An automated intelligence tool known as the “speech clock” can estimate a person’s biological age and detect early signs of cognitive decline, Alzheimer’s disease, and frontotemporal dementia by analyzing short voice recordings, according to a study published in Science Advances and covered by German wave.
How the Speech Clock Analyzes Biological Age Through Voice
The automated tool evaluates hundreds of acoustic and linguistic features from audio recordings to measure how someone speaks alongside what they say. Researchers trained the model using data from 2,928 Spanish-speaking participants ranging in age from 18 to 88, with an average age of 65, across Argentina, Chile, Colombia, México, and Peru. The study group included healthy adults, individuals with mild cognitive impairment, patients with Alzheimer’s disease, and people with various forms of frontotemporal dementia. A model of artificial intelligence analyzed these voice recordings from nearly 3,000 people across five Latin American countries.
The system evaluates speech rate, pauses, tone, emotional content, vocabulary choice, semantic precision, and the overall organization of verbal output. By combining these metrics, the machine learning models estimate chronological age and identify discrepancies known as the “speech age gap.” When this gap widens, a person’s voice sounds chronologically older than their actual birth date, which correlates with accelerated aging markers observed in clinical tests. The gap also relates to epigenetic aging, a method of measuring biological age based on chemical changes in DNA.
Clinical Comparisons and Biomarker Links Across Latin America
Healthy participants consistently displayed the narrowest speech age gaps, whereas wider gaps appeared progressively among patients with Alzheimer’s disease and frontotemporal variants of dementia. According to findings highlighted by Euronews, greater speech age acceleration also mapped directly to lower scores in global cognition, executive function, functional capabilities, and multiple forms of memory. Those with older-sounding speech profiles performed worse even on non-verbal tests, including certain visual memory assessments, suggesting the tool captures broader cognitive processes rather than just pronunciation or vocabulary.
Additional biological markers reinforced these patterns. Brain magnetic resonance imaging scans revealed greater structural loss and altered brain activity in participants with older-sounding speech profiles. Furthermore, higher speech ages linked directly to elevated levels of p-Tau217, a key blood protein utilized as a biomarker for Alzheimer’s disease, as noted by Canal26.
Social Context and Non-Invasive Diagnostic Potential
Beyond neurological markers, the research demonstrates that vocal aging reflects accumulated life experiences. Agustín Ibáñez, professor of brain health at the Trinity College Dublin School of Medicine, stated that the voice captures chronological time alongside signals from cognition, systemic biology, and the accumulated social environment. The study found that an accelerated speech age correlated with unfavorable lifelong social contexts, such as lower economic status, poor nutrition, and limited healthcare access. “Nuestra voz parece contener mucha más información sobre el envejecimiento de la que habíamos reconocido hasta ahora,” asegura Ibáñez.
Traditional assessments for biological aging and neurodegeneration typically require expensive magnetic resonance imaging scans, molecular assays, or specialized clinical evaluations. In contrast, the speech clock offers a non-invasive, low-cost alternative that can be administered remotely on standard devices. This accessibility holds particular value for Latin American populations, which have historically lacked adequate representation in dementia research. “Esto plantea la posibilidad de que pueda ser usada como método complementario de diagnóstico,” señala Ibáñez.
Frequently Asked Questions About the Speech Clock
Is the speech clock currently used as a diagnostic test for dementia?
No, the study authors emphasize that the speech clock is not yet a diagnostic test. Additional clinical validation across diverse languages and cultures is required before healthcare providers can implement it in routine practice.
Can the tool predict who will develop cognitive decline in the future?
The current research does not establish whether an older-sounding speech profile can predict which healthy individuals will eventually develop mild cognitive impairment or dementia later in life.
What specific tasks did study participants perform during the voice recordings?
Participants completed standardized exercises that included narrating an animated short film, retelling a story after a 20-minute delay, and naming items belonging to specific semantic categories.
Keep reading