Elsevier

Psychiatry Research

Volume 81, Issue 1, 19 October 1998, Pages 1-8
Psychiatry Research

Role of REM sleep and dream affect in overnight mood regulation: a study of normal volunteers

https://doi.org/10.1016/S0165-1781(98)00089-4Get rights and content

Abstract

To test that REM sleep and/or dreaming aid in the overnight regulation of negative mood, 60 student subjects, selected to have no current or past episodes of depression, were tested with the Profile of Mood States (POMS) before and after two nights of laboratory sleep. There was a significant overnight effect of sleep on the Depression scale (Dep) both on a sleep-through night and a night of REM awakenings for dream recall. Pre-sleep Dep was significantly correlated with the affect in the first REM report. Although Dep scores were truncated due to the screening criteria, a subgroup of the 10 highest scorers differed from the 50 low scorers in the distribution of dream affect categories across the night. Low scorers displayed a flat distribution of positive and negative affect in dreams, while those with some pre-sleep depressed mood showed a pattern of decreasing negative and increasing positive affect in dreams reported from successive REM periods. This suggests that dreaming may actively moderate mood overnight in normal subjects.

Introduction

There is now an extensive literature on the effects of sleep on morning mood in normal subjects and in various patient groups. These studies have been conducted under conditions of intact sleep, partial sleep deprivation, selective sleep stage (REM) deprivation and total sleep loss. Overall, the findings suggest that overnight reductions in negative mood vary with the amount and kind of intervening sleep and with the presence of major depression.

When sleep is intact, normal subjects generally experience an improvement in morning mood. Kramer (1994)has reported a regular post-sleep reduction in the level of morning `unhappiness' on the Clyde Mood Scales (Clyde, 1985). In contrast, some 50–75% of depressed patients show diurnal variation in mood, most of whom are `evening types' who feel better at night and more depressed following a night's sleep (Stieglitz et al., 1988).

When normal subjects are sleep deprived for one night, their positive mood scores are significantly lower in the morning (Colecchia et al., 1997). In contrast, Wu and Bunney (1990)summarized the recent data showing 60% of depressed patients have a positive change in mood following one night of total sleep deprivation, although most relapse quickly. Among those who are `evening types' on the Hamilton Depression Rating (HDR) Scale (Hamilton, 1967), this mood improvement effect is even stronger. Reinink et al. (1990)report up to 90% improve on self-rated morning mood after a night of sleep loss. Sixty percent also respond positively to 3–5 nights of late night, partial sleep deprivation (Dessauer et al., 1985; Holsboer-Trachsler and Ernst, 1986), a manipulation which results in a higher proportion of non-REM sleep loss than of NREM sleep.

When normal subjects are selectively REM sleep deprived throughout the night, they have been found to have increased feelings of anxiety and irritability on projective tests (Clemes and Dement, 1967). In contrast, Vogel et al. (1975)showed a strong beneficial effect from extended REM loss in a group of hospitalized depressed patients. In this study 50% of the sample improved sufficiently, following an average of 3 weeks of REM sleep deprivation, to be discharged and to remain well for extended periods of time.

If, ordinarily, a full night of sleep is followed by an improvement in negative mood, at least two explanations may account for why this is not found in the depressed: (1) the depressed have higher levels of negative mood to be regulated; and (2) their intervening sleep is not normal. The differences in sleep of those with major depression from normal sleep have been well documented to include poorer sleep efficiency, reduced amounts of delta sleep, and a shift in the distribution and an increase in the activity of REM sleep (Gillin et al., 1979; Reynolds and Kupfer, 1987). Many patients show an early onset of REM sleep and a higher level of eye movement activity (REM density) than observed in normal subjects.

Not only is the REM sleep shifted and intensified, the dream content associated with REM is also reported to be affected by major depression. Those who are severely depressed are reported to have poorer dream recall and blunted dream affect (Riemann et al., 1988; Armitage et al., 1995) while those who have less severe depression, especially females, have higher rates of unpleasant dreams when compared to those with other psychiatric diagnoses, or to normal subjects (Beck and Ward, 1961; Hauri, 1976; Cartwright, 1992). It was suggested that this negative dream style might account for the higher rate of depression among women and their vulnerability to recurrent episodes. An additional finding linked the presence of more elaborate dreaming to the early REM sleep characteristic of the depressed (Cartwright and Lloyd, 1994).

In summary, an explanation of why the depressed show a short-term positive effect from being kept awake all night, or from missing the second half night of sleep, may be that these manipulations eliminate some or all of the `dysfunctional' REM sleep, and/or reduce the experience of unpleasant dreams. On the other hand, the more lasting positive effect of extended total REM deprivation reported by Vogel et al. (1975)may be that their method of inducing REM loss actually provokes more active dreaming. Their method involved awakening patients at the first signs of REM sleep for six consecutive nights followed by a seventh night of ad lib sleep, a routine that was repeated for several weeks until a waking change in mood was noted. It is likely that on each seventh night a REM rebound effect occurred resulting in earlier, more active REM, with an intensification of the associated dream affect. If the dreams of these hospitalized patients were of the `blunted affect' type prior to their experimentally induced sleep loss, their improvement in waking mood may be due to the activation of more intense dreaming during the REM rebound rather than to the deprivation.

This suggestion is in line with the hypothesis that REM sleep and/or dreaming activity regulates negative affect overnight when this is at normal levels, but that this effect is reduced when REM sleep is absent or when clinical depression is present. In depression the pre-sleep negative mood level may be higher than can be accommodated in sleep, especially when this is curtailed by early morning awakening. Then negative affect may remain unresolved at the end of the sleep period and so affect waking mood. This model suggests that there is a range within which waking negative mood triggers an active regulatory process to take place during sleep, but when this level is either too low or exceeded, mood regulation fails to occur (Cartwright, 1979). In other words there should be both a floor and a ceiling effect.

As a test of this model, a study was designed to examine whether overnight mood reduction occurs when there is a `normal' level of depressive mood (i.e. unhappiness), whether there is a gender difference in this effect, whether REM sleep is involved, and whether the processing of negative affect in dreams, as indicated by their frequency and pattern of distribution across the night, is associated with morning mood regulation.

Six general questions were explored in this study: (1) Do non-depressed subjects show a significant decrease in depressive mood following sleep? (2) Is there a gender difference in this effect? (3) Is the difference related to REM sleep variables, such as: REM percent, number of REM episodes, latency to the first REM and eye movement density? (4) Does depressed mood prior to sleep have an effect on the affective component of the first dream? (5) Is the distribution pattern of negative dreams across the night relted to the pre-sleep depressive mood? and (6) Is the post-sleep reduction in the level of morning mood related to a pattern of decreasing frequency of negative affect dreams as the night progresses?

Section snippets

Subjects

Volunteers for a two-night study were solicited through notices posted at a major medical college. The flyer offered a fee of $100 for participation. Student subjects were screened first by phone to exclude those with a self-report of present or past episode of major depression, or current use of an anti-depressant or other medication that would affect normal sleep architecture. Subjects were then interviewed, using a structured clinical interview for non-patients (SCID-NP) and the HDR Scale.

Results

Table 2 shows the four POMS depression scale scores: (1) for Night 1 prior to sleep; (2) Morning 1 on awakening; (3) Night 2 prior to sleep; and (4) Morning 2 on awakening. A split-plot analysis of variance (Kirk, 1968) was conducted with gender as the between-subjects factor and Night vs. Morning and replication (Night-Morning 1 vs. Night vs. Morning 2) as the within-subjects factor.

Results of the analysis (Table 3) indicated a significant change in scores from night to morning but this effect

Discussion

When subjects are screened to eliminate those with present or past depression episodes, the range of depressive mood scores is necessarily restricted. Nonetheless, the study shows that there is a significant reduction overnight in negative mood both when sleep is intact and when it is interrupted for REM awakenings. The fact that the small subgroup with some self-reported unhappiness continued to regulate mood despite REM interruption could not be explained on the basis of the sleep parameters.

Acknowledgements

This study was supported by a grant to the first author from the National Institute of Mental Health, MH-50471.

References (25)

  • Clyde, D., 1985. Clyde Mood Scale, Clyde Computing Service, Key Largo,...
  • Colecchia, E., Berardi, A., Hirschfeld, U., Kosslyn, S., Leproult, R., Levin, T., Riffle, A., Van Cauter, E., 1997....
  • Cited by (126)

    • Dreaming

      2023, Encyclopedia of Mental Health, Third Edition: Volume 1-3
    • Changes in affect

      2023, Encyclopedia of Sleep and Circadian Rhythms: Volume 1-6, Second Edition
    • Physiological feelings

      2019, Neuroscience and Biobehavioral Reviews
    View all citing articles on Scopus
    View full text