Evaluation of Family Treatment Models for Eating Disorders

James Lock


Curr Opin Psychiatry. 2011;24(4):274-279. 

In This Article

Systematic Studies of Family based Treatment for Anorexia Nervosa

Most of the recent systematic studies of family treatments for eating disorders have focused on anorexia nervosa. The most recent study compared FBT with a developmentally tailored individual therapy called adolescent focused therapy (AFT).[16••] The rationale for the study was to compare an approach that focused on parental management of anorexia nervosa symptoms (FBT) and an approach that was designed to use the therapeutic relationship to foster independence, autonomy, and self-management of anorexia nervosa symptoms (AFTs).[38] These two approaches each had credible evidence to suggest that each could be effective for adolescent anorexia nervosa based on a smaller study (n = 37) that had compared versions of them previously.[14] The larger study was designed to be powered sufficiently to test two active treatments (n = 121) and involved two clinical research sites (Stanford University and The University of Chicago). Participants in this study were adolescents (age 12–18 years) who had a DSM-IV-R diagnosis of anorexia nervosa exclusive of the amenorrhea criterion.[39] Weight thresholds for diagnosis were established using Centers of Disease Control BMI percentiles for age and sex.[40,41]

The results of the study found that those treated with FBT had significantly greater improvements in mean percentile BMI [AFT = 23.4 (2.8), FBT = 31.4 (2.8), P = 0.048] and greater changes in eating-related psychopathology as assessed by the Eating Disorder Examination (EDE) total score at the end of treatment [AFT = 1.20 (0.15), FBT = 0.71 (0.16), P = 0.03], but these differences were not detected at follow-up. In contrast, remission rates (normal weight and normal EDE) did not differ significantly at the end of treatment (EOT) (P = 0.55), but at 6-month and 12-month follow-up those in FBT demonstrated higher rates of remission than those in AFT (AFT = 23%, FBT = 49%, P = 0.02). The likely reason for not detecting a significant difference in remission rates at the end of treatment was study power. Despite not finding a significant difference in remission rates at EOT, the effect size (5) was similar to that found at all follow-up points. On the contrary, the lack of significant differences on continuous measures of BMI percentile and EDE at follow-up likely reflects a tendency for general improvements on these measures (regression to the mean) but not improvements to clinically meaningful levels (remission). In addition, differences in relapse rates from remission status at the EOT differed between the two groups (AFT = 40%, FBT = 10%) by 12-month follow-up. Hospitalization rates were greater in AFT (31%) than FBT (18%). Taken together, these results supported the main hypothesis of the study, which was that FBT would be more successful in helping adolescents remit from anorexia nervosa than AFT.

During the past 5 years, other important studies evaluating the role of FBT for child and adolescent eating disorders have been published that have helped to understand the role of therapeutic alliance in FBT, longer-term outcomes in FBT, the amount of FBT needed, predictors of response to FBT, utility of FBT of preadolescents, and the dissemination of FBT.

Therapeutic alliance in FBT for both the affected adolescent and parents are important in helping to initiate treatment response and to keep families from dropping out of treatment.[42] However, therapeutic alliance alone in FBT is insufficient over the course of therapy to sustain clinical gains. One study found that behavioral response (early weight gain) was a better predictor of psychological improvement and therapeutic alliance at the end of treatment.

Two studies of longer-term outcome in RCTs found that participants successfully treated with FBT remained well 4–5 years after treatment.[43–45] In each of these studies, approximately 80–85% of patients treated no longer met diagnostic criteria at follow-up.

One relatively large study (n = 86) compared a 6-month and 12-month dose of FBT and found no differences in outcome related to dose at the end of treatment or follow-up.[15] In this study, patients received a shorter dose of 6 months and 10 sessions or a longer dose of 12 months and 20 sessions of FBT. Patients in both groups had similar changes in BMI (shorter: baseline = 17.0, EOT = 19.5; longer: baseline = 17.3, EOT = 19.5) and improvements in the restraint subscale of the EDE (shorter: baseline = 2.76, EOT = 1.62; longer: baseline = 2.64, EOT = 1.42). However, subgroups of patients who had higher levels of obsessive compulsive symptoms or who came from nonintact families did better with the longer treatment.

A post-hoc study of predictors of reaching normal weight by EOT in FBT found that gaining about 4 pounds by week 4 of FBT was 90% predictive of this outcome.[46•] This finding suggests that, when FBT works, it does so relatively quickly – a finding consistent with other behavioral treatments.

Finally, a single case series (n = 32) examined the use of FBT with patients younger than 13 and found that it was feasible and acceptable, and that patient outcomes were similar to those achieved with adolescent subjects.[47] Patients in this study demonstrated significant weight gain from a mean of 86% ideal body weight (IBW) to 99% IBW at EOT. Significant declines were also noted on the EDE.

There are several studies that suggest that FBT can be disseminated and achieve similar outcomes in diverse settings. The first of these studies was conducted by Loeb et al.[48] at Columbia University. Clinicians were trained using a basic workshop, utilized the manual, and had supervision from one of the authors of the manual. The authors report that dropout was low (25%) and the patients gained weight (from a mean of 82% at baseline to 94% at EOT) and significantly improved scores on the EDE. Researchers in Australia who were trained in FBT found that the use of FBT reduced rehospitalization rates (from an average of 2.08 admissions in 2002 to 1.27 admissions in 2006, P = 0.0001).[49] Couturier et al.[50•] in Canada evaluated the fidelity of the clinicians who had been trained in FBT. They found most patients demonstrated similar clinical improvements in physical and emotional health compared with those treated in published trials. In addition, therapists demonstrated good fidelity to the first phase of FBT (72%) though adherence lessened in the second (47%) and third (54%) phases. The most recent study was conducted in Brazil, where clinicians were trained and treated a series of clinically referred patients.[51] They found that patients and families accepted FBT (82%) and that most patients normalized their BMI from baseline (mean = 16.4) to follow-up (20.8) (P = 0.012) and significantly improved eating-related psychopathology on the EDE (baseline = 2.81, follow-up = 1.22, P = 0.069).