UCLA Research Brief
A new study led by UCLA investigators shows that Verbal Reaction Time (VRT), the amount of time it takes a person to respond verbally, can be a marker of sleepiness in older adults. The study, which measured participants’ voice data through standardized cognitive assessments, show how VRT can passively detect excessive sleepiness, especially among older individuals using sedative medications.
Why it matters
Sleepiness is a major contributor to safety risks in daily life, but is often underreported or unnoticed, especially among older adults. Excessive sleepiness contributes to motor vehicle crashes, cognitive impairment, and falls, especially in older adults using sedative medications like benzodiazepine receptor agonists (BZRAs). Current methods to assess sleepiness are often intrusive or impractical for real-world use. This study provides a scalable way to detect sleepiness to help identify at-risk individuals before accidents or health declines occur.
What the study did
Researchers studied adults aged 55 and older with a history of insomnia and BZRA use, recruited from a deprescribing clinical trial. Participants completed memory tests via a mobile app, which recorded their verbal responses. The team measured verbal reaction time (VRT), the delay between the start of recording and the first spoken word, comparing it to participants’ self-reported sleepiness. The researchers then used advanced tools to study the link between how quickly people started speaking and how sleepy they felt. They also tested whether a computer model could accurately predict someone’s sleepiness based on their voice.
What they found
The model successfully predicted participants’ self-reported sleepiness based on their voice recordings. People who took longer to start speaking after a prompt also reported feeling sleepier. The computer model used in the study was able to correctly identify different levels of sleepiness with strong accuracy, achieving an F1-score of 0.80 ± 0.08. (The F1-score is a measure of how well the model balances accuracy and consistency; 1.0 is perfect, and 0 means it failed.) The voice analysis method also reliably detected when someone was speaking versus silent, with 92.5% accuracy. The results suggest that voice timing could be a useful, low-effort way to monitor sleepiness—especially outside of clinical settings.
What’s next
The team plans to expand this approach to larger and more diverse populations and explore integration into everyday technologies like smartphones and telehealth platforms. Future research may also investigate how voice-based markers can monitor medication effects or detect early signs of cognitive decline.
From the experts
“This study shows that something as simple as how quickly someone starts speaking can tell us a lot about their level of alertness,” said Dr. Tue T. Te, lead author of the study and a researcher at the David Geffen School of Medicine at UCLA and the VA Greater Los Angeles Healthcare System. “It opens the door to using voice as a passive, scalable tool for monitoring sleepiness during everyday activities.”
About the study: Predicting subjective sleepiness during auditory cognitive testing using voice signaling analysis. Published July 1, 2025, in Sleep Science and Practice. DOI: https://doi.org/10.1186/s41606-025-00141-y
