Back in April, I wrote an article on PBB discussing the empirical support for various treatments for anorexia nervosa (AN). The bulk of the article focused on family-based treatment (FBT) for child and adolescent AN, as this treatment has the greatest degree of research supporting its utility for this particular diagnosis. In recent months, I have had a number of pleasant interactions with the folks at Maudsley Parents (http://www.maudsleyparents.org; follow @JaneCawley on Twitter) and have come away from those conversations feeling compelled to spread the word about the data on this matter.
If you have not read my prior FBT article, I would recommend consulting it for a detailed description of the nature and structure of the treatment. Additionally, I would recommend reading Lock and le Grange (2005) for a summary of the nature of the treatment and earlier research supporting its efficacy in the treatment of eating disorders (see our References page for the entire citation). That being said, the context within which to view the utility of this treatment is one of hope amidst frustration. Empirical support for treatments for AN is elusive and, as a result, many look upon our ability to treat the disorder with a mix of emotions ranging from anger to despair. Fortunately, results of FBT, which was designed at the Maudsley Hospital in London, have offered reason for greater optimism, at least with respect to adolescents. The goal of today's article is to provide a greater degree of detail regarding the data supporting the efficacy and effectiveness of FBT in the treatment of eating disorders in order to help readers develop a clearer sense of why many clinicians and researchers believe we are seeing the beginning of a better age in the treatment of previously treatment resistant diagnoses.
The rationale for discussing this data stems from a the fact that some controversy - or at least some concern - surrounds FBT. The early empirical work on this approach, as one might expect, was conducted by a single research group comprised of the researchers who developed the treatment. As such, good science dictates that independent replications by researchers not associated with treatment originators at a location outside the Maudsley Hospital are needed in order to optimize confidence in the results. Fortunately, such data now exist. Additionally, some individuals find it troubling that FBT involves parents assuming control of the adolescent's eating behaviors for the first six months of treatment, as early theoretical models of AN centered around the idea that the disorder represents a misguided search for autonomy from an enmeshed family (e.g., Minuchin et al., 1978). Loeb and colleagues (2007) noted that FBT takes an atheoretical stance on the origins of AN and focuses instead on what approaches are most likely to increase the chances of recovery. Along these lines, they utilize the family as a resource for guidance and authority in treatment and their results make it difficult to argue with that decision. Other criticisms of FBT center on the relatively brief nature of the treatment (one year) and its outpatient nature, as most treatment models for AN involve a chronic course involving frequent inpatient hospitalizations. In this case, the data are all that matters, and I will present empirical evidence indicating that individuals not only recover at a high rate through FBT, but maintain these gains at long term post-treatment follow-ups. Finally, some critiques of FBT stem from the fact that its methods do not easily map onto a standard theoretical orientation (e.g., psychodynamic, cognitive behavioral). As a scientist, I find this critique baffling. Although much of my training is from the CBT perspective, my devotion is to empirical investigation not CBT in particular, and if a treatment not associated with my typical theoretical viewpoint demonstrates efficacy and/or effectiveness through rigorous scientific investigations, it is unclear how the theoretical orientation of the treatment is relevant. Think of it this way: if somebody discovered a cure for a previously incurable form of cancer and that treatment did not involve any dangerous side effects or unethical procedures, would it matter to you whether the treatment was consistent with what you expected from it? If it would - which is reasonable - I suspect it would nonetheless matter much less than the fact that it works.
Okay, so let's discuss the evidence indicating that FBT not only works, but works outside of the Maudsley hospital itself and results in long-term improvements. This review will focus on recent research performed outside of the Maudsley Hospital, but please note that prior research on this treatment exists as well.
Loeb et al., 2007
In an open trial conducted in a university clinic, 20 male and female adolescents (ages 12-17) were treated with manualized FBT. All participants either met full criteria for AN or were subthreshold. Subthreshold AN required that the participant meet the weight loss criteria (below 85% of expected body weight) but only have missed one or two menstrual periods rather than the three required for amenorrhea or that the participant miss the weight cutoff but lose enough weight to qualify for amenorrhea. As such, subthreshold AN did not represent "mild" cases. Participants in this study exhibited a significant increase in body weight and a significant decrease in dietary restraint and eating concerns. They did not, however, exhibit statistically significant improvements in shape or weight concerns or depression symptoms. Depression symptoms at baseline were mild, however, so improvement on that factor was not a primary aim of treatment. Only two of the 20 participants self-initiated early termination from treatment, meaning that FBT did a solid job of overcoming the attrition issues that plague AN treatment.
On the whole, these findings were fairly strong. They demonstrate that the manualized form of FBT can be implemented in a setting outside of the Maudsley hospital with a decent degree of success. The lack of improvement on weight and shape concerns is an issue; however, the fact that the participants improved substantially on the more dangerous symptoms of AN is important to consider as well. Additionally, the small sample size vastly diminished the statistical power of the analyses, rendering it difficult to detect even moderate sized effects. As such, a larger sample size could help reduce fears of false negatives.
le Grange, Crosby, Rathouz, & Leventhal (2007)
In this study, the authors examined the utility of FBT in the treatment of bulimia nervosa (BN). In prior articles, we have discussed the evidence supporting the utility of a number of other treatments for BN, including CBT, interpersonal psychotherapy, and dialectical behavior therapy (as well as the lack of evidence for other treatments such as equine assisted psychotherapy). Here, the authors presented compelling data indicating that FBT can be useful in treating BN as well.
This study was a randomized controlled trial (RCT) comparing FBT to supportive psychotherapy in the treatment of adolescent BN. 80 participants were randomly assigned to one of the two treatments and the results for the two approaches were compared immediately post-treatment and at a six month follow-up. Immediately following treatment, more individuals in the FBT condition (39%) were binge-purge abstinent than in the supportive psychotherapy group (18%). Abstinence rates decreased somewhat at 6-month follow-up, but again individuals in the FBT group were more likely to be binge-purge abstinent. Post-treatment analyses also revealed that individuals in the FBT group exhibited significantly greater reductions in objective binge episodes, subjective binge episodes, all compensatory behaviors, dietary restraint, weight concerns, shape concerns, eating concerns, and global eating disorder symptoms. The two groups did not differ in degree of improvement on self-esteem or depression.
Clearly, these results are very promising. The authors utilized strong methodology (RCT) to compare two active treatments and FBT demonstrated superior efficacy in a wide range of eating disorder related outcomes. FBT was not compared to any empirically supported treatments for BN, so of course caution should be used in interpreting the degree to which FBT offers as much promise as those treatments; however, this study does provide a rationale for budding confidence and future research.
Rhodes, Baillee, Brown, & Madden (2008)
In another RCT, Rhodes and colleagues (2008) sought to examine whether the addition of a parent-to-parent consultation early in treatment would improve outcomes in the treatment of AN. 20 families were randomly assigned to standard FBT and FBT with the added parent component. Results indicated that the addition of the parent-to-parent consultation was associated with significantly greater weight restoration in AN patients.
Paulson-Karlsson, Engstrom, & Nevonen (2009)
In the most recent study I could find on this matter, Paulson-Karlsson, Engstrom, and Nevonen (2009) tested the efficacy of FBT in the treatment of AN in a hospital in Sweden. 32 female adolescents who met criteria for AN and their parents received FBT. At 36-month follow-up, 75% of the participants were in full remission. In addition to the elimination of their eating disorder symptoms, the participants reported experiencing less distance from their families and a less chaotic familial environment. The lack of a comparison group obviously renders this study far from perfect, but the results are nonetheless highly impressive.
The above mentioned studies do not represent the entirety of the work in this area, but they provide a fairly comprehensive glimpse at the evidence supporting the utility of FBT in the treatment of adolescent eating disorders, particularly with respect to AN. Skepticism regarding treatments is a healthy scientific mindset and clearly we at PBB demonstrate this rather regularly when we discuss treatments for which evidence is lacking. That being said, skepticism towards the utility of FBT for eating disorders appears unwarranted based upon the data. Concern that symptom improvements are not more substantial are entirely legitimate; however, the complete lack of empirically supported alternatives for this particular subset of individuals renders the gains seen through FBT highly meaningful. As Daniel le Grange (2006) said in a commentary replying to an article criticizing the positive views of FBT (Bergh et al., 2006): "it makes little sense to dump the car that starts every other morning before you have purchased a new and potentially more reliable model!" In other words, FBT represents progress - impressive progress at that - with a population for whom we previously had no treatments with strong empirical support. Its imperfections do not render it useless or irrelevant, nor do they bestow upon less supported treatments any measure of enhanced efficacy. FBT marks an emerging dawn after a long dark stretch devoid of strong treatment research on AN. We should embrace these results even as we challenge them, test alternatives, and strive for something better.
If you would like to learn more about family-based treatment, we recommend the following items, all of which are available through our online store of scientifically-based psychological resources:
- Treatment Manual for Anorexia Nervosa: A Family-Based Approach by James Lock, Daniel le Grange, Stewart Agras, and Christopher Dare
- Eating with Your Anorexic: How My Child Recovered Through Family-Based Treatment and Yours Can Too by Laura Collins