Sleep Patterns Linked to Dementia Risk, Study Finds

Machine learning analysis of brain waves during sleep may help identify those at high risk of developing dementia.

Mar. 20, 2026 at 2:04am by Ben Kaplan

Got story updates? Submit your updates here. ›

A study led by researchers at UC San Francisco and Beth Israel Deaconess Medical Center found that when a person's "brain age" estimated from sleep signals using EEG exceeds their actual age, the risk of developing dementia increases. The study analyzed brain wave patterns during sleep in over 7,000 participants aged 40-94, and found that certain sleep EEG features like delta waves, sleep spindles, and sudden large spikes were associated with either higher or lower dementia risk.

Why it matters

This research provides new insights into how sleep physiology is linked to cognitive health and aging. The ability to detect dementia risk from sleep EEG signals could enable earlier intervention and prevention, as sleep disorders are often treatable. The findings also raise the possibility that improving sleep health may influence brain aging.

The details

The study used a machine learning model to integrate 13 different features of brain waves from EEG recordings during sleep. Participants were followed for 3.5 to 17 years, during which time about 1,000 developed dementia. The researchers found that for every 10-year increase in "brain age" compared to actual age, dementia risk rose by nearly 40%. Conversely, if brain age was lower than actual age, dementia risk was lower. Specific sleep EEG patterns like delta waves, sleep spindles, and sudden large spikes were linked to either higher or lower dementia risk.

  • The study was published on March 19, 2026 in JAMA Network Open.
  • Participants were aged between 40 and 94 and were followed for 3.5 to 17 years.

The players

Yue Leng

Associate professor of psychiatry at the UCSF School of Medicine and senior author of the study.

Haoqi Sun

Assistant professor of neurology at Beth Israel Deaconess Medical Center and first author who developed the machine learning model.

Robert J. Thomas

Co-author from Beth Israel Deaconess Medical Center who developed the machine learning model with Sun.

M. Brandon Westover

Co-author from Beth Israel Deaconess Medical Center who developed the machine learning model with Sun.

UC San Francisco (UCSF)

The university where Leng is a professor and which led the research study.

Got photos? Submit your photos here. ›

What they’re saying

“Brain age is calculated from sleep brain waves. We know that brain activity during sleep provides a measurable window into how well the brain is aging.”

— Yue Leng, Associate professor of psychiatry, UCSF School of Medicine

“Better body management, such as lowering body mass index and increasing exercise to reduce the likelihood of apnea, may have an impact. But there's no magic pill to improve brain health.”

— Haoqi Sun, Assistant professor of neurology, Beth Israel Deaconess Medical Center

What’s next

The researchers noted that because sleep EEG signals can be collected noninvasively, brain age could eventually help detect dementia risk using wearable technologies in non-clinical settings.

The takeaway

This study provides important new insights into the connection between sleep physiology and cognitive aging, raising the possibility that improving sleep health could help slow brain aging and reduce dementia risk.