Blue is the new black

Dr Gupta's paper adds to the swell of doubts about the current DSM classification of depressive disorders. Numerous concerns have been raised by psychiatrists about “major depression”, including myself (Parker 2008), as well as by non-psychiatrists – with Horwitz and Wakefield's 2007 book The Loss of Sadness exemplifying logical challenges to DSM-defined “major depression”.

In this issue Gupta offers another challenge to such DSM classification. He argues that “depression”, “major depression”, “major depressive illness”, and “major depressive disorder” are better defined and distinguished by “objective physical signs”. By so variably defining “depression”, Gupta unfortunately leaves himself open to claims of ineffability as faced by the proponents of “major depression”. In essence, what exactly are we talking about when we use the term “depression” or the diagnosis of “major depression” so variably.

Sir Denis Hill (1968) argued that “depression” can exist as a disease, a reaction, a normal state, a posture and even as an existential dilemma. In essence, “depression” is not an entity. We should not talk about “it”. It is more appropriate to view “depression” as no more than describing a domain of psychological perturbation. At that level, we should expect some generic defining features (not necessarily symptoms) – such as a depressed mood, lowered self-esteem, increased levels of self-criticism, and feelings of hopelessness and helplessness. As such mood states are a universal human experience, we need then to define differences between such states and clinical conditions, and there has been a dramatic change in building such distinctions in recent times. The DSM model is to require the individual to have a set number of prescribed criteria, be depressed for a defined period and also – as a consequence of the depression – have a level of impairment. Such severity, duration and impairment criteria appear reasonable at face value but we can question (as Gupta does) how well they operate in practice. In essence, do they distinguish between “normal depression” and “clinical depression” so as to neither risk under-diagnosing nor over-diagnosing clinical syndromes? This issue (i.e. the actual boundary of clinical depression) needs to be faced before considering whether “depression” should or should not be defined by physical signs.

Until the twentieth century there was the longstanding binary view, as detailed by Altschule (1967). He suggested that there was early biblical differentiation of depressions – which contrasted those that came “from God” and those that reflected causes that were more of “the world”. In essence, Altschule argued that Christian leaders (such as John Cassian, Isidore of Seville, Bede) recognized the distinction between “irrational” depression (which emerged frequently without any antecedent stressor and was therefore seemingly from God) and “rational” depression which followed life event stresses and was therefore explainable as a consequence of “worldly” stressors. The binary model of such separate “endogenous” and “reactive” depressions remained dominant until the mid-twentieth century. Such a distinction had important implications in terms of defining “meaningful” depressive disorders, weighting the “irrational” and effectively normalizing the “rational” or “reactive” disorders.

Thus, until DSM-III emerged in the 1980s, clinical depression was largely equated with “endogenous” depression (now more commonly termed melancholic depression). These were the patients who generated rich phenomenological descriptions of clinical depression (Jackson 1986), who were most likely to be admitted to psychiatry units and who dominated outpatient practice. Epidemiological studies that so focussed on such a depressive condition returned a low prevalence. In fact, when the first antidepressant (imipramine) was discovered, the manufacturer was reluctant to release it as it was judged that the low prevalence of “depression” would not generate a viable market.

By contrast, those with a non-endogenous or non-melancholic depressive disorder were frequently viewed as “neurotic” or having slight conditions, and therefore not having a meaningful clinical disorder, as the latter group can include those with severe states. Some correction was clearly needed.

The consequential unitarian DSM-III model effectively consolidated melancholic depression and a number of non-melancholic conditions into a single principal category, “major depression”. As Gupta observes, only a small set of mood-related items are necessary for an individual to achieve a diagnosis of “major depression”, and with the criteria descriptors kept at a low level of inference, it is not particularly difficult for a depressed individual to meet criteria for major depression. While prior to DSM-III “clinical depression” was estimated as having a 2–5% lifetime risk (being essentially dominated by those with melancholia and bipolar depression), post-DSM-III studies led to the commonly quoted estimate of a 1 in 4 chance for women and a 1 in 6 chance for men of having clinical (or “major”) depression over their lifetime. When clinical depression is broadened to include the DSM minor groupings and the sub-syndromal disorders, lifetime rates would suggest that “clinical depression” is both a broad church (like “depression”, allowing diseases, disorders and normal reactions) and rishing a ubiquitous experience (Parker 2007).

The low cut-off thresholds predictably raise substantive questions. Firstly, are our current cut-off criteria valid in defining “clinical depression”? The two contrary positions need to be considered. Firstly, as the low-level cut-off accords “clinical depression” status to a much higher percentage of individuals with a depressive condition, it redresses the suggestion that non-melancholic disorders may be trivial and therefore not worthy of serious consideration – or even insurance or hospitalization benefits. Secondly, and conversely, it invites the criticism that non-substantive states (including sadness, normal depression and human misery) are “pathologised”. This is a debate in and of itself.

The more important issue – in relation to Gupta's paper – is how we view “major depression”. Unfortunately, it is commonly viewed as an entity – and researchers seek the cause and pursue treatments for “major depression” as if it were a homogeneous condition. Further, it has accrued gravitas and status, and it is interesting to consider how far this status was advanced merely by the adjectival descriptor “major”. Further, those who accord it high status, frequently make statements such as “major depression is a disease … (or illness)”. There are two problems here. Firstly, in positioning “major depression” as a single homogeneous entity. Secondly, in according it “disease” or “illness” status of necessity.

It is far wiser to view “major depression” as a domain diagnosis, akin to saying that a patient has clinical dyspnoea and then pursuing whether that domain diagnosis is caused by a particular condition (e.g., asthma, pneumonia, a pulmonary embolus) or might merely be a normal reaction (e.g., to excessive exercise). Similarly, pain may be normal and reactive, or pathological and indicative of an underlying medical condition. Merely describing dyspnoea or pain as “major” would not be particularly helpful in pursuing cause or treatment.

Thus, if we accept that “major depression” is no more than a domain diagnosis encompassing differing depressive sub-types, we are obliged to then identify the clinically meaningful sub-types to allow pursuit of their differential causes and their consequential differential treatments. Here the task becomes extremely difficult as attempts to define “meaningful” depressive sub-types (endogenous vs. reactive; melancholic vs non-melancholic) have failed to convince the scientific community over the decades.

Our research group has certainly argued that melancholia is a distinctive sub-type with specific and over-represented clinical features, with observable psychomotor disturbance being a key marker and a “signal” of perturbed neurobiological processes implicating neurocircuits linking the basal ganglia and the prefrontal cortex, and that this condition responds better to physical treatments (such as drugs and ECT) than to the psychotherapies. In an earlier review (Parker and Hadzi-Pavolvic 1996), we detailed historical descriptions (going back to ancient times) that positioned melancholia as more of a movement disorder than a mood disorder, with observable psychomotor disturbance being a key feature. We operationalized observable psychomotor disturbance via a set of 18 CORE signs and in several studies, established that the CORE measure differentiated melancholic from non-melancholic depression to such a degree that it made any need for defining melancholic symptoms somewhat redundant. In essence, observable psychomotor disturbance (as operationalized) was a “necessary and sufficient” feature to define melancholia. This model therefore argues that psychomotor disturbance is the defining feature of melancholia and while it is best measured as an observable sign, there are symptom correlates. Patients with a melancholic depression tend to report marked anergia and anhedonia, and with both mood and energy worse in the morning, while they tend to become more insular, describe poor concentration and have a non-reactive and anhedonic mood state. By contrast, those with a non-melancholic depression do not have such observable physical signs, and they are less likely to report such prototypic symptoms. Thus, physical signs are present – and effectively define –one depressive sub-type (melancholia) and are absent in residual non-melancholic clinical conditions.

Returning then to Gupta's paper, rather than argue that “major depression” is an illness with objective physical signs, we would favour the argument that any such construct as “major depression” or “clinical depression” is no more than a domain diagnosis encompassing constituent disease states, with predictably varying clinical features. Within that grouping there exists a melancholic sub-type which may be identified more precisely from observable physical signs than by DSM-IV symptoms. In essence, major depression should not be defined by physical signs but one of its constituent members (melancholia) may well be. If such a modification of Gupta's model is valid, then testing of the proposition is warranted to advance diagnostic clarification. Gordon Parker Scientia Professor School of Psychiatry, University of New South Wales Executive Director, Black Dog Institute Correspondence: Prof. G. Parker Black Dog Institute Prince of Wales Hospital Randwick 2031, Australia Tel: [phone] E-mail: [email protected]

Acknowledgements

None Statement of interest The author has no conflict of interest with any commercial or other associations in connection with the submitted article.