In an intriguing discovery that could revolutionize the way we measure brain aging, researchers have developed a novel tool called a 'speech clock.' This innovative technology analyzes subtle characteristics of a person's speech to estimate how quickly they are aging. The study, published in Science Advances, involved 2,928 Spanish-speaking participants from five Latin American countries and found that differences between a person's actual age and their speech-predicted age were associated with cognitive impairment and dementia.
The study sheds light on the concept of biological age, which refers to how old our bodies appear to be based on factors such as cellular damage, changes in organs, and the effects of aging. Unlike chronological age, biological age can differ significantly from our actual years lived. The research team analyzed speech recordings from participants across various health conditions, including healthy individuals, those with mild cognitive impairment, Alzheimer's disease, and frontotemporal dementia.
Using machine-learning models, the researchers examined hundreds of acoustic and linguistic characteristics in the recordings, including speech timing, pitch, vocabulary, word choice, language complexity, emotional expression, and the number of words used. The model was then trained to predict a person's chronological age based on these speech patterns. The predicted age was compared with the participant's actual age to calculate their 'speech age gap.'
The findings revealed that healthy individuals generally had smaller speech age gaps, while those living with cognitive impairment tended to have larger ones. This difference was particularly pronounced among people with frontotemporal dementia, especially those whose condition affected language-related abilities. Additionally, larger speech age gaps were associated with poorer performance in several cognitive assessments, including tests measuring executive functioning, memory, and overall cognitive ability.
Furthermore, among participants with Alzheimer's disease, larger speech age gaps were associated with higher levels of p-tau217, a protein in the blood linked to Alzheimer's-related brain changes. This suggests that speech patterns may reflect biological changes associated with the disease rather than simply differences in how people communicate.
The potential application of this research is significant. A speech clock could offer a relatively low-cost, low-effort way of monitoring aspects of cognitive aging. Unlike many existing biological aging models, which require complicated procedures such as MRI scans, blood samples, molecular tests, or specialized clinical assessments, speech analysis could be considerably easier to carry out. This non-invasive method could potentially be more comfortable for individuals undergoing repeated assessments and could be performed in various settings.