA antagonists (ketamine and high-dose dextromethorphan), MDMA, anticholinergic-containing Datura plants, cannabis, and Salvia divinorum. A fictional drug, "kaopectamine," was included as a validity check; respondents reporting its use were excluded. Participants were explicitly queried about diagnoses sometimes associated with visual abnormalities (seizure disorders, migraine, schizophrenia or other psychotic disorders, visual impairments) and completed four Likert-scale items intended to detect psychotic ideation. The visual-experience module asked whether, when drug-free (not intoxicated, not within three days of use, and not in trance or sleep-deprived states), respondents had ever experienced any of nine listed phenomena (for example, halos or auras, movement of stationary objects, trailing afterimages, enhanced colours, open-eye patterns or formed objects). Endorsers of any of the first seven phenomena received follow-up questions on frequency, phenomenology, and a free-text description of the most vivid occurrence; they were also asked whether these experiences had been troublesome enough to prompt consideration of professional treatment and whether a specific precipitating event appeared to trigger onset. Inclusion criteria required completion of the questionnaire and reported past hallucinogen use; exclusion criteria included incomplete submissions, rapid duplicate submissions from the same IP address, self-reported poor English comprehension, endorsement of the fictional drug, reporting no hallucinogen use, or responses suggesting description of acute drug effects rather than drug-free phenomena. For analysis, estimated counts of drug exposures were log10-transformed because they were not normally distributed. The investigators focused on two main outcome measures: the number of types of unusual visual experiences endorsed (NUMSYMPT) and the number of experiences occurring constantly or nearly constantly (NUMCONSTANT). Statistical methods included Fisher's Exact Test for categorical variables, Poisson and binomial regression models with backwards elimination to identify independent predictors, calculation of odds ratios with 95% confidence intervals, chi-squared tests to assess model significance, and analyses conducted in R.

Over the 80-day recruitment period, 3,139 responses were collected (19.4% of 16,192 who viewed the information sheet), of which 2,679 (85.3%) met inclusion criteria. Exclusions were due to unfinished responses (283), duplicate submissions (126), difficulty understanding the questionnaire (31), apparent descriptions of acute drug effects (17), and no history of hallucinogen use (3). Among the included respondents, the sample was 89.5% male with a mean age of 21.6 ± 3.7 years (range 13–77). Most lived in the United States (68.9%), and respondents reported a median of five different drugs used (out of 15 listed). Two hundred and twenty-four respondents (8.4% of 2,679) reported at least one diagnosis associated with unusual visual experiences; these individuals were excluded from the main analyses except where psychosis history was specifically examined. In the analysed sample of 2,455 participants without complicating diagnoses, 1,487 (60.6%) reported at least one of the nine queried drug-free visual phenomena. Five hundred and eighty-seven (23.9%) endorsed at least one experience on a constant or near-constant basis. In free-text responses, 278 participants (11.3%) described additional visual experiences beyond the nine listed. Relationships between reported visual experiences and drug exposures were examined using regression models. A Poisson regression predicting NUMSYMPT from log10-transformed exposures to individual drugs was significant; higher exposures to LSD, LSA, psilocybin-containing mushrooms, DXM, ketamine, 2C-E, DPT, and Salvia were each significant predictors of a greater number of symptom types. For example, each log10-unit increase in LSD exposures was associated with an expected increase of 1.23 additional types of experiences. A Poisson model predicting NUMCONSTANT found log10-transformed exposures to 2C-E, LSD, DXM, LSA, and AMT to be significant predictors; estimated increases in the number of constant experiences per log10-unit exposure were 1.79 for 2C-E, 1.53 for LSD, 1.51 for DXM, 1.33 for LSA, and 0.67 for AMT. Binomial models predicting presence versus absence of any experience were also constructed, with significant predictor drugs and odds ratios reported in the paper's tables. Regarding clinical impact, 104 of 2,455 respondents (4.2%) said their visual experiences had been sufficiently troublesome to prompt thoughts of treatment; 70 considered treatment within the first two months after symptom onset, 14 within the two months prior to the questionnaire, and 20 during both periods. Only 27 participants (1.1%) had actually sought professional treatment. Presence of constant symptoms markedly increased the likelihood of considering or seeking treatment (odds ratio 9.76, 95% CI 6.18–15.82, p < 0.001). Thirteen-point-nine per cent (262 of 1,487) of respondents who reported visual experiences felt that a specific episode precipitated onset. Among these, 72.1% reported no prior occurrence of the symptoms. Of those reporting a precipitating episode, 45 (17.2%) considered or sought treatment; six attributed non-drug events in the week before onset (for example illness or loss of consciousness). Free-text descriptions of precipitating episodes often mentioned unusually high dose or strong effects (45 participants, 17.2%), an acute dysphoric response to the drug (26, 9.9%), or first exposure to the specific drug (22, 8.4%). A comparison between a subgroup identified as HPPD-like (45 participants who reported sudden onset after drug exposure and were treatment-seeking) and 33 otherwise-excluded individuals reporting a history of psychosis found no significant differences in age or cumulative drug exposures. However, those with a past diagnosis of psychosis were more likely to endorse psychotic-ideation items: believing others were spying on them (79% vs 44%, OR 4.97, 95% CI 1.66–16.56, p = 0.002), believing they had special powers (64% vs 27%, OR 4.70, 95% CI 1.65–14.24, p = 0.001), or feeling influenced by an outside force (54% vs 22%, OR 4.12–4.70, 95% CI 1.42–12.65, p = 0.004). Visually, those with psychosis were less likely to report movement of stationary objects (MOVEMENT: 42% vs 73%, OR 0.27, 95% CI 0.092–0.77, p = 0.0096) and less likely to report constant HALOS or PATTERNS compared with the HPPD-like subgroup.

Using an online convenience sample of hallucinogen-experienced individuals, the investigators found that drug-free visual experiences reminiscent of acute hallucinogen effects were commonly reported. Sixty-one-point-seven per cent of included respondents endorsed at least one such experience when not recently intoxicated, and 4.2% reported symptoms sufficiently distressing or impairing that they had considered or sought treatment. The authors emphasise that, because the study lacks objective testing and a drug-free comparison group, it cannot determine how many reports reflect objectively measurable visual dysfunction versus heightened attention to subtle normal phenomena, nor can it estimate prevalence in the general population. Despite these limitations, statistically significant relationships emerged between cumulative drug exposures and both the number and persistence of reported visual experiences. LSD was the most robust predictor, consistent with its prominence in prior case reports, and exposure to several other drugs (including certain phenethylamines, tryptamines, DXM, ketamine, LSA and Salvia) was also associated with greater reporting of symptoms. The authors note possible confounding because symptomatic individuals may cease use, obscuring exposure–response relationships, but also observe that many free-text accounts described symptom onset during intoxication with persistent effects thereafter. The paper addresses the possibility that many reports might represent undiagnosed psychosis. Comparison of the HPPD-like subgroup with respondents reporting past psychosis found different profiles of paranoid ideation and visual symptom types, which the authors interpret as arguing against prodromal or undiagnosed psychosis explaining most HPPD-like reports in this sample. The investigators also discuss diagnostic difficulty in distinguishing HPPD from other causes of visual disturbance, such as persistent migrainous aura without headache. Key limitations highlighted by the authors include reliance on self-report, the non-representative convenience sample drawn from an enthusiast website, potential selection bias favouring those with visual complaints, inability to make clinical diagnoses, and possible inaccuracies in recalled drug doses and timing. Given these caveats, the authors conclude that while the findings are preliminary, they suggest that post-hallucinogen visual changes may be more common than previously thought and that HPPD could represent a severe manifestation of a broader, drug-triggered susceptibility. They recommend more objective testing of visual function in hallucinogen-using populations to clarify mechanisms and prevalence.