A THEORETICAL BACKGROUND - SEARCH ACTIVITY CONCEPT
Numerous investigations on sleep psychophysiology in psychiatry have provided science with interesting empirical data. However, there are many contradictions in these data and the most important and substantial questions are still far from being solved. The only way to overcome these contradictions and to elucidate the role of sleep alteration in the general pathogenesis of mental disorders is to develop an integrative theory of the basic sleep functions, which are equally relevant for humans and animals, for psychology and physiology, for mental health and for psychopathology. Obviously, this is a difficult task. It would be too optimistic to expect that such an integrative theory would appear at once in its final and comprehensive version. Thus, it is worth producing different versions of such a theory for broad discussion, to provide in the future a holistic picture from all of these puzzles. The task of the present chapter is to use one such theory (search activity concept) for the revision of some contradictions and for the discussion of some theoretical assumptions.
The search activity concept 1-5 deals mostly with the function of rapid eye movement (REM) sleep, although the alteration of NREM sleep seems also to be very important for understanding the mechanisms of mental disorders. However, NREM sleep deficiency is an unambiguous and general finding in many psychiatric patients with different diagnoses, while data on REM sleep are much more contradictory. The discussion of the nature of such contradictions may be very productive.
Search activity is defined as "activity that is oriented to changing the situation (or at least the subject's attitude to it) in the absence of the precise prediction of the outcome of such activity 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 between the behavior and its outcome enabling the subject to correct his behavior according to the outcome.
Many forms of behavior do not include search activity. Renunciation of search is opposite to search activity and encompasses depression and neurotic anxiety in humans, freezing in animals and learned helplessness in both species. Neurotic anxiety, according to psychodynamic conceptions, is a consequence of the repression of ah unacceptable motive from consciousness. The repression can be regarded as a purely human variant of renunciation of search, of the modes of realization of the unacceptable motive in behavior, and modes of integration of this motive with other, realized, behavioral orientations. The difference between productive emotional tension (the normal anxiety of a healthy individual in a state of stress), which is instrumental in mobilizing all psychological and physical resources to overcome obstacles, and unproductive emotional tension, which hampers successful activity, is determined by the presence or absence of search activity in the structure of emotional tension6.
Panic reaction can imitate search activity, but a subject in a state of panic does not consider the real outcome of his behavior and is unable to correct it. The common clinical condition of panic is depression. Stereotyped behavior is based on the precise prediction of the outcome and is not flexible; thus it also does not fit with the idea of search activity.
I have proposed this new classification of behavior because our investigations have shown that search activity has a very important psychobiological meaning 1;6-13. All forms of behavior that include search activity, but differ in other respects, increase body resistance, while absence of search decreases body resistance and predisposes a person to mental and psychosomatic disorders. Persons with psychosomatic illness have a more pronounced tendency to give up than do healthy subjects, especially in stressful conditions and after facing failure. We concluded that the process of search activity in itself, independent of the pragmatic results of such activity (i.e. whether it is successful or not) and independent of the accompanying emotions, protects persons from somatic disorders.
Herein lies the basic difference between the concept of search activity and the concept of coping behavior 14 that must be successful in order to be protective.
The protection of search behavior reflects a particular self-service and a positive feedback: search activity requires a lot of effort and stimulates a person to enter unpredictable and potentially dangerous situations, and prevents exhaustion in such situations. All details of the investigations performed on animals and humans as well as all aspects of the theoretical discussion can be found in the above-mentioned publications.
The biochemical mechanisms of search activity have been insufficiently studied but it is possible to suggest that the brain's monoamine system is closely connected with search behavior. Learned helplessness accompanied by somatic disturbances emerges when the brain monoamine level drops 15. An artificial reduction of brain catecholamine levels by tetrabenazine speeds passive behavior whereas the prevention of its depletion by monoamine oxidase (MAO) inhibitors raises stress resistance, restoring the animal's ability for an active reaction to stress 16. In animals that cannot control the stressful situation the brain level of norepinephrine drops particularly low and they show the greatest distress 17,18. Finally, the locus ceruleus plays a major role in the organization of various forms of search behavior 19.
On the basis of the preceding facts the following hypothesis has been advanced 8-20. Search activity can begin in the presence of a certain critical level of the brain monoamines that are utilized in the course of search. Search activity itself, once it starts, stimulates the synthesis of the brain monoamines and ensures that they remain at adequate, or even excessive, levels. The task of such a system with positive feedback (brain monoamines -behavior - brain monoamines) is to support search activity.
At the same time, a stable high level of brain monoamines causes a hyposensitivity of the inhibitory alpha2-adrenoreceptors. As a result, in order to switch on the negative feedback that regulates homeostasis of the brain monoamines, its level has to be excessively high.
In a state of renunciation of search the above-mentioned positive feedback between
brain monoamines and behavior does not function. Moreover, in this state, which manifests itself in depression and maladaptive emotional tension, "the monoamine expenditure climbs. The requirement for brain monoamine production and expenditure decreases because the behavior that requires brain monoamines is absent. At the same time, the low level of brain monoamines by itself does not stimulate its own production. The brain monoamine system is only partly homeostatic. If the appropriate behavior (search activity) is absent, activation of the brain monoamine system is not required. When the level of brain monoamines is constantly low the inhibitory oc2-adrenoreceptors become adapted to it and maintain this low level. If search activity is of such great biological relevance, and the absence of search, and especially renunciation of search in stressful conditions, leads to illness and death, why has the latter regressive behavior survived in the course of evolution and what makes it emerge?
All higher mammals, including humans, at the early stages of ontogenesis inevitably experience helplessness determined by the relatively slow development of the central nervous system and all mechanisms (nervous, hormonal and vegetative) that ensure subsequent search behavior. Naturally, during the early stages of ontogenesis such a state cannot be called renunciation of search. It is normal and inevitable and the only accessible form of a defense reaction for an immature organism. Nevertheless, the organism thus acquires an early experience of passive reaction, an experience of helplessness. Both this experience and the ways to overcome it are of colossal relevance to the individual's entire subsequent life. In the case of the correct attitude of the primary group, above all the mother, this early experience of helplessness can be successfully and painlessly overcome. However, all injuries of early childhood, from physical separation from the mother to insufficient emotional contact with her and the feeling of insufficient protection as a result of strained relations between the parents, can consolidate the experience of primary helplessness and produce a tendency to give up. The imprinting mechanism may lead to the development of a state of renunciation of search in an adult, especially if the emotional problems encountered in some way resemble the conflicts of early childhood. Thus, readiness for the development of mental and psychosomatic illnesses is formed. It is exactly from these general biological positions, the relations between early helplessness and the formation of search activity, that the present author proposes considering the Freudian theory of the role of early psycho-traumatizing situations in the entire subsequent development of the individual. Regressive behavior in the case of neuroses and psychosomatic illnesses, according to Freud, is indeed a regression towards a biologically earlier state of helplessness that assumes a form of renunciation of search.
However, if search activity is so important for survival, and renunciation of search is so destructive and dangerous for health, it would be reasonable to assume a natural brain mechanism that can restore search activity after temporary renunciation of search. I believe that dreams in REM sleep fulfill this function. A covert search activity in dreams' compensates for the renunciation of search in previous wakefulness and ensures the resumption of search activity in subsequent wakeful-ness. I base this claim on the following findings:
(1) Different forms of animal behavior that contain search activity (self-stimulation, fighting) suppress REM sleep in subsequent sleep without a restorative rebound effect 21-22. This means that less REM sleep is required after such overt behavior.
(2) Renunciation of search produced by the direct stimulation of the hypotha-lamus causes an increase in REM in subsequent sleep 1. This corresponds to data 23 that REM sleep increases along with a moderate reduction of norepinephrine system activity, while the excessive decrease or increase of this activity suppresses REM sleep (see discussion in references 12, 20). This means also that search activity can start in REMS sleep with a lower level of brain monoamines than is required in wakefulness. This conclusion corresponds to data 25 that in REM sleep alpha2-adrenoreceptors are much more sensitive than in wakefulness; thus the fine regulation of brain monoamines can be performed on their lower level.
(3) If, during REM deprivation, the subject is involved in active behavior (exploration, active defense reaction), the REM rebound effect during subsequent sleep is substantially reduced 26.
(4) Depression, neurotic anxiety and generalized learned helplessness as manifestations of renunciation of search are accompanied by increased REM sleep requirement (REM sleep latency is decreased while the duration of the initial REM episodes is increased) 27-29. A correlation was detected between learned helplessness and REM sleep percentage 30-31.
(5) Both REM sleep and search activity in wakefulness are characterized by a regular and synchronized hippocampal theta-rhythm. Moreover, the more pronounced the theta-rhythm in wakefulness, the less pronounced it is in subsequent REM sleep 26. REM sleep in animals regularly contains pontogeniculo-occipital (PGO) spikes, which are electroencephalographic (EEG) signs of neural activation. In wakefulness, PGO correlates with orienting activity 32. The presence of the PGO spikes in REM sleep may mean that the animal is predisposed to react to novel stimuli, including spontaneous changes of dream content.
(6) When a special part of the nucleus ceruleus (nucleus ceruleus aleph) in the brainstem 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 orienting (or search) activity 33.
(7) The same psychological variables that predispose the subject to renunciation of search - trait anxiety, fixation on obstacles according to Rosenzweig 34, motivation of avoidance of failures - have a tendency to correlate with REM percentage 8.
(8) REM sleep is increased in students who display a stable emotional tension both before and after an examination, even with its positive outcome (when the stress situation no longer exists) as compared with students who display emotional tension only before an examination 2. It is possible to suggest that this stable emotional tension, which is not compatible with the actual situation, is similar to neurotic anxiety and represents renunciation of search.
(9) Healthy subjects with normal search activity during wakefulness are characterized by active participation in their own dreams. This active participation correlates with heart rate acceleration and eye movement density in REM sleep 35.
(10) 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 36. It can be suggested that the characters' activity represents search activity in dreams. One must stress that, in the case of healthy subjects, the decrease of unhappiness after sleep correlates only with the above-mentioned dream variables, not with the sleep physiology. As I see it, this means that, as long as the dream is functionally effective and is able to restore search activity, REM sleep requirement does not increase. It is increased only when dreams are going to lose their restorative capacity, for instance in clinical or pre-clinical disorders (depression and subsyndromal depression, neurotic anxiety, narcolepsy).
The dreams of healthy subjects represent a very specific kind of search activity, which, however, is compatible with the above-mentioned notion of search activity: the subject is active in his dreams but is unable to make a definite forecast according to the outcome of dream events. The dream does not apparently fit solely those parts of the notion emphasizing the permanent consideration of the behavioral outcome. This is true only if the dream is not self-reflective and does not include self-control, i.e. if activity in the dream is a chaotic activity. However, it was demonstrated 37 that, in the vast majority of spontaneous dreams, the dreamer is moderately self-reflective and effective in dream control.
Dreams provide a good opportunity for search activity after giving up. First, the subject is separated from reality while sleeping, including those aspects of reality that caused renunciation of search. Thus, in dreams a person is free to start from the beginning. Second, within his dream, the dreamer can be very flexible in his behavior: it may be an attempt to solve an actual problem in a metaphoric manner, or it can be an attempt to solve another problem, one that displaces the actual problem 38, since not the topic of search, but the search process itself is the main restorative factor. Moreover, in dreams, one is not restricted by logical and conventional rules while manipulating the problems. One is able to use image thinking, which is polysemantic by nature and as a result much more flexible than logical thinking and, unlike the waking consciousness, can avoid contradictions 5,20,24,39,40. Image thinking is free from the probability forecast 41,42. Since we assume that the final aim of the dream work is not the real solution of the actual problem but only the restoration of search activity, all the above features contribute substantially to the restoration.
The proposed explanation of the functional significance of the REM sleep in the restoration of search activity helps us to understand a very well-known phenomenon in human sleep, the "first-night effect" (FNE). It is an alteration of sleep structure observed on the first night of the multinight sleep study relative to subsequent nights 43,44. Compared with later nights, the first night in the laboratory is characterized by increased REM sleep latency, combined with a moderate REM sleep reduction, especially in the first part of the night, usually without the subsequent REM sleep rebound. Sleep latency is also often increased and sleep stages shift, while sleep efficiency is decreased. FNE reflects the natural increase of the sleeper's state of vigilance in an unfamiliar sleeping environment as part of the adaptive orientation in the new situation 45, and is much less prominent in healthy subjects sleeping at home where they are reacting only to the monitoring equipment 44,46. Such adaptive orientation includes search activity (orienting behavior, as I have stressed, is a manifestation of search activity); thus according to the search activity concept, the requirement of REM sleep in the night sleep has to be decreased. That is exactly the case. It means that the FNE can be used for the estimation (even quantitative estimation) of the subject's ability to display search activity.
Another theoretical outcome of the search activity concept is the conclusion that REM, sleep may be efficient or inefficient according to its main function - restoration of search activity 5,35. If dreams do not contain a covert search activity, and moreover, if it is a continuation of the renunciation of search in dreams, it is possible to suggest REM sleep as inefficient. REM sleep inefficiency displays itself not only in the stable feeling of helplessness and giving up in dreams, in the passive position of the dreamer facing the dangerous events in the dream, but also in the decrease of dream reports after awakenings in REM sleep and in the impoverishment of these reports (fewer images and events). Such features characterize, for instance, dreams of depressed patients 38,47,48. As mentioned before, there are some physiological correlations of the subjects' imaginative activity in their own dreams. Thus, it is possible to use such physiological variables for the estimation of REM sleep functional activity even without awakenings.
In the following sections of this chapter I will discuss alterations of sleep psychophysiology in different mental disorders using the frame of the search theory.
POST-TRAUMATIC STRESS DISORDERS
Complaints about sleep disorders - difficulty falling or staying asleep, increased number of spontaneous awakenings, recurrent distressing dreams (nightmares) - belong to the main diagnostic features of post-traumatic stress disorders (PTSDs). In most cases these complaints are confirmed by polysomnography: the sleep of PTSD patients is characterized by disturbances in sleep continuity, including increased sleep latency, increased number of awakenings, especially awakenings in REM sleep during nightmares, and low sleep efficiency 49-54. However, data on sleep structure in PTSDs, and especially data on REM sleep variables, are very ambiguous and contradictory.
First, PTSDs are characterized by a high variability of REM sleep latency. In comparison to healthy control subjects, REM latency in PTSD patients has been reported to be either shortened or lengthened 55-60. Correspondingly, REM sleep time was either increased 61 or decreased 57. In some investigations, there was no difference in the mean REM time and REM latency between healthy subjects and combat veterans 62.
PTSDs are often accompanied by major depression (MD) or other depressive disorders such as dysthymia 54, 63, 64. Thus, it has been reasonable to assume that the difference between PTSD patients in REM sleep variables may be determined by the presence pr absence of PTSD co-morbidity with depression 54. Comparison of PTSD and MD patients has shown that REM sleep time was reduced in PTSDs; however, this reduction was presumably a side-effect of the decreased sleep duration 54. In contrast to healthy subjects, the decrease of the total sleep time in PTSD patients was not accompanied by a compensatory increase in the percentage of REM and slow-wave sleep (SWS). At the same time, REM latency does not differ significantly between both clinical groups. However, it was difficult to draw a definite conclusion from this study that PTSD patients display REM sleep latency typical for depression. First, depression in the group of MD patients was not severe, and only 47% of these patients had REM latency less than 60 min. On the other hand, 20% of PTSD patients suffered from co-morbid MD, and 40% of these patients had prior depressive episodes. Moreover, within the PTSD group there was no significant trend for REM latency to be reduced in patients with the current co-morbid depression. Thus, the suggestion of depression as a reason for the high variability of REM latency in PTSDs was neither confirmed nor refuted.
The investigation of Woodward and colleagues 65 seems to give a more definite answer to this question. Two groups of PTSD patients - one with and one without co-morbid depression - demonstrated similar mean REM sleep latencies, although the variability of REM latency was higher in patients with a co-morbid depression. Thus it is not depression itself, or at least not only depression, which determines the decrease of REM latency in some PTSD patients and causes heterogeneity of REM latency in all groups of these patients. The authors concluded that "a (unknown) factor emerges in PTSD which exerts a specific effect upon REM sleep timing and amount. ... This factor is related to the apparent increase in both tonic and phasic REM phenomena in sleep of PTSD patients".
In discussing this topic, it is also worth emphasizing that some core physiological features of PTSD are the converse of those in depression. PTSD patients in comparison to healthy subjects, and contrary to depressed patients, are characterized by the decreased basal cortisol level and by the increased negative feedback regulation of cortisol excretion (see reference 66). Patients with PTSDs demonstrate augmented cortisol suppression in response to dexamethasone, while depressed patients are characterized by non-suppression of cortisol in this test 67,68. PTSD patients have more, and depressed patients have fewer, type II glucocorticoid receptors in circulating lymphocytes than do normal controls 69,70. Platelet alpha2-adrenergic receptor binding, platelet MAO activity and platelet serotonin uptake are decreased in PTSD patients in comparison to normal subjects 71. The increased activity and reactivity of the monoamine system in PTSDs is confirmed by increased nocturnal noradrenergic excretion 53. Thus, it is possible to suggest an exaggeration of the normal adaptive physiological response to stress in PTSD - although without any real stressful condition - and this exaggerated response to stress is the opposite to what may be found in depression. This exaggeration of the normal response is confirmed in the investigation of the FNE 45. According to the REM sleep variables, and especially to REM latency, PTSD inpatients were bidirectionally sensitive to the degree of familiarity they associated with the sleep laboratory conditions. In comparison to non-ill trauma-free controls, in-patients familiar with the laboratory environment demonstrated reduced FNE (increased extinction of the orienting reaction), while patients unfamiliar with this environment demonstrated enhanced FNE. At the same time MD patients, in contrast to normal subjects, often demonstrate a reduced FNE in the unfamiliar situation (see below). Thus, PTSD symptomatology is very complicated. On the one hand, depression is a common PTSD co-morbidity and there are some factors that make REM sleep variables in isolated PTSD similar to those in depression. However, on the other hand, PTSDs differ from depression according to many physiological variables. We will show that this distinction has real roots in the clinical picture and in the pathogenesis of PTSDs.
The clinical picture of PTSDs includes, on the one hand, a persistent re-experiencing of the past traumatic events manifested as intrusive distressing recollections (images, thoughts, perceptions, dreams, dissociations like illusions, hallucinations and flashback episodes) accompanied by increased physiological reactivity on exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event. It can be assumed that persistent symptoms of increased arousal such as irritability, hypervigilance, exaggerated startle reaction, prominent orienting reaction and FNE, difficulty falling and staying asleep, although they are usually considered separately, belong to the same group of symptoms - group I.
On the other hand, the same patients (and this is worth emphasizing) display persistent avoidance of stimuli associated with the trauma and numbing of general responsive-ness (conscious avoidance and repression of thoughts, feelings, memories and activities associated with the trauma; diminished interest in any forms of activity and interpersonal relationships, as in depression; a hopeless view on the future). This group II of symptoms is obviously opposite to group I, and both groups compose a contradictory clinical picture, where symptoms of group I are continuously displaced by symptoms of group II and vice versa, or even partly appear together.
However, it is possible to integrate these different symptoms in one holistic frame. The central point of the PTSD diagnosis is the feeling of helplessness caused by the traumatic event. It is the central feature of the entire clinical picture. Fear of death and fear of ego-destruction produces helplessness because it is a very distressing and disorganizing emotion - there are no clear ways to cope with a fear of death, if the psychological defense mechanisms become insufficient 72. The feeling of helplessness is very difficult to resist. It is a state of giving up, of renunciation of search, with all negative outcomes on human behavior and health, as shown in the Introduction. By considering helplessness as a corner-stone of PTSDs, it is possible to explain all group I symptoms as attempts to cope with this feeling of helplessness and loss of control which accompanied the traumatic event. Hypervigilance, and exaggerated startle reaction, reflect the mobilization for coping. However, unfortunately, this coping is irrelevant and cannot be successful, because the object of coping is elusive: it is the traumatic event that already happened in the past, and the subject is unable to win back. In principle, the re-experience of trauma may help to re-examine the traumatic event and to include it in a more broad and polydimensional picture of the world where trauma is able to find its definite and restricted place and will no more cause general helplessness. However, for this re-examination, the subject has to be well equipped with skills of the right-hemispheric polysemantic way of thinking 39,73, and exactly these skills are underdeveloped in PTSD patients 74. In these patients traumatic experiences are fragmental, not integrated in the holistic picture of the world, and reappear in a very rigid form of primitive and frustrating images in wakefulness, in altered states of consciousness and in nightmares. Such coping is stereotypic in its nature; however, it can include a very strong physiological mobilization. The position of the subject in this state of coping can be very active, but the final goal - a freedom from helplessness and from a feeling of being a victim - cannot be achieved. It is like a fight with a shadow where you never can win. Traumatic events, instead of being coped with, restore every time the same distressing frame and do not help to overcome helplessness. It is a key point of this unsuccessful coping.
Group II symptoms reflect an attempt to avoid the negative experience with which the subject is unable to cope. To avoid does not mean to solve, but only to conserve the problem. Trauma after such avoidance remains isolated from the normal integrative function of the brain and from the larger associative network of memory. Actually, these group II symptoms display a giving-up state in front of traumatic experience after the unsuccessful coping. However, giving up only increases helplessness, and it forms a vicious circle. Coping and avoidance work in opposite ways towards the same .goals, but both are unsuccessful. The continuous oscillations between unsuccessful attempts to cope and giving up (avoiding) can explain the complicated and contradictory picture of PTSDs and their sleep alterations.
For instance, Woodward and co-workers45 proposed that helplessness and not clinical depression may be a factor related to increases in both tonic and phasic REM phenomena and reduced REM latency. Conversely, the predominance of group I symptoms (active coping) may be responsible for increased REM latency in some PTSD patients and in the exaggeration of FNE. The same patient may demonstrate different sleep structures at different time points.
By taking into consideration the role of REM sleep and dreams in balancing behavioral attitudes and restoration of search activity, it is reasonable to pay attention to the dream content. In contrast to the increased variability of REM latency and REM sleep time, eye movement (EM) density in REM sleep in PTSD patients seems to be unambiguously increased 54,62,66. According to the above-mentioned relationships between EM density and dream content 35,75, it is possible to suppose that the increase of EM density in PTSD patients correspo'ficts to their typical complaints of traumatic dreams and nightmares 59,76-79. However, direct investigation of this topic in the sleep laboratory does not confirm this proposition. According to Greenberg and associates 55, most REM awakenings in Vietnam War veterans with PTSD symptoms led to content-less reports. Kramer and colleagues 81 have also shown a lower than normal (around 50%) dream recall rate after spontaneous awakenings from REM sleep. In the investigation of the Lavie group 49,82 these findings have been substantially enlarged, and they deserve elaboration. The authors systematically collected dreams from REM sleep in well-adjusted and poorly adjusted Holocaust survivors, and compared them to the control group of aged persons without the Holocaust experience. The estimation of the level of adjustment was based on clinical interview regarding all main areas of life: marital and familial problems, problems at work, social relationships, somatic and mental problems, and general satisfaction in life. It is worth stressing that the well-adjusted group presented not only fewer somatic complaints man the poorly adjusted group, but also even fewer complaints than the control group. Members of the well-adjusted group were characterized by stronger ego-forces, in comparison to the poorly adjusted group. Their emotional feelings were positive. Recollection of the Holocaust contents occurred infrequently and was under their control.
The less well-adjusted group was disturbed by their Holocaust experience. According to the clinical picture, the less-adjusted group displayed a moderate PTSD complex, including decreased sleep efficiency in comparison to both other groups. In sleep structure, including REM sleep variables, no differences were found among the groups. (This is not very surprising if we consider the variability of sleep in PTSD patients.) The most prominent differences between groups were in dream reports and in dream content. The control group reported a normal percentage (80%) of dream content when awakened from REM sleep. The percentage of reports in the less well-adjusted group was 50.7% - equal to the percentage reported by the PTSD patients of Kramer and associates 81, and Bleich and co-workers 83. This percentage is also approximately equal to the percentage of dream recall (54.6%) presented by mentally ill patients (with depression, anxiety, hypochondria and hysterical neurotic disorders 84). At the same time, the well-adjusted group had the lowest dream rate (33.7%). Moreover, in contrast to both other groups, well-adjusted survivors were unable to realize not only the content of the dream, but also even the fact of dreaming. The survivors' dreams, particularly those of the well-adjusted group, were less complex and salient in comparison to dreams of the control group. The number of anxious dreams was high in both groups of survivors and especially high in the poorly adjusted group. However, only the less well-adjusted survivors experienced nightmares resulting in spontaneous awakenings. In dreams of the less-adjusted group the dreamers were usually the victims, while in the well-adjusted group dreams were characterized by hidden hostility in which survivors were not directly involved. This is a very important point, because according to the search activity concept a feeling of being a victim means the continuation of giving up and the renunciation of search from wakefulness into dream, and this feeling reflects the functional insufficiency of the dream. Only the well-adjusted survivors demonstrated indifferent responses toward their dream content after awakening, even if the content was stressful. Similar to controls, and in contrast to the less well-adjusted survivors, the well-adjusted subjects dreamed about their present life rather than about the past.
Dreams of both groups of survivors contained more danger to the dreamer, more anxiety and hostility, and more interpersonal conflicts than dreams of the control group, although all these features were more prominent in the less-adjusted group. By taking into consideration this similarity, and the relative similarity in the low number of dream reports, Kaminer and Lavie 49 concluded that both groups, and especially the well-adjusted one, were characterized by the repression of dreams, and the stronger the repression, the better was the psychological adjustment. However, this explanation looks very doubtful. First, it has already been stressed that according to many important features (such as general physical and mental health, sleep efficiency, psychological response to the negative dream content during sleep and after awakening, position in dream - victim versus non- victim) the well-adjusted group of survivors is closer to the healthy control group than to the poorly adjusted group, while according to the number of dream reports it is most different from the control group. Second, it is very difficult to accept the statement that repression is a healthy defense mechanism. As has been shown previously, repression is a very specific human form of renunciation of search, according to the psychodynamic concept of neurotic free-floating -anxiety. We have shown 85 that in psychologically maladapted somatic and psychosomatic patients there is a positive correlation between repression (measured according to Plutchik and associates 86) and the D scale of the Minnesota Multiple Personality Inventory (MMPI), which measures depression and neurotic anxiety. This correlation was absent in psychologically adapted subjects; thus it is possible to suggest that high levels of repression caused increased anxiety. If anxiety required repression for its compensation, such positive correlation would be present in psychologically adapted subjects. In addition, the high strain of repression characterized essential hypertension. This means that repression does not belong to the most adaptive defense mechanisms. Moreover, repression increases together with anxiety after REM/dream deprivation 80,87. This means that in normal subjects dreams are used for the compensation of repression and for the prevention of anxiety, and it would be very paradoxical to assume that the repression of dreams serves psychological health. It is possible to suggest repression of the functionally insufficient dreams to explain the lack of dream reports in the poorly adjusted group of PTSD patients; however, it looks very doubtful that repression can serve psychological health and is used by well-adjusted survivors.
The search activity concept provides another explanation of the extremely low number of dreams in the well-adjusted group. Members of this group appear to have a very strong search activity. They are usually successful and active business people, with high social achievements 49; they cope with problematic situations well, perhaps better than people without the Holocaust experience. This resembles hyper-compensatory activity and can explain their normal health as well as a decreased dream requirement. Thus, the number of dream reports is decreased in this group not owing to the dream repression but to the high search activity during wakefulness. However, in the less-adjusted group the low number of dream reports reflects suppression of the functionally insufficient REM dreams 35. Thus, dreams in FTSDs are either anxious and hostile (coping) or repressed (avoidance of experience).
DEPRESSION
Sleep in depression has been investigated in more detail than sleep in any other psychiatric disorder and there is a general consensus regarding the sleep structure in depression. Slow-wave sleep (SWS) deficiency, especially in adults; decreased sleep duration caused by the increased sleep latency; awakenings during the night sleep and early morning awakenings; decreased REM sleep latency and the redistribution of REM sleep with its concentration in the first half of the night; increased EM density in the first cycle -various combinations of these features characterize the sleep of depressed patients 64,88-91. However, what these alterations mean and the cause-and-effect relationship between them and other clinical symptoms of depression is still an open question.
There is contradiction in the literature regarding REM sleep variables: they have been claimed to normalize during remission 92,93, but this finding has not been confirmed in other investigations 94,95. Thus, the alteration in REM sleep variables may be a trait or state marker of psychopathology 96-97. REM sleep deprivation was used as an effective treatment for depression 98; thus REM sleep may even play a role in the pathogenesis of depression. Alternatively, REM sleep redistribution may reflect either an altered circadian rhythm 99 or an attempt by the brain to compensate for depression. This latter proposition appears to be more explanatory and ties in with theories concerning the adaptive function of REM sleep in healthy subjects, especially in the domain of mood 28,36,100. The search activity concept also fits this proposition because depression is a typical state of renunciation of search that requires search activity in dreams for compensation and for the restoration of overt search behavior in the subsequent wakefulness 5. According to this point of view, the decreased REM latency and increased REM duration and EM density in the first sleep cycle reflects an increased requirement in REM sleep. This increased REM sleep pressure in depression was confirmed in our recent investigation 101: while in normal subjects and in schizophrenic patients the incorporation of wakefulness in the first sleep cycle increases REM latency in proportion to the incorporated wakefulness, it does not happen in depression. The absence of the FNE, according to REM sleep variables, in any patients with severe depression 102-105 can be considered not only as an outcome of the diminished orienting reaction in depression, but also as a sign of the increased REM pressure. In our investigations 106,107, FNE was present only in 44% of patients with MD. According to our criteria, FNE was present if REM sleep latency on the first night of study was at least 30 min longer than REM sleep latency on the second night. FNE was absent in most patients with mood-congruent psychotic features (delusional depression) highly resistant to the antidepressant drugs and with more previous hospitalizations (as was shown also by Ansseau and co-workers 89), and clinical improvement in these patients without FNE was achieved only after electroconvulsive therapy (ECT). Thus, absence of the FNE characterized depressed patients in a more severe clinical state. However, and especially interesting, depressed patients who exhibited FNE also demonstrated increased REM sleep pressure. The delayed first REM sleep period in the first night was increased, in comparison to the first REM sleep period in the subsequent nights of patients with FNE as well as in comparison to the first REM sleep period in all nights of patients without FNE. In depressed patients the first REM sleep period duration is usually increased, in comparison to healthy subjects, but in patients with FNE the first REM sleep period on the first night was especially high and occupied 44% of REM sleep. Moreover, REM sleep latencies on the second and third nights in patients who exhibited FNE were significantly shorter than REM sleep latency of the corresponding nights of the group without FNE. In patients with FNE total REM sleep was increased in the second night compared with the first night. In addition, patients with FNE showed an increased number of short cycles (less than 40 min) in all nights. By taking into consideration all these data, it is possible to suggest that depressed patients who display the FNE are also characterized by the high REM sleep requirement; however, they are more flexible in their physiological responses, in particular in the response of their REM sleep system, in comparison to patients without FNE. Being in an unfamiliar environmental condition, they are able to react to this condition with a FNE, like healthy subjects - by increasing the REM sleep latency. However, presumably due to the high REM sleep requirement, they demonstrate a 'rebound effect' not typical for healthy subjects: a shift of REM sleep to the first cycle and an increase of REM sleep in the subsequent nights. I cannot exclude that they are less resistant to the treatment partly owing to this flexibility of orienting and REM sleep systems.
In parallel with the increased REM sleep requirement, it could be suggested that REM sleep in depression has an adaptive function. This suggestion is partly based on the relationship between REM sleep quality and SWS restoration during night sleep. As was stressed previously, depression in general is characterized by SWS deficiency, and SWS is often restored only after successful treatment 102,108. Patients who display FNE exhibit more SWS in their night sleep 106. In addition, SWS in depression is often redistributed compared to healthy subjects: it predominates in the second, and rarely even in the later cycles. We have shown 109 that in REM sleep episodes just before such an "explosion" of SWS in the third or fourth cycle, EM density is increased in comparison to all previous REM episodes, while such an explosion does not happen in depression with a flattened EM density distribution. Thus, it is possible to speculate that in some particular cases REM sleep, characterized by enhanced EM density, contributes to the improvement of the mental state of depressed patients and determines the psycho-physiological condition that allows SWS to appear. This conclusion is in agreement with data of some recent investigations: the increase of the positive dream content from the first to the last REM sleep periods is related to the remission from depression 27; at the same time the increase of REM density from the first to the last sleep cycle in depressed patients (a relatively rare finding) is accompanied by mood improvement from evening to morning 119.
However, the concept of the restorative and compensatory function of REM sleep at the first glance contradicts some experimental data, and this contradiction requires explanation. First, this concept does not correspond, to the beneficial effect of REM sleep deprivation 98 . Second, the duration of the first REM sleep period and its EM density are usually increased in depression 91; however, this does not contribute to mental health improvement. When EM density is highest in the first cycle, mood does not change from evening to morning, or even becomes worse. While EM density in the fourth cycle correlated positively with mood restoration, REM sleep duration in the first cycle correlated negatively with the same variable 110, 118. Also, the high EM density in the first REM sleep period, even if it exceeds EM density in any other cycle, does not predict or determine the increase of SWS in the second cycle. As stressed above, only in the second part of the night, after the second cycle passed, was any significant increase of SWS preceded by enhanced EM density. Previously, we have shown 2 that not the first, but the second REM sleep period increased in healthy students in the process of adaptation to stress. Thus, it is possible to suggest that the first REM sleep period is functionally different and less efficient in comparison with the subsequent REM sleep periods.
According to Cartwright and colleagues 27-28, it looks as if the first REM period serves as a bridge between the emotional state of the preceding wakefulness and its regulation in the subsequent REM sleep periods. In healthy non-sensitive subjects without strong negative pre-sleep mood dream activity, the first REM period is usually short and less elaborate, and corresponds to the short REM duration and low EM density. A shift toward more intense dreams in the first REM period may represent a response to excessive levels of pre-sleep negative affect, and in such dreams this affect is still not compensated. A normal dynamic in normal subjects with a high pre-sleep negative affect is the change from the initial negative affect in the first dream to increasing positive affect in subsequent dreams. According to Cartwright and Lloyd 111, depressed patients who had an enhanced dream-like quality of mentation during the first REM sleep period showed decreased Beck Depression Inventory scores at follow-up assessment. An enhanced dream-like activity may indicate that the brain starts to deal with negative affect in the first REM period. Such dealing is a necessary but insufficient condition for adaptation and mood regulation. The sufficient condition may be the restoration of search activity in the subsequent dreams. However, when speaking about resistant depression such restoration is usually not available. There are doubts whether the normal adaptive process even starts in the first REM period, in spite of its increased duration and enhanced phasic activity. There are even more doubts about the functional sufficiency of subsequent REM periods characterized by flattened EM density, in contrast to normal subjects who display an increase of EM density from cycle to cycle 91. In depressed patients the number of dream reports after awakenings in REM sleep during the night is decreased, dreams are shorter and simpler, and contain fewer images than dreams of healthy subjects 112. We have found that in patients with MD, subjective estimation of sleep latency correlated with EM density in the first REM sleep period and subjective estimation of the number of awakenings correlated with the total EM scores 113. In healthy subjects, as has been shown previously 35,75, EM density correlates with dream reports. Thus, it is possible to suggest that in depression psychic activity in REM sleep is often not perceived as dream mentation. Dreams may be repressed from the consciousness if they contain failures and feelings of helplessness. As a result, REM sleep is considered subjectively as wakefulness. This may be one of the reasons for the underestimation of sleep duration in depression. However, at the same time it is a sign of the proposed functional insufficiency of REM sleep that it does not provide an opportunity for the restorative search activity in dreams 5,35. Thus, the increased need for REM sleep in depression is combined with the functional insufficiency of REM sleep.
The insufficient dreams cannot restore search activity; moreover, such dreams may conserve or even increase the renunciation of search. The functional insufficiency of REM sleep and dreams may explain the positive outcome of REM deprivation in depression.
Next article>>>
