Psilocybin therapy for females with anorexia nervosa: a phase 1, open-label feasibility study

Anorexia nervosa (AN) is a severe, often chronic and deadly psychiatric disorder with substantial morbidity and mortality, including an estimated 18-fold increase in mortality and elevated suicide risk. Despite these consequences, adult AN lacks proven treatments that reliably reverse core psychopathology and has no approved pharmacological therapies; fewer than half of patients recover, relapse rates approach 50% and about 20% develop a chronic course. Previous research indicates altered serotonergic function in AN and altered 5-HT2A receptor activity, and contemporary studies of psilocybin suggest that a single administration can produce rapid and enduring neural and psychological changes via serotonergic (and downstream dopaminergic and glutamatergic) mechanisms. This raises the possibility that psilocybin therapy, when paired with psychological support, might disrupt rigid, ego-syntonic patterns and improve emotional processing and cognitive flexibility relevant to AN symptoms. Peck and colleagues report a Phase I, open-label feasibility study that aimed to evaluate the safety, tolerability and acceptability of a single 25 mg dose of pharmaceutical-grade synthetic psilocybin (COMP360) administered with psychological support to adult females meeting DSM-5 criteria for AN or partial remission (pAN). In addition to primary safety endpoints, the investigators explored preliminary signals of efficacy on eating-disorder-specific psychopathology and related measures over a 3-month follow-up, to inform the design of future controlled trials.

This was an open-label pilot study conducted at a single academic clinical research institute. Ten female adults aged 18–40 who met DSM-5 criteria for AN (including restricting and binge–purge subtypes) or pAN were enrolled. Recruitment ran from April to December 2021; 158 people initially expressed interest and the final sample comprised self-referred participants. Key medical exclusions included BMI < 16 kg m-2, medical instability, recent cardiovascular disease and pregnancy. Psychiatric exclusions included current or past psychotic disorder, bipolar disorder, recent substance use disorder, substantial suicide risk and certain personality disorder screens. Participants on serotonergic medications underwent supervised washout before dosing. Intervention consisted of a single oral 25 mg dose of COMP360 (five 5 mg capsules of synthetic psilocybin) administered in a therapeutic setting with two psychologists present. Preparatory psychotherapy comprised two sessions with the lead study psychologist in the 2 weeks before dosing; integration sessions occurred on day 1 and day 7, with an additional integration session at the day-1 visit lasting 60–90 minutes. Participants remained at the clinic for 8 hours after ingestion and completed an altered-states rating (5D-ASC) before discharge once baseline had returned. For participants 6–10 the team introduced a standardised breakfast on arrival to control for hypoglycaemia. Follow-up assessments were conducted the morning after dosing (day 1), at 1 week, 1 month (primary psychological endpoint) and at 3 months (the 3-month visit was telehealth for seven participants). Primary safety assessments included adverse event (AE) monitoring, changes in vital signs, electrocardiograms (ECGs), clinical laboratory tests and suicidality (Columbia Suicide Severity Rating Scale, C-SSRS), measured at baseline (day –1), day 1 and 1-week follow-up. Secondary and exploratory efficacy measures (n = 10) included the Eating Disorder Examination (EDE, clinician interview; primary efficacy instrument), EDE-QS, Eating Disorders Inventory (EDI), BMI, body-image scales (PASTAS, BISS), anxiety (STAI-T), preoccupations/rituals (YBC-EDS), functional impairment (Clinical Impairment Assessment, CIA), depressive symptoms (QIDS-SR), readiness/motivation (RMQ), and patient acceptability items modelled on prior psychedelic studies. Self-report psychedelic intensity was measured via the 5D-ASC. Assessments focused on mean changes from baseline to 1 month, with additional 3-month data and individual case-series presentations. Given the exploratory nature, no power calculation was performed. The investigators used paired t-tests (two-sided α = 0.05) to evaluate mean changes at 1 month and reported Cohen’s d effect sizes; they did not correct for multiple comparisons. Individual trajectories for EDE and BMI were presented as a case series. Data capture was via Qualtrics and analyses used SAS v9.4.

Primary safety and tolerability: All 10 treated participants completed dosing and follow-up; there were no serious adverse events. Acute psilocybin effects were generally well tolerated, and no clinically significant changes were observed in ECGs or vital signs. Two participants developed asymptomatic hypoglycaemia on the dosing day, both detected in follow-up laboratory assessments and both resolving within 24 hours without intervention; both had been given a standardised breakfast on arrival. No other clinically significant laboratory abnormalities were reported. Suicidality assessments (C-SSRS) showed no increases in suicidal ideation or suicidal behaviours in the post-dosing period; one participant with a prior history of major depressive disorder reported increased suicidal ideation at 3 months that the investigators judged unrelated to study participation. Reported AEs were mild and transient, most commonly headache, nausea and fatigue. Secondary outcomes—eating-disorder psychopathology and related measures: Group-level analyses showed significant reductions in some EDE subscales. Weight concerns decreased from baseline to 1 month (P = 0.036, Cohen’s d = 0.78) and to 3 months (P = 0.04, d = 0.78). Shape concerns decreased at 1 month (P = 0.036, d = 0.78) but not significantly at 3 months (P = 0.081, d = 0.62). Eating concern and dietary restraint subscales did not change significantly. At 3 months, 4 of 10 participants (40%) had global EDE scores that fell to within one standard deviation of community norms; the investigators treated this as clinically meaningful. Mean BMI changes across the sample were not statistically significant and were variable between participants; five participants showed BMI increases at 3 months in the range 0.4–1.2 kg m-2. No correlation was observed between baseline participant characteristics and outcomes. Other secondary/exploratory measures: At the primary 1-month endpoint, the sample showed significant reductions in trait body-image anxiety (PASTAS; P = 0.04, d = 0.76), trait anxiety (STAI-T; P = 0.036, d = 0.78) and preoccupations/rituals related to eating and shape (YBC-EDS; P = 0.043, d = 0.75). Functional impairment related to disordered eating (CIA) decreased significantly from baseline to 1 month (P = 0.041, d = 0.75). Mean changes in depressive symptoms (QIDS) were not statistically significant. Response to treatment was heterogeneous across participants. Patient experience and acceptability: Qualitative and acceptability items indicated that most participants regarded the psilocybin experience as meaningful: 90% reported feeling more positive about life endeavours, 80% ranked the experience among the top five most meaningful of their lives, and 70% reported a shift in personal identity or quality of life. Notably, 90% stated that one dosing session was insufficient. No participant required anxiolytic rescue medication during the dosing session. Measures of altered consciousness (5D-ASC) and further phenomenological data were reported in Extended Data and are to be detailed in a forthcoming qualitative manuscript.

Peck and colleagues conclude that, in this small open-label sample, a single 25 mg dose of pharmaceutical-grade psilocybin administered with psychological support was safe and generally well tolerated in female participants with AN or pAN. The lack of serious adverse events and the transient, mild nature of treatment-emergent AEs are viewed as promising given the physiological risks inherent to the AN population. The two episodes of asymptomatic hypoglycaemia prompted the investigators to hypothesise a relationship to prolonged fasting on the dosing day rather than a direct pharmacological effect of psilocybin; they note clinicians should monitor blood glucose in malnourished participants and consider requiring caloric intake before dosing. Regarding clinical signals, the investigators emphasise heterogeneity: a subset (4/10) showed clinically meaningful reductions in EDE scores that fell to near-community levels by 3 months, and group-level improvements were observed for weight concerns, some body-image and anxiety measures, and functional impairment at 1 month. However, mean BMI did not change significantly across the sample and did not consistently track with symptom reductions. The authors suggest that weight restoration may require targeted nutritional rehabilitation alongside any psychotherapeutic gains, and they quote a participant reflecting on the difficulty of translating increased motivation into sufficient caloric intake for physical recovery. The report acknowledges several important limitations: small sample size, open-label single-arm design, absence of a control condition, lack of power analysis and no correction for multiple comparisons. Selection bias is possible because all participants were self-referred and may have had positive expectations; the sample lacked gender, racial and cultural diversity and included many participants with mild to moderate illness rather than more severe cases. The investigators also note that the single-dose design may be less potent than multi-dose protocols used in other psilocybin trials and that the brief adjunctive therapy may limit behavioural change. Strengths listed include use of a quantified, pharmaceutical-grade psilocybin preparation and the structured therapeutic model. For future research, the authors recommend larger, adequately powered randomized controlled trials in more diverse and severe samples, exploration of mechanisms and moderators of response, optimisation of dose and number of administrations, and investigation of adjunctive treatments (including nutritional rehabilitation) to enhance physical recovery and clinical outcomes. They conclude that while preliminary findings support safety and tolerability, efficacy remains unproven and requires RCT confirmation.