A machine-learning “speech clock” linked older-sounding speech to brain age, epigenetic aging, cognition, dementia markers and social adversity across five Latin American countries.

Listen to someone speak and you can usually make a decent guess at their age. The pitch, the pace, the pauses and even the words people choose all shift over the decades. A large international study suggests those changes carry far more information than a rough estimate of birth year. They may reflect how quickly a person’s brain and body are aging.

The research, published in Science Advances and led by scientists at Trinity College Dublin and the Global Brain Health Institute, built what the team calls a “speech clock”: a machine-learning model that estimates a person’s age from hundreds of acoustic and linguistic features of their speech.

“Our voice appears to contain much more information about aging than we previously recognised,” said senior author Agustin Ibanez, Professor in Brain Health at the Global Brain Health Institute and Trinity’s School of Medicine. “It captures both the passage of chronological time and signals coming from cognition, the brain, systemic biology, and even our accumulated social environment.”

Measuring the gap

The study analysed speech from 2,928 Spanish-speaking participants in Argentina, Chile, Colombia, Mexico and Peru. They included healthy adults as well as people with mild cognitive impairment, Alzheimer’s disease and several forms of frontotemporal dementia. The models drew on features such as speech rate, pauses, pitch, emotional content, vocabulary and semantic precision.

The key measure was the speech age gap: the difference between the age the model predicted from someone’s speech and their actual age. A person whose speech sounds older than their years has a positive gap, a possible sign of accelerated aging. The approach mirrors other biological aging clocks, which compare a person’s chronological age with an age estimated from their DNA, blood or brain scans.

The first widely used epigenetic clock was published in 2013, and since then researchers have built clocks from blood proteins, brain imaging and other data. Each tries to capture the same idea: that people age biologically at different rates, and that measuring that pace could help predict disease and test interventions.

A voice that tracks the body

Older-sounding speech was associated with several independent markers of accelerated aging. It was linked with brain age estimated from structural and functional neuroimaging, and with epigenetic aging measured by three separate DNA-methylation clocks. Epigenetic clocks track chemical tags on DNA that change predictably with age and are among the most widely used measures of biological aging.

A larger speech age gap was also associated with poorer global cognition, executive function, functional ability and several measures of memory. Notably, those links appeared even on cognitive tests that did not involve language, suggesting the speech clock captures something broader than verbal skill.

The gap widened from healthy participants to those with Alzheimer’s disease and frontotemporal dementia, and the combined speech measure distinguished between groups better than any single speech feature. In people with Alzheimer’s disease, a larger gap correlated with higher blood levels of p-tau217, a protein fragment that has emerged as one of the most accurate blood markers of Alzheimer’s pathology, and tracked clinical and cognitive function.

Frontotemporal dementia, which often begins earlier in life than Alzheimer’s disease, can affect language and behaviour early in its course, so differences in speech are perhaps unsurprising there. Blood tests for p-tau217 have advanced rapidly, and in 2025 the United States Food and Drug Administration cleared the first blood test to aid in diagnosing Alzheimer’s disease, which relies in part on that marker.

The social side of aging

The researchers also examined the social exposome, the accumulated social and environmental conditions a person experiences across life, including education, finances, food insecurity, access to health care and early-life experiences. Among healthy participants and those with Alzheimer’s or other dementias, faster speech aging was associated with a more adverse social exposome.

That finding fits a broader body of research showing that disadvantage gets under the skin, accelerating biological aging and raising dementia risk. Dementia research has long been concentrated in wealthy, mostly English-speaking countries, and the release notes that Latin America, a region with high inequality and a rapidly aging population, remains underrepresented. This study, conducted entirely in five Latin American countries, helps address that gap.

“The broader finding is nevertheless striking in that a person’s voice may provide a remarkably compact readout of multiple dimensions of aging,” Ibanez said. “From chronological age to brain aging, epigenetic aging, cognition, Alzheimer’s-related pathology, social exposures, and dementia phenotypes, information traditionally obtained through very different and often expensive measurements appears to converge, at least partly, in the way we speak.”

Cheap, remote and repeatable

Part of the appeal is practical. Speech can be recorded remotely, repeatedly, non-invasively and at low cost, unlike MRI scans, blood draws or specialist assessments. That could make it valuable in places with limited access to advanced diagnostics.

“This raises the possibility that something as simple and accessible as speech clocks, maybe combined with biomarkers, could eventually complement much more expensive measures of aging,” Ibanez said.

The researchers stress that the speech clock is not a diagnostic test for dementia. The study was cross-sectional, capturing each participant at a single point in time, so it cannot show whether older-sounding speech predicts future decline. Longitudinal studies, research in other languages and cultures, and tests in more natural speaking situations are all needed. Privacy questions around voice data will also need careful handling if such tools reach wider use.

If those hurdles are cleared, the payoff could be significant. “If confirmed longitudinally and across populations,” Ibanez said, “speech could ultimately become one of the most scalable tools for monitoring healthy and accelerated aging, potentially transforming an everyday human behaviour into a window onto the biology of aging.”

---

Notes

1. Science Advances, 30 September 2026. DOI: 10.1126/sciadv.aef9864.
2. Trinity College Dublin. “Your voice may reveal how fast and how well you’re aging – new research.” EurekAlert!, 30 September 2026. https://www.eurekalert.org/news-releases/1145291.

Featured image: Photo by cottonbro on Pexels.com