Do the currently proposed DSM-5 criteria for anorexia nervosa adequately consider developmental aspects in children and adolescents?

The purpose of this article is to discuss the proposed criteria of the fifth version of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) for anorexia nervosa (AN) and to compare these with an alternative proposal which is based on a broader conception of the AN phenotype (Hebebrand and Bulik, in press). The proposed DSM-5 criteria seem to only insufficiently resolve the problems inherent to the current classification of AN because (1) the A criterion does not include a reference to allow the clinician to decide if the (young) patient meets the weight criterion, (2) the AN patient first must have evolved the cognitive capacity for complex abstract reasoning in order to fulfill the criteria B and C (Bravender et al. in Eur Eat Disord Rev 18:79-89, 2010), (3) physical symptoms of starvation including the neuroendocrine dysfunction characteristic of AN are not a diagnostic requirement, and (4) the subtypes are not helpful for classification of younger patients who almost all have the restricting type. On these grounds the proposed DSM-5 criteria will perpetuate the diagnostic tradition of a high percentage of patients who are subsumed under the diagnosis of eating disorders not otherwise specified (EDNOS), thus hampering both clinical practice and research. The use of our recently proposed alternative criteria for AN would result in most children and adolescents with an AN-like phenotype receiving a diagnosis of AN. Accordingly, our proposed criteria would be readily applicable to children, adolescents and adults.