The problem with borderline personality disorder

In the late 1980s, the ICD-10 Working Party on Personality Disorders had little evidence on which to base its decisions and, understandably, followed the lead of the DSM, with its well-funded and popular third and subsequent editions.

When the Working Party came to the sensitive subject of individual personality disorders, it found that the evidence for “borderline personality disorder” was insufficient for it to be included. But a lobby of supporters did not allow this, and eventually two extra personality disorder groupings were included under the heading of “emotionally unstable personality disorder” (F60.3) – an “impulsive type” (F60.30), characterized by a “tendency to act unexpectedly” and to show “quarrelsome behaviour” and an “unstable and capricious mood”; and a “borderline type” (F60.31), characterized by uncertain self-image, unstable relationships, efforts to avoid abandonment, and recurrent self-harm.

We have yet to see much evidence that the impulsive type (F60.30) has been used in practice. On the contrary, the borderline type is by far the most commonly used personality disorder diagnosis, so much so that the original splitting of the “emotionally unstable personality disorder” into two groups has been forgotten entirely.

In the ICD-11 revision group, more than two decades later, the same conclusion was reached: borderline personality disorder was not considered to be a suitable diagnosis for inclusion and was ignored, as indeed were all other categories of personality disorder in the new dimensional system¹. But, as with the ICD-10, the borderline diagnosis was not to be spurned by others. There was general dissatisfaction with its omission², and a strong appeal for it to be included in some form. Thus, the “borderline pattern specifier” was added as a compromise³.

How do we explain that, after two revision groups decided to exclude this condition as unsatisfactory, borderline personality disorder continues to be supported as a diagnosis? The standard explanations are that it is useful in clinical practice, is widely used, and gives options for treatment, unlike other personality disorders. However, the same could be said, almost exactly, of the diagnosis of neurasthenia between 1870 and 1990 (it appeared apologetically in the ICD-10), which has now been recognized to be redundant, as it was vaguely defined, was so prevalent that it lacked discrimination, and became toxic through criticism and stigma.

These same concerns apply to borderline personality disorder. It is like a large bubble wrap over all personality disorders, easily recognized on the surface but obscuring the disorders that lie beneath. Personality abnormality is identifiable through traits that are persistent, exactly as normal personality traits. The features of borderline personality disorder are not traits, but symptoms and fluctuating behaviours⁴, and – like many symptomatic conditions – improve steadily over time⁵. When borderline symptoms are examined in factor analytic studies, they are scattered over a range of both personality and other mental disturbance, and have no specificity⁶.

All attempts to find a borderline trait have failed. While borderline symptoms appear coherent when examined in isolation, they disappear into a general personality disorder factor when modelled alongside other personality disorder symptoms⁷. Borderline personality disorder symptoms strongly align with all other personality disorder symptoms, and the borderline personality disorder diagnosis is better conceptualized as moderate to severe personality pathology in general⁶. Gunderson and Lyons-Ruth may have been on to something when they identified the core of borderline pathology as interpersonal hypersensitivity, a symptom-behaviour complex present in most personality disorders⁸.

An unsatisfactory diagnosis leads to imperfect treatment. Although it appears that there are many treatments available for borderline personality disorder, their value evaporates on analysis. While the treatments are complex, often time-consuming and well-constructed, they are no more effective than good psychiatric care, which now, in our current passion for three-letter acronyms, is called SCM (structured clinical management) or GPM (general psychiatric management). There is confusion over who should receive SCM and GPM and who needs the more complex interventions of dialectic behavioural therapy (DBT), mentalization-based therapy (MBT), transference-focused psychotherapy (TFT), cognitive behavioural therapy (CBT) and cognitive analytic therapy (CAT). Wheeling out stepped care as an answer sounds good but, because the diagnosis is so defective, nobody knows where stepped care is to begin.

An argument might be made that, while criticisms of the borderline personality disorder diagnosis are valid, the term is familiar to clinicians and could be seen as a synonym for moderate to severe personality pathology and lead to appropriate treatment with structured psychotherapy. The problem with this argument is that the term is a major source of stigma. Patients identified as having borderline personality disorder are seen as more difficult to manage even when their behaviour is the same as other patients without the label⁹. Access to treatment for other psychiatric disorders – such as attention-deficit/hyperactivity disorder, substance use disorder or mood disorders – as well as for physical disorders may also become more difficult. The label borderline personality disorder devalues all other symptoms, so that they can be more easily disregarded. This, in turn, increases the sense of alienation that many patients with personality problems already feel.

We argue that the solution is to drop the borderline personality disorder diagnosis and replace it with a more transparent system of describing personality pathology. Since borderline personality disorder diagnoses are highly correlated with overall moderate to severe personality disorder, assessing the level of severity of patient dysfunction is the first step. Many patients with moderate or severe personality disorder will have features now called “borderline”, such as emotional dysregulation, interpersonal hypersensitivity and impulsive behaviours, but not everyone. Some will have prominent social and emotional detachment, others perfectionism and stubbornness, or self-centeredness and a lack of empathy. These patients, with personality features described over many centuries, are largely ignored by treating personality disorders with a focus on so-called borderline features.

The new ICD-11 personality disorder classification allows this broader assessment. The dimensional classification of severity – which is divided into personality difficulty and mild, moderate and severe personality disorder – means that clinicians are encouraged to assess overall severity before focusing on specific symptoms and behaviours. The five domains (negative affectivity, detachment, dissociality, disinhibition and anankastia), similar to the Big Five in normal personality, allow a more nuanced description of these symptoms and behaviours, going beyond those encompassed within borderline personality disorder, particularly in the detachment and anankastia domains.

This should lead clinicians to consider the whole spectrum of personality pathology in their patients, rather than losing interest when the borderline personality disorder criteria have been ticked off. A sophisticated formulation would hopefully lead to a range of interventions rather than standard protocol-driven treatment given to everyone. It might also encourage research around treatment for those with non-borderline personality disorder symptoms and traits.

In conclusion, borderline personality disorder may best be seen as a transitional diagnosis which drew attention to patients suffering from moderate to severe personality disorders and encouraged structured psychotherapies to be tested. However, it has now emerged that the diagnosis is not related to specific personality traits, is overinclusive, and does not lead to specific treatments beyond structured clinical care. Its domineering presence in the field means that assessment and treatment of other personality pathology is discouraged, and the whole concept of personality dysfunction is stigmatized. It is time for borderline personality disorder to lie down and die.