Study Reveals Speech Patterns May Predict Cognitive Decline

Research conducted by a team from Baycrest Hospital, the University of Toronto, and York University suggests that natural speech patterns could serve as indicators of cognitive decline and potentially signal early signs of dementia. Published recently, the study highlights how traits such as prolonged pauses, reliance on filler words, and difficulties in word retrieval may reflect deteriorating brain health among older adults.

The investigation, led by Jed Meltzer, senior author of the study, was motivated by a keen interest in understanding how changes in speech can reveal cognitive decline. Meltzer explained in an interview with CTV News that the research involved two cognitive assessment experiments: one in 2019 with 67 healthy adults aged 65 to 75, and another in 2023 with 174 participants aged 18 to 90.

Participants underwent standard mental ability tests, including colour-word interference tasks. They were also recorded while describing complex images depicting everyday scenarios. Meltzer stated, “We analyzed the speech samples with a very powerful computer algorithm to see if we could find aspects of their speech that predicted their mental abilities.”

The analysis employed advanced technology developed by Winterlight Labs, a Toronto-based company, which specializes in detecting cognitive impairment through speech analysis. The algorithm examines several factors, including the complexity of sentences, the frequency of repeated words, and the ability to use rare vocabulary.

This approach is pioneering, as noted in a press release from Baycrest, marking one of the first studies to establish a direct link between natural speech patterns and cognitive function. The findings revealed that participants exhibited notable difficulties with timing-related aspects of speech, such as excessive pauses and the use of filler words like “um” and “ah,” which signal cognitive deterioration.

Meltzer pointed out that while frequent use of filler words does not automatically indicate cognitive delay, an increase over time could warrant attention. He remarked, “If you do it more than you did five years ago, maybe you want to take a look at your cardiovascular health.”

Implications for Early Detection

Although Meltzer emphasized that the research does not provide a definitive diagnosis of cognitive decline, it offers a valuable early detection method. This can assist healthcare professionals in monitoring gradual changes in brain health over time. “If you test someone once, they may have dementia, or they may not… but you can’t definitively diagnose someone from one examination,” he explained. “However, if you test somebody twice, with one year in between, if someone is undergoing a serious decline, that will be detectable by the comparison between the two tests.”

This promising development raises hopes that the Baycrest team’s findings could play a role in evaluating the effectiveness of dementia treatments and medications.

Modifiable Risk Factors

Regarding the prevention of cognitive disorders such as dementia, Meltzer identified several modifiable risk factors that individuals can address. These include reducing social isolation and loneliness, as well as increasing physical activity. He stated, “The single best thing you can do is go on a walk with your friend.”

Other beneficial factors may include bilingualism and higher education, which, while they may not halt brain damage, contribute to greater resilience against it. On the other hand, aging and genetic predispositions are non-modifiable risk factors that significantly influence cognitive health.

As research continues to explore the intricate relationships between speech, cognition, and health, findings from this study may help pave the way for innovative approaches to early detection and intervention in cognitive decline.