Improvement in OCD symptoms associated with serotoninergic psychedelics: a retrospective online survey

OCD is a chronic, often disabling psychiatric disorder featuring intrusive obsessions and repetitive compulsions. Standard treatments combine cognitive and behavioural therapy with high-dose selective serotonin reuptake inhibitors (SSRIs), but many patients have incomplete response, long delays to benefit and 30–40% non-response. This gap has renewed interest in alternative approaches, including classic serotoninergic psychedelics such as LSD and psilocybin, but clinical and preclinical evidence for their efficacy in OCD is limited, long-term outcomes are poorly characterised, and placebo/expectation effects are difficult to rule out in open or uncontrolled work. Buot and colleagues therefore conducted a retrospective, anonymous online survey in people with OCD symptoms to identify which psychoactive substances users perceived as changing their OCD, to examine whether any improvement related to features of the acute subjective experience (intensity, pleasantness), mindset and setting, and to document the occurrence and frequency of subsequent intakes. The survey targeted a broad set of substances and sought to explore determinants of both the magnitude and persistence of any self-reported therapeutic effects in a real-world sample.

The investigators deployed a bilingual (French and English) anonymous online survey from September 2021 to March 2022, advertised on OCD support organisation websites, psychedelic research and user forums, and related social media pages. Inclusion criteria were age >18 years, presence of OCD symptoms, and at least one lifetime use of listed psychoactive substances (including LSD, MDMA, psilocybin, DMT, ayahuasca, ketamine, 2C-B, Salvia, mescaline, ibogaine). Participants provided informed consent and data were collected in REDCap; IPs and identifying data were not recorded. The questionnaire comprised sections on sociodemographics, health and OCD history, current therapies and medications, lifetime substance use, and detailed questions about the single substance that the participant judged to have produced the strongest change in OCD symptoms. OCD severity was assessed with the short Obsessive Compulsive Inventory (OCI‑r). Change in OCD was measured across six domains (negative emotions, obsessions, compulsions/rituals, anxiety, acceptance, avoidance) on a −100 (worsening) to +100 (improvement) slider. Persistence of change, mindset (expectations), setting (context of intake), reported dose/onset/duration, whether the substance was taken again and frequency were collected. Acute subjective effects were rated on 13 positive/alteration items and 21 unpleasant items on 0–100 scales; a pleasantness score was computed as 100 minus the mean unpleasantness. Data processing included translation alignment, duplicate checks, exclusion of incomplete or incoherent responses, and grouping substances into five categories: classic psychedelics (LSD, psilocybin, DMT, mescaline, ayahuasca), entactogens (MDMA, ecstasy), ketamine, delirogens (Salvia, Datura) and novel psychoactive substances (2C‑B). Persistence and frequency categories were collapsed to ensure cell sizes adequate for regression. Intensity of general acute effects was used as a proxy for dose after showing a positive association with reported dose in a subset. Statistical analyses used chi-squared tests and one-sample t-tests for initial comparisons. Subsequent analyses focused on LSD and psilocybin users (n = 90). Linear regressions tested predictors of magnitude of OCD change and of pleasantness; an ordinal regression model examined predictors of persistence. Occurrence of subsequent intake was modelled with logistic regression and intake frequency with ordinal regression among repeat users. Regressors included substance, mindset, setting, intensity and pleasantness of acute effects, OCI‑r score, magnitude and persistence where appropriate. Multicollinearity was checked using generalized variance inflation factors (GVIF). Multiple testing correction used the Benjamini–Hochberg procedure and standard regression diagnostics and parallel‑slopes checks were applied.

From 227 initial respondents, 33 were excluded (no consent, incoherent or incomplete), leaving 174 participants who met the study criteria (either clinician diagnosis, OCI‑r > 18, or both). The sample was predominantly English respondents (76%), mean age 29 (range 18–65), 54% female, mean OCI‑r 36.37 (SD 12.82). Most reported no therapy (58%) or CBT (20%), and either no medication (52%) or antidepressant use (30%). Lifetime use of substance categories in the sample was high for classic psychedelics (84%), entactogens (72%) and ketamine (49%); smaller proportions reported NPS, delirogens or ibogaine. Within classic psychedelics users, 92% had used psilocybin mushrooms and 77% LSD. When asked whether any substance produced a change in OCD, classic psychedelics were selected as the most impacting by 66% of their users, compared with 15% for entactogens and 7% for ketamine; a chi‑square test showed these proportions differed by substance category (χ2(2) = 113.5, P < 0.001). Mean change scores were significantly positive only for classic psychedelics (mean ± SE = +42.62 ± 3.79, t(95) = 11.24, P < 0.001), while entactogens and ketamine did not produce significant mean changes. Within classic psychedelics, psilocybin (mean +44.22 ± 4.66) and LSD (mean +40.23 ± 5.95) both showed significant mean improvements. Analyses focused on LSD and psilocybin users (n = 90). Thirty‑six per cent had expected OCD improvement before intake and 66% reported taking the substance in no particular context. Reported onset was most commonly 30–60 minutes and duration 5–12 hours. Across participants, acute intensity averaged 67.07 (SE 1.61) and pleasantness 73.91 (SE 1.92). In a subset where dose could be estimated, intensity of acute effects correlated positively with measured dose (β = 0.02, t = 3.48, P = 0.001), supporting intensity as a dose proxy. A linear regression (n = 89) found that greater magnitude of OCD improvement was significantly associated with both higher intensity (β = 0.87, t = 3.86, Padj = 0.003) and greater pleasantness (β = 0.76, t = 3.96, Padj = 0.003) of the acute experience. Substance (psilocybin vs LSD), mindset, setting and baseline OCI‑r did not significantly predict magnitude. A separate model found no significant predictors of pleasantness. Reported persistence of improvement clustered at two modes: less than one week (33%) and more than three months (33%), with 21% reporting 1 week–1 month and 12% reporting 1–3 months. An ordinal regression (n = 90) did not identify any significant predictors of persistence among substance, intensity, pleasantness, magnitude of improvement or OCI‑r. Seventy‑seven per cent of classic psychedelic users reported taking the substance again; 47% of repeat users did so at most three times per year, 14% once a month and 16% at least once a week. Logistic regression showed that occurrence of subsequent intake was predicted only by the magnitude of OCD improvement (β = 0.03, z = 2.92, Padj = 0.03), indicating higher likelihood of re‑use when benefit was larger. Among the 69 repeat users, ordinal regression indicated that longer persistence of improvement (> 3 months) was associated with a greater probability of lower intake frequency (being in the at most three times per year category; β more3months‑less1week = -1.76, z = -2.53, P = 0.01). Other regressors (substance, intensity, pleasantness, magnitude) were not significant for frequency. In the subgroup taking the substance at least weekly, 57% reported microdosing regimens, 21% regular doses and 21% gave no dose information.

Buot and colleagues interpret their findings as supporting a perceived therapeutic association between classic serotoninergic psychedelics (principally psilocybin and LSD) and reductions in self‑reported OCD symptoms in a community sample. Classic psychedelics were the only substance category to show a significant mean improvement in the survey, and within that class psilocybin and LSD did not differ significantly in their reported effects. The authors emphasise that in their sample the magnitude of reported symptom improvement correlated with the intensity of the acute psychedelic experience, and that intensity itself was associated with higher reported doses, suggesting that moderate to high doses may be required to obtain measurable benefit. The authors note that pleasantness of the acute experience also correlated with improvement, but they caution this may reflect a carry‑over or global‑impression bias in retrospective self‑ratings rather than a causal mechanism. Mindset and setting did not emerge as predictors in their models, but the investigators acknowledge that their measurement of these factors was coarse and may have lacked sensitivity. Persistence of benefit varied widely; approximately 30% of respondents reported effects lasting beyond three months, but no baseline or acute‑experience variables predicted persistence in their analyses. The authors stress the imprecision of persistence estimates because the survey did not capture the time elapsed between intake and survey completion, limiting interpretation. Limitations the authors acknowledge include reliance on retrospective, self‑selected online reporting subject to recall and selection bias, lack of temporal anchoring for persistence estimates, possible carry‑over effects in subjective ratings, and inability to infer causality from observational data. They also note that approximately 10% of participants reported worsening of OCD after use and that risk factors for adverse outcomes require assessment in controlled settings. Regarding safety and misuse, the sample did not show evidence of high‑frequency problematic use; repeated use appeared linked to experienced benefit rather than to hedonic effects. For future research, the authors recommend well‑controlled clinical trials to determine efficacy and durability of effect, more fine‑grained study of the phenomenology of acute experiences and their therapeutic relevance, and specific evaluation of different symptom dimensions (obsessions, compulsions, depressive comorbidity). They also highlight neurobiological hypotheses such as drug‑induced neuroplasticity as one potential mechanism worth investigating in relation to sustained clinical effects.