summary: Researchers have developed a machine learning model that can estimate a person’s “brain age” simply by analyzing their sleep.
The study reveals that if a person’s brainwaves appear “older” than their actual age, the risk of developing dementia is significantly increased. This breakthrough suggests that sleep tracking may serve as a non-invasive early warning system for cognitive decline, years before symptoms appear.
Important facts and statistics
- “Brain age” gap: for everyone 10 year increase With estimated brain age compared to actual age, the risk of dementia is almost 40%.
- Protect your youth: Conversely, participants whose brainwaves made them appear “younger” than their actual age had a significantly lower risk of dementia.
- Large scale data: The study analyzed approximately one year of EEG recordings. 7,000 participants (40-94 years) were followed for up to 17 years across five different cohorts.
- invisible pattern: Traditional sleep metrics (such as total sleep time and sleep stages) have not shown any association with dementia. The risk is 13 microstructural features Brain waves identified by AI.
- Important brainwave markers:
- Delta wave: Associated with deep and restorative sleep.
- Sleep spindle: Rapid bursts of activity are essential for memory consolidation.
- kurtosis: Sudden large spikes in brain activity. This is surprisingly lower Dementia risk.
sauce: U.C.S.F.
Machine learning analysis of brain waves recorded during sleep could help identify people at high risk of developing dementia, according to a study led by the University of California, San Francisco and Beth Israel Deaconess Medical Center in Boston.
This study found that when a person’s “brain age”, estimated from sleep signals using brain waves, exceeds their chronological age, the risk of dementia increases.
For every 10 years of increase in brain age relative to chronological age, dementia risk increased by nearly 40%. Conversely, if your brain age is lower than your chronological age, your risk of dementia will be lower.
This study JAMA network open March 19th.
The researchers used a machine learning model that integrated 13 microstructural features of brain waves from EEG recordings. Data came from approximately 7,000 participants enrolled in five studies.
The participants ranged in age from 40 to 94 years old, and none had dementia at the start of the study. They were followed for 3.5 to 17 years, during which time approximately 1,000 participants developed the disorder.
Researchers have discovered that analyzing the subtle patterns of brain waves during sleep can provide insights that traditional sleep metrics often miss. Previous pooled analyzes across several participant cohorts found no significant associations between dementia risk and traditional sleep measures, such as time spent in different sleep stages or overall sleep efficiency.
“Broad sleep metrics do not fully capture the complex and multidimensional nature of sleep physiology,” said senior author Yue Leng, MBBS, PhD, associate professor of psychiatry at UCSF School of Medicine.
Brain wave patterns associated with cognitive health
Several sleep EEG patterns that contribute to brain age are known to play a role in brain health and memory. These include delta waves, which form rolling wave patterns associated with deep sleep, and sleep spindles, short, fast-frequency brain activity associated with memory consolidation.
One of the most notable findings was that sudden large spikes seen in brain waves, known as kurtosis, were associated with a lower risk of dementia.
Researchers also found that the relationship between “older” brain age and dementia risk remained significant even after accounting for factors such as education, smoking, BMI, physical activity, and other health conditions and genetic risk factors.
Possibility of early detection
Because sleep EEG signals can be collected non-invasively, brain age could eventually be used to detect dementia risk in non-clinical settings, including through the use of wearable technology, the researchers said.
“Brain age is calculated from sleep brain waves,” Ren says. “We know that brain activity during sleep is a measure of the extent of brain aging.”
The findings also raise the possibility that improving sleep health may impact brain aging. Leng noted that earlier studies found that treatment for sleep disorders can alter sleep-related brain wave patterns.
“Better physical management, such as lowering body mass index and increasing physical activity to reduce the likelihood of apnea, may have an impact,” said lead author Dr. Haoki Sun, assistant professor of neurology at Beth Israel Deaconess Medical Center, who developed the model with two co-authors*. “But there is no magic bullet for improving brain health.”
Co-author:* Robert J. Thomas, MD, and M. Brandon Westover, MD, PhD, of Beth Israel Deaconess Medical Center, worked with Sun to develop the machine learning model. See the paper for other authors.
Funding: National Institute on Aging (R21AG085495 and R01AG083836). National Science Foundation (2014431); National Health and Medical Research Council (GTN2009264); American Academy of Sleep Medicine.
Answers to key questions:
answer: Not necessarily. This means that the brain’s electrical activity during sleep shows signs of accelerated aging and wear and tear. The study found that this pattern was detected even in people who were cognitively healthy at the time, and could serve as a “predictive window” rather than a diagnosis of current illness.
answer: During sleep, the brain performs important “housekeeping” tasks, such as removing toxic proteins such as amyloid beta, which is associated with Alzheimer’s disease. Slow or disrupted brain waves may indicate that these repair processes are failing, leading to the accumulation of long-term damage.
answer: That’s not all amount Sleeping, but quality About brain waves. Although there is no “magic pill,” researchers suggest that managing health factors such as BMI and exercise that reduce symptoms such as sleep apnea may help maintain healthier brainwave patterns and slow brain aging.
Editorial note:
- This article was edited by the editors of Neuroscience News.
- Journal articles were reviewed in full text.
- Additional context added by staff.
About this sleep and dementia research news
author: Suzanne Lee
sauce: U.C.S.F.
contact: Suzanne Lee – UCSF
image: Image credited to Neuroscience News
Original research: Open access.
“Machine Learning-Based Sleep EEG Brain Age Index and Dementia Risk: A Meta-Analysis of Individual Participant Data” Haoqi Sun, Sasha Milton, Yi Fang, Hash Brown Taha, Shreya Shiju, Robert J. Thomas, Wolfgang Ganglberger, Matthew P. Pase, Timothy Hughes, Shaun Purcell, Susan Redline, Katie L. Stone, Authors Christine Yaffe, M. Brandon Westover, Yue Ren. JAMA network open
DOI:10.1001/jamanetworkopen.2026.1521
abstract
Machine learning-based sleep EEG brain age index and dementia risk: a meta-analysis of individual participant data
importance
The microstructure of sleep electroencephalograms (EEG) is closely related to cognition and changes with age. However, its multidimensional nature makes it difficult to interpret using traditional approaches. The machine learning-based EEG Brain Age Index (BAI) measures the deviation between sleep EEG-based brain age and chronological age.
objective
To clarify the relationship between sleep BAI and the onset of dementia in community-dwelling people.
data source
This individual participant data (IPD) meta-analysis pooled sleep study data from five community-based longitudinal cohorts. These cohorts include the Multi-Ethnic Study of Atherosclerosis (MESA; 2010-2013), the Atherosclerosis Risk in Communities (ARIC) Study (1987-1989), the Framingham Heart Study-Oscendary Study (FHS-OS; 1995-1998), and the Male Osteoporotic Fracture Study (MrOS; 1995-1998). 2003-2005), and the Study of Osteoporotic Fractures (SOF; 2002-2004).
Research selection
Adults (18 years and older) without dementia at the time of polysomnography were included.
Data extraction and synthesis
BAI was calculated using interpretable machine learning incorporating sleep EEG features extracted from the central channel of a nighttime home-based polysleep test. Fine-Gray models were used to assess the association between BAI and incident dementia within each cohort, accounting for death as a competing risk. We then used random-effects meta-analysis to pool cohort-specific estimates. The analysis was conducted from March 2024 to September 2025.
Main results and measures
Dementia onset or potential dementia was determined for each cohort, with death as a competing risk.
result
This meta-analysis included 7,105 participants in MESA (n = 1,802, mean [SD] Age, 69.3 years old [9.0] year. 956 women [53.1%]), ARIC (n = 1796; 62.5 [5.7] year. 918 women [51.1%]), FHS-OS (n = 617; 59.5 [8.9] year. 318 women [51.5%]), MrOS (n = 2639 men) [100%]; 76.0 [5.3] year), and SOF (n = 251 women [100%]; 82.7 [2.9] Year) Cohort. Median (IQR) time to dementia was 4.8 (4.2-5.6) years in the MESA cohort (n = 119). [6.6%]), 16.9 (14.9-19.8) years in the ARIC cohort (n = 354) [19.7%]), 13.1 (8.5-16.2) years in the FHS-OS cohort (n = 59) [9.6%]), 3.6 (1.3–7.1) years in the MrOS cohort (n = 470) [17.8%]), 4.6 (4.2-5.2) years in the SOF cohort (n = 86) [34.3%]).
Across the cohort, each 10-year increase in BAI was associated with a 39% increased risk of developing dementia (hazard ratio) [HR]1.39 [95% CI, 1.21-1.59]; P< .001) after adjusting for covariates. These associations remained after additional adjustment for comorbidities and apnea-hypopnea index score (HR, 1.31). [95% CI, 1.14-1.50]; P< .001) and apolipoprotein E ε4 (HR, 1.22) [95% CI, 1.02-1.45]; P= .03), which was consistent across gender and age groups.
Conclusion and relevance
In this IPD meta-analysis, higher sleep EEG-based BAI was associated with higher risk of developing dementia. These findings highlight the need to evaluate the predictive value of BAI as a non-invasive digital marker for early detection of dementia in the community.
