The Effect of Psychedelics on Individuals with a Personality Disorder: Results from two Prospective Cohort Studies

Personality disorders (PDs) are enduring patterns of maladaptive intra- and interpersonal functioning that frequently co-occur with depression, anxiety, suicidality, emotion-regulation problems and impaired cognitive flexibility. Existing psychotherapies and adjunctive pharmacotherapies show only modest effectiveness for many PD presentations, and regulatory-approved pharmacologic treatments specific to PDs are lacking. Classic psychedelics (for example psilocybin, LSD, DMT/ayahuasca, mescaline) act primarily at the 5-HT2A receptor and, in therapeutic contexts, have shown promise for a range of psychiatric disorders and for improving constructs relevant to PD pathology such as emotion regulation and cognitive flexibility. Despite this, people with PDs are frequently excluded from psychedelic-assisted therapy (PAT) trials, leaving a gap in knowledge about safety and therapeutic effects for this population. Gordon and colleagues set out to address that gap by analysing prospective data from two naturalistic cohort studies of psychedelic use. The investigators aimed to characterise short- to medium-term changes in suicidality, depressive symptoms, anxiety, cognitive flexibility and emotion-regulation (expressive suppression and cognitive reappraisal) among people who self-reported a PD diagnosis and who used psychedelics outside a controlled clinical trial. The paper reports two analyses: Study 1 pooled data from three online cohort studies of classic psychedelic use and focused on suicidal ideation, depression and trait anxiety up to 4 weeks post-use, while Study 2 used a larger prospective cohort of planned psilocybin use to examine depressive symptoms, trait anxiety, cognitive flexibility and emotion-regulation up to 2–3 months post-use.

Both analyses used prospective, naturalistic cohort data collected via the online platform Psychedelic Survey between March 2017 and December 2019. Study 1 combined three anonymous cohort studies that recruited individuals planning to use classic psychedelics in naturalistic or ceremonial settings; inclusion criteria were age >18, adequate English comprehension and intention to take a classic psychedelic. Participants completed baseline surveys roughly 1 week before use and follow-ups at 2 weeks and 4 weeks. From a parent sample of 2,250, 21 participants self-reported a formal PD diagnosis and completed baseline plus at least one follow-up, and these 21 comprised the Study 1 PD subsample. Study 2 drew from a large cohort of individuals planning psilocybin use; of 8,006 respondents 457 reported a lifetime PD diagnosis and 55 completed baseline and at least one post-use assessment (2–4 weeks and/or 2–3 months). The investigators measured suicidal ideation (Study 1 only) using the Suicidal Ideation Attributes Scale (SIDAS); depressive symptoms with the QIDS-SR-16 in Study 1 and the Beck Depression Inventory-II (BDI-II, with the suicide item removed) in Study 2; trait anxiety with the shortened STAI-trait (Study 1) and other trait measures in Study 2; and in Study 2 additionally assessed cognitive flexibility and emotion-regulation using the Emotion Regulation Questionnaire (ERQ) subscales expressive suppression and cognitive reappraisal. Demographic and diagnostic history data were collected at baseline. Analyses focused on within-subject change from baseline to follow-up(s). For suicidal ideation, changes were reported by SIDAS risk categories (0 none, 1–20 low, 21–50 high) and individuals scoring 0 at baseline were excluded from change analyses. Clinically meaningful worsening in depression was defined using established minimal clinically important difference thresholds (>28.5% increase on QIDS-SR-16; >13.49% increase on BDI-II). Normality was assessed with QQ plots, Shapiro–Wilk and Kolmogorov–Smirnov tests; paired t-tests were used for normally distributed change scores and Wilcoxon matched-pairs signed-rank tests for non-normal data. Effect sizes were computed as Hedges' g for t-tests and r_s for Wilcoxon tests. Correlations between changes in cognitive measures and changes in anxiety/depression were computed using Pearson or Spearman coefficients as appropriate. False discovery rate correction (Benjamini–Hochberg) was applied separately for sets of time-point comparisons and for correlation analyses. The analyses were conducted in Microsoft Excel and GraphPad Prism. The extracted text does not clearly report detailed dosing, setting characteristics beyond 'naturalistic' or 'ceremonial', or which specific cognitive flexibility measure was used in Study 2.

Study 1 (n=21 with self-reported PD) reported mixed but predominately favourable short-term changes in suicidality, depressive symptoms and trait anxiety following naturalistic classic psychedelic use. Suicidal ideation resolved completely in 20% (n=3) of participants from baseline to 4 weeks, while one individual (6.67%) showed an increase that remained within the low-risk category; no participant increased into the high-risk category. A Wilcoxon matched-pairs test found a significant reduction in SIDAS scores at 4 weeks compared to baseline (median baseline 9 vs median 5 at 4 weeks), z=0.63, p=0.009, p-FDR=0.015, with a large effect size r_s=0.86. Regarding depressive symptoms (QIDS-SR-16), from baseline to 2 weeks 78.6% (n=11 of those assessed) showed reductions and 64.3% (n=9) met the study's threshold for clinically significant reduction; three participants (21.4%) increased in depression scores and two (14.3%) had clinically significant increases. From baseline to 4 weeks, 71.4% (n=10) demonstrated reductions with 61.3% (n=9) clinically significant reductions; two participants (14.2%) had clinically significant increases (a 33.3% rise). Trait anxiety decreased for the majority: at 2 weeks 63.6% (n=14) improved, 22.7% showed no change and 13.6% (n=3) increased; at 4 weeks 62.5% (n=10) improved, 12.5% showed no change and 25% (n=4) increased. The extracted text reports effect-size ranges in other sections consistent with moderate-to-large reductions in anxiety and depression (for example Hedges' g values reported elsewhere in the document), but full statistical detail for each paired comparison in Study 1 was limited by sample size and missing data. Study 2 (n=55 with self-reported PD completing baseline plus follow-up) found sustained reductions in depressive symptoms up to 2–3 months after planned psilocybin use. Wilcoxon tests indicated significant decreases in BDI-II scores at 2–4 weeks (median baseline 23 vs median 7 at 2–4 weeks; z=0.12, p<0.001, p-FDR=0.002, r_s=0.52) and at 2–3 months (median baseline 23 vs median 9.5 at 2–3 months; z=0.29, p<0.001, p-FDR=0.002, r_s=0.71). Expressive suppression (ERQ subscale) did not change significantly at either follow-up, despite large effect-size estimates in the extracted tests (non-significant Wilcoxon results: p>0.1 after correction). For cognitive flexibility, Study 2 reported a transient small increase at 2–4 weeks (the extracted text gives a small effect estimate) but changes in cognitive flexibility were not significantly correlated with changes in depression (r_s=-0.17, p=0.220, p-FDR=0.293) or trait anxiety (r=0.11, p=0.472, p-FDR=0.472). By contrast, increases in cognitive reappraisal were associated with improvements in both depressive symptoms (r_s=-0.37, p=0.007, p-FDR=0.028) and trait anxiety (r=-0.33, p=0.020, p-FDR=0.04). The extracted text also notes that several participants experienced transient worsening: non-trivial proportions showed increased anxiety (Study 1: 13.6%–25.0%; Study 2: 11.5%–16.3% depending on timepoint) and some had clinically significant worsening of depressive symptoms (Study 1: roughly 14% at different timepoints; Study 2: about 8%–10%). Finally, no instances of post-use increases in suicide risk category were observed across the samples. Across both studies the principal pattern was overall reduction in depressive symptoms and trait anxiety for the majority of participants, with some individuals experiencing worsening; correlations suggest that improved cognitive reappraisal coincided with symptom reductions. The extracted text does not provide detailed adverse-event narratives, dosing data or PD subtype-specific results.

Gordon and colleagues interpret these prospective, naturalistic findings as preliminary evidence that psychedelic use may be associated with improvements in suicidality, depressive symptoms and trait anxiety among people who self-report a PD diagnosis, with effects evident up to 4 weeks in Study 1 and up to 2–3 months in Study 2. The investigators highlight the clinically important observation that no participant escalated into a high-risk suicidality category and that 20% showed complete resolution of suicidal ideation by 4 weeks. At the same time, they acknowledge that several participants experienced transient or clinically meaningful worsening of anxiety or depressive symptoms, underscoring that adverse outcomes are possible in this population. The authors situate their findings within a limited literature that includes case reports and small trials suggesting potential benefits of PAT for individuals with PD traits, but note that PDs are often excluded from contemporary clinical trials because of safety concerns. They therefore present their data as an early, ecologically valid contribution addressing the safety and therapeutic questions relevant to PD populations. Key limitations are emphasised. The samples were small, diagnoses of PD were self-reported rather than clinically confirmed, participants were not recruited as PD cohorts in the parent studies, and psychedelic use occurred in uncontrolled, naturalistic settings. The design is susceptible to regression to the mean, attrition, expectancy effects and sampling bias. Important stratifying information was not collected: the studies did not differentiate PD subtypes, nor did they include PD-specific symptom measures or standardised clinician-rated outcomes. Dosing, setting and concomitant treatments were not described in detail in the extracted text. For these reasons the investigators caution against overinterpretation and call for clinical trials and larger observational studies that include rigorous PD assessment, longer follow-up, and procedures to monitor and manage suicide risk. Regarding implications, the study team suggests that the apparent association between increased cognitive reappraisal and symptom reduction points to emotion-regulation processes as a potential mechanism worth investigating. They recommend future research to examine psychedelic effects by specific PDs, to collect clinician-rated and PD-specific measures, and to evaluate the safety and efficacy of controlled psychedelic-assisted therapies in this population, including exploration of repeated dosing and longer-term outcomes. The authors frame their conclusions as preliminary and stress the need for caution given the complexity and safety considerations of treating individuals with PDs.

Following naturalistic psychedelic use among individuals who self-reported a PD diagnosis, the investigators did not observe any instances of clinically significant worsening of suicidal ideation, while a minority experienced worsening anxiety or depressive symptoms. Most participants showed reductions in suicidal ideation, anxiety and depression, transient increases in cognitive flexibility and sustained increases in cognitive reappraisal, with improvements in reappraisal correlating with symptom reductions. Given substantial methodological limitations—most notably small, self-selected samples, self-reported diagnoses and uncontrolled settings—the authors conclude that these results are preliminary and that controlled clinical trials are required to determine safety and potential therapeutic efficacy of psychedelics for people with PDs.