The problem with the criteria

First, adjustment disorder is poorly delineated in both DSM-IV-TR and ICD-10. ⁴ The boundary between adjustment disorder and normal adaptive stress is not addressed, although the requirement that dysfunction must be present in the ICD descriptor is a tacit attempt to deal with this. The differentiation from other psychiatric disorders such as major depression and generalised anxiety is a further problem since the criteria are underdeveloped and rudimentary. Apart from specifying that adjustment disorder requires a stressor and that the symptoms resolve within 6 months of termination of the stressor or its consequences, no assistance is offered with regard to the nature or configuration of the symptoms. Instead, ICD-10 ⁴ and DSM-IV-TR ⁵ state that the diagnosis cannot be made when the symptom threshold for another condition is reached. In the case of major depression in DSM-IV-TR, that threshold is reached after 2 weeks, a duration that, in clinical practice, seems short.

One of the consequences of the inadequate criteria of adjustment disorder is that some of the structured diagnostic interviews have failed to include it and those that have, such as the Structural Clinical Interview for DSM-IV (SCID-I), ^{Reference First, Spitzer, Miriam and Williams6} only diagnose it after other diagnoses have been excluded. In light of these considerations, the low prevalence of adjustment disorder found in the Fernandez et al study is not surprising since SCID-I was used, which is derived from the principles of ICD-10 and DSM-IV that classify adjustment disorder as a residual category. A similar problem also arose in the ODIN study, which also found an unexpectedly low prevalence ranging from 1 to 1.9% in the general population. ^{Reference Ayuso-Mateos, Vázquez-Barquero, Dowrick, Lehtinen, Dalgard and Casey7}

With regard to its seemingly low prevalence, some studies have found a discrepancy between adjustment disorder when diagnosed clinically as compared to using a structured interview. Clinical diagnosis has identified a higher prevalence for adjustment disorder, which, when structured interviews are used, is replaced by major depression. For instance, among new psychiatric out-patients, adjustment disorder was diagnosed in 36% of those seen, but this dropped to just over 11% using SCID, ^{Reference Shear, Greeno, Kang, Ludewig, Frank and Swartz8} while among a population assessed following self-harm, a clinical diagnosis of adjustment disorder was made in 31.8% and major depression in 19.5%, but when using SCID the proportions were changed to 7.8% and 36.4% respectively. ^{Reference Taggart, O'Grady, Stevenson, Hand, Mc Clelland and Kelly9}

This raises the question of the utility of the current crop of structured interviews in evaluating adjustment disorder. Although these are generally seen as the gold standard in psychiatric research, they are based simply on cross-sectional assessment of symptom numbers and their minimum duration, whereas adjustment disorder is a longitudinal diagnosis based on aetiology and outcome. So the construct of adjustment disorder is not captured within the framework of the context-free, cross-sectional approach of the current classifications and their associated diagnostic schedules. This is one issue that can only be addressed when due prominence is given to adjustment disorder in the revisions to DSM and ICD that take account of aetiology and course, and when new interview schedules are developed.

What are the research implications?

Does it matter that adjustment disorder is regarded as a subsyndrome, that it appears to be uncommon in epidemiological studies and that it is underresearched? One of the consequences of regarding adjustment disorder as a subclinical category is that it is viewed as mild in comparison to other full-blown conditions and less worthy of research than other disorders. Although Fernandez and colleagues found that in terms of severity it lay between major depression and no psychiatric disorder, other studies have pointed to adjustment disorder as a much more serious condition, particularly in respect of suicidal behaviour. One psychological autopsy study found that adjustment disorder was the most common diagnosis, ^{Reference Manoranjitham, Rajkumar, Thangadurai, Prasad, Jayakaran and Jacob10} while among those presenting to emergency departments following self-harm it was the most common psychiatric diagnosis. ^{Reference Taggart, O'Grady, Stevenson, Hand, Mc Clelland and Kelly9}

One of the consequences of the lack of attention to adjustment disorder in mental health research is that the condition is underrecognised and may be mistaken for major depression ^{Reference Casey, Maracy, Kelly, Lehtinen, Ayuso-Mateos and Dalgard11} and treated accordingly. As noted by others, adjustment disorder is being eclipsed by major depression over time and the authors observe ^{Reference Strain and Diefenbacher12} that this is not necessarily because of changes in its prevalence but due to a changed culture of diagnosis consequent on a change in the culture of prescribing due to the wide availability of antidepressants.

The study by Fernandez et al provides evidence for both of these propositions. The authors point to the low recognition rate by general practitioners, with only 2 of 110 cases being identified. In addition, 45% of those diagnosed with adjustment disorder by structured interview were prescribed an antidepressant. Furthermore, data on prescribing from the USA ^{Reference Olfson and Marcus13} show that antidepressants are the most commonly prescribed medications and their use in the general population has nearly doubled over a 10-year period from 5.84% in 1996 to 10.12% in 2005. This represents an increase from 13 million to 27 million persons. Antidepressant use in individuals with adjustment disorder showed the biggest increase from a rate of 22.26/100 to 39.37/100 annually. Worryingly, the use of antidepressants in treating adjustment disorder is not founded on any strong evidence and although there have been a few randomised trials, none was double-blind and most of the focus has been on herbal remedies. The use of brief psychological therapies is the recommended treatment but studies are also limited in number and quality. Furthermore, there is a possibility that no specific treatment is required since adjustment disorders are by definition self-limiting conditions, and one study comparing antidepressants, placebo, supportive psychotherapy and a benzodiazepine found that all four treatments were associated with significant improvement. ^{Reference de Leo14} Clearly, mistakenly offering services for a condition that may not require them has significant service planning and financial implications and warrants further study.

DSM-5 and ICD-11

The problems outlined above, such as the low level of research interest in adjustment disorder, its conflation with other diagnoses, inappropriate treatment and the inadequacy of the measurement of the disorder in the current diagnostic interview schedules could be resolved in ICD-11 and DSM-5 if a change to the current status of adjustment disorder is initiated. This would involve a number of alterations to the diagnostic criteria. These have been detailed elsewhere, ^{Reference Baumeister, Maercker and Casey15} and foremost among these is changing the status of adjustment disorder from a residual category to a full syndromal category. A system of symptom weightings and directing more attention to the cognitive proximity between the stressor, the symptoms and mood reactivity should be considered. Regarding adjustment disorder as a failure of adaptation is another avenue that has also been suggested. ^{Reference Maercker, Forstmeier, Pielmaier, Spangenberg, Brähler and Glaesmer16} A more difficult task will be deciding on diagnostic criteria that recognise the favourable longitudinal course that is generally the hallmark of adjustment disorder. This may require a combined dimensional and categorical approach to classification, as suggested for other categories. ^{Reference Slade and Andrews17}

A further challenge will be delineating adjustment disorder from normal stress responses and this should take into consideration the impact of symptoms on functioning, based on the nature of the stressor, the personal and interpersonal context in which it has occurred and cultural norms with regard to such responses. Ultimately, the upgrading of adjustment disorder into a full syndrome will entail the development of diagnostic criteria which will be incorporated into pre-existing structured interviews such as the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) ^{Reference Wing, Babor, Brugha, Burke, Cooper and Giel18} and SCID, allowing comparisons along the borders of adjustment disorder while utilising specialised statistical tools to examine the latent structure of the construct.

At this point there is some information on the proposed changes to adjustment disorder in DSM-5 including its inclusion in a genre of stress-related disorder, ^{Reference Strain and Freidman19} which would lead to harmonisation with ICD-11. The addition of a subtype of posttraumatic stress disorder that does not meet all the criteria for this condition is also appropriate clinically. However, there would appear to be no plans to upgrade its subclinical status and without this the criteria will, most likely, continue to be poorly delineated. A recent editorial ^{Reference McHugh and Slavney20} highlighted what the authors described as a ‘rote-driven essentially rule-of-thumb approach to the diagnosis and treatment of patients’ that the tick-box approach of DSM fosters. Nowhere is this more apparent than in the approach to the classification and relegation of adjustment disorder behind other disorders crossing a symptom threshold. This must be rectified so that the common condition is accorded appropriate recognition in the revised classifications. Thereafter a renewed interest in adjustment disorder and its management will follow and ultimately inform service planning and treatment decisions, correcting the deficiencies noted by Fernandez and colleagues. ^{Reference Fernández, Mendive, Salvador-Carulla, Rubio-Valera, Luciano and Pinto-Meza1}