TL;DR
Turn quiet afternoons into listening time
- Thousands of audiobooks, podcasts and originals
- Listen on your phone, tablet or Echo — also offline
- Cancel anytime
Researchers developed a machine-learning model that estimates chronological age from hundreds of speech features. In 2,928 Spanish-speaking participants, the difference between estimated and actual age was associated with biological and brain-aging measures, cognitive performance and dementia groups. The researchers say the model is not a diagnostic test and needs longitudinal and broader validation.
Researchers have developed a machine-learning “speech clock” that estimates a person’s chronological age from how they speak, according to a study published in Science Advances. In 2,928 Spanish-speaking participants, the gap between estimated and actual age was associated with measures of brain and biological aging, cognitive performance, social conditions and dementia, though the model is not a diagnostic test.
The study included participants from Argentina, Chile, Colombia, Mexico and Peru, including healthy adults and people diagnosed with mild cognitive impairment, Alzheimer’s disease or forms of frontotemporal dementia. The model analyzed hundreds of acoustic and language features rather than relying on one characteristic of a voice. These included speech rate, pauses and pitch, as well as emotional content, vocabulary, meaning and the amount and organization of spoken material.
Researchers used those features to estimate chronological age and calculate a speech age gap: the difference between a participant’s actual age and the age predicted by speech. Larger gaps, indicating speech that appeared older than expected, were associated with measures of biological aging, including brain age assessed through structural and functional neuroimaging and aging estimates from three DNA-methylation clocks.
The gap was also associated with lower scores in global cognition, executive function, daily functioning and several types of memory. The researchers reported links with performance on non-language cognitive measures as well. Healthy participants had the lowest gaps on average, with larger gaps across the Alzheimer’s and frontotemporal dementia groups. Among people with Alzheimer’s disease, the measure was associated with higher levels of plasma p-tau217, a blood biomarker linked to Alzheimer’s pathology. These are associations in the study; they do not show that speech changes cause disease or predict an individual’s future condition.
A Low-Cost Window on Aging
Many measures used to study aging involve MRI scans, blood samples or specialized assessments. Speech can be recorded repeatedly and remotely, giving researchers a potential way to track changes at lower cost and with less burden for participants. If further studies support the findings, speech measures could eventually complement existing assessments, particularly in places where access to advanced diagnostic tools is limited.
The research also draws attention to a possible source of health information that is available during ordinary conversation. The authors found that speech age was associated with several distinct measures, from cognition to biological aging. That pattern makes the approach interesting for research, but it does not establish that one speech recording can reliably measure a person’s overall health or replace clinical evaluation.
How the Speech Clock Was Studied
The researchers combined features describing how people speak and what they say in a machine-learning model. It then estimated age and compared that estimate with each participant’s chronological age. The study also compared speech age gaps with clinical and biological measures collected from participants.
Speech age gaps were associated with a more adverse social exposome—a combination of factors across a person’s life, including education, financial conditions, food security, access to health care and early-life experiences. The study’s five-country Latin American sample adds evidence from a region the report describes as historically underrepresented in dementia research. The researchers say the findings could help explore aging measures that do not depend only on expensive technologies.
““Our voice appears to contain much more information about aging than we previously recognized.””
— Agustin Ibanez, senior author and professor of brain health at the Global Brain Health Institute and Trinity College Dublin’s School of Medicine
Prediction Is Not Diagnosis
The study was primarily cross-sectional, so it cannot establish whether an older-appearing speech profile predicts who will later develop cognitive decline or dementia. The reported associations also do not show that the speech clock can diagnose a condition in an individual. The researchers emphasize that it is not yet a diagnostic test.
Its performance also needs evaluation in additional languages and cultures, and in more natural speaking situations. The supplied study report does not give enough detail to establish how accurately the model would estimate age for people outside the studied sample or how well it would work in routine clinical use.
Testing Across Time and Languages
The researchers say longitudinal studies are needed to test whether speech age gaps change before later cognitive decline or dementia. Further validation in additional populations, languages and everyday recording settings would help establish how broadly the method applies.
Until that evidence is available, the speech clock should be understood as a research measure. The study raises the possibility that speech could complement more costly measures of aging, but whether it can support monitoring or clinical decisions remains to be tested.
Key Questions
What does the speech clock estimate?
It uses acoustic and language features in a person’s speech to estimate chronological age. Researchers then compare that estimate with the person’s actual age to calculate a speech age gap.
Does the speech clock diagnose dementia?
No. The researchers state that it is not a diagnostic test. The study found associations between speech age gaps and dementia groups, but it did not establish that the model can diagnose an individual.
What did a larger speech age gap mean in the study?
A larger gap meant the model estimated an age above the participant’s chronological age. In this study, larger gaps were associated with several aging and health measures; the result does not establish a cause or predict an individual’s future health.
Who took part in the research?
The study analyzed 2,928 Spanish-speaking participants from Argentina, Chile, Colombia, Mexico and Peru. Participants included healthy adults and people with mild cognitive impairment, Alzheimer’s disease and forms of frontotemporal dementia.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
