Psilocybin-induced decrease in amygdala reactivity correlates with enhanced positive mood in healthy volunteers

Kraehenmann and colleagues frame the study around the amygdala's central role in serotonergic circuits for emotion processing and its consistent hyperactivity in major depression. Earlier research indicates that modulation of serotonin neurotransmission—such as with selective serotonin reuptake inhibitors (SSRIs)—reduces amygdala hyperreactivity and is associated with shifts toward more positive mood. Psychopharmacological and clinical evidence has further suggested that the hallucinogen psilocybin, a 5-HT1A/2A/2C agonist, can acutely induce positive mood changes and alter neural responses to emotional stimuli, but whether these effects include modulation of amygdala reactivity and how such modulation relates to mood change remained uncertain. This study set out to determine whether a single acute oral dose of psilocybin (0.16 mg/kg) reduces amygdala reactivity to negative stimuli in healthy volunteers and whether any change in amygdala activity relates to changes in mood. Using a placebo-controlled, double-blind, randomised cross-over design, the investigators combined task-based fMRI during an established amygdala reactivity paradigm with validated mood questionnaires to connect neural effects with subjective affective changes. The authors hypothesised that psilocybin would decrease amygdala reactivity to negative stimuli and increase positive mood state.

Twenty-five healthy, right-handed volunteers (16 males; mean age 24.2 ± 3.42 years), mostly students or university-educated, were recruited and screened by medical history, physical exam, routine blood tests, ECG and urine drug/pregnancy screens. Most participants reported no prior hallucinogen use. The study used a randomised, double-blind, placebo-controlled cross-over design: each subject received oral psilocybin (0.16 mg/kg) and placebo in two imaging sessions separated by at least 14 days. Mood measures were collected before and 210 minutes after each administration using the Positive and Negative Affect Schedule (PANAS) and the state portion of the State-Trait Anxiety Inventory (STAI). Ethics approval and informed consent were obtained; additional screening and procedural detail were reported in supplementary material (not included in the extract). During fMRI, participants performed a modified amygdala reactivity task composed of alternating blocks of emotional picture discrimination (negative and neutral pictures) interleaved with shape discrimination baseline blocks. A simple motor task was also included to test for nonspecific global drug effects on BOLD signals. BOLD data were analysed in SPM12b. The authors defined left and right amygdala anatomical ROIs using the Automated Anatomical Labeling atlas via the WFU PickAtlas tool and conducted a second-level voxel-wise repeated-measures ANOVA with drug (psilocybin, placebo) and emotion (negative vs shapes; neutral vs shapes) as within-subject factors, followed by paired t-tests for planned comparisons. For the amygdala ROI the a priori significance threshold was p < 0.05 family-wise error (FWE) corrected for the amygdala (small volume correction), with an initial voxel-level threshold of p < 0.001 and extent threshold k = 0 voxels. Parameter estimates were extracted from left and right amygdala ROIs for each emotion condition and session and analysed with repeated-measures ANOVA (factors: emotion, laterality, drug) and Bonferroni-corrected paired t-tests. Additional analyses extracted baseline (shape) BOLD to test whether psilocybin increased baseline amygdala activity. An exploratory whole-brain ANOVA (drug × emotion) used a whole-brain FWE threshold of p < 0.05. Finally, Pearson correlations and a multiple regression investigated relations between right amygdala change (psilocybin minus placebo) and five rating change scores (PANAS positive, PANAS negative, STAI state anxiety, and two Altered States of Consciousness imagery subscores); predictors were mean-centred and diagnostic checks were performed. Specific image acquisition parameters and further preprocessing details were referenced to the Supplementary Material and are not present in the extracted text.

Behaviorally, psilocybin produced a pronounced increase in positive affect on the PANAS (Bonferroni-corrected p = 0.001), while PANAS negative affect and STAI state anxiety showed no significant change (Bonferroni-corrected p = 0.87 and p = 0.37, respectively). In the voxel-wise whole-brain ANOVA there was a significant main effect of drug localised to the right amygdala (peak MNI 27, -4, -19; F1,72 = 27.25; Z = 4.25; FWE-corrected p < 0.001). No significant drug × emotion interaction survived correction. Planned paired t-tests showed that psilocybin significantly attenuated right amygdala activation to negative pictures (peak 24, -4, -22; Z = 4.38; FWE-corrected p = 0.001) and to neutral pictures (Z = 4.60; FWE-corrected p < 0.001). ROI-based analyses corroborated a significant main effect of drug (F1,24 = 19.45; p < 0.001) without a drug × emotion interaction (F1,24 = 0.29; p = 0.59), and revealed a drug × side interaction (F1,24 = 6.24; p < 0.05) indicating a preferential reduction of right over left amygdala activation. Extracted parameter estimates illustrate the effect sizes: for the negative vs shapes contrast the right amygdala mean BOLD parameter estimate was 0.36 ± 0.28 under psilocybin versus 0.58 ± 0.23 under placebo (p < 0.001). The left amygdala showed a smaller reduction to negative pictures (psilocybin 0.49 ± 0.30; placebo 0.62 ± 0.26; p < 0.05). There were no significant differences between psilocybin and placebo in amygdala activation during the baseline shape task for either right (t = 0.05, p = 0.96) or left (t = -1.20, p = 0.24) amygdala, indicating that the observed reductions were not driven by increased baseline activity. The separate motor task activated primary motor cortex in both placebo (-39, -22, 65; Z = 7.03; FWE p < 0.001) and psilocybin (-36, -19, 53; Z = 6.70; FWE p < 0.001) sessions, with no significant difference between sessions. Complementary ROI tests in motor cortex were non-significant (t = 0.36, p = 0.72). A Savage-Dickey Bayes factor t-test supported the null for motor cortex effects: the null hypothesis of no psilocybin effect was six times more probable than the alternative. Whole-brain paired comparisons additionally revealed psilocybin-related decreases in bilateral occipital gyri, lingual gyrus, fusiform gyrus and temporal gyri (all FWE-corrected p < 0.05). Follow-up extraction showed these regional decreases were driven by reduced responses to negative stimuli (all Bonferroni-corrected p < 0.04) rather than by increased activation to the baseline shapes condition (all Bonferroni-corrected p > 0.44). Critically, changes in right amygdala reactivity (psilocybin minus placebo) were inversely correlated with changes in positive affect (psilocybin minus placebo) (r = -0.46, p < 0.05), such that greater attenuation of amygdala activation was associated with larger increases in positive mood. None of the other rating variables correlated with amygdala change (all p > 0.1). A multiple regression including the five rating change scores confirmed positive affect as the sole significant predictor of right amygdala BOLD change (model F = 5.44, p = 0.03, adjusted R2 = 0.17; positive affect beta = -0.46, t = -2.33, p = 0.03).

The investigators interpret their findings as evidence that acute psilocybin administration markedly reduces task-induced amygdala activation during emotional picture processing and concomitantly increases positive mood in healthy volunteers. The right amygdala showed a stronger drug-related attenuation than the left, and the observed neural effect was specifically associated with increases in positive affect. These data are presented as consistent with a mechanistic framework in which serotonergic modulation—likely via 5-HT2A receptor stimulation—can reduce limbic reactivity and thereby relate to improved mood. The authors note that no drug × emotion interaction was found and that psilocybin reduced amygdala responses to both negative and neutral images. They caution that this pattern does not support a valence-specific effect and suggest the blocked-design fMRI method, which has limited temporal resolution compared with electrophysiology, might have missed transient valence-specific effects reported in EEG studies. Lateralisation of the effect to the right amygdala is discussed in the context of prior evidence showing preferential right-sided amygdala modulation by serotonergic agents; however, the authors acknowledge that lateralisation findings across studies remain divergent and inconclusive. Beyond the amygdala, whole-brain decreases were observed in visual cortical regions. The researchers discuss the possibility that psilocybin-induced visual perceptual alterations and changes in baseline excitability might interact with stimulus-induced responses in visual cortex, but they emphasise that in this dataset the decreases were driven by reduced responses to the negative stimuli rather than by increased baseline activity. The authors propose that reduced amygdala output to visual cortex could contribute to the diminished stimulus-induced visual responses, and they recommend future connectivity studies to investigate amygdala–visual cortex interactions. At the neurochemical level, the discussion highlights the likely role of serotonergic mechanisms: psilocybin (via its active metabolite psilocin) acts at 5-HT2A receptors and also engages 5-HT1A/2C receptors, which are present on GABAergic interneurons in the amygdala and can indirectly inhibit postsynaptic firing. The authors therefore suggest that increased serotonergic tone in the amygdala may underlie the observed inhibition of amygdala reactivity, while acknowledging that precise synaptic mechanisms remain to be clarified. They note that 5-HT2C contributions are less likely given the absence of anxiety changes and prior pharmacological findings. Finally, the authors position their results within translational and clinical contexts: reductions in amygdala reactivity are a putative biomarker of antidepressant action, and the present acute effects of psilocybin on amygdala activity and positive mood are consistent with rapid antidepressant-like properties reported in recent clinical trials. Nonetheless, they caution that responses in healthy volunteers may not generalise to patients with depression and call for clinical studies to directly assess psilocybin’s effects on mood and amygdala reactivity in depressed populations. Key limitations acknowledged include the inability of the blocked fMRI design to resolve fast valence-specific dynamics and the inconclusive nature of lateralisation findings; the authors also recommend further connectivity and receptor-specific studies to elucidate mechanisms.