Scientists use “sleep age” to infer long-term health

Summary: Sleep age, the predicted age that correlates with a person’s sleep health, can be a predictor of overall health and mortality risk.

source: Stanford

The numbers tell a story. From your credit score to your age, the metrics predict a variety of outcomes, whether it’s your likelihood of getting a loan or your risk of heart disease. Now, researchers at Stanford Medicine have described another telltale sign — one that can predict mortality. It’s called the sleep age.

Sleep age is a predicted age that correlates with a person’s health based on sleep quality. For example, if you analyze the sleep characteristics of dozens of 55-year-olds and average them, you will have an idea of ​​what sleep looks like at that age. For example, someone who is 55 years old and sleeps soundly at night with benign REM cycles could theoretically have a sleep age of 45 years.

Sleep expert Emmanuel Mignot, M.D., Ph.D., and colleagues analyzed about 12,000 studies, each focusing on an individual who reported characteristics of their sleep — such as chin and leg movement, breathing and heart rate. Their goal was to develop a system that ages a person’s sleep and, using machine learning, identifies the variations in sleep most closely associated with mortality.

Generally speaking, people sleep differently at different ages, with changes in sleep quality being one of the first and best-documented signs of aging and ill health. The good news: The age of sleep isn’t set in stone. We have the power to make it better.

The study, led by Mignot, Craig Reynolds, professor of sleep medicine at Stanford Medicine, appeared July 22 in npj Digital Medicine. I spoke with Mignot, who has been studying sleep for 30 years, about why sleep age matters, how it’s calculated, and what the study’s findings mean for our health.

Why study the age of sleep?

When you sleep, you’re disconnected from sensory inputs—ideally, you’re not bothered by the noisy outside world or bright lights.

During sleep, not only does the brain go through an automatic program, but heart rate and breathing also change, and variations in these can be early predictors of health disturbances. We spend about a third of our lives sleeping, so it is an essential component of our overall well-being.

It is well known that in almost any disorder sleep is one of the first things to be disturbed. For example, about five or 10 years before other symptoms appear in patients with Parkinson’s disease, a specific sleep disorder occurs during which the patient violently acts out dreams, screams, or punches a wall.

What was the most important finding of the study?

Our main finding was that sleep fragmentation – when people wake up multiple times during the night for less than a minute without remembering it – was the strongest predictor of mortality.

Although we see an association in the data, it is not known how it contributes to mortality. This is different from a person being aware that they are waking up, which occurs during sleep disorders such as insomnia.

Determining why sleep fragmentation is so detrimental to health is something we plan to explore in the future.

Can we measure our own sleep age? Can it be improved?

The code is available to doctors and researchers, but the average person would probably have trouble running it through a computer. Nevertheless, it is not deterministic. There is a huge variation.

Even if you have a greater sleep age than your chronological age, this does not mean that the risk of mortality will be higher. You see people smoking and drinking alcohol at 90 years old and you wonder, “How did this person survive this long?” There is always a huge natural variation.

Going to bed and waking up at regular times is key to improving your sleep. This means not sleeping, but making sure you are fully rested. It’s a different amount for everyone, and often the window varies slightly – for example, being a night owl versus an early bird.

Steady exposure to light – preferably outside light – during the day, keeping the sleep environment dark at night, exercising regularly but not too close to bedtime, not drinking alcohol and caffeine before bed and avoiding heavy evening meals, all of these contribute to healthy sleep. And of course, make sure any sleep disorder is treated.

How did you calculate the age of sleep in this study?

We used a machine learning program to predict sleep ages by feeding sleep study data and each person’s age into these programs. This tells us what the average sleep looks like at a certain age.

The algorithm recognizes patterns in the data and uses this information to predict the age of sleep. Once the algorithm is built, we can use it to set additional sleep ages. For some people, their sleeping age appears much older than their chronological age.

Sleep age is a predicted age that correlates with a person’s health based on sleep quality. Image is in the public domain

We can use the difference between their chronological age and their sleep age to predict mortality based on the idea that a higher sleep age is an indicator of a health problem. And we did find that people with a higher sleep age than their actual age were at increased risk of mortality based on the sleep of patients who later died.

We know from other studies that poor sleep occurs in conditions as diverse as sleep apnea, neurodegeneration, obesity, and chronic pain. It is not known how poor sleep causes, worsens, or results from these conditions.

See also

This shows a synapse map

What are the next steps with your research?

I hope to use sleep studies to better predict and treat disease before it manifests in death. This study included only 12,000 people. In the future, we will try to predict the future occurrence of heart attacks, strokes and Alzheimer’s disease, which cause mortality.

We’re working with scientists at Harvard University to collect 250,000 sleep studies. Much of the data in this larger set was collected 10 years ago, allowing us to make better predictions about mortality.

Can you imagine if we could use sleep studies to predict a person’s heart attack risk and then use that information to initiate early interventions? That would be a big deal.

About this news about sleep and mortality research

Author: Emily Moscow
source: Stanford
Contact: Emily Moskal – Stanford
Image: Image is in the public domain

Original Research: Free access.
“Age estimation from sleep studies using deep learning predicts life expectancy” by Andreas Brink-Kjaer et al. npj Digital Medicine


Age estimation from sleep studies using deep learning predicts life expectancy

Sleep disturbances increase with age and are predictors of mortality. Here, we present deep neural networks that estimate age and mortality risk using polysomnograms (PSGs).

Aging was modeled using 2,500 PSGs and tested in 10,699 PSGs from men and women in seven different cohorts aged between 20 and 90 years. Age was estimated with a mean absolute error of 5.8 ± 1.6 years, while the main measures of sleep assessment had an error of 14.9 ± ± 6.29 years.

After controlling for demographics, sleep, and health covariates, each 10-year increase in age estimation error (AEE) was associated with an increased all-cause mortality rate of 29% (95% confidence interval: 20–39%). An increase from −10 to +10 years of AEE translates into an estimated reduced life expectancy of 8.7 years (95% confidence interval: 6.1–11.4 years).

Greater AEE is mostly reflected in increased sleep fragmentation, suggesting that this is an important biomarker for future health independent of sleep apnea.

Leave a Comment