The need for a conceptual framework in psychiatry acknowledging complexity while avoiding defeatism

The notion – often put forward nowadays – that what we call mental disorders are just convenient constructs that may or not appropriately reflect what exists in the real world is very likely to be misinterpreted by several psychiatrists (contributing to their current frustration about the status of our profession), by many colleagues within the medical community (reinforcing their skepticism towards our discipline), by the general public (already sensitized by the recent debates following the publication of the DSM‐5), and by people with mental health problems and their carers (further discouraging them from seeking our help and listening to what we say). The primary (but not the only) problem is that the difference between this notion (which acknowledges both the existence of mental illness and the limitations of current diagnostic categories) and the radically “constructivist” position of Szasz and others (stating that current diagnostic categories are just a theoretical fiction or a myth, i.e., that there is no such thing as mental illness except by metaphor) is not easy to grasp for someone who does not have a philosophical background.

Much easier to understand and less destructive is the notion that many of what we call mental disorders, although not qualifying at the moment as proper “disease entities”, are indeed patterns of observed signs and reported symptoms that trained clinicians have been able to recognize for decades in a variety of clinical contexts and in the community (although also noticing their frequent co‐occurrence as well as the existence of intermediate and subthreshold forms) and have been managing with a degree of success that, although less than optimal, is actually comparable to that achieved by many other branches of medicine for the conditions they deal with.

It is certainly true that several diagnostic concepts in psychiatry have changed to some extent through the years and that some of them have disappeared along this way. Also, several diagnostic categories have been split or lumped in a way that is questionable. There is surely much room for improvement in our diagnostic practices. However, it would be difficult for me to identify a substantial difference between the history and the current characterization of, say, the mental disorder called depression versus the non‐mental disorder called migraine. Both of them are defined syndromally, and mainly on the basis of what the person reports; both of them have an unclear and certainly heterogeneous etiopathogenesis; both of them have been classified and subtyped differently along the decades; and both of them have various clinical presentations, including formes frustes, and fuzzy diagnostic boundaries. It would also be difficult for me to accept that, imagining to turn the clock back ten thousand years and allow human civilization to develop again – as K. Kendler proposes as a thought experiment in this issue of the journal1 – the pattern of depression would be less likely to emerge and be identified than that of migraine (unless, of course, the nature itself of human beings were to be totally different).

True, the project launched in the early 1980s to validate DSM‐III categories by elucidating their “specific” etiopathogenetic underpinnings2 seems to have failed, but the picture that has gradually emerged during the past 35 years does represent in itself a prominent scientific advance, that the use of the DSM‐III and its successors has not obstructed. We know today that the etiopathogenesis of most or possibly all patterns of mental disorder is very complex, involving the interaction of a multiplicity of biological, intrapsychic, interpersonal and socio‐cultural factors. We also know that several of these factors are not specific for individual DSM/ICD categories. This complexity is not only due, as frequently stated, to the fact that the brain is a much more complex organ than the others we have in our body, but more crucially to the fact that mental disorders are not merely “brain diseases”, but actually emerge at the interface between that complex organ which is the brain and the even more complex world of interpersonal relationships in which we are all immersed.

For some patterns of mental disorder, e.g. eating disorders, the role of sociocultural factors in shaping their psychopathological identity is already obvious, but even for patterns such as psychotic disorders there may be some distance between any neurobiological mechanisms that we are likely to elucidate and the level at which their psychopathological identity emerges. So, taking for granted that these patterns can be fully “explained” at the neurobiological level, and feeling defeated or blaming our discipline because we are unable to do so, may be inappropriate, and the elucidation of the “higher‐order processes”³ which are involved may be crucial (see, for instance, Howes and Nour⁴ in this issue of the journal). Furthermore, several different neurobiological processes may have a role in each of the limited number of patterns of mental disorder that human beings are able to express, and the same neurobiological process may be involved in several of those patterns.

I am also not very keen of the distinction between “utility” and “validity” of psychiatric diagnoses. There is an extensive overlap between what is called today “utility” and what used to be called “predictive validity”. If the utility of a diagnostic entity resides in its ability to predict further course and response to treatments, then the ascertainment of that utility is an intrinsic component of the “validation” process delineated by Robins and Guze5. And it would be appropriate to pay some attention to that component because, if the project of validating our current diagnostic entities by elucidating their specific etiopathogenetic underpinnings may have failed2, other components of the above validation process may have been less unsuccessful, although also requiring a refinement. Otherwise, all the clinical research of the past 35 years may risk to be thrown into wastebasket, which would probably be a mistake.

On the other hand, we have to distinguish between the “utility” of a given diagnostic category and the “utility” of a whole diagnostic system. The DSM and ICD may be not sufficiently “useful” for ordinary clinical practice, in the sense that they may have features which discourage their use by clinicians. We have indeed some evidence6 that a substantial proportion of psychiatrists worldwide do not use formal diagnostic systems in their ordinary practice, or use them just as “coding systems” (i.e., they use the ICD codes in clinical records and other similar documents, but do not have in mind the ICD descriptions when they use those codes, or have never read those descriptions). Certainly something should be done, and to some extent is being done7, in this respect.

I think that psychiatrists worldwide, and the people with whom they interact daily (colleagues of other medical disciplines, other mental health professionals, politicians, administrators, journalists, patients, carers, residents, students), need today a conceptual framework which explicitly acknowledges the above complexity and the oversimplifications which may have occurred, while avoiding to indulge in a pessimism that may be excessive and destructive.

Mental disorders may not be “disease entities” in the proper philosophical sense, but a large proportion of them are certainly not theoretical fictions. They are patterns of observed signs and reported symptoms that trained psychiatrists are able to recognize and manage, often successfully, in clinical settings and in the community. We do not have laboratory tests on which to base our diagnoses, but this means that psychiatrists are expected to be very skilled clinicians, and that high‐quality clinical training is even more important in psychiatry than in other medical disciplines.

It is not true that there has been no progress in etiological research in psychiatry in the past 35 years. On the contrary, we have learnt that the etiopathogenesis of most mental disorders is very complex, involving the interaction of a multiplicity of biological, intrapsychic, interpersonal and sociocultural factors, that research is gradually identifying and weighing. No simple explanations are to be expected, though the complex models which may emerge will need to be made understandable by all the above‐mentioned stakeholders.

Neurobiological mechanisms are likely to be involved in most or all mental disorders, but the level at which the psychopathological identity of these disorders emerges may be higher than that of the brain machinery, and the elucidation of the higher‐order (e.g., psychological, cultural) processes which intervene may be crucial. Therefore, a dialogue should be kept between the neurosciences and other (anthropological, psychological, social) sciences when exploring the etiopathogenesis of what we should probably accustom ourselves to more exactly conceptualize, following the latest Kraepelin8, as “patterns of mental disorder”.

Mario Maj
Department of Psychiatry, University of Naples SUN, Naples, Italy