REM and NREM Dream Content

REM and NREM dream content differences may hold clues to function.

Posted Jun 19, 2013

A recent paper in Sleep Medicine (Uguccioni, G., Golmard, J.-L., de Fontréaux, A.N., Leu-Semenescu, S., Brion, A., Arnulf, I.Fight or flight? Dream content during sleepwalking/sleep terrors vs rapid eye movement sleep behavior disorder. Sleep Medicine, volume 14, issue 5, year 2013, pp. 391 – 398) reports some hard-to-get, unique data on enacted dreams, or dreams that patients appear to enact while asleep. The patient may start to vocalize while asleep as if talking to someone and then thrash around as if fighting some opponent etc The patient may actually get out of bed and run around the room as if trying to escape from a threat and so on. These types of dreams typically occur as part of the symptom complex associated with REM Behavior Disorder (RBD) or with sleepwalking or with sleep terrors.

Enacted dreams are uniquely important for a science of dreams as they give us a very unusual glimpse into dream content. Here we can see dreams being enacted as if on a stage while the patient is asleep and dreaming. We do not have to rely only on the patient’s verbal report of a dream after he awakes.

In this study of enacted dreams the investigators were interested in, among other things, the question of whether the content of REM dreams differed from the content of dreams that occurred in other stages of sleep—non-rem dreams. The investigators could get at REM dreams via analysis of enacted dreams reported by the RBD patients and at NREM dreams via analysis of the enacted dreams reported by the sleepwalkers/sleep terror patients who typically enact their dreams in NREM stages of sleep.

Thirty-two patients with sleepwalking/sleep terrors (SW/ST) and 24 patients with REM Behavior Disorder (RBD) were consecutively recruited and interviewed concerning their enacted dreams. These were dreams experienced during a night of sleep videosomnography as well as recent and lifetime dreams involving parasomniac events and enacted dreams. A total of 121 dreams were analyzed in both groups (sleepwalking group, n=74 dreams; RBD group, n=47 dreams). Dream content was then scored using primarily the standardized Hall vs de Castle criteria and Revonsuo’s threats scale.

There was a significantly higher level of aggression in the RBD than in SW/ST subjects but these aggression levels were not correlated with daytime waking behaviors. Sleepwalkers had more frequent misfortunes in their enacted dreams than patients with RBD (28% vs. 8%, p=0.01) and tended to exhibit less frequent aggression (17% vs. 33%, p=0.06) during the enacted dreams. In dreams with aggression both RBD and sleepwalkers/sleep terrors (SW/ST) patients were more often victims rather than aggressors. In dreams with misfortunes the sleepwalkers mostly fled from a threat/disaster while patients with RBD counter-attacked when assaulted. The settings for SW/ST enacted dreams tended to be familiar/domestic while the settings in RBD dreams were less familiar.

In summary, aggression (such as counter-attacks) is rare (8%) in SW/ST enacted dreams, but common (reported by at least 33% of dreamers) in RBD. Given that SW/ST sleep parasomnias are associated with a dissociated state between N3 sleep and arousal, and RBD is associated with REM sleep, it is reasonable to suggest that the enacted dreams reflect brain state differences in these two sleep states. As the authors point out in their discussion, content analyses of thousands of dreams from healthy participants reveal strikingly similar content differences for REM versus N2/N3 sleep dreams. It is a remarkable and now fairly well-established fact that REM dreams appear to specialize in simulation of aggressive interactions while N2/N3 dreams specialize in simulation of non-aggressive and friendly interactions.

The authors conclude that “These striking (dream content) differences suggest that cognitive

processes during non-REM and REM sleep parasomnias simulate different types (natural

disaster vs. human/animal aggression) of threats. Taken together, these threats may be the

exacerbation of systems that train humans to appropriately react during the daytime to a wide

spectrum of dangers.”

The authors suggest that their data is at least partially consistent with Revonsuo’s threat simulation theory of dream content because both SW/ST and RBD dreams appeared to have simulated threats to the dreamer with the RBD patients reacting with aggression and the SW/ST reacting with flight or awakenings. But note that the threat simulation hypothesis cannot explain the key question as to why the REM dreamers responded with aggression and the SW/ST with flight. In REM sleep there is some diminution of prefrontal inhibition on limbic brain sites. While this physiologic fact can account for the mechanics of the aggressive responses in REM dreams, it does not explain the question as to why such responses should occur while we are asleep in the first place. Why are limbic circuits dis-inhibited in REM in the first place? The dream content data suggests that the reason is that Mother Nature wants the organism to simulate aggression-related behaviors while asleep. But again, why only for REM dreams? If activating aggression circuits while asleep is adaptive, why not do so in NREM states as well?