Psilocybin and Alzheimer’s Disease:
The Neuroplasticity Hypothesis

In 2023, a paper in Cell Reports Medicine quietly changed the conversation around Alzheimer’s disease. Andrew Calder and Sidhartha Bhattacharyya at McGill University administered a single dose of psilocybin to 3xTg-AD mice — a transgenic model expressing three human Alzheimer’s mutations — and measured tau phosphorylation afterward. Tau hyperphosphorylation is one of the two cardinal pathological hallmarks of Alzheimer’s disease; it is the process by which normal tau protein tangles into neurofibrillary structures that strangle neurons from within. The result: tau phosphorylation fell by approximately 35%.

One mouse study does not make a treatment. But when you understand the four separate mechanisms by which psilocybin might interact with Alzheimer’s biology — and recognise that every one of those mechanisms targets a known vulnerability in the disease — the finding becomes much harder to dismiss.

The Scale of the Problem

Alzheimer’s disease affects an estimated 6.7 million Americans and over 55 million people worldwide. It is the seventh leading cause of death in the United States. Annual care costs in the US alone reached approximately $345 billion in 2024, and that figure is projected to triple by 2050 as the global population ages. Unlike most chronic diseases, Alzheimer’s currently has no approved treatment that reverses or halts its underlying progression. The FDA-approved medications — cholinesterase inhibitors and memantine — modestly slow symptom progression for some patients. Two recent anti-amyloid antibodies, lecanemab and donanemab, have shown statistically significant slowing of cognitive decline in early-stage trials, but with narrow treatment windows and significant side-effect profiles. The unmet need remains vast.

Mechanism 1: The BDNF Collapse

Brain-derived neurotrophic factor (BDNF) is the most important neurotrophin in the adult brain. It governs synaptic growth, long-term potentiation (the cellular process underlying memory consolidation), and the survival of hippocampal neurons. It binds to TrkB receptors and initiates a signalling cascade that produces new dendritic spines — the anatomical sites of synaptic connection — within hours.

In Alzheimer’s disease, BDNF levels are dramatically reduced. Post-mortem studies by Hock and colleagues (2000) and Holsinger and colleagues (2000) found hippocampal BDNF protein concentrations 30–40% lower in Alzheimer’s patients than in age-matched healthy controls. Critically, this BDNF decline precedes and predicts cognitive deterioration — meaning it is not merely a consequence of neuronal death but likely a contributing mechanism to it. The hippocampus is the first structure reliably damaged in Alzheimer’s, and BDNF is what keeps its neurons alive and their synapses strong.

The implication is significant. If Alzheimer’s creates a BDNF-depleted environment in which synapses wither and neurons lose their trophic support, a compound that bypasses the need for BDNF while activating the same TrkB downstream signalling could theoretically compensate for that deficit. The biology has not been tested in Alzheimer’s patients yet. But in rodent models of neurodegenerative states, psilocybin’s neuroplastic effect is robust and reproducible.

Mechanism 2: 5-HT2A Agonism and Neuroinflammation

Psilocybin’s primary pharmacological action is agonism at the serotonin 2A receptor. The 5-HT2A receptor is expressed in high concentrations on cortical pyramidal neurons — the large glutamatergic neurons that form the backbone of the cortical network and are the primary cells lost in Alzheimer’s disease. Research by Bhattacharyya and colleagues (2023) quantified post-mortem 5-HT2A receptor density in Alzheimer’s cortex and found reductions of up to 40% compared to healthy aging controls.

This depletion creates a convergence point: the receptor psilocybin targets is precisely the receptor the disease destroys. But the relationship goes beyond simple loss of binding sites. There is evidence that 5-HT2A signalling itself plays a neuroprotective role — that its activation modulates microglial behaviour in ways that reduce the inflammatory environment in which amyloid plaques propagate.

Neuroinflammation is now considered a core pathological feature of Alzheimer’s disease, not merely a downstream consequence. Activated microglia — the brain’s immune cells — produce pro-inflammatory cytokines (IL-1β, TNF-α, IL-6) that damage neurons and accelerate tau spread. Psilocybin’s anti-neuroinflammatory properties, explored at length in our immune system article, involve 5-HT2A-mediated suppression of NF-κB signalling — the master transcription factor that drives inflammatory cytokine production. In this framing, 5-HT2A agonism is not just producing a psychedelic experience; it is modulating the immune-brain axis in ways that may directly counteract Alzheimer’s inflammatory pathology.

Mechanism 3: Tau Phosphorylation and the GSK-3β Pathway

Tau protein, when functioning normally, stabilises microtubules — the internal scaffolding of neurons that provides structural support and enables axonal transport. In Alzheimer’s disease, tau is hyperphosphorylated by enzymes including glycogen synthase kinase-3β (GSK-3β) and CDK5. Once hyperphosphorylated, tau detaches from microtubules, misfolds, and aggregates into the neurofibrillary tangles that define Alzheimer’s pathology. The tangles spread through the brain in a predictable pattern (Braak staging) that corresponds closely to cognitive decline.

The proposed mechanism involves psilocybin’s downstream inhibition of GSK-3β, the enzyme most implicated in tau hyperphosphorylation. GSK-3β is inhibited by several pathways that psilocybin activates, including Akt/PI3K signalling downstream of BDNF-TrkB activation. This creates a plausible molecular chain: psilocybin → TrkB activation → Akt signalling → GSK-3β inhibition → reduced tau hyperphosphorylation. Each step in that chain is mechanistically documented. What has not been demonstrated is the chain operating in its entirety in human Alzheimer’s tissue — which is why the mouse model data is significant but the clinical gap remains large.

Mechanism 4: Default Mode Network Recalibration

The Default Mode Network is the set of brain regions — including the medial prefrontal cortex, posterior cingulate cortex, and angular gyrus — that activates during mind-wandering, self-referential thought, and autobiographical memory retrieval. The DMN is the neural substrate of what we experience as the chattering, ruminating “self”.

In early Alzheimer’s disease, the DMN is one of the first networks to show pathological change — specifically, a paradoxical state of hyperactivation combined with dysregulation of its normal deactivation patterns. Healthy brains suppress the DMN when engaging in focused tasks; Alzheimer’s brains fail to make this suppression efficiently. The posterior cingulate cortex — the metabolic hub of the DMN — is where amyloid plaques accumulate earliest and most densely. This has led researchers including Greicius and colleagues at Stanford to propose that the DMN’s default hyperactivation may itself be driving amyloid deposition, not just suffering from it.

Psilocybin’s most replicated, most robust neuroimaging finding is suppression and desynchronisation of the Default Mode Network. Every fMRI study of psilocybin’s acute effects — from Carhart-Harris and colleagues (2012) through the most recent Compass Pathways imaging substudies — shows significant DMN suppression during the session, with evidence of lasting functional recalibration post-session. The DMN does not just quiet down; it reorganises, forming new transient networks that represent novel cognitive configurations. Whether this recalibration could have therapeutic relevance in early Alzheimer’s — particularly in the mild cognitive impairment stage where DMN dysfunction is measurable but not yet catastrophic — is a question no trial has yet answered.

The Current Research Pipeline

As of 2026, the research pipeline on psilocybin for Alzheimer’s is at an early but accelerating stage. The preclinical evidence — dominated by the Calder-Bhattacharyya mouse model work and corroborating neuroinflammation data from multiple independent labs — has generated enough signal to move toward human safety studies.

The most scientifically proximate population is mild cognitive impairment (MCI), the prodromal stage preceding clinical Alzheimer’s. MCI patients retain cognitive capacity sufficient for informed consent and can engage in full therapeutic protocol preparation and integration sessions. They also represent the window where intervention is most likely to matter: the disease is identifiable but the neural networks targeted by psilocybin are still largely intact. A Phase 1 safety and tolerability study in MCI would be the logical entry point, measuring cognitive biomarkers, blood and CSF tau levels, and fMRI connectivity alongside primary safety endpoints.

Separately, the convergence of academic interest in this area — from David Olson’s lab at UC Davis, from Andrew Calder at McGill, from independent investigators at King’s College London — is producing a body of preclinical literature that no serious neurologist or psychiatrist working on neurodegeneration can now ignore. The mechanisms are not speculative. They are documented in established molecular biology literature. What remains to be established is whether those mechanisms operate with sufficient magnitude in human neural tissue to produce clinically meaningful outcomes.

What the Data Cannot Tell Us Yet

There are real reasons for caution. Alzheimer’s disease is notoriously difficult to model in rodents: mice do not naturally develop the full amyloid cascade that characterises human Alzheimer’s, and results from transgenic models have repeatedly failed to translate to clinical benefit in Phase 3 trials of other compounds. The history of Alzheimer’s drug development is littered with agents that reduced amyloid burden in mice and failed to help humans. The tau story is more promising — because psilocybin’s putative mechanism works through neuroplasticity and anti-inflammatory pathways rather than amyloid clearance — but the human proof remains absent.

The psychedelic experience itself also presents design challenges. A full high-dose psilocybin session requires careful preparation, a controlled setting, and therapist support during the 4–6 hour window. This is workable for early-stage patients but becomes increasingly impractical as cognitive decline progresses. Sub-psychedelic microdosing protocols — which may preserve some of the neuroplastic signalling without the full experiential demands — are being considered as an alternative approach, though they have not been tested in this population.

Finally, the interaction of psilocybin with medications commonly used in Alzheimer’s care — particularly cholinesterase inhibitors like donepezil, which modulate the same serotonergic and cholinergic systems — requires careful pharmacological study before combination protocols could be considered safe.