Original title: Remote Identification of APOE-ε4 Carrier Status in Cognitively Normal Adults via Speech Acoustics: Cross-Sectional Observational Study
Maryam Tavakoli, Mehrdad Dadgostar, Jordan R. Green, David H. Salat +6 · October 2026
People carrying a gene linked to Alzheimer's disease can have subtle motor changes long before memory loss. Researchers recorded eighty adults without memory problems speaking into their personal phones. A computer model analyzed sound patterns like pitch and tone across different talking exercises. Free, unscripted speech worked best, letting the system tell gene carriers apart from noncarriers with eighty-nine percent accuracy.
A mechanic listens to the sound of an idling car engine to find a worn belt before it snaps. The car runs normally around town, so you cannot see any damage just by watching it drive down the street. In this paper, adults show no signs of memory trouble in daily life, but subtle changes in how their vocal muscles move reveal an underlying genetic risk when they speak.
Alzheimer's disease causes brain changes years before memory problems appear, but testing genetic risk usually requires clinical visits and lab work. Because people can record their voices from home on their own phones, acoustic testing could become an easy, low-burden way to spot risk early. This helps researchers identify preclinical warning signs before noticeable symptoms begin.
The abstract notes that this was a cross-sectional study of 80 participants, so it does not track changes over time or prove outcomes long-term. The models internally distinguished carriers from noncarriers, meaning external validation on separate populations is still needed. In addition, the system worked far better on spontaneous speech than on scripted tasks, and combining multiple speech tasks together actually degraded classification performance.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Remote Identification of APOE-ε4 Carrier Status in Cognitively Normal Adults via Speech Acoustics: Cross-Sectional Observational Study — Maryam Tavakoli, Mehrdad Dadgostar, Jordan R. Green et al. — JMIR Formative Research — doi:10.2196/89830
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.