HMN 2025: How Verbal response time reveals hidden sleepiness in older adults

A model new study led by UCLA investigators reveals that Verbal Reaction Time (VRT), the time frame it takes a person to answer verbally, could possibly be a marker of sleepiness in older adults. The study, which measured contributors’ voice data by means of standardized cognitive assessments, reveals how VRT can passively detect excessive sleepiness, significantly amongst older folks using sedative medicines.

The work, “Predicting subjective sleepiness all through auditory cognitive testing using voice signaling analysis,” is published throughout the journal Sleep Science and Practice.

Sleepiness is a severe contributor to safety risks in daily life, nevertheless is often underreported or unnoticed, significantly amongst older adults. Excessive sleepiness contributes to motor vehicle crashes, cognitive impairment, and falls, significantly in older adults using sedative medicines like benzodiazepine receptor agonists (BZRAs). Current methods to guage sleepiness are generally intrusive or impractical for real-world use. This study provides a scalable technique to detect sleepiness to help set up at-risk folks sooner than accidents or properly being declines occur.

Researchers studied adults aged 55 and older with a historic past of insomnia and BZRA use, recruited from a deprescribing scientific trial. Participants completed memory exams via a cell app, which recorded their verbal responses. The employees measured verbal response time (VRT), the delay between the start of recording and the first spoken phrase, evaluating it to contributors’ self-reported sleepiness.

The researchers then used superior devices to test the hyperlink between how quickly of us started speaking and the way in which sleepy they felt. They moreover examined whether or not or not a computer model may exactly predict someone’s sleepiness based on their voice.

The model effectively predicted contributors’ self-reported sleepiness based on their voice recordings. People who took longer to start speaking after a instant moreover reported feeling sleepier. The computer model used throughout the study was able to appropriately set up utterly totally different ranges of sleepiness with sturdy accuracy, reaching an F1-score of 0.80 ± 0.08. (The F1-score is a measure of how correctly the model balances accuracy and consistency; 1.0 is right, and 0 means it failed.)

The voice analysis methodology moreover reliably detected when someone was speaking versus silent, with 92.5% accuracy. The outcomes advocate that voice timing may probably be a useful, low-effort technique to watch sleepiness—significantly exterior of scientific settings.

“This study reveals that one factor as simple as how quickly someone begins speaking can inform us heaps about their stage of alertness,” talked about Dr. Tue T. Te, lead creator of the assessment and a researcher on 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 software program for monitoring sleepiness all through frequently actions.”

The employees plans to extend this technique to larger and additional varied populations and uncover integration into frequently utilized sciences like smartphones and telehealth platforms. Future evaluation might look at how voice-based markers can monitor drugs outcomes or detect early indicators of cognitive decline.