Classification of anxiety disorders: Changes,conundrums and consequences

DSM Digest

Anxiety disorders occupy a unique position between normality and severe mental illness. This poses challenges for psychiatric taxonomy.

Anxiety disorders as the shrinking domain of psychopathology

Although pathological anxiety is a ubiquitous phenomenon, the psychopathological ‘territory’ encompassed by the anxiety disorders in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) has shrunk. This is due to the removal of obsessive-compulsive disorder (OCD), acute stress disorder and posttraumatic stress disorder (PTSD) from the group of anxiety disorders – largely because pathological anxiety is not necessarily the key feature of these conditions.

The DSM-5 has made some welcome changes to the anxiety disorders. For example, a problematic and often artificial distinction between unexpected and situational (expected) panic attacks is de-emphasised and clinicians do not have to assign primacy to panic disorder when it co-occurs with agoraphobia. The diagnostic criteria for generalised anxiety disorder (GAD) remained the same as in DSM-IV, which was deemed positive relative to the unjustified proposals to broaden the concept of GAD (Starcevic and Portman, 2013).

As part of the ‘harmonisation’ process, it has been proposed that OCD and PTSD should also be classified separately from the anxiety disorders in the International Classification of Diseases, 11th Revision (ICD-11). Interestingly, the ICD-11 draft document (‘Beta Draft’) refers to ‘anxiety and fear-related disorders’, thereby suggesting that anxiety and fear may be different, albeit related emotions. This is important because it revives an unresolved question (e.g. Krueger, 1999; Vollebergh et al., 2001) of whether ‘fear disorders’ (e.g. panic disorder and phobias) differ from ‘anxious-misery’ or ‘distress/dysphoric disorders’ (e.g. GAD, dysthymia and major depression). If they do, should they be grouped together?

Boundary issues

For decades, concerns about boundaries have plagued the conceptualisation of anxiety and related disorders. Fortunately, the diagnostic thresholds have generally not been decreased in the realm of the DSM-5 anxiety disorders, with the boundaries between normal fear and anxiety disorders thereby being maintained. Admittedly, these boundaries are not always clear and more work is needed to strengthen them.

The boundary between anxiety disorders and depression has not been addressed adequately in the DSM-5 (Malhi and Henderson, 2013). Also, it is a paradox that a common clinical presentation of pathological anxiety and depression does not have a diagnostic designation in DSM-5, not even among the conditions for further study. In the ICD-11 draft document, such a diagnosis is envisaged, although it appears to be a residual category for presentations that are diagnostically subthreshold for specific depressive and anxiety disorders. The extent of arbitrariness and uncertainty is reflected both in the naming and classification of this disorder: whereas in ICD-10, it was referred to as ‘mixed anxiety and depressive disorder’ and classified among other anxiety disorders, the ICD-11 draft proposes ‘mixed depressive and anxiety disorder’ and classifies it as a depressive/mood disorder.

One of the key problems in psychiatric classification is the lack of clear criteria that would stipulate the degree of relatedness between disorders required for them to be classified together and degree of distinctness necessitating their separate classification. This is particularly pertinent to the anxiety disorders: for almost every argument to move OCD and PTSD out of the group of anxiety disorders, there is a counterargument to classify them there. However, we lack tools to assess the relative weight of these arguments, which would tell us whether some are more important than others. This leaves the door wide open for arbitrary decisions about group memberships; unsurprisingly, such decisions are often imposed by groups with strong but not evidence-based views or individuals with personal agendas.

Implications for clinical practice

Clinical practice often follows its own rules that are largely based on logic and practicality. From this perspective, it will not matter so much which conditions are classified as anxiety disorders and how they are referred to. What matters is whether these disorders are recognised and treated adequately. Our diagnostic systems, as imperfect as they are, still provide a solid basis for various anxiety syndromes to be identified. What should be kept in mind, however, is that DSM- or ICD-based diagnostic assessments are like snapshots, capturing only the cross-sectional psychopathology. Considering the fluid boundaries between the anxiety disorders and their changing pattern of presentation over time, the only valid approach is a longitudinal diagnosis (Starcevic, 2008). Within this framework, a cross-sectional anxiety diagnosis should have no more than a provisional status.

With regards to the treatment of anxiety disorders, it is diagnosis-driven only to some extent. Most evidence-based treatments for these conditions, whether psychological (e.g. techniques based on cognitive-behavioural therapy such as exposure) or pharmacological (e.g. selective serotonin reuptake inhibitors), are transdiagnostic. When selecting treatments, the specific situational factors and social context, as well as personality-related issues, co-occurring disorders and response to any previous treatment, are often more important than the actual anxiety diagnosis. This is another reminder for a need to fully espouse personalised medicine in psychiatry.