The study

The brain has a secret age. And it reveals it while we sleep

Research has demonstrated the existence of a brain age, distinct from the anagraphic age, and calculable by analysing the electrical activity of the brain during sleep

by Maria Rita Montebelli

21 April 2026

3' min read

Translated by AI

Versione italiana

Key points

Two-speed ageing
Artificial intelligence reveals secrets of the brain
But how to slow down brain ageing?

3' min read

Translated by AI

Versione italiana

In the stillness of the night, when the world shuts down, our brain continues to work. It does not go into stand-by, as we have imagined for years; on the contrary, it enters a phase of intense and refined activity, during which it releases valuable traces about its state of health. Traces that today, thanks to artificial intelligence, can finally be read. And which have revealed a surprising truth: some people's brains age faster than their bodies.

Two-speed ageing

This was discovered in research, conducted between the University of California, San Francisco and the Beth Israel Deaconess Medical Center in Boston and published in JAMA Network Open, which demonstrated the existence of a brain age, distinct from the age of registry, that can be calculated by analysing the electrical activity of the brain during sleep. When this age is higher than the registry age, the risk of developing dementia increases significantly. In fact, the numbers speak for themselves: for every ten years of difference, this risk increases by almost 40 per cent. In contrast, a brain that appears younger than the age written on the identity card would seem to be more protected against this eventuality.

American researchers came to this conclusion by following almost seven thousand people between the ages of 40 and 94 for a period that, in some cases, exceeded fifteen years. At the beginning of the study, none showed signs of dementia. Over time, however, about one thousand participants developed the condition. A fate, the data suggest, clearly readable in the electrical brain waves of their nights.

Artificial intelligence reveals secrets of the brain

The turning point of this research lies in the way sleep has been analysed. For years, studies have focused on general parameters: sleep duration, sleep efficiency, distribution between the different phases (Rem, non-Rem). Useful indicators, but a bit grossier. The 'stars and stripes' researchers, on the other hand, used an electroencephalogram and an artificial intelligence model based on machine learning, capable of detecting very fine details: thirteen microscopic characteristics of brain waves, which escape traditional observation. And it is in this invisible texture that the true age of the brain is written.

Some signals are particularly significant. The delta waves, slow and deep, typical of the most regenerative sleep, are crucial for the recovery of brain functions. So-called sleep spindles, short bursts of high-frequency activity, are linked to memory consolidation and learning processes. Then there are more complex phenomena, such as sudden spikes of electrical activity that seem to be associated with a lower risk of dementia. It is as if the brain, during sleep, engages in sophisticated maintenance. And the way it does it can reveal how much it is ageing.

One of the most striking aspects of the study is the robustness of the result. The correlation between brain age and dementia risk was significant, even taking into account variables such as level of education, smoking, body mass index, physical activity, concomitant diseases and genetic predisposition. In other words, what happens in the brain during sleep provides information that is independent of other risk factors for dementia and not influenced by either lifestyle or genetics.

The potential spin-offs of this study are considerable. Since electroencephalography is a non-invasive diagnostic technique, it is possible that in the future the risk of dementia could be predicted long before the onset of symptoms, even on an outpatient basis. Wearable devices and 'home' technologies could turn sleep into a predictive and possibly preventive tool, providing a low-cost and easily accessible window of observation on brain health.

But how to slow down brain ageing?

The most important question remains open: once it has been discovered that the brain ages faster than the rest of the course, can this process be slowed down? To date, unfortunately, there are no simple solutions or miraculous interventions. However, the study suggests a direction for the future. Improving sleep quality could directly influence the mechanisms of brain ageing. Intervening on factors such as body weight, sedentariness or disorders such as sleep apnoea not only helps improve general well-being, but could have direct effects on future cognitive health.

Every night, our brain tells a story, with a metric of electrical impulses, rhythms and imperceptible variations that we are just beginning to decipher. Inside these 'Thousand and One Nights' stories may lie one of the most important keys to predicting and possibly slowing cognitive decline.

Sleep, in short, is not an 'empty' time, a passive pause between two days of leisure or work. Rather, it is a living archive, on which the brain checks up, night after night. Learning to decipher the language of these nocturnal coupons means gaining a valuable advantage: (pre)see earlier what we usually discover too late: cognitive decay.