You are on page 1of 6

The Effect of Dream (Stage REM) Deprivation on Adaptation

to Stress
Ramon Greenberg, MD, Richard Pillard, MD and Chester Pearlman, MD
The role of dreaming in the assimilation and mastery of new experiences
was examined in this study. Previous work had shown that a film of an autopsy can evoke measurable psychologic and physiologic indices of anxiety. Adaptation to the experience was indicated by lower levels of anxiety
during a second viewing of the film. We examined the effect of dream
(Stage REM) deprivation on adaptation to the second viewing. Between
the first and second viewings, 9 subjects were dream deprived, 5 had normal sleep and 6 were awakened from non-REM sleep. For those who
showed a significant anxiety response to the first viewing, the dream deprived group showed significantly less adaptation to the second viewing
than the other two groups. These results support the hypothesis that
dreaming aids adaptation to anxiety-provoking stimuli.

The concept that dreaming serves to fulfill


wishes, presented by Freud in the Interpretation of Dreams, eclipsed all previous theories
about dreaming, but subsequent writers have
frequently commented on the limitations of this
hypothesis as a total theory of dream function.
Maeder(l), Garma (2) and Piaget (3) have
proposed that dreaming is involved in adaptation to traumatic (anxiety-arousing) experiences. French and Fromm (4) elaborated this
concept into a theory that dreaming is concerned with the formulation of solutions to current "focal conflicts" of the dreamer.
Studies based on current psychophysiologic
concepts of sleep have also led to the idea that
dreaming during stage REM sleep* is involved
in processing new experiences. Breger (5) formulated dreaming as a working over and inteFrom the Boston Veterans Administration Hospital and
the Boston University School of Medicine.
Supported by Research Scientist Development Award
Grant No. MH-32896.
Received for publication Sept 7, 1971; revision received
Nov 1 I, 1971.
Address for reprint requests: R. Greenberg, MD, Psychiatry Research, Boston VA Hospital, 150 S Huntington
Ave, Boston, Mass 02130.

gration of recently perceived input into already


existing structures. Gaarder's (6) hypothesis
also included the idea of adding recent experiences to existing structures for adaptive purposes. Dewan (7) characterized REM as a
programming process for the brain. We recently presented a similar hypothesis (8). We
stated that daily experiences may arouse repressed conflicts, memories or feelings. In the
dream, these new experiences interact with the
previously repressed material with a resultant
reinstitution of characterologic defense patterns
that previously have been used to deal with this
kind of experience.
The aim of the present study was to investigate the role of dreaming (Stage REM sleep) in
adaptation by normal persons to an anxietyprovoking experience. Previous work (9) had
shown that viewing a film of a medical autopsy
usually aroused anxiety in a normal medically
unsophisticated subject. Most normal subjects
* Mental content has frequently been recorded following
awakening from Non-REM sleep but, for most persons, it is
easily distinguishable from the dream experiences associated with REM sleep.

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

257

GREENBERG ET AL

subsequently adapted to the experience so that a


second exposure produced much less anxiety.
In this study, we compared the influence of
REM-deprived sleep, normal sleep and interrupted (but not REM-deprived) sleep on the
subjects' capacity to adapt to the second viewing
of the film. From our hypothesis about the
function of dreaming, we predicted that subjects
deprived of REM sleep between the first and
second viewing would be more anxious and
show less adaptation to the second viewing than
those who had normal sleep or interrupted
sleep without REM deprivation.
METHOD
Twenty college students (male and female) were involved
in this study. Those who had interrupted sleep were paid
$30, and those who had undisturbed sleep received $25.
Subjects began the study with a session at the Psychopharmacology Laboratory at Boston University School of Medicine. There they became familiar with the psychophysiologic measurements which involved attachment of electrodes
to measure heart rate, respiration rate, skin potential and
finger sweat (FSP). They also were given the Psychiatric
Outpatient Mood Scale (POMS) test (10). After these
measurements, they were shown a short, interesting film
about computer generated tones.* Physiologic measurements were made for 3-minute periods before and during
the film. Following the film the POMS was again administered. That night, they reported to the Sleep Laboratory at
the Boston VA Hospital where elcctroencephalograph
(EEG), electrooculograph (EOG) and electromyograph
(EMG) recording was performed throughout a night of
undisturbed sleep. This was a practice session to acquaint
subjects with the procedure and to allay nonspecific apprehension.
A week later, each subject returned to the Psychopharmacology Laboratory where the same procedure was repeated, except that on this occasion the movie was an
8-minute excerpt from a medical film, "Basic Autopsy Procedures." It showed a pathologist examining a cadaver,
making an abdominal incision, a scalp incision, sawing and
chiseling away the calvarium and making an incision along
the spine. Sound was turned off to increase opportunity for
fantasy. That night, they again reported to the Sleep Laboratory where 5 were allowed undisturbed sleep, 9 were

"A Pair of Paradoxes" by RN Shepard and EE Zajac.


Furnished through the courtesy of Bell Telephone Laboratories.
258

REM deprived by being awakened at the first signs of


REM sleep, and 6 were awakened from non-REM sleep
the same number.of limes as a paired member of the REMdeprived group had been awakened. The next morning they
returned to the Psychopharmacology Laboratory where the
same film and measurements were repeated. The psychophysiologist did not know the subjects' sleep group.
Following this session, each subject was interviewed to elicit
his feelings and impressions of the experience.

RESULTS

We first assessed the stressfulness of the autopsy film. Increases were registered in heart
and respiration rates, skin potential, finger
sweat and those factors on the POMS which
describe "tension anxiety" and "disgust
shock." Of these, the increases in FSP and
POMS were significant (P = <.01) (Figure
1).* By contrast, most subjects did not develop
any change in these variables when they saw
the computer tone film. Thus, the stressful
nature of the autopsy film in comparison to the
computer tone film was clearly demonstrated in
the increases in FSP and POMS scores; thus we
assumed the tests were measuring an increase in
anxiety.
Most subjects also reported that the first
viewing of the autopsy film bothered them.
During the period of time between viewing the
film and reporting to the Sleep Laboratory,
they felt a) "Shocked and surprised," b) "1
couldn't be alone after watching it," c) "It was
a shock treatment. . . ," d) "I was tempted to
ask you to stop it," "Repulsive. . . horrible."
One subject requested that the film be stopped
halfway through.
The POMS and FSP results showed the
stress reaction for most but not all of the subjects. Those who failed to show greater anxiety
with the first autopsy film than with the computer film were dropped from further analysis.
For the POMS this included 2 control sleep, 1
* POMS "tension/anxiety" scored as described by
McNair et al (10) with some items added to reflect feelings
of disgust and shock. FSP scored from a photographic density scale described in reference 9.

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

EFFECT OF DREAM DEPRIVATION

Fig 1. Comparison of responses of all subjects to computer tone and autopsy films on
the POMS and FSP scores.

control awakening and 1 REM-deprived subject. For the FSP this included 1 control
awakening and 4 dream-deprived subjects.
Without a definable measure of stress from the
first viewing of the autopsy film, we assumed
that the effects of dreaming on the stress could
not be assessed. The excluded subjects were not
necessarily the same for the POMS and the
FSP. This seems understandable because people manifest anxiety in different ways.
The next question involved the difference in
sleep patterns between REM-deprived and
control groups. The REM-deprived group had
7 to 19 minutes of REM sleep. The control
awakening and control subjects had from 53 to
97 minutes of REM. Sleep parameters of the
control awakening and unawakened subjects
did not differ significantly and these groups
were combined for comparison with the REMdeprived group. The combined control group
had an average of 360 minutes of sleep and the
REM-deprived group averaged 265 minutes.
oControl (N = 8)
Dreom Deprived (N = 8)

t
Sleep

AUTOPSY
Fig 2. Comparison of POMS anxiety scores for
REM-deprived and control groups on the first and
second viewings of the autopsy film.

Of this 95 minutes difference, about 70 is accounted for by the difference in REM.


We now turn to a comparison of the effects of
viewing the autopsy film a second time on the
combined-control and REM-deprived groups
(Figure 2). (There was no difference between
the control awakening and undisturbed sleep
groups.) Using the tension-anxiety scores of the
POMS, both groups had a significant increase
in anxiety in response to first viewing the
autopsy (t = 3.92; df = 15; P < .002) and a
lesser increase to the second viewing (t = 1.90;
P < .10). Thus, both groups showed some
adaptation to the second viewing. The REMdeprived subjects were more anxious, however,
especially on the postfilm measure (/ = 2.14;
P < .05).
Another way to test this effect is by a 2 x 2
analysis of covariance using the scores after the
first autopsy viewing as the covariate. This
permits an estimate of whether the change in
anxiety for the two groups was different. Analysis of the data in this fashion shows that
while both groups have an increase in anxiety
on the second viewing (F = 3.42; P < .10), the
REM-deprived group is made more anxious
than the control by the second viewing
(F = 5.41; P < .05), almost approaching the
levels they showed after the first viewing. The
interaction term was not significant, showing
that the anxiety increase for the two groups was
not of different magnitude.
A final way of viewing the data is to compare
the difference between responses to the first and
second viewing for each subject. Six of the 8
control subjects were affected less by the second
viewing than by the first, 1 showed no change,

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

259

GREENBERG ET AL

and 1 was made more anxious. In contrast, 5 of


the 8 R EM -deprived subjects were made more
anxious by the second viewing than by the first,
2 were less affected, and 1 showed no change.
The Mann-Whitney U test showed this difference to be significant (P < .05).
The FSP scores were analyzed in the same
fashion. Our criterion for excluding subjects
who failed to show greater increase in FSP to
the autopsy than to the tone film left only 4
subjects in the REM-deprived group. The results parallel those on the POMS test, but because of the small number of subjects the difference between control and REM-deprived
groups is nonsignificant.
DISCUSSION

This study was designed to explore the relationship of dreaming to adaptation to an anxiety-provoking situation. The results showed
that, with a stress that clearly induced anxiety,
a second experience of the stress revealed significantly less adaptation by the REM-deprived
group than by the control group. That is, the
REM-deprived group was significantly more
anxious following the second viewing than the
control group. Our hypothesis suggests the following explanation of this finding: When an
individual meets a situation which is stressful
for him, the stressfulness is due to the arousal of
memories of prior difficulties with similar situations. The person's initial defensive reaction is
usually of an emergency or generalized type
(such as global denial or repression). Then,
during the dream experience, these feelings
from the past and the current stressful stimulus
are integrated, and the individual's characteristic defenses for that particular set of emotions
and memories are used to deal with the current
threat. If the stress is re-experienced, he now
has available his characteristic (for him most
efficient) means of dealing with the threat.
Thus, re-exposure to the stress should not produce the initial degree of anxiety.
Another possible explanation for the poorer
260

adaptation is that the subjects were in a sensitized state due to increased drives as a result of
the REM deprivation procedure. While this
interpretation is consistent with classic psychoanalytic concepts of dream function, there is
no evidence confirming the development of
increased drives with 1 day of REM deprivation. A study in our laboratory revealed that
rats REM deprived for 1 day showed neither an
increase in exploratory behavior nor increased
bar pressing for food. The Rorschach protocols
in our study of REM deprivation in humans
also failed to show a consistent increase in drive
state (8).
Although the results show a significant impairment of adaptation by the REM-deprived
subjects, some adaptation did occur. A possible
explanation for this is that they all did some
dreaming during the night. In a similar study,
Breger et al (11) found that REM-deprived
subjects reported dream fragments incorporating elements from the stressful film. This
implies an immediate attempt to deal with the
anxiety-provoking stimulus in the dream. It
would be extremely difficult to evaluate just
how much dreaming is necessary to deal with
the stressful situation. It is also possible that
some REM-deprived subjects continued to use
emergency defenses (eg, intellectual understanding of the situation) which permitted some
reduction in anxiety.
Two other observations in this study are
worth noting. There was a tendency for anxiety
and fatigue in the REM-deprived subjects to be
increased prior to viewing the second autopsy
film. Most studies of REM deprivation have
not shown a generalized increase in anxiety
(12, 13). Why, then, did we observe some
increased anxiety? The unique aspect of this
REM-deprivation study is that the subjects
were presented! with an anxiety-provoking
stimulus prior to the REM deprivation. The
increase in anxiety may well represent defective
handling of the anxiety aroused by the stimulus. In other words the individual is presented

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

EFFECT OF DREAM DEPRIVATION

with a specifically labeled task and, without


REM, the anxiety continues. The fatigue might
represent unusual efforts to deal with the anxiety due to prevention of the usual mechanismie, dreaming. While this explanation is
speculative, it fits our hypothesis and also offers
a new way of studying REM deprivationie,
by presenting subjects with a task before the
deprivation procedure. For example, wearing
inverting prisms has been shown to lead to an
increase in REM. Would REM deprivation
during the period of wearing the prisms lead to
impaired adaptation?
In discussing the psychoanalytic psychology
of adaptation, Joffe and Sandier (14) emphasized the continual development by the ego of
new organizations of the "ideal state of the
self" to preserve the feeling of safety and to
avoid being traumatically overwhelmed. A
stressful experience is a demand for such an alteration, and a traumatic experience involves
a demand for greater change than the individual is capable of at the time. It is a truism that
successful adaptation requires time. The
reasons for this are not so clear. Experiments
with the learning of cognitive material suggest
that such learning is completed soon after the
learning experience. Some defensive operations
occur in a similar close relation to emotionally
stressful experiences. Thus, why adaptation or
emotional learning requires so much more time
is puzzling. Thinking about the stressful
experience or discussing it with other people
may facilitate adaptation, btit anxious brooding, usually accompanied by sleep disturbance,
indicates the limitations of conscious thinking
as a mechanism of adaptation. It is also commonly observed that adaptation occurs relatively independently of the conscious action of
the subject. It "just happens." In attempting to
determine when it happens, one might propose
that during waking life unconscious processes
of psychic adaptation occur which are analogous to other processes of growth and homeostasis. In our opinion, such a simple model

does not do justice to clinical experience. The


vivid example of the posttraumatic dream suggests another model. Posttraumatic dreams reveal, over a period of time, an evolution from
repetition of the traumatic event to a change in
the content which reveals some of the elements
from the past which have been aroused by the
traumatic event. The past and present become
woven together and the traumatic event is
gradually mastered. When this fails to happen,
as in the repetitive war dream, the patient
continues to have a symptomatic neurotic illness.
SUMMARY

In this study we have examined the hypothesis that a critical intervening process in
adaptation to a stressful situation is dreaming
or the REM stage of sleep- We studied a group
of volunteer subjects who, after adaptation to
the laboratory, conditions, were shown a
stressful movie on 2 consecutive days. During
the night between these two viewings, some
subjects were REM deprived, some awakened
an equivalent number of times during NREM
sleep, and some allowed to sleep undisturbed.
Psychologic and physiologic measures of anxiety were obtained in relation to the two
viewings. The results showed that the subjects
who were REM deprived showed significantly
less habituation to the second viewing than the
control subjects. These findings were discussed
in relation to the hypothesis that REM sleep
serves to integrate memories of similar experiences with the current stress, allowing the use
of the individual's characteristic defenses. When
the film is viewed for a second time, those subjects who have been allowed to dream were
made less anxious than the REM-deprived
group because they could now use their characteristic, and probably more efficient, defenses to
deal with the anxiety-provoking aspects of the
film. REM-deprived subjects, on the other
hand, showed a decreased ability to adapt to the
specific stress.

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

261

GREENBERG ET AL

REFERENCES

8. Grccnberg R, Pearlman C, Kavvliche S, et al:

1. Maeder, AE: The dream problem. Nerv Mcnt


Dis, Monograph No. 22, 1916
2. Garma A: The traumatic situation in the genesis
of dreams. Int J Psychoanal 27:134-139, 1946
3. Piaget J: Play, Dreams and Imitation in Childhood. New York, W. W. Norton & Company,
Inc, 1951
4. French TM, Fromm E: Dream Interpretation:
A New Approach. New York: Basic Books, Inc,
Publishers, 1964
5. Breger L: The function of dreams. J Abnorm
Psychol, Monograph No. 641, 1967
6. Gaarder K: A conceptual model of sleep. Arch
Gen Psychiatry 14:263-70, 1966
7. Devvan E: Programming (P) hypothesis for
REM sleep, Sleep and Dreaming, International
Psychiatric Clinics. Edited by E Hartman. Boston, Little, Brown and Company, 1970, pp
295-307

262

9.

10.

11.
12.

13.

14.

The effects of dream deprivation. Br J Med


Psychol 43:1-11, 1970
Pillard R, Atkinson K, Fisher S: The effect of
different preparations on film induced anxiety.
Psychol Record 17:35-41, 1967
McNair DM, Lorr M, Dropplcman LF: Manual, Profile of Mood States. San Kiego: Educational Industrial Testing Service, (In press)
Brcger L, Hunter I, Lane RW: The effects of
stress on dreams. Psychol Issues 27:
1971
Kales A, Hodemaker FS, Jacobson A, et al:
Dream deprivation: an experimental reappraisal. Nature 204:1337-8, 1964
Sampson H: Psychological effects of deprivation
of dreaming sleep. J Nerv Mcnt Dis 143:305317, 1966
Joffe WG, Sandier J: Comments on the psychoanalytic psychology of adaptation. Int J Psychoanal 49:445-54, 1968

Psychosomatic Medicine, Vol. 34, No. 3 (May-June 1972)

You might also like