Prediction of hallucinogen persisting perception disorder and thought disturbance symptoms following psychedelic use

After a long hiatus, research into classic serotonergic psychedelics (for example psilocybin, LSD, mescaline, and DMT) has expanded rapidly because these compounds induce profound alterations in perception, mood and cognition and show therapeutic promise for several psychiatric conditions. Despite generally favourable safety profiles in controlled trials, case reports and naturalistic surveys have raised concerns about rare but potentially persistent adverse outcomes, notably prolonged psychotic reactions or increased schizotypal traits, and enduring visuoperceptual disturbances clinically recognised as hallucinogen persisting perception disorder (HPPD). HPPD is commonly described in two subtypes: transient “flashbacks” (type 1) and chronic visuoperceptual aberrations (type 2), such as motion trails, halos and intensified colours. Observational studies of naturalistic psychedelic use are particularly important because clinical trials typically exclude individuals considered at-risk, leaving uncertainty about frequency and predictors of iatrogenic effects in broader populations. Zhou and colleagues designed a prospective, naturalistic study to examine changes in delusional ideation and magical thinking surrounding a planned psychedelic experience, and to estimate the prevalence and baseline predictors of HPPD-type visual symptoms. The investigators hypothesised that baseline delusional ideation and magical thinking would correlate positively with lifetime psychedelic use and performed exploratory regression analyses to test whether demographics, trait absorption, psychiatric history, aspects of drug use or features of the acute experience predicted HPPD-type effects or shifts in schizotypal measures over a four-week follow-up period.

This was a large-scale prospective online survey recruiting English-speaking adults who planned to use a psychedelic (psilocybin, LSD/1P-LSD, ayahuasca, DMT/5-MeO-DMT, salvia, mescaline, or iboga/ibogaine) within three months. Sign-up was via a purpose-built website and recruitment channels included social media and drug-related websites. Data were collected at three timepoints: baseline (one week before the planned experience), a sub-acute assessment (one day after the experience for acute-experience details), and follow-ups at two and four weeks post-experience; HPPD-type symptoms were assessed only at the 4-week endpoint. The baseline survey captured demographics (age, gender, education, employment), lifetime psychedelic use, psychiatric history, and psychometric measures: the Peters’ Delusions Inventory (PDI) for delusional ideation, the 15-item Magical Ideation Scale (MIS) for magical thinking, and the 15-item Modified Tellegen Absorption Scale (MODTAS) for trait absorption. An investigator-constructed HPPD checklist asked about nine specific visual symptoms (for example intensified colours, positive afterimages, motion trails, halos, macropsia/micropsia); participants endorsing any effect answered follow-up questions about distress and whether symptoms were attributable to medical causes. Acute-experience variables included drug type, dose, setting (retreat vs non-retreat) and concomitant non-psychedelic drug use. Statistical analyses used Spearman correlations to test associations between lifetime psychedelic use and baseline PDI/MIS scores, with partial correlations adjusting for age. Longitudinal changes in PDI and MIS were tested with Friedman and Wilcoxon signed-rank tests, and between-group comparisons (novice vs experienced users) used Mann–Whitney U tests. Change scores were dichotomised (>0 versus ≤0) to define increases in PDI or MIS and analysed with logistic regression including predictors that passed collinearity checks (variance inflation factor < 10). A separate logistic regression model examined predictors of reporting at least one HPPD symptom at 4 weeks; a multinomial model explored specific psychiatric diagnoses. Analyses were performed in Python and statistical significance was set at P < 0.05. The extracted text does not clearly report precise dosing regimens, nor detailed measures of prior dosing frequency or purity of substances used.

At baseline 654 participants completed the survey, with 315 and 212 participants completing the two- and four-week follow-ups respectively. Mean age was 28.9 years (SD = 10.5); 74.2% were male (n = 485), 77.1% had university-level education or above (n = 504), and 51.2% were employed (n = 337) while 39.1% were students (n = 256). Participants were categorised as “novice” (< 6 lifetime psychedelic experiences; n = 250, 38.2%) or “experienced” (≥ 6 experiences; n = 404, 61.8%). About a third (33.0%) reported a lifetime psychiatric diagnosis. During the indexed acute experience, most reported using LSD/1P-LSD (n = 143, 54.0%) or psilocybin/magic mushrooms/truffles (n = 108, 40.8%); 17.7% (n = 67) reported the acute experience occurring in a retreat setting. Cross-sectionally at baseline, weak but statistically significant positive correlations were observed between lifetime psychedelic use and delusional ideation (Spearman r_s = 0.11, P = 0.01) and between lifetime use and magical ideation (r_s = 0.12, P = 0.003). These relationships remained after adjusting for age (PDI r_s = 0.11, P = 0.004; MIS r_s = 0.13, P = 0.001). The distress component of the PDI did not correlate with lifetime psychedelic use (r_s = 0.01, P = 0.06). Longitudinally, the overall sample showed a statistically significant reduction in delusional ideation after the psychedelic experience (χ2 = 15.47, P ≤ 0.001). Pairwise comparisons showed decreases from baseline to 2 weeks (Z = -3.60, P ≤ 0.001) and from baseline to 4 weeks (Z = -5.19, P ≤ 0.001). Novice users exhibited significant reductions in PDI scores from baseline (mean = 5.89, SD = 3.87) to 2 and 4 weeks (4-week mean = 3.91, SD = 3.21). No significant changes in magical ideation were observed for experienced users over the 4-week period. Regarding increases in schizotypal measures, 23.8% of participants showed increased delusional ideation at 4 weeks (mean change 1.8 points), but a logistic regression did not identify significant predictors for persistent PDI increases. Increased magical ideation occurred in 28.1% of participants; the authors report that higher baseline trait absorption was associated with increased odds of reporting greater magical ideation after the experience. The extracted text gives an odds increase figure of 6.0% per absorption unit (P < 0.001) but the associated confidence interval as extracted appears inconsistent and is not clearly reported. At the 4-week endpoint, 68 of 212 respondents (32.1%) endorsed at least one HPPD-type visual symptom; roughly half of these reported two or more effects (n = 35). The most frequently reported symptoms were intensified colours (n = 29, 18.6%) and positive afterimages (n = 28, 17.9%); macropsia (n = 6, 3.8%) and micropsia (n = 5, 3.2%) were least common. There was no significant difference in HPPD reporting between novice and experienced users (P = 0.64). Only two participants (< 1% of the total sample; 2.9% of those reporting any HPPD effects) described their symptoms as distressing; both were experienced users with > 10 prior uses. One additional participant endorsed all nine symptoms but did not report distress. In a multivariable logistic regression predicting any HPPD-type symptom at 4 weeks, several baseline factors were associated with altered odds: male sex was associated with lower odds compared with female (odds reduced by about 63.4%; reported 95% CI [0.15, 0.91]), each year of higher age reduced odds by approximately 6.2% (95% CI [0.90, 0.97]), and higher baseline absorption increased odds modestly (reported increase about 3.4% per absorption unit; 95% CI [1.01, 1.06]). A reported history of a psychiatric diagnosis substantially increased the odds of reporting at least one HPPD-type effect (authors indicate an increase of about 202%), though specific diagnostic categories did not emerge as significant predictors in a secondary model. Drug type (binary LSD vs non-LSD), dose and polydrug use were not identified as strong predictors in the extracted results. The authors note that the secondary logistic model had a low pseudo R-squared of 0.015, indicating limited explanatory power.

Zhou and colleagues interpret their findings as showing that, in this naturalistic prospective sample, delusional ideation decreased over the month following a planned psychedelic experience, which contradicts the simple hypothesis that increased lifetime psychedelic exposure causes elevated schizotypal traits. They suggest alternative explanations for the observed baseline association between lifetime use and higher schizotypy, including self-selection (persons with higher trait schizotypy or greater risk-taking being more likely to try psychedelics), self-medication for latent pathology, or a transient post-use reduction in schizotypal thinking that might rebound later. The authors caution that the observed short-term decreases could represent temporary ‘‘honeymoon’’ effects and call for studies examining longer follow-up intervals to determine durability. The cohort showed higher-than-average baseline magical ideation relative to general populations, and lifetime use correlated with magical thinking at baseline; the investigators note that prior research has also reported group-level shifts in metaphysical beliefs after psychedelic use, but underscore that correlation does not imply causation. Trait absorption emerged as a consistent predictor: it predicted increases in magical ideation and was associated with HPPD-type symptoms, which the authors link to absorption’s relationship with imagination and hallucination-proneness. While higher absorption may be associated with some adverse perceptual aftereffects, the authors observe that it is also predictive of types of acute subjective experiences thought to be therapeutically meaningful. HPPD-type visual phenomena were relatively common (about one-third at 4 weeks) but almost never reported as distressing, implying that clinically relevant HPPD is much rarer—likely under 1% in this sample—consistent with prior naturalistic surveys and lower than the DSM‑V prevalence estimate for HPPD. The authors propose that features of the acute setting (for example eyes-closed experiences or use of eyeshades in therapeutic contexts) might reduce risk of persistent visual disturbances and note that age, female sex and psychiatric history were associated with greater odds of HPPD-type effects. Proposed mechanisms include age-related windows for perceptual learning, sex differences in pharmacokinetics or dose-to-weight effects, and greater neurobiological sensitivity in people with psychiatric histories. However, they acknowledge that the drug-type variable (LSD vs non-LSD) was dichotomous and may have failed to capture dose–frequency nuances. Key limitations acknowledged include reliance on self-report and online recruitment leading to potential sample bias, lack of data on age at initiation, prior dose history, baseline prevalence of HPPD-like symptoms, and substance purity, and substantial attrition between baseline and follow-ups which may bias prevalence estimates. The authors also caution that some extracted numerical details and confidence intervals in their reported models are unclear in the provided text. Overall, they recommend larger, better-controlled studies, including more granular measurement of dose, setting and prior use, and argue for careful screening and aftercare when expanding access to psychedelic treatments.

In this prospective naturalistic sample Zhou and colleagues report a short-term reduction in delusional ideation four weeks after a planned psychedelic experience, despite cross-sectional associations between lifetime psychedelic use and higher baseline schizotypal traits. HPPD-type visual effects were commonly reported at four weeks but were almost never described as distressing, indicating that clinically significant HPPD was very rare in this cohort. Younger age, female sex, a history of psychiatric diagnosis, and higher baseline trait absorption were associated with increased odds of reporting HPPD-like symptoms. The authors call for larger, better-controlled studies and careful clinical screening and aftercare to better characterise causal relationships and risk factors as psychedelic use expands.