The Outcome of Borderline Personality Disorder: Good for Most But Not All Patients

One of the reasons why clinicians are reluctant to diagnose borderline personality disorder is the perception that patients with this disorder are doomed to chronicity. As discussed by Zanarini et al. (1) in this issue of the Journal, it is now well established that while outcome is heterogeneous, most patients do well, with the majority no longer meeting diagnostic criteria over time. Zanarini et al. buttress this conclusion with a unique 16-year prospective follow-up study of a large cohort of patients with borderline personality disorder. However, they emphasize that remission (defined as not meeting criteria for a formal diagnosis) is not equivalent to recovery. Personality disorders are amalgams of traits and symptoms, and even when symptoms remit, problematic traits can produce difficulty. That is probably why remission is more common than full recovery.

Should we be concerned about this finding? On the one hand, “former” borderline patients may continue to have problems that require treatment. This group would then be diagnosed with personality disorder not otherwise specified, which DSM-5 (www.dsm5.org) refers to as “personality disorder, trait specified.” Yet few patients with borderline personality disorder require lifelong treatment. The research of Zanarini et al., confirmed by 10-year prospective data from the Collaborative Longitudinal Personality Disorders Study (2), shows that most of these patients eventually get a life, find a place in the world, and, most importantly, stop wanting to kill themselves. In fact, the prognosis of borderline personality disorder is much better than that of bipolar disorder, a condition with which it is so often confused (3).

The chronic suicidality that characterizes borderline personality disorder is one reason why patients can be stigmatized by psychiatrists, who are not always comfortable with the view that treating these patients means accepting a calculated risk (4). Moreover, psychiatrists spend more time with the sickest group of borderline patients, who are more likely to be chronically suicidal. This gives them the impression that most cases are like that.

Also in this issue, Soloff and Chiappetta (5) describe outcome in a high-risk group of seriously suicidal patients. This sample is different from the one described by Zanarini et al., which probably explains why half the patients in this group had a poor psychosocial outcome. However, the findings of Soloff and Chiappetta do not contradict the robust finding that most borderline patients usually recover with time. Rather, they point to a subgroup with poor outcome that is particularly likely to show continued suicidal behaviors.

One of the unanswered questions about chronically suicidal patients is whether we can predict who is most at risk for mortality. This is a problematic issue. In spite of the fact that most borderline patients who threaten suicide never kill themselves, clinicians may opt for “safety” (i.e., hospitalization). My own research group (6) reported a 10% rate of death by suicide after 27 years of follow-up, but the mean age at death was as late as 38 years. This contrasts with the early age at which suicidal behaviors are most prominent in borderline personality disorder and suggests that clinicians could afford to be less concerned about younger patients who come to the emergency department with suicidal ideas or actions and more concerned about older patients who do not recover, even after extensive treatment. Psychiatrists should also be aware that self-harm by cutting is not usually suicidal behavior but an attempt to regulate dysphoric emotions (7).

Borderline personality disorder is not, as once thought, a life sentence (8). Borderline patients do not deserve the bad reputation that some psychiatrists associate with the diagnosis. Clinicians get the wrong impression about chronicity because so many patients they see present in an emergency setting, since the sickest patients remain in the mental health system, and since those who remit may not return to clinical attention.

Cautious optimism about borderline personality disorder is also supported by evidence that borderline patients can be treated effectively, particularly with specific forms of psychotherapy (9, 10). The main problem these days is that borderline personality disorder is not always diagnosed accurately or separated from primary mood disorders (3). To manage these patients effectively, one first has to recognize the disorder.