by Michael D. Anestis, Ph.D.
A study published in the current issue of the Journal of Consulting and Clinical Psychology by Jamie Bedics of California Lutheran University and David Atkins, Katherine Comtois, and Marsha Linehan of the University of Washington caught my eye recently. The addressed some important issues, particularly the role of therapeutic alliance in treatment outcome for dialectical behavior therapy (DBT) relative to community treatment for borderline personality disorder (BPD). Importantly, treatment outcome involved not only symptoms (non-suicidal self-injury; NSSI) but also factors related to general intrapsychic change (introject).
Before getting into the study itself, I want to take a moment to highlight a point raised by the authors early in their manuscript. They noted that DBT has been shown to be superior to treatment as usual and treatment by expert clinicians with respect to suicide attempts, emergency/inpatient treatment, NSSI, anger, depression, and social and global adjustment (Kleim, Kroger, & Kosfelder, 2010; Lynch, Trost, Salsman, & Linehan, 2007). Those are, without question, important variables and I have argued many times that people who trivialize the importance of symptom reduction relative to broader variables like quality of life are missing the mark. I've also noted that DBT is, without question, my favorite form of treatment to administer as a clinician, so I have no desire to bash the treatment or its creators. That being said, given the extensive literature on the efficacy and effeciveness for DBT, particularly with respect to patients with BPD, I can't help but wonder why we don't see data on the imapct the treatment has on the specific targets of the treatment modules (mindfulness, interpersonal effectiveness, emotion regulation, and distress tolerance) or on the other symptoms of BPD (e.g., affective lability, inconsistent sense of self, dissociation). Either those variables aren't being measured or the results are not great. Either way, that information should be readily available by now.
In this paper, the authors note that a frequent complaint against DBT is that it is soley focused on symptom change rather than more general intrapsychic change, with the idea being that if behaviors shift, that does not mean that the underlying pathology has been resolved. To address this, they proposed to examine the impact of DBT on introject, defined as "an spect of an individual's personality that consists of self-directed actions including cognitive self-appraisals, and verbal and physical actions directed towards the self." (p.66) Introject is not a construct with which I am particularly familiar, so my discussion of its meaning will be relatively minimal here.
The authors noted that some researchers have hypothesized that the therapeutic relationship in DBT may serve as a mechanism through which intraphychic change my arrise and that other researchers have repeatedly noted that therapeutic alliance is preditive of treatment outcome (e.g., Castonguay, Constantino, & Grosse Holtforth, 2006; click here for a PBB post contradicting the proposed impact of therapeutic alliance). Given these points, they also sought to consider the role of therapeutic alliance across outcomes, not only in DBT, but also in patients receiving community treatment from expert clinicians.
To do this, the authors utilized data from a longitudinal randomized controlled trial in which 101 adult females were randomly assigned to receive either DBT or community treatment from expert clinicians. All participants were diagnosed with BPD through a structured diagnostic interview. Additionally, to be included, each participant had to endorse a past history of self-injury (at least two suicide attempts or NSSI within the past 5 years and a minimum of one NSSI incident within the past 8 weeks). Participants with a lifetime history of schizophrenia, schizoaffective disorder, bipolar disorder, psychotic disorder not otherwise specified, or mental retardation were excluded. Community therapists were nominated by leaders in the mental health community as particularly adept in the treatment of difficult clients and they self-identified as "eclectic" or "mostly psychodynamic." Community treatment consisted of at least one individual session per week and additional treatment prescribed as needed.
Therapeutic relationship was assessed during the active phase of treatment at 4 months, 8 months, and 12 months. Introject was assessed at 2 weeks, 4 months, 8 months, 12 months, and post-therapy follow-up at 16 months, 20 months, and 24 months.
The authors had 5 main hypotheses and I will describe the results relevant to each below the description of the hypothesis itself.
1. DBT patients would show a more affiliative introject during the course of treatment.
The two treatment conditions did not differ on levels of affiliative introject at the onset of treatment.
As anticipated, DBT patients saw an increase in affiliative introject across treatment and follow-up. Importantly, the patients in DBT improved more on this outcome than did individuals receiving treatment by community experts. Equally important, DBT patients, on average, shifted from a hostile to affiliative introject prior to the end of treatment whereas individuals in community treatment did not approach this point on average until the end of 1-year follow-up. The authors also provide data on individual clusters within these measures and I'll refer readers to the actual study for that level of detail.
2. DBT therapists would be perceived by patients as demonstrating greater levels of affirmation, protection, and control during treatment.
This hypothesis was confirmed, with DBT participants also reporting that DBT therapists exhibited higher levels of controlling behavior early in treatment while granting greater levels of autonomy as treatment progressed.
3. Higher therapist affiliation would be associated with increased introject affiliation.
When the authors examined the impact of therapist affiliation on introject measured at the same moment or the next assessment period in DBT, the effects were non-significant; however, when they looked at specific clusters within and across treatment, there were some significant findings. Specifically, DBT participants who reported higher levels of therapist "active love" at one time point reported higher "self-love" at the next assessment point and, similarly, DBT participants who reported higher "therapist protect" at one time point reported higher "self-protect" at the next assessment point. Additionally, DBT participants reported a stronger, positive association between "therapist affirm" and next period ratings of "introject self-affirm" in contrast to participants receiving community treatment who, on average, reported lower levels of "self-affirm" in assessment periods following high ratings of "therapist affirm." This last finding speaks to the possibility that the manner in which DBT therapists express affirmation may have stronger (in fact, postive relative to problematic) effects on patients' self-affirmation.
4. Higher therapist affiliation would be associated with less frequent NSSI.
There was no main effect of therapist affiliation on NSSI. In other words, the therapeutic relationship itself is not the determining factor on self-injurious behavior, an important point consistent with earlier PBB conversations emphasizing the importance of specific therapeutic techniques relative to "common factors" such as the therapeutic alliance (which many people use to claim that all treatments are equal when administered by quality therapists). There was, however, a significant affiliation by treatment interaction. In other words, the impact of relationship on NSSI depended upon the type of treatment being received. Specifically, DBT patients who perceived their therapists as more affiliative reported less NSSI throughout treatment and follow-up and this effect was stronger than in the community treatment condition. Importantly, in the community treatment condition, patients who reported higher "therapist affirm" also reported HIGHER levels of NSSI. In other words, when DBT therapists use affirmation, they appear to do so in a manner that reduces NSSI behaviors whereas when community expert clinicians use affirmation, they may actually reinforce and increase the likelihood of future NSSI behavior. In DBT, the relationship between therapist and patient is one of the primary means through which behavior is shifted in that, when a client self-injures, he or she is not able to contact the therapist for 24 hours and the next session is spent primarily conducting chain analyses to understand why the behavior occurred and how it can be avoided in the future through alternative adaptive response patterns. Contrastingly, when the patient calls for coaching in crisis or avoids using the behavior between sessions, substantial affirmation is given, thereby modeling effective means for building intimacy in relationships (e.g., encouraging healthy behavior and not rewarding harmful behaviors through greater levels of care and attention than are experienced in response to positive behaviors). It is possible that community clinicians were responding to their clients' self-harming behaviors through increased expressions of caring and concern, thereby granting such attention more in response to NSSI than in response to healthy behaviors and, in effect, increasing the value of NSSI behaviors to the clients.
5. The simultaneous use of emancipating, affirming, proecting, and controlling behavior (the DBT dialectic) by the therapist would predict improved outcome.
Although there was no main effect for the DBT dialectic (e.g., simultaneous emphasis on accepting the client as she is while noting the need to change), there was an interaction effect. Specifically, DBT patients who reported that their therapists used this approach also reported less NSSI whereas community patients who reported their therapists used this approach reported more NSSI. Now, as the authors note, it could be that the DBT pattern does not fit the approach used by community experts and that, as such, it might represent poorly conducted treatment. Regardless, it is certainly supportive that therapists who are perceived by their patients as behaving in a manner consistent with how DBT is intended to be administered tend to have patients who end up decreasing their use of NSSI.
On the whole, I think this study had a TON of strengths to it, in that it expanded the type of outcomes assessed, used an impressive study design (RCT, longitudinal approach with multiple follow-ups), and strong analytical approach (hierarchical linear modeling). That being said, I can't help but wonder why outcomes often assessed in DBT trials (e.g., suicide attempts, re-hospitalization) were not included here. If they did not have the data, that makes perfect sense. If there was not enough power, that makes sense. If they simply did not report on data they had access to and enough variability to analyze, that's a problem. Additionally - and this is a limitation that applies to almost all clinical trials - I can't help but wish the comaprison treatment was one unified approach administered by experts in that particular approach, thereby allowing for comparisons of incremental validity using the same sample rather than relying on effects aggregated in meta-analyses in which few if any of the studies actually involved the direct comparision of treatments.
What are your thoughts?
Studies cited in this post:
Bedics, J.D., Atkins, D.C., Comtois, K.A., & Linehan, M.M. (2012). Treatment differences in the therapeutic relationship and introject during a 2-year randomized controlled trial of dialectical behavior therapy versus nonbehavioral psychotherapy experts for borderline personality disoders. Journal of Consulting and Clinical Psychology, 80, 66-77.
Castonguay, L.G., Constantino, M., & Grosse Holtforth, M.G. (2006). The working alliance: Where are we and where should we go? Psychotherapy: Theory, Research, Practice, Training, 43, 271-279.
Kliem, S., Kroger, C., & Kosfelder, J. (2010). Dialectical behavior therapy for borderline personality: A meta-analysis using mixed-effects modeling. Journal of Consulting and Clinical Psychology, 78, 936-951.
Lynch, T.R., Trost, W.T., Salsman, N., Linehan, M.M. (2007). Dialectical behavior therapy for borderline personality disorder. Annual Review of Clinical Psychology, 3, 181-205.
Dr. Anestis is a post-doctoral fellow with the Military Suicide Research Consortium