Reduced Brain Responsiveness to Emotional Stimuli With Escitalopram But Not Psilocybin Therapy for Depression

Major Depressive Disorder (MDD) is common and disabling, and current guidelines recommend psychotherapy for mild cases and pharmacotherapy or combination treatments for more severe cases. Selective Serotonin Reuptake Inhibitors (SSRIs) are widely used but have limitations: only moderate efficacy, a 4–8 week onset of effect, and substantial discontinuation rates. A frequently reported phenomenon with SSRIs is ‘‘emotional blunting’’—a reduced range or intensity of emotional experience—which has been linked in prior work to diminished limbic responsiveness, particularly in the amygdala. By contrast, early clinical trials of psilocybin therapy for depression report rapid, large symptomatic improvements and subjective reports of increased emotional connection; psilocybin acts primarily as a 5-HT2A receptor agonist and has been associated with acute and subacute changes in brain function and connectivity in both healthy and depressed samples.

The fMRI task was an eight-minute blocked emotional face paradigm (fear, happy, neutral, and rest blocks) with passive viewing. Imaging used a Siemens 3T scanner with multiband EPI (TR=1250 ms, 3 mm isotropic voxels); anatomical scans used standard MPRAGE parameters. Pre-processing and statistical analysis were performed with FSL: motion correction, 6 mm smoothing, registration to MNI152, high-pass filtering, and first-level GLMs modelling fear, happy and neutral blocks convolved with a gamma haemodynamic response. An extended set of 24 motion regressors was included. Contrasts compared each face type to fixation and an all-faces summary contrast. Group-level random-effects analyses used FLAME-1 with cluster correction (Z>2.3, p<0.05). Additional planned analyses included an anatomical bilateral amygdala mask and exploratory correlational and moderation analyses relating BOLD changes to clinical outcomes.

In the psilocybin group there was minimal post-treatment change overall, with a significant increase in activation to neutral faces on the post-therapy visit (t[24]=-3.17, p=0.004), a result that survived the corrected alpha threshold. Amygdala-focused analyses showed a reduction in response to fear faces in the escitalopram group (t[20]=2.33, p=0.031) but this did not survive correction for multiple comparisons across the amygdala tests. Baseline comparisons confirmed no significant pre-treatment differences between groups in task-evoked BOLD. Exploratory correlations between region-wise BOLD changes and clinical change scores were non-significant. Moderation analyses yielded different patterns by group: in the psilocybin arm, change in emotional function (LEIS) predicted QIDS change (Z=-2.24, p=0.025) while brain-activity change did not; in the escitalopram arm, brain-activity change predicted BDI outcomes (Z=1.97, p=0.048) and the interaction between brain change and LEIS was strongly significant (Z=3.14, p=0.002), suggesting the brain–clinical relationship depended on degree of emotional blunting.

Key limitations acknowledged by the investigators include the timing difference between groups at the post-treatment scan (escitalopram scans coincided with acute drug presence while psilocybin scans occurred three weeks after the final dosing session), which could mean transient post-psilocybin effects were missed. They also note that the emotional-face paradigm may not be the most sensitive marker of psilocybin's antidepressant action on this timescale and that some amygdala findings did not survive correction for multiple comparisons. The authors conclude that their fMRI results are consistent with existing theories of SSRI action and support a different neural mechanism for psilocybin therapy, with implications for understanding how these two treatment modalities produce antidepressant effects.