Is Sleepwalking Acting Out a Dream?

01What is actually happening during an episode

The old way of describing sleepwalking was to say the person is unconscious and running on autopilot. The better description, and the one the recordings support, is that they are in two states at once.

Sleep is not one switch for the whole head. Different regions can be in different states at the same time, and a sleepwalking episode appears to be exactly that. The clearest look anyone has had came from a young man being monitored with electrodes placed inside the brain while doctors investigated his epilepsy. During that recording he had a confusional arousal, and the team caught it. Terzaghi and colleagues reported in the journal Sleep that the motor and cingulate cortices showed a local arousal, awake in the middle of the night, while the frontoparietal associative cortices showed increased delta activity, the slow waves of deep sleep. Both patterns were there before the episode began and stayed there all the way through it.

Read that plainly and the whole thing makes sense. The parts that move a body were up. The parts that reason, plan, evaluate, and file a memory were still asleep. Nobody had to choose between awake and asleep, because both were true in different places at once.

That is one person, in one recording, with epilepsy, and it should be held as a vivid demonstration rather than a population finding. But it lines up with the wider picture. A scoping review of dreams and dreamlike mentation in non-REM parasomnias by Longe, Omodan, Leschziner, and Rosenzweig in the Croatian Medical Journal, covering 16 studies and 195 adults, describes sleepwalkers as "dream-walking" because of local arousals and a wake-like activation in motor and limbic regions, alongside a preserved and sometimes increased inhibition of the frontoparietal network. Motor system on. Emotional system on. The executive on top of it still off.

This also answers the question that bothers people most about what they watched. How can someone navigate a dark staircase, turn a door handle, and step around a chair while asleep? Because the systems doing that work are not the ones that are asleep. The route was never being thought through. It was being executed by the part of the brain that executes routes.

02Why the eyes are open and nobody is home

The open eyes are what make an episode so unnerving, and they are also the single most misread detail. Watching somebody look straight at you is such a strong social signal that the mind fills in a person behind it. There is no person behind it in the sense you mean.

In the sleep laboratory, Oudiette and colleagues at the Pitie-Salpetriere hospital in Paris recorded a woman with sleep terrors opening her eyes, turning her head, and looking around with a scared expression while the recording still showed her in quiet deep sleep. She screamed a moment later. The eyes were open and the brain was not awake. Those are two different measurements, and only one of them is visible from a doorway.

So, the blank look is not a person ignoring you or a person in a trance. It is a face without the machinery behind it that would normally answer a question, recognize a voice as belonging to somebody in particular, or notice that anything is unusual. It is genuinely eerie. It is also, in itself, ordinary for this kind of episode.

03What was actually going on in their mind

Here is where most pages stop too early, and it is the reason you are reading this one. The standard line is that sleepwalking happens outside dreaming, so there is nothing to report. That was the consensus for decades. Then somebody asked.

Oudiette's team studied 43 adults referred to their sleep unit for severe sleepwalking and sleep terrors, people whose episodes were dangerous, frequent, or disturbing enough to send them to a specialist. Their average age was 26 and they had been having episodes for about 20 years. All of them had an overnight video sleep study, and 38 of the 43 could be interviewed properly about what had been going through their minds at the moment of an episode. Twenty-seven of those 38, which is 71 percent, could describe at least one dreamlike scene tied to an episode somewhere in their history.

So, there is something in there. What it is, though, is not what the phrase "acting out a dream" makes you picture.

Across the 106 scenes they collected, 95 percent were a single visual image, not a story with a beginning and an end. One scene, seen once, and gone. Only one report in the entire collection had the sprawling, image-filled, odd quality of a full dream, and it came from the youngest patient in the group, an 11-year-old boy. The rest were flashes.

They were also, overwhelmingly, unpleasant. Misfortune appeared in 54 percent of them, meaning something bad happening through chance or circumstance rather than through anybody's intent. Apprehension, fear, or terror was the emotion in 84 percent. Aggression turned up in 24 percent, and in almost every one of those the sleeper was the victim rather than the aggressor. There was no sexual content at all, and no scenes of good fortune.

The examples the researchers published are worth reading, because they explain the behavior in a way no summary does. A woman saw a truck about to run her over, and leapt out of bed and off a mezzanine. A woman saw her baby in danger, and grabbed her baby and ran out of the room with her. A woman felt herself locked in a box and suffocating, and pushed at the walls around her. A man was hurtling down an icy slope toward a ravine, and shouted and leapt out of bed. A woman saw the ceiling collapsing, and jumped up and pushed at the wall.

Every one of those is a single frightening image plus the most obvious physical response to it. That is not a dream being performed. It is something much smaller and much more immediate, and the body doing the only sensible thing.

Now the limits, because they matter as much as the numbers and this is a small literature. These were people referred for severe episodes, so they are not a cross section of everybody who has ever sleepwalked. The reports were gathered retrospectively, covering a lifetime, so memory has had years to work on them. The authors themselves write that dreamlike recall was occasionally associated with episodes and that the exact frequency of the association is still unknown. And in the sleep unit, on the night of the study, not one of the 43 patients remembered any mental content in the morning, because the episodes they had in the lab were minor compared with the ones they had at home.

The scoping review makes the same point about the field as a whole. Of its 16 studies, only three collected dream content immediately after an episode caught on a sleep study. Everything else was a person recalling, later, something that happened while most of their brain was in deep sleep. That is thin evidence by design, not by neglect, and it is why the honest answer to your question has a shape rather than a single word.

04Why it looks so exactly like enactment from the doorway

There is one finding that explains the illusion better than anything else, and it comes from a study that put the two conditions side by side.

Uguccioni and colleagues compared 32 people with sleepwalking or sleep terrors against 24 people with REM sleep behavior disorder, collecting the mental content associated with their behaviors over their lifetimes and again on the morning after a video sleep study. Ninety-one percent of the sleepwalking group and 87.5 percent of the REM group recalled an enacted dream, so on that crude measure the two look the same. The difference is in what the scenes contained.

Forty-two percent of the sleepwalkers' scenes were set in their own bedroom. In the REM group that was exceptional.

Sit with that for a second, because it is the whole answer to why an episode looks like a person acting something out. The scene the sleepwalker is reacting to is largely the room they are actually standing in, with a threat added. The ceiling that is collapsing is their ceiling. The box they are trapped in is their bed. So, of course the movement fits the furniture. They are not navigating a dream world and colliding with a real one. They are moving through the real room, which is also, at that moment, most of the scene.

The rest of the comparison separates the two conditions cleanly. Threat was in 70 percent of the sleepwalking scenes and 60 percent of the REM ones, so both are frightening. But the sleepwalkers' threats were misfortunes and disasters, while the REM group's were human and animal aggression. And the responses were opposite. Sleepwalkers mostly fled from a disaster, with a quarter fighting back when they were attacked. Three quarters of the REM group counterattacked when they were assaulted. The REM group's dreams were also longer, more complex, and less bizarre.

Our companion guide to acting out dreams while asleep takes the REM column of that table and goes properly into it, including how it is confirmed and what it can signal. This page is the other column.

05Why they will not remember it in the morning

You will probably have tried, at some point, to ask them about it. And you will have got a blank, or an argument, or a joke. The gap between how much of it you are carrying and how little they are carrying is one of the harder parts of living with somebody who does this.

The reason is in the recording described earlier. Laying down a memory you can retrieve later is work done by the parts of the brain that were still in slow-wave sleep. The motor regions were awake enough to walk. The regions that would have written the night down were not awake at all. So, there is often nothing to remember, in the sense that nothing was ever stored. This is not denial, not embarrassment, and not a person covering for themselves.

What does sometimes survive is a fragment, and that is what the studies collected. A single image, an emotion, a sense of having needed to get out. Four of Oudiette's patients had the feeling of having dreamt with no image at all, only a pure fear. If the person you watched offers you something that small and odd the next day, that is what a genuine recollection of this looks like. It will not be a story.

There is a practical version of this. Pressing somebody for details of an episode they cannot access does not recover anything, and it can leave them feeling accused of something they were not present for. It is kinder, and more useful, to tell them plainly what you saw and let them keep whatever fragment they have.

06The other direction, and it matters more clinically

Everything above pushes toward reassurance, and it should. But there is a mistake that runs the opposite way, and because it is the less common one it almost never gets said, so here it is once.

If somebody in your household is regularly moving, shouting, punching, or grabbing in their sleep in the last hours before morning, and they wake up alert and can tell you a dream that matches exactly what their body was doing, that is a different event, and calling it "just sleepwalking" is the error worth avoiding. That pattern is the one that earns an assessment, and it is the only parasomnia in this family with a documented association with later neurological conditions. Waving it away is more costly than worrying about the wrong thing.

Oudiette's team say as much from inside their own data. The overlap in dream content, they write, means the distinction between sleepwalking and REM sleep behavior disorder can be difficult when it rests on an interview alone, and a video sleep study is what helps in doubtful cases. They also list the features that push a clinician one way rather than the other: age, since disorders of arousal skew young and the REM condition skews middle-aged or older, and walking, which is common in sleepwalking and exceptional in the REM condition.

None of which you are meant to sort out yourself from a hallway. The point is simply that both errors exist. Reading dream enactment into an ordinary childhood pattern frightens people for nothing. Reading an ordinary childhood pattern into genuine dream enactment loses time that is actually worth having.

07How common any of this is

Numbers help at one in the morning, so here are the good ones. Stallman and Kohler pooled 51 studies covering 100,490 people for PLOS ONE and found an estimated lifetime prevalence of sleepwalking of 6.9 percent, with a confidence interval running from 4.6 to 10.3 percent. Within the previous twelve months, the rate was 5.0 percent in children and 1.5 percent in adults.

Roughly one person in fifteen has sleepwalked at some point. One child in twenty has done it in the past year, and only about one adult in seventy. The steep drop between those two figures is the shape of the reassurance: this is overwhelmingly something that happens for a while and then stops.

It is also worth knowing that these episodes are hard to catch even under laboratory conditions. The scoping review notes that complex behaviors show up in only about 30 to 35 percent of sleep studies done on people who genuinely have a non-REM parasomnia. If a clinician ever orders a study and nothing happens on the night, that is not a verdict. It is how often these things cooperate.

08What is worth noticing, which is not the same as a checklist

There is no list of warning signs on this page, because a list at three in the morning turns an anxious person into their own diagnostician and that has never once helped anybody. What is genuinely useful is much smaller: a few plain observations that make a future conversation with a clinician better, if you ever have one.

Note roughly when it happened, in relation to when they went to sleep. Not the exact minute. Early in the night or close to morning is the distinction that carries information. Note whether they said anything, and whether they could be steered back to bed without much resistance. Note, in the morning, whether they remember anything at all, and if so what shape it took, an image or a story. Note whether anything unusual preceded it: a short night, a fever, a stressful week, an unfamiliar bed.

Those last ones show up in the research as the things patients themselves point to. In Oudiette's group, the precipitating factors patients named were stress or an argument in 58 percent, sleep deprivation in 28 percent, something specific from the day before such as a horror film, an action film, or a thriller in 16 percent, alcohol in 14 percent, and sleeping at somebody else's house in 5 percent. Sixteen percent could name nothing at all. Those are patient attributions rather than proven causes, and the honest reading is that they are worth noticing rather than worth acting on.

The other thing worth doing is the least interesting thing on this page, and it is the one that actually reduces harm: look at the route. Stairs, the front door, a window, anything sharp or hot left out overnight. A person in this state can navigate a familiar room competently and cannot evaluate whether the thing they are heading toward is a good idea. Handling the route in daylight is worth more than anything you can do while it is happening.

09Nightmares, night terrors, and sleepwalking, in one paragraph

Since these three get bundled together constantly, here is the quick set of boundaries. A nightmare is a REM event, usually in the second half of the night, and the person wakes frightened and can tell you the dream in detail. A night terror is a non-REM event from the same family as sleepwalking, usually early, and the person is unreachable during it and untroubled by it in the morning. Our guide to what nightmares mean works through that comparison properly, so there is no point repeating it here.

And the difference that this page exists for, said once and plainly: sleepwalking is an incomplete arousal out of deep non-REM sleep, and REM sleep behavior disorder is a REM event in which the muscle shutoff that normally keeps the body still has failed. They look almost identical from a doorway and they are opposite events under the surface.

10When it is worth speaking to a clinician

Not because of one episode, and not because of anything on this page. The conversations worth having are the practical ones.

If somebody has been hurt or nearly hurt, if the episodes are frequent enough to be wearing on the household, if a person is leaving the house, if this has newly started in adulthood, if it arrived alongside a new medicine or a change in dose, or if it comes with loud snoring, gasping, witnessed pauses in breathing, or heavy daytime sleepiness, those are all worth raising. Sleepiness in the daytime is a real one: nearly half of Oudiette's patients scored in the abnormal range on a standard sleepiness scale, which suggests the nights are costing something even when the person insists they slept fine.

An appointment for this is mostly a history. What time, how often, what happened, who saw it, what changed recently, what medicines are being taken. A sleep study comes into it when the picture is atypical or when the behavior carries real risk of injury. Never stop or adjust a prescription to test a theory about a night. Bring the dates to whoever prescribed it.

If somebody is being injured tonight, or is at immediate risk, that is urgent rather than something to schedule.

11Common questions about sleepwalking and dreams

Do sleepwalkers dream?

Not in the way the word usually means, and not nothing either. Sleepwalking arises out of deep non-REM sleep, which is the least dreamlike part of the night, and it is not the REM stage where long story-like dreams live. But when researchers interviewed 38 adults with severe episodes, 71 percent could recall at least one dreamlike scene tied to an episode at some point in their lives, and 95 percent of those scenes were a single visual image rather than a narrative. So, brief mental content does occur.

Is sleepwalking the same as acting out a dream?

They are different events that look alike from across the room. Dream enactment is a REM phenomenon, usually in the last third of the night, where the body performs what a full dream is doing and the person wakes able to describe that dream. Sleepwalking is a partial arousal from deep sleep early in the night, where the mental content, when there is any, is a flash rather than a plot, and the person typically remembers nothing.

Why do their eyes look open if they are asleep?

Because sleep is not a single switch for the whole brain. An intracerebral recording of a confusional arousal showed the motor and cingulate cortices in a local waking state while the associative cortices stayed in the slow waves of deep sleep. Eyes can be open, the body can move and navigate, and the parts that recognize people and evaluate a situation can still be asleep.

Why does it always happen an hour or two after they fall asleep?

Because the deepest non-REM sleep is concentrated at the front of the night, and these episodes are incomplete arousals out of it. In one study of 43 adults with severe episodes, events fell within the first two hours of sleep in 88 percent of patients. That timing is the single most recognizable feature of sleepwalking and one of the more reassuring ones.

Does what they did during the episode mean anything?

Not as a message. In the studies, the content clusters around misfortune, apprehension, and being threatened rather than around anything personal, and it is usually set in the sleeper's own bedroom. A person who ran was reacting to an image of something falling or chasing, not expressing a wish or revealing a secret. Reflecting on a remembered fragment is a reasonable thing to do; treating the behavior as a coded statement about the person is not supported by anything.

They cannot remember it. Should I keep asking?

There is usually nothing there to recover, so pressing rarely helps. The brain regions that store retrievable memories were still in deep sleep during the episode. Tell them what you saw, once, without interrogation, and let them keep whatever fragment they have. In the research, four patients had only a feeling of having dreamt with no image at all.

Is sleepwalking a sign of a brain disease?

Ordinary sleepwalking is not, and it is common: pooled across 51 studies, about 5 percent of children and 1.5 percent of adults had an episode in the previous year. The association with later neurological conditions belongs to REM sleep behavior disorder, which is a different, REM-based condition confirmed by an overnight video sleep study, and which our page on acting out dreams while asleep covers in full.

How would anyone tell the two apart properly?

By history first, and by a video sleep study when the history is not clear. Clinicians weigh the time of night, whether the person walks, whether they wake alert with a matching dream or confused with nothing, and the person's age. The researchers who collected the mental content of both conditions say plainly that an interview alone can make this difficult and that a video sleep study helps in doubtful cases. It is not something to settle from a doorway.

If you came here after watching somebody walk past you with the lights off behind their eyes, the honest summary is that what you saw was almost certainly a partial arousal out of deep sleep rather than a dream being performed, that a brief frightening image may well have been in there and it was a flash rather than a story, that the memory of it is yours alone because theirs was never written down, and that the most useful thing you can do is not tonight's vigil but tomorrow's look at the stairs and the front door. DreamTold treats dreams as a window into your own mind rather than a code to be cracked, so if a fragment does survive the night and stays with you, look it up in the dream dictionary and think about it there. This guide is educational and meant for reflection, not medical advice, and it cannot diagnose anything. If somebody is being hurt, if episodes are frequent, if this started newly in adulthood, or if it arrived alongside a change in medication, speak with a qualified clinician.

12Sources

Every claim on this page that comes from somewhere else, with the somewhere else.

• PLOS ONE: Stallman and Kohler pooled 51 studies covering 100,490 people for PLOS ONE
• PubMed: Uguccioni and colleagues compared 32 people with sleepwalking or sleep terrors against 24 people with REM sleep behavior disorder
• PubMed Central: a scoping review of dreams and dreamlike mentation in non-REM parasomnias by Longe, Omodan, Leschziner, and Rosenzweig
• PubMed Central: Oudiette and colleagues at the Pitie-Salpetriere hospital in Paris
• SLEEP, Oxford Academic: Terzaghi and colleagues reported in the journal Sleep