Sleep is essential — researchers are trying to work out why

When neurobiologist Dragana Rogulja began studying sleep in her laboratory at Harvard Medical School in Boston, Massachusetts, in 2013, she had no idea what her experiments would tell her about how sleep affects the body. She wasn’t assuming — as many people do — that it’s a function mainly of the brain. She was just starting her lab and decided to throw the net wide, asking the question: if animals are sleep deprived, how long will it take until they die — and when they die, what is the specific cause?

She expected it would take years to answer these questions. But in six months, Rogulja’s postdoctoral assistant, Alexandra Vaccaro, found a tantalizing clue.

“We started looking at lifespan and saw that the moment when sleep-deprived animals would start dying was pretty much always the same,” Rogulja says. “When they lost about 90% of their sleep each night, they started dying within about ten days.” Those studies were in flies and mice — but the implications for other animals, including humans, was intriguing.

The average adult spends more than 20 years of their life sleeping, and the consequences of not getting enough are clearly harmful. But surprisingly little is known about why it’s necessary. That is, in part, because only in the past couple of decades have the tools become available to help scientists to understand the fundamental biological function of sleep. These include optogenetics, which involves directing laser light to specific neurons to wake a person or put them into deep sleep, and focused ultrasound, which has emerged in the past five years as a way to view the deep-brain neural oscillations that occur during sleep.

With the help of such tools, researchers are beginning to understand that sleep does more than just give our brains and bodies a respite from the rigours of the day. When we sleep, our genes, metabolism and hormones are regulated through increasing or slowing of their activity.

Finding homeostasis

What’s the most important function of the brain? It might be tempting to say a specific task, such as communicating, finding food or avoiding predators. But none of these roles is possible if the brain’s cellular machinery isn’t functioning reliably. Indeed, it is maintaining that reliability that might well be the main purpose of sleep, says Keith Hengen, a neuroscientist at Washington University in St. Louis, Missouri.

Nature Outlook: Sleep

“This was like a light bulb for me a long time ago,” Hengen says. In terms of machine learning, “if you have a deep-learning network, and you continue to let it learn, it’ll exhibit catastrophic forgetting — the wheels come off and the thing stops. It just fails.”

Biology has solved this problem of brain overload by resetting to a ‘critical point’ — a process that occurs mainly during sleep. During the day, the brain is consumed with the many things it has to do, such as learning, coordinating movement and recognizing faces. These tasks all change connections in the brain, essentially undermining the platform that is set up for that learning. Sleep, Hengen contends, allows the brain to reset. According to this hypothesis, the restoration, which occurs through a common biological process known as homeostasis, is the ‘set point’ at which the brain is able to maximize information-processing potential¹.

Hengen compares homeostasis to the familiar workings of a home heating system. A drop in a room’s temperature below the set point causes the heater to turn on, restoring the environment to the desired level of warmth.

Likewise, when the brain is further from its critical point, a person will feel more tired and need to fall asleep to reset. The closer the brain is to its critical point the more likely the person is to stay awake, because they will have plenty of capacity for doing and learning.

Sleep has several mechanisms that help the brain to reset after the stimulus and onslaught of information of each day. According to a 2024 study, parts of the hippocampus go silent during sleep, preparing its neurons for use again the next day². And a 2017 study³ found that, in the brains of mice, sleep restores the balance of junctions between neurons called synapses, which grow during the day and shrink during sleep.

Rogulja’s team confirmed that sleep is indeed important for homeostasis, but that the organ in which this takes place is not the brain.

After determining the ten-day timeline for sleep deprivation and death in flies, Vaccaro, who had previously studied sleep deprivation and ageing, began looking for markers of ageing in these insects. She and her colleagues found that the concentrations of reactive oxygen species (ROS) in the guts of flies were at their highest when the animals started dying. Small amounts of ROS are beneficial — they regulate the body’s immune response and help cells to defend against pathogens. But without sleep these molecules accumulate to toxic levels⁴.

“We looked at tissues all over the body,” Vaccaro says. “We looked at the brain and didn’t see anything. The one tissue that showed the most damage when the flies started to die was in the gut.”

The same happened in mice. When the animals were sleep deprived, then allowed to sleep for small periods, during that time, they stopped expressing genes for fat uptake. Organelles called mitochondria sense this nutrient deficiency and signal cellular enzymes to make ROS to stimulate the proliferation of cells in the gut that are better at absorption, Rogulja says. After just one day of restricting sleep in mice, Rogulja could see fats in the animals’ intestines that weren’t entering circulation. After five days, their guts were filled with fat although their bodies were starved of nutrients.

In rats, most genes seem to be upregulated while the animals are awake and downregulated during sleep. Genes that are upregulated during periods of sleep deprivation are responsible for functions such as energy metabolism, hormone reception and protein synthesis⁵. In a 2023 study, researchers in Brazil found that fruit flies that were sleep deprived showed overexpression mostly of genes that affected metabolism, glucose, triglyceride and levels of the hormone dopamine⁶.

It’s brainwashing

For decades, the main purpose of sleep was thought to be maintaining brain health. Babies and teenagers sleep for such long periods of time, according to this theory, because there is so much brain development going on at those stages.

It’s not just the amount of sleep we get that is important, however. Sleep quality, particularly the amount of deep sleep, might drastically improve memory consolidation. And researchers have found a way to artificially increase deep, or slow-wave, sleep in humans — by delivering acoustic stimulation in the form of soft pulses of ‘pink noise’ to a sleeping person through a headset. Unlike the familiar white noise, which contains equal parts of the sound spectrum, pink noise contains more low- than high-frequency sounds.

In a small study⁷, half of the participants received pink-noise acoustic stimulation and the other half wore headsets playing a different sound pattern, as a control. Researchers monitored brain activity with an electroencephalogram and found that pink-noise recipients had increased slow-wave activity. Importantly, they also had increased memory retention the next day.

“We asked them to remember word pairs during the day and then they slept,” says co-author Phyllis Zee, a neurologist and sleep-medicine specialist at the Feinberg School of Medicine in Chicago, Illinois. Overall, Zee says, those that had pink-noise stimulation performed almost 30% better on memory tests than did those who listened to the control sounds.

Sleep is thought to not only help memories to take root, but also clear the brain of anything superfluous. Neuroscientist Maiken Nedergaard at the University of Rochester in New York, has a theory of how that cleaning occurs.

In 2013, Nedergaard and her colleagues reported⁸ that one benefit of sleep might be the removal of neurotoxins from the brain. The brain’s glymphatic system is made up of what Nedergaard describes as a “doughnut-shaped tunnel” surrounding cerebral vessels that create a cell wall. In this tunnel, cerebrospinal fluid flows with interstitial fluid to remove proteins, including amyloid-β and tau, which are thought to have a role in Alzheimer’s disease.

“What we showed is, every time the heart beats, you have the arteries expanding and moving the vessel wall and putting fluid into the brain,” she says. “It’s a very beautiful biological system.” The paper reported that the interstitial space in mice grew by about 60% when they were sleeping. This drastically increased the amount of exchange of cerebrospinal fluid and interstitial fluid as well as the rate of clearance of amyloid-β.

This finding sent Nedergaard and her colleagues searching for the drivers of that fluid clearance. The heart is one way, but Nedergaard says it’s more likely to account for movement of fluid than its clearance. In February⁹, she and her colleagues reported that norepinephrine, which induces vasoconstriction, is the main promoter of clearance through “spontaneous, rhythmic constriction and dilation of arteries”. And regular oscillations of norepinephrine levels in the brain occur during deep, slow-wave sleep.

“Sleep is the period where our brain doesn’t just rest, but it does all the housekeeping,” Nedergaard says. “The glial cells are the cleaning crew that come in when the brain is quiet and do the things that other organs do when we are awake. It’s a function of the brain that’s probably not compatible with wakefulness.”

Nedergaard’s theory about protein clearance during sleep has recently been disputed as an incomplete explanation, especially because it pertains to the link between sleep and Alzheimer’s disease. The most prominent challenge to her ideas has come from biophysicist Nick Franks at Imperial College London. In May 2024, Franks and his colleagues used a fluorescent tracer to show that brain clearance is actually reduced during sleep¹⁰.

Nedergaard defends her version of what happens during sleep. Franks, she contends, “misdefines brain clearance as the movement of a tracer from one location to another within the brain. In contrast, the correct definition of brain clearance is that the tracer must exit the brain entirely. This is akin to moving garbage from the kitchen to the bedroom and calling the house clean.”

Franks disputes this characterization and maintains that his results directly contradict Nedergaard’s. “We showed that tracer injected into the brain was retained at higher concentrations during sleep, compared to waking — in other words, less was cleared,” he says, adding that his data “cannot be reconciled with her conclusions about clearance during sleep”.

Moreover, Nedergaard says that her ideas about clearance align with a well-established theory about why people sleep — the synaptic homeostasis hypothesis. This says, in essence, that sleep is the price that animals pay for brain plasticity. During the day, the brain consumes energy, takes in noise and other stimulus, and spends time learning. During sleep, brain activity is reduced, and cells are restored.

It does a body good

It’s intuitive to view sleep as a function of the central nervous system. But that can’t be entirely the case, because creatures without central nervous systems also show sleep-like behaviour.

More from Nature Outlooks

In 2020, researchers found that hydra, a freshwater polyp with no brain, also enter a languid state at night¹¹. These creatures show a need for sleep homeostasis as well, resting longer the day after being sleep deprived.

There is other evidence that sleep’s function goes beyond the central nervous system. Sleep promotes the release of certain hormones, for example. This process has various physiological effects, including lowering the risk of cardiovascular disease and some cancers. Inadequate sleep can slow production of beneficial hormones, including cortisol and growth hormone. And the hormone melatonin can suppress breast cancer cells’ ability to proliferate, survive, spread and develop resistance to drugs¹².

Zee and her colleagues’ study on the impact of slow-wave sleep found that people who listened to pink noise had lower spikes in morning cortisol levels and improved parasympathetic activity — slowing of the heart rate and lowering of the blood pressure. During sleep, parasympathetic activity gives the cardiovascular system a much-needed break. “We know that if you don’t sleep well, your heart rate and blood pressure go up and you are at greater risk for cardiovascular disease,” Zee says.

Despite myriad studies, there is still no consensus on why sleep is needed for survival. But the consequences of not getting enough of this fundamental physical process bear out the importance of continuing to seek answers.