§ Claim under review · Study
"Scientists think psilocybin could address the core neurobiology of autism, based on a recent study (Whelan et al., medRxiv) finding that a low dose of psilocybin increased communication between several large-scale brain networks in autistic participants."
Verdict
Partially accurate but misleading
Confidence
MediumSummary
The study being cited is real. Researchers at King's College London gave 37 autistic and 30 non-autistic adults low doses of psilocybin (2 mg or 5 mg) or a placebo and scanned their brains. At 5 mg, connectivity between certain large brain networks went up in autistic participants and down in non-autistic participants, which is an interesting group difference. But the study did not test whether psilocybin helps anyone, and the authors say so directly: it was not a clinical trial and measured no clinical outcomes. The idea that scientists think psilocybin could address the core neurobiology of autism does not come from the study. The authors framed the work around mental health conditions that often accompany autism, and around choosing doses for future research, not around treating autism itself. Two things the post leaves out: the paper is a preprint that has not been peer reviewed, and the research was funded by a commercial psilocybin company whose co-founder is a listed author. The post's own caption is more careful than the headline claim, correctly noting this does not show psilocybin treats or cures autism.
The readings
key figures from the evidencetotal participants in PSILAUT study
FDR-corrected P value for higher-order/attentional network connectivity increase correlating with autistic traits
psilocybin dose producing largest opposite-direction connectivity shifts
Why this verdict
Evidence
The cited study is real. The PSILAUT study at the Institute of Psychiatry, Psychology and Neuroscience, King's College London, used a pseudo-randomised, double-blind, cross-over, case-control design. Data were collected from January 2023 to August 2024. A total of 67 participants were included: 30 non-autistic adults (mean age 30.0, 50% male) and 37 autistic adults (mean age 28.6, 51% male), matched on age, sex and IQ, all with an existing DSM-IV, DSM-5 or ICD-10 autism diagnosis.
Each participant received a single oral dose of 2 mg psilocybin, 5 mg psilocybin, or inactive placebo on separate visits at least one week apart. The 5 mg dose produced the largest shifts in functional connectivity in both groups, but in opposite directions. In autistic participants, 5 mg psilocybin increased between-network connectivity (integration) of higher-order and attentional networks. In non-autistic participants, connectivity between the same networks decreased. There were no group differences in network connectivity at baseline under placebo. Across the whole sample, the size of the psilocybin-induced increase in connectivity between higher-order networks (β = 0.33, T = 2.16, FDR-corrected P = 0.036) and attentional networks (β = 0.36, T = 2.43, FDR-corrected P = 0.036) correlated positively with autistic traits measured by the Autism Quotient.
The authors' own stated conclusion is that functional brain networks supporting mood, emotion and cognition are more responsive to low-dose psilocybin in autistic adults than in non-autistic adults, and that future applications of psilocybin in autistic people should include evaluation of low doses. The framing question stated in the paper is that adult autism is commonly complicated by co-occurring mental health difficulties, and that it was unknown whether the autistic brain responds differently to psilocybin. The paper explicitly states there was no evaluation of clinical efficacy and that PSILAUT was not a clinical trial under UK regulations.
Findings
✓ What's accurate 5
- The study exists and the authors and venue are correctly identified: Whelan et al., medRxiv.
- A low dose of psilocybin (5 mg) did increase between-network connectivity of higher-order and attentional networks in autistic participants.
- The effect differed by group: connectivity between the same networks decreased in non-autistic participants.
- The study is genuinely early-stage, and the post's caption correctly says it does not show that psilocybin treats or cures autism and that more research is needed.
- Altered brain connectivity is a genuine and long-standing area of autism research.
≈ What's misleading 7
- Unsupported scope expansion: the claim that "scientists think psilocybin could address the core neurobiology of autism" does not appear in the source. The authors' conclusion is about psilocybin's potential relevance to co-occurring mental health difficulties in autistic adults, and about dose selection for future studies. The study measured how the autistic brain responds to a serotonergic probe, not whether psilocybin addresses autism's core features.
- Framing shift from probe to therapy: the study explicitly states it did not evaluate clinical efficacy and was not a clinical trial. Presenting it as evidence that a drug "could address" autism's neurobiology converts a mechanistic measurement into an implied therapeutic direction.
- Contradicted by the same group's own background framing: the researchers' earlier work notes that pharmacological trials targeting the core features of autism have failed, which is the opposite of the confident framing in the claim.
- Omitted qualifier (evidence status): the source is an unreviewed preprint. The claim text does not convey this, though it does name medRxiv.
- Omitted qualifier (conflict of interest): commercial funding by a psilocybin developer and authorship by a company co-founder are not disclosed in the post.
- Omitted qualifier (dose and direction): the effect was specific to 5 mg, not to "a low dose" generally, and the headline result is a group difference in direction of change, not simply "more communication."
- Attribution vagueness: "scientists think" attributes a collective belief to the research community with no identified source holding that view.
? What's uncertain 5
- Whether the preprint has since passed peer review or been published in a journal. No peer-reviewed version was located.
- Whether "increased network integration" in autistic adults corresponds to any clinical benefit. The preprint notes that increased integration has been associated with clinical utility in other conditions, but this is an inference across conditions, not a measured outcome in this study.
- Whether any individual researcher has publicly stated the "core neurobiology" framing. No such statement was found, but the search was not exhaustive across interviews and conference talks.
- Full effect sizes, confidence intervals, blinding integrity, and the complete limitations section were not retrieved in full text, only the abstract-level statistics.
- Whether the post's "core neurobiology" language originates with the account or was added in the claim's restatement. The post caption itself as recorded is notably more cautious than the claim sentence.
Sources
6 of 6 linked to recordsWhelan TP, Dimitrov M, França LGS, Ellis CL, Moruzzi F, Ponteduro FM, Kangas J, Khalil N, Ge Y, Mulcrone N, Ivin G, Batallé D, Daly E, Malievskaia E, Puts NA, Murphy DGM, McAlonan GM. "Distinct response of resting-state brain networks to psilocybin in autism." medRxiv, posted 22 August 2026
Whelan TP et al. "The 'PSILAUT' protocol: an experimental medicine study of autistic differences in the function of brain serotonin targets of psilocybin." BMC Psychiatry. 2024;24:319
ClinicalTrials.gov NCT05651126, "Psilocybin in Adults With and Without Autism Spectrum Disorder"
King's College London press release announcing the study
Science (AAAS), "Could psychedelic drugs improve the mental health of autistic people?"
Whelan TP et al., "Bridging the translational neuroscience gap: the 'shiftability' paradigm"