Article Text

Download PDFPDF

Family therapy improved outcomes in patients with early onset and short history anorexia nervosa
  1. Blake Woodside, MD, MSc, FRCPC
  1. University of Toronto Toronto, Ontario, Canada

Statistics from

OpenUrlCrossRefPubMedWeb of Science

Question Compared with individual supportive therapy, can family therapy for 1 year improve the outcome at 5 years for patients who have been treated in hospital for anorexia nervosa and bulimia nervosa?


Randomised controlled trial with 5 years follow up.


A hospital eating disorders unit in London, UK.


80 patients (age range 14 to 55 y) who had anorexia nervosa (n=57) or bulimia nervosa (n=23) and were discharged from hospital after treatment. Patients were classified into 4 groups according to the features of their illness: early onset (≤18 years of age) and short history (<3 years) of anorexia nervosa; early onset and long history (≥3 years) of anorexia nervosa; late onset of anorexia nervosa (>18 years of age); and bulimia nervosa. Follow up was 91%.


Patients in each group were allocated to individual or family therapy: 11 and 10 patients for early onset and short history of anorexia nervosa; 9 and 10 patients for early onset and long history of anorexia nervosa; 7 and 7 patients for late onset of anorexia nervosa; and 10 and 9 patients for bulimia nervosa.

Main outcome measure

A good outcome was defined as having a body weight within 15% of the average body weight and regular menstrual cycles.

Main results

In patients who had early onset and a short history of anorexia nervosa, those who received family therapy were more likely than those who received individual therapy to have a good outcome at 5 years (p≤0.02) (table). In patients with late onset anorexia, a non-significant trend towards a better outcome at 5 years existed for patients who received individual therapy compared with those who received family therapy (57% v 29%). There were no differences between treatments for the number of good outcomes in patients who had bulimia nervosa or those with early onset and a long history of anorexia nervosa.

Family v individual therapy in patients with early onset and short history of anorexia nervosa


For patients with early onset and a short history of anorexia nervosa, family therapy led to more patients having good outcomes (body weight within 15% of average body weight and regular menstrual cycles) at 5 years than individual therapy.


This study by Eisler et al expands on a critical investigation in the late 1980s which examined the efficacy of 2 treatments for patients who have anorexia nervosa and bulimia nervosa.1 Family therapeutic techniques had not previously been subjected to rigorous scrutiny; the results of the original study were one of the first pieces of empirical evidence supporting the use of family therapy for these illnesses.

Briefly, the original study showed that family therapy was the preferred treatment after leaving hospital for patients with anorexia nervosa who were ≤18 years of age and had a short duration of illness. For older patients, individual therapy led to a better outcome.

The current report with 5 years of follow up confirms that the observed improvements have persisted. The overall recovery rate at 5 years is about one third of patients, with an additional one fifth partially recovered, and nearly 40% still quite ill. Although the news of better outcomes for family therapy in younger patients with early onset and a short history of anorexia nervosa is encouraging, the rates of serious illness at 5 years are of concern, as is the nearly 5% mortality rate.

For clinicians, the most important issue is the choice of treatment. It must be noted that the investigators do not recommend family or individual therapy as an initial outpatient treatment. These are recommendations for treatment after the patient has been nutritionally rehabilitated and has had weight restored. Clinicians who intend to do work of this nature should carefully review the descriptions of the treatments.

Finally, the authors correctly conclude that a less favourable outcome, which was associated with additional treatment, was likely a marker for more severe or treatment resistant illness. Given that chronicity may be a marker for treatment resistance,2 clinicians must provide patients with anorexia nervosa with the best possible treatment as early as possible in the course of the illness.



  • Source of funding: Medical Research Council of Great Britain.

  • For correspondence: Dr G F M Russell, Hayes Grove Priory Hospital, Prestons Road, Hayes, Kent BR2 7AS, UK. Fax +44 (0)181 462 5028.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.