Psychiatria Danubina, 2015; Vol. 27, Suppl. 1, pp 7–20 Conference paper © Medicinska naklada - Zagreb, Croatia CLASSIFICATION IN PSYCHIATRY: FROM A SYMPTOM BASED TO A CAUSE BASED MODEL? Dylan Pritchard Clare College Cambridge, Cambridge, UK SUMMARY The assumption that eventually the classification in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM) will incorporate aspects of causation uncovered by research in neuroscience is examined in view of the National Institute of Mental Health’s NIMH Research Domain Criteria (RDoC) project. I argue that significant advantages of maintaining the classification system, focussed on grouped descriptions of symptoms, are often undervalued or not considered. In this paper I will challenge the standard view that the transition from the purely symptom based approach is an inevitable and desirable change. Key words: classification – symptoms - mental illnesses - biomarkers * * * * * INTRODUCTION logies are still in their infancy, it is apparent that they have worth when it comes to identifying linkages bet- A standard view exists in much of psychiatry that ween the physical brain and mental disorders. More new classification in the American Psychiatric Association’s insights are occurring at the biochemical level with Diagnostic and Statistical Manual of Mental Disorders alterations in neurotransmitter composition being shown (DSM) will incorporate aspects of causation uncovered to be strongly linked with various mental disorders, by research in neuroscience. Currently the DSM aims to such as dopamine with the positive symptoms of schizo- minimise the influence of notions of causation and phrenia (Seeman 2011). These findings suggest to some aetiology, omitting any deeper theoretical underpinning. that molecules can be the cause of a disease. A third Arguments in favour of the standard view focus on two level for many researchers is genetics. With the increa- main factors, first of all that the introduction of sing ease with which a genome can be sequenced and causation is necessary to ensure classification is stable large scale genomic analyses, genetic and genomic and “carving nature at its joints” (Regier 2009), and se- linkages to mental disorder are being demonstrated. condly that future research will empirically demonstrate Huntington’s disease is the prime example of this as the that adding aetiology into classification will bring ad- HD gene has been shown to be the necessary and suffi- vantages in treatment. Much of the literature addresses cient cause of the disease (Walker 2007). Accordingly, when and how such a shift to include causes in classi- the introduction of new techniques for visualising the fication will occur. Significant advantages of main- living brain coupled with subsequent findings sugges- taining the classification system, focussed on grouped ting causal linkages, have brought forward the debate descriptions of symptoms, are often undervalued or not around the consequences of such research for psychia- considered. In this paper I will challenge the standard tric classification. view that the transition from the purely symptom based These advances have led many prominent figures in approach is an inevitable and desirable change. psychiatry to advocate an increased role for such Calls for such a shift are common in the field of physical causal theories in the DSM. Nancy Andreason, psychiatry. There is a desire for an alteration in the as a member of the task force for the development of underpinnings of the DSM to more closely reflect the DSM-III, and DSM-IV in an editorial of the American new research being conducted into the linkages between Journal of Psychiatry suggests that, “psychiatry is not mental disorder and physical processes within the brain. only founded on diagnoses that are validated by clinical The calls for a change are largely driven by recent description and epidemiological criteria, but it is technological advances in neuroscience. One of the challenged by the opportunity to probe more deeply into levels this is occurring at is advances in analysing brain mechanisms and perhaps to reach very fundamental structure, with the increased use of computerised tomo- levels of knowledge about etiology that will have graphy (CT) scanning and functional magnetic reso- profound implications” (Andreason 1995). In a piece by nance imaging technology (fMRI) to visualise the the chairs of the committee developing the most recent structure of the brain in living patients. These techno- DSM (DSM-5) (American Psychiatric Association logies have allowed researchers to visualise the brains 2013), the belief in an eventual move towards an of people with mental disorders such as in schizophrenia aetiologically based classification is again stressed with and find significant differences when compared to a focus on investigating, “how advances in neuroscience healthy individuals (Wright 2000). Whilst these techno- and behavioural science over the past two decades have S7 Dylan Pritchard: CLASSIFICATION IN PSYCHIATRY: FROM A SYMPTOM BASED TO A CAUSE BASED MODEL? Psychiatria Danubina, 2015; Vol. 27, Suppl. 1, pp 7–20 both increased our knowledge of etiologies and widened to less radical versions such as that of Dominic Murphy the gaps in our existing diagnostic system. A primary which allow disorders to be defined by causes on a goal for revising DSM-5, then, is to more fully variety of levels from the genetic and biomolecular incorporate research from the past two decades in an through to the psychological and social. The common attempt to build an empirical foundation” (Kupfer theme is that research evidence from studies in sub- 2008). The chairs of the DSM-5 committee in their disciplines of neuroscience will eventually prove too summary of the continuing evolution of the DSM significant to be ignored. As these views are projections emphasised the importance of advances in neuroscience into the future there is little consensus on what this prioritising, “The opportunity to evaluate the readiness would entail for the current mental disorders in DSM-5. of neuroscientific advances in pathophysiology, gene- Research may radically shift, or simply reaffirm the tics, pharmacogenomics, structural and functional current classes in psychiatry. The consensus amongst imaging, and neuropsychology” (Kupfer 2008). There is the proponents of a role for neuroscientific theory in a clear assumption within psychiatry that classification classification is that the classification systems of mental based on description alone is simply waiting to be disorders will define disorders using underlying causes. superseded by what is seen as an inherently superior The justifications for this approach revolve around system incorporating aetiology to produce its defini- two key arguments. The broader philosophical argument tions. However, I contend that when the potential advan- states that classifications based purely on symptoms are tages of cause based classifications are weighed against misguided as they are not representing an aspect of the the oft overlooked benefits of the description based nature of mental disorders. Neuroscience holds that system, it becomes clear that the description based mental disorders are products of the brain and classi- model is the superior option for classifying the majority fication should reflect this by including some references of mental disorders. to structure and theory linking structure to disorder. More confidence could be had in the disorder categories BACKGROUND if they could be shown to reflect distinct causal processes. This would lead to classifications which Current classification in psychiatry is based on des- would be more stable over time, in the sense that they cription of symptoms, the descriptive or atheoretical would be less prone to changing across the editions of approach. It aims to reduce the involvement of causes, the DSM. Treatment of mental disorders would be an approach enshrined in the DSM since the third improved if the categories were made more stable in edition in 1980. For instance, the previous second edi- this way as the targets for treatments would be less tion of DSM had included the neuroses as one of its prone to change. The second argument arising from this nine major diagnostic groups” (Kupfer 2008). These is that achieving a classification based on causes of were explicitly defined in the causal terms of psycho- disorders would result in improved treatments available analytic theory as an anxiety which was, “felt or expres- to patients acting at the level of the causes of disease. It sed directly, or ... may be controlled unconsciously and is assumed that studying causation in neuroscience automatically by conversion, displacement and various would lead to the development of new treatments acting other psychological mechanisms” (American Psychia- at the causal level. By classifying disease with reference tric Association 1968). The notion of causation was to its neuroscientific causes, such new interventions included in the core definition of what the disorder was would be optimally targeted, improving the treatment of with the reference to various psychological mechanisms patients. The advantages of a more accurate represen- which caused the manifestation of symptoms. The third, tation of the nature of mental disorders and conse- fourth and fifth editions, with their atheoretical aims, quently the development of treatments acting at these classify very differently.