There are currently three major psychotherapeutic approaches to the management of borderline personality disorder (BPD): the psychodynamic, the cognitive-behavioral, and the supportive. There are special varieties within each: e.g., transference-focused psychotherapy (psychodynamic) or dialectic behavioral therapy (cognitive-behavioral). Though differing in basic conceptions and in methodology, all approaches aim at the amelioration of both the symptom-aspects that dominate the clinical picture at the outset, and the personality difficulties that remain apparent after the symptoms have been alleviated. The term “management” implies a focus on the more serious aspects of the borderline picture. These can be pictured hierarchically as to their level of seriousness, and there is universal agreement about the nature of this hierarchy. Therapists must pay attention first to suicidal and self-mutilative behaviors. Next, one deals with any threats to interrupt therapy prematurely. Third in order of seriousness: non-suicidal symptoms such as (mild to moderate) depression, substance abuse, panic and other anxiety manifestations, or dissociation. Psychopharmacological treatment will often be used adjunctively to help control any target symptoms, which usually fall into such categories as cognitive-perceptual, affect dysregulation, or impulsive/ behavioral dyscontrol. Therapists must then be alert to any signs of withholding, dishonesty, or antisocial tendencies, since these have an adverse effect on prognosis. When all these disruptive influences are (to the extent possible) dealt with, therapists will next take up milder symptoms such as social anxiety or lability of mood. Throughout this initial process, the personality-disorder attributes of BPD will become more apparent, and will usually emerge with greater clarity, once the serious symptoms have been dealt with. The management issues will gradually be supplanted with the overlapping and enduring personality issues: inappropriate anger, abrasiveness, manipulativeness, demandingness, jealousy, “all-or-none” thinking and the extreme attitudes (idealization/devaluation) that accompany such thinking, masochistic traits, etc. Under ideal circumstances, the borderline patient will have “graduated” toward a higher level of function, where (acute) management issues have been adequately dealt with or have receded into the background. Psychotherapy, individual and group, becomes the dominant intervention, with such goals as psychic integration, skills training, and the fostering of long-range ambitions relating to friendships, partner choice, and work.
Within the domain of personality disorders, borderline personality disorder (BPD) has attracted the greatest attention; the literature devoted to it is more voluminous than that devoted to any of the other recognized personality disorders. This attention and large literature has more to do with the challenging nature of the condition and the difficulties attendant upon treating it, than to its frequency. In fact, the pooled prevalence for BPD has been estimated as 1.16% in the general population, which is less than the percentages for the antisocial, histrionic, obsessive-compulsive, and avoidant types (1).
Part of the difficulty in treating patients with BPD lies in the complexity of the condition, which is (in the beginning phases at least) accompanied almost invariably by one or more symptoms (assigned in the DSM-IV to Axis I), alongside its pathological personality traits, as enumerated in Axis II. Among the latter, some are themselves symptomlike: particularly the self-damaging acts (such as wrist cutting) and suicidal gestures (or actual attempts) that so often accompany BPD. Obviously these self-damaging behaviors require not only immediate attention but also skillful handling, since they are (initially, anyway) ego-syntonic in the borderline patients, and therefore not as easily discouraged as such painful and disruptive behaviors would be in ordinary persons.
As for the Axis I disorders that will be noted with regularity in BPD patients, the common ones include serious depressive episodes, bipolar II disorder, eating disorders (anorexia and/or bulimia), panic and other anxiety disorders, abuse of alcohol and other substances, and dissociative disorders. Also common in certain subgroups of BPD patients is post-traumatic stress disorder (PTSD). The subgroup most likely to experience concomitant PTSD is that of BPD patients who had been the victims of incest by an older family member, especially before age ten. Females are much more likely to have been subjected to such experiences than are males, and this may account in part for the female preponderance in samples of BPD patients (2–4).
The role of mood disorder in BPD has been a matter of debate ever since BPD won official status as a diagnosis in 1980 (DSM-III). Akiskal (5) and Stone (6) have taken the position that a significant proportion of BPD patients developed their personality disorder as an expression or offshoot of an underlying mood disorder, often of a recurrent depressive or a bipolar II type. The BPD diagnosis may become manifest in late adolescence or in the early 20s, before the bipolar disorder finds full clinical expression. This evolution may be more common in certain samples of BPD patients; less evident in others. Other authors (7), in contrast, did not find compelling evidence for the equivalence of BPD and any mood disorder.
The fact remains that many BPD patients experience serious depressive episodes at some point in their life course and that the suicide rate in long-term (10 to 30 year) followup studies is high, ranging from 3% (8) to 10% (3,9). Stanley and Brodsky (10) made the important point that suicidal behaviors in BPD are not always the expression of mood disorder, although the coexistence of BPD and major depression augments the risk for the more serious (including fatal) forms of suicidal behavior. The wrist cutting and other forms of self-mutilation common in BPD, and the minor, non-lethal suicidal gestures are often done to relieve unbearable interpersonal tension (rather than to end one’s life) and are followed by the patient’s feeling better. This sequence of events, which strikes most clinicians as paradoxical, influences the management of BPD, since it may be desirable not to hospitalize those who engage in self-mutilation by way of relieving tension, whereas those whose suicidal behaviors arise out of chronic feelings of depression and hopelessness may indeed require inpatient care.
The proper management of patients with BPD also requires taking into account the totality of their personality. Clinicians will rarely, if ever, encounter a borderline patient in whom BPD is the only diagnosable “category” (à la DSM) of personality disorder. Instead, as Oldham et al (11) have demonstrated, the panoply of pathological traits manifested by borderline patients will fulfill category-based criteria for other personality disorders as well: often one or two additional types; sometimes three or more. Most commonly, these accompanying types will be within DSM’s “dramatic” cluster (cluster B): histrionic, narcissistic, or antisocial. Depressive-masochistic traits – not a category recognized in DSM, but well-described in the psychoanalytic literature (12) – are also commonly encountered, as are those of avoidant and (to a lesser extent) schizotypal personalities. Psychotherapy must be tailored to the admixture of traits noted in each BPD patient, assessed in dimensional terms (i.e., how much of each personality category is discernible in each case, and how intensely so). Both from the standpoint of amenability to psychotherapy and of ultimate prognosis, certain combinations prove easier and more rewarding to work with, and enjoy better long-term outcomes, while other combinations are more daunting and with less likelihood of success. BPD with depressive-masochistic features, for example, belongs generally to the more favorable group (such patients tend to be more introspective and motivated); BPD with narcissistic features has an intermediate prognosis; BPD with antisocial features will usually prove the most difficult to work with and carries a gloomy prognosis (13). BPD patients who lie and are generally dishonest will present enormous hurdles in the treatment, since such behavior is designed to evade or disparage, rather than cooperate with, the therapist (14). Other personality configurations that augur a poor prognosis, though not as ominously so as the antisocial, are the hypomanic and the paranoid (15). Among the BPD patients with marked paranoid “comorbidity” are those exhibiting pathological jealousy. Such patients often live at the very edge of delusion (or at times, on the other side) and cannot be reasoned out of their mistrust by argumentation, no matter how skillful, nor by proofs, no matter how compelling. One must try, in psychotherapy, to get below the surface where the mistrust resides, down to the lower layers of the psyche, where one will find the vast insecurity and self-doubt that feed the jealousy. Given the tendency of borderline patients with such jealousy to externalize, and to resist looking into themselves, this is no easy task. Some useful guidelines for dealing with such patients have been provided, nevertheless, by White and Mullen (16).
Psychopharmacological treatment will often be used adjunctively in the management of BPD to help control any target symptoms, which usually fall into such categories as cognitive-perceptual, affect dysregulation, or impulsive/behavioral dyscontrol. This aspect is not covered in the present review. It can be stated in brief, however, that the medications of choice for the cognitive-perceptual symptoms are atypical antipsychotics in low doses. For both affect dysregulation and impulsive/behavioral dyscontrol, the best approach is to use a selective serotonin reuptake inhibitor (SSRI). If an SSRI does not prove effective, one would then use a mood stabilizer such as lithium or valproate (17).
VARIETIES OF PSYCHOTHERAPY IN CURRENT USE FOR TREATING BORDERLINE PATIENTS
It has been customary in recent years, for didactic purposes, to divide the psychotherapeutic approaches to BPD into three main groups: psychoanalytically-oriented, cognitive- behavioral, and supportive. These approaches rest on differing theories and psychological foundations, such that practitioners trained in one approach tend to feel more “at home” in the techniques and tactics peculiar to that approach. In actual practice, and especially in the frequent crisis situations that characterize the course of BPD in the early phases, many therapists will find themselves relying upon tactics and interventions borrowed from an approach other than their primary one. Such shifts require a healthy measure of flexibility, besides lending an air of eclecticism to the practice of many who work intensively with borderline patients.
The psychodynamic therapies
The psychoanalytically-oriented or “psychodynamic” methods are based on the assumption that unconscious forces and conflicts are buffeting the borderline patient and are responsible for the sharply polarized attitudes and the often wildly oscillating behavioral patterns seen in BPD. Borderline patients, for example, tend to swing from adoration to contempt toward key figures in their lives (manifesting, in so doing, the defenses of idealization and devaluation). These attitudinal shifts may be triggered unpredictably via quite minor events in their interpersonal lives, such that other people experience the borderline person as unintegrated and existing only in extremes. The psychodynamic approach strives to promote psychic integration through the careful examination of the polarized attitudes (maintained outside awareness via the defense of splitting), making use of the fact that the therapist will soon be experienced, via the transference, as harboring the various strong feelings (of love, hatred, lust, jealousy, envy…) which stem actually from the borderline patient, but which for the time being exist outside the patient’s awareness and get draped, so to say, around the shoulders of the therapist, via the defense of projective identification. During this process, strong countertransference feelings will be elicited in the therapist. But these feelings become highly useful material for the therapist to process and ultimately interpret back to the patient, by way of increasing the patient’s conscious awareness of his basic conflicts and attitudes. The aim of the psychodynamic approach is, in general, to effect an integration of the hitherto disparate and split-off elements in the patient’s psyche, in hopes this will lead to more appropriate attitudes toward other people and to more modulated behaviors in everyday life.
In the early phases of work with borderline patients, the therapist may be confronted with all manner of life-threatening or self-mutilative behaviors, interpersonal crises, disruptions in the treatment, and the like, necessitating the use of various supportive interventions. When suicidality or extreme impulsivity are present and necessitate hospitalization, psychodynamic considerations take second place to limit-setting and supportive interventions, until stability is restored. Gunderson (18) has given an excellent overview of the stepwise progression of treatment modalities that are brought into play, as the hospitalized BPD patient moves from dangerous to more calm and reflective modes of interaction. Gabbard (19) has drawn attention to the manner in which supportive interventions are routinely used at various points along the way in most psychodynamic therapies with BPD patients.
In recent years a number of treatment guidelines, or “manuals”, have been developed for the major therapeutic approaches, including the psychodynamic. Kernberg et al (20), for example, have created a guideline for a transferencefocused psychotherapy (TFP) for BPD, in which supportive interventions are avoided except in situations of considerable urgency. In England, Bateman and Fonagy (21) have published a comprehensive guideline for their mentalizationbased treatment, which aims at fostering the capacity for reading the mental states of self and others more accurately (i.e., improving mentalization), so as to develop a more coherent sense of self and a better regulated set of emotions in relation to the external world. In their book the reader will find, however, an excellent summary of all the competing and widely used approaches to the management of BPD.
The psychodynamic approach of Kohut (22), known as self psychology, was developed primarily for ambulatory patients, especially those with narcissistic personalities. Some of the patients described by Kohut et al (23) exhibit the borderline personality organization as described by Kernberg, although Kohut and his collaborators expressed the view that the “borderline states” are not analyzable (23,24). The self-psychological approach utilizes, in any case, supportive interventions (expressions of sympathy, validation of the patient’s perceptions of others, etc.) to a greater extent in working even with the milder (and ambulatory) end of the borderline spectrum. In working with borderline patients, Kohut sought to convert such patients over time into an “analyzable narcissistic personality” – if the therapist can, as Kohut put it, “stand the heat” of the stormy emotions that will surely emerge in the beginning of the work (25). Gunderson, Kernberg, and Bateman and Fonagy, in contrast, have extensive experience with BPD patients, whether ambulatory or hospitalized, and in their numerous publications make a compelling case for the efficacy of a psychodynamic approach in a respectably high proportion of borderline patients. Many clinical vignettes attesting to this impression are included in my book on treatable and untreatable personality-disordered patients (13). One can, at all events, speak of a spectrum of psychodynamic approaches, insofar as supportive interventions are either used or avoided. In Kernberg’s TFP, supportive interventions are avoided insofar as possible, because these may “undermine working within the transference-countertransference paradigm and often lead to enactments of the countertransference” (14). Instead, TFP relies on such interventions as clarification (of confusing or poorly understood communications of the patient), confrontation (of the often paradoxical and polar-opposite attitudes BPD patients express concerning themselves and other people), and ultimately interpretations (aimed at making the patient aware of the split-off aspects of these disparate attitudes, so as to promote psychic integration). The presumption is that the important (and often turbulent) dynamic constellations in the borderline patient’s everyday life will eventually play themselves out within the transference relationship, there to be brought to light, clarified, and modulated along more adaptive paths. Gunderson’s approach, especially with hospitalized BPD patients, is more accepting of supportive interventions, especially at the outset; the ultimate aim is otherwise similar to that of TFP. Kohut’s approach, as mentioned, is less adapted to the more severe borderline patients, and in any case relies importantly on supportive measures by way of building a therapeutic alliance.
The cognitive-behavioral therapies
The cognitive-behavioral approaches to the treatment of borderline patients place emphasis on observable behaviors and on the psychic schemata or “inner scripts” (habitual patterns of thought concerning the self and the interpersonal world, built up during one’s developmental years). The behaviors and their underlying schemata have, in the evolution of BPD, become maladaptive for a variety of causative factors: hereditary predispositions, humiliations and other psychological hurts experienced at the hands of one’s caretakers, and in some cases, outright traumata stemming from physical or sexual abuse. The early maladaptive schemata pertinent to BPD, outlined by Beck and Freeman (26), include such basic assumptions as (among others): “I’ll be alone forever; no one will be there for me”; “I’m a bad person; I deserve to be punished”; “No one would love me if they really got to know me”. These assumptions reflect some important themes relevant to BPD patients: namely, fear of abandonment, conviction of unlovability, and exaggerated guilt. The cognitive distortions typical of BPD patients involve (as we noted in discussing the dynamic approaches) polarized all-or-none attitudes, which Beck refers to as “dichotomous thinking”. Though transference is not often discussed as such by cognitive- behavioral therapists, the phenomenon can readily be understood in cognitive language, as the responses on the part of the patient toward the therapist that are based on early embedded beliefs and expectations derived from past experience, rather than on the therapist as an actual and realistically-perceived individual. Therapy focuses on decreasing the tendency to dichotomous thinking, helping the patient develop better control over his emotions and impulses, and strengthening the patient’s sense of identity.
In the last 25 years, the dialectic behavior therapy (DBT) developed by Marsha Linehan has become increasingly popular and respected as a method particularly for minimizing the tendency among BPD patients to make suicide gestures or attempts, or to indulge in other forms of self-harm. She and her colleagues have created a manual for carrying out DBT (27). This treatment approach includes the use of one individual session per week with a therapist, along with a weekly group session oriented toward skills training. Phone calls to the therapist are permitted (whereas they are discouraged in TFP), but with the proviso that the therapist will speak with the patient who is about to self-cut or make a suicide gesture, in the expectation that their conversation will lead to an ability on the part of the patient to gain self-control and to refrain from the self-damaging act. If the patient calls after having made the self-damaging act, then the call is cut short, as a means of reducing the “secondary gain” that might otherwise have been derived from the therapist’s listening to what the patient had already in fact done. As for the dialectic nature of the individual sessions, the term refers to the patient-therapist dialogue directed at reconciling the polarized opposite feelings and attitudes of the patient, with the goal of achieving a synthesis (27). This involves the same kind of psychic integration and development of more appropriate behavior patterns, that are sought as goals in the psychodynamic therapies. Since many BPD patients had been abused and mishandled in childhood, in situations where the parents used intimidation and denial to paper over the reality of what was being done to their children, validation during the therapy becomes an important healing intervention: reassuring the borderline patient that certain negative experiences (neglect, unjustified punishment, incest…) really did happen, and that the patient was not “crazy” for thinking so. This intervention is often a crucial ingredient in the overall approach of DBT.
The supportive psychotherapies
Although Arnold Winston et al have expressed the view that supportive therapy is the “shell that fits over most theoretical orientations” (29), the literature on supportive therapy in relation to BPD is sparse. Appelbaum and Levy (30) have written on the application of supportive therapy to the treatment of BPD, in what becomes an amalgam of dynamic and supportive interventions of the sort earlier outlined by Rockland (31). Rockland mentioned such techniques as sympathetic listening, education, encouragement, limit setting, exhortation (to do or to refrain from certain behaviors), reassurance, advice, and validation. Appelbaum has been developing a manual of supportive psychotherapy for BPD, which will supplement those already published for TFP and DBT. Transference feelings are recognized in the supportive approaches, but transference interpretations are not used.
SHORT-TERM VERSUS LONG-TERM PROSPECTS
Many of the recent follow-up reports concerning BPD patients have centered on the issue of self-damaging or parasuicidal acts. Linehan et al (32) showed, for example, that BPD patients treated with DBT showed less likelihood to commit such acts by the end of a year’s therapy, as compared with patients offered “treatment as usual”. The latter involved fewer hours of therapy per week than DBT, so it is difficult to know how much improvement the DBT-treated patients owed to the method per se, as opposed to the greater time spent with each of them. In the last few years, other approaches have in any case achieved comparable results (reduction of self-harm after a year’s therapy). This has been demonstrated for TFP (33) and for Peter Tyrer’s manual-assisted cognitive behavior therapy (MACT) (34). These studies, taken in the aggregate, suggest that wellplanned time-intensive therapy, carried out by skilled therapists using a variety of approaches, can achieve rewarding results with some of the symptomatic aspects of BPD (i.e., those that center on self-damaging acts), and can do so within the first year of therapy.
Although the new manuals focus on the first year to year-and-a-half, psychotherapy with borderline patients is a much longer enterprise. This is because one must aim not merely at the reduction of self-damaging acts and other troubling symptoms (such as depression, eating disorders, panic, dissociation and the like), but at improvement in the borderline patient’s ability to function in work, at developing and maintaining friendships and, if possible, at fostering some measure of success in intimate relationships. These goals will seldom be met short of five or ten years of therapy, and long-term (10 years or more) followup studies are necessary to appraise the results of psychotherapy.
A HIERARCHY OF TASKS IN THE MANAGEMENT OF BORDERLINE PATIENTS
Strictly speaking, the term management, in relation to psychiatric treatment, conveys the notion of establishing a set of remedies for fairly acute or severe disorders. One “manages” acute symptoms, whereas one applies psychotherapeutic principles to the amelioration of maladaptive personality traits. Since BPD represents an agglomeration of symptoms and maladaptive traits, and often is heralded by serious or even life-threatening symptoms, management issues may dominate the scene at the outset, even as the verbal psychotherapy per se is getting underway. Gradually, psychotherapeutic issues come to dominate the scene, as the symptom aspects subside and become less disruptive.
A number of prominent clinicians who deal extensively with borderline patients, such as Kernberg and Linehan, have outlined hierarchies of tasks with which the therapist will be confronted as the work unfolds. Given the nature of BPD and the similarities between one large group of borderline patients and another, these hierarchies are quite similar. They will consist of a series of tasks, wellordered as to their urgency, creating in this way a list of problems that therapists should pay attention to as they go about treating a borderline patient. Here is one convenient arrangement of these therapeutic tasks:
Make sure suicidality (in whatever form(s): ideation, threats, gestures, attempts) is explored and adequately dealt with.
Deal promptly with the patient’s threat to interrupt treatment prematurely.
Inquire about and treat any severe non-suicidal symptoms (e.g., panic, anorexia, dissociation, depression or hypomania, substance abuse, compulsions).
Be alert to any signs of gross withholding of important information or any signs of dishonesty, or of antisocial tendencies.
If the foregoing tasks are taken care of, attend to less disruptive symptoms that may be present (e.g., social anxiety, mild depression).
Focus on personality traits that cause significant trouble at work or in interpersonal relationships (hostility, abrasiveness, bitterness, jealousy, manipulativeness, demandingness…).
Focus on less disruptive personality traits, including those that are more troublesome to the patient than to others (shyness, “moral masochism”, obsequiousness, unassertiveness…).
Focus on long-range occupational, educational (if pertinent), and interpersonal goals; appropriateness of partner choices, and attitudes toward family members.
In this hierarchical schema, management relates primarily to steps 1 through 5. Borderline patients who improve and progress toward the latter steps are generally the ones who evolve into persons who no longer meet DSM criteria for BPD (35,36). These are the borderline patients whose therapy takes on more and more the qualities of psychotherapy with those better-functioning patients who were called psychoneurotic in the older literature and whose personality disorder would now be better characterized with a milder label, such as one of the cluster C (anxious cluster) disorders.