Dreaming in REM and NREM Sleep

I wrote this during the second semester of my senior year at Johns Hopkins University for Sleep, Dreams and Altered States of Consciousness with Dr. Richard Allen.

Dreaming in REM and NREM Sleep

Carlos Macasaet

200.368 – Sleep, Dreams and Altered States of Consciousness
Dr. Allen
24 March 2005

The most fascinating part of human sleep, at least to writers and poets is dreaming. Sleep in humans is characterised by five distinct brain states, stages one through four, collectively referred to as non-REM (NREM) sleep and REM or rapid eye movement sleep. The term stage W is often used to refer to the wake state. REM sleep is characteristically different from NREM sleep in that the brain activation patterns more closely resemble brain activation in the waking state. In addition, humans experience muscle atonia during REM sleep, which is not present in NREM sleep. When researchers first studied this phenomenon, they noticed that when woken, subjects reported dreaming during REM sleep. This led to the theory that REM is responsible for dreaming. However, it had also been observed that dreams could occur during NREM sleep as well. Further research has shown that dreams occurring in REM sleep are quantitatively and qualitatively different from dreams occurring in NREM sleep. This paper will investigate those differences as well as the underlying brain mechanisms that accompany dream production.

REM sleep is characterised by selective activation of the limbic system, in particular the amygdala, which is associated with aggression and fear, and the hypothalamus, which regulates, among other things, sexual arousal. The limbic system is often called the “emotional centre” of the brain. During this period, the brain undergoes periodic reductions in forebrain serotoninergic activity during REM sleep (McNamara, et al.).

NREM sleep is characterised by a global decrease in cerebral energy metabolism relative to REM and waking states. There is more activity in the frontal cortices as compared to REM sleep and spindling and delta activity accompanies temporal lobe activation (McNamara, et al.).

The study by Smith, et al. 2004, studies the motivational and emotional aspects of dreaming. The limbic system plays a primary role in motivation and emotion and Smith et al. hypothesised that dreams during REM sleep would be more emotional and goal-oriented than NREM dreams. They assessed motivation by how many times a subject expressed a desire to change his or her current state or the state of the dream to some other state. The emotional aspect of the dream was assessed using a Likert scale of common emotions.

Smith, et al. performed their study by waking subjects several times during REM and NREM periods and asked them to report on their dreams, if any. They further hypothesised that initial reports of REM dreams, taken within 15 seconds of waking, would be biased toward more concrete features such as visual imagery and that more abstract features, such as goals and overall affect would become more apparent after some reflection. To test this, they asked the subject to report on their dreams a second time after they provided their initial report.

Smith, et al. found that reports on REM dreams were significantly longer than reports on NREM dreams. Based on analyses of report content, they found that REM motivation was significantly greater than NREM motivation. Total emotion was more intense in REM than in NREM and negative emotions were significantly more intense in REM than NREM.

McNamara, et al. performed a similar study, to investigate the nature of social interactions during REM and NREM dreams. First, they noted that social aggression is associated with reductions in serotoninergic activity as well as hypothalamic and amygdalar activation. They suggested that REM dreams would be characteristically more aggressive than NREM dreams. They also performed their study by waking subjects at various REM and NREM periods during the night and asking them to report on their dreams. In addition, they had subjects report on their activities at randomly distributed times during the day when they were awake.

McNamara, et al. found that REM reports were more likely to involve social interactions than NREM reports and twice as likely to involve social interactions than wake reports. In general, dream reports had significantly more social interactions than wake reports. Furthermore, aggressive interactions were three times as likely to be reported in a REM report than a wake report and twice as likely to occur in a REM dream than a NREM dream.

They also studied the role of the dreamer in each social interaction. REM dreams involved significantly more self-initiated aggression than friendliness. NREM dreams involved significantly more self-initiated friendliness than aggression. The likelihood of a friendly interaction was more or less the same across the three states. In addition, the dreamer was never reported to be an aggressor in NREM sleep.

McNamara, et al. also found that wake reports contained significantly more familiar characters than either REM or NREM, but more characters appeared in dream reports than wake reports. In addition, roughly the same number of characters participated in aggressive interactions in the two dream states.

Their findings suggest that there is an active NREM process that inhibits aggressive social impulses while promoting friendly impulses. Similarly, there must be a REM process that promotes aggression. They propose that there are two distinct processes that produce dreams. The study also gives rise to some new questions. Why are dreams particularly concerned with social interactions proportionately more so than waking life experiences? Why should REM sleep be particularly concerned with aggressive interactions while NREM sleep is concerned with friendly interactions?

Finally, the study by Takeuchi, et al. studies REM and NREM dreams in a different way. They propose that dreams may be affected by prior REM and NREM periods as well as external stimulation during NREM periods. They performed their study by inducing sleep onset REM periods (SOREMPs) and sleep onset non-REM periods (NREMPs) in healthy individuals. To do this, they used the Sleep Interruption Technique (SIT). The purpose of the study was to investigate the quantitative and qualitative differences between SOREMP dreams and NREMP dreams.

They found that dream reports were very likely in SOREMP and that recall failure and reports of no dreams were related to NREMP. Furthermore, they found that the amount of REM significantly contributed to the likelihood of dream recall in SOREMP and that the amount of stage W significantly contributed to the likelihood of dream recall in NREMP. Finally, they found that compared to NREMP dreams, SOREMP dreams were clearer, more explicit, more memorable, more vivid, noisier, more dynamic and had more activity. When they eliminated results that may have been contaminated by arousal factors such as body movements, non-descending stage shifts or the appearance of stage W, they found that all of the SOREMP reports contained dreams and none of the NREMP reports contained dreams.

All of these studies highlight the quantitative and qualitative differences between REM and NREM dreams. In addition, they suggest that different mechanisms produce dreams in REM and NREM sleep. Dreams that occur during REM sleep tend to be more emotional and motivational probably because of the limbic system activation that characterises REM sleep. NREM dreams are less likely to occur and are less elaborate than REM dreams. In addition, they may be influenced or prompted by the presence of external stimuli.

Things need not have happened to be true. Tales and dreams are the shadow-truths that will endure when mere facts are dust and ashes, and forgot.
- Neil Gaiman


McNamara, P., McLaren, D., Smith D., Brown, A. & Stickgold, R. (2005). A “Jekyll and Hyde” Within: Aggressive Versus Friendly Interactions in REM and Non-REM Dreams. Psychological Science, 16(2), 130-136.

Smith, M. R., Antrobus, J. S., Gordon, E., Tucker, M. A., Hirota, Y., Wamsley, E. J., et al. (2004). Motivation and affect in REM sleep and the mentation reporting process. Consciousness and Cognition, 13(3), 501-511.

Takeuchi, T., Miyasita, A., Inugami, M. & Yamamoto, Y. (2001). Intrinsic dreams are not produced without REM sleep mechanisms: evidence through elicitation of sleep onset REM periods. Journal of Sleep Research, 10(1), 43-52.

2 comments to Dreaming in REM and NREM Sleep

Leave a Reply