There is a popular misconception out there about what causes repetitive nightmares. Most people seem to think that frequent nightmares are caused by some sort of emotional trauma or haunted past. But that is not really the case.

Certainly there is a class of nightmare sufferers who have experienced trauma and their nightmares are clearly linked to that trauma. But there is another class of sufferers who do not fit the stereotype

There are many people who experience frequent nightmares who have no special histories of trauma –emotional or otherwise. So we need to look at other potential causative factors. 

In my 2008 book on nightmares (McNamara, P. (2008). Nightmares: The science and solution of those frightening visions during sleep. Westport, CT: Praeger Perspectives ), I suggested that one overlooked factor was the balance between REM and NREM sleep amounts the individual experiences on a nightly basis. 

The normal balance between REM and slow wave NREM sleep (SWS) was modeled by Borbély and Wirz-Justice back in 1982. Their model has been widely successful in capturing the need to catch-up on lost sleep when the balance between REM and NREM is upset due to sleeplessness/insomnia or some other perturbation. 

Normal sleep is under homeostatic control and requires a balance between REM/NREM sleep amount and intensity. In the Borbely/Wirz-Justice ‘two-process’ model of sleep regulation a sleep need process (Process S) increases during waking (or sleep deprivation) and decreases during sleep. This part of the model indexes restorative aspects of sleep. Process S is proposed to interact with input from the light-regulated circadian system (Process C) that is independent of sleep and wakefulness rhythms. Slow-wave delta activity (SWA) is taken as an indicator of the time course of Process S because SWA is known to correlate with arousal thresholds and to markedly increase during the previous waking period and during the rebound period after sleep deprivation in all mammals studied. Once a threshold value of Process S is reached (i.e., once the appropriate amount and intensity of SWS is reached), Process C will be activated. Simulations using the model’s assumptions show that the homeostatic component of sleep falls in a sigmoidal manner during waking and rises in a saturating exponential manner during sleep.

The two-process model predicts that both REM and NREM are under homeostatic control and mutually inhibit one another’s expression. Too much REM is associated with reductions in SWS and vice versa. Borbély’s model tells us how sleep intensity is adjusted depending on sleep need. If an animal goes without sleep for too long the urge to sleep builds up and once sleep is initiated the animal spends an inordinate amount of time in SWS and only after that is accomplished is REM sleep made up.

In the case of nightmares, the available data suggest that there is too much REM and too little SWS and thus the REM/SWS balance is lost. A recent report from a Hungarian group in the European Archives of Psychiatry and Clinical Neuroscience (Simor P, Horváth K, Gombos F, Takács KP, Bódizs R.; Disturbed dreaming and sleep quality: altered sleep architecture in subjects with frequent nightmares. Eur Arch Psychiatry Clin Neurosci. 2012 Apr 24. [Epub ahead of print]) the authors investigated the sleep architecture of 17 individuals with frequent nightmares and 23 control subjects. Now it is important to note that these data are extremely valuable because it is so hard to get polysomnographic studies on frequent nightmare sufferers for a variety of reasons.

In the Simor et al study the authors claim that they found that the subjects with nightmares evidenced a reduced amount of slow wave sleep, increased nocturnal awakenings, and longer durations of REM sleep (“that was mediated by heightened negative affect”). I have not yet read the paper so the above summary is based on the abstract. The basic finding however is clear: people with frequent nightmares have a measureable decrease in slow wave sleep and a increase in REM sleep and the latter was related with negative affect. 

Whether or not one can argue that the imbalance between SWS and REM in nightmare sufferers is causative or contributory it may be that one avenue of treatment for frequent nightmares is to restore the REM/NREM balance. That can be done pharmacologically or sometimes merely by shifting sleep schedules (i.e. treating the nightmare disorder as a disorder of circadian rhythms) and observing strict sleep hygiene habits. I do not wish to make things sound so simple but there may be a small number of nightmare sufferers who respond to these simple interventions.


Borbély, A. A., & Wirz-Justice, A. (1982). Sleep, sleep deprivation and depression: A hypothesis derived from a model of sleep regulation. Human Neurobiology, 1(3), 205-210.

You are reading

Dream Catcher

Sleep and the Social Brain

The sleeping brain uniquely functions to process social information.

You Really Need to Be Tracking Your Dreams

Attention to your dreams produces gold over time.

Sleep, Genetic Conflict, and Human Longevity

Non-optimal sleep durations impact sperm quality and perhaps longevity.