A six-week, double-blind trial comparing psilocybin with the antidepressant escitalopram in 59 patients with long-standing moderate-to-severe depression has found both treatments reduce the brain’s tendency to interpret ambiguous social cues negatively. Conducted by researchers including Marieke A. G. Martens of the University of Oxford, the study also involved Robin Carhart-Harris and Catherine J. Harmer and was published in Translational Psychiatry. The trial used a facial expression recognition task to track changes in emotional processing.
Participants were randomly allocated to two groups. One received two 25-milligram doses of psilocybin three weeks apart, alongside a daily inactive placebo for six weeks. The other group took daily escitalopram for six weeks and also received two tiny one-milligram psilocybin doses as an active placebo during the dosing sessions to maintain blinding. Both groups received identical psychological support throughout the trial, including preparatory and integration sessions with therapists.
The researchers employed a computer-based facial expression recognition task, where faces displaying six basic emotions were shown for half a second and gradually intensified. Participants had to identify the emotion quickly and accurately, with the task repeated at the six-week mark.
At baseline, participants exhibited a negative affective bias, correctly identifying negative emotions while struggling with positive ones. After six weeks, both treatment groups showed a reduction in this bias, with no statistically significant difference between the two. Across both groups, participants became less accurate at recognising negative emotions compared with baseline, but they made fewer errors misclassifying faces as negative. Reaction times improved for recognising both negative and positive emotions, and by the end positive emotions were recognised faster than negative ones.
Brain imaging from earlier phases of the same patients showed different patterns for the two drugs. Escitalopram reduced activity in the amygdala when participants viewed faces, a response not observed with psilocybin. Despite these differing neural effects, the behavioural outcomes on the facial expression recognition task were similar.
When the researchers looked for links between cognitive shifts and overall depression symptoms, the six-week reduction in negative affective bias did not predict lower depression scores at six weeks for either group.
In a follow-up at ten weeks, the escitalopram group showed a link: a smaller tendency to misclassify positive faces as negative at six weeks predicted lower depression scores at ten weeks. This finding supports the theory that SSRIs progressively improve mood as patients engage with the world through a less negative lens. The psilocybin group did not show the same relationship.
The researchers acknowledged several limitations. The trial did not include a purely inactive placebo group, complicating distinctions between the drugs’ chemical effects and the psychological support provided. There may also have been a sampling bias, with volunteers more inclined to receive psilocybin potentially influencing responses and symptom reporting.
Future work will aim to include active placebo controls to improve blinding in psychedelic research and to assess emotional processing earlier in treatment, potentially within days of psilocybin administration.
Overall, the study concludes that both treatments shifted how patients interpreted emotional information, revealing a shared cognitive outcome even as their neurobiological pathways differed. The authors of the paper are Marieke A. G. Martens, Bruna Giribaldi Cunha, David Erritzoe, David Nutt, Robin Carhart-Harris, and Catherine J. Harmer.
