There are promising early hints that psilocybin might help Alzheimer's and dementia patients, but no rigorous controlled trial (the gold standard for proof) has been completed and published yet. Running these trials is brutally hard: people with dementia can't reliably consent, recruiting enough patients takes years, and following them safely for months after treatment is medically complex. Even well-funded trials starting now won't finish, analyze, and publish results by the end of 2026.
The Alzheimer's-psilocybin signal is currently at case-report or early preclinical stage. Controlled trials in this population face extraordinary recruitment challenges (informed consent with cognitive impairment), safety hurdles, and lengthy IRB timelines. Even aggressively funded trials initiated in 2025 would not complete enrollment, treatment, follow-up, and peer review within 30 months. The prediction strips the unfalsifiable 'grant funding' component of the original and focuses on the replication claim, which is concrete and checkable. 76% confidence adjusted to 73% to account for possibility of fast-tracked open-label pilot data being published.