By Michael D. Anestis, M.S., Psychotherapy Brown Bag

I have noted on a number of occasions that I am a believer in empirical data, regardless of whether the numbers support ideas consistent with my beliefs. 

Today, I would like to provide you with some evidence of this point by discussing a new study just published in Behavior Therapy by Debra Safer, Athena Hagler Robinson, and Booil Jo of Stanford University (2010).

In this study, the authors discussed a recently conducted randomized controlled trial (RCT) of dialectical behavior therapy (DBT) for binge eating disorder (BED...excuse all the acronyms!). 

Prior work looking at the utility of DBT in treating binge eating behaviors has provided highly promising results; however, those studies either involved single case study designs or RCTs in which DBT was compared to a waitlist control group. 

In the study I'll discuss today, the authors compared DBT to an active treatment control group and I'll explain why this is an important difference.

Before describing today's study, let me provide a little background information. 

A substantial amount of research has produced strong evidence that binge eating behavior is often motivated by a desire to reduce negative emotions.

 In other words, people get upset and use binge eating as a method of distracting themselves from and reducing the intensity of negative affect (click here and here for descriptions of some of this evidence). 

Because of this, researchers have investigated whether DBT, which focuses on teaching clients how to better tolerate and regulate negative emotions, might be an effective intervention (click here for a description of this early evidence). 

Any time a new treatment is investigated or an established treatment is investigated for a new diagnosis, early studies take a conservative approach, comparing the treatment in small samples to a waitlist (individuals on the waitlist then typically receive the same therapy once space opens up, assuming the treatment appears to work). 

If the treatment produces better results than the waitlist, this supports the notion that it is better than nothing. 

Obviously, while that is a positive, it falls far short of our goals for any treatment. 

Because of this, the next step in studying a treatment involves comparing it to an active control treatment.  In this type of study, participants either receive the new treatment or an alternative treatment that does not share any of the same specific techniques. 

By demonstrating that the treatment outperforms another active treatment, researchers can develop confidence that, when individuals get better, it is as a result of the treatment itself and not due to common factors of psychotherapy (e.g., therapeutic alliance, therapeutic optimism).

Okay...on to today's study.  Safer and her colleagues recruited 101 participants and randomized them to receive either 20 sessions of DBT or 20 sessions of an active control group treatment (ACGT).  DBT was conducted according to the protocol designed by Telch and colleagues (2001), which involves only group sessions and does not include the interpersonal effectiveness module. 

ACGT was designed to resemble supportive psychotherapy for depression and emphasized discussions of the participants' strengths and efforts to increase self-esteem. By pure chance, the participants in the DBT group exhibited higher levels of depression than did the participants in the ACGT group, but no other between group differences were noted prior to the onset of treatment.  Assessments were made prior to treatment, immediately following treatment, and 3, 6, and 12 months post-treatment.

As I hinted at above, the results of the study were fairly surprising. 

On the one hand, dropout rates were much lower in the DBT group (4%) than in the ACGT group (33.3%).  Additionally, participants in the DBT group responded to treatment much more quickly than did individuals in the ACGT group, as evidenced by the fact that 64% of DBT participants reported no binge eating immediately following the end of treatment as compared to 36% in the ACGT group (p < .05). 

On the other hand, as time went by, the results painted a much different picture.  There were no differences between the two treatment groups at 3 (51% for DBT vs 53% for ACGT), 6 (52% vs 43%), and 12 (64% vs 56%) month follow-up on binge eating abstinence.  In other words, although those percentages do not look equal to one another, the data indicate that any differences between them failed to reach statistical significance, meaning that both treatments did a relatively good job of reducing binge eating, but neither outperformed the other on that outcome. 

Additionally, DBT did not appear to have a substantial impact on emotion regulation skills and, in fact, there was a small between group difference favoring ACGT on this outcome.


So what does all of this mean?  First of all, although this is only one study, it represents the only direct comparison of DBT to an active treatment for BED and the evidence does not support the notion that DBT is superior to a supportive psychotherapy approach in the long term with respect to remaining abstinent from binge eating relative.  Now remember, DBT does outperform waitlists and the abstinence rate post-treatment up to 12-months in this study was solid, but these findings indicate that the results might not be due to the mechanisms thought to drive DBT. 

In fact, DBT did not appear to have the intended effect on emotion regulation skills, meaning that its utility in treating binge eating may have more to do with common factors than with its focus on emotion (neither treatment specifically focused on ways to manage eating habits, so such skills do not account for the benefits either).  Substantially fewer people dropped out of DBT and the two that did dropped out because of family emergencies (n = 1) and a move out of the area (n = 1), so it does appear that DBT is a more acceptable treatment to clients with BED, a point that is not insignificant as it indicates that a greater number of people are likely to actually benefit from treatment. 

Additionally, the effects of DBT were faster than the effects of ACGT, a point that might be of importance for particularly severe cases.  That being said, these results are not consistent with the overwhelmingly positive results from earlier trials and they raise questions as to whether this approach is a viable alternative to other empirically supported treatments for BED such as cognitive behavioral therapy and interpersonal psychotherapy.

A trial like this is a bit of an eye opener.  On a personal level, I am a firm believer in DBT and it is easily the form of therapy I have most enjoyed administering as a therapist. Additionally, I have conducted a substantial amount of research on the impact of emotions on binge eating behaviors and, as a result, have spent a good deal of time considering the potential utility of DBT for BED. 

The bottom line, however, is that reality sometimes does not conform to our expectations.  This is why science is such an incredibly useful component of life.  Without systematic investigations of topics like this, we are left to do what seems reasonable rather than what actually produces the best results.  As you can see, even those of us whose professional lives are devoted to these topics can be taken aback by the numbers, which means that the numbers are remarkably important. 

This one study should by no means cause us to cease investigating ways to implement DBT protocols capable of producing reliably strong results in the treatment of BED, but it should keep us from completely accepting that this seemingly logical approach is a valid direction to take prior to attempting other approaches with a greater level of support.

I would be interested in hearing your thoughts on this.  Why do you think DBT failed to impact emotion regulation skills in this sample?  Was it a problem with the study itself or a reflection of DBT simply being less useful for this particular diagnosis relative to others (e.g., BPD)?  Are there ways in which you would have conducted this trial differently?


If you would like to learn more about these topics, I hope that you will consult our online store for scientifically-based psychological resources.

Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University and an incoming resident at the University of Mississippi Medical Center.

See original for multiple links:
Treating binge eating disorder: A new study looks at dialectical behavior therapy (DBT)

emotional eating, eating and stress, binge eating disorder, binge eating behavior