Body mass index (BMI) does not predict responses to psilocybin

Spriggs and colleagues frame their study around inter-individual differences in 5-HT2A receptor density and how these may relate to body mass index (BMI). Prior work has reported positive correlations between BMI and 5-HT2A binding in several cortical regions, genetic links between 5-HT2A polymorphisms and body mass, and altered receptor density in disorders such as anorexia nervosa. Body weight-adjusted dosing has been common in psilocybin research because weight is a conventional pharmacological covariate, but body composition and receptor-level differences raise uncertainty about whether weight-adjusted dosing is necessary or optimal for producing consistent acute and long-term psychological effects from psilocybin. To address this gap, the investigators pooled data from three studies at the Centre for Psychedelic Research, Imperial College London, in which a fixed high dose of 25 mg psilocybin was administered in a supportive setting. The analysis aimed to assess whether BMI predicts (1) the intensity and qualitative features of the acute psychedelic experience, indexed by the Altered States of Consciousness (ASC) questionnaire and the Emotional Breakthrough Inventory (EBI), and (2) longer-term changes in well-being indexed by the Warwick–Edinburgh Mental Well-being Scale (WEMWBS). Given limited prior expectations, both frequentist and Bayesian hypothesis testing were used so that evidence for or against the null hypothesis could be quantified.

The pooled dataset comprised three trials: an open-label feasibility trial in treatment-resistant depression (D-OL), a randomised controlled trial comparing psilocybin to escitalopram in moderate–severe depression (D-RCT), and a healthy volunteer (HV) study in psychedelic‑naïve individuals. All dosing sessions occurred at the same research facility under ethical approvals noted in the original reports. Participants were adults aged 18–85 and physically healthy; exclusion criteria included current or family history of psychotic disorder, MRI contraindications, pregnancy, and substance misuse. With one exception, clinical participants discontinued serotonergic medication prior to dosing. For consistency, analyses were restricted to each participant's first 25 mg psilocybin session. D-OL participants received 10 mg then 25 mg one week apart and only the first 25 mg was used here. In D-RCT, 30 participants received two 25 mg doses separated by three weeks (only those receiving 25 mg were included in this pooled analysis); the HV study delivered a 25 mg dose following a blinded 1 mg dose one month earlier. Across all studies participants were supported by two guides, received preparatory and integration sessions, reclined with eye mask and headphones, and were exposed to a music playlist designed for therapeutic inward focus. Primary measures included BMI calculated from measured height and weight, the 42-item ASC (yielding total and three subscales: oceanic boundlessness, OBN; dread of ego dissolution, DED; visionary restructuralization, VRS), the Emotional Breakthrough Inventory (EBI; administered in D-RCT and HV), and WEMWBS measured within 10 days pre-dose and exactly 2 weeks post-dose in D-RCT and HV. Statistical analysis was post hoc: after pooling, frequentist linear regressions tested BMI as a predictor of acute (ASC, EBI) and longer-term (WEMWBS) outcomes while controlling for mean-centred age, sex and study, with baseline WEMWBS added when appropriate. Complementary Bayesian linear regressions were run using default JSZ priors to compute Bayes Factors (BFs) comparing nested models to quantify evidence for or against inclusion of BMI; follow-up Bayesian t-tests were used where relevant. Analyses used R tooling and plots were generated with ggplot2.

The pooled sample included 77 participants who provided ASC data; EBI analyses were restricted to D-RCT and HV (N = 56) and WEMWBS analyses had additional dropout (N = 51). There were no significant differences between studies in age or BMI, and sex distributions did not differ by study. Across the pooled sample BMI did not correlate significantly with age, but females had lower BMI than males (female mean 24.1 kg/m2, SD 4.8; male mean 26.5 kg/m2, SD 4.36; t(57.58) = 2.241, p = 0.03). Age, sex and study were included as covariates in all reported models. When testing BMI as a predictor of the acute psychedelic experience, frequentist linear regressions controlling for age, sex and study found no significant effect of BMI on ASC total score or on the OBN and VRS subscales. The DED subscale showed a marginal, non-significant trend (p = 0.076). Bayesian analyses provided strong evidence favouring the null model over models including BMI for ASC total, OBN and VRS; DED was the sole exception where evidence was inconclusive. Additional model comparisons showed that study—rather than BMI—was the strongest predictor of DED. Bayesian t-tests comparing studies indicated strong evidence for lower DED in D-RCT versus HV (BF = 22.07), moderate evidence against a D-OL versus HV difference (BF = 0.36), and inconclusive evidence between the two depression trials (BF = 1.76). For emotional breakthrough, the frequentist regression with covariates was not significant (p = 0.904). Bayesian regression gave very strong evidence for the null model (BF02 = 30.91) and moderate evidence against including BMI (BF12 = 2.60). Regarding longer-term outcome, BMI did not predict change in well‑being: the regression controlling for age, sex, study and baseline WEMWBS produced b = 0.27 (95% CI −0.28 to 0.82), p = 0.325. Bayesian model comparison provided moderate evidence against BMI as a predictor (BF12 = 2.051). By contrast, indices of the acute experience did predict well‑being: ASC-OBN significantly predicted improvements in WEMWBS (b = 0.10, 95% CI 0.03 to 0.18, p = 0.009) and EBI also predicted WEMWBS (b = 0.10, 95% CI 0.03 to 0.18, p = 0.005). Bayes Factors supported ASC-OBN (BF31 = 6.50) and EBI (BF31 = 8.95) as meaningful predictors, while providing weak evidence against BMI's contribution in models that included these acute-experience predictors (BF23 ≈ 1.85–1.86).

Spriggs and colleagues interpret the pooled results as providing evidence that BMI does not predict either the acute phenomenology or short-term improvements in well‑being following a fixed 25 mg dose of psilocybin administered in a supportive setting. Bayesian analyses in particular were used to argue that the findings represent evidence of absence for BMI effects rather than simply an absence of evidence. The only exception was a marginal signal for greater dread of ego dissolution (ASC-DED) at lower BMI, but further analyses implicated study context rather than BMI as the primary driver of that variation. The authors position their findings against prior literature that has typically used body weight-adjusted dosing and reported mixed associations between body weight and acute intensity; their pooled fixed-dose analysis therefore adds new evidence relevant to dosing strategy. They note that fixed dosing would considerably simplify logistics, standardisation and potential large-scale roll-out of psychedelic-assisted therapy, and might be advantageous where body composition makes weight-based scaling problematic. The discussion also references PET studies linking 5-HT2A receptor binding and subjective intensity, and acknowledges that receptor-level relationships remain complex and warrant further PET work to reconcile pharmacokinetic and pharmacodynamic factors with subjective outcomes. Several limitations are acknowledged. The sample included relatively few individuals at the extremes of BMI (15 participants were classified as obese and only three were underweight), limiting inference about tails of the distribution. No pharmacokinetic measures were collected, so the study cannot address how drug concentration or metabolism interacted with BMI. The analysis is also limited to a single fixed dose (25 mg) and to psilocybin specifically, so results may not generalise to other doses, compounds or administration contexts. Finally, although analyses controlled for study, the three included trials differed in design and prior participant exposure, which could introduce heterogeneity. In terms of implications, the authors highlight that mystical-type experiences (ASC-OBN) and emotional breakthrough (EBI) predicted improvements in well-being irrespective of BMI, reinforcing the role of the acute subjective experience as a mediator of therapeutic gain. They suggest that fixed high dosing—at least at 25 mg within a supportive therapeutic environment—may be a feasible approach that does not disadvantage participants across a range of BMIs, while recommending further research in broader populations, across doses, and with pharmacokinetic and PET measures to refine understanding.