Effects of psilocybin versus escitalopram on rumination and thought suppression in depression

Major depressive disorder (MDD) commonly involves maladaptive coping strategies such as rumination and thought suppression, which are linked to maintenance, recurrence and greater severity of depressive episodes. Thought suppression denotes effortful attempts to avoid distressing thoughts, which often backfire and increase intrusive thinking, while rumination denotes repetitive, self-focused negative thinking. Earlier research indicates both strategies are associated with poorer problem-solving and emotional processing. Selective serotonin reuptake inhibitors (SSRIs) such as escitalopram are a first-line pharmacological treatment for MDD but have only partial response rates (around 50-60%) and can cause side-effects including emotional blunting. Classic serotonergic psychedelics, notably psilocybin, combined with psychological support, have been proposed to relax entrenched cognitive habits and increase psychological flexibility, and prior qualitative and non-randomised data have suggested reductions in self-rumination after psychedelic therapy. However, no randomised trial had directly compared psilocybin with an established SSRI on standardised clinical measures of rumination and thought suppression while attempting to maintain blinding and providing an active comparator. Barba and colleagues report exploratory analyses from a previously conducted randomised clinical trial (N = 59) that compared psilocybin therapy (COMP360) with 6 weeks of escitalopram for MDD. The present paper focuses specifically on the 22-item Ruminative Response Scale (RRS) and the White Bear Suppression Inventory (WBSI) as outcomes. The primary hypothesis was that psilocybin therapy would produce greater reductions in RRS and WBSI at the 6-week primary end-point than escitalopram. Secondary hypotheses examined whether reductions were driven by participants meeting conventional treatment-response criteria (≥50% reduction on QIDS-SR-16) and whether acute subjective experiences or subsequent psychological insight following psilocybin sessions related to changes in rumination and suppression.

The analyses derive from a randomised, controlled trial registered as NCT03429075. Ethics approvals and regulatory authorisations were obtained from relevant UK bodies and the trial received psilocybin (COMP360) under a Schedule 1 licence. The intention-to-treat sample comprised 59 adults with MDD, randomised 30 to psilocybin therapy and 29 to escitalopram. Mean age was 41 years; 34% were women and 85% White. Some participants had recently discontinued psychiatric medication or psychotherapy per trial stopping rules; a small number did not complete all dosing visits for reasons including adverse effects and COVID-19 restrictions. Participants attended six visits over 6 weeks. The psilocybin arm received two supervised oral doses of 25 mg psilocybin on dosing visits separated by 3 weeks; the escitalopram arm received a 1 mg ‘‘active placebo’’ psilocybin dose on those same dosing visits to standardise expectation, combined with daily oral escitalopram (reported elsewhere as 10–20 mg/day) taken in capsule form. Between the first and second dosing days participants took one capsule each morning, increasing to two each morning from week 3; capsules contained inert filler for the high-dose psilocybin group and escitalopram for the low-dose psilocybin condition. Psychological support modelled on acceptance and commitment therapy was provided before, during and after dosing and during integration visits. Primary outcomes for these analyses were RRS (22 items, scored 22–88) and WBSI (14 items reported here, scored 15–75), measured at baseline and at 6-week follow-up; internal consistency at baseline and follow-up was reported (RRS α = 0.81 and 0.94; WBSI α = 0.71 and 0.92). Depressive symptoms were assessed with the QIDS-SR-16 and treatment response at week 6 was defined as ≥50% reduction from baseline. Acute subjective measures following psilocybin sessions included the Ego-Dissolution Inventory (EDI), Emotional Breakthrough Inventory (EBI) and Challenging Experience Questionnaire (CEQ); the Psychological Insight Scale (PIS-6) was administered at week 6 to quantify insight after the experience and integration. Statistical analyses used intention-to-treat data. Two-way mixed ANCOVAs tested time (baseline to 6 weeks) by treatment condition effects on RRS and WBSI, with centred baseline scores as covariates. Three-way mixed ANCOVAs tested time×condition×response (responder versus non-responder) interactions. Pairwise follow-up comparisons reported Cohen's d effect sizes but were not corrected for multiple comparisons; supportive mixed-model and non-parametric analyses were reported in supplementary materials. Bivariate Pearson correlations assessed associations among changes in QIDS-SR-16, RRS and WBSI; Spearman correlations (two-tailed) were used for non-normally distributed acute measures. Statistical significance was set at P < 0.05.

Primary analyses found significant time by condition interactions for both outcomes. For rumination (RRS) there was a main effect of time (F(1,56) = 7.72, P = 0.007) and a time×condition interaction (F(1,56) = 4.58, P = 0.037). Pairwise tests showed no significant RRS change in the escitalopram group from baseline to 6 weeks (mean difference −1.00, P = 0.16, d = 0.10), whereas the psilocybin group showed a significant reduction (mean difference −7.76, P < 0.001, d = 0.63). For thought suppression (WBSI) there was a strong within-participant time effect (F(1,57) = 19.79, P < 0.001) and a time×condition interaction (F(1,57) = 5.88, P = 0.019). The escitalopram group showed no significant WBSI change (mean difference −2.85, P = 0.162, d = 0.32), while the psilocybin group showed a significant reduction (mean difference −9.70, P < 0.001, d = 0.87). Secondary analyses that incorporated treatment response produced a nuanced pattern. For RRS the time×response interaction was significant (F(1,54) = 23.50, P < 0.001) while the time×condition×response interaction was not. Both escitalopram responders and psilocybin responders showed significant reductions in rumination (escitalopram responders mean difference −7.00, P = 0.013, d = 0.62; psilocybin responders mean difference −12.72, P < 0.001, d = 0.82). Non-responders in either condition did not show significant RRS changes. For WBSI a significant three-way interaction was observed (time×condition×response: F(1,54) = 8.42, P = 0.005). Only psilocybin responders showed a significant decrease in thought suppression (mean difference −13.95, P < 0.001, d = 0.91); escitalopram responders did not (mean difference −1.50, P = 0.575, d = 0.18), and non-responders in either arm showed no significant WBSI change. Correlational analyses showed expected baseline interrelations among measures in both groups: baseline RRS correlated with WBSI, and baseline QIDS-SR-16 correlated with both RRS and WBSI. Change scores revealed that in the psilocybin group reductions in depressive symptoms (ΔQIDS-SR-16) correlated with reductions in both RRS (r = 0.48, P = 0.007) and WBSI (r = 0.49, P = 0.01). In the escitalopram group ΔQIDS-SR-16 correlated with ΔRRS (r = 0.39, P = 0.014) but not with ΔWBSI (r = −0.04, P = 0.926). In the psilocybin arm, ΔRRS and ΔWBSI were strongly correlated (r = 0.66, P < 0.001); this association was absent in the escitalopram arm (r = 0.18, P = 0.354). Acute subjective measures during psilocybin dosing were related to subsequent change. Higher maximum ego-dissolution (EDI) scores correlated with greater reductions in rumination (ΔRRS r = −0.44, P = 0.014) and thought suppression (ΔWBSI r = −0.41, P = 0.024). Emotional breakthrough (EBI) and challenging experience (CEQ) did not show significant correlations with ΔRRS or ΔWBSI, though trends were reported for WBSI. Psychological insight measured at week 6 (PIS-6) correlated strongly with reductions in both rumination (r = −0.69, P < 0.001) and thought suppression (r = −0.56, P < 0.001) in the psilocybin group; no such relationships were observed in the escitalopram group. The extracted text does not provide full tables of means and standard errors, but reports the key effect sizes and P-values above.

Barba and colleagues interpret their findings as evidence that psilocybin therapy produces broader changes in maladaptive coping strategies than a 6-week course of escitalopram. Specifically, psilocybin led to significant reductions in both rumination and thought suppression at the 6-week end-point, whereas escitalopram responders showed reductions only in rumination; escitalopram did not produce measurable decreases in thought suppression even among clinical responders. The authors suggest these differential patterns may reflect distinct mechanisms of action: psilocybin's 5-HT2A receptor agonism may acutely dysregulate entrenched patterns of spontaneous neuronal activity in higher cortical regions, opening a window of plasticity that, when combined with psychological support, can facilitate insight, acceptance and relinquishing of avoidant cognitive habits. By contrast, escitalopram's effect on increasing synaptic serotonin and downstream 5-HT1A-mediated dampening of limbic responsivity may confer resilience and reduced stress reactivity but be less effective for overturning defensive cognitive strategies such as thought suppression. The investigators note that reductions in rumination correlated with reductions in depressive symptoms in both arms, suggesting rumination is a treatment-sensitive feature of depression irrespective of modality. For thought suppression, however, the association with clinical improvement was observed only in the psilocybin arm, and reductions in suppression were related to higher acute ego dissolution and to psychological insight following dosing, implying a role for altered-state and integrative processes in producing that specific change. Several limitations are acknowledged. These analyses were exploratory tertiary outcomes not preregistered and were not corrected for multiple comparisons, so type I error is a risk. The sample was modest in size and lacked diversity, limiting generalisability, and subgroup analyses were underpowered. Blinding integrity is questionable: the escitalopram arm received a 1 mg psilocybin dose to equalise expectation but acute effects of 25 mg psilocybin are conspicuous and could have led to correct condition guessing and expectancy biases. Both arms received substantial psychological support, modelled on acceptance and commitment therapy, so pharmacological effects cannot be disentangled from context and therapy; the authors emphasise the likely synergy between psilocybin and therapeutic support. The 6-week duration may have limited escitalopram's capacity to show full effects, as SSRIs can require longer courses to impact rumination in prior studies. Concurrent measurement of patient-rated scales at the same time points prevents causal inference about temporal ordering among symptom change, rumination and suppression. Lastly, safety conclusions are tentative given small trial sizes and the need for larger real-world data to characterise uncommon adverse events. The authors recommend replication in larger, more diverse samples, assessment of moderator variables such as baseline cognitive control or emotional regulation, longer-course SSRI comparisons, and use of objective measures (for example neuroimaging and additional time points) to probe mechanisms and mitigate expectancy and blinding concerns. They also suggest that the observed effects on rumination and thought suppression may help explain preliminary signals of psilocybin efficacy in related conditions characterised by repetitive or avoidant thought patterns, and propose further testing of theoretical models such as REBUS to link neurobiology, altered states and changes in maladaptive cognitive domains.