Clinical Aspects of Borderline Personality Disorder

Author: Deborah Daniels Carver, MD, Payne Whitney Clinic

Abstract: The majority of patients with borderline personality disorder (BPD) can improve in time if given appropriate treatment. In a study known as PI-500--one of the largest studies with the best follow-up--two thirds of the patients were "clinically well" at follow-up. Results of this study suggest that factors associated with the best outcomes for patients with BPD include high intelligence (IQ >130); unusual artistic talent; physical attractiveness (women only); and obsessive-compulsive features that enhance self-discipline, work-orientation, and the ability to structure leisure time. Regarding treatment, psychobiologic principles suggest the impulsive behavior should be ameliorated by enhancing serotonergic activity with selective serotonin reuptake inhibitors. Because there may be a relationship between affective instability and hyperresponsiveness of the noradrenergic and cholinergic systems, agents that stabilize catecholaminergic function, such as monoamine oxidase inhibitors, lithium, or carbamazepine, may be helpful in treating labile BPD patients. Antipsychotics, mood stabilizers, anxiolytics, and antidepressants can also play a role in the pharmacologic management of BPD. Depression is a frequent chief complaint in patients with BPD. It can be a positive sign of growth during treatment, or it can foreshadow a suicide attempt. Understanding of the etiology of a patient's depressive symptoms is crucial to providing the appropriate treatment. Although there appears to be an association between BPD and bipolar disorder, the exact nature of this relationship is unclear. It has been observed, however, that while not every adolescent with BPD develops bipolar disorder, most adolescents with bipolar disorder also have features of BPD. [Medscape Mental Health 2(9), 1997. © 1997 Medscape, Inc.]

Introduction

Impulsive and frequently self-destructive acts of patients with borderline personality disorder (BPD) can pose a tremendous challenge for clinicians treating them. Advances in the understanding of BPD and its management, however, are improving the prognosis and outcome. During a symposium entitled "Clinical Aspects of Borderline Personality Disorder," presented at the 150th Annual Meeting of the American Psychiatric Association in San Diego, California, 4 leading clinicians highlighted major developments in BPD: Michael H. Stone, MD, of the New York State Psychiatric Institute, discussed new insights into the course of the illness and outcomes identified in long-term, follow-up studies. Larry J. Siever, MD, of the Bronx VA Medical Center, reviewed the current understanding of the biologic correlates of BPD and their implications for pharmacotherapy. Harold W. Koenigsberg, MD, of New York Hospital-Cornell Medical Center, presented a systematic approach to the psychotherapy of patients with BPD and depression. Finally, Clarice J. Kestenbaum, MD, of the New York State Psychiatric Institute, looked at the relationship between BPD in adolescence and bipolar disorder in adulthood.

Course of Borderline Personality Disorders

The first follow-up studies on patients with BPD in the 1970s were of brief duration (3-5 years), and generally employed broad and unsystematic diagnostic criteria. The results obtained led to the belief that the life course for patients with BPD was unfavorable and scarcely distinguishable from that of schizophrenia.

In the 1980s, however, several studies with larger numbers and longer intervals refuted these earlier findings. Research teams headed by McGlashan[1], Plakun and colleagues[2], Stone and others[3], Stone[4], Paris and coworkers[5], and Kroll and Ogata[6] systematically diagnosed patients with BPD using either the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III) or Gunderson criteria, and then followed these patients for 10 to 25 years from the time of initial presentation. During the symposium, Stone presented highlights and insights from the study known as PI-500.[3,4,7] This was the largest study of its kind in the 1980s, and it had the highest follow-up rate (95%). The results, like those from other recent studies,[1,2,5,6] have led researchers to uniformly conclude that the majority of patients with BPD show improvement at follow-up, although more in the sphere of their professional and career activities than in their social relationships and interactions.[7]

In the PI-500 study, two thirds of the patients were "clinically well" at follow-up. As many as 20% were asymptomatic, working, and able to sustain a gratifying long-term sexual relationship. The marriage rate in the PI-500 population, however, was only half the national average reported for people from a similar culture and of a similar age. Moreover, the percentage of women with BPD who went on to have children was only 25% of the general-population average, and the number of men with BPD who had children was negligible. At follow-up, the majority of these patients still met the criteria for one of the "milder" personality disorders (histrionic, avoidant, or obsessive-compulsive).

Stone has compiled a list of factors that can be used to predict how BPD patients will fare. Patients with BPD whose outcomes were distinctly better than average tended to show: (1) high intelligence (IQ higher than 130); (2) unusual artistic talent; (3) physical attractiveness (women only); and (4) obsessive-compulsive features that enhance self-discipline, work-orientation, and the ability to structure leisure time. BPD patients who were also substance abusers had better outcomes if they maintained participation in groups such as Alcoholics Anonymous.

BPD patients whose outcomes are "distinctly worse" than average also share common features. These patients are more likely to have a history of (1) transgenerational incest and (2) parental brutality. Moreover, they tend to have (3) schizotypal or antisocial features and (4) untreated substance-abuse problems. BPD patients who have "poor" outcomes can be divided into 2 broad categories: patients who have chronic impairment and patients who complete suicide.

The suicide rate in the PI-500 study was 8.5% in an average 16-year follow-up period. This rate is equivalent to that found by Paris and coworkers[5] but is more than twice that of the McGlashan study.[1] Stone attributes the lower suicide rate in the McGlashan patients (26 suicides) to their older average age, because the peak age for suicide in the US ranges from ages 20 to 29 years. Of the 19 patients in the PI-500 study who completed suicide, 13 had a history of affective illness, 7 had a history of alcohol abuse, 5 had a history of incest, and 8 had at least 1 first-degree relative with a diagnosed psychiatric illness (primarily substance abuse or affective illness).

The PI-500 study by Stone and others,[3] and other studies by his colleagues,[1,2,5,6] show that although patients with BPD may not "routinely recover," the majority do improve in time with appropriate treatment. This majority, however, coexists alongside a minority of patients who either commit suicide or live with significant chronic impairments. While there is no accurate way to distinguish between the 2 groups at initial presentation, studies such as PI-500 highlight features that may serve as useful predictors of outcome.

Pharmacotherapy of the Borderline Patient

Besides being able to recognize the factors most likely to influence the course and outcome of BPD, clinicians need to apply emerging principles of psychobiology to appropriately select the most effective treatment approach. Siever and Davis[8] and Siever's[9] multidimensional approach to the psychobiology of personality disorders is one that can be used as a framework to guide pharmacotherapy for BPD. Siever proposed that most clinical and personality disorders--that is, Axis I and Axis II--can be classified on 4 dimensions: (1) cognitive/perceptual organization, (2) impulsivity/aggression, (3) affective instability, and (4) anxiety/inhibition.

Using Siever's model,[8] people with a disturbance in cognition or perception would be vulnerable to psychosis. Those with problems of impulsivity would be quick to respond to internal or external stimuli. Those with affective instability would be predisposed to marked and rapidly reversible shifts in mood. Finally, those with high anxiety would be excessively sensitive to punishment or rejection. In Siever's schema, patients who meet DSM-IV criteria for BPD would thus be seen as having difficulties with impulsivity/aggression and affective instability.

Siever[9] proposed that these dimensions are regulated by specific neurotransmitter systems. To date, several preclinical studies have linked behavioral inhibition to the serotonin system. In one such study by Valzelli,[10] rats with lesions created to disrupt serotonin production were shown to have an increase in unmodulated aggression, including unrestrained killing of mice. More recently, Coccaro and coworkers[11] looked at serotonin activity in patients with personality disorders. Coccaro's group measured the prolactin response to fenfluramine--a promising agent for evaluating overall central 5-HT (serotonin [5-hydroxytryptamine]) activity. Fenfluramine stimulates the release of endogenous central stores of presynaptic 5-HT and activates postsynaptic 5-HT receptors. The investigators also found that central 5-HT function was lower in patients with BPD than in controls, and they determined that reduced 5-HT function and a history of suicidal and/or aggressive behaviors were significantly related.[11]

Although further research is needed, these preliminary results linking serotonin to impulsivity may have significant implications for treatment, now and in the future. Before initiating any pharmacotherapy based on research findings, however, it is important to take a careful family history and personal mental health history, including previous response to medications. Then, all comorbid Axis I and Axis II disorders should be evaluated and fully treated.

Even with optimal management of all existing comorbid conditions, patients with BPD frequently manifest a persistent tendency toward impulsivity and/or affective instability that requires further management. Theoretically, impulsive behavior should be ameliorated by enhancing serotonin. Therefore, medications that enhance postsynaptic serotonergic activity, like selective serotonin reuptake inhibitors (SSRIs) or lithium, would be reasonable choices. As patients with BPD appear particularly sensitive to side effects of medications, it is important to begin with low doses and to increase them very gradually. Siever[9] further cautioned not "to give up too soon," as these patients seem to have less serotonin at baseline and thus may be slow responders.

Although not discussed in depth, Siever[9] also alluded to a possible relationship between affective instability and hyperresponsiveness of the noradrenergic and cholinergic systems. Agents that stabilize catecholaminergic function, such as monoamine oxidase (MAO) inhibitors, lithium, or carbamazepine, may thus be particularly helpful in treating labile BPD patients. Tricyclic antidepressants, which enhance noradrenaline, are not recommended because they can worsen affective lability and are lethal in overdose.

To the extent that affective instability is part of a bipolar spectrum, patients should not be started on antidepressant treatment without the prophylaxis of a mood stabilizer. However, Siever[9] emphasized that not all impulsivity is a part of bipolar illness. Therefore, in BPD patients without a clear history of bipolar mood shifts, it is acceptable practice to begin antidepressant treatment, as long as close follow-up is maintained.

Antipsychotics, mood stabilizers, anxiolytics, and antidepressants can all play a role in the pharmacologic management of BPD. At times of crisis, low-dose neuroleptics, in particular, may prove very useful. Once a patient is stabilized and the crisis is past, however, Siever[9] recommended tapering medications wherever possible. This way, there will be something left to use when the next crisis comes--a pretty sure thing for a borderline patient.

Advances in the Psychotherapy of Patients with BPD

Shifting from pharmacotherapy to psychotherapy, Koenigsberg[12] stressed the need to understand and address the depression that is so common in patients with BPD. According to Koenigsberg, depression is a frequent chief complaint in patients with BPD. Depression in these patients can mean many things: At one extreme, depression can be a positive sign of growth during treatment, and at the other, it can foreshadow a suicide attempt. Therefore, a thorough understanding of the etiology of a patient's depressive symptoms is crucial to providing the appropriate treatment.

Many patients with BPD will have comorbid Axis I major depression. Depending on the study, estimated comorbidity rates range from 14% to 83%.[12] According to Koenigsberg,[12] 50% of BPD patients will develop a major depression at some point during treatment. It is important to understand the context in which the depression occurs and to determine the duration and severity of the symptoms. This information is needed to validate the diagnosis, as some patients, especially those with histrionic character traits, will exaggerate symptoms in such a way that they appear to meet criteria for a DSM-IV major depressive episode, although in reality they do not.

The primitive defense mechanisms and fragmented internalized self-representations of BPD patients can lead to feelings of depression. Intense needs to be in control may lead to feelings of helplessness and hopelessness via the mechanism of projective identification. Through projective identification, depression can become a powerful measure of interpersonal communication with the therapist.[12]

Within the constellation of borderline personality organization, there are many different personality styles that contribute to patients' triggers for depressive feelings. Histrionic patients, for example, are very sensitive to rejection and thus particularly susceptible to feelings of depression when they believe they are being ignored or shunned. Moreover, patients with BPD tend to live turbulent lives and may have very real environmental problems--such as problems with work, love, or money--contributing to their depression. In addition to these problems, substance abuse is a frequent problem in this population, either masking or precipitating depressive symptoms.[12]

Depression also may be a favorable result of the therapeutic process. As patients with BPD improve, they begin to integrate their fragmented internalized self-images. This engenders guilt or remorse about how they have treated others, particularly their therapist, which in turn leads to feelings of depression. As the therapeutic work progresses, they may also anticipate the termination of treatment and the loss of the therapist and become depressed as a consequence.

Given the multiple etiologies for depressive symptoms, Koenigsberg[12] advocated a systematic approach to diagnosis and treatment: First, determine from which realm or realms the depression originates. Consider the full spectrum of possible etiologies including (1) biologic, (2) phenomenologic, (3) characterologic, (4) environmental, and (5) transferential.

If a depression is biological, the decision whether to use medication is a complex and important one. There will be situations in which medications are warranted. In these instances, the complications created by the symbolic meaning of medication for both patient and physician must be addressed. Ultimately, Koenigsberg[12] concluded, medication issues, along with the rest of the work with the depressed borderline patient, lies within the transference.

Is Today's Borderline Adolescent Tomorrow's Bipolar Adult?

Concluding the symposium with an examination of the relationship of BPD to bipolar disorder, Kestenbaum[13] noted that although not every adolescent with BPD will go on to develop bipolar disorder, most adolescents with bipolar disorder also have features of BPD. Many children and adolescents diagnosed with bipolar disorder later in life are discovered to have had disturbed interpersonal relationships consistent with borderline personality. The theory that BPD may be a prodrome for bipolar disorder is a relatively new concept, Kesterbaum noted, "because one needs to think about it to look for it." Citing Siever's research establishing strong links between molecular biology and behavior, Kestenbaum[13] concluded that there is genetic loading in patients with BPD who go on to manifest bipolar illness. These are the children and adolescents who appear irritable, angry, impulsive, or restless, and frequently have relatives with the same characteristics. Thus, it is critical to take "a real family history." It is not enough, Kestenbaum stressed, to ask if anyone was "manic" or "depressed" because people don't tend to think in these terms. Rather, ask if "anyone has a short fuse" or "gets really sad"; look for traits and not diagnoses.

In making the diagnosis of either BPD or bipolar disorder, it is important to remember that psychiatric disorders manifest differently at different ages. For example, Kestenbaum listed 6 criteria that differ substantially in the borderline child and in the borderline adolescent. The borderline child may demonstrate (1) disturbed interpersonal relations, (2) distorted reality, (3) panic/anxiety, (4) impulsivity, (5) neurotic symptoms, and (6) erratic development. Those children with BPD who go on to develop bipolar disorder are also likely to exhibit frequent masturbation or genital irritability and early signs of hypersexuality. The borderline adolescent, by contrast, is more likely to evidence the characteristics used to define BPD in the DSM-IV:

  • intense and unstable interpersonal relationships;

  • potentially self-damaging impulsivity;

  • marked identity disturbance;

  • recurrent suicidal behavior, gestures, or threats.
Although Kestenbaum has not completed a study focused specifically on the transition from borderline personality disorder to bipolar disorder, Decina and associates[14] and Kestenbaum[15] examined the clinical features of 31 offspring of adults with bipolar illness. First, they found that children of bipolar parents (experimental group) were significantly more likely to merit a psychiatric diagnosis (16 of 31) than were children of controls (1 of 18). Moreover, of the 15 experimental children who did not meet full criteria for a formal diagnosis, 7 shared characteristics too minimal to classify but nonetheless indicative of a psychiatric problem. Second, mean performance IQ was much lower for the experimental group than for the control group, while verbal IQ was the same. Third, the children of bipolar parents were significantly more likely to be left-handed than control children. Finally, when given Rorschach tests, children of bipolar parents identified color more often than movement, a finding associated with high levels of affective expression.

From this and other studies she has conducted, Kestenbaum concluded that children of bipolar parents are more likely to display affective instability; hypersensitivity to loss; and irritable, histrionic, or impulsive behavior.[15,16,17] They are frequently very bright, creative, artistic, and sensitive people, however. Thus, while many appear to have features of BPD, Kestenbaum concluded, bipolar children and adolescents are "very gifted, treatable, and worthwhile people."

About the Author

Dr. Carver is a Clinical Associate in Psychiatry at the Payne Whitney Clinic, New York Hospital-Cornell Medical Center in New York, N.Y.

Online Resources