Two Programs to Combat Relapse in Anorexia Nervosa

Relapse after successful initial treatment for anorexia nervosa (AN) is a common problem, and one of the reasons that the prognosis for such patients is often poor. The first step in the recovery process involves nutritional rehabilitation and weight restoration, often requiring day hospital or inpatient treatment. Afterward, relapse rates can range from 9% to 65%.

Jacqueline C. Carter, PhD, and colleagues at York University and Toronto General Hospital, Toronto, Canada,  recently compared two maintenance approaches for patients with AN who had achieved a minimum body mass index (BMI) of 19.5 kg/m2 and had controlled their binge eating and purging symptoms after participating in a specialized hospital-based program (Int J Eat Disord 2009; 42:202).

In the non-randomized clinical trial, two maintenance treatment conditions for AN after weight restoration were compared: individual cognitive behavioral therapy (CBT) and “maintenance treatment as usual” (MTAU). The latter was an assessment-only naturalistic control condition. The patients were treated with one of the two approaches for one year, and assessments were made before and after the first weight-restoration treatment as well as at three-month intervals during the year. The study participants were 88 consecutive female patients who met DSM-IV criteria for AN at the time of admission to the inpatient or day hospital program at Toronto General Hospital.

CBT group

Patients in the CBT treatment group were eligible to receive 50 individual therapy sessions with an experienced psychologist. The sessions lasted 45 minutes, and all patients concurrently received fluoxetine or a placebo. Most patients participated in at least 38 sessions.

The treatment was manual-based and involved strategies to address behavioral dysfunction related to eating and weight that increase the risk of relapse; cognitive restructuring techniques pertaining to eating and weight; and application of a schema-based approach to address a wide range of relevant issues beyond eating and weight, for example, interpersonal problems or self-esteem.

Maintenance treatment as usual group

Patients in this group were advised to seek follow-up care “as usual.” One option was to attend the aftercare program at the treatment center, which consists of three 90-minute group therapy sessions per week for up to 12 weeks. Patients who complete this first phase may then choose to attend a second phase consisting of one 90-minute group therapy session per week for up to six months. While patients were free to seek follow-up care in the community, they could also choose not to seek any therapy if desired.

How both groups fared

Of the 46 participants in the CBT group, eight were withdrawn as “treatment failures” because of deterioration in their condition; 10 dropped out prematurely before relapsing; and two were withdrawn for missing too many sessions or for not complying with medications. Twenty-six completed the entire one-year CBT treatment protocol. Of the 42 participants in the MTAU condition, 12 dropped out of the study within the year because they were unwilling to continue to participate in the assessments. There was no significant difference in dropout rates between the two groups.

When relapse was defined as a BMI ≤ 17.5 kg/m2 for 3 months, the time to relapse, based on survival analysis, was significantly longer in the CBT group than in the MTAU group. After one year of treatment, 24.4% of the CBT group but 50% of the MTAU group had relapsed.

When relapse was defined as either a BMI ≤ 17.5 kg/m2 for three months or resumption of regular binge eating and/or purging behavior for 3 months, time to relapse was significantly longer in the CBT group. At one year, 32.5% of the CBT group and 65.6% of the MTAU group met criteria for relapse. An evaluation of the medication study indicated that fluoxetine had no impact on relapse.

Limitations and future studies

The authors note that an important limitation of their study was that the participants were not randomly assigned to treatment groups; thus, the study’s results should be carefully interpreted and considered as preliminary findings. Only one significant difference emerged between the groups: The CBT group reported significantly higher baseline scores on the EDI Drive for Thinness subscale. Because participants were not randomized to the study conditions, it is possible that selection bias influenced the results.

On the positive side, the results of the study offer further preliminary support for use of CBT for treatment of AN. Because it is essential to develop effective maintenance treatment approaches for patients with AN, the authors state that a next important step is to conduct a large-scale randomized controlled study of CBT for AN in which CBT is compared to a psychotherapy control condition that controls for “common factors” but does not contain the specific therapeutic strategies of CBT.