Introduction
In their editorial ‘What is a mental/psychiatric disorder? From DSM-IV to DSM-V’, Stein et al. (Reference Stein, Phillips, Bolton, Fulford, Sadler and Kendler2010) propose several changes to the definition of mental disorder that currently appears in the introductory section of DSM-IV. They begin their commentary by noting that the definition of mental disorder is important because it guides us in distinguishing normal distress from disorder and thus in deciding which conditions should be in the DSM, and also in helping to set the threshold within a category that represents the boundary between disorder and non-disorder. There are also broader reasons for the definition's importance. As the inclusion in DSM of a section for ‘Other Conditions That May Be a Focus of Clinical Attention’ makes clear, psychiatry has functions other than treatment of disorder. It can use its knowledge to relieve the symptoms of normal distress or enhance human potential. Keeping these roles clear is important, and the definition of mental disorder is the main place that the DSM shows that psychiatry understands and observes this distinction. In fact, when diagnosing disorder, the clinician must frequently go beyond diagnostic criteria to make a judgment about disorder versus non-disorder, as in the frequent application of ‘not otherwise specified’ (NOS) diagnoses. A definition of mental disorder in the introduction offers some necessary guidance as to what is legitimate and what is not in such cases.
Analysis of proposed changes to definition
To facilitate examination of the proposed changes, we have depicted them using the format adopted by the DSM-IV Options Book (Task Force on DSM-IV, 1991) (see Table 1) and also depicted our own suggested revised version of the Stein et al. proposal (Table 2). We consider each of the criteria in turn.
Table 1.Stein et al.'s proposed changes to the DSM-IV definition of mental disorder
Table 2.Suggested revisions to the Stein et al. proposed definition of mental disorder
Criterion A
The only proposed change to criterion A involves moving ‘clinically significant’ from modifying the phrase ‘syndrome or pattern’ to modifying the distress or impairment component of criterion B. We view this change as without substantial impact because, in either position, it refers to the level of harm caused by the syndrome. This change does make the definition more consistent with the way the clinical significance criterion is currently deployed in diagnostic criteria sets.
Criterion B
The first change (i.e. clarifying that distress or disability is a consequence of the syndrome and not merely associated with it) makes sense. It will not substantially impact the definition because the causal nature of this relationship is widely assumed. The second change is the removal of ‘a significantly increased risk of suffering death, pain, disability, or an important loss of freedom’. We agree that this clause should be amended, given that there are no known mental disorders that cause death directly, without prior distress or disability (although for a definition including physical disorder, the addition of ‘death’ may be necessary); ‘pain’ is a form of distress in the intended broad sense; and loss of freedom is presumably a form of disability, as Stein et al. have argued. However, some provision for dramatically increased risk should be retained. Stein et al. correctly emphasize the importance of not confusing risk of disorder with disorder itself because many normal conditions elevate risk of harm, creating a danger of massive false positives. However, the definition's logic requires that only those increased risks that are consequences of internal dysfunction can be considered disorders. Such conditions can be legitimate disorders, as in physical medicine when persistently elevated blood pressure or an early stage malignancy prior to occurrence of harmful consequences is considered a disorder because of the likelihood of later harm from the dysfunction. One proposal being considered for DSM-V is to include a clinically defined pre-psychotic prodrome as an indicator that the individual is at significant risk of developing a psychotic disorder (Carpenter, Reference Carpenter2009). Although this proposal is, in our opinion, ill-advised because of the high potential of false positives (Heckers, Reference Heckers2009), if an objective laboratory marker that is a manifestation of a current dysfunction that is not yet causing distress or disability were to become available that reliably predicted the development of a psychotic disorder, that could be a legitimate category of disorder. The omission of all mention of future risk from the definition would seem to preclude its being considered a disorder. We therefore recommend that the phrase in question be replaced with ‘or substantial increased risk of future distress or disability’. However, to avoid confusing risk factors with disorders, we recommend that an additional sentence be added to clarify that syndromes without current distress or disability that carry with them significantly increased risk of future distress or disability should only be considered a disorder if due to a dysfunction (i.e. ‘Increased risk of distress or disability is not in itself a disorder unless due to a dysfunction’).
Criterion C
This DSM-IV criterion partially operationalizes the requirement (in criterion D) that the syndrome must represent a dysfunction in the individual; that is, that even if a syndrome causes distress or impairment (i.e. criterion A and B are met), it would not be considered a disorder if it did not also indicate that something has gone wrong in the individual (Wakefield, Reference Wakefield1992a, Reference Wakefieldb; Wakefield & First, Reference Wakefield, First, Phillips, First and Pincus2003). The DSM-IV definition provides a single clear example of a potentially distressing and disabling syndrome that would not be considered a disorder; that is, an expectable and culturally sanctioned response to a particular event, such as intense grief occurring after the death of a loved one. The proposed version splits this into two types of syndromes that should not be considered disordered because they do not represent a dysfunction in the individual: (1) expectable responses to common stressors and (2) syndromes that might seem to be pathological were it not for a consideration of the cultural context in which they occur (e.g. hearing the voices of dead relatives in a culture in which this is a culturally sanctioned experience). The motivation for this change seems to be a desire to emphasize consideration of cultural context in determining whether the syndrome represents a dysfunction. One problem is that the original DSM-IV wording (‘culturally sanctioned response to a particular event’) has been retained, effectively limiting consideration of cultural context to responses to events rather than a more general consideration of whether a particular behavior is culturally sanctioned, such as the behavior noted in their example (i.e. trance states in religious rituals). We therefore suggest that this phrase be replaced by ‘or to a culturally sanctioned behavior or belief’.
Criterion D
We agree that the term ‘psychobiological’ better captures the complexity of the interaction between biological and psychological factors that is often the case in mental disorders. However, it seems to us premature to assume in the very definition of mental disorder that every such disorder must involve a biological dysfunction. Note that this is different from saying that every psychiatric disorder involves brain processes, which is obviously true. However, to use one worn but illuminating analogy from cognitive science, every software program runs in hardware, yet not every software malfunction is a hardware malfunction; there is no chip failure despite a real software failure because the failure is at a higher level of description. Analogously, according to some etiological theories, there could be some disorders for which the dysfunction takes place at a psychological level of description involving the interaction of meanings, and no solely physiological level of description reveals a malfunction at that level. Thus, rather than violating theory neutrality, we suggest keeping the current DSM-IV phrasing. Because of the centrality of the ‘dysfunction’ criterion to the logic of the definition, we also suggest moving this criterion up to appear as the second sentence, immediately following criterion A.
Criterion E
Although this proposed change is motivated by a desire to simplify a wordy criterion, we contend that something important has been lost in the process. Many DSM disorders include behaviors that may put the individual in conflict with society. These so-called ‘vice disorders’ (First, Reference First2008), such as kleptomania, conduct disorder, substance dependence, and some of the paraphilias often involve behaviors that are illegal and thus may bring the individual into conflict with society. What makes these entities psychiatric disorders is the fact that they are a manifestation of an underlying dysfunction. We believe that this important requirement gets lost in the succinct phrase ‘not solely a result of’ because it does not explicitly indicate those circumstances in which it would be appropriate to consider these ‘conflicts with society’ disorder worthy. Given the historical importance of this issue, we would recommend that the DSM-IV wording be retained as making a clearer statement.
Criteria F and G
By recommending the inclusion of these two new criteria to the DSM-V definition, the authors seem to be confusing the purpose of having a definition of mental disorder in the introduction to the DSM with the need to establish criteria for deciding whether or not to include proposed new disorders or retain old ones in the DSM-V. Criteria F and G require that disorders included in the DSM have both diagnostic validity and clinical utility. Although these are important qualities in developing particular diagnostic categories and thus may usefully inform the DSM-V Task Force's decision making, they are not relevant to the key question addressed in the definition of mental disorder, namely, what makes a condition a mental disorder versus some other kind of human problem? Therefore, if a comment about validity and clinical utility is to be included, then rather than including it in the definition, we recommend that it be added as a codicil, along the following lines:
At this stage in the development of psychiatry, direct identification of an internal dysfunction (along with knowledge of etiology) is often lacking. Where the existence of a dysfunction is inferred or presumed, consideration should be given to the degree of validation of the category on the basis of key validators (e.g. prognostic significance, evidence of psychobiological disruption or prediction of response to treatment), with an understanding that construct-valid non-disordered conditions can also pass such validation tests. The clinical utility of a diagnosis should also be considered, keeping in mind, however, that clinical utility often only emerges in time as research on a category reveals the nature of the disorder and its effective treatments.
You have
Access
Introduction
In their editorial ‘What is a mental/psychiatric disorder? From DSM-IV to DSM-V’, Stein et al. (Reference Stein, Phillips, Bolton, Fulford, Sadler and Kendler2010) propose several changes to the definition of mental disorder that currently appears in the introductory section of DSM-IV. They begin their commentary by noting that the definition of mental disorder is important because it guides us in distinguishing normal distress from disorder and thus in deciding which conditions should be in the DSM, and also in helping to set the threshold within a category that represents the boundary between disorder and non-disorder. There are also broader reasons for the definition's importance. As the inclusion in DSM of a section for ‘Other Conditions That May Be a Focus of Clinical Attention’ makes clear, psychiatry has functions other than treatment of disorder. It can use its knowledge to relieve the symptoms of normal distress or enhance human potential. Keeping these roles clear is important, and the definition of mental disorder is the main place that the DSM shows that psychiatry understands and observes this distinction. In fact, when diagnosing disorder, the clinician must frequently go beyond diagnostic criteria to make a judgment about disorder versus non-disorder, as in the frequent application of ‘not otherwise specified’ (NOS) diagnoses. A definition of mental disorder in the introduction offers some necessary guidance as to what is legitimate and what is not in such cases.
Analysis of proposed changes to definition
To facilitate examination of the proposed changes, we have depicted them using the format adopted by the DSM-IV Options Book (Task Force on DSM-IV, 1991) (see Table 1) and also depicted our own suggested revised version of the Stein et al. proposal (Table 2). We consider each of the criteria in turn.
Stein et al.'s proposed changes to the DSM-IV definition of mental disorder
Suggested revisions to the Stein et al. proposed definition of mental disorder
Criterion A
The only proposed change to criterion A involves moving ‘clinically significant’ from modifying the phrase ‘syndrome or pattern’ to modifying the distress or impairment component of criterion B. We view this change as without substantial impact because, in either position, it refers to the level of harm caused by the syndrome. This change does make the definition more consistent with the way the clinical significance criterion is currently deployed in diagnostic criteria sets.
Criterion B
The first change (i.e. clarifying that distress or disability is a consequence of the syndrome and not merely associated with it) makes sense. It will not substantially impact the definition because the causal nature of this relationship is widely assumed. The second change is the removal of ‘a significantly increased risk of suffering death, pain, disability, or an important loss of freedom’. We agree that this clause should be amended, given that there are no known mental disorders that cause death directly, without prior distress or disability (although for a definition including physical disorder, the addition of ‘death’ may be necessary); ‘pain’ is a form of distress in the intended broad sense; and loss of freedom is presumably a form of disability, as Stein et al. have argued. However, some provision for dramatically increased risk should be retained. Stein et al. correctly emphasize the importance of not confusing risk of disorder with disorder itself because many normal conditions elevate risk of harm, creating a danger of massive false positives. However, the definition's logic requires that only those increased risks that are consequences of internal dysfunction can be considered disorders. Such conditions can be legitimate disorders, as in physical medicine when persistently elevated blood pressure or an early stage malignancy prior to occurrence of harmful consequences is considered a disorder because of the likelihood of later harm from the dysfunction. One proposal being considered for DSM-V is to include a clinically defined pre-psychotic prodrome as an indicator that the individual is at significant risk of developing a psychotic disorder (Carpenter, Reference Carpenter2009). Although this proposal is, in our opinion, ill-advised because of the high potential of false positives (Heckers, Reference Heckers2009), if an objective laboratory marker that is a manifestation of a current dysfunction that is not yet causing distress or disability were to become available that reliably predicted the development of a psychotic disorder, that could be a legitimate category of disorder. The omission of all mention of future risk from the definition would seem to preclude its being considered a disorder. We therefore recommend that the phrase in question be replaced with ‘or substantial increased risk of future distress or disability’. However, to avoid confusing risk factors with disorders, we recommend that an additional sentence be added to clarify that syndromes without current distress or disability that carry with them significantly increased risk of future distress or disability should only be considered a disorder if due to a dysfunction (i.e. ‘Increased risk of distress or disability is not in itself a disorder unless due to a dysfunction’).
Criterion C
This DSM-IV criterion partially operationalizes the requirement (in criterion D) that the syndrome must represent a dysfunction in the individual; that is, that even if a syndrome causes distress or impairment (i.e. criterion A and B are met), it would not be considered a disorder if it did not also indicate that something has gone wrong in the individual (Wakefield, Reference Wakefield1992a, Reference Wakefieldb; Wakefield & First, Reference Wakefield, First, Phillips, First and Pincus2003). The DSM-IV definition provides a single clear example of a potentially distressing and disabling syndrome that would not be considered a disorder; that is, an expectable and culturally sanctioned response to a particular event, such as intense grief occurring after the death of a loved one. The proposed version splits this into two types of syndromes that should not be considered disordered because they do not represent a dysfunction in the individual: (1) expectable responses to common stressors and (2) syndromes that might seem to be pathological were it not for a consideration of the cultural context in which they occur (e.g. hearing the voices of dead relatives in a culture in which this is a culturally sanctioned experience). The motivation for this change seems to be a desire to emphasize consideration of cultural context in determining whether the syndrome represents a dysfunction. One problem is that the original DSM-IV wording (‘culturally sanctioned response to a particular event’) has been retained, effectively limiting consideration of cultural context to responses to events rather than a more general consideration of whether a particular behavior is culturally sanctioned, such as the behavior noted in their example (i.e. trance states in religious rituals). We therefore suggest that this phrase be replaced by ‘or to a culturally sanctioned behavior or belief’.
Criterion D
We agree that the term ‘psychobiological’ better captures the complexity of the interaction between biological and psychological factors that is often the case in mental disorders. However, it seems to us premature to assume in the very definition of mental disorder that every such disorder must involve a biological dysfunction. Note that this is different from saying that every psychiatric disorder involves brain processes, which is obviously true. However, to use one worn but illuminating analogy from cognitive science, every software program runs in hardware, yet not every software malfunction is a hardware malfunction; there is no chip failure despite a real software failure because the failure is at a higher level of description. Analogously, according to some etiological theories, there could be some disorders for which the dysfunction takes place at a psychological level of description involving the interaction of meanings, and no solely physiological level of description reveals a malfunction at that level. Thus, rather than violating theory neutrality, we suggest keeping the current DSM-IV phrasing. Because of the centrality of the ‘dysfunction’ criterion to the logic of the definition, we also suggest moving this criterion up to appear as the second sentence, immediately following criterion A.
Criterion E
Although this proposed change is motivated by a desire to simplify a wordy criterion, we contend that something important has been lost in the process. Many DSM disorders include behaviors that may put the individual in conflict with society. These so-called ‘vice disorders’ (First, Reference First2008), such as kleptomania, conduct disorder, substance dependence, and some of the paraphilias often involve behaviors that are illegal and thus may bring the individual into conflict with society. What makes these entities psychiatric disorders is the fact that they are a manifestation of an underlying dysfunction. We believe that this important requirement gets lost in the succinct phrase ‘not solely a result of’ because it does not explicitly indicate those circumstances in which it would be appropriate to consider these ‘conflicts with society’ disorder worthy. Given the historical importance of this issue, we would recommend that the DSM-IV wording be retained as making a clearer statement.
Criteria F and G
By recommending the inclusion of these two new criteria to the DSM-V definition, the authors seem to be confusing the purpose of having a definition of mental disorder in the introduction to the DSM with the need to establish criteria for deciding whether or not to include proposed new disorders or retain old ones in the DSM-V. Criteria F and G require that disorders included in the DSM have both diagnostic validity and clinical utility. Although these are important qualities in developing particular diagnostic categories and thus may usefully inform the DSM-V Task Force's decision making, they are not relevant to the key question addressed in the definition of mental disorder, namely, what makes a condition a mental disorder versus some other kind of human problem? Therefore, if a comment about validity and clinical utility is to be included, then rather than including it in the definition, we recommend that it be added as a codicil, along the following lines:
At this stage in the development of psychiatry, direct identification of an internal dysfunction (along with knowledge of etiology) is often lacking. Where the existence of a dysfunction is inferred or presumed, consideration should be given to the degree of validation of the category on the basis of key validators (e.g. prognostic significance, evidence of psychobiological disruption or prediction of response to treatment), with an understanding that construct-valid non-disordered conditions can also pass such validation tests. The clinical utility of a diagnosis should also be considered, keeping in mind, however, that clinical utility often only emerges in time as research on a category reveals the nature of the disorder and its effective treatments.
Declaration of Interest
Dr First consults with pharmaceutical companies to provide diagnostic training for clinical trials, and receives book royalties from APPI and Wiley-Blackwell. In the past 12 months, he has consulted with Cephalon, GlaxoSmithKline, Memory Pharmaceuticals, Worldwide Clinical Trials, and i3 research. Dr Wakefield receives book royalties from Oxford University Press.