Implications for the Treatment of Substance Abuse

By Marsha M. Linehan

The purpose of this chapter is to describe a behavioral treatment approach designed specifically for chronically parasuicidal individuals meeting criteria for borderline personality disorder (BPD).

Both the overlap between substance abuse and BPD as well as common correlates between the two disorders suggest that Dialectical Behavior Therapy (DBT), in whole or in part, might be effective as a treatment for substance abuse in general and for substance abusers who also meet criteria for BPD in particular.

The chapter will first give a brief overview of the overlap between substance abuse and BPD. Second, it will describe the central elements of a broad-band, behaviorally based treatment (DBT) for BPD.

Finally, it will briefly describe the empirical evaluations of the effectiveness of this treatment regime.

BORDERLINE PERSONALITY DISORDER AND SUBSTANCE ABUSE

Overlap

Impulsiveness in areas that are potentially self-damaging is part of the criteria for BPD. Substance abuse counts as half of this criterion in both DSM-III-R (American Psychiatric Association 1987) and DSM-IV (American Psychiatric Association Task Force on DSM-IV 1991).

Thus, it is not surprising that, compared with individuals with all other personality disorders (except antisocial disorder) and with psychiatric patients with no personality disorders, individuals meeting criteria for BPD score higher on substance abuse scales (Pitts et al. 1985; McCann  et al. 1992), more commonly report a history of substance abuse (Akiskal et al. 1985), and also meet criteria for current substance abuse (Loranger and Tulis 1985; Zanarini et al. 1989; Dulit et al. 1990; Koenigsberg et al. 1985).

In the study conducted by Zanarini and colleagues (1989), 84 percent of 50 borderline outpatients had met criteria for substance abuse/dependence at some point in their lives. In the study conducted by Dulit and coworkers, 67 percent of BPD patients met criteria for substance abuse disorder.

When substance abuse was not used as a criteria for BPD, the incidence dropped to 57 percent, still a significant portion of the population. Koenigsberg and colleagues (1985), reporting on a sample of 2,462 inpatients and outpatients, found that 21 percent of borderline patients had a primary Axis I diagnosis of substance abuse.

Furthermore, two studies suggest that BPD individuals are more likely to be polydrug abusers, usually combining drug and alcohol abuse, than are non-BPD individuals (Kosten et al. 1989; Nace et al. 1983). The high incidence of substance abuse among borderline individuals suggests that, at least some of the time, patterns diagnosed as indicative of BPD may be better viewed as sequelae of substance abuse patterns.

Indeed, Zweben and Clark (1990-1991) argue that the overlap between substance abuse and BPD may be a result of correlates and consequences of substance abuse that masquerade as BPD symptoms.

For example, they note that instability of mood, inappropriate expression of anger, chaotic interpersonal relationships, impulsiveness, and persistent feelings of emptiness or boredom (all criteria for BPD) are characteristic patterns found among serious substance abusers and may clear up with sufficient abstinence.

However, although the BPD-substance abuse overlap is substantial, not all substance abusers meet criteria for BPD. In a study of 64 female psychiatric outpatients meeting criteria for substance abuse, Vaglum and Vaglum (1985) found that 66 percent also met criteria for BPD.

In a more restricted sample, Nace and colleagues (1983) found that 13 percent of 94 consecutive admissions to an alcohol treatment program also met criteria for BPD. Tousignant and Kovess (1989) found that one-third of substance abusers showed a high number of BPD traits.

BPD substance abusers, however, are uniformly more disturbed than those abusers who do not meet criteria for BPD. Studies have shown that they are more commonly comorbid for depressive disorders, have more frequent suicide attempts and accidents, and score higher on  impulse dyscontrol and antisocial tendencies and lower on reality testing (Kosten et al. 1989; Inman et al. 1985).

Of particular interest for treatment is the overlap between suicidal behaviors and both substance abuse and BPD. Suicidal behaviors, including threats, “gestures,” and suicide attempts, represent one of the eight (or nine in DSM-IV) criteria for BPD. Parasuicidal behavior (all acute, intentional, nonfatal, self-injurious behavior, including suicide attempts) has been called the behavioral specialty of BPD (Gunderson 1984).

Although much of this behavior is without lethal intent, the percentage of those followed after an index treatment who eventually die by suicide is at least 5 to 10 percent (Stone et al. 1987; Paris et al. 1987; Frances et al. 1986).

Similarly, substance abuse is associated with an increased risk of suicidal behaviors, including both completed suicides and parasuicide (see Lester 1992). Roy and Linnoila (1986) estimated that 18 percent of alcoholics subsequently complete suicide. What is not clear at this point is whether the high rate of suicidal behaviors among substance abusers is mediated by the concomitant BPD associated with the substance use disorders.

Achieving treatment success with both BPD and substance abuse has been notoriously difficult. Although there have been few randomized controlled trials investigating treatment for BPD, followup studies of individuals who have received substantial inpatient and outpatient psychiatric care suggest that current treatments are marginally effective at best when measured 2 or 3 years following treatment (Perry and Cooper 1985; Tucker et al. 1987).

In contrast, there have been many controlled trials of treatments for both alcohol and drug abuse. Although efficacy of treatments has been demonstrated, lasting positive outcomes have been difficult to achieve (Marlatt and Gordon 1985; Miller and Rollnick 1991).

Substance abuse combined with BPD may be particularly difficult to treat. Nace and colleagues (1985) reported that at several years following treatment for substance abuse, these individuals have more severe problems remaining than do non-BPD substance abusers. 

OVERVIEW OF DBT

DBT was developed from a combined motivational and capability deficit model of BPD. The idea was twofold: (1) borderline individuals lack important interpersonal, self-regulation (including emotional regulation), and distress tolerance skills, and (2) personal and environmental factors inhibit the use of behavioral skills that the individual does have and often reinforce inappropriate borderline behaviors.

The emphasis on capability enhancement is similar to that in substance abuse treatment models that stress relapse prevention (Marlatt and Gordon 1985). The emphasis on changing motivational factors is similar to the motivational interviewing proposed by Miller and Rollick (1991) and to aversive conditioning models (although DBT provides a greater emphasis on reinforcement than on punishment).

DBT presumes that attention to both skill acquisition and behavioral motivation is essential. In developing the treatment, however, it quickly became apparent that (1) skill training to the extent believed necessary is extraordinarily difficult, if not impossible, within the context of a therapy oriented to reducing the motivation to die or act in a borderline fashion, and (2) sufficient attention to motivational issues cannot be given in a treatment with the rigorous control of therapy agenda needed for skill training.

From this, the idea developed to split the therapy into three components, one that focuses primarily on skill acquisition, one that focuses primarily on motivational issues and skill strengthening, and one designed explicitly to foster generalization of skills to the everyday life outside the treatment context.

The three modes in standard outpatient DBT are psychosocial groups (for skill training), individual psychotherapy (addressing motivational issues and skill strengthening), and telephone contact with the individual therapist (addressing generalization).

Within each treatment mode, DBT is characterized by a philosophy of dialectics, a biosocial theoretical perspective, a hierarchy of treatment targets specific to the mode, and a set of treatment strategy groups. Space here is too brief to give a detailed description of each component of the treatment.

The interested reader is referred to the treatment manual and associated updates (Linehan, 1993a, 1993b). 

Theoretical Base

As one might suppose from the title of the treatment, DBT flows from a dialectical philosophical position. “Dialectic” is used here in two contexts, that of persuasive dialogue and relationship and that of the fundamental nature of reality.

From the point of view of dialogue and relationship, it refers to change by persuasion and by making use of the opposition inherent in the therapeutic relationship, rather than by formal impersonal logic.

Thus, unlike analytical thinking, dialectics is personal, taking into account and affecting the total person, and it does not seek absolute truth but instead attempts to facilitate the construction or evolution of truth over time.

As a world view, dialectics convey these coexisting multiple tensions that must be addressed within the therapeutic relationship, as well as the emphasis in DBT on (1) a systems perspective (asking always, “What is being left out of our understanding here?”), (2) searching for synthesis and balance (to replace the rigid, often extreme, and dichotomous response characteristics of suicidal and borderline patients), and (3) enhancing comfort with ambiguity and change, which are viewed as inevitable aspects of life.

The overriding dialectic for the therapist is the necessity of acceptance of the patient as he or she is within the context of simultaneously trying to produce change. Treatment strategies are polarized into those most related to acceptance and those most related to change, although it is this very polarization that is the mot of many therapeutic failures.

DBT requires that the therapist balance use of these two types of strategies within each treatment interaction. The author has proposed elsewhere (Linehan 1993a) that BPD is primarily a systemic disorder of emotion regulation.

Characteristics of this dysregulation include a high sensitivity to emotional stimuli, intense response to even low-level stimuli, and a slow return to baseline combined with an inability to modulate emotional states.

Borderline behavioral patterns either function to remediate negative emotional arousal directly (a view similar to one that sees substance abuse as self-medication) or indirectly (e.g., by eliciting help from the environment) or are inevitable outcomes of unregulated and unstable emotionality. In short, borderline patterns either are attempts to solve problem emotions or are problematic sequelae, either of the initial emotions or of the dysfunctional attempts to reduce emotionality.

The author has further hypothesized that this pattern of dysregulated  emotion and behavior is a result of an initial temperamental disposition to emotionality, and perhaps inadequate modulation, combined with an invalidating rearing environment.

Such an environment is characterized by a tendency to disregard emotional experiences, especially negative ones, and oversimplify the ease of solving difficult problems, and it puts a high value on positive thinking. Although such attitudes are certainly beneficial for some, if not most, this type of environment invalidates the experiences of vulnerable individuals and does not take seriously their communications, especially when such communications have to do with nonpublic events and with difficulties in meeting social expectations.

Invalidating environments, especially physically and sexually abusive families, contribute to the development of emotion dysregulation and fail to teach the child how to label and regulate arousal, how to tolerate emotional distress, and when to trust one’s own emotional responses as reflections of valid interpretations of events.

At the adult level, borderline individuals adopt the characteristics of the invalidating environment. Thus, they tend to invalidate their own affective experiences, look to others for accurate reflections of external reality, and oversimplify the ease of solving life’s problems.

This oversimplification leads inevitably to unrealistic goals, an inability to use reward instead of punishment for small steps towards final goals, and self-hate following failure to achieve these goals. The shame reaction, a characteristic response to uncontrollable and negative emotions among borderline individuals, is a natural result of a social environment that “shames” those who express emotional vulnerability.

These two polar extremes, vulnerability versus invalidation, represent the central dialectical dilemma of the borderline patient and therapist.

Treatment Targets

Treatment targets for individual DBT therapy and for DBT as a whole are the same and are hierarchically arranged as follows:

1. Reducing high-risk suicidal behaviors (parasuicide and high-risk suicide ideation and plans);

2. Reducing therapy-interfering behaviors-all responses or behaviors of both the patient and the therapist that make therapy progress or continuation difficult (e.g., missing or coming late to sessions, phoning at unreasonable hours, refusing to collaborate or work in sessions, remaining interpersonally aloof or too clinging, invalidating the other, and not returning phone calls);

3. Reducing behavioral patterns serious enough to substantially interfere with any chance of a reasonable quality of life (serious substance abuse would qualify here);

4. Behavioral skill acquisition (skills in emotion regulation, interpersonal effectiveness, distress tolerance, and selfmanagement, as well as a number of “core” [mindfulness] abilities to observe, describe, participate spontaneously, be nonjudgmental, focus awareness, and focus on effectiveness);

5. Reducing posttraumatic stress responses related to previous traumatic events;

6. Increasing self-respect; and

7. Meeting other goals of the patient.

With respect to each target, the task of the therapist is first (and many times thereafter) to elicit the patient’s collaboration in working on the target behavior, then to apply the relevant treatment strategies described below.

Attention to each target within individual therapy, ordinarily involving direct and focused work on the behaviors relevant to the target, is jointly determined by the hierarchy list above and by the behaviors and problems that have surfaced since the last session or during the current session. Thus, treatment is oriented to current behaviors.

Therapy is somewhat circular in that target focal points revolve over time. The hierarchy of targets (i.e., what is attended to) is somewhat different in group skills training and in phone calls. In skills training, as one might imagine, skills acquisition is the top priority.

The only behaviors that would take precedence are behaviors that threaten to destroy skills training (e.g., aggression toward other group members or not coming to sessions).

Behavioral skills are taught in modules concentrating on mindfulness skills (observation, description and spontaneous participation, nonjudgmentalness, focused attention, and  “doing what works”), interpersonal effectiveness for conflict situations, emotion regulation, and distress tolerance.

During telephone calls, generalization of skills is the top priority, preceded only by threats to the patient’s life. Thus, during phone calls the focus is always on, “What skills could you use here?”