Borderline Personality Disorder Research Paper

Outline

I. Historical Development of the Concept of Borderline Personality Disorder

II. Borderline Personality Disorder Core Symptoms and Character Styles

III. Demographic and Data-Based Studies of Borderline Personality Disorder

IV. Etiology and Relationship to Other Disorders

A. Psychoanalytic Hypotheses

B. BPD as an Affective Spectrum Disorder

C. BPD as Posttraumatic Stress Disorder Secondary to Childhood Sexual and Physical Abuse

D. BPD as an Impulse Spectrum Disorder

V. Course of Borderline Personality Disorder

VI. Treatment of BPD

I. Historical Development of the Concept of Borderline Personality Disorder

More than one decade after the development and publication of DSM-III, borderline personality disorder (BPD) remains the most controversial category in the nomenclature. Disagreement persists regarding the term itself, the particular diagnostic criteria established for BPD by DSM-III and DSM-IV, the scope of applicability, and the extent of overlap with Axis I and other Axis II disorders. Ultimately, this degree and intensity of dispute reflect both the range of difficulties in identifying and working with those persons designated as borderline, as well as the more basic question of validity: whether the BPD construct describes a meaningful unitary syndrome that corresponds to an actually existing state of affairs. While this latter question can certainly be asked of any of the personality (Axis II) disorders, something about the borderline concept seems to have engendered the strongest controversy.

At least one major reason for the ongoing disputes is the fact that the very concept of borderline was born out of attempts to explain the clinical observation that certain patients seemed to do very poorly in psychodynamic psychotherapy. Thus, from the very first, this category was used to describe a disparate group of patients who had two things in common: they responded to psychotherapy by developing transient psychotic symptoms and they did not meet classical definitions of schizophrenia. It is not that they did not necessarily improve; many obsessional patients, for example, did not improve with psychotherapy. Rather, it is that these patients worsened in psychotherapy with a fairly specific pattern of acting out that showed up most dramatically in the development of severe transference problems. The difficulty confronting the predominantly psychoanalytic theoreticians and skilled therapists was how to fathom the nature of these patients who gave promise of being good psychotherapeutic cases, yet deteriorated during the course of a psychotherapy. Thus, the very origins of the borderline concept arose in the context of a clinical puzzle.

The solution to the puzzle, keeping in mind that American psychiatry held a much more encompassing concept of schizophrenia in the 1940s and 1950s than at present, was to conceptualize these patients who became worse in psychotherapy as having a schizophrenic core underlying the neurotic facade. This notion was given concrete expression in a paper by Hoch and Polatin in 1949 describing the new category of pseudoneurotic schizophrenia. The construct fit neatly into a psychoanalytic model that postulated a spectrum of psychopathology based upon increasing primitiveness of defense mechanisms, extending in an unbroken chain from mild neurotics at one end to deteriorated schizophrenics at the other. The pseudoneurotic patient served as the missing link, bridging neurosis and psychosis, and thus serving as visible proof of the continuity connecting mild and severe psychiatric disorders.

The problem with the pseudoneurotic schizophrenia construct was that the patients did not go on to develop the more classical symptoms of hallucinations and delusions nor the deteriorating course that is the usual outcome of schizophrenia. Nevertheless, the observation that there existed a group of patients who appeared neurotic, but worsened with intensive psychotherapy, was a valid finding that outlived the misleading label attached to it. The focus of what might be wrong with these difficult-to-treat patients shifted away from schizophrenia to consideration of severe character pathology, described as borderline states by Knight in 1953 and as the psychotic character by Frosch in 1964. In addition, the joint U.S.-U.K. diagnostic studies carried out in the mid-to-late 1960s demonstrated convincingly that many patients diagnosed as schizophrenic by American psychiatrists fit much better with manic-depressive and personality disorder symptoms and outcome. This diagnostic realignment tightened the diagnostic criteria for schizophrenia, thereby further emphasizing the differences between borderline conditions and schizophrenia.

In 1968, Grinker and colleagues published the results of their study of 58 hospitalized patients who fell into a broadly defined notion of borderline syndrome. These patients had difficulties in interpersonal relationships, transient losses of reality testing under stress, angry and depressive affects, and deficient self-identities. Cluster analyses of the data, primarily of measurements of ego functions, produced four major clusters. There was a “core” borderline group, two groups defined as bordering upon the psychoses and neuroses, and a fourth group embodying certain “as-if” features, most notably absence of a core self-identity. Grinker’s study, the first to utilize psychometric instruments and statistical analyses, moved the borderline concept away from the realm of schizophrenic spectrum disorders and provided the basis for future empirical studies that continued the attempt to define the still vague borderline syndrome.

It is instructive that in the next series of studies carried out by Gunderson and Singer in 1975, the primary diagnostic concern was still to demonstrate that borderlines were different than schizophrenics. At the same time that empirical studies were focusing on narrowing the construct of borderline, Kernberg developed a broader notion of borderline, based upon a fusion of ego psychology and object relations theory, to designate a form of personality organization that was characterized by the use of primitive ego defenses (denial, splitting, projective identification), intact reality testing (with transient regressions under stress), and identity diffusion. Kernberg’s construct of borderline personality organization includes the milder as well as the more severe forms of character pathology, and, in essence, encompasses most of the patients presently grouped under the Cluster B (dramatic, unstable)personality disorders: histrionic, narcissistic, borderline, and antisocial.

This was the state of affairs while the DSM-IV committee developed inclusion and exclusion criteria for the personality disorders. There were four competing and overlapping concepts of borderline, and the final result represented some degree of compromise between the various groups. Since ideological and economic considerations, in addition to empirical studies and clinical lore, influenced the final product, it is important to define these considerations in some detail. The four overlapping concepts of borderline were as follows: (1) A residual model based upon the schizophrenic spectrum concept, using the term borderline to designate those persons, usually relatives of schizophrenics, who displayed odd, eccentric thinking and schizoid interpersonal relationships; this group was given the term schizotypal personality disorder. (2) An affective disorder model, which considered BPD as an affective spectrum illness displaying prominent features of mood instability with a predominance of depression, anger, and preoccupations with suicide. (3) An empirically derived model based primarily on the research of Gunderson, with diagnostic symptoms placed into five major groupings: impulse/action patterns (including self-destructive behaviors); ego-dystonic, transient psychotic episodes; mood instability with primarily negative affects; disturbed but intense interpersonal relationships; and an unstable sense of self. (4) A psychoanalytic concept based primarily on the work of Kernberg, but encompassing theoretical formulations by Mahler relating to difficulties in the separation/individuation phase of child development.

The final configuration of BPD adopted was most influenced by Gunderson’s work, but nevertheless showed the strains inherent in a compromise between points of view that are ideologically very divergent. The results were the creation of several new personality disorders within Axis II, not based upon empirical studies, but with each reflecting to some extent components that were once loosely connected to the borderline concept. Essentially, in dividing the broad territory of the borderline syndrome, as this concept evolved during a 40-year span, the cognitive disturbances that had long been noticed were placed in the schizotypal personality disorder, the milder dramatic and attention-seeking traits were placed into the histrionic personality disorder, self-centeredness and entitlement became the core of the narcissistic personality disorder, and the affective symptoms of mood instability and negative affectivity (depression, anger, anxiety), along with impulsivity, were given prominence in the borderline personality disorder.

Borderline personality disorder was defined by DSM-III-R as a condition marked by a pervasive pattern of instability of mood, interpersonal relationships, and self-image, beginning by early adulthood and present in a variety of contexts, as indicated by at least five of the following:

1. A pattern of unstable and intense interpersonal relationships characterized by alternating between extremes of overidealization and devaluation.
2. Impulsiveness in at least two areas that are potentially self-damaging, e.g., spending, sex, substance use, shoplifting, reckless driving, binge eating.
3. Affective instability: marked shifts from baseline mood to depression, irritability, or anxiety, usually lasting a few hours and only rarely more than a few days.
4. Inappropriate, intense anger or lack of control of anger, e.g., frequent displays of temper, constant anger, physical fights.
5. Recurrent suicidal threats, gestures, or behavior, or self-mutilating behavior.
6. Marked and persistent identity disturbance manifested by uncertainty about at least two of the following: self-image, sexual orientation, long-term goals or career choice, type of friends desired, preferred values.
7. Chronic feeling of emptiness or boredom.
8. Frantic efforts to avoid real or imagined abandonment.

The revision of DSM-III-R into DSM-IV was completed by late 1993. Although the BPD construct did not undergo any major alterations, several changes were instituted which served to correct the overemphasis in DSM-III on the close relationship between BPD and the affective disorders and the omission of cognitive deficits. Criterion 3 (Criterion 6 in DSM-IV), which outlined the affective symptoms seen in BPD was changed to reflect reactivity of mood; this serves to emphasize the difference between the mood disturbances seen in BPD and the relatively situationindependent mood disturbances characteristic of the endogenous affective disorders (major depression and manic-depressive illnesses). Complementing this more accurate delineation of the type of mood disorder seen in BPD was the inclusion of a new criterion to reflect the specific cognitive disturbances of BPD. The DSM-IV calls for a ninth criterion as follows: Transient stress-related paranoid ideation or severe dissociative symptoms. There were a few additional changes to the original eight criteria, but these are relatively minor, either reflecting grammatical alterations in the interest of clarity or the result of low sensitivity/specificity ratings for a few items on further field testing. Thus, the description of the identity disturbance in Criterion 6 was reworded and the construct “boredom” was dropped from Criterion 7.

II. Borderline Personality Disorder Core Symptoms and Character Style

The clinical description of a psychiatric disorder does not correspond exactly to that disorder’s diagnostic criteria in DSM-III. The main reason for this is that a clinical description needs to be a full and rich portrayal of the condition under question, whereas the requirements for diagnostic criteria are vastly different. Diagnostic criteria must aim for those characteristics of an illness that capture a few of its core symptoms while avoiding overlap with neighboring conditions. For example, as indicated above, while boredom may very well be a characteristic mental state in BPD, it was also found in histrionic and narcissistic personality disorders and therefore was of little specific diagnostic value. It did not help discriminate between BPD and other Cluster B personality disorders. In addition, diagnostic criteria must have acceptable validity and reliability. The issue of validity of psychiatric disorders, especially of personality disorders, is a troublesome one, since there are not external validators. The construction of DSM-III had paid major attention, some would say excessively so, to reliability issues. For example, certain factors that most workers would agree are characteristic of a disorder, such as the psychological defense of splitting in BPD, were not included in the diagnostic criteria because of a preference for behavioral rather than psychological phenomena, presumably because assessment of behaviors permits greater agreement as to whether they are present or not as compared to psychological constructs.

As indicated at the beginning of this research paper, there remains considerable controversy about the core characteristics and boundaries of BPD. Workers in the field have tended to bring to the evaluation of BPD their own theoretical and clinical perspectives in the evaluation of borderlines. In addition, some of the core characteristics of BPD, such as an increase in dissociative phenomena, appear to be changing in the past decade, a possibility that raises the question of the cultural influences and even faddish quality of some of the symptoms.

Most workers would agree that BPD is a relatively severe personality disorder, seen primarily in young adults, that presents with a characteristic cognitive style, mood disturbances, problematic interpersonal relationships, negative and deficient sense of self, and a variety of dramatic and impulsive behaviors usually of a self-injurious nature. These diagnostic features represent points distributed on a continuum of personality traits with somewhat arbitrary use of social norms to determine cut-off scores separating normal from pathological. Because of this, some workers in the field have advocated use of a dimensional rather than categorical model for the personality disorders, but a categorical model has always been adopted because it is easier to use in clinical work.

The cognitive style seen in borderline individuals encompasses three overlapping features. First, borderlines tend to have altered states of consciousness; these are usually referred to as dissociative states, and vary in intensity, density, and duration. They run the gamut from brief periods of self-absorption to fugue states lasting hours. The person may be partially or fully amnestic for some of the dissociative episodes. Second, borderlines tend to split their universe into good and bad, black and white. They have difficulty conceptualizing a person, including themselves, or an event, as encompassing positive and negative features. They tend to swing between the opposite poles of idealization and devaluation in their affections toward others. Third, borderlines tend to have impressionistic and global rather than precise and focused perceptions. They tend to be intolerant of unpleasant thoughts and images and to interrupt these processes with impulsive action, dissociation, and drug and alcohol use. There is a tendency toward imprecision and exaggeration, with a loss of salient detail. All of these disturbances are increased under conditions of stress.

The affective disturbances are characteristically mood instability or lability. Mood is typically reactive to environmental circumstances, but this must be taken to include the borderline’s own thought processes too. Negative affects, such as sadness, anger, and anxiety predominate the emotional landscape, but too literal adherence to this description would belie the positive affects and interpersonal warmth that borderlines can exhibit.

Problematic interpersonal relationships are a hallmark of borderlines. Their relationships are characteristically intense, stormy, and conflictual. Dependency needs, power struggles, and the idealization/devaluation swings described earlier tend to complicate most meaningful relationships. Victimization and entitlement themes in which the borderline alternates between being exploited by others and demanding reparations from others for damages incurred are frequent patterns seen in this disorder.

Borderline individuals tend to have a deficient sense of self, and what enduring image of themselves they may have is usually negative. A deficient sense of self refers to the absence of a stable sense of core identity, of knowing who you are. A certain degree of this is expected in adolescents and young adults in Western culture, but the borderline problem with identity, by definition, must go beyond the norm for this age group. Borderlines will take on different roles and personality characteristics, depending upon the dominant features of the group they are associating with. This has been referred to as the “as-if” personality, first described by Helene Deutsch in 1942. When not caught up in a persuasive group identity, borderlines tend to have very negative notions about themselves, ranging from dislike to contemptuous loathing.

Finally, borderlines characteristically are dramatic and impulsive in their actions. The patterns of impulsivity include directly self-injurious behaviors as well as an assortment of either ill-considered or risk-taking behaviors that also may be seen as self-destructive. Alcohol and drug abuse, bulimic eating disorders, promiscuity, and attraction to predatory partners are among the impusive actions seen in borderlines. As with the other core features of borderlines, the self-injurious behaviors range from infrequent and mild delicate cutting of the wrists to deep cutting of the limbs, torso, and genitals, as well as occasional ingenious use of cigarettes, lighters, caustic solutions, and hot irons to burn themselves. Suicide threats and attempts are also hallmarks of borderlines, most frequently but not exclusively with prescription as well as nonprescription medication overdoses. There are many more threats and gestures than serious attempts, leading to the use of the term “para-suicide” to describe these provocative actions of borderlines, but often the differentiation between manipulative and serious attempts is not at all clear.

III. Demographic and Data-Based Studies of Borderline Personality Disorder

There are no accurate measures of the prevalence of BPD in the community. Most estimates range from 0.5 to 1%, but may go higher as a broader concept of borderline, such as that used by Kernberg, is applied. The prevalence of the disorder in clinical settings is influenced by the type of clinical population under consideration. An average across studies indicates that the general prevalence of BPD is 10-15%, in inpatient settings about 20%, among outpatients with a personality disorder 30-35%, and among inpatients with a personality disorder 60-65%. Prevalence figures alone may be deceptive; it is possible that borderlines in an inpatient setting may have little similarity to outpatients who have never needed hospitalization. In most studies, excepting those done in VA and prison settings, 60-75 % of BPD are women.

Although DSM-III diagnostic rules do not permit differential weighting of the different criteria, most studies have demonstrated that several items contribute disproportionately to diagnostic efficiency. The presence of two, or at most three, specific criteria (impulsivity, unstable-intense interpersonal relationships, and self-injurious behaviors) predict most strongly the diagnosis of BPD, although once again, the type of clinical setting (inpatient or outpatient) will influence this finding.

There is considerable overlap (20-60%) between BPD and the other personality disorders, especially those of Cluster B, as well as schizotypal and dependent personality disorders. This finding continues to raise the question of whether personality disorders are discrete entities truly different from each other or reflect points on a continuum of serious character pathology. There are several Axis I disorders that have substantial overlap with BPD. These are alcohol and substance abuse disorders, bulimia, and the mood disorders, primarily dysthymia and major depression. To some extent, this finding reflects overlapping criteria (e.g., substance abuse is listed as a criterion for BPD), the heterogeneity of the BPD concept, and the fact that traits such as impulsivity and mood lability do express themselves in a wide array of behaviors.

IV. Etiology and Relationship to Other Disorders

Since it appears that BPD is not a unitary disorder, and since diagnostic threshold can be met in a polythetic system by fulfilling any five of eight (or nine, under DSM-IV) criteria, it is highly unlikely that a unitary etiology will be found for this or other Cluster B personality disorders. Theories about the etiology of BPD tend to follow major trends of interest in the behavioral sciences in general. Thus, the predominance of psychoanalytic constructs as explanatory hypotheses of human health and illness has given way to a variety of biological-genetic models in the past decade. Even the recent robust correlations between childhood sexual abuse and adult BPD symptoms are increasingly explained more in terms of long-lasting neurophysiological alterations of stress-response systems rather than in terms of psychodynamic mechanisms. The major theories of the etiology of BPD are as follow:

1. Psychoanalytic model of stage-specific difficulties
2. Deficit model (Masterson; Adler)
3. Conflict model (Kernberg)
4. BPD as an affective spectrum model
5. BPD as post-traumatic stress disorder secondary to childhood sexual and physical abuse
6. BPD as an impulse spectrum disorder

A. Psychoanalytic Hypotheses

Based upon Mahler’s theories of the importance of successful resolution of the rapprochment subphase of the separation/individuation processes in toddlerhood (ages 15-30 months), several overlapping psychodynamic hypotheses were advanced to explain those BPD features that were thought to represent the consequences of rapprochment failure. These features were the mental operation and defense of splitting, identity diffusion, and deficiencies in object constancy and object relationships. Differences of opinion and emphasis exist between various psychodynamic theories: Masterson has suggested that the mother of the borderline is herself borderline and establishes emotionally impossible conditions for the toddler to achieve age-appropriate separation and individuation, thereby resulting in the development of a borderline personality in the child. Adler has emphasized the borderline child’s inability, under circumstances similar to those described by Masterson, to form internalized soothing, holding introjects, such that the borderline child (and adult) lacks basic ego functions such as frustration tolerance, stable self-object relationships, and methods for calming itself during periods of stress. Kernberg has postulated the likelihood of an excessive aggressive drive in the infant that interferes with the fusion of sexual and aggressive drives; Kernberg’s model therefore sees borderline pathogenesis as the result of a complex interaction between infant and caregiver rather than as unilaterally caused by a “not-good-enough” mother.

The basic problem with the psychoanalytic hypotheses regarding etiology of BPD is shared by psychodynamic explanations of behavior in general: first, difficulty in operationalizing and thereby in testing various theories and second, a lack of specificity whereby certain postulated mechanisms at best appear to be general risk factors (e.g., parental psychopathology) rather than the specific and inevitable cause of a particular outcome. This latter problem, of course, applies to all unitary theories of etiology. Finally, the nature of the psychodynamic hypotheses are such that supportive evidence comes primarily from retrospective rather than prospective studies, and from individual case studies in which the investigator testing the hypothesis is also the therapist commited to the hypothesis.

B. BPD as an Affective Spectrum Disorder

The observation that borderline patients are frequently depressed, and the prominence of mood instability in the symptom picture, have led to the hypothesis that an affective disorder underlies the borderline condition. Attempts to validate this hypothesis examined a variety of biological markers, familial patterns, follow-up data, and pharmacological responses. The initial findings, varying somewhat from study to study, were that from 20 to 60% of borderline patients met diagnostic criteria for an affective disorder, usually major depressive episode. This was not particularly surprising since the diagnostic criteria for BPD were slanted toward affective type symptoms. The studies have shown that patients with depression and borderline patients who were concurrently depressed resembled each other in regard to several biological markers of depression, such as the dexamethasone suppression test, REM latency time, and thyroid stimulating hormone response to thyrotropin, but the resemblances fell away with “pure” borderline patients, i.e., borderline patients who were not depressed.

Similar results were found in the family pedigrees of borderline patients. Borderline patients with concurrent depressions had a greater prevalence of relatives with affective disorders. However, this finding is true for most of the Axis II disorders, namely, that there is a higher prevalence of depressed persons in the families of patients with any personality disorder and depression. On the other hand, borderline patients without depression tend to have increased familial linkages to other disorders, namely, borderline and antisocial personality disorders, and alcoholism and drug abuse. Studies of pharmacological efficacy with borderlines have demonstrated minimal benefit from antidepressants, even with depressed borderlines, except for some amelioration of depressive symptoms. Lithium therapy has not proven valuable in treating BPD. There have been some indications that monoamine oxidase inhibitors are effective in reducing core borderline symptoms, thereby supporting the atypical depression model of BPD, but these findings have never been sufficiently replicated to be more than suggestive. Finally, the long-term follow-up studies have shown that most borderline patients do not go on to develop depressive syndromes, again arguing against a causal linkage between BPD and affective disorders.

Despite the fairly clear evidence that BPD is not a variant of affective disorders, most studies do show that a certain percentage of borderline patients have a recurrent affective disorder (either depressive or bipolar type II, i.e., depressions and hypomanias) and evolve into a typical affective disorder pattern after the dramatic borderline symptoms recede in the 30s. Thus, it seems likely that a subclass of borderlines has a primary affective disturbance.

C. BPD as Posttraumatic Stress Disorder Secondary to Childhood Sexual and Physical Abuse

There has been an increasing awareness of the frequency of childhood sexual abuse in the life history of many psychiatric patients. This awareness has paralleled a growing public consciousness of domestic violence of many types. The question remains unresolved as to whether child abuse and other forms of violence have indeed become more common recently, reaching epidemic proportions, or whether the social taboos that maintained silence over such assaults have been lifted, with the result of greater case-finding and reporting of such episodes. Among psychiatric patients, rates of childhood sexual abuse range between 25 and 80%, depending on the population surveyed and the survey methods. Surveys from such varied locations as state hospitals, community hospitals, outpatient clinics, and emergency rooms have been consistent in these findings. Reported rates are highest for borderline personality disorder, in the order of 50-80%. In the borderline population, there also appears to be a correlation between severity of certain types of symptoms, such as self-injurious behaviors and dissociative episodes, and the severity of the childhood sexual abuse experiences, as judged by age of first abuse, frequency and duration of abuse, degree of force and violence employed, and absence of ameliorative factors in the life of the child. The correlations between abuse and borderline symptomatology have been robust enough to lead several workers to hypothesize that most patients who have been diagnosed BPD are really suffering from PTSD and that this latter diagnosis makes better scientific and social sense, removing the stigma that has been attached to a BPD label. The case is strengthened by the logic of borderline symptoms, such as dissociation, as a learned response of the abused child to the horrors of the abuse experience, a response that was once adaptive, but has now become generalized as a response to all emotional flooding. In a similar way, self-injurious behavior seems to make sense as an expression of the self-hatred that the abuse victim directs inwardly.

There are several obvious problems to the linear causal chain that links childhood abuse to borderline symptomatology. The major problems relate to specificity between abuse and outcome. Patients with many psychiatric diagnoses, as well as many persons who do not have psychiatric symptoms have histories of childhood sexual abuse. Only a percentage of abused persons develop the BPD or PTSD picture. Conversely, not everyone with BPD has a history of childhood abuse. In addition, the abused child was most likely raised in a chaotic home with many other disturbing features, such that it is not valid to single out the experience of sexual abuse as the cause of adult problems. There are also considerable methodological problems related to the very sensitive nature of the topic and the fact that most of the research and clinical work are based upon retrospective reports of abuse in childhood. The methodological problems slice both ways; there are persons who have been abused and who deny it, and there are patients who may distort, exaggerate or invent abuse histories. There is no easy resolution to these issues, but, in general, the detailed reports by patients about their abuse appear to have credibility and are accepted by most researchers and health care workers. The particular diagnostic question discussed here about the overlap of BPD and PTSD, however, is less an issue of data than definition of causal relationships in human behavior. Thus, it appears that childhood sexual abuse and the disturbed environment in which the abuse occurred function as general risk factors predisposing to increased severity of many types of psychiatric and physical illnesses. Within the BPD population, there does appear to be a large subgroup whose symptoms and personality styles were profoundly affected by the experiences of childhood sexual abuse and whose symptoms can be understood as a form of PTSD. It needs to be kept in mind that PTSD is still a fairly vague concept encompassing many types of traumas and responses, and that most persons suffering from PTSD do not show borderline symptoms.

D. BPD as an Impulse Spectrum Disorder

Although it sounds tautological to say that a syndrome characterized by impulsivity may be an impulse spectrum disorder, more is implied in the statement than meets the eye. Essentially, such a hypothesis raises the question of whether there is a group of disorders that share some common features in addition to impulsivity, such as familial linkage, associated psychiatric disorders, and underlying neurophysiological mechanisms. Family studies have shown an increased rate of alcoholism, substance abuse, and antisocial personality in the relatives of borderline personality. Other disorders considered related to problems with impulsivity include compulsive gambling, bulimia, intermittent explosive disorder, and the other Axis II personality disorders within Cluster B (histrionic and narcissistic). Studies are presently under way to investigate serotonergic and dopaminergic mechanisms that may have some linkage to impulsive behaviors.

It is well recognized that the notion of “impulsivity” is very vague, such that the various conditions being considered as impulse disorders may turn out to have very little in common beyond surface appearances. Conceptual clarification concerning what the terms “impulsive” and “compulsive” mean, and how these relate to the notion of “addiction,” will be necessary if the hypothesis regarding impulse spectrum disorder is to be of any practical use.

V. Course of Borderline Personality Disorder

The initial delineation of borderlines as encompassing a group of difficult treatment cases combined with the finding of a poor outcome on short-term follow-up led to a fairly pessimistic outlook for patients with this diagnosis. Patients who were diagnosed in their late teens or early 20s as borderline were still doing poorly 2 to 5 years later, with ongoing self-injurious behavior and suicide attempts leading to multiple hospitalizations. It was not until the late 1980s that follow-up studies covered the 10- to 20-year period after initial hospitalization. Surprisingly, the outcome was much more favorable than the early studies indicated. In several independent studies from different parts of the country, it became clear that between 50 and 60% of BPD patients were doing fairly well as they moved into their 30s. Another 30-40% of patients showed varying levels of disability. Suicide rates ranged from 8 to 15 % on 10-year follow-up. The largest follow-up series of patients was reported by Stone, who traced 502 of 550 patients (of whom 193 met DSM-III criteria for BPD) who had been hospitalized on an intensive long-term psychotherapy ward at New York State Psychiatric Institute during the years 1963-1976. As judged by Global Outcome Scores (GAS), 63% of the BPD patients were in the good to recovered categories, another 16% had made a fair adjustment, 12% were doing poorly, and 9% suicided. Less favorable outcome was correlated with the presence of major affective disorder, antisocial personality, and a pattern of alcohol and drug abuse. Poor outcome was not correlated with self-mutilative behaviors in the early years of the illness. Patients with a history of childhood neglect or sexual abuse tended to do less well than patients without these histories. Finally, there was not a good overall correlation between outcome and psychiatric treatment; some patients with very good outcomes had minimal treatment following index hospitalization and some patients with extensive treatment had poor outcomes. It is possible that averaging the outcome data washes out a treatment effect, but this remains to be demonstrated.

VI. Treatment of BPD

Three has been as much controversy about the treatment of BPD as there has been about the diagnosis. To a large extent and with some overlap, treatment modalities have tended to follow etiological hypotheses. As one might expect with a condition that drew its initial delineation from a group of difficult-to-treat patients, no single modality has yet demonstrated clear-cut superiority or even effectiveness. Studies designed to evaluate treatment of BPD have been plagued by the usual problems of therapy outcome research: differing characteristics of the patient population, despite use of DSM-III criteria; difficulty in determining what constitutes evidence of improvement; difficulty in establishing control groups.

Psychodynamic psychotherapy has been the standard and accepted form of treatment of BPD, despite the many problems that arise in this form of treatment. In a sense, the BPD population, comprising primarily young verbal adults who are dysfunctional but nonpsychotic, have appeared to be the obvious if not ideal candidates for psychotherapy. Close to 50% of psychotherapy patients seen in private practice and at most outpatient clinics will have a diagnosis of BPD or a related Axis II Cluster B (narcissistic or histrionic) disorder. While there has been no canon defining a specific therapeutic protocol for BPD (or any other disorder), the work of Kernberg has been most influential in guiding the theory and practice of psychotherapy with borderlines. The therapy has tended to be a mix of supportive and exploratory work, with special attention paid to avoiding becoming enmeshed in ill-advised rescue attempts and other acting out features that are the hallmarks of borderline patients. The outcome results of the Menninger psychotherapy project reported by Wallerstein and Stone’s follow-up study suggest that it is impossible to predict, from patient characteristics alone, which patients would benefit most from supportive and which from exploratory psychotherapy, nor is there evidence that ultimate outcome is better with exploratory than supportive psychotherapy. A single study by Stevenson and Meares employing a 12-month psychotherapy regimen that utilized a written protocol based upon self-psychology demonstrated significant improvements across a broad range of measurements. Patients served as their own controls (pre- and post-treatment measures); a separate control group of patients was not used.

There has been increasing interest in cognitive-behavioral treatment (CBT) modalities for BPD. The essence of these modalities is a focus on recognizing and eliminating the factors that reinforce self-injurious behaviors, and learning and practicing new behaviors that will enhance the quality of life of the patient. Therapy is not directed toward underlying psychodynamic causes, since the assumption of CBT is that self-injurious behavior is a learned behavior that has become relatively independent of the specific causes that originally inspired it. CBT is done individually and in groups. Techniques that are taught and practiced include behavioral skill training, contingency management, cognitive restructuring, exposure to emotional cues, distress tolerance, interpersonal skills, and emotional regulation. Linehan and colleagues reported significant improvement in self-injurious and parasuicidal behaviors in a group of SIB borderlines in CBT compared to a group receiving treatment as usual. The improvements were not accompanied by changes in severity of reported depression, suicidal ideation, or reasons for living.

The relationship of BPD to PTSD in those borderlines who experienced sexual abuse in childhood suggests that a PTSD-oriented treatment program should be helpful. To date, this has not been the case, most likely because no overall effective program for the treatment of PTSD has been demonstrated. The treatment of PTSD usually includes group therapy, desensitization techniques, and pharmacological agents. There has been a proliferation of incest and sexual abuse treatment groups, some of which seem to be very helpful and some of which have a deleterious effect on some group members. No controlled studies have been reported. Pharmacological treatment of PTSD is in its infancy; different medications have been reported to be effective with particular components of PTSD, especially the sleep disturbance and depressions that accompany PTSD, but no agents appear to interrupt the flashbacks and intrusive imagery that form the hallmark of this disorder.

The pharmacological treatment of PBD is widely used, but relatively disappointing. Tricyclic antidepressants are effective only in alleviating depressive symptoms in those borderlines who are also depressed. Monoamine oxidase inhibitors have been reported to reduce the target symptom of rejection-sensitive dysphoria, but a controlled study is still wanting. There have not been controlled studies of the efficacy of the specific serotonin reuptake blockers to date. Lithium has not appeared to be of special benefit. There are mixed reports on the benzodiazepine anti-anxiety agents; there may be some benefit to the anti-anxiety properties, but several studies have reported a worsening of impulsive behaviors in BPDs taking these agents. In addition, long-term use of benzodiazepines would not be indicated in patients with significant alcohol or drug abuse histories. The single class of medications that has demonstrated significant short-term effectiveness in several key borderline symptoms has been low-dose anti-psychotics, but here the benefits must be weighed against the serious long-term side effects of these agents. A study of Soloff and associates in 1993 failed to replicate the positive findings of their earlier study reporting improvement in borderline patients with the use of antipsychotic medications.

There has been a recent trend away from long hospitalizations for borderline patients. While much of the driving force toward brief hospitalizations in all medical fields has been concern about rising medical costs, there has also been growing awareness of the deleterious rather than helpful effect of prolonged hospitalization of borderlines. Although there are undoubtedly some patients who benefit from a controlled hospital environment that prevents major self-destructiveness, the general experience has been that borderline patients continue their self-injurious behaviors in the hospital. This behavior sets up major conflicts with staff regarding proper responses to patients who challenge staff to prevent them from hurting themselves. Placing patients on one-to-one or constant observation has seemed to encourage rather than discourage self-injurious acts. The broad, but not unanimous consensus recently is that hospitalizations should be kept as brief as possible within the boundaries of responsible patient care, with the option of brief rehospitalizations seen as preferable to lengthy hospital stays.

Bibliography:

1. Druck, A. (1989). “Four Therapeutic Approaches to the Borderline Patient.” Jason Aronson, Northvale, NJ.
2. Gunderson, J. G. (1984). “Borderline Personality Disorder.” American Psychiatric Press, Washington, DC.
3. Herman, J. L. (1992). “Trauma and Recovery.” Basic Books, New York.
4. Kernberg, O. (1984). “Severe Personality Disorders.” Yale University Press, New Haven, CT.
5. Kroll, J. (1988). “The Challenge of the Borderline Patient.” WW Norton, New York.
6. Kroll, J. (1993). “PTSD/Borderlines in Therapy: Finding the Balance,” Norton, New York.
7. Linehan, M. M., Armstrong, H. E., Suarez, A., Allmon, D., and Heard, H. L. (1991 ). Cognitive-behavioral treatment of chronically parasuicidal borderline patients. Arch. Gen. Psychiatry 48, 1060-1064.
8. Links, P. S. (1990). “Family Environment and Borderline Personality Disorder.” American Psychiatric Press, Washington, DC.
9. Paris, J. (1992). “Borderline Personality Disorder: Etiology and Treatment.” American Psychiatric Press, Washington, DC.
10. Soloff, P. H., Cornelius, J., George, A., Nathan, S., Perel, J. M., and Ulrich, R. F. (1993). Efficacy of phenelzine and haloperidol in borderline personality disorder. Arch. Gen. Psychiatry 50, 377-385.
11. Stevenson, J., and Meares, R. (1992). An outcome study of psychotherapy for patients with borderline personality disorder. Am. J. Psychiatry 149, 358-362.
12. Stone, M. (1990). “The Fate of Borderline Patients.” Guilford, New York.
13. Wallerstein, R. S. (1986). “Forty-Two Lives in Treatment.” Guilford, New York.