Effect of psilocybin versus escitalopram on depression symptom severity in patients with moderate-to-severe major depressive disorder: observational 6-month follow-up of a phase 2, double-blind, randomised, controlled trial

Major depressive disorder (MDD) is a leading cause of global disability and carries a high risk of relapse after initial remission. Previous trials have shown that one or two high doses of psilocybin administered with psychological support can produce rapid antidepressant effects that may persist for months, but long-term outcomes have rarely been compared directly with established treatments such as selective serotonin reuptake inhibitors (SSRIs) plus psychotherapy. Erritzoe and colleagues set out to examine the six-month naturalistic follow-up of a Phase II, double-blind, randomised trial that contrasted two high-dose psilocybin sessions plus psychological support (psilocybin therapy; PT) with a six-week course of escitalopram combined with matched psychological support (escitalopram treatment; ET). The aim was to evaluate sustained changes in depressive symptom severity and secondary measures including work and social functioning, psychological connectedness, flourishing, and meaning in life, and to characterise potential differences between the two treatment arms over six months after the trial endpoint.

This report describes the six-month follow-up of a double-blind randomised controlled trial (ClinicalTrials.gov NCT03429075) in which 59 participants with moderate-to-severe unipolar MDD were randomised 1:1 to PT (n = 30) or ET (n = 29). Eligible participants were adults judged physically healthy and diagnosed with MDD; they discontinued any pre-trial antidepressants prior to baseline. All participants received preparatory and integrative psychotherapy and approximately 20 hours of in-person therapeutic support during the trial, delivered using the ACE (Accept, Connect and Embody) framework, which draws on processes from Acceptance and Commitment Therapy. In the PT arm participants received two oral 25 mg doses of COMP360 psilocybin (with therapeutic support during each 6–8 h dosing day) separated by three weeks, plus daily placebo capsules. The ET arm received two 1 mg psilocybin doses at the corresponding visits and a six-week course of escitalopram (10 mg daily for three weeks then 20 mg daily for three weeks) with matched psychological support. Participants in ET were not required either to stop or continue escitalopram after the trial. No safety monitoring of adverse events was performed during the follow-up period. Follow-up assessments were completed remotely via monthly online questionnaires for QIDS-SR-16 (primary outcome: 16-item Quick Inventory of Depressive Symptomatology Self-Report) and every three months for secondary measures: Work and Social Adjustment Scale (WSAS), Watts' Connectedness Scale (WCS), Flourishing Scale (FS), and Meaning in Life Questionnaire (MLQ). At six months participants also retrospectively reported any psychiatric medications, psychotherapy, or psychedelic use received since trial end. Statistical analyses included omnibus linear mixed-effects models (equivalent to two-way repeated measures ANOVA) to test Time × Condition interactions for each outcome. Multiple-comparison correction using the Benjamini–Hochberg false discovery rate (FDR) was applied to the first and second sets of analyses. To address missing data and potential attrition bias the investigators performed supplementary conservative single imputation (assigning worst follow-up scores to PT and best to ET) and additional models controlling for treatment-seeking during follow-up (use of psychiatric medications, non-pharmacological therapy, and psychedelic drugs). Sensitivity and supplementary within-arm analyses were also reported; these were treated as exploratory and not FDR-corrected.

Of 59 randomised participants (30 PT, 29 ET), 25 in the PT group and 21 in the ET group completed the six-month follow-up assessments. Average missing data for QIDS-SR-16 across follow-up was 13% (range 3–29%) in PT and 26% (range 14–38%) in ET; other measures showed similar differential missingness. During follow-up 37 of 59 participants (62.7%) reported receiving some form of additional treatment, with no significant differences between arms in initiation of medications, psychotherapy, or psychedelic use. On the primary outcome at six months both conditions showed sustained within-arm improvements in depressive symptoms. The mean between-condition difference in QIDS-SR-16 at six months was 1.51 (95% CI -1.35 to 4.38; p = 0.311), indicating no statistically significant difference between PT and ET on the primary self-report depression measure at that timepoint. However, linear mixed models without imputation detected a significant overall Time × Condition interaction for QIDS-SR-16 and post-hoc tests revealed a significant between-arm difference at one month, where PT showed larger early reductions (PT mean change baseline→1 month: -7.63, p < 0.001, Cohen's d = 1.55; ET: -4.04, p < 0.001, d = 0.74). Single-arm analyses indicated large within-arm effects at three and six months for both treatments (ET baseline→3 months: -6.59, p < 0.001, d = 1.33; ET baseline→6 months: -6.83, p < 0.001, d = 1.35; PT baseline→3 months: -6.48, p < 0.001, d = 1.24; PT baseline→6 months: -5.27, p < 0.001, d = 1.33). The authors note that remitter rates at one month were higher in PT (details in supplementary materials). Secondary outcomes favoured PT on functioning, connectedness, and meaning in life. For WSAS (work and social adjustment) a significant Time × Condition interaction was observed and PT showed substantially larger improvements: PT baseline→6 months mean change -12.73 (p < 0.001, d = 1.69) versus ET -3.10 (p = 0.002, d = 0.66). The reported between-condition mean difference on WSAS at follow-up was -7.46 (95% CI -12.4 to -2.47; p < 0.001). On the MLQ (meaning in life) PT produced larger gains (PT baseline→6 months +6.98, p < 0.001, d = 1.02; ET +2.12, p = 0.16, d = 0.31), with a between-condition mean difference of 4.86 (95% CI 0.67 to 9.05; p = 0.021). Connectedness (WCS) likewise improved more in PT (PT baseline→6 months +21.99, p < 0.001, d = 1.22; ET +10.95, p = 0.003, d = 0.74), with a between-condition mean difference of 11.02 (95% CI 1.25 to 20.83; p = 0.033). For flourishing (FS) no significant Time × Condition interaction was detected; both arms showed within-arm improvements. The investigators tested robustness to missing data and potential confounds using conservative imputation and models controlling for additional treatments during follow-up; the pattern of greater improvements on functioning, connectedness and meaning in PT remained present in these sensitivity analyses. Safety during follow-up was not assessed in this report.

Erritzoe and colleagues interpret the findings as showing that both a brief intensive course of psilocybin therapy and a six-week course of escitalopram combined with substantial psychological support were associated with sustained reductions in self-reported depressive symptoms up to six months. They emphasise that PT exhibited superior sustained improvements on psychosocial functioning, psychological connectedness and meaning in life, outcomes that the authors consider clinically relevant and sometimes more meaningful to patients than symptom reduction alone. The investigators discuss several possible mechanisms for PT's sustained effects, including neurobiological changes (neuroplasticity), psychologically transformative experiences enabled by the psychedelic state, and the role of psychotherapy in consolidation and relearning. They also consider that the relatively intense psychotherapy delivered in the ET arm (approximately 20 h including two long dosing-day sessions) may have potentiated escitalopram's effects, and that escitalopram itself may enhance neuroplasticity and thereby augment psychotherapy. The authors argue the two 1 mg psilocybin doses given in the ET arm (intended as placebo-like) were unlikely to account for findings based on typical microdosing regimens and timing. Several limitations acknowledged by the authors temper their conclusions. Missing data—higher in the ET arm—treatment-seeking during the follow-up period, lack of granular information about adjunct treatments, and reliance on retrospective self-report introduce uncertainty about causal attribution. The coronavirus pandemic overlapped with follow-up for many participants and may have influenced outcomes. The ET regimen represented a relatively brief SSRI course rather than prolonged maintenance dosing, limiting its generalisability as a ‘gold-standard’ comparator. The investigators also note possible psychometric issues with the QIDS-SR-16 and that the study was not optimally powered to detect small between-arm differences, raising the possibility of Type II error. Finally, longitudinal adverse-event or side-effect monitoring was not performed during follow-up. Given these limitations the authors characterise the results as preliminary and recommend replication and more granular measurement of post-trial treatments and safety in future studies.