Dialectical Behavior Therapy for Personality Disorders

Introduction

Dialectical behavior therapy (DBT) was originally developed by Linehan [1–3] to address the treatment needs of actively suicidal individuals. It was then further refined to treat suicidal individuals with borderline personality disorder (BPD). Over the past few years, a number of adaptations have been made to DBT to enhance its effectiveness with other clinical disorders, including substance use disorders comorbid with BPD, eating disorders, and antisocial behavior. This paper first gives a brief overview of standard DBT treatment and then focuses primarily on research over the past 2 years on the application of DBT in the treatment of personality disorders. For a more extensive review of other empiric research on DBT, interested readers are referred to reviews by Koerner and Dimeff [4], and Koerner and Linehan [5].

Overview of dialectical behavior therapy

Dialectical behavior therapy combines change strategies from cognitive and behavioral therapies with acceptance strategies adapted from Zen teaching and practice; it is a synthesis of both validation and acceptance of the patient, on the one hand, with persistent attention to behavioral change on the other. The change procedures consist of systematic and repeated behavioral analyses of dysfunctional response chains, training in behavioral skills, contingency management to weaken or suppress disordered responses and strengthen skillful responses, cognitive restructuring, and exposure-based strategies aimed at blocking avoidance and reducing maladaptive emotions. The acceptance procedures consist of mindfulness (eg, attention to the present moment, assuming a nonjudgmental stance, and focusing on effectiveness) and a variety of validation and stylistic strategies [6]. In its standard form, the treatment is provided in once-weekly individual psychotherapy and group skills training sessions, skills coaching phone calls with the primary therapist (when needed), and weekly team meetings of all DBT therapists aimed at reducing therapist burn-out and increasing therapists’ adherence to the treatment model and competence in treating these patients. Individual sessions are based on clearly prioritized targets and focus on reducing maladaptive response patterns (life threatening behaviors, behavioral patterns that interfere with or threaten therapy, severe Axis I disorders, and patterns that preclude a reasonable quality of life) and enhancing motivation for skillful behaviors. The foci of specific sessions are determined by the patient’s behavior and problems since the previous session. Skills training is psychoeducational and teaches mindfulness, distress tolerance, emotion regulation, interpersonal effectiveness, and self-management skills.

Data on this standard model indicate that DBT is more effective than treatment-as-usual (TAU) in the community in terms of reductions in parasuicidal behavior, completion of treatment, and fewer inpatient hospital days [7]. Since this first controlled trial, recent replication studies of standard DBT to treat BPD, as well as research with adaptations of the standard model, have been conducted. These are summarized below and outlined in Table 1.

Recent research studies

Dialectical behavior therapy for borderline personality disorder: 6-month treatment programs

The original clinical trial of Linehan et al. [7] provided one year of treatment. Because most of the treatment gains in DBT were made in the first 4 months of treatment, it is reasonable to ask if a shorter length of treatment would be effective. Two recent studies have examined six-month treatment programs. Koons et al. [8] randomly assigned 20 women meeting criteria for BPD to either standard DBT or to standard outpatient psychotherapy administered in the Veterans Administration psychiatric clinics. The control treatment in this study was quite intensive: all women received weekly individual therapy and there was an optional supportive, psychoeducational group. Individuals assigned to DBT and to TAU received equivalent amounts of individual therapy hours. Findings indicate that DBT was more effective at lowering suicidal ideation, depression, hopelessness, and expression of anger compared with TAU. There were no significant differences in reduction of parasuicidal acts (both suicide attempts and nonsuicidal self-injury) between the two groups, although this may be attributable to the low base rate for this behavior; subjects were not required to have recent parasuicides to be included in the study.

Stanley et al. [9] conducted a nonrandomized pilot study comparing 6 months of standard DBT with TAU (standard psychiatric treatment in the community with no restrictions) for 30 women meeting criteria for BPD [9]. Similar to the Koons et al. [8] study, DBT was more effective than TAU in lowering suicide ideation. DBT was also superior to TAU in reducing the frequency of parasuicide acts as well as urges both to commit suicide and to self-mutilate. In contrast to Linehan et al. [7], who found greater decreases in suicide attempts in DBT versus TAU, Stanley et al. [9] found decreases in nonsuicidal parasuicide (self-mutilation) only. There were no suicide attempts in either condition during the course of the study. Taken together, these studies have important implications in two areas: 1) results suggest that support for DBT can be replicated by research teams other than the team of the treatment developer, and 2) they suggest that DBT has potential effectiveness as a shorter treatment, at least when patients are not selected for being highly suicidal at the start of treatment. Lack of follow-up data for either of the 6-month studies, however, warrants caution in interpreting these findings.

Inpatient dialectical behavior therapy for borderline personality disorder

In Germany, Bohus et al. [10] have investigated DBT for inpatient settings under the hypothesis that treatment for BPD can be accelerated by an initial course of intensive inpatient treatment as a precursor to standard outpatient treatment. In this uncontrolled trial, 24 suicidal women meeting criteria for BPD were assessed at pretreatment and after 3 months of inpatient DBT. DBT in this trial consisted of three stages. The first stage covered an analysis of the targeted behaviors (life-threatening behaviors, behaviors interfering with treatment, and other severe problematic response patterns), including their antecedents and consequences. The second, and the longest, stage covered DBT behavioral skill acquisition and contingency management of problematic behavior. The third stage covered preparation towards discharge, which included attention to social supports outside the hospital and referral to outpatient DBT. Findings from this innovative study showed a significant decrease in depression, anxiety, and frequency of parasuicide at the end of the 3-month inpatient treatment. A randomized trial is currently underway to determine if this 3-month phase of intensive treatment does indeed aid in the effectiveness of standard DBT immediately following it.

Manual-assisted treatment for parasuicide and cluster B disturbance

A group in the United Kingdom has recently developed a manual-assisted cognitive-behavioral therapy (MACT) [11]. This treatment, in the form of six booklets, includes elements of DBT, including chain analyses for maladaptive behavior such as parasuicide, basic cognitive techniques for handling negative emotions, and distress tolerance strategies. This treatment was compared with TAU, which was standard psychiatric treatment for parasuicidal behavior with no restrictions. Thirty-four individuals took part in the study, with 18 assigned to the MACT group and 16 assigned to the TAU group. Although the main focus of the treatment was to reduce parasuicide episodes, no significant between-condition differences in frequency were found. MACT was superior to TAU, however, in reducing the severity of parasuicide episodes, the number of hospital admissions following subsequent parasuicide episodes, and in reducing depression. Although this was not a implementation of standard DBT, the study suggests that components of DBT strategies may be easily disseminated and understood without intensive treatment.

Treatment of substance dependence in individuals with borderline personality disorder

Linehan et al. [12] adapted standard DBT to better fit the treatment needs of women with a dual diagnosis of BPD and substance dependence. Modifications include alteration of the target hierarchy to specifically include substance use and behaviors functionally related to substance use, addition of a set of “attachment” strategies to keep patients engaged in therapy, the tapered use of replacement medication (ie, methylphenidate or methadone) for individuals with stimulant or opiate dependence, and some minor alterations in the skills being taught to better fit the problems associated with drug abuse. Twenty-eight women meeting criteria for BPD and current substance dependence were randomly assigned to 1 year of DBT or TAU in the community. TAU consisted of alternative substance abuse and mental health counseling or drug abuse programs in the community. Results indicated the superiority of DBT on several outcome measures, including higher proportions of days abstinent for those assigned to the DBT condition over the whole treatment year as measured by structured interviews as well as urinalyses, higher treatment retention, and greater social and global adjustment at a follow-up assessment at 16 months. All subjects decreased their frequency of parasuicide and reduced state and trait anger; there were no between condition differences. An interesting post-hoc analysis among subjects assigned to DBT, indicated that patients treated by therapists rated as adherent to DBT had a higher proportion of clean urinalyses at the end of treatment than patients seeing therapists that were less consistently adherent to DBT. This finding suggests that training, adherence to DBT, and therapist competence may be important predictors of outcome and should not be overlooked in future studies.

Treatment of antisocial behaviors

Dialectical behavior therapy has also been applied in forensic settings. The appeal of DBT for this population can be attributed to its focus on life-threatening behaviors (both suicidal and homicidal), its focus on behaviors that interfere with a treatment program, or with a therapist’s motivation and willingness to treat a patient, and the high prevalence of personality disorders, especially antisocial personality disorder and BPD, within this population [13,14]. McCann et al. [14] adapted DBT for forensic inpatient units with a violent and severely mentally ill population. Their modifications to the standard DBT protocol include the following: expansion of the category of life-threatening behaviors to include those specifically related to homicide and interpersonal violence, specific attention to the reinforcement of honest recording of maladaptive behaviors, testing skills acquisition with exams and role-play quizzes, targeting emotional insensitivity of antisocial patients, and inclusion of a “graduate” level skills group following a standard completion of DBT skills training. This graduate level group entitled DBT Graduates’ Crime Group is designed to increase empathy for victims and prevent relapse of violent behavior. What is perhaps most unusual about these modifications is that the inpatients with antisocial personality disorder themselves contributed most to the modifications.

A 20-month nonrandomized pilot study was conducted with 35 patients comparing the modified DBT with TAU within the forensic inpatient ward. Approximately 50% of the patients met full criteria for BPD, the other 50% each had at least three BPD symptoms. In addition, one third of the participants met criteria for antisocial personality disorder. Twenty-one of the patients received DBT and 14 received TAU. As compared with TAU, the DBT group had significantly decreased levels of maladaptive coping and significantly increased levels of adaptive coping over the 20-month period. Moreover, the DBT group showed a trend toward significant decreases in depression, hostility, paranoia, and psychotic behaviors. This study highlights the potential utility of the use of DBT with a broader spectrum of personality-disordered individuals as well as its utility in more diverse settings, such as a forensic inpatient ward.

Similarly, Trupin et al. [15] have used DBT in the treatment of incarcerated female juvenile offenders. Participants had an average of six prior offenses. Information on the prevalence of personality disorders was not reported. DBT was implemented in two mental health cottages at a juvenile detention center. The staff on one DBT mental health cottage (DBT1) received extensive DBT training (80 hours), whereas staff in the other mental health cottage (DBT2) received an abbreviated training of 16 hours. The two cottages were also significantly different at the beginning of the study; residents of the DBT1 cottage exhibited more severe mood and thought disturbances as well as a higher frequency of parasuicide episodes. Key modifications to standard DBT included the addition of a self-management module to skills training, specifying “unit-destructive behaviors” and targeting them after life-threatening behaviors, and focusing special attention on offense related behaviors. Over the 10-month study duration, juveniles in the DBT1 cottage had a significant reduction in problem behaviors whereas no significant changes were found in the DBT2 cottage. In addition, staff in the DBT2 cottage actually showed a significant increase in punitive actions toward the adolescent girls over the course of the study, a result not found in the DBT1 cottage. Although the mixed results of this quasi-experimental study are difficult to interpret unambiguously, they highlight the need for further research on the issue of training in DBT.

Is the use of dialectical behavior therapy in the community ahead of its time?

Both Scheel [16•] and Westen [17] have suggested that DBT has been disseminated into the community more widely and faster than empiric support warrants. The rapid spread of DBT certainly cannot be accounted for solely by the few studies showing treatment efficacy. Swenson [18•] articulates reasons why this may have occurred. For one, DBT is used to treat a population of individuals with severe pain and dyscontrol. Thus, any treatment that appears effective, as well as cost-effective, is accepted wholeheartedly. Second, Swenson [18•] argues that DBT addresses the issue of therapist support. Clinicians who deal with BPD patients experience significant stress and a therapy that includes a therapist consultation group as a mandatory component is incredibly appealing. Third, quite simply, there is a lack of other options available to practitioners dealing with BPD patients. Since the first controlled trial of DBT [7], only one other randomized trial of a psychosocial treatment for BPD has been published [19]. Bateman et al. [19] have developed an 18-month, partial hospitalization treatment program, based on object-relations theory, and demonstrated efficacy in a randomized trial comparing it to a community control condition consisting of a community treatment that prohibited individual therapy but provided all other community supports. The widespread attention to DBT is due in part to the scarcity of other cost-effective, efficacious treatments available to clinicians dealing with highly difficult-to-treat populations.

Any new treatment shown efficacious in clinical trials must stand up to rigorous examination of grounds for empiric support. With respect to DBT, it is particularly important to ask whether findings to date demonstrating DBTs efficacy might be due to several non-DBT specific factors such as allegiance factors, the provision of stable individual therapy, rather than DBT per se, the high amount of structure provided, and the training involved in behavioral treatment, in general, and DBT, in particular. Each of these will be addressed in turn.

Is DBT only effective implemented at the University of Washington or by Linehan [1–3]? Replication by other investigators is, of course, of the essence in science. As illustrated in this paper as well as other reviews [4,5], Linehan’s findings have been replicated by other research groups. The high risk of the patient population and length of treatment that must be offered together with the time needed for researchers to be trained in a new treatment technique has limited the speed of replications. Interestingly, an argument against Linehan’s results being due simply to allegiance can be found in her own research. In an earlier study described in the DBT treatment manual [2], Linehan et al. tested their hypothesis that adding DBT skills training to stable, individual therapy in the community would greatly enhance the outcomes of those treatments. Linehan was one of the principle skills trainers in the study. Results did not support the hypothesis, at least as tested by Linehan, suggesting that allegiance may not be sufficient by itself to produce positive outcomes with highly suicidal BPD patients.

If DBT is efficacious, why is it so? Is it simply that DBT is particularly good at insuring that patients receive stable, individual therapy of any sort? Although stability in the treatment of BPD is not to be underrated, a few studies have tentatively discounted this claim. Koons et al. [8] provided individual therapy to all participants in their study and still found favorable results for DBT. In the Linehan et al. [7] study, not only did patients assigned to DBT and TAU have equivalent number of therapeutic hours, but the number of treatment hours received did not predict outcome. Similarly, one could argue that DBT is effective because it provides a highly structured and stable environment for disordered individuals. This is a reasonable hypothesis; however, even inpatient settings such as described in the Trupin et al. [15] and McCann et al. [13,14] studies, which theoretically have a lot of structure for all individuals, found some support for DBT. More research is needed to test this hypothesis further by comparing DBT with a more tightly controlled alternative treatment with a similar amount of structure and stability.

Finally, it can be argued that positive results have been found for DBT because, as opposed to TAU in the community, DBT clinicians have had superior training and supervision. Again, this is a reasonable hypothesis that cannot be discounted. In fact, one major feature that distinguishes between treatment research conducted within academia and TAU in nonacademic settings is training of therapists and monitoring of the treatment delivered. However, to discount DBT for this reason is to discount most psychosocial treatments that have been shown to be effective for a whole host of mental disorders. Perhaps what needs to occur is not the “naturalizing” of DBT but the improvement of training, supervision and monitoring in community therapeutic settings.

Conclusions

Interest in DBT has peaked in recent years despite only relatively few studies providing empiric support for its superiority to other forms of treatment. In this paper, we have provided information on seven research studies. Although the level of empiric support for DBT is still inadequate, each of these studies offers some further support for DBT as a efficacious treatment for personality disorders, specifically BPD and antisocial characteristics. Many of these treatments used adaptations of DBT, which further illustrates its generalizability across settings and populations. Although these studies all provide support for the idea that DBT is effective, we have yet to learn why DBT works. The next generation of research studies must address this question. DBT is a compilation of current behavioral treatments combined with extensive validation and acceptance of the patient packaged within a format that provides clear guidelines and expectations, structure and support for both patients and treatment providers. Although the guiding premise has been that both explicit behavioral change strategies and equally explicit acceptance strategies are both equally important in the treatment, this is an empirical question yet to be answered. It is also unclear whether the DBT skills training component is essential to efficacy. DBT is defined as a community of therapists treating a community of patients. The guidelines for how the treatment team members are to interact with each other forms a part of the treatment itself. Is the focus on the community of therapists and the requirement of team application of the treatment really essential? Are the proscribed guidelines an important component of treatment effectiveness? Common sense suggests that a treatment program that reduces the sense of burnout of therapists, provides emotional support and consultation to increase competence, and rigorously requires a compassionate and nonjudgmental attitude among the treatment providers would enhance efficacy. Treatment providers regularly report that this is the case. As is always the case with new treatments and new lines of research, many more studies are needed.