In practice, long-term individual psychotherapy is the cornerstone of most treatments for BPD. Howard et al. showed that borderline patients require longer-term therapy to achieve benefits than do other types of patients. Hoke found that those who remained in stable psychotherapy for 2 or more years showed more improvement than those who used therapy more intermittently. Several controversies exist in the vast literature on psychotherapies for borderline patients, but areas of agreement also exist. From this literature, the following areas of agreement regarding essential components of treatment can be identified:
• Providing a stable treatment framework
• Having highly active and involved therapists
• Establishing a connection between the patient’s actions and feelings
• Identifying adverse effects of self-destructive behaviors
• Paying careful attention to countertransference feelings
In addition to these areas of consensus, very strong evidence indicates that, regardless of the therapeutic approach or the therapist’s level of experience, most individual psychotherapies end with the borderline patient’s dropping out. This is usually a result of the patient’s sense of being misunderstood or mistreated but is also often caused by antagonism toward therapy on the part of the patient’s significant others.
Based on the literature and our clinical experience, the following five simple principles are useful:
1. Therapists should identify, confront, and treat a comorbid substance abuse disorder or clear-cut major depression. Follow-up studies suggest that resolution of a comorbid substance abuse disorder may greatly improve the course of the BPD.
2. Clinicians need to develop a means for differentiating nonlethally motivated self-harm from true suicidal intention. Studies have shown that the lifetime risk of suicide in borderline patients is quite high (about 10%), and for the subsample with major affective disorder and untreated alcoholism, this rate may be considerably higher. Most of the time, however, the borderline patient’s thoughts or threats of suicide do not result in attempts; if they do, the attempts are not potentially lethal.
3. The therapist should stress that psychotherapy is a collaborative enterprise, while establishing from the outset that safety is an important issue. It must be made clear to the patient that the therapist is neither omnipotent nor omniscient.
4. The therapist must manage countertransference, a major component of therapeutic intervention in individual psychotherapies (supportive or exploratory) with borderline patients. Gabbard and Willkinson proposed that the source of countertransference feelings is the patient’s projection of unwanted aspects of the self onto the therapist. When the therapist is able to hold, contain, and “clarify” the patient’s projections, rather than responding in ways that activate old forms of interaction, the patient’s self becomes transformed by the corrective effect of the new interaction in the therapy relationship. At its most extreme, attention to and containment of countertransference hatred can safeguard against suicidality.
5. As implied by all of the above principles, therapists should set a low threshold for seeking consultation.
Because of the documented capacity of borderline patients to regress in unstructured therapies, it has generally been accepted that formal psychoanalysis is contraindicated for most of these patients. Some case reports of psychoanalysis have been published, but the patients in these samples would not be considered borderline by current criteria. The reason for this relative contraindication is the proclivity for psychotic or regressive transferences and uncontrolled acting-out in an unstructured treatment like psychoanalysis. For the occasional exception to this rule, careful consideration and expert consultation should be prerequisite.
Controversy exists within the domain of psychodynamic psychotherapy about the role of early interpretation and the management of negative transference. Kernberg was most articulate in identifying the need for early confrontation and interpretations of primitive transferences in here-and-now situations. Kernberg in particular emphasized the need to identify the aggressive motives that exist in the here-and-now so as to make their inappropriateness visible and dystonic. At times, this involves drawing patients’ attention to the sadistic and controlling motives behind their behaviors. Because interpretations are often transformed or experienced as attacks, linking statements that anticipate such reactions are often needed. The approach just outlined emerges from a two-person psychology in which one person is expected to be healthy and capable of neutralizing the transference and the other is expected to be ill and to form a transference neurosis that can respond to interpretation. Some clinicians believe that interpretations can be useful even early in therapies, but they need to be used carefully, awaiting those periods in sessions when an alliance with the patient is formed.
Buie and Adler and Chessick, following developments in self psychology by Kohut, anticipated in many ways the paradigm shift in therapeutic models for borderline patients suggested by the trauma data. More specifically, they argued that interpretations of aggressive themes are at best ineffectual and at worst harmfully disruptive in the early phase of treatment. They emphasized the need to validate the real role of bad parenting in the patient’s past as a motivating force for the patient’s aggression. These authors advocated reserving early transference interpretation to the development of intolerance-of-aloneness themes. Observations involve the need to reserve working through some of the borderline patient’s central issues until later phases of therapy, when a secure alliance has been established and the patient has a stable living and working situation. Included among these issues are the interpretations of negative transference and the recall of and reaction to memories of childhood abuse.
The effectiveness of this more empathic supportive psychodynamic approach has gathered impressive empirical support. Independent studies by Wallerstein, Gabbard et al., and Waldinger and Gunderson all testified to the importance of the supportive relationship-building aspects within dynamic psychotherapies with borderline patients. Most significant is a study from Australia by Stevenson and Meares in which the outcomes of 30 patients who received an Adler/Kohut type of psychodynamic psychotherapy were prospectively examined. The patients were seen for 1 year in twice-weekly sessions by closely supervised trainees. After 12 months of therapy and on extended follow-ups, the patients had significantly reduced impulsivity, affective instability, anger, and suicidal behavior. Patients were making fewer office visits and were engaging in fewer self-harm behaviors and less drug abuse. Thirty percent (9 of 30) of the patients no longer had symptoms that met criteria for BPD. Of note, only 16% of the BPD patients dropped out of this therapy. The effectiveness of this therapy was given further support when the outcomes were compared with those of 30 waiting-list control borderline patients who received treatment as usual. These results are very impressive, but the study’s methodological limitations preclude strong conclusions about efficacy.
The evidence of frequent childhood sexual and physical abuse in patients with BPD also underscores the importance of helping the patient form a secure attachment and a reasonably trusting alliance. In so doing, therapists often need to acknowledge and empathize with the patient’s experience of being victimized and reframe the patient’s anger and manipulative behaviors as understandable adaptations.
Those who have written about psychoanalytic psychotherapy with borderline patients agree that basic personality change is a legitimate goal. They also agree that such changes require that clinicians with psychoanalytic training or supervision and with experience and interest in BPD conduct two or more therapy sessions per week for years. Although the results of such therapies are not well documented, a detailed exposition of five successfully completed therapies that lasted about 5 years has validated the potential for basic changes. That study showed that long-term treatments can enable patients to assume independent functioning without ongoing therapies, even though problems in identity formation and self-esteem persist.
Intensive psychoanalytic psychotherapies constitute only a small fraction of the treatment actually given to borderline patients. Most borderline patients are seen in supportive psychotherapies, usually once a week or less, in which the primary focus is on the reality problems of daily life, and the patient and therapist have relatively little opportunity to examine and work through developmental issues or transferences. Nevertheless, within this less intensive form of psychotherapy, the same demands for saving interventions and the same accusations of cruel withholding are predictable strains on therapists. Psychotherapy is expected to last for a long time, sometimes serving maintenance functions and sometimes tapering off into an as-needed schedule after 3-5 years. The goals of this form of psychotherapy are directed at improving the patient’s adaptation to his or her life circumstances and diminishing the likelihood of self-destructive responses to expectable interpersonal frustrations.
Under the broad rubric of supportive therapy, Dawson and MacMillan proposed a model of psychotherapeutic “relationship management.” Relationship management psychotherapy is based on the principle of “first do no harm.” This therapy proposes that the self system of borderline patients is overly context bound. As a result of the patient’s lack of a consistent self system, relationships are a constant attempt to resolve self-attributed polarities such as good versus bad and competent versus incompetent. Patients resolve the ambiguity about themselves by eliciting responses from the therapist that will confirm a clearer and more stable identity for themselves; for example, by evoking directive “doctorly” responses from a therapist, the patient can assume the role and identity of being incompetent. A major task in this therapy is for the therapist to avoid the distortion of roles that the patient will try to impose. The therapist is advised always to focus on the process, rather than the content, in the therapy session.
Clinics with training functions or limitations on available reimbursement often force time-limited (10-20 sessions) strategies for borderline patients. These time-limited psychotherapies are focused on specific situational or interactional problems. This sort of strategy (i.e., one with specific time limits and focused subjects) may be particularly well suited for borderline patients who have a history of dropping out of more ambitious treatments and for those who present with concerns about being engulfed and overwhelmed or becoming too dependent. Some patients move from short-term treatment into a long-term therapy for which they were initially unmotivated. Silver has written of short-term, intermittent therapy offered on a long-term basis. Such clinical experience suggests that there may be a role for time-limited psychotherapies.
Two patterns of family involvement can help clinicians plan family interventions: overinvolvement and neglect. Borderline patients who are from overinvolved families are often actively struggling with a dependency issue by denial or by anger at their parents. Whether denied or reviled, these dependency needs are often being actively gratified by the family.
Such families benefit from active participation in treatment, and recent evidence suggests that their borderline offspring may have a particularly good prognosis. To exclude the family from involvement in the index borderline person’s treatment leads the parents to withhold support and, moreover, causes the patient to feel as though participation in therapy is disloyal to the parents and will lead to abandonment. Sometimes family sessions clarify that the parents’ alleged resistance to separation represents projections by borderline patients behind which are their own fears of separation.
Even in families that have been neglectful, dysfunctional, or abusive, therapists should appreciate the strength of the persisting attachments. In these families, conjoint sessions should be approached slowly. Childhood histories of abuse in patients with BPD often cause them to fear that their families, whom they view as malevolent and powerful, will convince the treaters that they are “bad” or “crazy.” In such instances, helping patients maintain some distance from abusive families is an important step in their feeling safe. However, this may encourage an idealized view of therapy or of the therapist as offering an asylum from or an alternative to the patient’s family. A supportive and educational approach with family members often is needed to ease their wariness and begin the process of developing an alliance. On occasion this will give way to the disclosure of family problems that are initially denied, sometimes even family secrets involving knowledge of or participation in abuse. When the family’s acknowledgment is associated with shame or regret, the patient’s alienation may diminish significantly. In any event, the primary task of family meetings is support of the borderline patient’s treatment. In many instances, parental meetings must take place with a family therapist other than the patient’s therapist. The two therapists in such cases should be in communication.
Interest in the use of psychoeducation and skills training approaches for families with borderline members is growing. These approaches use multiple family groups in ways that offer cross-family support and education and are very cost beneficial. Preliminary evidence indicates very positive responses from participants and both better communication and less alienation. These approaches are often very welcome to families who feel supported and respected. Individual psychodynamic therapies usually should be reserved for those families who have learned skills of listening and communicating without hostility.
Group therapies appear useful for most borderline patients. Although these patients have difficulty entering and remaining in such treatment, the chances of doing so are enhanced if participation is strongly urged from the beginning by concurrent individual therapists. Concurrent group and individual therapies offer a holding environment that permits the patient to be angry (have a “bad object”) without needing to leave treatment.
The presence of peers in a therapy group has several benefits not available in individual therapy. Peers are more able than a therapist to confront maladaptive and impulsive patterns without being dismissed as trying to control the patient. Groups are also very effective in identifying dependent needs or manipulative behaviors and making them more syntonic and dystonic, respectively. At the same time, the group provides a set of peers with whom feelings and personal problems can be communicated without harmful repercussions. Groups provide a field to study others’ methods of coping, and borderline patients often find that it is easier to identify a maladaptive pattern of coping in another person than in themselves. Groups also provide a social networking function that may extend to the development of new and better relationships outside of, as well as within, the therapy. Empirical investigations have begun to support the value of group therapy for borderline patients. Marziali and Munroe-Blum studied a 30-session interpersonal group therapy approach and compared it with individual psychotherapy. Although those borderline patients who remained in either treatment had better outcomes (especially on suicidality and hospitalizations) than those who dropped out, the group therapy emerged as equally effective and far more economical than the individual therapy.
Budman also tested interpersonal group therapy for patients with either borderline, avoidant, or obsessive-compulsive personality disorders. He concluded that borderline patients could make significant and cost-effective improvements in their interpersonal and social functioning from this therapy. He also concluded, however, that the presence of too many borderline patients in any group was harmful for other, nonborderline participants. Both Budman and Marziali and Munroe-Blum noted that group therapy is less burdensome to the therapist than individual therapy and that the support of a cotherapist and ongoing supervision are often useful.
Group therapy, usually with concurrent individual therapy, appears to be useful for borderline patients. Specialized approaches involving a more intensive level of outpatient or community-based care are being developed and need to be empirically tested.
Linehan developed dialectical behavior therapy (DBT) specifically for patients with BPD. The empirical support for its efficacy and the clarity of its implementation have inspired the widespread use of this modality. Linehan hypothesized that the primary impairment in patients with BPD is a constitutional dysregulation of emotion interacting with an invalidating environment. The DBT she developed includes manualized individual and group therapy components. The group therapy focuses on teaching patients behavioral coping skills, whereas the individual therapy focuses on six goals for change that are arranged with hierarchical priority:
1. Suicidal behaviors
2. Behaviors that interfere with therapy
3. Behaviors that interfere with quality of life
4. Behavioral skill acquisition
5. Posttraumatic stress behavior
6. Self-respect behaviors
This approach is unique because the individual therapist focuses on all suicidal behaviors as a priority. Individual therapists are on call for crises, which are used to coach patients in use of skills. In contrast, other therapies have tended to view suicidality and self-destructiveness as symptoms that will subside after underlying problems resolve. An important aspect of DBT is its use of a consultation group. The group’s co-therapists and the individual therapist meet weekly to help one another maintain a balance between acceptance and a call for change.
The results from a randomized controlled trial comparing 23 patients who received DBT for 1 year with 21 patients who received treatment as usual are impressive. Patients receiving DBT had a significantly lower attrition rate during the year of the trial than did the group receiving treatment as usual (16.7% vs. 58.3%). DBT patients also had significantly fewer and less severe episodes of parasuicidal behavior and fewer days of inpatient hospitalization than did the group receiving treatment as usual. The two conditions, however, did not differ on measures of depression, hopelessness, suicidal ideation, or reason for living at completion of the trial.
The approach developed by Linehan testifies to the importance of making adaptations to usual behavior therapies. She encourages flexibility and, like dynamic therapies of the Adler/Kohut type, emphasizes the role of validation and empathy. Much of the effort is directed at developing and maintaining the therapeutic relationship. She indicates that for patients with BPD, the relationship is the means to provide therapy, but the relationship also is the therapy. Insofar as the latter is true, the distinctions between the therapeutic action of DBT and dynamic therapy become blurred.
Cognitive-behavioral therapies for patients with BPD are beginning to be developed. The cognitive therapy approach is characterized by Beck and Freeman and J. E. Young. These clinicians postulate that borderline patients have disturbed cognitions that develop early in their lives, have maladaptive consequences, are self-perpetuating, and are (albeit difficult to change) the targets for cognitive therapies. The usual strategies of cognitive or cognitive-behavioral therapy must be modified for borderline patients. Needed modifications include the use of greater flexibility in the therapy and greater attention to the creation and maintenance of a therapeutic relationship.
The specific maladaptive cognitive schemas proposed by J. E. Young for BPD are abandonment and loss, unlovability, dependence, subjugation, lack of identification, mistrust, inadequate self-discipline, fear of losing emotional control, guilt and punishment, and emotional deprivation. Beck offered a different, but overlapping, dissection of the borderline patient’s disturbed cognitions. Beck and Freeman stressed three “basic assumptions” held by borderline patients:
1. “The world is dangerous and malevolent.”
2. “I am powerless and vulnerable.”
3. “I am inherently unacceptable.”
According to Beck, dichotomous thinking is particularly common and problematic in borderline patients, who tend to evaluate experience in terms of mutually exclusive categories rather than seeing it as part of a continuum. Beck postulated that the extreme evaluation of the situation led to the extreme emotional responses and actions of these patients. The extreme evaluation was also accompanied by a rapid shift to the opposite view. Beck felt that the combination of the basic assumptions, dichotomous thinking, and an unstable sense of identity were the particularly problematic cognitive features of the borderline patient. Beck suggested that reducing or eliminating dichotomous thinking is an early goal of therapy, but that part of the therapist’s attention needs to be invested in establishing a trusting and collaborative working relationship.
Although cognitive approaches offer promising means of operationalizing interventions that allow for research testing, little has been done to date. Tyrer and Davidson describe promising results from a brief six-session manual-assisted cognitive therapy directed at reducing intentional self-harm behaviors. The approach is educational and directive. Eighteen borderline patients randomly assigned to manual-assisted cognitive therapy had a reduced frequency of self-harm and depression compared with 14 patients who received usual treatments.
Revision date: July 8, 2011
Last revised: by David A. Scott, M.D.