Nineteen hours after a single supervised 5-gram oral dose of psilocybin-containing mushrooms, an octogenarian Japanese-American woman with a decade of Alzheimer's disease and five years of near-total speech loss spoke again. Her first sustained words were spontaneous, autobiographical, and directed at the people in the room.
The episode is described in a case report published 28 May 2026 in Frontiers in Neuroscience by a team at the Medical Department of Associação Cruz de Ankh in São Paulo. It is not a clinical trial. It is not a treatment. It is a single patient, a single high dose of an Enigma-strain mushroom, and a narrow window of improvement that the authors themselves call transient.
At baseline, the patient had advanced Alzheimer's disease with markedly reduced spontaneous communication, predominantly monosyllabic speech, chronic urinary incontinence, executive dysfunction, dysphagia, dependent mobility, and a flat affect, according to the case report. About 19 hours after the dose, spontaneous autobiographical speech returned. In the days that followed, she recovered urinary continence, improved ambulation, regained the ability to dress herself, and showed emotional responsiveness, sustained social interaction, contextual memory retrieval, preserved working memory for social context, and spontaneous conversational engagement.
The improvements faded. The authors are explicit that the changes do not represent disease reversal and that the underlying neurodegenerative process remains unaltered.
The acute phase was not benign. The team reports clinically suspected hyperthermia, profuse sweating, a prolonged deep sleep-like state, and signs of autonomic activation during the hours after the dose, in an 80-year-old patient with a decade of neurodegeneration. None of those events are characterized as minor in the paper. The authors flag safety, contraindications, and drug interactions in this population as unresolved.
Case reports sit at the bottom of the clinical-evidence hierarchy for a reason: they describe one patient, in one setting, with no control and no randomization. They can still be useful when they put a construct on the table that was not previously thinkable. The construct here is named in the paper's own framing: residual functional capacity may persist in late-stage neurodegeneration and may become transiently accessible under specific neuromodulatory conditions.
Psilocybin is already known to transiently alter large-scale brain network dynamics and to induce plasticity-related mechanisms in preclinical models, per the authors' review of adjacent work. The case report is a doorway into that adjacent body of work, not a substitute for it. The authors' own hedge is in the abstract: the findings do not imply disease reversal, and they are calling for controlled study designs rather than clinical use.
The case does not show that psilocybin treats Alzheimer's disease. It does not show that any patient or family should attempt this. It does not show a sustained effect, a dose response, a safety profile in this population, or a mechanism in humans. The Enigma strain is not pharmacologically standardized. The 5-gram oral dose is high. Consent, capacity, and risk in advanced dementia are ethical questions the case report gestures at but does not resolve.
Independent expert reaction from an Alzheimer's clinician, a geriatric psychiatrist, and a psychedelic-medicine researcher not involved in the case would be needed to weigh baseline plausibility, mechanism, and next-step study design. The wider psilocybin-in-dementia literature, including prior case series and ongoing trials in the United Kingdom and the United States, would frame this report within a field rather than as a standalone breakthrough.
What to watch next is whether any group designs a controlled, registered study in advanced or late-stage Alzheimer's that can test whether the residual-capacity construct holds across more than one patient, at a defined dose, with formal safety monitoring and a family-centered consent process. Until that exists, the case remains a research question, not an answer.