Psilocybin increases emotional empathy in patients with major depression

Empathy, encompassing emotional and cognitive components, is central to social cognition and interpersonal functioning and is frequently impaired in psychiatric disorders including major depression. Emotional empathy denotes vicariously feeling another person's affect, while cognitive empathy (theory of mind) refers to accurately recognising or labelling another's emotional state. Previous research in healthy volunteers showed that classic psychedelics, including psilocybin, acutely and subacutely increase emotional empathy but have not reliably affected cognitive empathy. Despite promising clinical effects of psilocybin in depression — the authors note a remission rate of 54% in their larger trial sample — mechanisms linking social cognition changes to therapeutic benefit remain unclear. Jungwirth and colleagues set out to test whether a single dose of psilocybin, given within a brief psychological support programme, produces enduring increases in empathy in patients with mild to moderate major depressive disorder. The primary hypothesis was that psilocybin versus placebo would increase emotional empathy for up to two weeks post‑administration; a secondary hypothesis queried whether improvements in empathy would be associated with reductions in depressive symptoms independent of treatment condition. The study therefore examines both valence-specific effects (positive versus negative stimuli) and associations between empathy change and clinical outcomes.

This was a randomised, double‑blind, placebo‑controlled, parallel‑groups trial embedded within a clinical study of psilocybin‑assisted therapy for major depression. Adults aged 18–60 meeting DSM‑IV criteria for a current mild to moderate depressive episode (MADRS 10–40) were eligible; detailed inclusion and exclusion criteria, recruitment procedures and baseline symptom severity are reported in the supplement. Psychiatric medication was washed out at least two weeks or five half‑lives prior to dosing and participants agreed to pause external psychotherapy for the study duration. Blinding was maintained for all study staff and participants except the resident pharmacist, with unblinding occurring only after database closure. The pharmacological intervention was a single oral dose of psilocybin at 0.215 mg/kg or placebo (mannitol). Psychological support comprised two preparation sessions (one and four days before dosing) and three integration sessions at +2, +8 and +14 days after dosing; all support sessions were delivered by the same study therapist and lasted about 60 minutes. Outcome measurement occurred at baseline (‑1 day) and at +2, +8 and +14 days post‑administration. Empathy was assessed with the Multifaceted Empathy Test (MET), which presents 40 photographs (20 positive, 20 negative) and yields scores for cognitive empathy, implicit emotional empathy (arousal rating on a 1–9 Likert scale) and explicit emotional empathy (‘‘how strongly do you feel for this person?’’ on a 1–9 Likert scale). Depressive symptoms were measured at each visit with the clinician‑rated Montgomery‑Åsberg Depression Rating Scale (MADRS) and the self‑report Beck Depression Inventory (BDI). Analysis followed an intention‑to‑treat (ITT) approach including all participants with at least one post‑treatment empathy or efficacy assessment. MET scores were not imputed; missing MADRS/BDI scores were handled using last observation carried forward to align with the primary efficacy analysis. Mixed‑effects analyses of variance were used for MET subscales with Greenhouse‑Geisser corrections where necessary; p‑values for ANOVA interaction terms were adjusted using the Benjamini‑Hochberg method. Post‑hoc t‑tests were adjusted for three comparisons, effect sizes reported as Cohen's d, and Pearson correlations (with Holm correction for multiple tests) and Fisher's z‑tests were used to assess associations between changes in empathy and changes in depressive symptoms. Analyses were performed in R and visualised with standard graphics packages.

Of 1152 individuals screened by mail or phone, 68 were invited for on‑site medical screening and 55 met criteria and were enrolled. Three withdrew before randomisation (COVID‑related logistics), leaving 52 participants who received study drug; one additional participant in the psilocybin arm withdrew before completing the first MET follow‑up, yielding an analysed sample of 51. Two patients were lost to follow‑up (one at +8 days, one at +14 days), and several MET assessments were excluded for technical issues (one at +2 days, four at +8 days, three at +14 days). Demographic and fuller baseline details are reported in the main table and supplement. Baseline MET scores were in a similar range to prior reports for depressed and healthy samples: cognitive empathy mean = 0.67; explicit emotional empathy mean for positive stimuli = 4.2 and for negative stimuli = 5.3. Mixed‑effects ANOVA revealed a significant interaction between treatment condition and visit for explicit emotional empathy (F(2.43,102) = 5.83; P = 0.006; generalized eta squared η2G = 0.01), whereas implicit emotional empathy (F(2.21,92.7) = 2.57; P = 0.11; η2G = 0.005) and cognitive empathy (F(3,126) = 2.39; P = 0.22; η2G = 0.013) did not show significant interactions. Post‑hoc testing indicated significant differences between psilocybin and placebo for explicit emotional empathy at each post‑treatment visit (P < 0.05). The largest numerical group difference occurred at +8 days: mean difference 1.12 points (95% CI [0.13, 2.11]; P = 0.027) with Cohen's d = 0.66. Analyses stratified by stimulus valence showed that the increase in explicit emotional empathy after psilocybin was driven by responses to positive stimuli across time‑points; no comparable effect was found for negative stimuli. When examining associations between changes in explicit emotional empathy and depressive symptoms, the placebo group showed significant correlations at +8 days on MADRS and at +14 days on BDI (both P < 0.05, corrected), whereas the psilocybin group did not show significant associations (all P > 0.05, corrected). Fisher's z‑tests comparing correlations between groups were non‑significant after correction.

Jungwirth and colleagues interpret their findings as evidence that a single dose of psilocybin, delivered within a brief preparatory and integrative psychological support framework, produces sustained increases in explicit emotional empathy in patients with mild to moderate major depression for up to 14 days. The effect was valence specific, occurring for positive but not negative stimuli, and persisted across the three post‑treatment assessments. The authors link the absence of change for negative stimuli to the baseline ‘‘negativity bias’’ commonly seen in depression, where patients are more responsive to negative emotional content. The investigators note that prior studies in healthy volunteers reported acute and subacute increases in emotional empathy after psilocybin and LSD, sometimes across both valences; however, the present placebo‑controlled clinical sample showed selective enhancement for positive emotions. They caution that, in this study, increased empathy did not correlate with improvements in depressive symptoms in the psilocybin group, so the data do not support the claim that empathy change is a necessary mediator of clinical benefit. The significant correlations observed in the placebo group indicate a complex, potentially bidirectional relationship between mood and empathy that merits further investigation. Mechanistic speculation offered by the authors includes psilocybin‑related modulation of limbic circuitry and altered functional connectivity within default mode network nodes implicated in self‑other distinction, which could plausibly increase receptivity to others' positive emotions. Clinically, enhanced emotional responsiveness to positive experiences might bolster therapeutic alliance, social functioning and group therapy dynamics, representing a potential resource‑efficient adjunct in psychotherapeutic contexts. The authors also suggest applicability to other disorders with impaired empathy. Several limitations are emphasised: empathy outcomes were secondary and exploratory, sample size was modest, follow‑up was limited to two weeks, measures were collected after only one preparation session, and the sample was restricted to mild–moderate depression limiting generalisability to severe cases. Additional methodological limitations include the lack of an independent rater for the MADRS, absence of measures of expectancy or blinding success, and technical losses of some MET data. Taken together, the authors recommend further controlled studies with longer follow‑up, broader clinical samples, additional control stimuli (for example neutral images) and designs that can better disentangle valence processing from general mood change and test mediation between empathy change and clinical outcomes.