The way your brain waves behave while you sleep may reveal how fast your brain is aging — and whether dementia is on the horizon. Researchers at UC San Francisco and Beth Israel Deaconess Medical Center developed a machine-learning model that estimates a person's 'brain age' from electrical signals recorded during sleep, and found that an older-than-expected brain age is a powerful warning sign.

The model, published in JAMA Network Open, combines 13 subtle features found within EEG recordings. Applied to about 7,000 people across five studies — participants aged 40 to 94, followed for 3.5 to 17 years, with about 1,000 developing dementia — the analysis showed that for every 10-year increase between estimated brain age and chronological age, the risk of developing dementia rose by nearly 40 percent. People whose estimated brain age was younger than their actual age had lower risk.

'Broad sleep metrics don't fully capture the complex multidimensional nature of sleep physiology,' said senior author Yue Leng, associate professor of psychiatry at UCSF. Traditional measures such as time in different sleep stages or sleep efficiency showed no meaningful link to dementia in earlier pooled analyses. Several of the EEG features the model uses are already known to support memory, including delta waves of deep sleep and sleep spindles, brief bursts believed to help the brain store memories. Notably, large sudden spikes in EEG signals — a feature called kurtosis — were associated with lower dementia risk.

The association held even after accounting for education, smoking, BMI, physical activity, other medical conditions and genetic risk factors. Because EEG can be recorded non-invasively — and increasingly by wearables — the researchers see a future where sleep-based brain age helps assess dementia risk outside the clinic. 'Brain age is calculated from sleep brain waves,' Leng said. 'We know that brain activity during sleep provides a measurable window into how well the brain is aging.'

Sources