Speech reveals how fast the brain ages

Aging clocks are measurement tools that estimate a person’s biological age based on biological data, independent of their calendar age. The best-known versions use DNA patterns. But DNA tests are expensive and hard to scale. Speech analysis is not: a recording is enough.

Researchers analyzed speech recordings from nearly 3,000 people across five Latin American countries. The group included healthy participants, people with mild cognitive impairment, and people with various forms of dementia, including Alzheimer’s disease and frontotemporal dementia. The study was published in Science Advances.

Speech age deviates from calendar age

Using acoustic features such as pitch, speech rate, and pause patterns, alongside linguistic features such as word choice and sentence structure, the researchers calculated what they called a speech age gap (SAG). A positive value means someone’s speech sounds older than expected for their age. This deviation was larger as cognitive impairment became more severe.

The speech age gap also correlated with biomarkers such as phosphorylated tau, a protein associated with Alzheimer’s disease. It also correlated with results from DNA aging clocks, suggesting the speech clock is measuring something biologically real.

Accessible and broadly applicable

What sets this research apart is accessibility. A microphone and software are sufficient. That makes the method potentially useful in countries or regions where comprehensive medical diagnostics are difficult to organize. The researchers also studied culturally diverse populations, which broadens the generalizability.

Limitations remain. The clock is cross-sectional: it measures at one point in time, not longitudinally. Whether it is sensitive enough to detect treatment effects is still unknown. That is a core problem for all aging clocks. But as a low-cost, scalable tool for research in underrepresented populations, this approach offers real added value.

Read the original article