Effect of psilocybin on empathy and moral decision-making

Empathy and moral behaviour are central to social functioning and are implicated in several psychiatric disorders. Empathy is described as having at least two components: cognitive empathy (understanding another's mental state, akin to Theory of Mind) and emotional empathy (sharing another's emotional state). Deficits in these domains are reported in conditions such as major depressive disorder, bipolar disorder, borderline personality disorder and psychopathy, and may worsen with illness progression. Moral decision-making is commonly studied using hypothetical moral dilemmas that pit utilitarian outcomes against deontological constraints; personal dilemmas that involve directly harming another person are more emotionally engaging and tend to elicit deontological responses. Serotonin (5-HT) has been implicated in both empathy and moral processes, but the specific contributions of different 5-HT receptor subtypes remain unclear. Previous pharmacological work showed that MDMA increases emotional but not cognitive empathy, and that manipulating 5-HT tone with SSRIs can increase harm aversion in moral judgements, suggesting receptor-specific effects worth clarifying. Pokorny and colleagues set out to test the acute effects of psilocybin — a serotonergic hallucinogen acting primarily at 5-HT2A and also 5-HT1A receptors — on multiple facets of empathy and on hypothetical moral decision-making in healthy volunteers. The principal hypotheses were that psilocybin would impair cognitive empathy for negative stimuli while increasing emotional empathy, and that it would reduce utilitarian choices in personal moral dilemmas. The study therefore examined cognitive empathy, explicit and implicit emotional empathy using the Multifaceted Empathy Test (MET), and moral choice behaviour using a moral dilemma task (MDT), in a double-blind, placebo-controlled, within-subject design.

Thirty-three healthy volunteers were recruited; after exclusions one participant was removed for failing to understand the MET, leaving 32 participants (17 men, 15 women; mean age 26.72 ± 5.34 years) for the MET analyses. The MDT was implemented later and completed by 24 participants (13 men, 11 women; mean age 26.63 ± 5.33 years). Eligible participants were aged 20–40, screened by physical exam, ECG, blood and urine tests (including drug screen and pregnancy test), and structured psychiatric interviews to exclude current or past neurological or psychiatric disorders and first-degree relatives with major psychiatric illness. Participants agreed to abstain from illicit drugs for at least 2 weeks prior to and during the study. The study used a double-blind, randomised, placebo-controlled, within-subject cross-over design with two sessions separated by at least 10 days. Each participant received oral psilocybin at 0.215 mg/kg in one session and matched placebo capsules (mannitol) in the other. Tasks (MET and MDT) were administered on a computer in a quiet room 160 minutes after ingestion, when the peak perceptual effects had largely subsided. Subjective effects were assessed with the 5D-ASC (5-Dimensional Altered States of Consciousness) at 360 minutes post-dose and mood with the Positive and Negative Affect Schedule (PANAS) before and 360 minutes after dosing. Participants were monitored by a physician until effects resolved. Measures: Trait empathy was assessed once at screening with the Interpersonal Reactivity Index (IRI). The MET (40 photos; 20 positive, 20 negative) yielded cognitive empathy (identify the emotion from four choices), explicit emotional empathy (self-rated concern or happiness on a 9-point scale), and implicit emotional empathy (self-rated arousal on a 9-point scale). The MDT comprised two matched sets of 22 vignettes including personal (avoidable and inevitable), impersonal and nonmoral dilemmas; responses were binary choices of utilitarian action versus harm avoidance. The 5D-ASC provided multiple subscale scores including changed meaning of percepts. PANAS measured positive and negative affect and seven arousal-related items. Statistical approach: Repeated-measures ANOVAs tested drug (psilocybin, placebo) and valence or dilemma category as within-subject factors, with order (placebo first, psilocybin first) as a between-subjects factor. ANCOVAs controlled for mood changes where relevant. Paired t tests compared nonmoral scenario accuracy between conditions. Regression analyses (backward stepwise) examined whether 5D-ASC scales, PANAS change scores, or IRI subscales predicted empathy changes; moderator analyses tested whether arousal moderated relationships. Analyses used a significance threshold of P < .05 (two-tailed). PANAS data were missing for one participant and noted as such.

Subjective effects and mood: Psilocybin produced robust alterations on the 5D-ASC, with a significant drug by scale interaction (F(10,310) = 25.06, P < .0001) and a main effect of drug (F(1,31) = 100.66, P < .0001). Posthoc tests showed that psilocybin increased all 5D-ASC scale scores relative to placebo except for anxiety. On the PANAS there was a drug × time × scale interaction (F(1,30) = 6.58, P < .05) and a drug × time interaction (F(1,30) = 17.34, P < .001); positive affect increased under psilocybin (P < .05) while negative affect did not change. Arousal increased under psilocybin (psilocybin mean = 0.25, SD = 0.39) relative to placebo (mean = -0.07, SD = 0.32; t(31) = -3.96, P < .001). MET outcomes: For explicit emotional empathy there was a significant main effect of drug (F(1,30) = 7.74, P < .01), with higher ratings under psilocybin than placebo; this effect did not interact with stimulus valence. Implicit emotional empathy was also increased by psilocybin (F(1,30) = 4.77, P < .05), again without a valence interaction. Cognitive empathy showed no significant main effect of drug (F(1,30) = 1.45, P > .2), indicating no psilocybin-induced impairment in identifying others' emotions. A valence main effect indicated more errors for negative versus positive stimuli (F(1,30) = 6.58, P < .05). Results remained broadly unchanged when controlling for positive or negative mood, except implicit emotional empathy lost significance when controlling for positive mood (F(1,30) = 2.62, P > .11). No differences were found between psilocybin-experienced and psilocybin-naïve participants on MET scores in the psilocybin condition. Associations: Multiple regression showed that 5D-ASC scales explained a significant portion of variance in the psilocybin-induced increase in implicit emotional empathy (F(1,31) = 11.23, P < .01; R2 = .27), with the changed meaning of percepts subscale as the sole significant predictor (Beta = .52, t(31) = 3.35, P < .01). Moderator analyses testing arousal as a moderator found that arousal did not significantly predict implicit emotional empathy nor moderate the relationship between changed meaning of percepts and implicit emotional empathy; the key predictor remained changed meaning of percepts. Trait empathy (IRI) and PANAS change scores did not significantly predict empathy increases in these models. MDT outcomes: No drug effects emerged on moral decision-making. There was no significant drug × category interaction (F(2,44) = 0.97, P > .3) and no main effect of drug on the proportion of utilitarian choices (F(1,22) = 0.37, P > .5). However, participants made more incorrect responses on nonmoral dilemma scenarios under psilocybin (placebo mean = 0.99, SD = 0.05; psilocybin mean = 0.91, SD = 0.16; t(23) = 2.98, P < .01), though the absolute error rate under psilocybin was low (about 9%). Including mood covariates did not alter the MDT results, and no differences emerged between psilocybin-experienced and naïve participants on MDT performance.

Pokorny and colleagues interpret the findings as evidence that a single moderate dose of psilocybin selectively enhances emotional empathy (both explicit concern and implicit arousal-related empathy) without altering cognitive empathy or hypothetical moral choice. The increase in implicit emotional empathy was associated with the 5D-ASC subscale changed meaning of percepts, which assesses heightened significance and emotional engagement with objects or surroundings; the authors suggest this altered sense of significance may also extend to perceiving and emotionally engaging with other people. Trait empathy did not predict the empathy enhancement, implying the effect may operate independently of baseline empathic disposition. Although psilocybin elevated arousal and positive affect, arousal did not statistically moderate the link between changed meaning of percepts and implicit emotional empathy, and mood changes did not account for most empathy effects, although implicit empathy lost significance when controlling for positive mood in one analysis. The authors situate these results within serotonergic receptor mechanisms: psilocybin's primary action at 5-HT2A receptors and partial modulation via 5-HT1A receptors might underlie the empathy enhancement. They note parallel findings with MDMA and LSD on emotional empathy and point out that prior MDMA work implicated 5-HT mechanisms but did not implicate 5-HT1A in MDMA's empathogenic effects; by contrast, psilocybin's combined 5-HT2A/1A action makes both receptor subtypes plausible contributors. The discussion also references neuroimaging work linking emotional empathy to regions such as the insula and amygdala and notes that psilocybin increases activity in frontomedial, frontolateral and temporal cortices and the insula, providing a possible neural substrate for increased emotional sharing. Regarding moral decision-making, the lack of effect on utilitarian choices is emphasised. The authors note that moral judgement and choice of action may be dissociable processes with different neural bases; prior SSRI work that increased harm aversion assessed moral judgement rather than action choice, which might explain divergent results. They also acknowledge alternative explanations: psilocybin might require higher doses to affect the networks underlying moral choices, or reduced emotional salience of personal dilemmas under psilocybin (via lowered amygdala or ACC responses) could obscure changes. The small increase in errors on nonmoral items under psilocybin might reflect mild attentional effects or tiredness given task order, and the authors concede such factors could have masked potential MDT effects. Limitations and implications discussed by the authors include that psilocybin acts at multiple 5-HT receptor subtypes so receptor-specific conclusions remain tentative; they call for studies combining psilocybin with selective antagonists to parse receptor contributions. The clinical relevance is highlighted cautiously: because emotional empathy is impaired in disorders such as depression and psychopathy, 5-HT2A/1A receptor agonists might have therapeutic potential for social cognition deficits, but further research is required. The authors also note that the empathy enhancement was not simply attributable to visual hallucinations or perceptual disturbances, given the lack of association with visual hallucination measures.