Sleepwalking exposes an odd feature of sleep: movement and awareness do not always return together. An adult can leave bed with their eyes open whilst remaining poorly responsive and later remember little. The important questions are whether anyone could be hurt, whether episodes are new, and whether another sleep problem needs attention. Those questions matter more than how unusual the behaviour looks.
When waking happens only part of the way
Sleepwalking belongs to a group of conditions called non-rapid eye movement, or NREM, parasomnias: unwanted experiences or behaviours arising during sleep. Greta Mainieri and colleagues’ 2023 review in Diagnostics describes these as incomplete arousals from NREM sleep. Becoming able to move does not mean becoming fully awake. The familiar division between sleeping and waking is less tidy here than it appears from outside.
The same review questions how much diagnostic weight should be placed on complete amnesia or an absence of conscious experience. Remembering a fragment does not, by itself, rule out an arousal disorder. Equally, forgetting an episode does not establish one. These are details for a clinical history, rather than a home diagnostic test.
That distinction helps explain why arguing with someone about what happened is unlikely to settle the question. A witness can describe movement and responsiveness; the sleeper can describe whatever they recall. Both accounts can be useful without either being a complete recording of the event.
Why a new adult episode deserves attention
Mayo Clinic’s 2024 overview places typical episodes early in the night, often within the first two hours. A person may sit up, walk, dress, or talk with a blank expression and little normal response. Sleepwalking is more common in children, but adult episodes also occur.
Mayo advises assessment when episodes begin for the first time in adulthood, recur frequently, disturb the household, cause injury, or affect daytime functioning. New onset is a reason to ask what is happening, rather than assume a childhood habit has returned. A single account of walking around does not identify the underlying cause.
Frequency also needs context. A rare episode involving an open window may present a greater immediate safety problem than several harmless episodes of sitting up. When describing the problem, include what the person actually did and what hazards were nearby, alongside how often it happened.
Triggers are clues, not proof of a cause
The NHS sleepwalking guidance, reviewed in April 2025, lists insufficient sleep, stress, alcohol, fever, certain medicines, and other sleep conditions amongst possible links. Sleep apnoea and restless legs syndrome are examples. A busy week or disrupted night may therefore be relevant, but does not explain every episode.
A diary can record bedtime, sleep quality, alcohol or caffeine, stress, and witnessed events. Its purpose is to give a GP a clearer history. It cannot show that a particular evening caused the behaviour. Recording ordinary nights as well as difficult ones makes the account more useful than collecting only dramatic examples.
The NHS also recommends seeking help where sleepwalking accompanies marked daytime tiredness or breathing that stops and starts during sleep. Those symptoms deserve their own mention. They should not disappear into a general explanation that someone is stressed or sleeping badly.
Night-time movement does not identify the sleep stage
Walking, shouting, or an apparent response to a dream can have more than one explanation. Mayo Clinic’s diagnostic guidance describes checking for conditions that may be confused with sleepwalking, including nighttime seizures, other sleep disorders, and panic attacks. That is a reason for careful evaluation, rather than labelling every unusual movement as sleepwalking.
Cleveland Clinic’s clinical overview explains that a clinician may ask a bed partner about the episodes and, where appropriate, consider an overnight sleep study. A study is not necessary for every person. The decision depends on uncertainty, injury risk, and the possibility of another condition. Cleveland Clinic notes that recordings can sometimes help, because an event may not occur during a laboratory night.
There is no obligation to capture an episode at the expense of safety. A written description of the time, movement, responsiveness, and aftermath is a reasonable starting point for a conversation. If recording is discussed, agree privacy boundaries beforehand and keep the material for the assessment.
Safety comes before explaining every episode
Mayo Clinic’s diagnosis and treatment guidance prioritises reducing injury: clearing trip hazards, putting sharp or fragile objects out of reach, and considering a ground-floor bedroom where possible. Door alarms may be useful where leaving the home is a concern. Any changes to doors or windows should preserve a safe emergency escape route.
The same guidance advises gently leading a sleepwalker back to bed. Waking someone is not inherently dangerous, but confusion and agitation can follow. A calm response makes more sense than shouting or trying to demand an explanation during the episode. Where there is immediate danger, protecting people takes priority over avoiding an awakening.
For an appointment, Mayo suggests bringing a sleep diary and a list of medicines and supplements, including doses. A household member’s account can help. This makes the consultation about observable events and possible contributors, rather than whether the behaviour seems embarrassing or strange.
More sedation is not a simple answer
A 2018 systematic review in Sleep Medicine Reviews identified medicines associated with sleepwalking; the strongest evidence concerned zolpidem and sodium oxybate. This does not mean every episode in someone taking a medicine is caused by that medicine. It does mean that a medication history belongs in the assessment.
There is an important medicine-specific warning. NIH’s MedlinePlus guidance for zolpidem, eszopiclone, and zaleplon warns about serious, potentially life-threatening activities whilst not fully awake, including sleepwalking or driving. If such behaviour occurs whilst taking one of these three medicines, the guidance is to stop that medicine and contact the doctor immediately. This specific instruction should not be extended to every sedative or to sodium oxybate; discuss other suspected medicine effects promptly with the prescriber rather than making changes independently.
For people seeking a non-drug approach, a 2023 systematic review in Sleep Medicine examined behavioural and psychological treatments. It found some support for approaches including multicomponent cognitive behavioural therapy, but most publications were case reports or case series. The review could not establish a single reliable treatment for everyone. Treatment discussions need to acknowledge that uncertainty.
What this means in practice
- Consider a brief sleep diary this week, including ordinary nights, to make any appointment more informative.
- Review the route from bed to the doorway for loose cables, fragile objects, or other avoidable hazards.
- Agree a calm household response, with gentle guidance towards bed where this is safe.
- Arrange a GP discussion for first adult onset, injuries, disruption, or associated daytime or breathing symptoms.
- Bring the full medicine and supplement list. Follow the urgent, medicine-specific warning above for zolpidem, eszopiclone, or zaleplon; discuss changes to other prescribed treatment with a clinician.
These are options for documenting the problem and reducing risk, consistent with the clinical guidance linked above. They are not a programme guaranteed to stop episodes or a substitute for assessment.
It is also reasonable to ask the clinician what would change the assessment: a new symptom, an injury, or a different pattern of events. That makes the next step explicit without asking the household to decide on a diagnosis.
What we don’t know
A September 2026 systematic review in Sleep Medicine Reviews included 14 treatment studies, mostly retrospective and uncontrolled. Differences between studies prevented meta-analysis, and certainty was very low for most outcomes. Reports of improvement therefore need cautious interpretation: they do not establish how much benefit a particular adult can expect.
We still lack strong comparisons that reliably tell clinicians which treatment suits which patient. That uncertainty should be part of shared decisions, especially when medicines are being considered. A plausible explanation for partial waking is not the same thing as a proven treatment.
Sleepwalking deserves a practical, unembarrassed conversation: describe what happens, make the surroundings safer, and seek assessment when the pattern or its consequences warrant it.
Photo: Lisa Anna on Unsplash. Illustrative bedroom image.
