A new machine-learning model that analyzes brain activity during sleep could help identify people at higher risk of developing dementia. Researchers from the University of California, San Francisco, and Beth Israel Deaconess Medical Center in Boston developed the approach to estimate “brain age” using overnight electroencephalography, or EEG, recordings.
The system calculates a person’s estimated brain age by examining electrical signals produced by the brain during sleep. Researchers found that dementia risk increased when the brain appeared biologically older than the person’s chronological age.
For every 10-year gap between estimated brain age and actual age, the likelihood of developing dementia increased by nearly 40%. By contrast, people whose estimated brain age was younger than their chronological age had a lower risk of dementia.
The findings were published in JAMA Network Open.
Machine Learning Estimates Brain Age From Sleep EEG Signals
The researchers created a machine-learning model that combines 13 microscopic features identified in EEG brain-wave recordings. They tested the model using data from approximately 7,000 participants across five separate studies.
Participants were between 40 and 94 years old, and none had dementia when their studies began. Researchers followed them for periods ranging from 3.5 to 17 years. During the follow-up period, approximately 1,000 participants developed dementia.
The analysis showed that subtle, highly detailed patterns in sleep-related brain waves may reveal information that conventional sleep measurements do not capture.
Earlier pooled analyses involving multiple participant groups found no meaningful connection between dementia risk and standard sleep measurements. These traditional measures include the amount of time people spend in each sleep stage and how efficiently they remain asleep throughout the night.
“Broad sleep metrics don’t fully capture the complex multidimensional nature of sleep physiology,” said senior author Yue Leng, MBBS, PhD, associate professor of psychiatry at the UCSF School of Medicine.
Sleep Brain-Wave Patterns May Support Memory
Several EEG features used to estimate brain age have previously been linked to memory and cognitive health.
One example is delta waves, slow, rolling electrical patterns commonly associated with deep sleep. Another is the presence of sleep spindles, brief bursts of rapid brain activity that may help the brain strengthen and store memories.
One of the study’s most notable findings involved large, sudden changes in EEG signals. This feature, known as kurtosis, was associated with a lower risk of developing dementia.
The link between older estimated brain age and increased dementia risk remained significant even after researchers adjusted for education, smoking, body mass index, physical activity, other health conditions, and genetic risk factors.
Sleep-Based Brain Age Could Support Earlier Dementia Risk Assessment
Because EEG testing is noninvasive, sleep-based brain age assessments could eventually help evaluate dementia risk outside traditional clinical settings. Future wearable devices may be able to record the brain signals needed for this type of analysis during sleep.
“Brain age is calculated from sleep brain waves,” said Leng. “We know that brain activity during sleep provides a measurable window into how well the brain is aging.”
The findings also raise the possibility that improving sleep health could influence brain aging. Leng noted that previous research has shown that treating sleep disorders can change brain-wave activity recorded during sleep.
“Better body management, such as lowering body mass index and increasing exercise to reduce the likelihood of apnea, may have an impact,” said first author Haoqi Sun, PhD, assistant professor of neurology at Beth Israel Deaconess Medical Center, who developed the model with two co-authors. “But there’s no magic pill to improve brain health.”
Study Authors and Funding
Co-authors Robert J. Thomas, MD, and M. Brandon Westover, MD, PhD, of Beth Israel Deaconess Medical Center, developed the machine-learning model with Sun. For a complete list of authors, please see the published study.
Funding was provided by the National Institutes of Health (R01NS102190, R01NS102574, R01NS107291, RF1AG064312, RF1NS120947, R01AG073410, RF1AG064312, R01NS102190, R01AG062531); National Institute on Aging (R21AG085495 and R01AG083836); National Science Foundation (2014431); National Health and Medical Research Council (GTN2009264); and the American Academy of Sleep Medicine.
Source: www.sciencedaily.com