Review of: Zeeck A et al, Front Psychiatry 2018;9:158. doi:10.3389/fpsyt.2018.00158

Type of study: Meta-analysis of randomized controlled trials

Psychotherapy is the main treatment for anorexia nervosa, but which type works best? Several therapies have good evidence in this population, but they differ in their models and methods, and head-to-head comparisons among them are rare. To overcome that limitation, this study used a technique called “network meta-analysis,” which evaluates different treatments based on how they measured up against a common comparison group. For example, suppose that CBT and family therapy have never been directly compared to each other but both have been compared to supportive therapy. A network meta-analysis would compare CBT to family therapy based on how each fared relative to supportive therapy.

Only a handful of therapies have good evidence to work in anorexia, and most of them were included in this study. Effective therapies had two common ingredients: a focus on weight restoration and work on psychosocial factors. It was in the psychosocial focus that the therapies differed, which ranged from skill building (CBT), relationship dynamics (focal psychodynamic therapy, interpersonal psychotherapy), family work, and supportive psychotherapy (specialist supportive clinical management). The family therapies empowered parents to re-feed their child, and then progressed to work on family dynamics (systemic family therapy) or adolescent development (family-based treatment and the Maudsley model) as normal weight was restored.

The result: No single therapy was more effective than the others in this analysis of 18 randomized controlled trials. The authors followed that up with another new-fangled technique, called “standardized mean change analysis,” which compared the degree of weight gain among all of the therapies after 1 year of treatment. This analysis allowed naturalistic studies to be included, bringing the total number of trials to 38. Again, no single therapy stood out, but weight gain was more rapid with inpatient vs outpatient treatment, and overall weight gain was greater in adolescent studies than it was for adults (inpatient: 1.4 lbs/week for adolescents, 1.2 lbs/week for adults; outpatient: 0.42 lbs/week for adolescents, 0.23 lbs/week for adults).

The authors suggested that some therapies may be superior for certain subgroups of anorexia. Most successful therapies for adolescents involved the family, while individual therapy was the mainstay for adults with anorexia. Adolescents with significant obsessive-compulsive symptoms had greater benefit with systemic family therapy than family-based treatment. For severe anorexia, the ­Maudsley model was more effective than specialist supportive clinical management.

TCPR’s Take
While the outcomes for these therapies were similar, this does not mean that any psychotherapy will work for anorexia. These are highly structured therapies with specific behavioral and psychological techniques. When making referrals, psychiatrists should look for therapists that use evidence-based methods, and adolescents may do better with a family approach. Once in therapy, weight gain of 0.23–1.4 lbs/week can be considered a successful outcome.