go to content

SLEEP DEPRIVATION IN DEPRESSION: AN INTEGRATIVE APPROACH

Vadim S. Rotenberg, E-mail: vadir@post.tau.ac.il,

INTERNATIONAL JOURNAL OF PSYCHIATRY IN CLINICAL PRACTICE 2003 VOLUME 7 PAGES 9-16

Received 8 March 2001; accepted for publication 29 May 2002

The question as to whether the beneficial effect of sleep deprivation in depression is related to the increased wakefulness or to the sleep suppression by itself, is made moot by the search activity concept. According to this concept, rapid eye movement (REM) sleep is functionally deficient in depression and increases feelings of helplessness and hopelessness instead of restoring mood and search activity Thus, REM sleep deprivation, either selective or not, is beneficial by breaking a vicious circle: depression in wakefulness.. .giving up (helplessness) in dream scenario.. .increased depression in the subsequent wakefulness. In addition, sleep deprivation is an important challenge for the depressed patient. The ability to confront this challenge and maintain wakefulness has a positive outcome on depression, especially when wakefulness is accompanied by active behavior. (Int J Psych Clin Pract 2003; 7: 9-16)

Keywords: sleep deprivation, REM sleep, search activity, depression, wakefulness

INTRODUCTION

Numerous investigations have shown that total sleep deprivation (SD) often produces a dramatic decrease in depressive symptoms within 24 h;1 however, the mechanism of the therapeutic effect of SD on depression is still not known. Although many different explanations of this mechanism have been proposed, not one of them is free from contradictions. The goal of the present paper is to perform a short but comprehensive review of the concepts explaining antidepressant response to SD, to elucidate the present contradictions and to propose an integrative concept free from contradictions.

SD AS A PROLONGATION OF WAKEFULNESS

The main question is whether the beneficial effect of SD is related to increased wakefulness or to the sleep suppression. In other words, whether sleep has a depressogenic effect or wakefulness has an antidepressant effect. There are many reasons to suggest that it is the additional duration of wakefulness that plays a key role in the reduction of depressive symptoms.2 According to Van den Hoofdaker,3 SD prolongs natural remission which starts in endogenous depression in evening hours. This point of view is confirmed by data that patients who display an increase of mood from morning to evening respond to SD better than patients who do not display such mood increase.4 However, what mechanism is behind this mood increase from morning to evening? Is it only a circadian variation controlled by biological rhythm? If such is the case, as many investigators believe, then it would be reasonable to expect that according to diurnal variation, mood would be worse in the morning hours immediately after SD, in comparison to the night hours during SD. That, however, is not the case. Following SD, mood peaks in the early morning hours, exactly when according to the diurnal rhythm, it should be at its lowest level. Of course, SD may interrupt the circadian rhythm; however, there is no reason to expect that this rhythm would be changed to the opposite one, which appears to be what happens. Thus, it is possible to speculate that the restoration of mood from morning to evening, as well as after SD, is initially related to the wakefulness itself.

I suggest that wakefulness by itself can improve mood if it is accompanied by some sort of goal-directed activity, such as washing, dressing, buying food and cooking, and participation in some social relationships. The cumulative effect of this activity, even when performed on a very low level, may be at least partly responsible for mood improvement. This suggestion is based on the well-known experience that if it is possible to stimulate a patient with major depression to be active in some directions, the depression often diminishes and mood improves, at least for a short period.

An additional factor is that SD by itself presents a strong challenge. A depressed patient often gives up in the face of life challenges. In the morning following SD, a patient finds that he has been able to maintain wakefulness during the entire night, which requires a great deal of effort. This experience may have a stimulating effect on the subject's self-esteem and contribute to the positive effect of the prolonged and active wakefulness.

The interpretation of the SD effect as an outcome of subject's activation is in some aspects similar to the view of Ebert and Berger.5 These authors suggested a neurobiological similarity between antidepressant SD and use of psychostimulants. SD, as well as amphetamine, decreases dopamine release and limbic metabolism if both of them are high (in SD responders); however, brain metabolism increases if it is low. Metabolism depends on dopamine receptor stimulation, and SD responders show signs of increased dopamine release before SD. In animal studies, SD effect on the dopamine system is similar to the effect of psychostimulants. However, although alteration of brain CA activity may play an important role in antidepressant SD effect and prolonged wakefulness, the similarity between SD and psychostimulant use is limited. First of all, a similar effect of SD and psychostimulants on brain CA activity was discovered in "healthy" animals.

In human studies, the SD effect on patients suffering from major depression is very different from the SD effect on healthy subjects. For one thing, in healthy subjects, SD usually does not increase mood. SD in healthy subjects may increase excitability, usually in combination with tiredness and sleepiness, but this is incomparable to the effect of psychostimulants. Secondly, while SD in SD responders has an obvious antidepressant effect, psychostimulants usually do not display such effect.

It is well known that SD is more effective in endogenous depression than in reactive/neurotic depression. The main difference between endogenous and exogenous (reactive) depression is in the domain of the reactivity to the environment. Endogenous depression is characterized by absence of reactivity to environmental changes or precipitating stress, and by the low sensitivity to life events.6 At first glance, this statement appears to contradict the author's argument. However, a low sensitivity to the external world and absence of precipitating stress also means that stressful events do not play an important role in eliciting endogenous depression, while they play an important role in eliciting reactive depression and anxiety. SD, as a prolongation of wakefulness, is much less effective in reactive depression and often has a negative outcome in anxiety disorders. Thus, a patient suffering from endogenous depression, in contrast to patients with exogenous/reactive depression and anxiety, is relatively protected from harmful emotional disturbances caused by the environment. In this condition, relationships with the. environment during wakefulness may be more positively stimulating and beneficial. There are anecdotal reports of sudden remission of endogenous depression after a strong and sudden stress, such as death or illness of family members, fire or financial catastrophe. Such reports, although anecdotal, show that there could be a positive outcome of stress in endogenous depression if stress leads to mobilization and urgent activity. Thus, it is possible to also suggest a positive outcome from everyday mild activity. This argument is consistent with data that a high vigilance level, high behavioral activity and low level of tiredness all are related to a favorable SD response.7-8

Anxiety and exogenous depression are also often accompanied by a high vigilance level and behavioral activity, which suggests that not every type of behavioral activity is related to a favorable SD response, and is an additional reason why the effect of SD in SD responders is difficult to consider as a simple psychostimulant effect. According to our previous investigations,9 patients suffering from endogenous depression have a favorable effect from antidepressant treatment if they demonstrate a first-night effect in the laboratory - a normal orienting reaction to the new environment. Thus, first-night effect is part of the normal active interrelationship with the environment, and this normal interrelationship is accompanied by the increased sensitivity to the treatment. It stands to reason that there would also be increased sensitivity to the "treatment" by SD.

The present approach is in opposition to the hypothesis that cerebral fatigue induced by SD might break the distressing state of hyperarousal.10 According to this hypothesis, hyperarousal by itself is a crucial factor in the pathogenesis of depression, and SD abolishes this hyperarousal causing increased sleepiness and decreased vigilance. However, this hypothesis does not explain why even a very short sleep in the morning after SD (which is unable to compensate sleepiness after total SD) causes an acute relapse of depression. The hyperarousal hypothesis is also contradicted by our data, showing that in depression, the lack of sleep duration and slow wave sleep (SWS) in night sleep does not decrease vigilance and does not increase sleepiness during the day. Even before SD, most depressed patients suffer from disturbed sleep quality (decreased sleep efficiency and lack of SWS), and it was shown that the more pronounced these disturbances, the more resistant are depressed patients to the treatment.12-13 The hypothesis of the abolished hyperarousal and increased fatigue after SD also does not fit with data that the phase advance of the j restorative sleep after total SD can stabilize the antidepressant effect of SD.14-15 However, this data on first glance seems to I be in contradiction with our hypothesis and requires a substantial discussion.

THE DEPRESSOGENIC EFFECT OF SLEEP

If our initial hypothesis is correct and the main component of the positive effect of SD is the prolongation of active wakefulness, then it is reasonable to expect that the longer the period of wakefulness the better will be the outcome of treatment. However, this paradigm seems to be in contradiction with the positive effect of the restorative sleep displacement after SD. It was shown that if on the next day after SD the patient is allowed to sleep between 17.00 and 24.00 h, the relapse is absent and the positive effect of SD is more stable, in comparison to the normal sleep schedule.15 At the same time, relapse of depression after the total SD was not prevented, either when SD was followed by sleep in the morning hours, or when SD was followed by sleep in the subsequent night.16

The positive antidepressant effect of selective rapid eye movement (REM) SD is different from that of SD. REM deprivation requires more time and more procedures for the positive outcome; however, this outcome is more stable. According to reference 17, chronic selective awakenings of depressed patients from REM sleep without interruption of SWS produced a gradual but sustained alleviation of depression. At the same time, the positive effect of REM SD does not correspond to the hypothesis of the shifted circadian rhythm as a main reason for depression, because REM SD does not affect circadian rhythm. It also does not correspond to the hypothesis that the activation of S-process (homeostatic process that determines the requirement in sleep after wakefulness and causes the decrease of vigilance) is the reason of SD positive outcome, because REM deprivation does not cause decrease of vigilance. In any case, positive outcome of the selective REM SD requires an integrative explanation in the context of SD because SD includes also REM deprivation. All these findings lead to the conclusion that the effect of SD is at least partly related to REM deprivation.

This conclusion seems to be in agreement with data that the relapse of depression after SD caused by naps may relate to REM sleep incorporation into day naps. According to Wiegand et al.18 6/12 patients relapsed after day naps - five of them had naps with REM sleep and only one without REM sleep. Of the six patients who did not relapse, three had naps with SWS only and three with REM sleep; however, REM sleep was very short (1-1.5 min). The authors concluded that REM sleep seemed to be associated with a higher risk of relapse. The revue of literature data performed by Wu and Bunney19 has shown that the longer the naps the more often appear relapses after SD, while longer naps are usually associated with REM sleep incorporation.

If relapses after SD are related to REM sleep incorporation in naps, then it is possible to speculate that the effect of SD is at least partly related to REM SD. Of course, REM sleep is not always incorporated in those naps that lead to the exacerbation of depression. Knowles et al2 presented a case of a profound relapse after a 15-min nap early in the morning (05.00 h). The explanation of the relapses must be relevant for all cases, including naps without REM sleep, but first of all, I do not suggest that REM SD is the single mechanism responsible for the positive outcome of SD in depression. Thus, the incorporation of REM sleep is not obligatory for the reverse effect of naps. If, for instance, the active interrelation- ship with the environment during wakefulness is important for the restoration of mood, as we have suggested in the first part of this presentation, then naps in the first part of the day may deteriorate mood by preventing this interrelationship. Secondly, it is possible that the deteriorative effect of REM sleep in night sleep and in day naps after SD is related to some particular features of dream experience. Usually, dream experience accompanies physiological REM state. However, it was shown in many investigations (see references 20-21) that dream experience can shift to the NREM sleep, especially after total or partial SD22 and especially in early morning naps. In such cases, the presence of dream experience in day naps may determine the relapse even if REM sleep by itself is absent.

SEARCH ACTIVITY CONCEPT - AN INTEGRATIVE EXPLANATION OF SD EFFICACY

The question is, why does dream experience, either in REM or in NREM sleep, abolish the antidepressant effect of SD? We are going to discuss this topic in the context of the search activity concept.23-25

SEARCH ACTIVITY AND BODY RESISTANCE

Search activity is defined as activity that is oriented to change the situation (or at least the subject's attitude to it) in the absence of the precise prediction of the outcome of such activity (i.e. without a definite probability forecast) but taking into consideration the results at each stage of the activity. According to this definition, search activity is a component of many different forms of behavior - self-stimulation in animals, creative behavior in humans, as well as exploratory and active defense (fight/flight) behavior in all species. In all these forms of activity the probability forecast of the outcome is indefinite, but there is a feedback from the behavior and its outcome enabling the subject to correct his behavior in accordance with the outcome.

The opposite psychobiological state - renunciation of search, encompasses neurotic anxiety and depression in humans, freezing and learned helplessness in animals, as well as panic and stereotyped behavior in all species. Reactions to panic can imitate search activity but a subject in a state of panic does not consider the real outcome of his behavior and correct it. Panic reflects renunciation of search. This conclusion is in accordance with clinical experience. The common clinical outcome of panic is depression.26-30

One of the best indications of search activity in animals such as rats or cats is a high-amplitude and well-organized hippocampal theta-rhythm. All forms of behavior that are accompanied by hippocampal theta-rhythm (orienting, learning, searching for food, searching for a sexual partner or for security, alarm) include search activity, whereas stereotyped behavior, panic and freezing are accompanied by the desynchronization of hippocampal electrical activity.

What was the need for a new classification of behavior based on the presence or absence of search activity? I suggest that search activity has a very important biological and psychological meaning. In research conducted together with V Arshavsky, we found that all forms of behavior that include search activity but differ in other aspects increase body resistance. At the same time, all forms of behavior that include renunciation of search as a common feature decrease body resistance, especially in stressful situations, and predispose subjects to psychosomatic disorders.31-34 All forms of behavior that include search activity reduce artificial pathology in animals, while all forms of behavior not accompanied by search activity facilitate artificial pathology - such as artificial epilepsy, anaphylactoid edema, Parkinson-like muscle rigidity, heart arrythmia, increased blood pressure, etc. Psychosomatic patients have a more pronounced tendency to give up than healthy subjects, especially in stressful conditions such as chronic failure.35 A careful clinical analysis has shown that renunciation of search in a form of clinically manifested depression or give up/giving up complex36 predicts the development of somatic disorders. We concluded that the process of search activity in itself, independent of its pragmatic results (i.e. whether it is successful or not) and independent of the accompanying emotions, protects the subject from somatic disorders. This explains many clinical paradoxes, including the well-known paradox of concentration camp survivors who even displayed an improvement in their health, e.g. disappearance of psychosomatic disorders.

I suggest that search activity protects health according to a biological law. Search activity itself requires much effort and stimulates the subject to confront stressful, potentially dangerous and exhausting situations. If search activity would not protect the subject's health in such a situation, the most active and creative members of the population would become exhausted and ill. Search activity and body resistance, then create positive feedback.

SEARCH ACTIVITY AND BRAIN MONOAMINES

Concerning the relationship of the brain monoamine system to search behavior, the following hypothesis has been developed.23,37 Search activity can start in the presence of a certain critical level of the brain monoamines, which are utilized in the course of search behavior. Search activity itself, once it starts, further stimulates the synthesis of the brain monoamines and ensures their availability. Thus, the more pronounced the search activity, the sooner the turnover of the brain monoamines can be compensated for by re-synthesis, in turn maintaining search behavior (positive feedback system). For search activity to begin, the brain monoamine concentration must exceed a critical level. If it drops below this level, search activity is canceled. In a state of renunciation of search, the above-mentioned positive feedback system does not function. Furthermore, in this state the monoaminergic expenditure becomes faster, and brain catecholamines (CAs - dopamine, norepinephrine) have a tendency to drop, as happens in learned helplessness.38 This may be explained by the fact that renunciation of search is usually combined with distress, which causes intense CA expenditure, without subsequent restoration due to the absence of search activity. Thus, according to this hypothesis, monoamine functioning completes a vicious circle - renunciation of search leads to a drop in the brain CA level, which in turn leads to the renunciation of search's becoming more prominent.

This hypothesis is supported by data pertaining to the depletion of brain CA in the case of uncontrollable aversive stimulation39 - a condition usually producing helplessness. Authors39 convincingly proved that if the aversive stimuli induce coping behavior (behavior containing search activity,23 this is accompanied by an intensified utilization and synthesis of CAs. When such behavior is impossible, the utilization of amines intensifies, exceeding their synthesis.

The utilization of brain CAs during search activity, as well as during renunciation of search, is accompanied by a high level of alertness and emotional tension; however, this alertness and emotional tension is functionally and qualitatively different.11 Search activity corresponds to the adaptive emotional tension. Renunciation of search (depression) displays itself in maladaptive emotional tension,33 presumably related to limbic and frontal hypermetabolism.40

REM SLEEP AND RESTORATION OF SEARCH ACTIVITY

If search activity is so important for survival and renunciation 1 of search so destructive, it would be reasonable to assume a| natural brain mechanism that can restore search activity after temporary renunciation of search. According to the search activity concept, REM sleep and dreams fulfill this function. A covert search activity in dreams compensates for the renunciation of search and ensures the resumption of search activity in subsequent wakefulness. This claim is based on the following findings:

1. Different forms of animal behavior that contain search activity (self-stimulation, fighting) suppress REM sleep in the subsequent sleep without restorative rebound effect.41-42 This means that less REM sleep is required in such situations.

2. Renunciation of search in animals causes an increase \i REM sleep in subsequent sleep.31 A correlation detected between learned helplessness and REM sleep percentage.43-44 Depression is accompanied by the increased REM requirement (REM sleep latency is decreased while the initial REM episodes are increased).45

3. If during REM deprivation, the subject is involved in active behavior (exploration, creative task decision, active defense reaction), REM rebound effect during subsequent sleep is substantially decreased.46

4. In animals REM sleep, as search activity in wakefulness, is characterized by regular and synchronized hipocampal theta-rhythm. Moreover, the more pronounced the theta-rhythm in wakefulness, the less pronounced it is in subsequent REM sleep.47 REM sleep in animals regularly contains PGO (ponto-geniculo-occipital) waves. In wakefulness, PGO correlates with the orienting activity.48 The presence of the PGO spikes in REM sleep means that the subject is predisposed to react to novel stimuli, including spontaneous change of dream content.

5. If the nucleus coeruleus in the brain stem is artificially destroyed and as a result muscle tone does not drop during REM sleep, animals demonstrate complicated behavior that can be generally described as search behavior.49

6. Healthy subjects are characterized by active participation in their own dream scenario and by the heart rate acceleration in REM sleep that correlates with self-participation.24

7. The more characters and descriptive elements that appear in the dream, and the more active the characters and the dreamer himself, the larger the decrease in the scale of unhappiness during the night. It can be suggested that active characters represent search activity in dreams, but one must stress that in the case of healthy subjects, the changes in the scale of unhappiness during the night can be correlated only with the above-mentioned dream variables, not with the changes in sleep structure. This means that as long as the dream is functionally effective (sufficient) and is able to restore search activity, REM sleep duration is not affected. It is increased only in the case of clinical disorders like depression, when dreams lose their restorative capacity.24

As I have stressed above, according to the search activity concept, renunciation of search is a common basic mechanism of learned helplessness and depression, and it may explain why learned helplessness was used as a model of depression.38 Passivity, lack of initiative, giving up in front of life challenges are the core symptoms of depression, and they correspond to the absence of search activity. At the same time, if the subject gives up, his/her self-esteem is going down, and negative beliefs about self-ongoing experience and future dominate. In such state it is difficult to enjoy life in all its aspects (food, sex, etc.), and motivations became blunted and flat. Sadness and anxiety are the natural components of this state mirroring for the subject all the above-mentioned features.

Thus, a patient in the depressive state requires the restoration of search activity in functionally efficient dreams. However, depression is characterized by the alteration of dream content and REM sleep physiology, which suggest the functional insufficiency of REM sleep.

Depression is also characterized by the low level of dreamed activity, decreased dream recall and short dream reports.51'52 A passive position of the dreamer in depression display itself in the domination of masochistic dream type.50 Greenberg53 proposed that dreams of depressed patients, in contrast to dreams of healthy subjects, are maladaptive, and this proposition is in a good agreement with some recent investigations. Thus, healthy subjects display a regular increase of eye movement (EM) density from the first to the last REM sleep period, while EM density correlates in healthy subjects with dream content54 - in particular, with subjects active participation in his own dreams.55 This normal dynamic of dream activity from the first to the last cycle may relate to the mood restoration during night sleep in healthy subjects.50 However, in depression eye movement density has usually an opposite dynamic - a decrease from the first to the subsequent REM periods.56 This dynamic of EM corresponds to the change of mood the latter being usually worse in the morning than in the evening. It was shown57 that when mood was subjectively estimated as being better in the morning than in the previous evening (20% of all nights) EM density increased from the first to the fourth cycle. In all other nights EM density was slightly higher in the first than in the subsequent cycles. In depressed patients subjective estimation of sleep latency correlated with EM in the first cycle, and subjective estimation of the number of awakenings correlated with the total EM scores.58 Thus, it is possible to suggest that in depression psychic activity in REM sleep is often not perceived as dream mentation, but is considered subjectively as wakefulness. It is an additional sign of functional deficiency of dream mentation in depression.

REM SD BREAKS A DEPRESSOGENIC VICIOUS CIRCLE - AN INTEGRATIVE HYPOTHESIS

If REM sleep is functionally deficient, if dreams do not restore search activity and can even increase renunciation of search (when subject is passive and helpless in his dreams), then REM SD may play a positive role breaking the following vicious circle: renunciation of search during wakefulness -giving up in dream scenario - further renunciation of search in the subsequent wakefulness. Functionally insufficient, depressogenic dreams may correspond to the sleep associated depressogenic process proposed by Wu and Bunney.19 Selective REM deprivation for the elimination of depression does not show an immediate effect typical for total SD, and requires longer time and many procedures for the following reasons:

1. During the initial period of selective REM deprivation dream activity is partly displaced from REM sleep to NREM sleep.22 As a result, selective REM deprivation is less effective than total SD in suppressing dream mentation.

2. As it was stressed previously, total SD produces a continuous wakefulness combined with a strong challenge for a person, and this combination may abolish depression.

In sum, the main hypothesis of the present review is that search activity concept provides an opportunity for integrating both factors (prolonged active wakefulness and REM SD) presumably contributing to the antidepressant effect of SD. Active wakefulness combined with a strong challenge (a task not to sleep despite high sleepiness in the night time) contains search activity, while suppression of the maladaptive dreams prevents subject from turning back to the renunciation of search.

Our approach to the problem can help to solve also some other contradictions. It is well known that depressed patients complain on the early morning awakenings, and this wakefulness in the early morning hours, while patient is still in the bed, is accompanied by especially harmful emotional feelings. At the same time, partial SD as a treatment is more effective when it is performed in the second part of the night, in comparison to the first part. By taking into consideration that in depression psychic activity in REM sleep is often estimated as wakefulness, it is possible to suggest that in early morning patients are often in the mixed state (wakefulness is mixed with functionally insufficient REM sleep). This insufficient REM sleep, not wakefulness by itself is responsible for negative emotional feelings. In this condition total awakening caused by SD is beneficial.

SD is limited in time for a natural reason. During SD, NREM and SWS are also deprived and such deprivation causes a strong sleep pressure (a pressure of S process).59 It can explain a strange combination of mood restoration and increased energy (antidepressant effect of SD) with increased subjective tiredness and sleepiness (effect of SWS deprivation60'61). If the subject is going to sleep in the morning or in the first half of the day, the positive effect of SD is abolished for two main reasons:

1. The first part of the day is naturally predisposed for the active interrelationship with the environment. It is a period when according to the biological circadian rhythm subject's activity in normal subjects is high. In depressed patients, before SD the activity in this part of the day is low, and according to my opinion not due to the shifted biological rhythm, but to the negative influence of the functionally insufficient dreams. If the real reason of the decreased mood and activity in the morning is the shifted circadian rhythm, then selective REM deprivation would not cause the restoration of mood, because this deprivation leaves circadian rhythm intact. In depressed patients not treated by SD this negative influence of dream experience is especially strong in the morning hours where the time distance from the previous dream experience is very small. Subjects slowly overcome this influence in the process of waking activity.

2. REM sleep and dream-like mental activity are more prominent in early morning hours according to the ultradian rhythm. Thus sleep in this part of the day opens gate for REM sleep and/or dream mentation including also dream mentation in NREM sleep. On other hand, sleep in the afternoon is usually characterized by the domination of NREM (predominantly by SWS), while dream mentation is less prominent in this stage in the afternoon. If total SD was performed in previous night the pressure of SWS became especially high in the afternoon and sleep in this period contains a high proportion of SWS. It may be a reason why relapse of depression was prevented when SD was followed by an acute 6-h phase advance of sleep.14 The period between 17.00 and 00.00 h is the only period in circadian rhythm relatively free from REM sleep and dream mentation. It is worth stressing that in healthy subjects with a functionally sufficient REM s spontaneous sleep in the afternoon is often followed by dysphoria - maybe because the functionally sufficient REM sleep is absent in this sleep.

The accumulation of SWS debt in the process of SD explains why it is impossible to use a long-lasting SD. I my point of view, SD has a positive effect only if it break vicious circle of renunciation of search and restores search activity. However, sleepiness caused by SWS deprivation during long-lasting total SD makes search behavior unavailable. For this reason REM SD produces a more stable positive outcome.

This approach based on the search activity concept also helps to explain some other paradoxes and surprising findings. It was shown62 that a dopamine reuptake inhibitor prevents the antidepressant effect of repeated SD. However it was shown recently that dopamine is a very important biochemical substrate of forebrain activity in REM sleep, activity that displays itself in dream mentation.63 It is reasonable to suggest that dopamine is a part of the "oil of search activity in dreams.37 If brain dopamine level and search activity in dreams are restored by mean of dopamine reuptake inhibitor, than it seems unreasonable to perform REM and SD in order to treat depression. In this condition, SD and dopamine reuptake inhibitors are "neutralizing" each other: dopamine reuptake inhibitors restore REM sleep related dream activity while SD suppress this restored compensatory mechanism.

Thus, in summary, our general conclusion is that the antidepressant effect of SD is based on two main omponents: (1) deprivation of functionally insufficient REM sleep; (2) active behavior during wakefulness and a successful confrontation with a challenge - with natural sleepiness.

ACKNOWLEDGEMENTS

My sincerest thanks to Dr Jim Fisch for valuable discussion of the topic and for personal assistance.

KEY POINTS

REM sleep is functionally deficient in depression.

REM SD, either selective or not, breaks the vicious cycle of helplessness.

Active wakefulness in the process of SD provides patient with a mobilizing challenge, which contributes to the antidepressant effect.

REFERENCES

1. Wirz-Justice A, Van den Hoofdakker RH (1999) Sleep deprivation in depression: What do we know, where we go? Biol Psychiatry 46:445-53.

2. KnowlessJB, Southmayd SE, Delva N et al (1979) Five variations of sleep deprivation in a depressed women. Br J Psychiatry 135:403-10.

3. Van den Hoofdakker RH (1990) Mechanisms in the clinical effect of sleep deprivation. In: CVE Syllabus and Scientific Proceedings in Summary Form, 143rd Annual Meeting of the American Psychiatric Association. Washington, DC: .

4. Haug HJ (1992) Prediction of sleep deprivation outcome by diurnal variation of mood. Biol. Psychiatry 31: 2718.

5. Ebert D, Berger M (1998) Neurobiological similarities in anti-depressant sleep deprivation and psychostimulant use: a psychostimulant theory of antidepressant sleep deprivation. Psychopharmacology 140: 1-10.

6. Mendels J, Cochrane (1968) The nosology of depression: the endogenous-reactive concept. Am J Psychiatry 124: 1-11.

7. Bouhuys AL, Van den Burg W, Van den Hoofdakker RH (1995)The relationship between tiredness prior to sleep deprivation and the antidepressant response to sleep deprivation in depression. Biol Psychiatry 37: 457-61.

8. Szuba MP, Baxter LR, Fairbanks LA et al (1991) Effect of partial sleep deprivation on the diurnal variation and the antidepressant response to sleep deprivation in depression. Biological Psychiatry30: 817-29.

9. Rotenberg VS, Hadjez J, Kimhi R et al (1997) First night effect in depression: New data and a new approach. Biol Psychiatry 42:267-74.

10. Van den Burg W, Beersma DGM, Bouhuys AL, Van denHoofdakker RH (1992) Self-rated arousal concurrent with the antidepressant response to total sleep deprivation of patients with major depressive disorder: A disinhibition hypothesis. J Sleep Res1: 211-22.

11. Kayumov L, Rotenberg V, Buttoo et al (2000) Interrelationships between nocturnal sleep, daytime alertness and sleepiness: Two types of alertness proposed. J Neuropsychiatry Gin Neurosd 12:86-90.

12. Rotenberg VS, Kayumov L, Indursky P et al (1997) REM sleep in depressed patients: different attempts to achieve adaptation. J. Psychosomatic Res 42: 565-75.

13. Thase ME, Buysse DJ, Frank E et al (1997) Which depressed patients will respond to interpersonal psychotherapy? The role of abnormal sleep profiles. Am J Psychiatry 154: 502-9.

14. Berger M, Vollman J, Hohagen F et al (1997) Sleep deprivation combined with consecutive sleep phase advance as a fast-acting therapy in depression: An open pilot trial in medicated and unmedicated patients. Am J Psychiatry 54: 870-2.

15. Riemann D, Hohagen F, Konig A et al (1996) Advanced versus normal sleep timing: Effects of depressed mood after response to sleep deprivation in patients with a major depressive disorder. J. Affect. Disord 37: 121-8.

16. Elsenga S, Van den Hoofdakker RH, Dols LCW (1995) Early and late partial sleep deprivation in depression. In: Stefanis C, SoldatosC, Rabavilas A (eds) Psychiatry: A world perspective, vol. 2.Amsterdam: Excerpta Medica: 374-9.

17. Vogel GW, Vogel F, McAbee RS, Thurmond AJ (1980) Improvement of depression by REM sleep deprivation: New findings and a theory. Archives of General Psychiatry 37: 247-53.

18. Wiegand M, Berger M, Zulley J (1987) The influence of daytime naps on the therapeutic effect of sleep deprivation. Biol Psychiatry22: 386-9.

19. Wu JC, Bunney WE (1990) The biological basis of an antidepressant response to sleep deprivation and relapse: Review and hypothesis. Am J Psychiatry 147: 14-21.

20. Koukkou M, Lehmann D (1993) A model of dreaming and its functional significance: The state-shift hypothesis. In: Moffitt A, Kramer M, Hoffmann R (eds) The function of dreaming. State University of New York Press: 51-118.

21. Nielsen TA (2000) A review of mentation in REM and NREMsleep: "Covert" REM sleep as a possible reconciliation of two opposing models. Behav Brain Sci 23: 85166.

22. Cartwright RD, Monroe LJ, Palmer (1967) Individual differences in response to REM deprivation. Arch Gen Psychiatry 16: 297-302.

23. Rotenberg VS (1984) Search activity in the context of psychosomatic disturbances, of brain monoamines and REM sleep functions. Pavlovian J Bio! Sci 19: 1-15.

24. Rotenberg VS, Arshavsky W (1979) Search activity and its impact on experimental and clinical pathology. Activitas Nervosa Superior(Praha) 21: 105-15.

25. Rotenberg VS (1993) REM sleep and dreams as mechanisms of search activity recovery. In: Moffitt A, Kramer M, Hoffmann R(eds) Functions of dreaming. Albany: State University of New York Press: 261-92.

26. Venger A, Rotenberg VS, Desyatnikov Y (1996) Evaluation of search activity and other behavioral attitudes in indefinite situation. Dynamische Psychiatrie/Dynamic Psychiatry 160/161:368-77.

27. Rotenberg VS, Kutsay S, Venger A (1998) Behavioral attitudes and distress in adolescents: Relationship to age and gender. Homeostasis 39: 57-64.

28. Rotenberg VS, Kutsay S, Venger A (2001) Behavioral attitudes and the level of distress in the process of adaptation to the new society. Stress Health 17: 187-93.

29. Rotenberg VS (2001) Variations and active versus reactive behavior as factors of the selection process. Behavior Brain Sci24: 553-4.

30. Rotenberg VS (2000) Anorexia nervosa: Old contradictions and a new theoretical approach. (Guest editorial). Int ] Psychiatry clin. Pract 4: 89-92.

31. Rotenberg VS (2000) Search activity: A key to resolving contra-dictions in sleep/dream investigation. Behav Brain Sci 23: 996-9.

32. Rotenberg VS, Alexeyev W (1981) Essential hypertension: A psychosomatic feature or a psycho-somatic disease? A differential analysis of cases in terms of search activity concept. DynamischePsychiatrie/Dynamic Psychiatry 68: 12940.

33. Rotenberg VS, Boucsein W (1993) Adaptive vs. maladaptive emotional tension. Genet Soc Gen Psychol Monogr 119: 207-32.

34. Rotenberg VS, Sirota P, Elizur A (1996) Psychoneuroimmunology: Searching for the main deteriorating psychobehavioral factor. Genet Soc Gen Psychol Monogr 122: 329-46.

35. Rotenberg VS, Korosteleva IS (1990) Psychological aspects of the search activity and learned helplessness in psychosomatic patients and healthy testees. Dynamische Psychiatric/Dynamic Psychiatry120: 1-13.

36. Engel G, Schmale A (1967) Psychoanalytic theory of somatic disorders. J Am Psychoanal Assoc 15: 344-65.

37. Rotenberg VS (1994) The revised monoamine hypothesis: Mechanism of antidepressant treatment in the context of behavior. Integral Physiol Behav Sci 29: 182-8.

38. Seligman MEP (1975) Helplessness: on depression, development and health. San Francisco: Freeman.

39. Anisman H, Zacharko RM (1982) Depression: The predisposing influence of stress. Behavior Brain Sci 5: 89-138.

40. Wu JC, Buchsbaum MS, Gillin JC et al (1999) Metabolic rate in ventral and anterior cingulated and medial prefrontal cortex predict antidepressant effects of sleep deprivation. Am J Psychiatry156: 1149-58.

41. Cohen H, Edelman A, Bower R, Delmont W (1975) Sleep and self-stimulation in the rat. Abstracts of the llth Annual Meeting of theAPSS, New York.

42. Putkonen P, Putkonen A (1971) Suppression of paradoxical sleep following hypothalamic defense reactions in cats during normal conditions and recovery from PS deprivation. Brain Res 26: 334 -47.

43. Adrien J, Dugovic C, Martin P (1991) Sleep wakefulness patterns in the helpless rat. Physiol Behav 49: 257-62.

44. Rotenberg VS (1996) Learned helplessness and sleep: Discussion of contradictions. Homeostasis 37: 89-92.

45. Reynolds III CF, Kupfer D (1988) Sleep in depression. In: WilliamsR, Karacan I, Moore (eds) Sleep disorders, diagnosis and treatment. New York: J. Wiley: 147-64.

46. Oniani TN, Lortkipanidze ND (1985) Effect of paradoxical sleep deprivation on the learning and memory. In: Oniani TN (ed.) Neurophysiology of motivation, memory and sleepwakefulness cycle,Tbilisi: Metzniereba: 4: 214-34.

47. Oniani TN, Lortkipanidze ND, Mgaloblishvili MM et al (1988)Neurophysiological analysis of paradoxical sleep deprivation. In:Oniani TN (ed.) Neurobiology of sleep-wakefulness cycle. Tbilisi:Metzmereba: 19-42.

48. Kuiken D, Sikora S (1993) The impact of dreams on waking thoughts and feelings. In: Moffut A, Kramer M, Hoffmann R (eds)The functions of dreaming. Albany: State University of New York Press: 419-76.

49. Morrison A (1982) Central activity states: Overview. In: BeckmanAL (ed.) The neural basis of behavior. New York: Spectrum: 3-17.

50. Kramer M (1993) The selective mood regulatory function of dreaming: An update and revision. In: Moffitt A, Kramer M, Hoffmann R (eds) The functions of dreaming. Albany: State University of New York Press: 139-96.

51. Armitage R, Rochlen A, Fitch Th et al (1995) Dream recall and major depression: A preliminary report. Dreaming 5: 189-98.

52. Riemann D, Low H, Shredil M et al (1990) Investigation of morning and laboratory dream recall and content in depressive treatment with trimipramine. Psychiatr J Univ Ottawa 15: 93-9.

53. Greenberg R (1977) On understanding sleep disorders and their psychopathology. McLean Hasp J II 3: 139-46.

54. Hong Ch, Potkin SG, Antrobus JS, et al (1997) REM sleep eye movement counts correlate with visual imagery in dreaming: a pilot study., Psychophysiology, 14: 377-81.

55. Rotenberg VS (1988) Functional deficiency of REM sleep and its role in the pathogenesis of neurotic and psychosomatic disturbances. Pavlovian ] Biol Sci 23: 13.

56. Benson EZ, Zarcone VP (1993) Rapid eye movement sleep in schizophrenia and depression. Arch Gen Psychiatry 50: 477-82.

57. Indursky P, Rotenberg VS (1998) The change of mood during sleep and REM sleep variables. Int J Psychiatry Clin Pract 1: 47-51.

58. Rotenberg VS, Indursky P, Kayumov L et al (2000) The relationship between subjective sleep estimation and objective sleep variables in depressed patients. Int J Psychophysiol 4: 63-7.

59. Borbely AA, Wirz-Justice A (1982) Sleep, sleep deprivation and depression. Hum Neurobiol 1: 205-10.

60. Agnew HW, Webb WB, Williams RL (1967) Comparison of stage four and 1-REM sleep deprivation. Percept Mot Skills 24: 851-8.

61. Rotenberg VS (1991) The competition between SWS and REM sleep as index of maladaptation to shift work. Homeostasis 33:235-8.

62. Benedetti F, Barbini B, Campon E et al (1996) Dopamine agonist amineptine prevents the antidepressant effect of sleep deprivation.Psychiatry Res 65: 179-84.

63. Solms M (2000) Dreaming and REM sleep are controlled by different brain mechanisms. Behav Brain Sci 23: 843-850.