In the past decade research in psychosomatic medicine has taken new directions with new emphases. Although the transition has not been explicit and the change has been gradual, the departure from past traditions is nonetheless significant.
In many ways, the current thrust is in response to a growing awareness of the limitations of previous psychosomatic research models which were primarily correlational and retrospective in approach. For several decades psychosomatic research in man focused primarily on the pshychological correlates of already established illnesses defined as psychosomatic, that is, illnesses in which psychological factors were felt somehow to play a role in the predisposition, onset or clinical course of the disease. These previous approaches are well exemplified by the formulations of Franz Alexander and his colleagues.2 The Alexander school came to dominate psychosomatic thought in the 1940s and 1950s, stimulating much research, influencing practice and pervading even popular notions. It is worth reviewing their research in order to understand better the current critique of it.
Alexander and his colleagues made their original observations on a relatively small group of patients with a variety of medical illnesses who were in psychoanalysis and psychoanalytic therapy. Drawing upon the work of earlier psychosomatic theorists and investigators, Alexander modified their concepts and synthesized a persuasive formulation of the psychological processes active in the illnesses he defined as psychosomatic. It is important to recognize that these formulations were not only psychoanalytic but also psychophysiological hypotheses.
According to Alexander, Flanders Dunbar8 was correct in her notion of specific personality factors associated with particular illnesses, but what was specific was not the profile of superficial personality traits but rather a particular underlying, unresolved neurotic conflict. This conflict when it became intensified exacerbate or precipitate the illness. The psychosomatic mechanism involved was not conversion, however, as Felix Deutsch6 and his coworkers believed. Rather the affective component of the conflict, denied external expression, was discharged excessively into internal visceral and vegetative pathways of the type described by Walter Cannon, and this chronic internal discharge resulted eventually in somatic disturbances and organic lesions. Alexander also considered some constitutional organ vulnerability to be a necessary, though not sufficient, cause of psychosomatic illness.
In one of his most influential formulations, he reported that patients with peptic ulcer were revealed by psychoanalytic investigation to be in severe conflict about their oral dependent strivings, which were inconsistent with their ego ideals. These strivings, denied gratification and external expression, were discharged through the hypothalamus along autonomic pathways to the stomach, which behaved as if continually hungry for mother's milk, with excessive acid secretion and eventual ulceration. As another example, in ulcerative colitis, the disappointment, resentment and sense of loss in the depended-upon loved one could not be expressed; the patient responded by a desperate effort to please the loved one in the archaic way learned during toilet training - by producing bowel movement after bowel movement, leading to explosive diarrhea and eventual ulceration. Similar formulations were developed for rheumatoid arthritis, bronchial asthma, essential hypertension, Graves' Disease and neurodermatitis.
RESEARCH BASED ON ALEXANDER'S HYPOTHESES
Alexander's hypotheses, which he felt were strongly supported by his psychoanalytic data, inspired a torrent of psychosomatic research. Most of it was concerned with confirming his psychoanalytic correlations rather than his psychophysiological ideas. The culmination of the correlational research has only recently been published in the volumes on Psychosomatic Specificity^ after long delays occasioned by efforts to cope with the complex control issues involved. The study indicates that a panel of analysts trained in Alexander's formulations could correctly match edited transcripts of psychiatric interviews with various types of psychosomatic patients to the correct medical diagnosis over three times more frequently than chance- and, in fact, twice as successfully as a panel of internists. There were variations in the analysts' success rate from illness to illness, but their overall results were statistically significant.
Yet, despite the rigorous efforts to edit all medical clues from the interview transcripts, certain control problems gradually have been perceived as insurmountable in studying patients with well-established illnesses. A chronic illness seriously affecting an organ system and occupying a patient's concerns cannot help but have subtle yet pervasive somatopsychic effects, coloring perception and mental imagery with the psychological issues evoked by his symptoms. For example, what latent conflicts are likely to be stirred up by the need to deal with explosive, uncontrolled diarrhea or with constant hunger pangs and the need to keep drinking milk? Again, does not the patient who is continually and uneasily aware of his hypertensive potential also sense at some level that becoming angry will raise his blood pressure and possibly harm him? Such patients could be expected to develop inhibitions around expression of hostility much like the coronary patient who said, "I am at the mercy of the first scoundrel who can make me angry!" Patients also develop fantasies about their illnesses and their causes; the analyst is misled if he takes the patient's after-the-fact symbolization of his illness for the psychic mechanism which caused it. As another example, many patients respond to illness and hospitalization with a sense of personal failure and helplessness, feelings which may color their retrospective description of the life events preceding the illness; it would be easy to mistake this psychic reaction for the emotional state which actually helped precipitate the illness.
We can hardly dismiss all the psychological correlates of psychosomatic conditions as psychic responses to illness; this is particularly hard to do in the case of personality traits which, as best as can be determined retrospectively, were prominent long before the illness broke out. Yet, it has been apparent for some time that correlative psychosomatic research would have to move to prospective studies. If it were only possible to find a biochemical or physiological marker to identify patients likely to develop the illnesses before they actually become manifest!
Arthur Mirsky was among the first to realize the advantage, indeed the necessity, of such an approach. Utilizing the fact that plasma pepsinogen levels roughly correlate with gastric secretion of acid and pepsin and that hypersecretors of acid are highly at risk for development of peptic ulcers Weiner, Reiser, Mirsky34 and their co-workers screened a large number of soldiers for plasma pepsinogen concentration, and so identified a hypersecreting group. Although none of the soldiers had yet developed ulcers, it was possible with some success to select by psychological testing the high pepsinogen subjects from the controls, using Alexander's formulations. However, it was necessary to modify in a number of ways Alexander's concepts in order to obtain a significant separation of groups. What was especially impressive was that a high percentage of those soldiers with both the personality constellation and high pepsinogen levels subsequently developed manifest peptic ulcers during the stress of basic training. As Weiner33 has pointed out, these promising results really called for replication on another sample which has not as yet been done. Nevertheless, these preliminary observations led Mirsky21 to revise Alexander's formulation and to suggest that the constitutional physiological trait of hypersecretion might be associated from birth with the temperamental trait of strong oral striving; if psychological conflict around the oral striving developed, the likelihood of peptic ulceration would be increased.
"HOT SPOTS" IN WOMEN
In many ways, it is a pity that this significant prospective research in peptic ulcer has not been replicated or pursued. Other centers, however, have attempted to use a similar physiological marker technique for prospective psychosomatic studies. Wallerstein and his colleagues32 applied Alexander's formulation for psychological predisposition to Graves' Disease to women with areas of increased radioactive iodine uptake in the thyroid not as yet associated with any manifest clinical symptoms. They found that there was a tendency for women with these subclinical "hot spots" to be martyr-like people who from an unusually early stage in life had taken on heavy responsibilities, a finding which was consistent with the views of Alexander, Ham and their group. In one randomly sampled group of normals, the ability of the investigators to select women with the thyroid abnormalities from controls was highly significant although in other groups they were less successful. A subsequent longitudinal study31 of a group of women with the hot spots indicated that the spots tended to wax and wane with the degree of nonspecific life strain with the implication that frank hyperthyroidism could well occur in the future.
This data raises some puzzling questions, however. The type of hyperthyroidism likely to occur in women with hot spots is probably toxic nodular goiter, a very different illness in its endocrinology from the Graves' Disease originally studied by Alexander's group. Furthermore, the waxing and waning of the hot spots in association with nonspecific life stress does not exactly support the specificity hypothesis, that is, that a specific neurotic conflict gives rise to a particular illness.
PSYCHOANALYSIS AND PSYCHOSOMATIC SYNDROMES
In other words, although it may be possible to demonstrate, even prospectively, that patients with certain personality constellations also happen to be more prone to certain illnesses, it is another assumption that a particular neurotic conflict exerts a specific influence on the clinical course of a disease. From another point of view, accepting that states of acute emotional distress can somehow increase the likelihood of illness in a vulnerable organ and that different personalities become anxious in response to different stresses is not the same as claiming that a particular neurosis perpetuates a specific illness. Indeed, it is this aspect of Alexander's theory dealing with pathogenesis that has fared the worst in contemporary research. If a specific area of neurotic conflict were really pathogenic in these disorders, it should be possible to demonstrate that resolution of the conflict through psychoanalysis or psychotherapy would alter the long-term clinical course of the disease. This has not really been demonstrated, however, in part because of the failure to do adequate medical followups of psychotherapeutically-treated psychosomatic patients and in part because our understanding of the natural history of the illnesses under study is still inadequate.
There is still no good evidence (with adequate followu ? and adequate controls) that psychoanalysis favorably alters the ultimate course of peptic ulcer disease. Again, many analysts have claimed to have cured essential hypertension with analytic therapy. But nonmalignant essential hypertension, according to urrent concepts, is a disease which has at least two stages. First, there is a labile phase in which blood pressure may rise for days or weeks in association with dietary indiscretions, life stresses and other undetermined factors but following which the blood pressure typically normalizes in response to rest, sedation, diet and other supportive measures designed to relieve stress. Many such patients, however, will eventually enter a second, fixed phase in which the diastolic pressure remains permanently elevated in the absence of specific antihypertensive agents. It appears that many of the apparently positive analytic results have been temporary remissions obtained in the labile phase; there is no evidence as yet from long-term followup that such analyzed patients are less likely than controls to have subsequent exacerbations of hypertension or are less likely to enter the fixed stage of the disease with fixed elevation of blood pressure.
As another example, ulcerative colitis can take several forms, each with a different course and prognosis. One form has been recently recognized to be a separate illness altogether, Crohn's disease of the colon. Another type is from the beginning exclusively restricted to the proctosigmoid colon and runs a relatively benign course compared to the diffuse disseminated form. Even such a careful followup study of psychotherapeutically-treated ulcerative colitis patients as that conducted by Karush, O'Connor and their associates13 failed to distinguish adequately the medical types of ulcerative colitis under study or to include a control group of colitis patients treated without psychotherapy; and so it is hard to determine whether psychotherapy does have an enduring effect on the course of the disease in its various forms.
As more is understood of the pathophysiology of the psychosomatic illnesses, the less tenable seem Alexander's psychophysiological formulations. He spoke, for example, of the hostility of the rheumatoid arthritis patient producing increased tension in the joint muscles leading to arthritic lesions. We know now, however, that the muscle tension is secondary to the painful inflammation and that the primary lesion is certainly elsewhere, possibly in the cells of the synovial membrane. Alexander's psychophysiological formulation of ulcerative colitis was similarly based on a misconception. Engel9 was the first to point out that Alexander assumed that severe diarrhea is the cardinal symptom in ulcerative colitis with ulcerations of the mucosa occurring as a result; in fact, the initial observable lesion is a change in the mucosa associated with bleeding, and diarrhea is a secondary manifestation. If Alexander's formulations were correct, one would expect patients with the spastic colon-irritable bowel syndrome to go on to develop ulcerative colitis, which they do not. As yet another example, the role of sympathetic tone in essential hypertension, critical to Alexander's formulation, is now seen as only one element in an extremely complex pathophysiological system to which significant contributions are made by hormonal factors such as renin, angiotensin, aldosterone and so forth.
BREAKING FROM THE PAST FOR THE PRESENT
The real break between the psychosomatic research of the present and the past occurs at this point with the appreciation of the enormous complexities of the illnesses previously considered psychosomatic. More and more, the key research problem appears to be the need to clarify experimentally the pathophysiological mechanisms by which psychological states can affect somatic illness. In this respect, diseases like peptic ulcer, ulcerative colitis, neurodermatitis, rheumatoid arthritis and the like are seen as psychosomatically unresearchable at this time because medical research has not yet determined the nature of the primary disorders. How can one study the pathways by which emotional states lead to flareups of rheumatoid arthritis if one does not know what system is basically awry in that illness? Is rheumatoid arthritis an autoimmune disease, a latent viral infection, an enzymatic defect of the synovial membrane cells or a collagen disease? The strategies of psychosomatic research would be quite different in each instance. If one wanted to measure the relevant somatic response in the arthritic to a given psychological stimulus, one would hardly know whether to take a blood sample, a muscle biopsy or a sample of synovial fluid.
What has happened is that many psychosomatic researchers for the time being have shifted their attention away from complex disease entities and have focused instead on studying psychological influences on somatic systems whose responses can be measured, systems which may eventually prove to be implicated in somatic illnesses of many types, not just those designated by Alexander's school. The scope of this research has become quite large. Here we will describe just a few examples to illustrate some of these new directions.
One of the major psychosomatic pathways under study is the neuroendocrine apparatus. We have learned in the past 15 years that hypothalamic centers regulate the secretion of virtually all the hormones of the anterior pituitary,17 and it has also become clear that psychological stimuli can have important influences on thes'e neuroendocrine control mechanisms.18 Because the group of hormones from the pituitary and from their target glands affect virtually every biochemical process in the body, it is obvious that the neuroendocrine system could play a significant role in somatic disease.
THE ROLE OF THE PiTUITARY-ADRENOCORTICAL AXIS
The hormone system whose psychoendocrinology has been studied the most extensively is the pituitary-ad renocortical axis. It was demonstrated persuasively in the 1950s that psychological stress could profoundly stimulate release of ACTH and Cortisol in experimental animals with a doubling or tripling of the excretion of urinary corticosteroids and that this pathway could also be activated by psychological stress in man. Yet the adrenal responses in human subjects proved much less consistent and generally less intense than in experimental animals. It gradually became apparent that the animal model of stress research failed to take into account the elaborate system of psychological coping mechanisms including the classical defense mechanisms, which are highly developed in man and which can minimize the emotional impact of stressful stimuli and inhibit the associated adrenocortical response. During the 1 960s much of the psychoendocrine research on the adrenocortical responses to stress in man has been devoted to studying the role of this psychological buffering system.27
Studies of college students coping with the stress of hospitalization on a research ward,19 of parents coping with the stress of dying children,35 of soldiers coping with the stress of basic training,26 and of women coping with the stress of impending biopsy of breast tumors14 ail have yielded very similar findings. The majority of subjects defended rather effectively against the impact of the stresses and showed slight, if any, elevations in 24-hour corticosteroid excretion. On the other hand, that minority of subjects who were less well defended and experienced affective distress (anxiety and/or sorrow) did manifest concomitant increases in 24-hour corticosteroid excretion, usually about 25 to 35 per cent above baseline values. Indeed, experimenters have generally been able to predict with statistical success the degree of adrenocortical response in subjects by assessing the relative effectiveness or ineffectiveness of their coping mechanisms.
It has been further shown that when these protective ego mechanisms collapse altogether, as in the states of severe emotional turmoil and panic often seen in acute schizophrenic reactions29 and some psychotic depressive illnesses,28 the adrenocortical response can be massive indeed with two- or threefold increases in Cortisol production rate.
Even the 25 to 35 per cent increase in Cortisol secretion noted in poorly defended normals under stress is, however, of considerable biological significance. From the psychosomatic point of view, the potency of Cortisol in influencing electrolyte balance, carbohydrate metabolism and responses suggests that the adrenocortical stress response may eventually prove to be important in upsetting homeostasis in patients with congestive heart failure or diabetes or in altering susceptibility to Infection.
The observations on the psychological factors modifying the adrenocortical stress response in man may have broader psychosomatic significance as well if they are typical of the way certain other homeostatic physiological systems can be upset. The relevant psychological dimension would be the disruption of the normal psychological mechanisms for coping with stress, with concomitant release of anxiety, grief and other painful affects which have physiological correlates.
TESTOSTERONE, GROWTH HORMONE AND EPINEPHRINE
Studies of psychological factors affecting other hypothalami ically-control led hormones have also been revealing. There is growing evidence from studies on experimental animals18 that psychological stress can reduce the secretion of testosterone; the evidence for similar responses in man is as yet quite preliminary,25 but this should prove to be a promising future area of research.
The secretion of growth hormone, too, has been shown to be highly responsive to psychological Influences. Growth hormone in the adult is concerned with many factors other than growth per se, especially the regulation of sugar and fat metabolism. Acute psychological stress will frequently release a substantial burst of growth hormone in animals,24 and similar responses have been observed in man.10 It is interesting that certain psychological states have also been observed to be associated with an inhibition of growth hormone secretion. Children suffering from the maternal deprivation-failure to thrive syndrome have been found to lack the usual growth hormone response to hypoglycemia;23 the normal response frequently returns after a period of supportive care. A similar inhibition of growth hormone responses has been observed in some adults suffering severe depressive illnesses.30
The release of epinephrine from the adrenal medulla is also controlled by posterior hypothalamic centers and has long been known to be discharged in states of anxiety. Cannon was the first to describe in detail the physiological consequences of this hormonal response to a psychological state. In contrast to some of the other psychoendocrine responses whose clinical significance is not yet known, the release of epinephrine has been shown to have direct effects in certain disease states. In the patient with an acute myocardial infarction, bursts of epinephrine in response to anxiety are likely to provoke serious arrhythmias, and part of the immediate care of such patients involves protection from pain with analgesics and alleviation of anxiety with sedation and support. Epinephrine also mobilizes sugar and free fatty acids; an interesting recent study has suggested that adrenergic blocking agents can help significantly in the medical management of unstable juvenile and adolescent diabetics.4
Another significant area of current research on potential psychosomatic pathways concerns the instrumental conditioning of the autonomic nervous system. Until recently, it was generally believed that in terms of learning theory the ANS responded only to classical or Pavlovian conditioning techniques. However, the exciting work by Neal Miller and his colleagues has demonstrated that under certain conditions ANS responses can be shaped by instrumental (i.e., reward and punishment) methods with a precision that had not previously been suspected. Using ingenious methods involving stimulation in the reward centers in the brain tied to immediate feedback of physiological measures, these workers have shown that it is possible to shape in the curarized animal the ANS responses which control heart rate, blood pressure and even blood flow in specific organs.20 Thus, rats rewarded for upward or downward fluctuations in blood pressure can learn· to sustain blood pressure elevations or depressions to a statistically significant degree. Even more surprisingly, rats can be similarly trained to raise skin temperature in one ear and lower it in the other.7 Miller's colleagues are pursuing similar investigations of systems regulating pulmonary function, bowel motility and others.
While this research is only beginning to be applied to humans, it obviously has considerable relevance for both etiology and treatment of disease. It suggests, for example, that patterns of organ behavior can not only be affected by emotional states but also by experience with possible learning of both maladaptive and adaptive physiological response patterns. Patients with hypertension, for example, might be trained to lower blood pressure or asthmatic patients might learn to dilate bronchioles.
Related to these other areas of psychoendocrine and psychophysiological investigation are the burgeoning studies in developmental psychophysiology. Some of this research has as its premise that just as there appear to be critical periods in development when psychological patterns can be imprinted, there also may be critical periods for the imprinting of physiological response patterns. Several investigators1,15 are exploring the possibility that early experiences in the rat can affect the nature of subsequent adrenocortical responses to stress. Hofer11 is investigating the role of early maternal separation in the subsequent balance of parasympathetic and sympathetic tone in regulating the heart. In the human neonate, developmental psychophysiological research has focused on identifying some of the inborn constitutional differences between infants in psychophysiological reactivity,5·16 traits which may some day help account for the differences between people in their susceptibility to certain illnesses.
THE STUDY OF SLEEP PHYSIOLOGY
Yet another field of investigation which has potential for clarifying psychosomatic mechanisms is the study of sleep physiology. The stage of dreaming or rapid eye movement (REM) sleep is associated with profound psychophysiological activation of many systems. Of particular interest for the psychosomaticist are the findings of Kales12 that the peptic ulcer patient secretes the bulk of his nighttime acid during the REM periods, and the observations that patients with nocturnal angina episodes22 experience them almost exclusively during REM periods.
Although we have touched only briefly on a few of the new areas of psychosomatic research, it can be seen that the field appears to have moved far from what traditionally had been regarded as the sphere of psychosomatic medicine. These studies are making major contributions to our understanding of psychosomatic processes, both normal and abnormal. There is every reason to believe that, with increased medical understanding of the pathogenesis of diseases such as atherosclerosis and rheumatoid arthritis, a linkage will eventually be made between the two growing spheres of research. If so, we can look forward to the development of a true science of psychosomatic medicine.
1 . Ader, R. Adrenocortical function and the measurement of "emotionality." Ann N.Y. Acad. Sci. 159 (1969), 791-805.
2. Alexander, F. Psychosomatic Medicine. New York: W. W. Norton, 1950.
3. Alexander. F. et al. Psychosomatic Specificity. Vol I. Chicago: University of Chicago Press, 1968.
4. Baker. L. et al. Beta adrenergic blockade and juvenile diabetes. J. Pediat 75 (1969), 19
5. Bridger. W.; Birns. B. and Blank. M. A comparison of behavioral ratings and heart rate measurements in human neonates. Psychosom. Med. 27(1965). 123.
6. Deutsch. F., ed. On the Mysterious Leap from the Mind to the Body: A Workshop Study on the Theory of Conversion. New York: International Universities Press. 1959.
7 DiCara, L. Plasticity in the autonomic nervous system. In Neurosciences, Second Study Program, ed. F. O. Schmidt. New York: Rockefeller University Press, 1971. 218.
8. Dunbar. F. Emotions and Bodily Changes. New York: Columbia University Press. 1954
9. Engel, G. Studies of ulcerative colitis. II: The nature of the somatic process and the adequacy of psychosomatic hypotheses. Am. J. Med. 16 (1954), 416
10. Green. W. A. et al. Psychological correlates of growth hormones and adrenal secretory responses of patients undergoing cardiac catheterization. Psychosom Med. 32(1970). 599.
11 Hofer, M. A. and Weiner, H. Development of mechanisms of cardiorespiratory responses to maternal deprivation in rat pups. Psychosom. Med. 33 (1971), 353.
12. Kales. A. and Tau. T. Sleep alterations and medical illness. In Sleep: Physiology and Pathology, ed. A. Kales. Philadelphia: Lippincott. 1969, 148.
13. Karush, A. et al. Response to psychotherapy in chronic ulcerative colitis. I1 Psychosom. Med. 30 (1968), 255; II. Psychosom. Med. 31 (1969), 201.
14. Katz, J. L. et al. Psychoendocrine aspects of cancer of the breast. Psychosom. Med 32 (1970), 1. 15. Levine, S. The pituitary-adrenal system and the developing brain. In Progress in Brain Research, Vol. 32. Pituitary. Adrenal and the Brain, ed. D, DeWied and J, Weijnen. New York: Elsevier. 1970.
16. Lipton, E. L; Steinschneider, A, and Richmond. J. B. Autonomic function in the neonate: VII: Maturational changes in cardiac control Child Dev. 37 (1966). 1.
17. Martini. L. and Ganong, W. F., eds. Neuroendocrinology. New York: Academic Press. 1966.
18. Mason. J. The organization of psychoendocrine responses, Psychosom. Med. 30(1968) 565
19. Mason, J. et al. Corticosteroid responses to hospital admission. Arch. Gen. Psychiat. 13 (1965), 1,
20. Miller, N. Learning of visceral and glandular responses. Science 163 (1969). 434.
21. Mirsky. I. A. The psychosomatic approach to the etiology of clinical disorders. Psychosom Med 19 (1957), 424.
22. Nowlin, J. B, et al. Association of nocturnal angina pectoris with dreaming Ann. Int. Med. 63 (1965), 1040.
23 Powell, G F et al. Emotional deprivation and growth retardation simulating idiopathic hypopituitarism. N. Eng J Med. 276(1967). 1279
24 Reichlin S. Hypothalamic control of growth hormone and the response to stress. In Endocrinology and Human Behavior, ed. R Michael London. Oxford University Press. 1968, 256
25. Rose. R. et al Androgen responses to stress. Psychosom. Med 30(1969). 418
26. Rose. R. M.. Poe. R O and Mason, J. W. Psychological state and body size as determinants of 17-OHCS excretion. Arch. Int. Med. 121 (1968), 406.
27. Sachar, E. J. Psychological factors relating to activation and inhibition of the adrenocortical stress response in man A review. In Progress in Brain Research, Vol. 32. Pituitary, Adrenal and the Brain, ed. D. DeWied and J. Weijnen. New York: Elsevier, 1970.
28. Sachar, E. J. et al. Cortisol production in depressive illness, a biochemical and clinical clarification. Arch. Gen Psychiat. 23 (1970). 289
29. Sachar. E. J. et al. Psychoendocrinology of ego disintegration. Am. J. Psychiat. 126 (1970). 1067-78.
30. Sachar. E. J.; Finkelstein, J. and Hellman. L. Growth hormone responses in depressive illness. Arch. Gen. Psychiat 25(1971), 263
31. Voth, H. et al. Thyroid "hot spots": their relationship to life stress. Psychosom. Med. 32 (1970), 561 .
32 Wallerstein. R. S. et al. Thyroid "hot spots": a psychophysiological study. Psychosom. Med 27 (1965). 508.
33. Weiner, H. Current status and future prospects for research in psychosomatic medicine. J. Psychiat. Res. S (1971). 479.
34. Weiner, H. et al. Etiology of duodenal ulcer. I: Relation of specific psychological characteristics to rate of gastric secretion (serum pepsinogen). Psychosom. Med. 19 (1957). 1.
35. Wolff C. T.; Hofer, M. A. and Mason. J. W. Relationship between psychological defenses and mean urinary 1 7-hydroxycorticosteroid excretion rates. Psychosom. Med. 26 (1964). 576.