Longevity & AgingAI Detects Early Dementia Risk by Analyzing Your Everyday Speech Patterns
Researchers from Baycrest, University of Toronto, and York University found that everyday speech patterns — including pauses, filler words like 'um,' and word-retrieval struggles — are strong indicators of executive function and early cognitive decline. Using AI to analyze natural conversations, the team could predict performance on cognitive tests with surprising accuracy. Because speech is part of daily life, this approach could enable frequent, unobtrusive monitoring for cognitive changes at home or in clinics, potentially detecting dementia risk years before traditional testing would catch it. The findings build on prior work showing faster speech correlates with stronger thinking skills in older adults, reinforcing speech timing as a meaningful brain health biomarker.
