Sexsomnia: What Psychiatrists Need to Know

June 16, 2026by Chris Aiken, MD1

Sexual behavior during sleep can end up in court

STUDY: Shapiro CM et al, Can J Psychiatry 2026

STUDY TYPE: Review

FUNDING: Independent

Background

Sexsomnia is a subtype of non-rapid eye movement (NREM) parasomnia, in the same family as sleepwalking and sleep eating. The affected person performs sexual behaviors during sleep, with no memory of the episode afterward.

This review is from the researchers who coined the term in 2003.

Summary

Common: lifetime prevalence of 7–10%.
Legal risks: 41% of forensic sleep referrals involved sexual assault.
Common triggers: physical contact with a bed partner (64%), stress (52%), and fatigue (41%).
Causative meds and substances: Mainly GABA-ergic, especially alcohol (15% of cases). Also z-hypnotics like zolpidem, sodium oxybate (which is a date-rape drug), gabapentin, pregabalin, and trazodone.
Other meds that cause parasomnia may be involved, but no linkage shown yet (eg, antipsychotics, serotonergic antidepressants, lithium, and propranolol).

Sexsomnia differs from sleep-related epilepsy (brief, stereotyped seizure movements) and REM behavior disorder (vivid dream-enactment, older adults, late-night timing).

Rule out malingering, dissociation, active mania or psychosis, and paraphilic disorders. In all of these, conscious awareness is preserved.

Practice Implications

If a patient or their partner reports unusual sexual behaviors at night, ask about amnesia for the episodes, other parasomnias, and the triggers listed above.
When there’s any legal context, consult a medico-legal expert and get a bed partner interview.

—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report

What’s Your Take? Share in Comments

Geoffrey Guenther

June 21, 2026 at 11:45 pm

This seems like a special form of what might be called “horizontal sleepwalking.” The subject is still asleep, but animated by the brain’s autonomous emergency response network. There are various names for the latter. Daniel Kahneman called this “System 1.” There are also Lacan’s “reactive unconscious,” Richard O’Connor’s “automatic self,” Timothy Wilson’s “adaptive unconscious,” There are several other appellations, but I prefer “Guardienne,” as brief and most descriptive of the network’s primary function.
The network is probably not only faster than the frontal cortex, but also autonomous, for shortest reaction times. It is likely sentient, or semi-sentient, logic-enabled. A conscience would impede it, so it probably has little or none. (Its lack of conscience would be problematic, but I strongly suspect that is what makes it the brain’s primary creative engine.)
If the Guardienne concept is valid, there are many consequences, which are explored as listed below:
⓵ https://www.researchgate.net/publication/331047505
The_Guardienne_Autonomous_Semi-sentient_Protective_Region_of_the_Human_Brain
⓶ https://www.researchgate.net/publication/350359019
The_Guardienne_Hypothesis_and_Deja_Vu
⓷ https://www.researchgate.net/publication/346137535
The_Guardienne_and_Bipolar_Disorder
⓸ https://www.researchgate.net/publication/353306319
Cerebral_Dualism_and_the_Guardienne’s_Roles_in_Alcoholism_Addiction_and_Relapse
⓹ https://www.researchgate.net/publication/357163812
The_Guardienne_and_Creativity
⓺ https://www.researchgate.net/publication/358151200
Phineas_Gage_and_the_Guardienne_-a_case_history
⓻ https://www.researchgate.net/publication/383648700
THE_GUARDIENNE_AND_DEPERSONALIZATIONDEREALIZATION_1
Some background findings:
Hans Stuyck, A. Cleeremans, E. Van den Bussche, The Aha! experience is not constrained by cognitive load. DOI: 10.1016/j.cognition.2021.104946

One comment

Geoffrey Guenther
June 21, 2026 at 11:45 pm

This seems like a special form of what might be called “horizontal sleepwalking.” The subject is still asleep, but animated by the brain’s autonomous emergency response network. There are various names for the latter. Daniel Kahneman called this “System 1.” There are also Lacan’s “reactive unconscious,” Richard O’Connor’s “automatic self,” Timothy Wilson’s “adaptive unconscious,” There are several other appellations, but I prefer “Guardienne,” as brief and most descriptive of the network’s primary function.
The network is probably not only faster than the frontal cortex, but also autonomous, for shortest reaction times. It is likely sentient, or semi-sentient, logic-enabled. A conscience would impede it, so it probably has little or none. (Its lack of conscience would be problematic, but I strongly suspect that is what makes it the brain’s primary creative engine.)
If the Guardienne concept is valid, there are many consequences, which are explored as listed below:
⓵ https://www.researchgate.net/publication/331047505
The_Guardienne_Autonomous_Semi-sentient_Protective_Region_of_the_Human_Brain
⓶ https://www.researchgate.net/publication/350359019
The_Guardienne_Hypothesis_and_Deja_Vu
⓷ https://www.researchgate.net/publication/346137535
The_Guardienne_and_Bipolar_Disorder
⓸ https://www.researchgate.net/publication/353306319
Cerebral_Dualism_and_the_Guardienne’s_Roles_in_Alcoholism_Addiction_and_Relapse
⓹ https://www.researchgate.net/publication/357163812
The_Guardienne_and_Creativity
⓺ https://www.researchgate.net/publication/358151200
Phineas_Gage_and_the_Guardienne_-a_case_history
⓻ https://www.researchgate.net/publication/383648700
THE_GUARDIENNE_AND_DEPERSONALIZATIONDEREALIZATION_1
Some background findings:
Hans Stuyck, A. Cleeremans, E. Van den Bussche, The Aha! experience is not constrained by cognitive load. DOI: 10.1016/j.cognition.2021.104946