Is the loss of Axis II a loss?

The move away from a multi-axial nosological model for both major classifications in psychiatry appears, on first impression, to have merit. There is little fundamental evidence to suggest that personality disorders are separate from other mental disorders and there is no doubt that they overlap. ^{Reference Clark1} Like other mental disorders, personality disorders have a heritable component that accounts for longitudinal stability and changes in mental state related to environmental stress. ^{Reference Zanarini, Frankenburg, Hennen, Reich and Silk2} Gene-environment interactions shape personality disorder over time in a similar manner to other mental state disorders. ^{Reference Caspi and Moffitt3} Furthermore, there is some evidence that all psychopathology shares underlying similar factors, regardless of diagnostic label. ^{Reference Caspi, Houts, Belsky, Goldman-Mellor, Harrington and Israel4} Why then should personality and the possibility of personality disorder be considered separately?

There is no doubt that the placement of personality disorder on a separate axis stimulated consideration of personality in clinical assessment and in research. The past three decades have seen a dramatic rise in the literature focusing on personality disorder, particularly borderline and antisocial personality disorder. ^{Reference Blashfield and Intoccia5} There has also been a bourgeoning of clinical trials of interventions in patients with personality disorder, again mostly focused on borderline personality disorder. ^{Reference Stoffers, Völlm, Rücker, Timmer, Huband and Lieb6} Related to this is the increasing recognition that personality disorder is not a diagnosis of exclusion from clinical services. These research and clinical outcomes are arguably related to the development of a nosological system with a separate specifier for personality disorder.

The differences between personality and mental state disorder

It can also be argued that if personality disorder is categorised as a mental state disorder, this may obscure differences that are clinically important. Most obviously, personality facets, traits or characteristics, however described, are present more or less persistently throughout the life course and personality disorder is best conceptualised as extremes of these facets leading to functional and societal impairment. This differs from most mental state disorders, in which persistence may be common but which have clearer onset and end. The underlying structure of personality disorder can usually be identified in childhood and adolescence (despite reluctance to make diagnosis early) and the essential features may be present as early as 3 years old with some prognostic capacity. ^{Reference Caspi and Silva7} Personality disorder is best conceived using a diathesis model, as a condition that is always present but may often be clinically quiescent for long periods without ever losing its essential elements. ^{Reference Zanarini, Frankenburg, Hennen, Reich and Silk2} Furthermore, the nature of stability and change in personality through the life course is increasingly being understood. This allows for recognition of extremes of personality disturbance at differing stages in life. Personality disorder obviously interacts with mental state disorder and may predispose to it or impair clinical course ^{Reference Tyrer8} but is not part of the same construct. This understanding is supported by increasing evidence that the course of mental disorders may differ in the presence or absence of a personality disorder. Since we have failed to find neurobiological differences that explain these differing courses of illness it may be better to focus on the personality component rather than the mental illness one.

Clinical implications of the loss of the multi-axial system

In busy clinical settings cross-sectional assessment of symptoms followed by instigation of treatment is the primary focus. Personality disorder, diagnosed as a cross-sectional polythetic construct (as it is described in DSM-5), ⁹ is unlikely to be prioritised over mental state disorder. However, when considered alongside mental state disorders, personality pathology may play an important role in diagnosis and management, and the additional time needed to undertake a personality assessment may pay off in the long run. Patients may have worse outcomes if it is ignored. The diagnosis of personality disorders remains a stigmatised one even among mental health professionals ^{Reference Newton-Howes, Weaver and Tyrer10} and it is likely to become more hidden if it is diagnosed on the same axis as other mental disorders.

A possible reason for the confusion between personality pathology and mental state disorders may be related to the unempirical categorical personality disorder diagnoses in both DSM-5 and ICD-10. ¹¹ By setting clear, if unvalidated, operational criteria for individual categories of personality disorder in DSM, the clinician has been tempted to choose one or the other rather than embracing both. The removal of these categories in the proposed ICD-11 classification ^{Reference Tyrer, Crawford, Mulder, Blashfield, Farnam and Fossati12} in favour of a single axis of severity makes the classification more attuned to a separate axis. These categories lack scientific accuracy and clinically imply that personality disorders are just types of mental disorders with sets of symptoms and specific behaviours that are identified using strict operational definitions. If these personality disorder diagnoses were indeed valid descriptions of personality pathology, then removal of Axis II would seem a sensible way forward.

However, evidence does not support this. Most of the categorical personality disorders are rarely diagnosed and poorly researched, with borderline personality disorder and antisocial personality disorder overshadowing all other personality disorder diagnoses. This in turn reinforces an oversimplified view of personality disorder creating a false impression that antisocial personality disorder is a diagnosis in forensic settings and borderline personality disorder a synonym for recurrent self-harm. Such diagnoses ignore the underlying traits that are the constant features of personality, the influence of these traits in mental state disorders and the usefulness of identifying personality pathology in children, adolescents and older age patients.

What’s in a name?

In considering the nosology of mental and behavioural disorders the terms Axis I and Axis II increasingly have become terms to describe categories such as depression, schizophrenia, anxiety disorders, borderline personality disorder and antisocial personality disorder as though they were similar entities. The term ‘serious mental illness’ encapsulates all diagnoses, including personality disorders and does not adequately discriminate Axis I and Axis II disorders. How then can we capture the notion of personality disorder as separate from other mental disorder in a scientific and clinically meaningful way?

We are not advocating a return to DSM-IV ¹³ with Axis I mental state categorical disorders and Axis II personality categorical disorders described as though they were similar entities that could be comorbid with each other. Instead, we argue that the terms mental state disorder and personality disorder are useful clinical constructs that allow a different perspective to be taken for these two psychopathologies. Such an approach allows for the recognition of development in the science of personality pathology, best seen as a unitary dimensional construct with multiple phenotypes. It also provides a useful nosology for clinicians and researchers.

Although the decision of the DSM and ICD classification not to use multi-axial systems has some merit, if the negative implications for personality disorder are not identified and addressed there may be negative clinical consequences for patients. There is a significant danger that clinicians will rarely address the impact of personality pathology in psychiatric patients. Since all psychiatric disorders are now diagnosed on a single axis it is unlikely that an Axis II will be re-introduced and we do not necessarily advocate for that. We support the concept that personality pathology will be addressed as a secondary diagnosis where possible.

We would argue that clinicians should consider mental state disorder and personality disorders in all patients. This may still be possible if the current Axis I like categories of personality disorder are replaced by a dimensional descriptive model that could be used with all patients. The new ICD-11 classification system proposes that all categories of personality disorders are replaced by a single dimension of personality dysfunction at varying levels of severity. ^{Reference Tyrer, Crawford, Mulder, Blashfield, Farnam and Fossati12} (A similar proposal for DSM-5 was rejected and relegated to Section III – emerging measures and models.) Such a model recognises that change in personality pathology occurs and allows personality status to be repeatedly assessed as a secondary diagnosis.

The importance of the impact of personality disorder on the long-term outcomes of mental illness is increasingly recognised. We are concerned that personality disorder may become a less important component of the clinical diagnosis in psychiatry resulting in a loss of important clinical information. Personality pathology incorporates a longitudinal developmental view of psychopathology and should be considered alongside mental state symptoms, intellectual functioning, neuropsychological functions and social deprivation in all patients presenting with mental disorders.