Glymphatic System: How the Brain May Clean Up During Sleep and What Remains Unclear

This research has given a cleaning pathway in the brain a name: the glymphatic system. It is one of the most exciting discoveries in brain research in recent years. And one of the most controversial.

In this article, I will take you through both. Through the data that point to a nightly cleanup. And through the 2024 study that calls exactly this into question.

What is the glymphatic system?

In short: The glymphatic system is a cleaning pathway of the brain. Brain fluid flows along the arteries into the tissue, mixes with the fluid between the nerve cells and drains away along the veins, carrying metabolic waste with it. Water channels made of aquaporin-4 on the astrocytes appear to facilitate this exchange.

Perhaps you have never wondered how your brain gets rid of its waste. Why would you? It works, after all, most of the time anyway.

In the rest of the body, this is the job of the lymphatic system. A fine network of vessels collects fluid and residues from the tissue and carries them to the lymph nodes. The brain tissue itself lacks such classic lymphatic vessels.

At the same time, your brain works without a break. Nerve cells fire, use energy and leave residues behind. These include protein fragments such as beta-amyloid and the protein tau. Both play a role in Alzheimer's disease.

For a long time, the question was: where does all this waste go?

Imagine the brain as a city whose main roads are the blood vessels. Alongside each road runs a narrow canal. Through this canal, brain fluid, known as cerebrospinal fluid, flows into the city. It rinses through the alleys between the houses and leaves the city through canals alongside the outgoing roads, the veins.

The name glymphatic is a coined word. It combines glia, the brain's helper cells, with lymphatic. The star-shaped astrocytes belong to these glial cells. For a long time, they were considered mere filler between the nerve cells. Today we see them more as caretakers that help regulate water, nutrients and signals.

In 2015, another piece of the puzzle fell into place. A team at the University of Virginia led by Louveau and Kipnis discovered lymphatic vessels in the meninges of mice, along the large venous sinuses. These vessels carried fluid and immune cells from the brain fluid to the deep lymph nodes in the neck. With that, the outflow had an address. You can read how the lymphatic system works in the rest of the body in the article Stimulating the lymphatic system.

Through the lens of clinical psychoneuroimmunology, KPNI for short, this is fascinating. Here, two systems touch that we long thought of as separate: cleaning and immune defense. Astrocytes can respond to disturbances with a state of alarm, which specialists call reactive gliosis. And the same lymphatic routes that drain fluid are also routes for immune cells. That makes the glymphatic system more than a sewage pipe. It is an interface between the nervous system and the immune system.

One distinction matters to me. The glymphatic system has mainly been described for metabolic waste that the brain itself produces. What role it plays for environmental substances such as heavy metals has been little studied so far. Different questions arise there, which I describe in the article Heavy metals in the brain.

And now you know why researchers talk so much about water when it comes to brain cleaning.

Why sleep could be so important here

Do you know those nights after which your head is as clear as a freshly aired room? And others after which it feels sluggish, even though you lay in bed for a long time?

In 2013, the Rochester team put forward an idea that has been traveling around the world ever since.

Imagine an old town with narrow alleys. During the day, people and delivery vans crowd them, and the street sweeper can barely get through. At night the alleys empty, and the cleaning crew has a clear run. That is roughly how this study describes the sleeping brain.

But what drives the fluid? A possible answer came from Copenhagen in 2025.

In this picture, the blood vessels work like a slow bellows. Each contraction and widening pushes a little more brain fluid into the tissue. Norepinephrine sets the beat.

You may know norepinephrine, also called noradrenaline, as a stress messenger. In the brain, it comes mainly from a small arousal center in the brainstem, the locus coeruleus. During the day, it keeps you awake and alert. During sleep, it seems to come and go in waves, and these waves could drive the pump.

What happens if your stress system does not power down at night? In humans, this has not been studied directly in this context. It is biologically plausible that a persistently high alarm level could disrupt this rhythm. That is a hypothesis, not a measurement.

This fits with an observation about the internal clock, and with it the hormonal system. In 2020, the Rochester team showed that inflow and clearance in mice follow their own daily rhythm. The peak came in the middle of the animals' rest phase. Precisely then, the aquaporin-4 channels were most neatly lined up along the vessels, and without aquaporin-4 the difference between day and night disappeared.

So perhaps what counts is not only how long you sleep, but also whether your sleep fits your internal clock. You can read how light sets this clock in Blue light and the circadian rhythm.

And with age? In old mice, the clearance of beta-amyloid was 40 percent worse than in young ones. The walls of the small arteries in the cerebral cortex pulsed 27 percent more weakly, and the water channels lost their orderly position. Whether the same applies in humans is open. But it would be one possible building block in explaining why the brain becomes more vulnerable over the years.

All three figures come from animal studies. They show mechanisms and cannot be transferred directly to humans.

A second perspective comes from a very different direction. In anthroposophic medicine, the alternation of waking and sleeping has long been described as a rhythm of breaking down and building up. That is a tradition of thought, not a study. Still, I find it remarkable that modern brain research is asking a similar question: what does the body clear away at night so that the head can be clear again during the day?

And now you know why a night spent working through often feels not only tiring, but also cluttered.

What has been measured in humans so far

In short: In humans, there is no direct view into sleeping brain tissue. There are, however, studies using contrast agent in the spinal fluid, PET images and measurements in cerebrospinal fluid and blood. Most of them suggest that sleep loss could slow the clearance of metabolic waste.

Mice are not humans. That is probably your first thought, too. It is a fair one.

Their brains are tiny, their sleep is organized differently, and many experiments were carried out under anesthesia. That is why the most exciting question is: what do we see in humans?

There is also a detail from the same study. People who reported sleeping less in everyday life already had more amyloid in several regions at the rested baseline measurement. That is an association, not proof of a cause.

The second Alzheimer's protein, tau, also responds to sleep. A team at Washington University in St. Louis led by Holth and Holtzman found around 90 percent more tau in the tissue fluid of mice while awake than during sleep. In humans, tau in the spinal fluid rose by more than 50 percent during sleep deprivation.

A second study from the same university showed something subtle. When only deep sleep was deliberately disrupted in 17 healthy people, without shortening sleep duration, amyloid in the spinal fluid rose the next morning. The team suspects that this is mainly due to altered nerve cell activity. So more production, rather than necessarily less clearance.

This matters for interpretation. More waste in the morning can have two causes: more was produced, or less was cleared away. Both probably play a role.

And now you know why researchers look so closely at spinal fluid after a sleepless night.

The 2024 counter-study: does the brain perhaps clean up less during sleep?

Perhaps in 2024 you stumbled across a headline that turned everything upside down. Sleep, it said, does not clean the brain at all, if anything quite the opposite. If you wondered back then what you were supposed to believe now: welcome to science as it works in practice.

The London group was not the first to have doubts. As early as 2018, a French team from Caen had observed with MRI and fluorescence that brain fluid flow in the tissue was active mainly in the waking state. Under anesthesia it was slowed, and the researchers found no expansion of the spaces between cells.

This points to a basic problem in the field. Many early studies used anesthesia as a substitute for sleep. But anesthesia is not sleep. And as early as 2018, five research groups had described in a joint analysis that anesthesia, the age of the animals and the way the dye was administered could explain contradictory results.

How the other side responded

In 2025, a team led by Plá and Nedergaard responded with a scientific commentary in the same journal. The title sounds politely pointed: A curious concept of CNS clearance. Among other things, the criticism concerns whether London measured genuine clearance from the brain or merely a redistribution within the tissue. There are also objections to the computational model, the anesthesia conditions and the injection technique.

I find one detail remarkable here. The 2018 joint analysis had also described that injections directly into brain tissue can themselves dampen glymphatic flow. The London dye was given in exactly this way, directly into the tissue. How strongly this affects the measurement is part of the dispute. The London group has responded publicly to the criticism. The debate has not been settled.

Controversies like this are nothing new in the field. In 2017, a team at the University of California, San Francisco, led by Smith and Verkman found that substances in the brain tissue of mice and rats travel mainly by diffusion, meaning slow spreading without flow. In their experiments, aquaporin-4 was not needed for this. A year later, by contrast, the joint analysis of the five research groups found less inflow without aquaporin-4. So the dispute over the water gates has already made the field more precise once before.

And what does imaging in humans measure?

A second point of contention concerns the measurement methods in humans. Injecting contrast agent into the spinal fluid is elaborate and not routine. That is why many studies use a more convenient MRI marker.

And now you know why I take the counter-study seriously and still stand up for good nights.

Sleep apnea: when the night is constantly interrupted

You snore loudly, your partner says. Sometimes it sounds as if you briefly stop breathing. In the morning your head is pounding, and during the day you fight against nodding off. And you think: but I spend eight hours in bed.

With obstructive sleep apnea, the throat collapses again and again during sleep. Breathing stalls, oxygen drops, and the brain briefly jolts awake. Often you notice nothing of this. But your sleep gets chopped up like a film with constant commercial breaks.

It is precisely the deep, calm phases, in which glymphatic flow appears strongest in animal studies, that are interrupted again and again. Whether this slows cleaning in humans is currently being studied intensively.

Individual studies paint a similar picture. In Busan, South Korea, the index was 1.30 in 24 people with sleep apnea and 1.62 in 24 healthy people. In Shenzhen, China, all groups with snoring or sleep apnea had lower values than healthy people. In 15 people with moderate sleep apnea, the markers changed after treatment, and results on cognitive tests improved.

Context remains important. These are observational studies using a marker whose meaningfulness is currently under critical review. We see a correlation, not proof. Regardless of the glymphatic system, though, suspected sleep apnea should be medically assessed anyway. You can read how that works in Recognizing sleep apnea.

And now you know why I also ask about snoring when someone comes to me with exhaustion and concentration problems.

Alcohol in the evening: a sleep aid with side effects

The glass of wine after a long day. Your body grows heavy, your head grows quieter, you fall asleep faster. It feels like a shortcut into the night.

This shortcut comes at a price. A review from the London Sleep Centre evaluated the known studies on alcohol and sleep in healthy people. At all doses, people fell asleep faster, and the first half of the night was calmer. The second half of the night, however, became more restless. REM sleep, the phase in which we dream most, began later and decreased overall at moderate and high amounts.

One detail is intriguing: in most studies, deep sleep in the first half of the night even increased. Whether this deep sleep under alcohol has the same cleaning quality has not been studied in humans.

A team at the University of Macau observed in mice that even moderate amounts slowed the inflow of brain fluid and impaired the clearance of beta-amyloid. With prolonged exposure, the water channels lost their orderly position, and the team describes the disturbance as lasting. Two mouse studies, two different results for small amounts. For large amounts, both point in the same direction.

Through the lens of toxicology, alcohol is a substance your body has to break down as a priority, mainly in the liver. I describe what happens in the process in Alcohol and the liver. You can read how even small amounts can affect the gut in Alcohol and the gut.

And now you know why you often lie awake early in the morning after an evening with wine.

Which sleep position is best for the brain?

In short: In a study on anesthetized rodents, glymphatic transport was most efficient when the animals lay on their side. In humans, there are only observational data on this. A recommendation to force a particular position during sleep cannot be derived from this.

Perhaps you have read that you absolutely must sleep on your side. And perhaps ever since, you have been lying awake at night wondering whether your position is the right one. Then this section is for you.

In humans, there are only observational data so far. In a case-control study, 45 people with neurodegenerative diseases slept on their backs for more than two hours per night considerably more often than 120 healthy comparison subjects. That is a correlation. Whether lying on your back contributes to the disease or the disease changes sleeping posture is something such a study cannot show. In addition, three of the researchers involved work for the company that manufactures the sleep monitoring device used.

My pragmatic advice: sleep in the position in which you sleep soundly through the night. If that is on your side, wonderful. If you notice that you snore more heavily on your back, that is a good reason to bring it up with a doctor.

And now you know why I do not prescribe a sleep position for you.

Three levers for your nights

Perhaps you are now wondering: and what do I do with all this? The good news: the most important levers are unspectacular. They need no device, no capsule and no detox tea.

How to protect your sleep in everyday life without driving yourself crazy is something I describe in detail in Putting sleep hygiene into practice.

More from the Detox Guide

Where this topic connects

Sleep, breathing and the brain are closely linked. These articles go deeper into the neighboring levels.

Frequently asked questions