Looking for balance between promoting adult separation anxiety disorder and overlooking it

A response by Silove and Manicavasagar (2013) to my commentary about separation anxiety disorder (SAD) in adults (Starcevic, 2013) is a welcome contribution to a debate on several important issues. This commentary aims to correct an apparent misreading of my earlier commentary and shed more light on the underlying issues.

First, there is no doubt that separation anxiety is an important component of human development and that it can play a role in the development of psychopathology. According to Silove and Manicavasagar (2013), my view is that separation anxiety ‘should not be regarded as a pathological response’. Nowhere in my commentary did I write or imply this. In fact, I specifically referred to the manifestations of pathological separation anxiety, but argued that separation anxiety had fuzzy boundaries with its pathological forms. With regard to the latter point, Silove and Manicavasagar (2013) seem to be in full agreement with me, as the lack of adequate criteria for distinguishing between normality and psychopathology is indeed one of the fundamental problems in all of psychiatry.

The need for adult SAD and its recognition

The presence of pathological forms of any phenomenon, including separation anxiety, does not imply that there is a disorder characterised by this phenomenon. While there are many adults with pathological separation anxiety, I am not convinced that they should all receive a diagnosis of SAD. Perhaps some of them warrant such a diagnosis, but many others do not. This should be ascertained on a case-by-case basis. Thus, I am not calling for adult SAD to be ‘abandoned’, I am only cautious about this diagnosis.

There are conditions ‘unrecognised’ by the DSM (Diagnostic and Statistical Manual of Mental Disorders) system that are nevertheless diagnosed by clinicians, perhaps in defiance of the DSM, but also because such conditions are seen in clinical practice and there is apparently a clinical need for the corresponding diagnoses. In Australia, examples of these diagnoses are mixed anxiety and depression and irritable bowel syndrome. With respect to adult SAD, it is possible that there has been a relatively low clinical need for this diagnosis, not necessarily a poor recognition, misdiagnosis and neglect.

Silove and Manicavasagar (2013) suggest that there are many adults with SAD who feel misunderstood and whose suffering is not correctly attributed to SAD. Why would that be? Is it that clinicians are ‘blind’ to their suffering or that patients do not voice their SAD-specific problems enough for SAD to be considered their core psychopathology? If the former is understood to be the problem, it is difficult to see why clinicians would be selectively insensitive to SAD-related suffering and selectively unable to recognise SAD; if the latter is the problem, revised criteria for the already existing diagnosis of SAD would not be likely to be of much help. Therefore, if SAD is truly the main psychopathology, I see no huge obstacles for it to be recognised as such, whereby the arbitrary DSM-IV criterion about the onset of SAD before the age of 18 years could safely be ignored. But in order for SAD to be recognised, patients need to be explicit about their symptoms. If patients are not clear about what they experience and what troubles them, they are likely to be misdiagnosed. This applies to adult SAD as much as it does to other disorders.

Upholding standards for diagnosing adult SAD

The standards for establishing diagnoses of mental disorders should also apply to adult SAD. For research purposes, this means using a validated interview-based diagnostic instrument. However, Silove and Manicavasagar (2013) state that their self-report Adult Separation Anxiety Questionnaire (ASA-27; Manicavasagar et al., 2003) is adequate for diagnosing adult SAD because it has been ‘calibrated against a semi-structured clinical interview’ and demonstrated to ‘converge with a measure developed independently by a separate research group’. If we accept this, then major depressive disorder (MDD) could be diagnosed by means of the scores on a self-report scale such as the Beck Depression Inventory (BDI-II; Beck et al., 1996) because higher BDI-II scores have been associated with the diagnosis of MDD (Beck et al., 1996) and BDI-II scores above a cut-off point have been found to predict a diagnosis of MDD (Arnau et al., 2001). No clinician, however, would make a diagnosis of MDD in this way. Why should a diagnosis of adult SAD be an exception?

While various instruments may indeed be used in the assessment of one disorder, the differences between them often account for the discrepant results. This might have been one of the reasons for the large differences between reported prevalence rates of adult SAD, which then begs the question of which instrument is most psychometrically sound.

Overlap, co-occurrence and predisposing role of SAD

Co-occurrence of the putatively distinct mental disorders is a serious problem. In my view, it is largely a consequence of the DSM-promoted process of splitting larger diagnostic entities. While adult SAD may have some relatively specific manifestations, it shares a number of symptoms with other anxiety disorders. As long as there is a symptomatic overlap (i.e. symptoms common to at least two disorders) between adult SAD and other anxiety disorders, especially panic disorder/agoraphobia, some of the high co-occurrence rates for adult SAD will be attributed to this overlap. Stating that ‘overlap between adult SAD and agoraphobia reflects true comorbidity, that is, the overlap does not arise primarily because of the sharing of common symptoms’ (Silove and Manicavasagar, 2013) implies a degree of certainty about the relationship between adult SAD and agoraphobia that does not exist. In addition, overlap is incompatible with the notion of ‘true comorbidity’, as it does imply ‘sharing of common symptoms’ and therefore spuriously inflates co-occurrence rates.

With regards to SAD ‘driving’ later psychopathology, the critical issue here is a lack of specificity. There is no evidence to suggest that SAD plays a role as a specific precursor to any psychopathology. It seems that SAD reflects a broad predisposition to the development of anxiety and other disorders, perhaps analogous to the personality trait of neuroticism.

Does treatment development always depend on having an agreed-upon diagnosis?

Silove and Manicavasagar (2013) argue that a prerequisite for ‘developing an effective treatment for a disorder is to establish its nosological status’. Even if one accepts this line of reasoning, it does not explain the lack of effective treatments for adult SAD. First, adult SAD is not a new diagnostic entity and its relatively restrictive diagnostic criteria should not have prevented those who see many SAD patients from developing treatments for it. Second, treatment approaches to children and adolescents with SAD could have been adapted to adults with SAD. Third, with transdiagnostic psychological therapies for anxiety disorders becoming increasingly popular (e.g. Farchione et al., 2012), perhaps a separate treatment package for adult SAD would not be necessary. Finally, treatments can be developed in the absence of a consensus about the diagnostic and nosological status of psychopathological entities. Good examples of this are treatments for the controversial disorders such as Internet addiction (Winkler et al., 2013) and chronic fatigue syndrome (Malouff et al., 2008).

Conclusion

The concept of adult SAD will be useful to the extent that it accurately reflects the psychopathology that patients present with and makes it easier for them to obtain help. Only time will tell and future research will hopefully show whether the revised diagnostic criteria for adult SAD in DSM-5 have succeeded in this. In the meantime, some scepticism and restraint about adult SAD would be a healthy antidote to the zealous and possibly distorting embracement of this diagnostic concept.

See Commentary by Silove and Manicavasagar, 2013, 47(8): 780–782.