How scientifically valid is the knowledge base of global mental health?BMJ 2008; 336 doi: https://doi.org/10.1136/bmj.39513.441030.AD (Published 01 May 2008) Cite this as: BMJ 2008;336:992
All rapid responses
I am pleased that Fekadu and Cleare have contributed (Rapid Response 14 May). It is crucial to debate the nature of the Western-led psychiatric enterprise, which presents itself as a benign, progressive endeavour based on scientific principles and with universal application.
Their contribution is particular welcome in that it provides bmj.com readers with the other side of the coin: the forms of reasoning and justification that underpin the commanding heights of the orthodox psychiatric mainstream.
Fekadu and Cleare write that "we agree with Summerfield that until firm etiological understanding becomes available to support the diagnosis of psychiatric disorders, their validity remains less certain" and concede that the main supposedly international classification systems represent Western frameworks of understanding of mental disorders "in some respects". But these are just ritual genuflections, and are not meant to introduce doubt about current practice nor the biomedical triumphalism this trades on. Their response is not an answer to the conceptual fallacies to which my paper pointed, rather a re-affirmation of business as usual.
They argue that "the consensus has been that the many benefits of applying the classification systems outweigh the disadvantages”. Whose consensus exactly? Anyone other than researchers and clinicians in the Western mental health industry? They argue further that the development of the 4th edition of DSM involved the participation of "1000 experts". How many of these "experts" thought it relevant to incorporate the enormous diversity of understandings that non-Western societies bring to bear on life, and on their definition of a person? How many of them had the anthropological and ethnographic grip to enable them to do so, even if they had wanted to? In DSM “our” categories are considered universal, whereas “theirs” are merely culture-bound. This is imperialistic thinking.
Fekadu and Cleare state that diagnostically standardised psychiatric disorders are "conditions that emerge with the documented history of mankind". This is in effect to assert that the remarkable inflation in the official number of psychiatric categories in DSM (from 104 categories in the 1st edition to 357 in the 4th) simply reflects the unearthing of timeless disorders that had always been there. Can anyone seriously believe this?
They write that "technically sound research method is the essential starting point for valid research". This is an assertion, an ideological declaration; it is not a finding. It assumes that ‘mental disorder’ can be seen as essentially outside society and culture, rather as the American Psychiatric Association avers that mental disorders will eventually boil down to brain disorders. The claim that local categories in various non- Western societies can be considered "psychiatric diagnostic equivalents" is one that relieves Western psychiatry of any obligation to examine the limits of its knowledge and epistemological traditions. "Psychiatric diagnostic equivalents" arise ineluctably from our determination to use Western categories as the basic template. Quantitative research methods based on Western paradigms are taken to other parts of the world, generating ‘findings’ which are interpreted within those same paradigms: there is a well remarked on circularity of thinking here.
As a senior professor of psychiatry remarked to me a year or two ago, validity has been sacrificed on the altar of "reliability". Psychiatric reliability means the deployment of standardised methods, but if these methods are based on invalid paradigms – my core contention- the very ground the whole enterprise stands on is unsound. Lack of validity cannot be redeemed by “reliable” methods.
Fekadu and Cleare assert that "many psychiatric disorders are associated with identifying neurochemical and anatomical changes". This is I am afraid an example of the self-aggrandising exaggerations I associate with the claims of the biomedical mainstream. Could they offer readers a single example that supports this claim. Schizophrenia? Depression? Post-traumatic Stress Disorder? (We are not talking about the handful of organic cerebral conditions with well-defined features, I presume. These aside, there is no evidence of specific and clinically significant neurochemical or anatomical associations for any particular diagnostic category in the entire psychiatric canon).
They state that "the most important issue in cultural validity of psychiatric instruments appears to be careful translation to make sure that the questions are understandable to participants.” So (non-Western) participants must understand “us”’, rather than the other way round. The problem in cross-cultural research is not accurate translation between languages, but accurate translations between worlds!
They defend the quality of the mainstream published literature, yet as I mentioned in the paper, at least 80% of all research publications on refugees and mental health (and I think the percentage would be still higher in the highest impact journals) are based entirely on Western instruments and categories. We are to assume that "depression" in Camberley or Canada is the same as "depression" in Cambodia. To how many Cambodians would this make sense? The statement that "there is no evidence to negate the validity of psychiatric research in developing countries" is, as I say, simply a call for business as usual, relegating non-Western viewpoints and realities to the status of mere epiphenomena.
The Rapid Response by Dr Kadri (18 May) makes some relevant points in this connection regarding “depression” in India..
I am afraid I must reiterate that in large part the published research literature is not worth the paper it is printed on; doubly unfortunate, then, that the paper it is printed on are the pages of Western medical and psychiatric journals, for this is what passes as the knowledge base of global mental health. Those authors asserting solemnly in the Lancet Global Mental Health series last September that every year up to 30% of the global population would develop a mental disorder appear to be living in a parallel universe. I agree that I am questioning the value of a lot of research in Western countries too. Are Fekadu and Cleare comfortable with the claims, endlessly recycled, that as many as one in four or six of the UK population are carrying a mental disorder at any one time?
Lastly, readers might care to review a telling paper by Skultans (2003) on the internationalisation of “depression”, not to the non-Western world, but to post Soviet bloc Europe- in this case Latvia. (1) Psychiatric language in Latvia has been invaded by the diagnosis of "depression", key promoters being the translation into Latvian of the ICD and the conferences organised by pharmaceutical companies to educate psychiatrists and GPs about the new diagnostic categories (who in turn educate their patients). This is a radical departure from the older language of somatic distress that was central to both Soviet psychiatry and to lay conceptualisations of distress. “Depression” is displacing the discourse of nervi, which was salient both in lay discourse and in what was brought to a medical consultation. For a doctor to ask about nervi was to invite a life story, to illuminate temporal and social aspects of the self. The doctor-mediated shift from nervi to depression, a move from somatoform to psychological framing, represents the internalisation of a heightened sense of accountability for one life's circumstances. Yet this is at the very moment when for most people the post-Soviet Latvian economy and society has lost much of its former sense of stability and security, with few opportunities for real change in life situation. Previously, talking about nervi was in part a critical commentary on Soviet dominated society, economics, etc, with disorder and unpredictability attributed to the society. The discourse of depression switches the focus inwards to the person, who is now to hold him or self individually responsible for events. Thus people who have in fact very little control over their now more insecure lives must blame themselves. The narrative structure of these new accounts of distress indicates that people have internalised the values of Capitalist enterprise culture and the responsibility for personal failure that goes with it. To say these people have depression decontextualises their lived life, whereas nervi embodied it. It is this shift in officially endorsed versions of personal identity, the shaping of a different kind of citizen (one rather convenient for the Western economic order, as it happens) that can be seen as evoked in the globalisation of the category of depression, and yet is absurdly invisible in mainstream psychiatric accounts.
1. Skultans V. From damaged nerves to masked depression: inevitability and hope in Latvian psychiatric narratives. Social Science & Medicine 2003, 56, 2421-31.
Competing interests: None declared
Competing interests: No competing interests
Many thanks to the author for such an insightful article on a difficult subject. Even in well-structured Western countries utlising standardised diagnostic criteria (such as DSM-IV or ICD-10) the diagnosis of even severe mental disorders such as psychosis can vary from country to country.
To assume that these classification systems are automatically applicable to the rest of the world's populations would be to take an over -simplified and simplistic view of global mental health issues.
Coming from an Asian background but also having had the opportunity of working in psychiatry in the UK for nearly nine years I have been fortunate enough to see both sides of the coin. Taking depression as an example, life events which may lead to this illness in the UK may not necessarily have such an adverse impact on the mental health of an individual in India due to religious and cultural beliefs. An explanation for an adverse event or misfortune in India would be along the lines of 'it was in my kismet (luck)' or 'it is my karma (fate)' or 'it was because of my previous sins (in a past life)'......etc. Furthermore social networks are (fortunately) still strong - families, friends and communities rally round in times of need. An individual in India may not even get the time nor space to analyse his feelings or symptoms (of depression). As a result of these factors even if he or she has 'a depressive episode' his or her experience of this illness can be very different when compared to an individual with the same illness living in a Western country and therefore by extrapolation ICD-10 or DSM- IV critria may not be useful in defining and diagnosing mental illness in such populations.
Interventions which can work in Western countries may therefore not necessarily work in the rest of the world. Furthermore, interventions (such as Cognitive behavioural therapy, Family therapy, etc) are usually not even available and nor are the infrastructures required to deliver treatments and therapies in place.
I fully agree with the author when he says - 'Here Western psychological discourse is setting out to instruct, regulate, and modernise, presenting as definitive the contemporary Western way of being a person. It is unclear why this should be good for mental health in Africa or Asia. This is medical imperialism, similar to the marginalisation of indigenous knowledge systems in the colonial era, and is generally to the disadvantage of local populations'.
In my opinion the take home message from this article is - 'What is good enough for the goose (in this case) may not be good enough for the gander'.
Competing interests: None declared
Competing interests: No competing interests
The observations made by Dr Summerfield in the May 3rd edition of the British Medical Journal 1 are not new, but they serve as reminders of the challenges facing psychiatric practice and research. Although his focus is on transcultural applicability, in questioning the validity of psychiatric classification and diagnosis in general, Dr Summerfield is questioning the value of current research not only in developing but also in Western countries. The implication of his argument, which is based on limited and selected literature, is that scarce resources in developing countries should not be allocated to treat disorders (i.e., mental disorders) that are not established to be valid. Dr Summerfield argues that what is scientifically valid is what is real for “the people being studied”. He concludes that because psychiatric diagnoses are not realities for study participants in developing countries, they are “pseudodiagnoses” and the interventions “unwanted distraction in the hard pressed lives of non- Western subjects”.
Although the arguments forwarded by Dr Summerfield are valid in so far as some of the limitations of psychiatric diagnosis and classification are concerned, they misrepresent the fundamental achievements of epidemiological research in developing countries and risk tracking back the discourse. We outline here the reasons for this view using the same specific headings as Dr Summerfield. We remark on the issue of psychiatric classification as it pertains to diagnosis in developing countries.
1) Validity of psychiatric classification and diagnosis
We agree with Dr Summerfield that until firm aetiological understanding becomes available to support the diagnosis of psychiatric disorders, their validity remains less certain. It is also true that psychiatric disorders as depicted in the fifth chapter of the International Classification of Diseases (ICD 10)2 and the 4th revision of the Diagnostic Manual of Mental Disorders (DSM-IV)3 in some respects represent Western frameworks of understanding of mental disorders. However, to view these systems to be entirely Western is simplistic. We also suggest that there are useful alternatives in understanding validity beyond the narrow definition presented by Dr Summerfield. Despite the well documented discussions about the benefits and disadvantages of psychiatric classification4-6, the consensus has been that the many benefits of applying the classification systems outweighs their disadvantages4 5. To reduce the current classification systems to simple committee decisions, as is intimated by Dr Summerfield, ignores and misrepresents the facts. The development of the DSM IV (which is a convergent system with chapter V of ICD-10) was a five year iterative process involving extensive literature review, data analysis and field trials in which over 1000 experts participated5 7.
We would also like to bring additional perspective in to the discussion of the validity of psychiatric disorders as it reflects on developing countries.
a)) Historical continuity of disorders -- Knowledge regarding the major psychiatric disorders has endured over time and place. What appears to change is not the description but the attribution and intervention. Syndromes of depression, anxiety, psychotic disorders, “hysteria”, organic brain disorders (for example dementia) were well documented from Africa, Asia and the Middle East even before we find clear documentation from the West5. This shows that psychiatric disorders, although now diagnostically standardised, are conditions that emerge with the documented history of mankind. Thus, not withstanding the inclusion of some syndromes dictated by the Western culture, the assertion that psychiatric disorders as described in the DSM/ICD system are just Western concepts (of the pathological) is incorrect.
b) Phasic nature of understanding -- Understanding of mental disorders has evolved over the years passing through stages. Thus in some developing countries, the current causal attribution is similar to erstwhile attributions in the West8. Nevertheless, qualitative work from East and Western Africa has identified local terminologies of psychiatric disorders or what one might call “psychiatric diagnostic equivalents”. These are consistent with what would be considered schizophrenia, mania, depression, anxiety and conversion/dissociative disorders in modern psychiatric nosology8. Similar concepts are also identified in Latin America 9.
c) ‘Relative’ Validity -- The lack of concrete aetiological factors - as are present in the example of tuberculosis - is one of the key challenges in relation to the validity of psychiatric diagnoses. Researchers over the years have relied on alternative methods of validation, such as convergent validity (convergence with a known measure tapping into an aspect of a disorder, such as disability and service use) and predictive validity (prediction of treatment response and course)10. These have good utility4 and have been used in some developing countries as part of the measurement of depression11-13. Thus, although psychiatric diagnostic entities lack confirmatory biological criteria, most are, “by virtue of information about outcome, treatment response and aetiology that they convey…are…invaluable working concepts for clinicians”4 and researchers. Therefore, the assertion by Dr Summerfield that psychiatric diagnostic entities are invalid because they lack pathognomonic biological concomitants is not supported by the evidence.
d) Predictability of human response — The experience of psychiatric disorders are in most instances understandable in the context of response to life events or stress; for example loss events precipitating depression and threatening events precipitating anxiety14 15. These are universal experiences with comparable natural paradigms. Some of these responses are mediated by established physiological mechanisms. As accumulating evidence demonstrates, many psychiatric disorders are associated with identifiable neurochemical and anatomical changes16-19. Thus, despite the possible variations in behavioural expression, underlying changes and mechanisms of mental disorder are likely to be universal.
2) Validity of Mental Health Research
It is true that the validity of research method goes beyond the technical. However, technically sound research method is the essential starting point for valid research. Ideally, one would apply locally developed instruments and diagnostic systems. This is not always possible or necessary in the real world. In fact, the studies that have attempted to develop locally applicable psychiatric instruments indicate the best approach may be using the existing templates rather than “reinventing the wheel”20-22. A well formulated ethnographic two-year study in Zimbabwe led to the development of a 14-item local screening instrument21. However, on careful reading of the findings, one might be surprised to see that no unique or unusual symptom outside the well recognised symptoms of common mental disorders (e.g. anxiety and depression) are included. The most important issue in cultural validity of psychiatric instruments appears to be careful translation to make sure that the questions are understandable to participants.
Some argue that because psychiatric syndromes are not well demarcated they cannot be valid entities4. As is raised here by Dr Summerfield, distinction from “ordinary” response to life events and stressors is a prominent area of dispute. However, this issue is not unique to participants from developing countries. Millions are exposed to stressful events and daily hassles anywhere in the world. Additionally, some long- term follow-up studies indicate that the impact of general societal changes on the prevalence of depression and anxiety disorders may be minimal23. On the other hand, there is a consensus emerging about the aetiology of mental disorders that predisposed individuals develop mental disorder in response to stress, and that this predisposition can be measured in terms of genetics and early environmental adversity 14 15.
The main study Dr Summerfield cites to substantiate his argument is a meta-analysis of studies on refugees24. A refugee population is atypical sample with unique and complex needs. Most of the assessments used in the included studies were also dimensional scales mostly measuring the level of distress rather than diagnostic entities. Dr Summerfield argues that because improvement was predicted by improvement of the social circumstances of the refugees what they were suffering should have been transient distress rather than a valid disorder. However, in a bio- psychosocial illness model, improvement in social circumstances is expected to lead to some improvement in mental distress. Additionally, it is not clear what proportion of the sample had diagnosable psychiatric disorder. This study, therefore, cannot be used to substantiate his arguments. Incidentally, we would note that this paper24 primarily reviewed studies of European and Middle Eastern refugees with only 716 (of 67 294 participants) originating from Africa, while the comments of Dr Summerfield pertain primarily to Africa and Asia.
3) Cultural variability
Identifying uniqueness and differences across cultures will always be important in epidemiological studies to understand fully aetiological factors25 26. But many of the differences identified in earlier studies, particularly regarding illness distribution, symptoms and course of psychiatric disorders, were due to methodological problems as confirmed in more robust subsequent studies27-29. In a community study that employed psychiatrists and social scientists, in which cases with symptoms of anxiety and depression were identified, the authors were surprised to find “patterning and types of psychiatric disorder among the Yoruba turned out to be qualitatively much like disorders with which we in North America and Europe are familiar”28 The interviewers in the same group met up with “cultural barriers great enough to influence their judgement” in only 5.5% (21/385). Thus, although careful consideration should be given to cultural distortions, the evidence suggests that cultural variability has only a modest influence on the diagnosis of the major psychiatric disorders.
4) Depression as an example
Even in some non-Western countries with the least contact with Western culture, depression can be easily identified. The view that depression is not a problem in developing countries like Africa6 is a thing of the past. In fact, one of the major findings of the pivotal International Pilot Study of Schizophrenia (IPSS) was the remarkable similarity in the symptoms that characterised depressed patients with psychotic symptoms in the different cultures studied, which included developing countries30. However, although physical/somatic symptoms commonly accompany depression31, there is a perception that in developing country settings patients with depression may more commonly present with somatic/physical symptoms than with depression32. Under these circumstances, relevant questioning establishes the diagnosis32. Among the Chinese, patients with depression are more likely to receive a diagnosis of neurasthenia, with up to 87% of cases with neurasthenia fulfilling criteria for depression33 . Such prominence of physical symptoms in presentation may lead to underestimation of the prevalence of depression in lay administered diagnostic instruments.
We should be reminded that depression is a very serious illness that is associated with serious risks with a potential to change an individual’s life course drastically. Working in a developing country setting, one of the authors had seen countless patients with classical depression as defined here in the West, and had witnessed the taxing personal and family consequences of depression..
We suggest that there is no evidence to negate the validity of psychiatric research in developing countries. Rather than just focusing on ethnographic work as suggested by Dr Summerfield, applicable epidemiological and interventional studies should continue in parallel with more ethnographic work, as it is being done currently. Of course mental health is but one of the many public health challenges facing developing countries. On the other hand other global health challenges cannot be addressed ignoring mental health. Furthermore, whether desirable or not, the impact of globalisation is substantial and Western concepts of the world have already permeated most parts. Preconceptions about health and disease are unavoidable. Researchers in developing countries should be aware of not only distortions due to local culture but also distortions created through the impact of globalisation.
The advantage of using systematic classification systems outweighs their disadvantage
Many of the psychiatric disorders included in the current classification systems are identifiable in developing countries
There is no evidence to suggest that findings of studies conducted in developing countries are less valid than those conducted in the West
Global health issues cannot be addressed without addressing mental health
1. Summerfield D. How scientifically valid is the knowledge base of global mental health? BMJ 2008;336:992-4.
2. WHO. The International Classification of Diseases (10th revision). Geneva: World Health Organisation, 1992.
3. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (4th edn) (DSM-IV). Washington, DC: American Psychiatric Association, 1994.
4. Kendell R, Jablensky A. Distingushing between the validity and utility of psychiatric diagnoses. Am J Psychiatry 2003;160:4-12.
5. Frances A, First MB, Pincus HA. The essential companion to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition: DSM -IV guidebook. Washington, DC: American Psychiatric Association, 1995.
6. Lehmann HE. Psychiatric concepts of depression: nomenclature and classification. Can Psychiatr Assoc J 1959;4(Special supplement):1-10.
7. APA. DSM-IV Resource Book. Washington, DC: American Psychiatric Association, 1996.
8. Ndetei DM, Mburu J. History of Psychiatry. In: Ndetei DM, Szabo CP, Okasah T, Mburu JM, editors. The African text book of clinical psychiatry and mental health. Nairobi: AMREF, 2006:6-10.
9. Incayawar M. Efficacy of Quichua healers as psychiatric diagnosticians. Br J Psychiatry 2008;192:390-1.
10. Guze SB, Robins E. Establishment of diagnostic validity in psychiatric illness: application to schizophrenia. Am J Psychiatry 1970;126:983-7.
11. Bolton P, Neugebauer R, Ndogoni L. Prevalence of depression in rural Rwanda based on symptom and functional criteria. J Nerv Ment Dis 2002;190:631-7.
12. Fekadu A, O'Donovan MC, Alem A, Kebede D, Church S, Johns L, et al. Validity of the concept of minor depression in a developing country setting. J Nerv Ment Dis 2008;196:22-8.
13. Mogga S, Prince M, Alem A, Kebede D, Stewart R, Glozier N, et al. Outcome of major depression in Ethiopia: population-based study. Br J Psychiatry 2006;189:241-6.
14. Kendler KS, Hettema JM, Butera F, Gardner CO, Prescott CA. Life event dimensions of loss, humiliation, entrapment, and danger in the prediction of onsets of major depression and generalized anxiety. Arch Gen Psychiatry 2003;60:789-96.
15. Kendler KS, Gardner CO, Prescott CA. Toward a comprehenisve developmental model for major depression in men. Am J Psychiatry 2006;163:115-24.
16. Belmaker RH, Agam G. Mechanism of disease. Major depression. N Engl J Med 2008;358:55-68.
17. Freedman R. Schizophrenia. N Engl J Med 2003;349:1738-49.
18. Muller-Oerlinghausen B, Berghofer A, Bauer M. Bipolar disorder. Lancet 2002;359:241-7.
19. Stein DJ. Obsessive-compulsive disorder. Lancet 2002;360:397-405.
20. Okulate GT, Olayinka MO, Jones OBE. Somatic symptoms in depresion: evaluation of their diagnostic weight in an African setting. Br J Psychiatry 2004;184:422-7.
21. Patel V, Simmunyu E, Gwanzura F, Lewis G, Mann A. The Shona Symptom Questionnaire: the development of an indigenous measure of common mental disorders in Harare. Acta Psychiatr Scand 1997;95:469-75.
22. Zilber N, Youngman R, Workneh F, Giel R. Development of a culturally-sensitive psychiatric screening instrument for Ethiopian populations, NIRP Research for Policy Series 20. Amesterdam: Royal Tropical Institute, 2004.
23. Murphy JM, Laird NM, Monson RR, Sobol AM, Leighton AH. A 40-year perspective on the prevalence of depression: the Stirling County Study. Arch Gen Psychiatry 2000;57:209-15.
24. Porter M, Haslam N. Predisplacement and postdisplacement factors associated with mental health of refugees and internally displaced persons. JAMA 2005;294:602-12.
25. Fekadu A, Shibre T, Alem A, Kebede D, Kebreab S, Negash A, et al. Bipolar disorder among an isolated island community in Ethiopia. J Affect Disord 2004;80(1):1-10.
26. Torrey EF. Prevalence studies in schizophrenia. Br J Psychiatry 1987;150:598-608.
27. Fekadu A, Kebede D, Alem A, Fekadu D, Mogga S, Negash A, et al. Clinical outcome in bipolar disorder in a community-based follow-up study in Butajira, Ethiopia. Acta Psychiatr Scand 2006;114:426-34.
28. Leighton AH, Lambo TA, Hughes CC, Leighton DC, Murphy JM, Macklin DB. Psychiatric disorders in West Africa. Am J Psychiatry 1963;120:521-7.
29. Orley J, Wing JK. Psychiatric disorders in two African villages. Arch Gen Psychiatry 1979;36:513-20.
30. Griffith EEH, Gonzalez CA, Blue HC. The basices of cultural psychiatry. In: Hales RE, Yudofsky SC, Talbott JA, editors. Textbook of psychiatry. 3rd ed. Washington, DC: American Psychiatric Press, 1999:1463- 92.
31. Cleghorn RA, Curtis GC. Psychosomatic accompainments of latent and manifest depressive affect. Can Psychiatr Assoc J 1959;4(Special supplement):s13-s23.
32. Patel V, Abas M, Broadhead J, Todd C, Reeler A. Depression in developing countries: lessons from Zimbabwe. BMJ 2001;322:482-4.
33. Parker G, Gladstone G, Chee KT. Depression in the planet's largest ethnic group: the Chinese. Am J Psychiatry 2001;158:857-64.
Competing interests: Dr Abebaw Fekadu has practiced psychiatry in a developing country setting in Africa for several years and is involved in ongoing collaborative research. None of the authors are part of the global mental health group and they are not involved with the WHO or the diagnostic committees of DSM and ICD.
Competing interests: No competing interests