SQU Med J, August 2010, Vol. 10, Iss. 2, pp. 180-186, Epub. 19th Jun 10 Submitted: 20th Feb 10 Revision Req.25th Apr 10, Revision Recd. 8th May 10 Accepted: 9th May 10 Somatisation is generally defined as the tendency to experience psychological distress in the form of somatic symptoms and to seek medical help for these symptoms, which may be initiated and/or perpetuated by emotional responses such as anxiety and depression. Multiple or unexplained physical symptoms cause substantial disability in patients, excess use of medical services, disappointment for therapists, and frustration for physicians. Somatisation is a common problem in primary health care leading to disproportionately heavy demands on health services. Patients with somatisation account for about 20% of the work load in general practice.1,2 Presenting psychiatric distress with physical symptoms has been recognised as one of the important factors associated with the reduced recognition and treatment of depression in primary health care.3 While their psychiatric distress remains unrecognised and untreated, patients with somatisation undergo long and costly investigations.4 This article aims to review the literature on the development of the concept of somatisation and how this definition is applied, from a cross-cultural perspective. Department of Family Medicine & Public Health, Sultan Qaboos University Hospital, Muscat, Oman Email: zakiyaq@squ.edu.om مفهوم اضطراب اجلسدنة من منظور ثقافات متعددة زكيه البو�شعيدية لتلك الطبية امل�شاعدة وطلب ج�شدية اأعرا�س هيئة على النف�شية بال�شغوط ال�شعور اىل املري�س نزعة اأنها على اجل�شدنة تعرف امللخ�ص: األعرا�س. وقد تبداأ اأو تت�شاعف هذه الأعرا�س بوجود القلق ااأو الكتئاب. تعترب اجل�شدنة م�شكلة منت�رشة ب�شكل وا�شع خ�شو�شا يف الرعاية ال�شحية الأولية. وتوؤدي هذه امل�شكلة اإىل تعدد زيارات املري�س ملراكز اخلدمات ال�شحية وت�شبب الإحباط للمري�س والطبيب. خالل ال�شنوات وتطبيقاتها با�شتخداماتها يتعلق فيما خ�شو�شا له، الطبية والت�شنيفات وتعريفه اجل�شدنة مب�شطلح مرتبط ر�شى عدم ظهر املا�شية ومالءمتها للمجتمعات غري الغربية. هذه امل�شكلة مرتبطة بتنوع التعاريف، خا�شة الفائدة والكفاية والتطبيق. تهدف هذه الدرا�شة ملراجعة الأدبيات اخلا�شة يف تطور مفهوم اجل�شدنة وكيف يطبق التعريف يف خمتلف املجتمعات. مفتاح الكلمات: ا�شطراب اجل�شدنة، مقارنة بني احل�شارات، الكتاب الت�شخي�شي والح�شائي لالأمرا�س النف�شية، الت�شنيف الإح�شائي الدويل لالأمرا�س، عمان. abstract: Somatisation is generally defined as the tendency to experience psychological distress in the form of somatic symptoms and to seek medical help for these symptoms, which may be initiated and/or perpetuated by emotional responses such as anxiety and depression. Somatisation has been recognised as a commonly encountered problem, especially in primary health care, contributing to frequent use of medical services and to frustration in both the patient and the doctor. In recent years, there has been a great deal of dissatisfaction with the terminology and classification of somatisation, and the way this definition is applied to non-Western cultures. This dilemma pertains to different aspects of the definitions, mainly their usefulness, adequacy and applicability. This article aims to review the literature on the development of the concept of somatisation and how this definition is applied from a cross-cultural aspect. Keywords: Somatization disorders; Cross-cultural comparison; Oman; Diagnostic and Statistical Manual of Mental Disorders; International Classification of Diseases. review The Concept of Somatisation A Cross-cultural perspective Zakiya Q Al Busaidi Zakiya Q Al Busaidi review | 181 neurological symptoms.12 These classifications in the DSM-III and DSM- IV have been criticised for being confusing and lacking a satisfactory organising principle. The combination of tightly defined specific categories and large vague non-specific subcategories was described as misleading.8 This method of categorisation has resulted in most of the literature concentrating on rare phenomena and ignoring more common forms of somatisation, especially those presenting to primary health care. Studies have shown that in primary health care patients with undifferentiated or not otherwise specified somatoform disorders are far more common than those with a named disorder.13,14 The concept of somatisation presented in the above definitions is based on the assumption that psychological symptoms of depression are more central than somatic symptoms. This assumption has being challenged lately by several studies, including the World Health Organization international study, in which the findings suggest that somatic symptoms are a core component of the depressive episodes.15 Another reason given for the limited acceptance of the present definitions is that the concept has been based on the mind-body dualism characteristic of Western medicine. In addition, it implies a causal relationship between psychological distress and its physical manifestation, an approach that is not used in a number of other cultures where the mind-body dichotomy cannot be applied in diagnostic and therapeutic approaches to the patient. The use of the concept in other cultures leads to a problem of translation and communication bearing in mind that the term “somatoform disorders” is difficult to translate into many languages, and that terminology can have important implications for treatment and outcome.14 In Arabic, for example, the word soma is translated as jasad, which means body, but there is no word for somatisation in Arabic dictionaries. In the medical literature, the Arabic term used to indicate somatisation is tajseed. When translated back to English, this word means 'embodiment' or 'embodying'.16,17 In addition, the classifications are not clear with regard to the possibility of these syndromes presenting as primary conditions, or whether they are only linked to psychopathology, i.e. to diagnoses Current Definitions of Somatisation A number of definitions and categories have been proposed for somatisation; among these are the Classification of Mental and Behavioural Disorders section of the International Classification of Diseases (ICD) and the Diagnostic and Statistical Manual of Mental Disorders (DSM). The ICD-8 and ICD-9 referred to psychic factors contributing to the pathogenesis of physical illnesses without associating them to psychosocial events or calling them psychosomatic.5,6 In the ICD-10, somatisation is defined as multiple, recurrent and frequently changing physical symptoms usually present for several years, (at least 2 years) before the patient is referred to a psychiatrist. It is associated with a long complicated history, negative investigations and fruitless operations with the patient’s refusal to accept the advice or reassurance of doctors. Marked depression and anxiety are frequently present and may justify specific treatment.7 Later, the term “unexplained somatic complaints” was introduced in the primary care version of the ICD-10 to describe patients presenting with any physical symptom and frequent medical visits in spite of negative investigations.6 This description is more practical as it has the advantage of having no aetiological implications. In addition, it was proposed to be less intrusive and more acceptable to patients. In spite of this, the term is not without cultural bias and uses negative investigations as an indicator to rule out organic disease.8,9 Somatoform disorders were first introduced in the DSM-III as a category for patients whose presenting symptoms suggest physical illness in the absence of demonstrable organic findings, and lack any known pathophysiological mechanisms. In the DSM-IIIR this new category was developed further to include seven disorders: body dysmorphia; conversion; somatisation; somatoform pain; undifferentiated somatoform disorders; somatoform disorders not otherwise specified, and hypochondriasis.10,11 In the DSM-IV, somatisation was defined as a poly-symptomatic disorder that begins before the age of 30 years, extends over a period of years, and is characterised by a combination of pain, gastrointestinal tract, sexual and pseudo- The Concept of Somatisation A Cross-Cultural Perspective 182 | QU Medical Journal, August 2010, Volume 10, Issue 2 of anxiety and depression. For example, the criteria defined by Goldberg and Bridges include psychiatric disorder as a necessity for diagnosing somatisation.1 In their criteria, they argue that a patient with somatisation must seek medical help for the somatic manifestation of psychiatric illness and does not in fact have psychological symptoms. Instead, these symptoms are attributed to a physical problem. In the previous definitions, somatisation is viewed as a way of presenting psychopathology. In their current form, these definitions of somatisation reflect the increasing emphasis in psychiatric nosology on creating distinct entities by turning forms of distress into disorders and labelling them.14 However, several authors have called for differentiation between expression of distress, primary somatisation and a formal psychopathologic state.12 In addition, recent studies have suggested that persistent somatic symptoms can occur without there being any conventional psychiatric disorder.9 Moreover, some authors prefer not to consider somatisation as a diagnosis, but as a behavioural phenomenon in which there is a disturbance of the sick person role of the patient and the healer role of the doctor.18 For this reason Cooper19 suggested that “In future studies, it will be important to examine social and cross-cultural issues by involving medical anthropologists from the start”, as the sick person role is partly determined by cultural factors. As a result of all the controversies mentioned above, a new term, "medically unexplained symptoms", has emerged in recent literature and is becoming widely acceptable. Medically unexplained symptoms (MUS) are defined as those physical symptoms having little or no basis in underlying organic disease.18 When organic disease exists, the symptoms are inconsistent with or out of proportion to the disease.20 People with MUS are not necessarily abnormal. Many people exhibit MUS, but seldom seek care.20 MUS becomes a problem when it leads to frequent health care seeking for feared but nonexistent physical illness.21 The term ‘medically unexplained symptoms’ is useful, over and above somatisation, because it represents a broader concept; has no aetiological implication; includes the full spectrum of severity, and is consistent with the presence of concurrent organic and psychiatric illness causing these symptoms.8 However, the term has been criticised for being based on a negative statement that withholds from the patient a positive explanation for their symptoms. The term implies that the symptom are either organic (medically explained) or medically unexplained which reinforces the mind-body dualism concept. In addition, it covers a very large group of patients since it does not include information about the severity and duration of the symptoms and thus their clinical significance;22 hence, the term still needs to be researched further especially for application in non-Western cultures. The DSM and ICD are currently being revised. The DSM-V is due to be released in 2013 while the ICD-11 release is expected in 2015. The DSM-V workgroup is proposing a series of changes to somatoform disorders. First, the disorders would be grouped together under one term entitled “Somatic Symptom Disorders”, which would include somatoform disorders, factitious disorders, and psychological factors affecting a medical condition. Second, because of their many common features, the group is proposing that hypochondriasis, pain disorders, somatisation disorders, and undifferentiated somatoform disorders be grouped together as Complex Somatic Symptom Disorders.23,24 The proposed classification has the advantage of shortening the duration of symptoms to six months. Moreover, it allows a diagnosis of somatic symptom disorder in addition to a general medical condition, whether the latter is a well-recognised organic disease or a functional somatic syndrome such as irritable bowel syndrome or chronic fatigue syndrome. However, professionals in the field have been voicing concerns that mind-body dualism remains inherent in the proposed criteria. Furthermore, the proposals for broadening of the criteria would bring many more patients under a possibly stigmatising mental illness diagnosis.25,26 Cross-cultural Aspects of Somatisation Culture is defined as the accumulated beliefs, practices, attitudes, and values shared by a social collective. It is the lens through which a person registers experiences that shape his/ her perceptions, understanding and reactions to events.27 In the perspective of health and illness, Zakiya Q Al Busaidi review | 183 attributions and illness behaviour might differ across cultures due to culture-specific characteristics.31,38,39 Moreover, epidemiological data has not shown major cross-cultural differences, and community surveys have shown a similar balance between psychological and somatic symptoms in Western and non-Western countries. In addition, clinical studies have shown that the somatic presentation of depression is common in all cultures.40 Recent research on mental disorders in primary care increasingly suggests that somatisation, instead, appears to be the norm in primary care across cultures.15 Moreover, while some authors have claimed that less sophisticated subjects are more likely to somatise, others have noted the absence of social and cultural differences between somatisers and psychologisers.41 This evidences suggest that differences cross-culturally and between classes exist at the level of presentation rather than with the experience of somatic symptoms. One of the major studies on the cross-cultural aspects of somatisation is the WHO collaborative study of psychological problems in general health care to examine the relationship between current somatic symptoms and current psychological distress. The study involved data from 15 participating sites in 14 countries and examined a large number of subjects (5,438) using the General Health Questionnaire (GHQ) and the Composite International Diagnostic Interview.39 The results of the study showed a generally similar pattern of association between physical symptoms and psychological distress among patients from Western and non-Western countries indicating that cultural factors might not affect the rate of somatisation, but might influence other factors related to it such as subsequent illness behaviour.14 The study found that somatisation is common in many countries, but its frequency depends on how the phenomenon is defined. Reports from the Arab World show a high incidence of patients complaining of depression and anxiety and a variety of somatic symptoms;42,43,44 however, most of these reports have relied on clinical impressions and psychological scales depending on items eliciting psychological rather than somatic symptoms. Generally, there has been little published on somatisation in the Arab World. Among the culture is recognised to play a fundamental role in defining the states that comprise health and illness, shaping the expression of somatic and psychosocial distress, identifying the aetiological factors believed to be responsible for illness, and shedding light on culturally appropriate help-seeking behaviour and social stigma.27,28 Kleinman and Good29 state that culture and society shape the meanings and expressions people give to sadness, unhappiness, hopelessness and lack of pleasure. They add that there are dramatic differences in the way different cultures express physical complaints associated with depressive illness. In addition, cultural factors determine which symptoms or signs are normal or abnormal. They act to mould diffuse emotional and physical changes into a pattern which is recognisable to the person suffering from it and to people around him/her.30,31 The presentation of symptoms to the health care system is also affected by culture, as patients are selective regarding what to present to the doctor. They select verbal presentation elements which they believe are relevant and culturally acceptable.32 Kleinman,33 in addition to other authors, explains that somatisation as an "idiom of distress" is more common in cultures where stigma is connected with psychiatric problems and the expression of emotional distress is inhibited.33,34,35 Other explanations given to support this point of view suggest that patients from non-Western cultures are less willing or less able to express emotional distress as they may lack the capacity to label an emotional state.32 This inability to express emotions has been suggested to be due to vocabulary inadequacies in the original languages of these communities. A series of studies by Mumford et al.36 support the traditional view mentioned above. According to the authors, many patients presenting to medical outpatient clinics in Pakistan complained of unexplained somatic symptoms. Another series of studies in China and Taiwan by Kleinman,37 found that somatisation was the main language of distress and was used to express psychological disorders. The traditional claim that somatisation is more typical of patients from non-Western and developing countries is, however, strongly challenged by contrary evidence. There is a growing body of evidence that somatisation is a universal phenomenon. However, the presentation, The Concept of Somatisation A Cross-Cultural Perspective 184 | QU Medical Journal, August 2010, Volume 10, Issue 2 few studies, a screening study using the Patient Health Questionnaire in Saudi Arabia identified the prevalence of somatisation to be 19.3%.45 On the other hand, a study by El-Rufaie et al.46 showed that 48% of a sample of psychiatrically referred Arab primary care patients presented with predominantly somatic expressions of mental disorders. This contrasts with 42% who presented with predominantly psychological expressions. This study showed that the patients involved use both psychological as well as somatic language to express their distress. A study assessing the prevalence of somatisation and psychological morbidity in patients presenting to primary health care in Oman showed that both somatisation and psychological morbidity are relatively common. The study population equally experienced psychological distress as well as somatic distress.16 These finding challenges the old belief that people in developing countries express their distress mainly somatically. The Challenges facing Cross-Cultural Studies Researchers studying disorders across cultures face the dilemma of applying the definition of a disorder derived from one culture context to others in which their applicability might be doubtful. This is especially true regarding somatisation because somatisation has often been seen from a number of different perspectives. Due to the diverse ways in which the phenomenon is conceptualised, research findings have tended to be conflicting. Both diversity and conflicting results are evident as regards the different aspects of the phenomenon presented in the literature. In addition to the diversity of definitions, there is a diversity of settings in which these studies were conducted, as somatisation in community and primary health care patients might differ from that of psychiatric patients. This diversity of definitions and settings makes it difficult to generalise or interpret accurately the results of these studies across different cultures.39 In addition to the differences in the definitions used in various studies, there are differences in the settings of these studies, which make the generalisation of the results rather difficult. Although psychological morbidity has been shown to be very high among patients seeking primary health care, it has been suggested that, among psychiatric outpatients, somatisation may represent a more severe and chronic phenomenon. This is in comparison to the rather acute and more self- limiting form that presents in primary health care and in the community;39 however, most studies have focused on patients at the most severe end of the spectrum of the phenomenon. Other studies show a difference in the characteristics of somatising patients between those who present to hospital clinics and those presenting to primary health care. Cross-cultural studies reveal specific biases in Western patterns of causal attribution that are not shared by many other cultures. For example, people in Western societies tend to overestimate the importance of the individual and of personality traits, and underestimate the importance of situational factors in making attributions for behaviour. However, this bias is culture-specific, since people in India, China and Japan, for example, are more likely to attribute behaviour to situational or contextual factors.47 Kirmayer et al.47 explain that one of the reasons for the impression that Western patients usually express psychiatric distress in terms of emotional symptoms is related to the fact that most of the research in the past was based largely on tertiary care patients. Hence there is a need to differentiate between studies done on community and primary health care patients and those on patients in psychiatric clinics or tertiary care. In addition, most of the studies published on somatisation use quantitative research methods while qualitative research may be more suitable for eliciting patients’ own beliefs and help-seeking behaviour across cultures. Conclusion There is a need to reform the present classifications and definitions of somatisation aimed at finding better and more practial nosological solutions that take cultural variety more into account. Such reforms have to be based on extensive research on the subject, in addition to literature review of more recent research. Adopting a cross-cultural perspective in addressing psychosomatic questions can benefit both basic knowledge and practical health care applications. From the perspective of basic knowledge, attention to cross-cultural Zakiya Q Al Busaidi review | 185 12. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th Ed. Washington DC: American Psychiatric Association, 1994. 13. Lloyd G. Somatisation: A psychiatric perspective. J Psychosom Res 1989; 33:665–9. 14. Kirmayer LJ, Robbins JM. Somatoform disorders: Personality and the social matrix of somatic distress. J Abnorm Psychol 1994; 58:481–8. 15. Simon GE, Gater R, Kisely S, Piccinelli M. Somatic symptoms of distress: An international primary care study. Psychosom Med 1996; 58:481–8. 16. Al-Lawati J, Al-Lawati N, Al-Siddiqui M, Anthony SX, Al-Naamani A, Martin RG, et al. Psychological morbidity in primary health care in Oman: a preliminary study. SQU J Sci Res: Med Sci 2000; 2:105–10. 17. Baalabki M, Baalabki R. Al-Mawrid dictionary: English- Arabic, Arabic-English. Beirut: Dar-El-Illm Lilmalayin, 1999. 18. Escobar JI. Transcultural aspects of dissociative and somatoform disorders. Psychiatr Clin North Am 1995; 18:555–69. 19. Cooper JE. The classification of somatoform disorders in ICD-10. New York: Springer, 1999. 20. Kellner R. Hypochondriasis and somatization. JAMA 1987; 258:2718. 21. Verbrugge LM, Ascione FJ. Exploring the iceberg. Common symptoms and how people care for them. Med Care 1987; 25:539. 22. Creed F, Guthrie E, Fink P, Henningsen P, Rief W, Sharpe M, et al. Is there a better term than “Medically unexplained symptoms” J Psychosom Res 2009; 68:5–8. 23. Proposed DSM-V criteria. From www.dsm5.org. Accessed Apr 2010. 24. Dimsdale J, Creed F, and on behalf of the DSM-V Workgroup on Somatic Symptom Disorders. The proposed diagnosis of somatic symptom disorders in DSM-V to replace somatoform disorders in DSM- IV—a preliminary report. J Psychosom Res 2009; 66:473–6. 25. The proposed DSM-V revision. From www.dsmwatch. wordpress.com. Accessed Apr 2010. 26. Schroder A, Fink P. The proposed diagnosis of somatic symptom disorders in DSM-V: Two steps forward and one step backward? J Psychosom Res 2010; 68:95–100. 27. Devins GM. Culturally informed psychosomatic research. J Psychosom Res 1999; 46:519–24. 28. Katon W, Ries RK, Kleinman AM. The prevalance of somatization in primary care. Compr Psychiatry 1984; 25:208–15. 29. Kleinman AM, Good B. Introduction: Culture and depression. Berkeley: University of California Press, 1985. questions can help to differentiate fundamental principles from culturally specific ones. Moreover, there is a growing awareness of the importance of integrating culturally relevant considerations in medical management and in education and training strategies so as to provide more effective health care.27,38 Acknowledgements I would like to thank Sultan Qaboos University, Oman, for granting me a scholarship to study for my PhD at the University of Nottingham (2001- 2005), on which part of this article is based. I would also like to thank my PhD supervisor, Dr. Margaret Oates, for her guidance and support. References 1. Bridges RN, Goldberg DP. Somatic presentation of DSM-III psychiatric disorders in primary care. J Psychosom Res 1985; 29:563–9. 2. Craig TK, Boardman AP, Mills K, Daly-Jones O, Darke H. The south London somatisation study. I: Longitudinal course and the influence of early life experience. Br J Psychiatry 1993; 163:579–88. 3. Cape J, McCulloch Y. Patients' reasons for not presenting emotional problems in general practice consultations. Br J Gen Pract 1999; 49:875–9. 4. Kirmayer LJ, Robbins JM, Dowrkind M, Taffe MJ. Somatisation and the recognition of depression and anxiety in primary care. Am J Psychiatry 1993; 150:734– 41. 5. Kramer M, Sartorius N, Jablensky A, Gulbinat W. The ICD-9 classification of mental disorders: A review of its development and contents. Acta Psychiatr Scand 1979; 59:241–62. 6. Sartorious N. Classification: An international perspective. Psychiatr Ann 1976; 6:22–35. 7. The ICD-10 classification of mental and behavioural disorders: Diagnostic criteria for research. Geneva: WHO, 1992. 8. Barsky AJ, Ettner SL, Horsky J, Bates DW. Resource untilization of patients with hypochondriacal health anxiety and somatization. Med Care 2001; 39:705. 9. Kendler KS. A psychiatric dialogue on the mind-body problem. Am J Psychiatry 2001; 158:989–1000. 10. Diagnostic and Statistical Manual of Mental Disorders. 3rd Ed. American Psychiatric Association; Washington DC: 1980. 11. Diagnostic and Statistical Manual of Mental Disorders. 3rd Ed, revised. Washington DC: American Psychiatric Association, 1987. The Concept of Somatisation A Cross-Cultural Perspective 186 | QU Medical Journal, August 2010, Volume 10, Issue 2 30. Helman C. Culture, Health and Illness. Oxford, UK: Butterworth-Heineman, 1990. 31. Kirmayer LJ. Culture, affect and somatisation. Transcult Psychiatr Res Rev 1984; 21:159–88. 32. Lipowski ZJ. Somatisation and depression. Psychosomatics 1990; 31:13–21. 33. Kleinman AM. Depression, somatisation and the "new cross-cultural psychiatry". Soc Sci Med 1977; 11:3–10. 34. Kleinman AM. Culture and depression. Cult Med & Psychiatr 1978; 2:295–6. 35. Mechanic D. Social and psychological factors affecting the presentation of bodily complaints. N Engl J Med 1972; 286:1132–9. 36. Mumford DB, Bavington JT, Bhatnagar KS. The Bradford somatic inventory: A multi-ethnic inventory of somatic symptoms reported by anxious and depressed patients in Britain and the Indo-Pakistan subcontinent. Br J Psychiatry 1991; 158:379–86. 37. Kleinman AM. Patients and healers in the context of culture, Berkeley: University of California Press, 1980. 38. Ono Y, Janca A. Rethinking somatoform disorders. J Psychosom Res 1999; 46:537–9. 39. Gureje O, Simon GE, Ustun TB, Goldberg DP. Somatisation in cross-cultural perspective: A World Health Organisation study in primary care. Am J Psychiatry 1997; 154:989–95. 40. Simon GE, VonKorff M, Piccinelli M, Fullerton C, Ormel J. An international study of the relation between somatic symptoms and depression. N Engl J Med 1999; 341:1329–35. 41. Bridges K, Goldberg DP, Evans B, Sharpe T. Determinants of somatization in primary care. Psychol Med 1991; 21:478–83. 42. Hamdi TI, Al-Hasani L, Mahmood A, Al-Hussaini AA. Hysteria: A large series in Iraq. Br J Psychiatry 1981; 138:177–8. 43. Racy J. Somatisation in Saudi women: A therapeutic challenge. Br J Psychiatry 1980; 137:212–6. 44. Zaidan Z, Alwash R, Al-Hussaini AA, Al-Jarrah M. Psychiatric morbidity in northern Jordan: A ten-year review. Med Sci 2000; 35–40. 45. Beker S, Al-Zaid KH, Al-Faris E. Screening for somatisation and depression in Saudi Arabia: A validation study. Int J Psychiatry Med 2002; 32:271– 83. 46. El-Rufaie OEF, Al-Sabosy MMA, Bener A, Abuzeid MSO. Somatised mental disorder among primary care Arab patients: 1. Prevalence and clinical and sociodemographic characteristics. Psychosom Med 1999; 46:549–55. 47. Kirmayer LJ, Young A, Robbins JM. Symptom attribution in cultural perspective. Can J Psychiatry 1994: 39:584–95.