Study: The Way a Person Speaks May Reveal Early Signs of Brain Aging

The way a person speaks may reveal far more about their health and aging process than previously thought, according to a new study that used an automated computer tool called a “speech clock.” A person’s speech age is not simply a matter of voice characteristics: according to the researchers, it reflects processes taking place throughout the body and could become an inexpensive, non-invasive and automated alternative for measuring biological aging.

“Our voice appears to contain much more information about aging than we previously assumed,” said Agustín Ibáñez, the study’s senior author from Trinity College Dublin, Euronews reports.

The automated program calculated a person’s speech age and compared it with their chronological age, indicating whether their speech sounded younger than expected, typical for their age, or older than expected, thereby providing clues about how well the aging process is progressing.

“It captures both the passage of chronological time and signals coming from cognitive functions, the brain, systemic biology and even our accumulated social environment,” he added.

The researchers analyzed audio recordings from nearly 3,000 Spanish-speaking adults from different countries across Latin America. Participants completed a series of tasks: they retold a short animated film, retold a story they had heard 20 minutes earlier, and listed objects belonging to specified categories.

Using machine-learning methods, the team examined hundreds of features reflecting how people speak and what they say — from speech rate and pauses to pitch, emotional tone, vocabulary and semantic accuracy — to estimate chronological age and calculate an individual “speech age gap.”

The “speech clocks” were able to distinguish healthy people from patients with dementia. Among healthy participants, the gap between speech age and chronological age was minimal, whereas in people with Alzheimer’s disease and various forms of frontotemporal dementia, it progressively increased.

People whose speech appeared “older” than would be expected for their chronological age showed signs of accelerated aging across a range of biological and clinical systems. Brain MRI scans showed more pronounced structural loss and altered activity.

A higher speech age was also associated with increased blood levels of p-Tau217, a key protein associated with Alzheimer’s disease.

Participants with “older” speech performed worse on tests of memory, concentration and problem-solving. This association was not limited to language-based tasks but was also observed in nonverbal tests, such as visual memory.

“More broadly, the finding is striking because a person’s voice can provide a remarkably compact snapshot across several dimensions of aging at once,” Ibáñez said.

Factors such as chronological age, brain aging, cognitive abilities, as well as phenotypes of Alzheimer’s disease and other forms of dementia, have traditionally been assessed using completely different and often expensive tools. However, he said, they all appear to converge in features of speech.

The authors note that many existing methods for assessing biological aging require MRI scanners, blood samples or specialized clinical examinations. By comparison, “speech clocks” offer an inexpensive, non-invasive, fully automated alternative that can be used remotely on ordinary devices without the involvement of specialized medical professionals.

At the same time, the researchers emphasize that “speech clocks” are not yet a diagnostic test for dementia and cannot determine whether an “older” speech profile predicts the subsequent development of cognitive impairment.