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

Abstract

A renewed interest in the use of psychedelics for treating obsessive compulsive disorder (OCD) has emerged in the last 20 years. But pre-clinical and clinical evidence remain scarce, and little is known about the factor determining the magnitude and persistence of the therapeutic effect. We therefore designed a retrospective online survey to explore, in the general population using psychoactive drugs, their impact on OCD symptoms. We also assessed the attitude of the participants towards the substance in term of frequency of intakes. In a sample of 174 participants, classic psychedelics were reported as the only substances effective at reducing OCD symptoms. In classic psychedelics users, symptoms reduction was associated with the intensity of acute effects, itself correlated to the dose. Reports on the persistence of the therapeutic effect varied from weeks to months, but we could not find any predicting factor. Finally, the occurrence and frequency of subsequent intakes, which seemed to be limited in our sample, were predicted by the magnitude and persistence of the therapeutic effect, respectively. Our observations support the hypothesis of classic psychedelics efficacy in reducing OCD symptoms but a careful evaluation of the persistence of this effect is still needed.

Figure 1

Flow chart showing steps in participants selection.

Figure 2

Changes in OCD symptoms induced by psychoactive drugs. (A) Proportion of users declaring the given substance (dark green), another substance (light green) or no substance (light blue) as the most impacting on OCD symptoms (irrespective of the valence of the changes). Proportions were compared using a χ² test (B). Same representation for substances belonging to the classic psychedelics category. (C) Magnitude of changes in OCD symptoms for each substance category, expressed in arbitrary unit. Responses were provided on a scale ranging from -100 (= worsening) to + 100 (= improvement). Green dots represent the magnitude of the change per participant, black dots and bars represent the means and standard errors across participant, respectively. Significance was assessed using one-sample t-tests. (D) Same representation for psilocybin mushrooms and LSD users. The magnitude of the changes for NPS, delirogens, ayahuasca, DMT and mescaline were not computed due to small sample size (n < 5). ***P < .001, **P < .01, * P < .05, ns non-significant.

Figure 3

Relating self-reported dose to intensity of acute effects. (A). Percentage of participants who declared knowing the substance dose (n = 60) or not (n = 30). Light purple data represent participants discarded because of uninterpretable dose (n = 7). (B). Distribution of doses for psilocybin (left panel) and LSD (right panel). Black bars represent means. (C). Dose of substance as a function of the intensity of the acute effects. Doses are expressed as z-scores to include both psilocybin mushrooms and LSD in the same analysis. One dot is one participant, the line and ribbon represent the estimated regression line and its 95% confidence interval. Note that β represents the regressor for intensity in the statistical model used to predict the dose.

Figure 4

Modeling the magnitude of OCD improvement. OCD improvement as a function of intensity (A) and pleasantness (B) of acute effects of psychedelics, assessed in 90 participants. One dot is one participant, the line and ribbon represent the estimated regression line and 95% confidence intervals. Note that β represents the regressor for intensity (respectively pleasantness) in the linear regression used to predict the magnitude of OCD improvement.

Figure 5

Modeling subsequent intake probability and frequency. (A) Repeated use of the substance. (B) Persistence of substance-induced OCD improvement. (C) Estimated probability of subsequent intake as a function of OCD improvement magnitude, as modeled using a logistic regression (n = 90). One dot is one participant; the line and ribbon represent the regression curve and associated standard error. D. Estimated distribution of subsequent intake frequency according to the persistence of OCD improvement, as modeled using an ordinal regression (n = 69).