AI Speech Clock May Reveal Signs of Biological Aging

AI Speech Clock May Reveal Signs of Biological Aging
————————————–
Researchers have developed a machine-learning “speech clock” that estimates a person’s age from subtle acoustic and linguistic patterns in their speech. A study of nearly 3,000 people from Latin America found that individuals whose estimated speech age was older than their actual age tended to show faster brain and biological aging, as well as greater cognitive decline.

Published in Science Advances, the study analyzed factors including speech rate, pauses, pitch, vocabulary, emotions and sentence structure. Larger differences between chronological age and speech age were associated with brain changes, faster epigenetic aging, poorer cognitive performance and higher levels of Alzheimer’s-related biomarker p-tau217.

The researchers said speech age was also linked to social factors such as education, financial conditions, food insecurity and access to healthcare. Because speech can be collected remotely and repeatedly at relatively low cost, the technology could eventually offer a scalable way to monitor aging. However, the researchers stressed that the speech clock is not currently a diagnostic tool for dementia, and further long-term studies are needed to determine whether it can predict future cognitive decline.