Disordered eating behaviours

Jones and colleagues reported an alarmingly high prevalence of disordered eating behaviours in a community sample of adolescent girls.¹ This study is a valuable addition to the research literature on adolescent dieting. However, the language they used in describing their findings may be easily misinterpreted.

On the basis of the percentage of girls surveyed who scored above a cut-off score on the Eating Attitudes Test-26 (EAT-26),² the authors stated that disordered eating attitudes and behaviours were present in over 27% of girls aged 12–18 years. Although the results provide information about the percentage of teenaged girls who show unhealthy dieting behaviours and are at increased risk of developing eating disorders, they do not provide information about the prevalence of disordered eating.

The authors of a recent review, one of whom was one of the authors of the EAT-26, concluded that the predictive validity of this instrument is poor because the prevalence of eating disorders is low (1 to 3%).³ They recommended that the instrument not be used to establish the prevalence of disordered eating behaviours unless it serves as the first part of a 2-part diagnostic screen and the second part involves a clinical interview with high scorers.

I do not want to minimize the importance of the findings of Jones and colleagues, but they could have facilitated a more accurate interpretation of the results had they noted that the majority of girls who scored above the cut-off score of 19 on the EAT-26 may not actually have a disorder. The percentage of survey participants who score above the cut-off on a self-report screen cannot be equated with the prevalence estimate of a psychiatric disorder.

Frank Elgar Department of Psychology Dalhousie University Halifax, NS