This review found that the Maudsley model of family therapy was a promising treatment for adolescents with anorexia nervosa, but evidence for other forms of family therapy was limited and methodologically weak. The reporting and methodology used in this review make it difficult to assess the reliability of these conclusions.

To review the efficacy of psychosocial interventions for eating disorders in adolescents.

MEDLINE and PsycINFO were searched from 1985 to the date of the review and the search terms were reported. The authors stated that manual literature searches were conducted, but did not provide further details. The reference lists of articles were also searched.

Studies of psychosocial treatments for adolescents with eating disorders were eligible for inclusion if they used waiting list, standard care, placebo pills, or alternative treatments as controls, and reported treatment outcomes. Included studies assessed family therapy (separated, conjoint, or short- or long-term) compared with alternative treatment in younger (under 17 years) and older (17 to 21 years) adolescents. The proportion of females ranged from 88.9% to 100% and mean ages ranged from 14.2 to 20.9 years, most studies assessed adolescents with anorexia nervosa. A wide range of outcomes were reported including body mass index (BMI), weight change, body image disturbance, bulimic symptoms, and a composite score (Eating Attitudes Test, or Eating Disorders Inventory).

The authors did not report how many reviewers performed the study selection.

Studies were classified according to their design. Type 1 studies were RCTs with blinded assessment; clearly described inclusion and exclusion criteria, and statistical methods; state of the art diagnostic methods; and adequate sample size. Type 2 studies were clinical trials which did not meet some of these standards.

Data for studies of younger and older adolescents were extracted separately. Standardised effect sizes were calculated (Cohen's d) for both within-group effects (change between pre- and post-treatment), and between-group effects.

The authors did not report how the data were extracted for the review, nor how many reviewers performed the data extraction.

Details of the individual studies and their results were reported narratively. Effect sizes for each outcome were reported in a table, but no meta-analysis was performed.

Twelve studies providing results for adolescents were included (497 patients). Only two studies were classed as Type 1. Details of a further 50 studies providing results for adults were presented for comparison purposes.

Two studies found that the Maudsley model, or a system based on this model, of family therapy was superior to individual therapy for anorexia nervosa (between-group effect sizes for weight change of 0.53 and 1.03). All except one study reported greater gains in weight for family therapy compared with other treatment controls.

The evidence base for psychosocial treatments of eating disorders in children and adolescents was limited and due to methodological limitations it was not possible to recommend which treatments worked for whom in most cases. Family-based therapies were a promising future treatment avenue for adolescents with anorexia nervosa.

This review searched two relevant databases and the authors attempted to locate other studies through manual searches. It was unclear if there were any language restrictions and relevant studies may have been missed. The inclusion criteria were not clearly reported and nor were the review methods, such as how many reviewers selected studies and extracted data. Studies were classified according to their design and some details of methodological flaws were reported in the text, but very little detail was presented for the basic assessment of study validity. There was also little attempt to synthesise the evidence in a cohesive way.

The authors conclusions were appropriately cautious, but the poor reporting of this review makes it difficult to assess their reliability.

Practice: The authors stated that for adolescents with anorexia nervosa a family-based intervention, based on conjoint family therapy was shown to be superior to eclectic individual therapy, but this might not apply to older adolescents who may live away from home. For those with bulimia nervosa it was unclear if family therapy or cognitive-behavioural therapy guided self-care was the best treatment.

Research: The authors stated that future studies should increase the ethnic and racial diversity of study populations, include more men, and investigate treatments for the full range of possible eating disorders. They should also be of better quality, have large sample sizes, use manualised treatments, have state-of-the-art diagnostic criteria, and have good reporting of methods and analyses.

Supported by grants from the National Institute of Mental Health (R01 61836 and R01 63758).

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.