Reliability of self-reported eating disorders: Optimizing population screening

The objective of this study was to assess whether short self-report eating disorder screening questions are useful population screening methods.

We screened the female participants (N = 2881) from the 1975-1079 birth cohorts of Finnish twins for eating disorders, using several short screening questions and three Eating Disorder Inventory (EDI) subscales. Comparing these measures with clinician-conducted semi-structured diagnostic interviews (N = 549) of Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) anorexia and bulimia, we calculated their sensitivities and specificities and drew receiver operating characteristic curves to further compare these items.

For current and lifetime bulimia, best tradeoffs between sensitivity and specificity were reached by addressing purging behaviors. For current and lifetime anorexia, the questions "Have you ever had anorexia" and "Has anybody ever suspected that you might have an eating disorder?" optimized tradeoffs between sensitivity and specificity. These questions generally outperformed EDI subscales.

Simple screening questions, although less than ideal, are at least as good as other available instruments for community screenings.