| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
SPECIAL ARTICLES |
From the Department of Psychosomatics and Psychotherapy, Hannover Medical School, Hannover, Germany.
Address reprint requests to: Prof. Dr. Friedhelm Lamprecht, Hannover Medical School, Department of Psychosomatics and Psychotherapy, Carl-Neuberg-Str., D-30625 Hannover, Germany. Email: Lamprecht.Friedhelm{at}MH-Hannover.de
ABSTRACT
In this review we trace the history of and professional discussion on psychological traumatization due to "railroad spine syndrome," "shell-shock syndrome," and "war neuroses," as well as the more or less endemic "posttraumatic stress disorder" of today. Psychological trauma engenders longlasting consequences in the biological, intrapsychic, and social organization of individuals. Medical experts have reported a shift in attention from exogenous to endogenous and back to exogenous causes, as indicated by new diagnostic systems (DSM-IV and ICD-10). As far as the relevant literature is concerned, the medical profession demonstrates the same partial amnesia as their patient counterparts. The purpose of this review is to overcome this fragmented memory and thus reach a more integrated view of what constitutes psychological trauma by reviewing trauma-related articles published in Psychosomatic Medicine. Moreover, we point out the direction in which research is desperately needed and ought to develop.
Key Words: posttraumatic stress disorder.
Abbreviations: DSM = Diagnostic and Statistical Manual of Mental Disorders;; EMDR = eye movement desensitization and reprocessing;; HPA = hypothalamic-pituitary-adrenal;; ICD = International Classification of Diseases;; NK = natural killer;; PTSD = posttraumatic stress disorder.
HISTORY
Unfortunately, war and violence are basic experiences in human history. Shay(1) notes that Homer’s (Iliad) contains powerful descriptions of war traumatization and stresses that the soldier frequently reacts to with disappointed withdrawal, grief, and feelings of guilt toward fallen comrades; he feels as if he is dead himself and occasionally is disposed to berserk-like rage. In fiction and poetry one can also find examples of how to cope with traumatization, such as in Oliver Twist by Charles Dickens, who tells of a boy having to come to terms with the early death of his parents. Daly (2) describes a witness of the Great Fire of London in 1666 who wrote, "How strange that to this very day I cannot sleep at night without great fear of being overcome by fire. And last night I lay awake until almost 2 o’clock in the morning because I could not stop thinking about the fire." Thought intrusions are described here in an exemplary way.
One of the first articles in medical literature on the subject of PTSD is by the English surgeon Erichsen (3), who attributed conspicuous psychological abnormalities following railway accidents to microtraumas of the spinal cord, which then led to the concept of the "railroad spine syndrome." The diverse range of symptoms described includes tiredness, anxiety, defective memory, irritability, sleep disturbance, nightmares, noise in the ear, dizziness, perceptual disorders, and pain in the limbs. He notes that speech can also be affected. This original connection drawn by Erichsen was later contradicted by the surgeon Page (4) and above all by Oppenheim (5), who, in his monograph on traumatic neurosis, first coined this term and then placed the main seat of the disturbance in the cerebrum. The term "trauma," which until then had been used exclusively in surgery, was thus introduced into psychiatry. Oppenheim describes (5) the frequent involvement of the heart: "The abnormal excitability of the cardiac nervous system is an almost constant symptom of traumatic neurosis, only in few instances is there serious cardiac disease." Myers (6) and da Costa (7) found "irritability of the heart" so frequently in soldiers with fighting experience that they gave it a diagnostic term of its own: "irritable heart" or "soldier’s heart."
The introduction of a public transportation system, the railway, which replaced the horse-drawn carriage, was associated with much horror at the outset. Personal liability and accident insurance were introduced at the same time, and railway accidents were among the very first accidents for which a person could sue for damages. Oppenheim (5) quotes Rigler and points to the fact that after the introduction of the personal liability law in Germany the number of railway invalids had increased nine-fold in the period between 1871 and 1876, as compared with the time when the railways were founded from 1835 to 1871. Simulation was extremely rare, but exaggerations were frequent. Oppenheim also mentions the problem that once legal action has commenced, the injured party naturally has an interest in appearing to be as ill as possible until the case is concluded. Evaluating the outcome, he therefore tends to make an unfavorable prognosis, "but in my experience the number of those is greater in whom, even after the end of the lawsuit, the suffering continues unchanged." Another term widely used in German-speaking countries around the turn of the century was that of the "Schreckneurose" (fright neurosis), coined by Kraepelin (8), who placed the personal shock in the foreground.
Another line of development regarding trauma etiology started at the Salpetrière in Paris, where Charcot (9) and Janet (10) pointed out the importance of traumatic experience for the origin of hysterical or dissociative symptoms. Janet underscored the role of dissociation as the central process in the genesis of posttraumatic symptoms. In this connection he did not speak of the unconscious but rather of a coconscious or subconscious. Thus, the therapeutic measures consisted mainly of integrative work on the memory and synthesis, although he was very skeptical of normal psychoanalytical work in the sense of catharsis and abreaction. According to Janet, emotions that are far too strong cannot be deposited anywhere and thus dissociate themselves from conscious and deliberate control. He claimed that the psychically traumatized were incapable of sorting what they had experienced within their narrative memory or as new terminology in their semantic-declarative memory. The automatism that occurs in the presentation of the symptoms is triggered by sensomotoric stimuli, and only with great difficulty can it be modified by new experiences. These considerations of Janet’s, explained in more detail in his Boston lectures, are surprisingly modern, yet they were not acknowledged by the psychoanalytical community. It was left to the modern trauma researchers van der Kolk and van der Hart (11), exactly 100 years later, to rescue from oblivion Janet’s work that is relevant for therapy.
PSEUDOLOGIA PHANTASTICA AND FALSE MEMORY SYNDROME
Another discussion at the end of the 19th century concerning false memory and external reality must be mentioned here because it appears again at the end of the 20th century in almost identical form. This discussion was initiated by Masson’s book The Assault on Truth (12). Although the exogenous causation stood in the foreground of the railway traumas, the psychoanalytical view places endogenous factors in the foreground, thus deemphasizing external reality. To point out this relative neglect of external reality, Paul Bernhard (13) surveys the sexual crimes on children under the age of 15 years, 36,176 cases of which are on record between 1827 and 1870, a fact that Freud knew but did not quote. Krutzenbichler (14) writes the following about the connection between lies, hysteria, and sexuality at the time when Freud returned to Vienna from Paris: "Sexual abuse is a social mass phenomenon, the children’s reports are mostly in accordance with the truth." In 1896, when Freud was almost 40, he tried to prove in a lecture at the Vienna Society for Neurology and Psychiatry that the large variety of hysterical symptoms his patients suffered were attributable to sexual abuse during their childhood. He based his argument on 18 cases he had treated. However, Freud met with an icy wind of rejection. After that he turned away from the trauma etiology of hysteria and developed his instinct theory and concept of transference. As a result of his theory, and the crumbling taboo, the specialist world reversed their original judgment concerning the mass phenomena of sexual abuse in the sense that "sexual abuse of children is reality only in the rarest of cases, and it is predominantly a veil of lies."
The debate in Paris during the second half of the 19th century resembles the present discussion to an alarming extent. When the specialist in forensic medicine Tardieu (15) documented the sexual abuse of children, this was dismissed by Fournier as "pseudologia phantastica" in children who falsely accuse their parents of incest. This reminds us of the debate about the so-called "false memory syndrome" at the end of the 20th century. After the pendulum had swung too far in the opposite direction and traumatization in early childhood was assumed to lie behind almost every symptom, fantasy reannounced itself with all its might, following the overestimation of external reality. In 1992 the "False Memory Syndrome Foundation" was founded with its main office in Philadelphia. Its members were psychiatrists and psychologists, and its task was to fight against an epidemic of dubious therapies (repressed memory therapy) that "tears apart thousands of families, scares patients and makes them fearful, and breaks the hearts of innocent parents and other relatives. It is actually the real mental health crisis of the Nineties." In her paper "Remembering Dangerously," Elisabeth F. Loftus (16) gives impressive examples of how, through psychotherapists and especially those who handle their patients very suggestively, emerging memories can be easily held to be true and can be used for false charges in court. There is no doubt that there are some such rare cases. As a psychotherapist with almost 30 years of professional experience, I must ask myself why in the first 15 years of my work I diagnosed almost no cases of real incest, whereas today I encounter real incest with physical violence on a daily basis. Of course, its emergence is facilitated by the media, and, without a doubt, because the focal point of our therapy is psychotrauma, we attract patients of this kind. However, the suspicion remains that during the 1970s and 1980s traumatization within families received too little attention.
Returning to history, the term "shell-shock" was introduced into the specialist literature by the British military psychiatrist C. S. Myers (17) in 1915. After discovering that the shell-shock syndrome had also been found in soldiers who had not participated in actual fighting, he differentiated between "shell concussion," in which a neurological disturbance could clearly be identified as a result of a physical injury, and the actual shell-shock syndrome, in which an emotional shock brought about by extreme stress was regarded as sufficient for its causation (18). Those soldiers in the latter group were in part, at least in Germany, discriminated against as being cowards and traitors, yet other countries also have had difficulties in accepting this diagnosis (19). Thus, the therapies during and after World War I were mainly designed to increase the soldiers’ willpower. In his controversial article "The Law is the Cause of Accident Neurosis," Bonhoeffer (20) saw the cause of persistent accident neuroses in an interplay between predisposition, constitution, degenerative inclination, and compensation. Thus, from his point of view the secondary gain seems to be one of the major causal factors.
In addition to war traumatization and railway accidents, natural disasters were also dealt with from a psychiatric point of view at the beginning of the 20th century. Stierlin (21) investigated the earthquake at Messina (1907) and found that 25% of the victims suffered from sleep disturbance and nightmares. Prasad (22) describes the emotional problems of the victims after a catastrophic earthquake in India. Adler (23) provides a good clinical description of posttraumatic consequences in the survivors of the Boston Coconut Grove fire, with emphasis on avoidance behavior, nightmares, and insomnia.
The Traumatic Neuroses of War, published in 1941 by Kardiner (24), is an outstanding contribution to the field of psychotraumatology with very detailed clinical descriptions. Kardiner developed the concept of "physioneurosis," thereby indicating bodily involvement right from the start. Trained in psychoanalysis, he initially tried to reconcile the concept of traumatic neuroses with Freud’s repressed sexual drive theory but obviously failed to do so and came to the conclusion that "the traumatic neuroses fared badly in psychoanalysis." What we today call flashbacks he described as "hallucinatory reproductions of sensations on the original occasion." He saw amnesia as a defensive process of the personality as a whole and as a collapse of ego resources. He also mentioned nightmares with a threat of annihilation or very aggressive dreams. The central process, according to Kardiner, is a contraction of the ego, an inhibitory process that he saw as the primary symptom, all others being secondary. The proneness to motor expression and the arousal to anger account for the irritability of these patients and for the difficulty in treating them. According to Kardiner, the sleep disturbances are due to an increased susceptibility to external stimuli, preventing the patients from falling asleep, and when sleep is accomplished the dream content awakens them. The sense organs generally seem to function, but their meaning has been modified. In contrast to Oppenheim (5), Kardiner (24) has a much more optimistic view as far as treatment is concerned, with a 60% success rate. However, he cautions that "the recovery of the amnesis usually accompanying the trauma ought not to be made the objective of treatment." From a forensic perspective, it is important that he notes "that predisposition alone cannot produce this disease." "In fact, the traumatic neurosis occupied a kind of no-man’s,land between the departments of organic neurology, internal medicine and psychiatry," and I would add psychotherapy. When all these disciplines are taken together we end up with psychosomatic medicine, and posttraumatic disorder could be a core example of psychosomatic medicine. Kardiner (24) differentiated the normal action syndrome from its alteration through trauma in terms of the symptomatology. This differentiation led to the term physioneurosis, underlining the predominance of physiological changes. At that time, diagnoses such as "combat fatigue" or "combat exhaustion" were also quite popular (25).
DIAGNOSIS OF PTSD
We now will briefly outline how the operationally defined diagnostic category of PTSD came into being and will then trace this development through the articles published in Psychosomatic Medicine. Numerous accounts of combat reactions during World War II and after the Korean War (see also Grinker and Spiegel, Ref. 26; Noble et al., Ref. 27; and Lifton, Ref. 28) led to the DSM-I diagnosis of "gross stress reaction," which was introduced in 1952 (29). This diagnosis did not require an operational definition. It seemed to be justified for times when extreme demands or stress situations had arisen as the result of acts of war or natural catastrophes. Such extreme situations were capable of causing abnormal behavior in people who had previously been completely normal. The underlying assumptions were that every individual had a breaking point and that there were limits to the amount of stress that each individual could tolerate. During the period that followed, there was an increase in the number of publications about the consequences of natural catastrophes and civil disasters, such as the Alaska earthquake of 1962 (30), the sinking of the Andrea Doria (31), the Mississippi tornado in 1953 (32), and the Bristol flood disaster (33), to mention but a few.
From the mid-1960s to the early 1970s, many reports and evaluations of the long-term consequences of incarceration in a concentration camp also appeared (34–37). At the end of the 1960s a group of experts from the American Psychological Association published DSM-II (38). "Gross stress reaction" was replaced by "transient situational disturbance." This even included psychotic reactions following overwhelming stress and all kinds of acute reactions with differing degrees of severity. Once again, an operational definition was missing. This lack of operational criteria, the limited reliability of the diagnosis, and many missing links led to a commission within the American Psychiatric Association in the mid-1970s, headed by Robert Spitzer, which had the task of creating more detailed symptom profiles for all the 265 categories. The subcommission for "reactive disorders," whose members included Nancy Anderson, Robert Lifton, Chaim Satan, Jack Smith, Robert Spitzer, and Lyman Wynne, then came up with the label of "posttraumatic stress disorder" as published in the DSM-III (39). Specified diagnostic criteria were described. The use of this diagnostic category is reflected by the increase in publications on PTSD during the years to follow: according to Medline, from 10 in 1980 to more than 200 in 1990, with a tremendous increase during the 1990s (see Figure 1). In DSM-III-R (40), special classes of traumatization are described: for example, being a witness to a violent act or accident is included, as is being informed that something bad has happened to a close relative or friend. The question of comorbidity arose because in some patients with PTSD, symptoms of depression, anxiety, or phobias were so severe that they could be diagnosed additionally.
|
In Europe the ICD-10 (42) is more popular, and the research criteria for PTSD here are as follows:
The symptomatology normally appears within 6 months after the traumatic experience. Although the posttraumatic stress disturbance forms the focal point of our observations, it must be mentioned here that other diagnoses may be made as a consequence of psychological traumatization, such as depressive illnesses, anxiety disorders, panic reactions, dissociative disorders, somatoform disorders, and substance-related disorders, as well as certain kinds of personality disorders, depending on the duration and the age at which the traumatization took place.
PSYCHOSOMATIC MEDICINE AND PTSD
It was the wish of the editor of Psychosomatic Medicine to place particular emphasis on the PTSD-related articles published in this journal and on how this topic is reflected in the research reported elsewhere. The first article by Zabriskie and Brush in 1941 (43) is the most extensive on the topic. It deals with psychoneuroses in wartime. The authors cover the literature available at that time and place particular emphasis on World War I soldiers, reporting functional nervous disease among 34% of British soldiers, 24% of Canadian soldiers, and 9.5% in American soldiers. No explanation for these differences is given in the original article. In Britain in 1918, 32,000 pensions were paid for functional nervous disease among soldiers, which increased 3 years later to 65,000; in 1939, 2 million pounds were still being paid as compensation payments and to soldiers with "shell shock." This particular article is more like a review article without a clear position, although it states that the experience with war led to an increase in knowledge about the etiology and treatment of neuroses. In 1942, Rado (44) gathered arguments against the concept of traumatic neuroses attributed to microstructural or molecular changes in the nervous system by putting forward psychodynamic considerations such as "flight into illness," "fixation on the trauma," and "secondary gain from illness" in terms of the pension. The symptomatology he described was along the lines of Kardiner (24). As therapy he attempts a kind of reframing in the sense of "what a guy you are" and "you went through all that and nothing happened to you, and chances are that nothing will." In 1943 Goldstein (45) argues against the term "war neuroses," preferring to speak of war neuroses as a "nervous breakdown due to war events." He differentiates three groups of disturbances due to war-related events: 1) acute nervous breakdown in the sense of a prolonged anxiety reaction, 2) stress-induced conversion, and 3) worsening or change of previously existing neuroses brought on by stress.
Goldstein goes on to stress the fact that many people who have been exposed to extreme war conditions survive without suffering any serious consequences. He also brings protective factors into the discussion, such as good physical condition, ego strength, meaningfulness, and degrees of freedom to act in the traumatic situation.
Murray (46) describes his experience as a psychiatrist in the air force and states that practically "all men develop anxiety in the battle situation." He quotes Captain David G. Wright from the air surgeon’s office, saying that up to 70% of successful combat men developed irritability and quick-flaring anger toward their crew mates, in contrast to their former behavior. He encourages particular treatment programs for veterans with persisting symptoms but gives no detailed description of a proposed treatment program. He also disagreed with Freud’s verdict "that there is no neurosis without a disturbance of the sexual function."
In their article from 1944, Grinker and Spiegel (47) place emphasis on a psychoanalytically derived short-term therapy. These authors also describe the practice of inducing a state of seminarcosis using sodium pentothal. This enables the patient to live through his traumatic battle experiences, thereby synthesizing the formerly fragmented emotions and impressions related to the trauma so as to construct a "memory which corresponds almost completely with the original experience." The main goal of treatment was to get the soldiers back to the front line as soon as possible; therefore, the maximum stay in hospital was 4 to 5 days. To achieve this goal, pressure was usefully increased by indicating "that relief from the fight was dishonorable, that family, friends, comrades, and countries expected the soldier to return and finish his job." This is in contrast to their statement that "War neuroses are caused by war. No one is immune from a war neurosis." Therefore, it is hard to understand how people traumatized by war events, seen as the major cause of illness, should return within 5 days to the environment that made them sick. However, the authors themselves express doubts regarding these statistics in their observation on the rate of relapse. Among the more seriously disturbed patients, they find stuporous and amnesic anxiety states "where the ego seems incapable of digesting the traumatic experiences except in small doses."
In his 1945 article (25), Saul identifies traumatic war experiences with the term "combat fatigue." He incorporates Cannons’ "fight-flight reaction" and the concept of controllability. The intensity of a man’s emotional response to combat depends on the strength of his "fight or flight reaction and on the strength of his forces to control." The overwhelming stress combined with the inability to act results in an increase in heartbeat, breathing rate, blood pressure, etc. Saul himself was a navy officer, and he illustrates his point by supporting case histories.
The 1946 article by Hunt (48) presents four case histories in detail, using a pentothal interview, thereby enabling the patients to abreact, "which they consistently failed to do in the history-taking interview." He summarized his observations: "Although the symptoms of these four patients were colored by the war setting, they represent in each instance the exacerbation of an old neurosis, and their illness does not deserve a special name. In the author’s experience, such is the case with all persistent combat reactions." This stance certainly is in contrast to that of many of the previous and future contributors in the field.
In his later article in 1947, Grinker (49) describes the short-term therapy as implemented in psychosomatic problems. One could gain the impression that the experiences and theoretical reflections in the 1944 article written with Spiegel (47) were mainly contributed by Spiegel from his time working with Kardiner. Grinker seems to be the specialist for short-term psychotherapy, whereas Spiegel was the outstanding trauma therapist.
Menninger (50) deals with the great frequency and variety of bodily symptoms without organic cause among veterans. He introduces the concept of "somatization reaction," a precursor of somatoform disorders. "The anxiety is relieved in such reactions by channeling the originating impulses through the autonomic nervous system into visceral organ symptoms and complaints." The anxiety is thus prevented from becoming conscious. After finding that "every army physician was confronted with a far greater number of patients with physical complaints in which no organic pathology was found than he saw in civilian life," Menninger broadened his concept to include all organ systems (cardiovascular, genitourinary, skin-reaction, etc.) in an attempt to reconcile bodily reactions with psychoanalysis.
In 1950 Selye and Fortier (51) introduced "the general adaptation syndrome" derived from animal studies, with the three phases of alarm reaction, resistance, and exhaustion. This concept has been very fruitful and stimulating, particularly as far as the involvement of the hypothalamic-pituitary-adrenal (HPS) axis is concerned during the different phases. The broad implications of the stress concept are demonstrated in the original figure reproduced from this Journal (Figure 2).
|
As an air force psychiatrist, Lifton (54) reports on the reaction patterns of American prisoners of war repatriated from North Korea who were brought home by ship. After such a long period of imprisonment (>2 years), with the last phase being in a Chinese indoctrination camp, on the ship they showed the previously existing response patterns of either "playing it cool" or an attitude of hostile resistance, or they were again cooperative as they had been with the Chinese authorities. They were all surprisingly unenthusiastic about their release; they lacked spontaneity; were restless, tense, and suspicious; and spoke in a monotonous tone with diminished affectivity. No numbers relating to PTSD symptoms are given, although one gets the impression that a high percentage would fulfill the criteria for the diagnosis of continuing personality change following extreme stress (ICD-10: F62).
It seems that with the end of the Korean War interest in PTSD-like symptomatology diminished, particularly in Psychosomatic Medicine, whereas in other journals attention was given to the long-term consequences of the Holocaust (55–57) and to accidents in everyday life (58, 59).
In 1968 Seligman (60) used the paradigm of unpredictable shock to produce chronic fear. By showing formerly traumatized patients a stressful film, Horowitz (61) could induce the return of images from their traumatizing experience. Weiss (62) demonstrated in an animal study his model of how one can investigate the effect of the psychological construct of controllability (predictable vs. unpredictable shock) on the somatic effects. He found that the animal group that unpredictably received the same amount (duration and intensity) of electric shock as the control group showed more severe stomach ulceration, a greater rise in body temperature, higher plasma corticosterone, greater body weight loss, and greater decrease of food and water intake. The last three articles might indicate a more experimental approach to studying the symptomatology of conditions related to PTSD.
In a broader sense, there were numerous reports in Psychosomatic Medicine related to stress and endocrinological changes (63–68). The topic of bereavement could also have some implications for PTSD (69–72). Differences between the sexes with regard to examination stress (73) and the resulting psychoneuroendocrine reactions were an important topic of study because they had implications for research questions in PTSD.
Although stress studies are given adequate attention in Psychosomatic Medicine, there are almost no studies on PTSD in a narrow sense in the 1970s and 1980s despite the increased interest in PTSD prompted by the Vietnam War. The Vietnam veterans are probably the best-examined population at risk of developing PTSD. Horowitz (74) developed in 1979 a psychometric test to assess the PTSD-specific symptoms of intrusive thinking and avoidance behavior with good psychometric properties, and his Impact of Event Scale is still one of the most widely used psychometric test instruments in PTSD research even today. Krystal et al. (75) developed an instrument for the measurement of alexithymia, the Alexithymia Provoked Response Questionnaire, and found a higher score in PTSD patients than in patients with affective disorders. Henry et al. (76) see similar neuroendocrine patterns of PTSD and alexithymia. From this data, Henry et al. derived the hypothesis that the sympathetic-adrenal medullary axis is controlled more by activation of the left hemisphere and the HPA axis more by the right. He makes this dissociation responsible for the failure of cortisol to rise in PTSD (and presumably in alexithymia) despite high anxiety accompanied by high norepinephrine levels. In a recent volume of Acta Physiologica Scandinavica, Henry (77) gives a more extensive outline of these theoretical assumptions.
With the introduction of DSM-III-R, as mentioned above, scientific interest in PTSD rose sharply. This is reflected by the PTSD-related articles published during the 1990s in Psychosomatic Medicine. Kolb (78) states in his editorial comment, "Although PTSD is accompanied by a wide variety of psychosomatic symptoms, surprisingly little has appeared on this subject in the journals addressed to psychosomatic medicine." In his "future dream of PTSD as a psychosomatic disorder," he stresses the importance of the provocative audiovisual presentation of such traumatizing events. Shalev et al. (79) review the PTSD literature with emphasis on studies of the auditory startle response and find data that show an impaired habituation of the skin conductance component of the auditory startle response, thus indicating altered responsiveness of the central nervous system to elementary stimuli. This is in line with Kolb’s (80) hypothesis that excessive stimulation in PTSD leads to neuronal changes with impairment of learning habituation and stimulus discrimination. Not only war-related traumas became an object of interest but also natural disasters and intrafamilial abuse (sexual abuse, physical abuse, or neglect), which had emerged out of the realm of taboo.
Neuroendocrinological changes are frequently described (81–85). Yehuda (53) reviews her own studies and others on the function of the HPA axis under different stress conditions both in humans and animals. HPA activation under acute stress is certainly an adaptive response, but if it lasts too long, the damaging effects become obvious. Prolonged elevated glucocorticoid concentrations might lead to neuronal degeneration in areas with high densities of corticoid receptors, as in the hippocampus, which then might be responsible for the reduced hippocampal volume found in PTSD patients (86). Thus, an adaptive response can become maladaptive (87). Conversely, the underactivity of the HPA axis seen after chronic stress exposure might increase the vulnerability of these individuals when facing future traumatizing events. Yehuda mentions a study by Ressnik on 19 rape victims. Among those who had a prior assault history, 9 of 11 developed PTSD, and these victims had very low cortisol levels compared with those who experienced rape for the first time. The latter had cortisol levels about twice as high, and of these only two of eight developed PTSD. Thus, the higher cortisol level in the latter group seems to have some protective effect. This is supported indirectly by a study on abuse-related PTSD and alterations of the HPA axis in women with chronic pelvic pain (82). They conclude that a lack of protective properties of cortisol may be of relevance for the development of bodily disorders in chronically traumatized patients. Rahe et al. (88) found saliva cortisol as the only reliable correlate of psychiatric disturbance in hostages freed from captivity in Iran. In a different study (81) on abuse-related PTSD in contrast to combat experience, the authors found an increase in norepinephrine and cortisol as well, explaining this by possible gender differences. Gender differences are becoming more and more important in PTSD research. Hawk et al. (85) investigated recent-onset PTSD after serious motor vehicle accidents. Urinary epinephrine and norepinephrine and cortisol levels were elevated significantly only in men 1 month after the accident. Greater emotional numbing predicted lower cortisol levels after 6 months, whereas intrusive thoughts predicted higher catecholamine levels at 1 month.
In his editorial comment, Prange (83) suggested that increased secretion of triiodothyronine acts directly on brain structures, thus contributing to additional activation and thereby sustaining arousal. "Perhaps this can happen when the emergency, having disappeared from the environment, persists in the brain itself and there is a rooted terror". Wang and Mason (84) in their study found the most significant relationship between the CAPS-2 (Clinician-Administered PTSD Scale) hyperarousal score and the total T3 in 65 Vietnam combat veterans. This could indicate a high thyroid/high hyperarousal subtype or a special phase in the course of PTSD.
During the 1990s, immunological changes also received attention in Psychosomatic Medicine (89–93). Even during peacetime, an estimated 6% to 7% of the US population has been exposed to a trauma each year ranging from motor vehicle accidents and crime to hurricanes and tornados (94). Delahanty et al. (89), in their investigation of workers at the crash site of US Air Flight 427, found the highest natural cell (NK) cell activity at time 1 (within 2 months after the crash) and the highest levels of intrusive thoughts at both time points in nonmorgue workers exposed to bodies or body parts. This underlines the importance of preparing workers for such a task by informing them in advance about what to expect in the aftermath of such a disaster. Ironson et al. (90) looked at people with a high score on the Impact of Event Scale (IES) after Hurricane Andrew using several immunological parameters. NK cell cytotoxicity was decreased, particularly in those with high intrusive thoughts and sleep problems. These two symptoms could be seen as possible mediators between posttraumatic stress symptoms and NK cell cytotoxicity. An increased white blood cell count was found in this and a later study (93) of Vietnam veterans, in whom they also report an elevated total T-cell count. Here it becomes obvious that a traumatic event has such an impact that almost every system examined is changed somehow during the development of PTSD. For instance, in a longitudinal study of coronary heart disease risk factors in Italy in 1975, 1980, and 1987, a disastrous earthquake occurred during the second time point of data collection. Participants seen after the quake had significantly lower serum uric acid levels compared with those seen before, reflecting an acute effect. In 1987, however, increased serum uric acid was found in the group still suffering from damage from the 1980 disaster (95). These developed more anxiety during a 1983/1984 low-level earthquake (96). Here again, acute and chronic effects are quite different.
Treatment issues surrounding PTSD have only found meager representation in Psychosomatic Medicine. Consequently, Shalev et al. (97) were invited to write a review about the treatment of PTSD. They identified 81 articles that they divided into biological and psychological categories, with the latter further divided into behavioral, cognitive, psychodynamic, and other treatment modalities. The authors highlight the difficulties in forming proper control groups, including survivors of a variety of events, and difficulties in identifying the broad spectrum of initial symptoms, severity, and duration of PTSD. In most studies no details are given on prior traumatization or on possible protective factors in the sense of Antonovsky’s (98) "salutogenesis." The authors argue in favor of combining biological, psychological, and psychosocial treatment modules. Rehabilitative goals should replace curative ones, particularly in patients with chronic PTSD. A reduction of PTSD symptoms is reported in many studies combined with improvements in the quality of life, but results are "often limited, and remission is rarely achieved." In this review no calculations were done for effect sizes. Rosch (99), president of the American Stress Institute, criticizes that no mention is made of eye movement desensitization and reprocessing (EMDR), which by 1997 was practiced by about 17,000 trained practitioners in the US, who had treated approximately 1 million patients with promising, and occasionally impressive, results. There have been critical articles about EMDR (100, 101) questioning the different components of this treatment technique but not the effectiveness. Thus, in the practice guidelines from the International Society for Traumatic Stress Studies (102), EMDR is described as an effective treatment with level A evidence (based on a review of seven published, randomized, controlled studies with overall large effect sizes).
Glover (103) adds that an assessment of the methodological difficulties of evaluating and measuring the numbing response vs. depression should have been included (104). Glover then points to his own hypothesis that the endogenous opiate system seems to represent the primary neuroregulatory system to "effect major changes in diverse behavioral systems in response to overwhelming external and/or internal stress" (105).
PTSD treatment studies that include physiological or endocrinological changes before and after treatment are very rare. Therefore, we think an in-depth case report would greatly contribute to the understanding of symptom formation, such as hypertension (106). Mann and Delon found that the disclosure of a rape three decades before led to a turning point in the victim’s essential hypertension, with average values of 128/93 mm Hg for 18 months after disclosure, indicating that repressed emotion about a 30-year-old trauma could sustain hypertension for such a long time. This underlines the additional contribution by the underlying psychodynamics to the formation of psychosomatic symptoms (107).
A very important study (108) uses the Vietnam Era Twin Registry to address the investigation of traumatic events, genetic and environmental factors, and their interplay for current physical health problems. More than 90% of the variance is seen to be attributable to inherited factors and unmeasured environmental experiences not shared by twin siblings. Thus, according to the authors, combat stress seems to contribute relatively little to the risk of physical illness, whereas in a different study (109), the same group of authors found in 1987 a nine-fold greater risk of reporting PTSD when comparing Vietnam veterans with their siblings with no combat experience. Although in the 1998 study (108) a rough estimate is given of the combat experience without intensity or duration, no indication is given about the prevalence of PTSD. Because it is now known that PTSD is accompanied by a wide variety of physical symptoms, one would have expected a much higher explained variance. However, the physical symptoms investigated, such as joint problems, hearing problems, hypertension, and skin conditions, are not the kind of physical symptoms that occur often with PTSD.
Two articles address Gulf War veterans. Roy et al. (110) report on a wide variety of symptoms without the emergence of a new or unique illness related to the Gulf War experience. Wolfe et al. (111) examined the relationship between the psychiatric status and the presented health problems of Gulf War veterans. They found higher rates of reported health problems in veterans with PTSD, but this disorder did not entirely account for the reported health problems. In this context Kaplan’s 1997 editorial (112) deserves consideration. He introduces arguments in favor of a much broader view of the cause of traumatization. The prevention of war, although seeming somewhat grandiose, should become an integral part of considerations about the cause of traumatization, not only for public health professionals but for all biopsychosocial scientists, to reestablish the context of "exposure" to war and violence. This would require the inclusion of the "upstream determinants such as distribution of power, privilege, and resources" to see how this makes an individual vulnerable or resilient to stress even in everyday life (113).
Traffic accidents are unfortunately all too common, either through direct involvement or as a witness. Mayou et al. (114) examined the long-term outcome of motor vehicle injury in a prospective investigation up to 5 years after the event. They found little change either in the outcome of the quality of life or in the effects on travel between the different evaluation time points (3 months, 1 year, and 5 years). The prevalence of PTSD remained constant over time, at about 10%, but early cases had remitted and were replaced by delayed-onset PTSD. PTSD after 5 years is predicted by the physical outcome, postaccident intrusive memories, and emotional distress and compensation procedures that have not been settled within 5 years. In a letter referring to this article, Malt (115) contradicts Mayou et al.’s claim that they present the first report on the long-term psychosocial outcome of motor vehicle accident injuries by quoting Thorson’s work from 1975 (116) and his own articles (117, 118). Malt and Olafsen (117, 118) found a lower prevalence of PTSD (6%) and a significant relationship between preaccident problems and a negative long-term outcome. To explain the difference in prevalence rates in Norway and Britain, they assume that questionnaires and interviews focusing only on the postaccident situation will certainly overestimate the effects of the accident on long-term follow-up.
A new instrument for measuring intrusive thoughts (the Intrusive Thought Questionnaire) and their impact on distress was introduced and validated in three different trauma populations (motor vehicle accidents, hurricane survivors, and recovery workers at an airline disaster) (119) with good psychometric properties in all groups (eg, with an internal consistency (Cronbach’s 
) ranging from 0.81 to 0.93. The authors see an advantage over the Impact of Events Scale because of an increased predictive power with regard to trauma-related outcomes. For studies with limited clinical resources, a self-rating scale for PTSD is offered (120). Good internal consistency and interrater reliability are reported.
Train drivers are often confronted with accidents involving suicidal behavior (121). Theorell et al. (122) in their study of "person under train incidents" investigate the medical consequences for subway drivers. Three measurement time points (3 weeks, 3 months, and 1 year) after the event included prolactin, cortisol, and various psychometric measurements. The main outcome variable was "sick leave." The drivers had a mild transitory psychophysiological activation indicated by increased prolactin levels and increased sick leave during the first 3 weeks. Between 3 weeks and 3 months there was no difference from the control group, but between 3 months and 1 year there was again an increase in sick leave that was best predicted by a high plasma cortisol level, a high depression score, and by the condition of the victim. Mildly injured or dead victims led to less absence from work for the drivers than seriously injured victims. This is, of course, what one would have expected because of the continuing involvement with the severely injured victim in their fantasy life.
A long-term follow-up study was made of third-grade children who were taken hostage at gunpoint at their school (123) for 2 hours. Acute stress reactions were found in all but one. After 2 months18 children had severe clinical symptoms (7 PTSD, 11 subclinical PTSD), and even in the children indirectly involved, 2 full and 7 subclinical PTSD cases (of 21) were found, indicating increased vulnerability at a younger age. Psychological debriefing did not have a preventive effect, but "children who were not debriefed had the worst outcome," which is most likely a selection effect. In another study (124), 60% of the children still had fully developed PTSD 14 months after a school shooting.
PTSD ought to receive more attention in other medical disciplines, for instance transplantation units, trauma clinics, and departments of gastroenterology, neurology, surgery (125), and gynecology to name a few. As an indicator for life-threatening illnesses, 12% to 19% of breast cancer patients who had undergone autologous bone marrow transplantation (2–62 months earlier) met DSM-IV criteria for PTSD. Less educated people started treatment later and at a more advanced disease state and had greater risk of developing PTSD symptoms (126).
Leserman et al. (127) investigated female patients with functional and gastrointestinal symptoms and found a very high prevalence of physical and sexual abuse in their history. Seng et al. (128) used Medicaid records to compare pregnancy outcomes among women with and without a history of PTSD. Their main findings are that women with PTSD are significantly more likely to be hospitalized for spontaneous abortion, excessive vomiting, and preterm contractions. As a former neurologist and consultant, I saw a young patient who had a subarachnoidal hemorrhage during intercourse and who, despite good clinical physical improvement, developed the full picture of PTSD.
Intrusive memories and loss of control in itself seem to be possible determinants of chronic stress (129). In this longitudinal study Baum et al. looked at the after-effects of the nuclear accident at Three Mile Island. Intrusive traumatic recollections found in depressive patients give an indication of prior traumatization in 48% of the cases examined. Comorbid PTSD was found in 13% of the total sample. Whether this really justifies an independent diagnosis or if the depression is a consequence of prior traumatization can be answered only by longitudinal studies. Cheasty et al. (130) provide evidence for the assumption that childhood traumatization can lead to depression in later life.
The importance of taking into account the cultural backgrounds of PTSD patients is particularly remarkable because suffering from and resilience to stress are overshadowed and influenced by cultural upbringing. This is demonstrated in the life of a Puerto Rican women (131). For more insight into this subject, the article by Marsella et al. (132) is very helpful, reviewing the ethnocultural aspects of PTSD and discussing the methodological issues involved in understanding the relationship between culture and PTSD.
With this brief overview covering the articles relevant to PTSD in Psychosomatic Medicine, it has become evident that the relative neglect of trauma-related topics in the 1970s and 1980s is compensated by numerous and important contributions during the 1990s. It has also become obvious that Kolb’s dream (78) to see PTSD as a psychosomatic disorder became reality. Unfortunately, it is also clear that even in the absence of war, in our profession we are confronted with traumatizing events on a daily basis.
Studies on the treatment of PTSD patients are virtually absent in Psychosomatic Medicine, even though good therapeutic methods are available (see also the practice guidelines from the International Society for Traumatic Stress Studies) (102). Although treatment issues have never been a main focus in Psychosomatic Medicine, we think that attention should be devoted to therapy studies including biological data and their possible changes through treatment interventions, thereby fulfilling the purpose stated in 1939 by the founders of Psychosomatic Medicine: "Emphasis is put on the thesis that there is no logical distinction between ‘mind and body,’ mental and physical. It is assumed that the complex neurophysiology of mood, instinct, and intellect differs from other physiology in degree of complexity, but not in quality. Hence, again divisions of medical disciplines into physiology, neurology, internal medicine, psychiatry, and psychology may be convenient for academic administration, but biologically and philosophically these divisions have no validity. It takes for granted that psychic and somatic phenomena take place in the same biological system and are probably two aspects of the same process, that psychological phenomena should be studied in their psychological causality with intrinsically psychological methods and physiological phenomena in their physical causality with the methods of physics and chemistry" (133).
Desiderata concerning future research in the field of posttraumatic stress disorder are as follows: 1) prevention of PTSD and treatment of the consequences of acute trauma, 2) risk factors and protective factors, 3) complex PTSD, 4) neurobiological research approaches, 5) pharmacotherapy, and 6) dealing with violence and traumatization in society.
It still remains unclear which therapeutic methods can contribute successfully to the prevention of PTSD after acute traumatization. Critical stress incidence debriefing (134) is widely used, especially in the after-treatment of members of emergency services after catastrophes and large accidents. A surprisingly high incidence of PTSD was found in the intervention group in several controlled studies on the effectiveness of debriefing. A systematic meta-analysis carried out in accordance with the criteria of the Cochrane Collaboration thus comes to the conclusion that debriefing cannot be recommended for acute intervention (135). Debriefing is usually offered as group intervention. The members of the group are given information on the psychic consequences of traumatization and the possible ways of coping with it, and they exchange details about their traumatic experiences. Among some members of the group, this clearly leads to a renewed flooding of traumatic memories. For this reason individual counseling after traumatic experiences, also in the sense of screening for the need for treatment, seems to be more promising. Studies that evaluate the relevant screening criteria empirically and also include psychometric questionnaires, for instance, are thus urgently required. This is important also for traumatization induced by medical procedures. Just taking emergency caesarean section (136) or cardiac surgery (125) as an example, we are only touching the tip of the iceberg.
Although a whole series of risk factors for the development of PTSD have been empirically validated (eg, female gender, severity of the trauma and type of traumatization, peritraumatic dissociation, previous psychiatric disorders, and young or old age), only few systematic surveys exist on protective factors (137, 138). People who stay well after natural disasters, motor vehicle accidents, and even rape should also be evaluated. Here it is precisely the specific promotion of intrapersonal and social resources that could complement measures of psychotherapeutic treatment and promote their effectiveness. The concept of posttraumatic growth provides an interesting conceptual starting point for this. Experience with posttraumatic growth has certainly been frequently observed after serious illness and traumatization (139). Starting points for therapy, which have not yet been used frequently enough, are available here for helping traumatized people to integrate their traumatization into their life history, both as a painful experience and as a personally valuable one in a salutogenetic sense. A taxonomy of disasters and their victims (140) has to be followed and diversified.
The inclusion of the diagnostic criteria of PTSD in the diagnosis classifications DSM-III and ICD-10 was a big step forward and certainly helpful to many of those affected. Nevertheless, the current diagnosis criteria cover only a small section of the overall spectrum of frequently occurring psychic and psychosomatic symptoms after traumatization. There has been repeated criticism from various quarters concerning the diagnosis criteria after DSM-IV and ICD-10 (141, 142). This criticism is directed, above all, against the relative overvaluation of criterion C (avoidance behavior) when compared with the other diagnosis criteria. In clinical practice, patients are frequently seen who fulfill all the other symptom criteria. The lack of diagnostic categories for the characteristic symptoms of patients who suffered serious sexual abuse or maltreatment during their childhood is far more problematic. Approaches for the classification of a complex PTSD such as disorders of extreme stress not otherwise specified (DESNOS criteria) were investigated in a field study for the preparation of DSM-IV (143, 144). However, a further empirical evaluation of these criteria is still urgently required.
NEUROBIOLOGICAL RESEARCH APPROACHES
It is to be expected that the research activities on the neurobiology of PTSD, which have rapidly increased, will lead to a considerable growth in knowledge about trauma-induced functional and morphological changes in the central nervous system in the next few years. In addition to an increase in individual items of knowledge, for instance in the field of endocrine organs and the immune system as well as neurotransmitters and the morphological classification of brain functions, the link-ing of the different investigation results will gain in im-portance because this will enable researchers to establish an integrated overall understanding of the disorders caused by psychological traumatization. At this stage it can already be stated that a comprehensive theory satis-fying this requirement must manage without the tradi-tional dichotomy of research methods (ie, the separation of body and mind) to do justice to the complexity of the problem (145). This development represents a very spe-cial challenge to psychosomatic medicine as a discipline that genuinely and traditionally nurtures an integrating scientific understanding of physical, mental, and social conditions.
In its effectiveness, the pharmacological treatment of PTSD is inferior to psychotherapeutic intervention (146), particularly where the trauma-specific symptoms of avoidance behavior thought intrusions, flashbacks, and nightmares are concerned. The effectiveness of the pharmacotherapy of PTSD could presumably be considerably improved by the development of drugs that specifically remedy the disturbed neurobiological functions. One possible group of substances for this is the opiate antagonists, which, according to initial surveys, can relatively specifically improve dissociative symptoms and the urge to injure oneself (147). Justified hopes exist with regard to the successful use of drugs for primary prevention after acute traumatization. According to initial reports, the use of excitation-damping substances, such as ß-blockers, could reduce the risk of developing a PTSD (148). Careful clinical studies of the application of drugs in the case of acute traumatization are therefore urgently needed.
We live in an age in which traumatization, if not real, is nevertheless omnipresent in the attempts at its cultural assimilation, such as in films, video games, etc. (149). The average youngster in the United States, until the age of 18, has seen up to 40,000 homicides on television. That this has some influence on behavior is a vague but more probable conclusion than to state that it has no effect whatsoever. One can gain the impression that this constant presentation of violence in the media represents an attempt to overcome and assimilate it, however effective this may or may not be. Our society is called on to find ways of solving the problem of the passing on from one generation to another of violence within families and within society (150, 151). Numerous criminals have been victims during their early life. In this connection the psychotherapeutic treatment of those who have committed crimes could play an important part. The development of treatment concepts for the psychotherapeutic treatment of these people and their empirical evaluation is at a very early stage. In the future it could gain social significance and make an exceptionally important contribution toward breaking out of the spiral of violence.
Received for publication February 6, 2001.
REFERENCES
