by Michael D. Anestis, M.S.
Whenever possible, I like to make an effort to focus my articles on research that actually examines to what degree a particular therapeutic approach works for a specific mental illness. This, of course, stems from my belief in the utility of empirically supported treatments (ESTs), treatments whose value is determined by their ability to demonstrate improvement in clients in a systematic, quantifiable way (see our registry of EST clinics for services in your area). Along those lines, today's article will focus on a new study published in the International Journal of Eating Disorders by Dr. Jacqueline Carter of the Department of Psychiatry at the University of Toronto and several of her colleagues (2009). In this study, the authors examined the utility of cognitive-behavioral therapy (CBT) as a maintenance treatment for individuals recovering from anorexia nervosa (AN).
AN, unfortunately, is characterized by discouragingly high rates of relapse (Carter, Blackmore, Sutandar-Pinnock, & Woodside, 2004). Although there is strong empirical evidence for the utility of family-based treatment for children and adolescents with AN, for many individuals, treatment consists of inpatient weight restoration followed by a series of ineffective approaches towards maintaining those gains. In other words, although weight restoration is pivotal in order to ensure the physical safety of the client, simply restoring weight does not address the underlying symptoms and, unfortunately, efforts to develop maintenance treatments capable of addressing that issue have, in large part, failed to live up to expectations. Carter and colleagues (2009), however, pointed to a recent randomized trial comparing CBT to nutritional counseling that found CBT to be more effective at maintaining weight restoration. Building off of that finding, the authors wanted to see if individuals who received one year of manualized CBT would maintain weight gains better than a group of individuals who received an assessment only "maintenance treatment as usual (MTAU)."
The sample in this study was a bit different than those seen in many other studies. Initially, 88 female patients who had met DSM-IV criteria for AN at intake but had achieved a body mass index (BMI) of at least 19.5 and ceased regular binge eating and purging through inpatient weight restoration and treatment were considered for inclusion in a trial testing the utility of the antidepressant fluoxetine in preventing relapse. 46 of those 88 patients chose to participate in that study, which found that fluoxetine did not decrease the likelihood of relapse in AN. In addition to the antidepressant medication, however, individuals in that study also received one year of manualized CBT. The remaining 42 patients - those who chose not to participate in that study - were instead involved in an assessment only MTAU condition. This trial, in other words, was not randomized. Instead, participants chose which condition they preferred to be involved in, which makes it impossible to rule out the possibility that any effects found are better accounted for by whatever variable made those individuals choose the more effective condition.
CBT maintenance treatment in this study consisted of an average of 38 sessions, 45 minutes each, for one year. Treatment involved three phases. The first phase focused on strategies to alter dysfunctional behaviors related to diet and weight that might otherwise contribute to relapse. The second phase focused on cognitive restructuring in an effort to alter the manner in which the participants viewed themselves, food, and the need to lose weight. The third phase involved the application of schema-based procedures that aimed to address a variety of life domains, including interpersonal relationships and self-esteem. Participants in the MTAU condition, on the other hand, were asked to seek follow-up care "as usual" at the center running the study. This protocol involved attending three 90-minute group therapy sessions per week for up to 12 weeks, followed by the option to attend one 90-minute group therapy session per week for an additional 6 months. Participants in this condition were able to seek other forms of therapy if they chose or to seek no therapy at all.
A fairly substantial number of individuals dropped out of both conditions, leading to a decreased sample size. This, however, represents a concern in all AN treatment research, not a failing on the part of the authors. For the purpose of this study, relapse was defined as achieving a BMI of 17.5 or less for three or more consecutive months or resuming regular binge eating and/or purging behaviors for a period of at least three months. Based on that definition, the authors reported the following results:
Time to relapse was significantly longer in the CBT group than in the MTAU group. In other words, when individuals in the CBT group did relapse, this occurred after a longer period of recovery than was typically seen in the MTAU group. Additionally, whereas only 32.5% of the individuals in the CBT group met criteria for relapse at the one-year follow-up, 65.6% of the individuals in the MTAU group had relapsed.
Importantly, as the authors themselves note, these results are imperfect and represent only preliminary evidence for the utility of CBT as a maintenance treatment for AN. Ideally, relapse rates would be lower than they were in this sample; however, a greater than 50% reduction in relapse relative to treatment as usual is substantial and clinically important. In order for confidence in these findings to increase, several things must occur. First, additional studies must replicate these effects. In order to ensure that these findings were not due to chance or some particular aspect of this particular sample, other researchers need to be able to produce similar findings in completely different groups of participants. Additionally, as Carter and colleagues (2009) noted, future studies should compare CBT to other active treatments in order to control for common factors. The comparison treatments should not, however, include the active ingredients of CBT (e.g., cognitive restructuring) so that the comparisons made can therefore determine the degree to which specific aspects of CBT truly contribute to the welfare of the participants. Also, future trials should be randomized in order to ensure that whatever makes an individual choose a particular treatment is not the actual difference maker (e.g., enhanced motivation for treatment) in therapy.
So what is the take home point from this study? Although treatment research for AN still lags behind that of many other mental illnesses (e.g., depression, panic disorder, obsessive-compulsive disorder, bulimia nervosa), there are many devoted researchers and clinicians working to address this problem and producing promising results. Individuals with AN are not in a hopeless situation and treatment does not need to be random or based purely on the intuition of a well-intended but potentially misguided clinician. Instead, there is evidence that particular approaches to addressing treatment and relapse prevention hold particular promise. If you are consumer of these resources, I hope you will take the opportunity to educate yourself on these options.
If you would like to learn more about anorexia nervosa, we recommend the following resources, all of which are available through our online store:
Treatment Manual for Anorexia Nervosa: A Family-Based Approach
Eating with Your Anorexic
Crave: Why You Binge Eat and How to Stop
Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University.
Subscribe in a reader
