To our knowledge, this is the first data report on the effects of psilocybin therapy in AN in a clinical research trial. This open-label pilot study examined the safety and tolerability of administering psilocybin therapy to participants with AN and pAN. We chose to include participants in partial remission because we were most interested in exploring potential changes in core ED psychopathology (versus weight), which can lead to treatment resistance and which can persist after weight restoration35. Additionally, pAN has been shown to be common, severe, persistent and undertreated3,35.
Psilocybin therapy, which includes psychological support by trained therapists, was found to be safe and well tolerated for the 10 participants who received treatment in this study. Most participants endorsed the treatment as highly meaningful and the experience as a positive life impact, and yet effects on ED psychopathology were highly variable, with a subset of participants demonstrating a robust response for a single-dose treatment. Results of this study are preliminary and inconclusive given its size and design. In this section, we discuss study findings related to primary outcomes, patient acceptability and qualitative perceptions as well as ED-specific psychopathology.
No participants in our study experienced any serious AEs, and all treatment-emergent AEs resolved within 24 h and without intervention (with the exception of one report of orthostatic heart rate that was reported at 3-month follow-up for one participant). Hypoglycemia occurred in two participants on the dosing day and resolved in 24 h. We hypothesize that this was related to a prolonged period of fasting on the dosing day (a common effect of psilocybin) versus any direct relationship to the drug, given the state of malnutrition and low carbohydrate stores associated with AN. Food was available on request to participants during the dosing day, but participants were not required to eat during the therapeutic experience. To our knowledge, there are no reports of psilocybin-induced hypogylcemia. However, individuals with AN often have reduced plasma glucose levels36. The incidence of hypoglycemia is clinically important and may indicate that attention to blood glucose levels before and after intervention may be warranted in participants with compromised nutritional status given the dangers associated with hypoglycemia in AN, including sudden death36. Both incidents of hypoglycemia occurred in participants who were given a standardized breakfast upon arrival. AN has a high prevalence of serious medical complications and physiological disturbances, which account for much illness-related death. The lack of negative incidences regarding safety in our study is promising for future research with the AN population; however, larger studies with a more diverse participant group continue to be needed to determine safety37.
Most participants self-reported positive changes 3 months after the psilocybin dosing. That the treatment was regarded as beneficial by most participants and that there were no dropouts are promising signs of engagement. Dropout rates in currently available treatments tend to be high38. Positive patient perceptions are also notable because they may suggest improved quality of life, which is clinically important for those with a potentially severe and enduring illness38. These self-reported data suggest that most participants perceived some benefit that may have been ED-non-specific in nature or not well captured by our assessment measures. Although our treatment included AN-specific therapeutic elements, adjunctive therapy was brief. Additional therapeutic methods, or extended therapeutic time, may be a useful consideration to facilitate further behavioral change and increased specificity, as has been employed in other psilocybin studies39,40. Our effect sizes and response rates were less robust than those reported for primary outcomes in psilocybin studies for other psychiatric disorders41,42,43. This may also be related to a heterogenous sample, a single-dose trial (compared to a two-dose design used in other studies), a lack of sensitivity in assessment measures or unique/specific features of AN that are not as readily addressed by psilocybin therapy or that require dosing adjustments. AN is a difficult-to-treat disorder with a complex physiology and physical recovery needs that differentiate it from other mental illnesses. Notably, 90% of participants reported that one dosing session was not enough, suggesting that an additional psilocybin experience(s) may be beneficial44.
Our exploratory analysis showed some significant reductions in ED-related psychopathology when the sample was aggregated; however, the results were highly variable among participants. Forty percent (4/10) of participants demonstrated clinically significant reductions34 in ED psychopathology (EDE) at 3-month follow-up, with scores decreasing from clinical ranges to within 1 s.d. of community normative values34. Within the responder group, ED psychopathology decreased precipitously and dropped below clinical levels within the month after the psilocybin dosing session. Three of the responders were not enrolled in any concurrent mental health services during the study period but had mental health treatment histories, and one was in longstanding, outpatient therapy that did not change during study enrollment. Symptoms continued to improve between 1-month and 3-month follow-up. Given the size and design of this study, these effects are preliminary and inconclusive. However, it is notable that we found such a robust response in a subset of participants in a single-dose trial of psilocybin delivered over a brief time period, because currently available treatments for adult AN result in only modest improvements in symptoms and often focus on weight and nutritional rehabilitation without adequately addressing underlying psychopathology38. Participants also experienced significant reductions in anxiety; however, mean changes in depression scores were not significant. Changes in general psychopathology may partially explain the effects on ED symptoms44.
We did not observe a significant effect on BMI over time, and results were highly variable among participants. BMI did not follow the same change trajectory as ED psychopathology for participants who showed reductions on core psychopathology. Of the four treatment responders, two showed positive BMI trends, one remained stable at a normal BMI and one showed a two-point decrease over time. It is also possible that a longer follow-up period is necessary to observe meaningful changes in BMI, which has been suggested by ED experts45,46. Notably, there are well-documented phenomena associated with AN that impede upon weight rehabilitation, including hypermetabolism and unusually high caloric requirements37. When queried about the lack of weight change, one treatment responder stated, ‘The irony is that now that I want to recover I can eat intuitively, but that is not enough to support physical recovery’. These findings may suggest that targeted nutritional rehabilitation emblematic of traditional treatment may be a necessary adjunctive treatment even when significant improvements in ED psychopathology are conferred.
Limitations of this study include its small sample size, the lack of a control comparison and the open-label design. Owing to the exploratory nature of this study, we did not conduct a power analysis or correct for multiple comparisons. As a result, these findings are not conclusive and should be interpreted with caution. Additionally, all of the participants were self-referred, which may have resulted in a selection bias. Similarly, suggestibility and expectations of positive outcomes related to positive popular media coverage surrounding psychedelics, and attending treatment at a well-reputed ED service (particularly for those who were nonlocal), may have resulted in a suggestibility that influenced positive outcomes. Our sample included a diverse range of severity; however, many participants had mild to moderate AN. More research is needed to evaluate psilocybin therapy for severe presentations. The study also lacked gender, racial and cultural diversity.
Strengths of this study included administration of a precise dose of pure synthesized psilocybin and the evaluation of a highly novel treatment modality for a treatment-resistant population. Larger, adequately powered, well-controlled trials with comparator treatments are needed to evaluate the efficacy of psilocybin therapy in a diverse sample of patients with AN. Future studies should further investigate mechanisms of action and moderators of treatment to discern how psilocybin may lead to changes in AN and whether psilocybin therapy is more suitable and effective for a specific subset of AN. Additionally, future studies are needed for optimization to identify adequate dosage, identify the optimal number of psilocybin administrations and investigate the need for possible adjunctive treatments that could optimize treatment outcomes.
In conclusion, results from this open-label, single-arm study suggest that psilocybin therapy is safe and tolerable in participants with AN; however, adequately powered, randomized controlled trials are needed to draw any conclusions.