CanJPsychiatry 2013;58(11):618–621
Perspective
The Importance and Limits of Harm in Identifying Mental Disorder Jerome C Wakefield, PhD, DSW1; Michael B First, MD2 1
Professor of Social Work, Silver School of Social Work, New York University, New York, New York; Professor of Psychiatry, Department of Psychiatry, School of Medicine, New York University, New York, New York. Correspondence: Silver School of Social Work, New York University, One Washington Square North, New York, NY 10003; [email protected].
2
Professor of Clinical Psychiatry, Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, New York; Research Psychiatrist, Division of Clinical Phenomenology, New York State Psychiatric Institute, New York, New York.
Key Words: Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition, DSM5, diagnosis, false positives, dysfunction, mental disorder definition, clinical significance, validity, harmful dysfunction Received, revised, and accepted March 2013.
The In Review articles in this issue on normality and disorder by Dr Rachel Cooper and Dr Derek Bolton explore the importance of a value component of harm in the concept of mental disorder. They focus on the Diagnostic and Statistical Manual of Mental Disorder’s clinical significance criterion, requiring that symptoms cause significant distress or role impairment, as the expression of the harm component. As Dr Bolton argues, harm in the form of distress or role impairment has always been intimately tied to the concept of disorder and treatment decisions; as Dr Cooper argues, without the harm requirement, any disliked anomaly may be labelled a disorder. Moreover, as Cooper argues, a harm requirement is not incompatible with a natural kinds approach to distinguishing among disorders or to a categorical approach to disorder; the lack of zones of rarity on the harm continuum does not preclude categorical underlying causal processes. However, neither paper systematically develops arguments regarding the other component of disorder, the requirement that the harm must be caused by underlying dysfunction. The dysfunction component distinguishes disorders from the many other negative conditions in life. Cooper’s identification of dysfunction with symptom severity ignores the fact that normal suffering can be severe, and Bolton’s attempt to encompass risk of harm within harm yields an implausibly expansive conception of disorder. While the harm component is essential, clarification of the dysfunction component of the concept of disorder, pursued in part 2 of this In Review in the December 2013 issue, is also essential to establishing a coherent and plausibly limited domain of psychiatric disorder within the broader arena of harmful conditions. WWW
L’importance et les limites des dommages dans l’identification d’un trouble mental Les articles In Review de ce numéro sur la normalité et le trouble, dont la Dre Rachel Cooper et le Dr Derek Bolton sont les auteurs, explorent l’importance d’un élément de valeur des dommages dans le concept d’un trouble mental. Ils se penchent sur le critère de significativité clinique du Manuel diagnostique et statistique des troubles mentaux, lequel dicte que les symptômes causent une détresse ou une incapacité fonctionnelle, constituant ainsi l’expression de la composante des dommages. Comme le défend le Dr Bolton, les dommages sous forme de détresse ou d’incapacité fonctionnelle ont toujours été intimement liés au concept de trouble et aux décisions de traitement; pour sa part, la Dre Cooper fait valoir que sans le critère des dommages, toute anomalie détestée peut être étiquetée de trouble. En outre, comme dit la Dre Cooper, une exigence de dommages n’est pas incompatible avec une approche des espèces naturelles pour distinguer parmi les troubles ou pour une approche catégorique du trouble; l’absence de zones de rareté sur le continuum des dommages n’empêche pas les processus catégoriques causaux sous-jacents. Cependant, aucun des deux articles ne développe systématiquement d’arguments à l’égard de l’autre composante du trouble, l’exigence que les dommages soient causés par une dysfonction sous-jacente. La composante dysfonction distingue les troubles de nombre d’autres affections négatives dans la vie. Quand la Dre Cooper identifie la dysfonction à la gravité des symptômes, elle ne tient pas compte du fait que la souffrance normale peut être grave, et la tentative de Bolton d’englober le risque de dommages dans les dommages produit une conception du trouble invraisemblablement élargie. Bien que la composante des dommages soit essentielle, la clarification de la composante dysfonction du concept du trouble, qui continue dans la 2e partie de cet In Review dans le numéro de décembre 2013, est également essentielle pour établir un domaine limité cohérent et plausible du trouble psychiatrique au sein de l’arène plus vaste des affections nuisibles. 618 W La Revue canadienne de psychiatrie, vol 58, no 11, novembre 2013
www.LaRCP.ca
The Importance and Limits of Harm in Identifying Mental Disorder
Each of the 2 articles1,2 in this first instalment of a 2-part In Review on the concept of mental disorder and the normal– disordered boundary are by philosophers of psychiatry (Dr Derek Bolton2 is also a clinical psychologist) with wide experience writing on diagnostic topics. They both focus their analyses on the value issues in determining diagnostic validity, especially the harm that a condition causes, as an indicator of its disorder status. However, they diverge on whether harm is enough for disorder or something more is needed. We comment briefly here on some salient points in each of the papers,1,2 and raise concerns and questions for further thought. Dr Rachel Cooper1 argues, first, that retaining the harm criterion for disorder (as she labels the Diagnostic and Statistical Manual of Mental Disorders [DSM], Fourth Edition’s, clinical significance criterion that requires that symptoms cause distress or role impairment) is essential for distinguishing between disorder and mere difference, such as harmless eccentricity or normal human variation. Cooper’s plea to retain the harm criterion is not a theoretical exercise. Early on, the DSM-5 Task Force stated its intention to remove the harm criterion and all reference to impairment in functioning from the diagnostic criteria.3 The aim was coordination with the World Health Organization’s goal of separating disorder diagnosis from assessment of functioning so as to avoid resting diagnosis on culturally relative conceptions of proper functioning.4 However, as the DSM-5 Task Force soon recognized, given that objective indicators of disorder status are lacking and that many disorders are defined entirely in terms of symptoms that can occur in people without disorders, such a separation is not possible in most categories.5 For example, how does one diagnose reading disorder without reference to the impact of the person’s reading ability on school or work functioning, or snake phobia—which involves an otherwise normal human fear—without considering its interference with role functioning? Cooper1 and Bolton2 are thus on solid ground in insisting that harm must remain a critical component of disorder. Nonetheless, DSM-5 did demote the role of harm, that is, distress or impairment, in the definition of mental disorder from a necessary condition, as advocated by Cooper and Bolton, stating instead that mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities. Perhaps portraying harm as frequently occurring but not essential for disorder is aimed at preparing the way for further separation of diagnosis and impairment; thus this will likely be an area of future controversy. It has become popular to deny that disorders are natural kinds that carve nature at its joints, especially by those who embrace the importance of values in identifying disorders. Cooper1 rejects the argument that because values enter into disorder judgments, disorders are not natural kinds. Elsewhere6 she uses a compelling analogy to the concept of weeds, which is certainly a value concept—weeds by definition are plants that are unwanted in a certain context, such as a garden—and yet refers to a set of natural kinds, www.TheCJP.ca
Highlights •
The value judgment that a psychological condition is harmful because it causes significant distress or social role impairment is essential to the concept of mental disorder because it distinguishes disorders from harmless anomalies and directs attention to conditions warranting treatment.
•
However, not all negative or harmful psychological conditions are psychiatric disorders; a further reference to underlying dysfunction is crucial to separating psychiatric disorder from the many negative psychological conditions and risks inherent in normal life.
•
The International Classification of Diseases and initial Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, aspiration to completely separate role impairment from disorder diagnosis is impossible at this stage of knowledge because harm to role functioning is sometimes necessary to distinguish disorder from normal variation.
namely, the species that constitute the unwanted types in that context. Thus a value criterion can delineate a domain that is then divided according to natural kinds. The concept of harmful dysfunction works this way as well. These days, the natural kinds approach to disorder is confronted by the objection that symptoms seem to be continuously distributed, whereas categorical disorders can exist only if there are discontinuities, or points of rarity, in symptom distributions that reveal natural categories rather than arbitrary division of a continuous dimension. However, even assuming no such points of rarity, there are still several possible reasons why disorders may be categorical natural kinds. First, it is always possible that there are additional variables that may yield a discontinuity underlying apparent continuity.7 Moreover, multiple continuous dimensions can interact to create unexpected inflection points that can represent a higher-order discontinuity.8 More basically, if disorder consists of 2 components, harm and dysfunction, continuity of harm or symptoms does not preclude discontinuity of dysfunction.9 Cooper responds to the points of rarity objection to disorder categories with an analogy to metal alloys. Alloys vary continuously among possible mixes of metals. However, we value certain alloys for certain purposes, such as making stealth airplanes. The interactions of the metals in particular alloys optimize the properties we value. Such alloys become distinct categories for us that have their own essences— even though there are no points of rarity in the possible mix of metals or even in the degree of the property in which we are interested. Superficial continuity and lack of points of rarity do not rule out higher-order discontinuities that can differentiate disorder from normality. Like the harmful dysfunction analysis, Cooper1 allows for 2 dimensions in evaluating the presence of disorder, symptom severity (basically intended to be indicative of dysfunction) and harm, although her paper focuses more on the harm The Canadian Journal of Psychiatry, Vol 58, No 11, November 2013 W 619
Perspective
aspect. Symptom thresholds determine dysfunction but, she argues, the dimensionality of symptom distributions means that the threshold must be set using many criteria, so there is not one unique solution to how to separate disorder from normality. To the degree that Cooper1 equates dysfunction with symptom severity, we would suggest that the situation is more complex; symptom severity can exist in emotionally intense, nondisordered conditions as well. However, her view is widely shared. For example, Darrel Regier, ViceChair of the DSM-5 Task Force, expressed early in the revision process that one goal was to create symptom severity scales for all disorders that would eventually allow identification of “statistically valid cutpoints between normal and pathological.”10 The problem is that many normal conditions are, by any objective standard, severe. For example, think of normal grief after losing a child, or in physical medicine, such conditions as pain during childbirth. We certainly want to help people with normal suffering, but there is nothing disordered about these conditions. A symptom severity scale alone that ignores the independent distinction between normal function and dysfunction cannot give us the science of disorder that we seek. Whereas Cooper focuses on harm but acknowledges dysfunction as a requirement for disorder, Bolton2,11 sets aside dysfunction as an independent criterion and attempts to construct an account that relies more or less wholly on harm—in the form of clinically significant distress or impairment. He links this account to a pragmatic understanding of disorder as whatever physicians (or other mental health professionals) treat. We agree with Bolton that social values must enter into the notion of medical, including mental, disorder. The vigorous development of the harm component of disorder in Bolton’s hands is a valuable and welcome corrective to a sterile, purely scientistic notion of disorder. However, Bolton’s view has the limitation that a pure harm account of disorder, divorced from any reference to failure of biologically designed functioning, is subject to compelling counterexamples and can yield massive false positive rates in clinical practice. Bolton rightly points to “the tight conceptual linkage between illness and distress and impairment of day-to-day functioning.”2, p 612 However, this link must be placed within the context of the distinction between normal and disordered functioning or else it yields endless counterexamples. Childbirth pain, normal grief over loss of a child, and teething pain are all distressing conditions for which people consult their physicians yet are not considered disorders; unwanted fertility and unwanted pregnancy are distressing conditions that are commonly treated by physicians and yet not considered disorders; and even sleep and childhood can be considered enormously impairing yet are obviously not disorders. What goes into the diagnostic manual, what should be treated, what is harmful in the sense of distressing or impairing, and what is 620 W La Revue canadienne de psychiatrie, vol 58, no 11, novembre 2013
a disorder are equated at various times by Bolton, yet these need not coincide. In arguing for his exclusive focus on harm, Bolton’s2 reading of the DSM’s definition of disorder fails to address the part of the DSM’s definition that concerns dysfunction, which involves an inference to an underlying pathological state that is generating the harm. Indeed, it is unclear how Bolton’s harm view would explain the DSM’s own distinction between disorders and V codes that list conditions that are not disorders but are often the target of professional intervention owing to associated distress or impairment— such as marital incompatibility, unemployment, being sexually abused, and possessing few occupational skills. It is the very problem of normal distress and impairment that makes the challenge of false positives so difficult and the clarification of the concept of mental disorder so necessary. Bolton2 suggests that, to keep the disorder threshold meaningful, the DSM’s requirement of clinically significant distress or impairment might have been translated into intolerable distress or impairment. No doubt an increased threshold (even if potentially quite subjective and variable across people, as is intolerable) would resolve some problems with overdiagnosis. However, it must be kept in mind that, just like physical pathology, psychopathology can at times be moderate or even mild. A pure harm view runs into problems here because the inclusion of mild cases leaves disorder encompassing all of human unhappiness, whereas a reference to malfunction of biologically designed processes allows a distinction, even between mild disorders and mild normal conditions. Lastly, Bolton2 endorses the use of risk of harm as itself a form of harm that justifies attributing pathology. As it happens, DSM-5 itself revised the definition of disorder to eliminate risk as a sufficient form of harm on the basis that risk of disorder is not disorder. The initially proposed category of psychotic risk syndrome was rejected as a disorder in DSM-5 in part for confusing risk of disorder with disorder, but, when reconceptualized as a mild psychotic disorder in its own right and renamed attenuated psychosis syndrome, it was incorporated into section 3 of DSM-5. More generally, seeing disorder as captured by clinical significance understood as harm plus risk of harm seems to send the concept of medical disorder spinning out of control to encompass much of life. Pregnancy imposes risks on women, which at some periods of history and in some places have been extremely high, yet pregnancy, though medically monitored, has never been considered a disorder. Men are at a much-heightened risk of dying earlier than women, yet being male is not a disorder. Shorter people are at higher risk of cardiovascular disease, and sedentary people running for a bus have a much-inflated rate of an immediate cardiovascular event, yet neither shortness nor running for a bus are disorders. Risk is omnipresent, and it is only the dysfunction constraint on disorder that keeps disorder from also being omnipresent. A conception of disorder as harm, and an extension of harm to risk of harm, www.LaRCP.ca
The Importance and Limits of Harm in Identifying Mental Disorder
goes way beyond the concept of disorder as we understand it. We noted in the Guest Editorial12 that the need for an analysis of disorder is partly driven by the common existence of harm that is not disorder. Both In Review papers in this issue1,2 illuminate the conceptual terrain and constructively explore the role of harm in distinguishing disorder from nondisorder. In addition, Cooper1 offers good reasons for not abandoning the dysfunction requirement or a natural kinds reading of disorder just because values are involved. However, it seems to us that these papers stop short of elaborating a sufficiently specific and satisfying account of dysfunction that explains what differentiates harms that are disorders from harms that are not disorders. The dysfunction requirement will be the focus of part 2 of this In Review, appearing in the December 2013 issue.
Acknowledgements
Dr Wakefield and Dr First have no funding or conflicts of interest to declare.
The Canadian Psychiatric Association proudly supports the In Review series by providing an honorarium to the authors.
References
1. Cooper RV. Avoiding false positives: zones of rarity, the threshold problem, and the DSM clinical significance criterion. Can J Psychiatry. 2013;58(11):606–611.
www.TheCJP.ca
2. Bolton D. Overdiagnosis problems in the DSM-IV and the new DSM-5: can they be resolved by the distress–impairment criterion? Can J Psychiatry. 2013;58(11):612–617. 3. Schneider ME. DSM-V Task Force weighing changes. Clin Psychiatry News. 2009;37(4):1–4. 4. Ustun B, Kennedy C. What is “functional impairment”? Disentangling disability from clinical significance. World Psychiatry. 2009;8:82–85. 5. Wakefield JC. Disability and diagnosis: should role impairment be eliminated from DSM/ICD diagnostic criteria? World Psychiatry. 2009;8:87–88. 6. Cooper R. What is wrong with the DSM? Hist Psychiatry. 2004;15(1):5–25. 7. Klein DF. A proposed definition of mental disorder. In: Klein DF, Spitzer RL, editors. Critical issues in psychiatric diagnosis. New York (NY): Raven Press; 1978. p 41–72. 8. Wakefield JC. The perils of dimensionalization: distinguishing personality traits from personality disorders. Psychiatr Clin N Am. 2008;31:379–393. 9. Wakefield JC. The concept of mental disorder: on the boundary between biological facts and social values. Am Psychol. 1992;47:373–388. 10. Greenberg G. Inside the battle to define mental illness [Internet]. Wired.com (magazine). San Francisco (CA): Condé Nast; 2011;19(1) [cited 2013 Sep 12]. Available from: http://www.wired.com/magazine/2010/12/ff_dsmv. 11. Bolton D. What is mental disorder? An essay in philosophy, science and values. Oxford (GB): Oxford University Press; 2008. 12. Wakefield JC, First MB. Clarifying the boundary between normality and disorder: a fundamental conceptual challenge for psychiatry. Can J Psychiatry. 2013;58(11):618–621.
The Canadian Journal of Psychiatry, Vol 58, No 11, November 2013 W 621
Perspective
The Importance and Limits of Harm in Identifying Mental Disorder Jerome C Wakefield, PhD, DSW1; Michael B First, MD2 1
Professor of Social Work, Silver School of Social Work, New York University, New York, New York; Professor of Psychiatry, Department of Psychiatry, School of Medicine, New York University, New York, New York. Correspondence: Silver School of Social Work, New York University, One Washington Square North, New York, NY 10003; [email protected].
2
Professor of Clinical Psychiatry, Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, New York; Research Psychiatrist, Division of Clinical Phenomenology, New York State Psychiatric Institute, New York, New York.
Key Words: Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition, DSM5, diagnosis, false positives, dysfunction, mental disorder definition, clinical significance, validity, harmful dysfunction Received, revised, and accepted March 2013.
The In Review articles in this issue on normality and disorder by Dr Rachel Cooper and Dr Derek Bolton explore the importance of a value component of harm in the concept of mental disorder. They focus on the Diagnostic and Statistical Manual of Mental Disorder’s clinical significance criterion, requiring that symptoms cause significant distress or role impairment, as the expression of the harm component. As Dr Bolton argues, harm in the form of distress or role impairment has always been intimately tied to the concept of disorder and treatment decisions; as Dr Cooper argues, without the harm requirement, any disliked anomaly may be labelled a disorder. Moreover, as Cooper argues, a harm requirement is not incompatible with a natural kinds approach to distinguishing among disorders or to a categorical approach to disorder; the lack of zones of rarity on the harm continuum does not preclude categorical underlying causal processes. However, neither paper systematically develops arguments regarding the other component of disorder, the requirement that the harm must be caused by underlying dysfunction. The dysfunction component distinguishes disorders from the many other negative conditions in life. Cooper’s identification of dysfunction with symptom severity ignores the fact that normal suffering can be severe, and Bolton’s attempt to encompass risk of harm within harm yields an implausibly expansive conception of disorder. While the harm component is essential, clarification of the dysfunction component of the concept of disorder, pursued in part 2 of this In Review in the December 2013 issue, is also essential to establishing a coherent and plausibly limited domain of psychiatric disorder within the broader arena of harmful conditions. WWW
L’importance et les limites des dommages dans l’identification d’un trouble mental Les articles In Review de ce numéro sur la normalité et le trouble, dont la Dre Rachel Cooper et le Dr Derek Bolton sont les auteurs, explorent l’importance d’un élément de valeur des dommages dans le concept d’un trouble mental. Ils se penchent sur le critère de significativité clinique du Manuel diagnostique et statistique des troubles mentaux, lequel dicte que les symptômes causent une détresse ou une incapacité fonctionnelle, constituant ainsi l’expression de la composante des dommages. Comme le défend le Dr Bolton, les dommages sous forme de détresse ou d’incapacité fonctionnelle ont toujours été intimement liés au concept de trouble et aux décisions de traitement; pour sa part, la Dre Cooper fait valoir que sans le critère des dommages, toute anomalie détestée peut être étiquetée de trouble. En outre, comme dit la Dre Cooper, une exigence de dommages n’est pas incompatible avec une approche des espèces naturelles pour distinguer parmi les troubles ou pour une approche catégorique du trouble; l’absence de zones de rareté sur le continuum des dommages n’empêche pas les processus catégoriques causaux sous-jacents. Cependant, aucun des deux articles ne développe systématiquement d’arguments à l’égard de l’autre composante du trouble, l’exigence que les dommages soient causés par une dysfonction sous-jacente. La composante dysfonction distingue les troubles de nombre d’autres affections négatives dans la vie. Quand la Dre Cooper identifie la dysfonction à la gravité des symptômes, elle ne tient pas compte du fait que la souffrance normale peut être grave, et la tentative de Bolton d’englober le risque de dommages dans les dommages produit une conception du trouble invraisemblablement élargie. Bien que la composante des dommages soit essentielle, la clarification de la composante dysfonction du concept du trouble, qui continue dans la 2e partie de cet In Review dans le numéro de décembre 2013, est également essentielle pour établir un domaine limité cohérent et plausible du trouble psychiatrique au sein de l’arène plus vaste des affections nuisibles. 618 W La Revue canadienne de psychiatrie, vol 58, no 11, novembre 2013
www.LaRCP.ca
The Importance and Limits of Harm in Identifying Mental Disorder
Each of the 2 articles1,2 in this first instalment of a 2-part In Review on the concept of mental disorder and the normal– disordered boundary are by philosophers of psychiatry (Dr Derek Bolton2 is also a clinical psychologist) with wide experience writing on diagnostic topics. They both focus their analyses on the value issues in determining diagnostic validity, especially the harm that a condition causes, as an indicator of its disorder status. However, they diverge on whether harm is enough for disorder or something more is needed. We comment briefly here on some salient points in each of the papers,1,2 and raise concerns and questions for further thought. Dr Rachel Cooper1 argues, first, that retaining the harm criterion for disorder (as she labels the Diagnostic and Statistical Manual of Mental Disorders [DSM], Fourth Edition’s, clinical significance criterion that requires that symptoms cause distress or role impairment) is essential for distinguishing between disorder and mere difference, such as harmless eccentricity or normal human variation. Cooper’s plea to retain the harm criterion is not a theoretical exercise. Early on, the DSM-5 Task Force stated its intention to remove the harm criterion and all reference to impairment in functioning from the diagnostic criteria.3 The aim was coordination with the World Health Organization’s goal of separating disorder diagnosis from assessment of functioning so as to avoid resting diagnosis on culturally relative conceptions of proper functioning.4 However, as the DSM-5 Task Force soon recognized, given that objective indicators of disorder status are lacking and that many disorders are defined entirely in terms of symptoms that can occur in people without disorders, such a separation is not possible in most categories.5 For example, how does one diagnose reading disorder without reference to the impact of the person’s reading ability on school or work functioning, or snake phobia—which involves an otherwise normal human fear—without considering its interference with role functioning? Cooper1 and Bolton2 are thus on solid ground in insisting that harm must remain a critical component of disorder. Nonetheless, DSM-5 did demote the role of harm, that is, distress or impairment, in the definition of mental disorder from a necessary condition, as advocated by Cooper and Bolton, stating instead that mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities. Perhaps portraying harm as frequently occurring but not essential for disorder is aimed at preparing the way for further separation of diagnosis and impairment; thus this will likely be an area of future controversy. It has become popular to deny that disorders are natural kinds that carve nature at its joints, especially by those who embrace the importance of values in identifying disorders. Cooper1 rejects the argument that because values enter into disorder judgments, disorders are not natural kinds. Elsewhere6 she uses a compelling analogy to the concept of weeds, which is certainly a value concept—weeds by definition are plants that are unwanted in a certain context, such as a garden—and yet refers to a set of natural kinds, www.TheCJP.ca
Highlights •
The value judgment that a psychological condition is harmful because it causes significant distress or social role impairment is essential to the concept of mental disorder because it distinguishes disorders from harmless anomalies and directs attention to conditions warranting treatment.
•
However, not all negative or harmful psychological conditions are psychiatric disorders; a further reference to underlying dysfunction is crucial to separating psychiatric disorder from the many negative psychological conditions and risks inherent in normal life.
•
The International Classification of Diseases and initial Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, aspiration to completely separate role impairment from disorder diagnosis is impossible at this stage of knowledge because harm to role functioning is sometimes necessary to distinguish disorder from normal variation.
namely, the species that constitute the unwanted types in that context. Thus a value criterion can delineate a domain that is then divided according to natural kinds. The concept of harmful dysfunction works this way as well. These days, the natural kinds approach to disorder is confronted by the objection that symptoms seem to be continuously distributed, whereas categorical disorders can exist only if there are discontinuities, or points of rarity, in symptom distributions that reveal natural categories rather than arbitrary division of a continuous dimension. However, even assuming no such points of rarity, there are still several possible reasons why disorders may be categorical natural kinds. First, it is always possible that there are additional variables that may yield a discontinuity underlying apparent continuity.7 Moreover, multiple continuous dimensions can interact to create unexpected inflection points that can represent a higher-order discontinuity.8 More basically, if disorder consists of 2 components, harm and dysfunction, continuity of harm or symptoms does not preclude discontinuity of dysfunction.9 Cooper responds to the points of rarity objection to disorder categories with an analogy to metal alloys. Alloys vary continuously among possible mixes of metals. However, we value certain alloys for certain purposes, such as making stealth airplanes. The interactions of the metals in particular alloys optimize the properties we value. Such alloys become distinct categories for us that have their own essences— even though there are no points of rarity in the possible mix of metals or even in the degree of the property in which we are interested. Superficial continuity and lack of points of rarity do not rule out higher-order discontinuities that can differentiate disorder from normality. Like the harmful dysfunction analysis, Cooper1 allows for 2 dimensions in evaluating the presence of disorder, symptom severity (basically intended to be indicative of dysfunction) and harm, although her paper focuses more on the harm The Canadian Journal of Psychiatry, Vol 58, No 11, November 2013 W 619
Perspective
aspect. Symptom thresholds determine dysfunction but, she argues, the dimensionality of symptom distributions means that the threshold must be set using many criteria, so there is not one unique solution to how to separate disorder from normality. To the degree that Cooper1 equates dysfunction with symptom severity, we would suggest that the situation is more complex; symptom severity can exist in emotionally intense, nondisordered conditions as well. However, her view is widely shared. For example, Darrel Regier, ViceChair of the DSM-5 Task Force, expressed early in the revision process that one goal was to create symptom severity scales for all disorders that would eventually allow identification of “statistically valid cutpoints between normal and pathological.”10 The problem is that many normal conditions are, by any objective standard, severe. For example, think of normal grief after losing a child, or in physical medicine, such conditions as pain during childbirth. We certainly want to help people with normal suffering, but there is nothing disordered about these conditions. A symptom severity scale alone that ignores the independent distinction between normal function and dysfunction cannot give us the science of disorder that we seek. Whereas Cooper focuses on harm but acknowledges dysfunction as a requirement for disorder, Bolton2,11 sets aside dysfunction as an independent criterion and attempts to construct an account that relies more or less wholly on harm—in the form of clinically significant distress or impairment. He links this account to a pragmatic understanding of disorder as whatever physicians (or other mental health professionals) treat. We agree with Bolton that social values must enter into the notion of medical, including mental, disorder. The vigorous development of the harm component of disorder in Bolton’s hands is a valuable and welcome corrective to a sterile, purely scientistic notion of disorder. However, Bolton’s view has the limitation that a pure harm account of disorder, divorced from any reference to failure of biologically designed functioning, is subject to compelling counterexamples and can yield massive false positive rates in clinical practice. Bolton rightly points to “the tight conceptual linkage between illness and distress and impairment of day-to-day functioning.”2, p 612 However, this link must be placed within the context of the distinction between normal and disordered functioning or else it yields endless counterexamples. Childbirth pain, normal grief over loss of a child, and teething pain are all distressing conditions for which people consult their physicians yet are not considered disorders; unwanted fertility and unwanted pregnancy are distressing conditions that are commonly treated by physicians and yet not considered disorders; and even sleep and childhood can be considered enormously impairing yet are obviously not disorders. What goes into the diagnostic manual, what should be treated, what is harmful in the sense of distressing or impairing, and what is 620 W La Revue canadienne de psychiatrie, vol 58, no 11, novembre 2013
a disorder are equated at various times by Bolton, yet these need not coincide. In arguing for his exclusive focus on harm, Bolton’s2 reading of the DSM’s definition of disorder fails to address the part of the DSM’s definition that concerns dysfunction, which involves an inference to an underlying pathological state that is generating the harm. Indeed, it is unclear how Bolton’s harm view would explain the DSM’s own distinction between disorders and V codes that list conditions that are not disorders but are often the target of professional intervention owing to associated distress or impairment— such as marital incompatibility, unemployment, being sexually abused, and possessing few occupational skills. It is the very problem of normal distress and impairment that makes the challenge of false positives so difficult and the clarification of the concept of mental disorder so necessary. Bolton2 suggests that, to keep the disorder threshold meaningful, the DSM’s requirement of clinically significant distress or impairment might have been translated into intolerable distress or impairment. No doubt an increased threshold (even if potentially quite subjective and variable across people, as is intolerable) would resolve some problems with overdiagnosis. However, it must be kept in mind that, just like physical pathology, psychopathology can at times be moderate or even mild. A pure harm view runs into problems here because the inclusion of mild cases leaves disorder encompassing all of human unhappiness, whereas a reference to malfunction of biologically designed processes allows a distinction, even between mild disorders and mild normal conditions. Lastly, Bolton2 endorses the use of risk of harm as itself a form of harm that justifies attributing pathology. As it happens, DSM-5 itself revised the definition of disorder to eliminate risk as a sufficient form of harm on the basis that risk of disorder is not disorder. The initially proposed category of psychotic risk syndrome was rejected as a disorder in DSM-5 in part for confusing risk of disorder with disorder, but, when reconceptualized as a mild psychotic disorder in its own right and renamed attenuated psychosis syndrome, it was incorporated into section 3 of DSM-5. More generally, seeing disorder as captured by clinical significance understood as harm plus risk of harm seems to send the concept of medical disorder spinning out of control to encompass much of life. Pregnancy imposes risks on women, which at some periods of history and in some places have been extremely high, yet pregnancy, though medically monitored, has never been considered a disorder. Men are at a much-heightened risk of dying earlier than women, yet being male is not a disorder. Shorter people are at higher risk of cardiovascular disease, and sedentary people running for a bus have a much-inflated rate of an immediate cardiovascular event, yet neither shortness nor running for a bus are disorders. Risk is omnipresent, and it is only the dysfunction constraint on disorder that keeps disorder from also being omnipresent. A conception of disorder as harm, and an extension of harm to risk of harm, www.LaRCP.ca
The Importance and Limits of Harm in Identifying Mental Disorder
goes way beyond the concept of disorder as we understand it. We noted in the Guest Editorial12 that the need for an analysis of disorder is partly driven by the common existence of harm that is not disorder. Both In Review papers in this issue1,2 illuminate the conceptual terrain and constructively explore the role of harm in distinguishing disorder from nondisorder. In addition, Cooper1 offers good reasons for not abandoning the dysfunction requirement or a natural kinds reading of disorder just because values are involved. However, it seems to us that these papers stop short of elaborating a sufficiently specific and satisfying account of dysfunction that explains what differentiates harms that are disorders from harms that are not disorders. The dysfunction requirement will be the focus of part 2 of this In Review, appearing in the December 2013 issue.
Acknowledgements
Dr Wakefield and Dr First have no funding or conflicts of interest to declare.
The Canadian Psychiatric Association proudly supports the In Review series by providing an honorarium to the authors.
References
1. Cooper RV. Avoiding false positives: zones of rarity, the threshold problem, and the DSM clinical significance criterion. Can J Psychiatry. 2013;58(11):606–611.
www.TheCJP.ca
2. Bolton D. Overdiagnosis problems in the DSM-IV and the new DSM-5: can they be resolved by the distress–impairment criterion? Can J Psychiatry. 2013;58(11):612–617. 3. Schneider ME. DSM-V Task Force weighing changes. Clin Psychiatry News. 2009;37(4):1–4. 4. Ustun B, Kennedy C. What is “functional impairment”? Disentangling disability from clinical significance. World Psychiatry. 2009;8:82–85. 5. Wakefield JC. Disability and diagnosis: should role impairment be eliminated from DSM/ICD diagnostic criteria? World Psychiatry. 2009;8:87–88. 6. Cooper R. What is wrong with the DSM? Hist Psychiatry. 2004;15(1):5–25. 7. Klein DF. A proposed definition of mental disorder. In: Klein DF, Spitzer RL, editors. Critical issues in psychiatric diagnosis. New York (NY): Raven Press; 1978. p 41–72. 8. Wakefield JC. The perils of dimensionalization: distinguishing personality traits from personality disorders. Psychiatr Clin N Am. 2008;31:379–393. 9. Wakefield JC. The concept of mental disorder: on the boundary between biological facts and social values. Am Psychol. 1992;47:373–388. 10. Greenberg G. Inside the battle to define mental illness [Internet]. Wired.com (magazine). San Francisco (CA): Condé Nast; 2011;19(1) [cited 2013 Sep 12]. Available from: http://www.wired.com/magazine/2010/12/ff_dsmv. 11. Bolton D. What is mental disorder? An essay in philosophy, science and values. Oxford (GB): Oxford University Press; 2008. 12. Wakefield JC, First MB. Clarifying the boundary between normality and disorder: a fundamental conceptual challenge for psychiatry. Can J Psychiatry. 2013;58(11):618–621.
The Canadian Journal of Psychiatry, Vol 58, No 11, November 2013 W 621
