What your voice
reveals about ageing

Can our speech show whether our brain and body seem older than our age?

SBR Project ID

#BBHM02

Date of project

2020–2026

Location

Argentina, Chile, Colombia, Mexico and Peru

Type of project

Observational Study

Categories

,

Tags

, , , ,

Lead institution and funders

Ask someone to describe a short film and you hear more than the story: how quickly they speak, how often they pause and which words they choose. All of these change as we get older, and with dementia.

Scientists can measure signs of ageing with ageing clocksA brain clock, for example, estimates age from brain scans. If the estimate is higher than a person’s real age, their brain appears older than expected.: tools that estimate how old a person’s brain or body appears and compare it with their real age. Many rely on brain scans or blood tests, which are costly and not available everywhere.

So we built a “speech clock”: a computer model that estimates a person’s age from hundreds of features of their speech. We used recordings of 2,928 Spanish speakers in Argentina, Chile, Colombia, Mexico and Peru. Some were healthy volunteers; others had mild cognitive impairmentMemory or thinking problems that can be measured on tests but are not severe enough to be called dementia., Alzheimer’s disease or frontotemporal dementiaA group of dementias caused by damage to the front and sides of the brain. Depending on the form, it mainly affects behaviour, movement or language..

The difference between the age the speech clock estimates and a person’s real age is their speech age gap. We wanted to find out whether this gap reveals anything about a person’s brain, body, health and life circumstances.

57 Million

People

worldwide were living with dementia in 2021, over 60% of them in low- and middle-income countries, according to the World Health OrganizationRead the World Health Organization’s dementia fact sheet, updated 3 July 2026..

2,928

People

who speak Spanish, from Argentina, Chile, Colombia, Mexico and Peru, took part in this study.

700+

Speech features

were measured from each person’s recordings, from pauses and pitch to the words they chose.

Follow this research

Get an email whenever we publish new findings on this project.

AIMS

What we want to know

Our study asked three questions.

We recorded people doing seven short speaking tasks and measured hundreds of features of their speech, from pauses and pitch to the words they chose. We trained a computer model to estimate each person’s age from these features alone, and checked how close its estimates came to their real age.

For each person we worked out their speech age gapThe age the speech clock estimated from a person’s speech, minus their real age. A positive gap means their speech seems older than expected for their age; a negative gap means it seems younger.: the age the speech clock estimated minus their real age. We compared this gap between healthy volunteers and people with mild cognitive impairmentMemory or thinking problems that can be measured on tests but are not severe enough to be called dementia. or Alzheimer’s disease. We also included two groups of people with frontotemporal dementiaA group of dementias caused by damage to the front and sides of the brain. Depending on the form, it mainly affects behaviour, movement or language.: forms that mainly affect behaviour or movement, and forms that mainly affect language.

We compared each person’s speech age gap with their brain age from MRI scans and their age estimated from chemical marks on their DNA. We also compared it with their blood level of p-tau217A form of the tau protein that can be measured in a blood sample. It is one of the most important blood markers of Alzheimer’s disease in the brain., their memory and thinking scores, and a score for their life circumstances. Not everyone had all of these measures.

Illustration of three clocks on a shelf, their sound-wave hands joined by a thread of sound waves

IMPORTANCE

Why it matters

A measure of ageing based on voice recordings could reach people that scanners and laboratories do not. Speech can be recorded remotely, repeatedly, without needles or scanners, and at very low cost.

Where the research is done matters too. People in Latin America have historically been underrepresented in dementia research. Many live where brain scans and specialist blood tests are hard to get. Because this study involved people in five Latin American countries, it provides evidence that advanced measures of ageing need not depend only on expensive technology.

If voice recordings can show whether the brain and body seem older than expected, measuring ageing could become possible far beyond specialist clinics.

Illustration of a clock on a table by an open window, its sound-wave hands reaching out towards distant villages and mountains

PROJECT

So we set up a study

Everyone in this study took part through the Multi-Partner Consortium to Expand Dementia Research in Latin America (ReDLat)A network of research centres that aims to expand dementia research in Latin America and the Caribbean. Opens the ReDLat project page on the Global Brain Health Institute website.. It recruits volunteers through memory clinics, hospitals, universities, community outreach and public-health campaigns, and every site uses the same standard tests and diagnostic procedures. Nobody was given a treatment: this is an observational study, and it is mainly cross-sectionalA study that mainly measures each person at one point in time, rather than following them over months or years..

Who took part

Group 1 · 1,504 people

Healthy volunteers

People who passed standard memory and thinking checks and could manage daily life independently.

Group 2 · 1,068 people

Alzheimer’s disease

People diagnosed with Alzheimer’s disease by specialist teams, using established clinical criteria.

Group 3 · 255 people

Dementia affecting behaviour or movement

People with forms of frontotemporal dementia that mainly change behaviour or movement rather than language. They include behavioural variant frontotemporal dementia, corticobasal syndrome and progressive supranuclear palsy.

Group 4 · 77 people

Dementia affecting language

People with one of two forms of primary progressive aphasia (nonfluent and semantic). In these forms of frontotemporal dementia, language problems are the main feature.

A further 24 people had mild cognitive impairmentMemory or thinking problems that can be measured on tests but are not severe enough to be called dementia.. Altogether, 67% of the people taking part were women, and their ages ranged from 18 to 88.

What we measured

Speech

Each person did seven short speaking tasks in Spanish. They described a short animated film, named words starting with a given letter (twice) or from a category such as animals (twice), and retold a short story straight away and again about 20–30 minutes later. Software turned the recordings into transcripts, which people then checked. More than 700 features were then measured automatically.

Brain age

MRI scans of brain structure, and of brain activity at rest, were analysed with brain clocksComputer models that estimate how old a brain looks from MRI scans. Here we used scans of brain structure and of brain activity at rest..

DNA age

Blood samples were analysed for chemical marks on DNA that change with age. Three established epigenetic clocksTools that estimate a person’s age from chemical marks on their DNA, called methylation, which change as the body ages., which we call DNA clocks, turned these into an estimated age.

Alzheimer’s blood marker

Blood levels of p-tau217A form of the tau protein that can be measured in a blood sample. It is one of the most important blood markers of Alzheimer’s disease in the brain..

Memory and thinking

Standard tests of memory, attention, mental flexibility, naming and overall thinking, plus questionnaires about how well people manage everyday life. Some tests use language, such as recalling a story; others don’t, such as copying and later recalling a drawing.

Life circumstances

A detailed questionnaire combined into one score, the social exposomeAll the social circumstances a person has been exposed to across their life, such as education, money, food and healthcare, combined into one score., covering education, food security, money, possessions, access to healthcare, work in childhood, how well-off people feel, childhood experiences, traumatic events and relationships.

Building the speech clock

We grouped the most useful speech features into seven families. Two describe how speech sounds: its timing, and its pitch and voice qualityWhether the voice tends towards the low or high end of its range, and how steady and clear it sounds from one vibration of the vocal cords to the next.. Five describe the words: how concrete, how specific and how common they were, their emotional tone, and the mix of words: how many were content words and how many were pronouns.

A computer model learned to estimate age from these seven families. The model was tested on people whose recordings had been held back from that round of training, so it could not simply remember them. We then worked out each person’s speech age gapThe age the speech clock estimated from a person’s speech, minus their real age. A positive gap means their speech seems older than expected for their age; a negative gap means it seems younger. and adjusted it for a known statistical bias: age estimates tend to drift towards the average, overestimating younger people and underestimating older people.

From recording to finding

1. Recording
Seven short speaking tasks
→
2. Speech features
Recordings cleaned and transcribed; more than 700 measures taken, and the most useful combined into seven family scores
→
3. Speech age
The model estimates the person’s age from their speech
→
4. Speech age gap
Estimated age minus real age, adjusted for a known bias
→
5. Comparisons
Gaps compared across groups and with the other measures

Steps 1 to 4 were done for each person; in step 3 the model had been trained on other people’s recordings. Step 5 compares people as groups.

Illustration of a clock whose sound-wave hands unravel into seven small stacks of coloured blocks

RESULTS

What we have found so far

Here is what the recordings of 2,928 people showed. In short, the way people spoke reflected more than their age. It was also linked to signs of ageing in their brain and body, to their memory and thinking, to dementia and to their life circumstances.

Open each question to see what we found.

Question 1Can the way a person speaks be used to estimate their age?

Yes. The speech clock picked up a moderate but clear age signal: on average, its estimate was about 9 years away from a person’s real age. Some families of speech features mattered more than others:

Pointing to younger-seeming speech

The mix of words. The share of content words, such as nouns, verbs and adjectives, and of pronouns, such as “I” and “he”. Together with speech timing, this family mattered most for younger-seeming speech.

Speech timing. Faster speech: more syllables per second, and shorter syllables.

How specific the words were. For example, saying “poodle” rather than “animal”.

Pointing to older-seeming speech

Pitch and voice quality. Whether the voice tended towards the low or high end of its range, and how clear rather than breathy or hoarse it sounded. It also covers tiny wobbles in pitch and loudness from one vibration of the vocal cords to the next. Together with emotional tone, this family mattered most for older-seeming speech.

Emotional tone. How much the words expressed negative feelings, and how much they reflected sadness, anger, fear, disgust or surprise.

How concrete the words were. How far people used words for things you can see or touch, such as “table”, or for ideas, such as “freedom”.

The timing and word-specificity scores each run in one direction, so faster speech and more specific words went with younger-seeming speech. For the other families, we report which way each family pointed as a whole, not each measure within it. For example, we cannot say whether a voice tending towards the low or the high end of its range went with older-seeming speech. The seventh family, how common the words were, was not linked to either direction in our analysis.

Question 2Is the gap between speech age and real age larger in people with memory and thinking problems?

Yes. Healthy volunteers had the smallest speech age gaps. People with mild cognitive impairment (a small group of 24) and every group with dementia had larger ones. The gap was larger in frontotemporal dementia than in Alzheimer’s disease, and largest of all in the forms that mainly affect language.

The speech age gap separated healthy volunteers from each dementia group better than any single type of speech measure, or even a combination of the best individual measures. The differences held after taking account of age, sex, education, country, an overall score for other illnesses and the number of languages people spoke, and when each country was left out in turn.

Question 3Is the speech age gap linked to brain age, DNA age, an Alzheimer’s blood marker, memory and thinking, and life circumstances?

Yes. People whose speech seemed older than their age also tended to have older-looking brains and DNA, lower memory and thinking scores and harder life circumstances. These measures go together, but this study can’t show that one causes another.

What older-seeming speech went with

Older-looking brain

On brain clocks from MRI scans of structure and activity, in healthy volunteers and in every dementia group. This was the strongest link.

Older-looking DNA

On all three DNA clocks, most consistently in healthy volunteers and people with Alzheimer’s disease.

More p-tau217

Higher blood levels of this Alzheimer’s marker across everyone tested, and within the Alzheimer’s group. There was no link within the other groups.

Lower memory and thinking scores

On tests of overall thinking, attention and mental flexibility, and several kinds of memory, including tests that don’t use language. The link was stronger for tests that do.

Less independence in daily life

Less ability to manage everyday activities on one’s own, in Alzheimer’s disease and in dementia affecting behaviour or movement.

Harder life circumstances

Less favourable life circumstances, in healthy volunteers and people with Alzheimer’s disease.

“Our voice appears to contain much more information about aging than we previously recognised.”

Professor Agustin Ibanez, Trinity College Dublin · Press releaseRead the press release about this study on EurekAlert.
Illustration of a clock with sound-wave hands beside a brain and a DNA ribbon, all striped in matching colours

UNEXPLORED

Unexplored questions

Like all good research, this study answered some questions and raised others. Explore the topics below to see what we still need to know.

Tap or click a pin to open each one.

Over time
Other languages
Conversation
Many languages
Mood and health
Recording
Time
Does older-seeming speech predict who will develop memory problems?
Most of our measurements come from a single point in time, so we can’t yet say whether older-seeming speech predicts later decline or dementia. Studies that follow people over several years are needed.
Language
Does the speech clock work in other languages?
Everyone in this study spoke Spanish. Earlier research on Alzheimer’s disease suggests that speech timing carries across languages better than features based on word meaning, so the clock needs testing in other languages and cultures.
Method
Would everyday conversation work as well as set tasks?
People did structured speaking tasks. Testing the clock on natural conversation would show whether it works with more everyday speech.
Language
Is speaking several languages linked to younger-seeming speech?
People who spoke more languages tended to have slightly younger-seeming speech, but this was only a trend, and language background was recorded for only some of the people taking part.
People
How much do mood and other illnesses affect the clock?
Sex, other illnesses and mood can all affect how people speak. Our group differences held after allowing for sex and an overall score for other illnesses. Future studies should look at illnesses and mood in more detail, to separate effects that only add noise from those that may be part of faster ageing itself.
Method
Do microphones, rooms and accents affect the results?
Recordings were made at many sites in different conditions, and automatic speech tools may carry biases towards some dialects or cultures. These need checking systematically.
Illustration of people in a boat with a large clock, drifting towards a coastline that fades into fog

FAQ

Frequently asked questions

Can a speech recording tell me whether I have dementia?

No. The speech clock is not a test for dementia. We found differences between groups of people, and it has not been tested as a way to diagnose one person. Because most of our measurements come from a single point in time, it also can’t show who will go on to develop dementia. Studies that follow people over time, and in other languages, are needed first.

What does it mean if someone’s speech seems older than their age?

In this study it was a pattern across groups of people, not a verdict on one person. A single person’s speech can be affected by many things, such as mood, health or the recording itself.

Why was the study done in Latin America?

People in Latin America have been underrepresented in dementia research. ReDLat brings together research centres across the region to expand that research, and this study used the data its teams collected.

Why can’t anyone download the speech data?

Because features of a person’s speech can identify them. Researchers can apply for the individual-level data through a formal data-sharing agreement approved by an ethics board, a process that can take up to 12 weeks. The analysis code and summary data are openly available.

Illustration of a glass clock with small lanterns floating among its sound-wave hands

TIMELINE

The long journey

January 2020

ReDLat begins

Leaders in dementia research from Latin America meet at the University of California, San Francisco, to launch ReDLat, the consortium through which this study’s data were collected.

January 2020
Illustration of five clocks along a winding road, their sound-wave hands moving on a little in each

IMPACT

What’s happened

Here’s some of what we’ve published and contributed to. We’ll keep adding to this as the work develops.

Research paper

Published in Science Advances

Hernandez et al., “Speech clocks decode dementia phenotypes, social exposome, and biological aging”, Science Advances 12, eaef9864, 30 September 2026, open access. Read the paperRead the original research paper: Hernandez et al., Science Advances, 30 September 2026..

Open code

Analysis code openly available

The code and summary data are shared on ZenodoOpen the study’s code and data on Zenodo, an open research archive. and GitHubOpen the study’s code repository on GitHub.. Researchers can apply for the individual-level data through a formal data-sharing agreement.

Illustration of a clock with sound-wave hands rising like a pop-up sculpture from an open journal

FUTURE

Hopes for the future

Through this research, we hope to:

  • develop low-cost ways to monitor healthy and faster ageing from speech recordings

  • see speech clocks, perhaps combined with other biomarkers, complement more expensive measures of ageing

  • help make brain health research fairer across countries and cultures

Illustration of many hands of different skin tones holding up a clock with sound-wave hands at sunrise

TEAM

Who are we?

Lead Investigator

Professor Agustin Ibanez

Universidad Adolfo Ibáñez, Chile
Universidad de San Andrés, Argentina
Trinity College Dublin, Ireland
Istanbul Medipol University, Türkiye
Pasqual Maragall Foundation, Spain

Co-investigators

Members of the research team gathered in a lecture room for a group photo
Hernan Hernandez
Universidad Adolfo Ibáñez, Chile
Lizeth Katherine Pedraza
Universidad de La Sabana, Colombia
Hernando Santamaria-Garcia
Pontificia Universidad Javeriana, Colombia
Hospital Universitario San Ignacio, Colombia
Universidad Nacional de Colombia, Colombia
Fundación Santa Fe de Bogotá, Colombia
Universidad de los Andes, Colombia
Sebastian Moguilner
Universidad Adolfo Ibáñez, Chile
Universidad de San Andrés, Argentina
Harvard University, USA
Agustina Legaz
Universidad Adolfo Ibáñez, Chile
Universidad de San Andrés, Argentina
University of California, San Francisco, USA
Pavel Prado
Universidad San Sebastián, Chile
Show 45 more
Jhosmary Cuadros
Universidad Adolfo Ibáñez, Chile
Universidad Nacional Experimental del Táchira, Venezuela
Universidad Técnica Federico Santa María, Chile
Lucía Amoruso
Universidad de San Andrés, Argentina
Basque Center on Cognition, Brain and Language, Spain
Ikerbasque, Basque Foundation for Science, Spain
Liset Gonzalez
Universidad Adolfo Ibáñez, Chile
Universidad de Concepción, Chile
Damián Dellavale
CONICET, Argentina
Trinity College Dublin, Ireland
Juan Pablo Espinoza–Puelles
Universidad Adolfo Ibáñez, Chile
Pontificia Universidad Católica de Valparaíso, Chile
Javier Palma Espinosa
Universidad Adolfo Ibáñez, Chile
Centro Interdisciplinario de Neurociencia de Valparaíso, Chile
Cecilia Jarne
Universidad Adolfo Ibáñez, Chile
CONICET, Argentina
Universidad Nacional de Quilmes, Argentina
Aarhus Universitetshospital, Denmark
Fabio Mattiussi
Universidad Adolfo Ibáñez, Chile
Universitat Pompeu Fabra, Spain
Matías Caccia
Universidad de San Andrés, Argentina
Alejandro Sosa Welford
Universidad de San Andrés, Argentina
CONICET, Argentina
Nicolás Pelella
Universidad de San Andrés, Argentina
Jeremías Inchauspe
Universidad de San Andrés, Argentina
CONICET, Argentina
Franco J. Ferrante
Universidad de San Andrés, Argentina
CONICET, Argentina
Universidad de Buenos Aires, Argentina
Gonzalo Pérez
Universidad de San Andrés, Argentina
CONICET, Argentina
Universidad de Buenos Aires, Argentina
Marcelo Adrián Maito
Universidad Adolfo Ibáñez, Chile
Universidad de San Andrés, Argentina
Guido Rocatti
Universidad de San Andrés, Argentina
Maria Eugenia Godoy
Universidad Adolfo Ibáñez, Chile
Universidad de San Andrés, Argentina
Trinity College Dublin, Ireland
Joaquin Migeot
Universidad Adolfo Ibáñez, Chile
Trinity College Dublin, Ireland
Paulina Orellana
Universidad Adolfo Ibáñez, Chile
Ariel Caviedes
Universidad Adolfo Ibáñez, Chile
Martin Bruno
CONICET, Argentina
Universidad Católica de Cuyo, Argentina
Leonel Takada
University of São Paulo, Brazil
Andrea Slachevsky
Centro Gerociencias, Salud Mental y Metabolismo (GERO), Chile
Universidad de Chile, Chile
Clínica Alemana-Universidad del Desarrollo, Chile
Maria I Behrens
Clínica Alemana-Universidad del Desarrollo, Chile
Universidad de Chile, Chile
Bárbara Bruna
Universidad de Chile, Chile
David Aguillon
Universidad de Antioquia, Colombia
Lina Zapata
University of California, San Francisco, USA
Fundación Valle del Lili, Colombia
Universidad Icesi, Colombia
Jose Alberto Avila-Funes
Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico
Nilton Custodio
Instituto Peruano de Neurociencias, Peru
Bruce Miller
University of California, San Francisco, USA
Maria Luisa Gorno-Tempini
University of California, San Francisco, USA
Trinity College Dublin, Ireland
Stefanie Pina Escudero
University of California, San Francisco, USA
Pablo Reyes
Pontificia Universidad Javeriana, Colombia
Hospital Universitario San Ignacio, Colombia
Kun Hu
Massachusetts General Hospital, USA
Harvard University, USA
Maira Okada de Oliveira
University of California, San Francisco, USA
University of São Paulo, Brazil
Carlos Coronel-Oliveros
Universidad Adolfo Ibáñez, Chile
Trinity College Dublin, Ireland
Universidad de Chile, Chile
Josephine Cruzat
Universidad Adolfo Ibáñez, Chile
Juan Felipe Cardona
Universidad del Valle, Colombia
Michael Corley
University of California San Diego, USA
Irene B. Meier
Davos Alzheimer’s Collaborative, USA
Vaibhav A. Narayan
Davos Alzheimer’s Collaborative, USA
King’s College London, UK
Enzo Tagliazucchi
Universidad Adolfo Ibáñez, Chile
Universidad de Buenos Aires, Argentina
Sandra Baez
Trinity College Dublin, Ireland
Universidad de los Andes, Colombia
Claudia Duran-Aniotz
Universidad Adolfo Ibáñez, Chile
Adolfo M. García
Universidad de San Andrés, Argentina
University of California, San Francisco, USA
Universidad de Santiago de Chile, Chile

Argentina

CONICET
National Scientific and Technical Research Council, Argentina

Universidad Católica de Cuyo
San Juan, Argentina

Universidad Nacional de Quilmes
Buenos Aires, Argentina

Universidad de Buenos Aires
Buenos Aires, Argentina

Universidad de San Andrés
Buenos Aires, Argentina

Brazil

University of São Paulo
São Paulo, Brazil

Chile

Centro Gerociencias, Salud Mental y Metabolismo (GERO)
Santiago, Chile

Centro Interdisciplinario de Neurociencia de Valparaíso
Valparaíso, Chile

Clínica Alemana-Universidad del Desarrollo
Santiago, Chile

Pontificia Universidad Católica de Valparaíso
Valparaíso, Chile

Universidad Adolfo Ibáñez
Santiago, Chile

Universidad San Sebastián
Santiago, Chile

Universidad Técnica Federico Santa María
Valparaíso, Chile

Universidad de Chile
Santiago, Chile

Universidad de Concepción
Concepción, Chile

Universidad de Santiago de Chile
Santiago, Chile

Colombia

Fundación Santa Fe de Bogotá
Bogotá, Colombia

Fundación Valle del Lili
Cali, Colombia

Hospital Universitario San Ignacio
Bogotá, Colombia

Pontificia Universidad Javeriana
Bogotá, Colombia

Universidad Icesi
Cali, Colombia

Universidad Nacional de Colombia
Bogotá, Colombia

Universidad de Antioquia
Medellín, Colombia

Universidad de La Sabana
Chía, Colombia

Universidad de los Andes
Bogotá, Colombia

Universidad del Valle
Cali, Colombia

Denmark

Aarhus Universitetshospital
Aarhus, Denmark

Ireland

Trinity College Dublin
Dublin, Ireland

Mexico

Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán
Mexico City, Mexico

Peru

Instituto Peruano de Neurociencias
Lima, Peru

Spain

Basque Center on Cognition, Brain and Language
San Sebastián, Spain

Ikerbasque, Basque Foundation for Science
Bilbao, Spain

Pasqual Maragall Foundation
Barcelona, Spain

Universitat Pompeu Fabra
Barcelona, Spain

Türkiye

Istanbul Medipol University
Istanbul, Türkiye

UK

King’s College London
London, UK

USA

Davos Alzheimer’s Collaborative
Wayne, Pennsylvania, USA

Harvard University
Harvard Medical School, Boston, USA

Massachusetts General Hospital
Boston, USA

University of California San Diego
La Jolla, USA

University of California, San Francisco
San Francisco, USA

Venezuela

Universidad Nacional Experimental del Táchira
San Cristóbal, Venezuela

Funders

National Institutes of Health
National Institute on Aging, Fogarty International Center and National Institute on Drug Abuse, USA

Alzheimer’s Association
USA

Rainwater Charitable Foundation
Tau Consortium, USA

Bluefield Project to Cure FTD
USA

Global Brain Health Institute
University of California, San Francisco, USA, and Trinity College Dublin, Ireland

ANID – Agencia Nacional de Investigación y Desarrollo
FONDECYT, FONDAP, FONDEF, ANILLOS and doctoral scholarships, Chile

Wellcome
including the BRAIN-CLIMA and GRACE-Epi awards, UK

Wellcome Leap
CARE Program

European Commission
Marie Skłodowska-Curie Actions (CliCBrain)

JPND – EU Joint Programme – Neurodegenerative Disease Research
JPND-Care and DISCeRN 2025

Davos Alzheimer’s Collaborative
USA

Agencia Nacional de Promoción Científica y Tecnológica
Argentina

Spanish Ministry of Science and Innovation
Ramón y Cajal programme, Spain

Alzheimer’s Society
UK

Illustration of people seen from behind studying a clock with sound-wave hands painted over a world map

CONTACT

Get involved

Questions about the project? Get in touch with the research team.

Research team

Lead investigator

Professor Agustin Ibanez

Global Brain Health Institute, Trinity College Dublin

✉ agustin.ibanez@gbhi.org

Media enquiries

Thomas Deane

✉ deaneth@tcd.ie

Stay connected

The ReDLat consortium Global Brain Health Institute

Follow this research

Get an email whenever we publish new findings on this project.

Join the conversation

If a voice recording could show that your brain seemed older than your age, would you want to know?

Should low-cost tools such as voice recordings be developed first for places where brain scans and blood tests are hard to get?

How would you feel about a computer analysing recordings of your voice to check on your health?

Illustration of two hands holding a glowing clock like a lantern, with ribbons of sound waves flowing out from it
Scroll to Top