I experience episodes of what I assume is sleep paralysis (lasting ~ 10-30 seconds) a few times a year, where I'm conscious of being somewhat awake but unable to move. I can hear, but not see, and am quite focused on my breathing which feels labored and uncomfortable (can be accompanied by a feeling my face is covered by a sheet). Sometimes, I've even been able to "jolt" myself awake if somehow muscle control is regained. Otherwise, I assume I drift off back to sleep.

Do we know yet what has gone wrong (if it can be assumed to be aberrant) and which neural circuits are involved? What i'm looking for is a specific explanation at the neural circuit level of how it can occur.

Is anyone aware of anyone conducting research into this or how it could be studied as it doesn't occur frequently and likely difficult to induce on demand?


2 Answers 2


Body paralysis occurs during REM sleep to keep your body from acting out your dreams. You experience sleep paralysis when the mind awakens from REM sleep before the body paralysis has subsided. I've had this happen before also, along with the "jolting" out of it you describe. It usually happens when I'm very tired and trying to stay awake. I try to stay calm but it's not easy as you know. There is a good description at WebMD.

  • $\begingroup$ Thanks for sharing, @Kurt. Yes, i'm aware of the general mechanism of muscle atonia during REM and that sleep paralysis is most likely to happen as you exit the ultradian sleep cycle. What i'm curious about and after is a more specific explanation at the neural circuit level of how it can occur, and whether that is known/being researched (I can't find fMRI of an SP episode for example)? Is it a brain "snafu" or is a vestigial evolutionary-inherited trait that other animals have use for perhaps? $\endgroup$ Nov 4, 2017 at 1:53
  • $\begingroup$ @Louis Leung. Thanks for the clarification. I have had trouble finding much more indepth research also. I'll be interested to see what others come up with. $\endgroup$ Nov 4, 2017 at 9:57

Sleep paralysis is common, and associated with poor sleep

Isolated sleep paralysis is a relatively common parasomnia. Lifetime prevalence rates vary quite a bit from study to study, and between populations, but aggregating them gives a rate of about 7%. There are a number of associated conditions, many which are also associated with disturbed or irregular sleep in general (e.g., being an undergraduate student, shift worker, or having a psychiatric disorder).

Sleep paralysis also occurs in narcolepsy, but here it may be different

Sleep paralysis is a component of the overall clinical presentation of narcolepsy. It may have a similar but related mechanism, since these patients also experience cataplexy (muscle atonia in response to an emotional stimulus). Isolated sleep paralysis means sleep paralysis that isn't associated with narcolepsy. Some of our understanding of isolated sleep paralysis comes form studies of narcoleptic patients and animal models of narcolepsy. Because of the substantial differences in sleep patterns in narcolepsy, it's unclear if findings from these studies provide useful information for isolated sleep paralysis.

You can induce sleep paralysis by forcing sleep onset REM

You can induce isolated sleep paralysis in healthy volunteers by inducing sleep onset REM, a sleep period where REM sleep quickly follows sleep onset, instead of following a march through the phases of non-REM sleep. This was done in the linked study by, essentially, by waking the volunteers before the onset of the second REM period. Since sleep onset REM is more common in sleep deprived individuals or individuals with disturbed or irregular sleep patterns, this may be one of the reasons we see more sleep paralysis in these individuals.

Neural pathways involved in sleep are complicated, but REM doesn't directly inhibit arousal

The circuitry can be understood from studies of circuits involved in sleep in general and REM sleep in particular. You can read about this in Principles of Neural Science, Ch 47. It's quite complicated. One interesting note, though, is that the general cortical pattern for REM sleep, as measured, for example, by EEG, is more similar to waking that it is to non-REM sleep. Both states (waking and REM sleep) involve inhibition of a set of thalamic neurons that orchestrate the synchronous cortical firing seen in non-REM sleep.

Importantly, while the transition from waking to non-REM sleep involves a clear interaction between the neurons of the reticular formation that mediate arousal and non-REM sleep, there doesn't seem to be a direct pathway from REM-on neurons to inhibition of the reticular formation. It is not surprising that, during sleep onset REM periods, you could have dissociation of arousal (mediated by the reticular formation) and REM-on mediated hyperpolarization of motor neurons in the spinal cord. This review briefly discusses a possible mechanism of hyperactivity or hypersensitivity of REM-on neurons or weakness of the (still unknown) inhibition of REM.

Suppress REM and you reduce the frequency of sleep paralysis

Treatment focuses on reassurance and treatment of any underlying sleep irregularity and disturbances as well as any underlying psychiatric conditions, but when used, medications that suppress or inhibit REM in general reduce the frequency of sleep paralysis episodes.

It's probably not a big funding draw

There is, clearly, some research on isolated sleep paralysis outside of narcolepsy. It's an interesting phenomenon, but since it isn't dangerous or important clinically, I don't expect it gets much funding.

  • $\begingroup$ Thanks a lot for the info and particularly the link to ISP study in 1992, it was an interesting read. I agree there seems intuitively a strong association with the epiphenomenon linked to REM states, so it would make sense to start any study with the major REM modulators. Anectdotally, I've also noticed that my SP generally occurs in the early morning, which correlates to the time of more common bouts of REM sleep during the cycle. As you rightly say, as it's not a major life-ruiner, there won't be much ISP study but we can leave this open until the answer is eventually found! $\endgroup$ Aug 13, 2018 at 22:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .