Kenneth S. Kendler in World Psychiatry, 2018
In their paper in this issue of the journal, Krueger et al1 consider two different approaches to the classification of psychopathology. Here I would like to focus on the conceptual and historical background of these approaches. What the authors call “authoritative” nosology – represented currently in the US by the DSM-5 system2 – evolved from classificatory efforts starting in the late 17th century, when large numbers of patients began to be collected in asylums in Central and Western Europe. These efforts were based on earlier attempts to classify general medical conditions, which were in turn heavily influenced by systems that classified animal and plant species as part of the beginning of zoology and botany as descriptive sciences3 . So, I agree with the authors that the DSM is an historically influenced document, but I see this more positively than they do. Medicine has worked by a gradual evolutionary process of the articulation of broad syndromes, many of which, with advancing knowledge, become divided up into more homogenous entities that then develop into what we might call “disorders” and eventually “diseases”. In psychiatry, this process has been slower and more difficult than in most areas of medicine, but still represents an accumulated wisdom that typically works pretty well in the real world of patient care. How well it serves the goals of research is another matter. As this brief history suggests, categories are inextricably intertwined with the world of clinical medicine. Individuals in care need to be given diagnoses because of the key dichotomies that exist in this world – to treat or not to treat, to discharge (from an emergency room) or hospitalize, to qualify for a particular treatment algorithm or not, to bill or not and, if to bill, with what specific code. This does not, of course, preclude quantitative measures, the focus of the nosologic approach advocated by the authors. These too are woven into the fabric of medicine. Think of temperature, pulse rates, fasting blood sugar, white blood cell counts and bone densities. These measures happily co-exist with the diagnostic world and are used nearly universally to monitor health and illness and guide therapy. I worry that underneath this debate about continua versus categories there is a confusion between the “levels” of underlying physiology/etiology and clinical manifestation. Let me illustrate this by a “thought experiment”: A steep south-facing slope in the high mountains received a heavy snowfall. The next morning dawns warm with a clear sky and strong sun. The temperature – a classical quantitative variable – at the lower levels of the snow pack starts to rise and melting increases gradually throughout the morning. Suddenly, in mid-afternoon, the snow pack starts to slide, ending in a dramatic avalanche. This example illustrates a natural quantitative process – snow melting with increasing temperature – and a dramatic threshold effect. If you work for the ski patrol to prevent avalanches, you need to understand both processes. Turning to medical applications, consider a femur with increasing levels of strain – a quantitative trait. At some point, the bone breaks with dramatic health consequences. Think of a coronary artery with increasing occlusions as cholesterol plaques increase. At some point, the blood flow and associated delivery of oxygen slips below a critical level. Heart tissue starts to die and a myocardial infarct occurs. I agree that taxometric methods provide at most modest evidence for discrete diagnostic categories in psychiatry. But I want to add to this discussion a different and informative perspective – within individual analyses. Like when seeing an avalanche, when seeing an acute patient presenting in the emergency room with a broken femur or an active myocardial infarction, it is difficult to conclude that one should only be concerned with the underlying quantitative process. Something clinically dramatic and “categorical” has occurred that calls for immediate intervention. Consider the following brief psychiatric vignettes: A vulnerable individual, who stopped his antipsychotic medication four weeks ago, over 48 hours transitions from a non-psychotic state to a full-blown psychosis characterized by active auditory hallucinations and persecutory delusions about which he is quite preoccupied. An individual with prior bipolar illness in good remission, after traveling across five time zones and experiencing several nights of poor sleep, the next day, “flips” into a fully syndromal mania. You observe a friend with panic disorder in a crowded restaurant go from a calm, collected state in less than a minute to one of acute distress with sweating, panting, shaking and fear of dying. While not all psychiatric disorders have such dramatic “avalanche-like” transitions, they are fairly common in clinical psychiatry and challenge the authors’ conclusions that there is little viable evidence that psychiatric disorders need to be understood from a categorical perspective. Let me turn to a quite different issue. I was concerned by the manner in which the authors characterize the DSM proWorld Psychiatry 17:3 - October 2018 241 cess: “group discussions and associated political processes”, manifesting “sociopolitical dynamics”, issuing ex cathedra decisions with the final diagnoses resulting from “presumed authority and fiat”. This tone will not aid interdisciplinary discourse. The authors imply that they are the objective scientists while those who worked on DSM are, by comparison, bogged down in political discourse and constrained by old-fashioned historical dictates. While this is not the place to discuss this in detail, any organized effort in science to develop classifications involves “sociopolitical dynamics”. Readers who think otherwise might consult a history of the decision of the International Astronomical Union to remove Pluto from the official list of planets4 . I want to conclude by talking about standards of diagnostic validation. At the risk of over-simplification, the Hierarchical Taxonomy of Psychopathology (HiTOP) program emphasizes psychometric methods in its typological proposals. Such methods have been key in the history of psychology, for example in the development of personality typologies and measures of various cognitive skills. So, it is sensible that they should be applied in the area of psychopathology. However, this approach differs considerably from the medical tradition emphasized by DSM. Put simply, the medical tradition wants diagnoses that tell us a lot about the patient – the course, the likely etiologic process, the best treatment, etc.. We organize our literature around our diagnoses, from cohort studies to randomized controlled trials. The specific articulation of this viewpoint in psychiatry was first given by Robins and Guze5 with their list of validators, substantially expanded since then. Since DSM-III, the role of the evaluation of validators in diagnostic change has, albeit somewhat unevenly, gradually increased. The main approach has been the use of literature reviews trying to summarize available information on validators. These questions were the specific focus of the Scientific Review Committee that evaluated every proposed diagnostic change in DSM-56 . The procedures developed for change in DSM-5 by the American Psychiatric Association’s Steering Committee are empirically rigorous and data driven7 . It is not surprising that the scientific disciplines of psychiatry and clinical psychology have developed different approaches to the creation and evaluation of diagnostic entities/ dimensions. Optimal communication between these two disciplines, however, requires an understanding of the similarities and differences in these approaches, the relative strengths and limitations of each approach, and the acceptance by both sides that each is likely to be able to contribute meaningfully to the difficult challenge of designing an optimal psychiatric classification.
Kenneth S. Kendler Virginia Institute for Psychiatric and Behavioral Genetics and Department of Psychiatry, Virginia Commonwealth University, Richmond, VA, USA
1. Krueger RF, Kotov R, Watson D et al. World Psychiatry 2018;17:282-93.
2. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 5th ed. Arlington: American Psychiatric Association, 2013.
3. Kendler KS. Psychol Med 2009;39:1935-41.
4. Zachar P, Kendler KS. Philos Ethics Humanit Med 2012;7:4.
5. Robins E, Guze SB. Am J Psychiatry 1970;126:983-7.
6. Kendler KS. Psychol Med 2013;43:1793-800.
7. First MB, Kendler KS, Leibenluft E. JAMA Psychiatry 2017;74:115-6.