Two Types of Family Therapy, and the Effects on Outcome


There is increasing evidence that family therapy is an effective option for adolescents with anorexia nervosa. If family therapy is effective, is conjoint family therapy even more effective? A five-year study in Great Britain reports that it is—with some reservations.

Dr. Ivan Eisler and colleagues at King's College, Maudsley NHS Trust, and Hayes Grove Priory Hospital, London, conducted a five-year follow-up study on a cohort of 40 adolescent patients who had received either "conjoint family therapy" or "separated family therapy." In conjoint family therapy (CFT), the entire family is seen together for treatment, versus separated family therapy (SFT), in which the teen is seen individually, and then the parents meet in sessions with the same therapist without their child. Thirty-eight patients agreed to be reassessed with personal or telephone interviews or questionnaires (J Child Psychol Psychiatry 2007; 48:552).

Outcome: mostly positive

The most encouraging finding was the generally good outcome for most of the patients five years after recovery. At follow-up, more than 75% of the patients had no eating disorder symptoms. The authors note that this finding is very consistent with other studies of teens treated with family therapy; for example, studies by Martin (1985) and Herscovici and Bay (1996) described good outcomes among 80% and 60%, respectively, of patients five years after recovery.

Dr. Eisler and colleagues also reported a low rate of relapse once recovery was achieved. In their study, 8% of those who achieved a healthy weight by the end of treatment relapsed. The authors added add that because 20% of the patients were not available for follow-up evaluation, the relapse rate may have been slightly higher.

Their third positive finding on follow-up was that only three of the 40 patients developed bulimic symptoms and only one patient had symptoms that warranted a diagnosis of bulimia nervosa. In other studies, in comparison, around 30% of patients developed a pattern of binge eating even while they were in the midst of recovery (Strober et al, 1997; Sunday et al, 1996).

Factors that could predict outcome

Some factors helped predict outcome: the amount and type of previous treatment; the duration of illness; the degree of weight loss at the beginning of treatment; the degree of maternal criticism; and a history of psychiatric disorder. Patients whose mothers made three or more critical comments during the initial assessment family interview had poorer outcomes.

Seeing the family as a unit vs. seeing patient and family separately

Overall, there were no differences in the long-term outcomes between the separated and conjoint family groups. However, patients from families with raised levels of maternal criticism had more eating disorders symptoms at follow-up.

Family therapy isn't for everyone. The authors also noted that a small group of families just cannot fit into any type of traditional family therapy. Parents in these families often feel particularly burdened by the problem and are generally self-blaming.

One parent, often the mother, may alternate between being closely involved in attempts to help her daughter, and then feeling overwhelmed by the enormity of the task. She then backs off completely and tends to avoid dealing with the problem.

Seeing this, the other parent may try to play down the seriousness of the problem to reduce the pressure on the spouse or may become more distant and absorbed in work. The sense of helplessness and anxiety generated by the situation often leads to increased criticism or hostility.

In such cases, clinicians may find that their attempts to identify the strengths and potential resources of the family are rejected by the family. With such families, joint family meetings early in treatment may simply reinforce a sense of ineffectiveness and helplessness in the family.

However, there is a hopeful note, according to the authors: later in the course of treatment, once-hostile family members may do much better in group sessions.