DSM-IV versus DSM-5: implementation of proposed DSM-5 criteria in a large naturalistic database

Abstract

Objective: Problems with the current DSM-IV eating disorder (ED) section have resulted in proposed changes toward the upcoming DSM-5 (http://www.dsm5.org/ProposedRevisions/Pages/EatingDisorders.aspx). We investigated consequences of these by implementing the proposal in a large naturalistic database.

Method: Patients were 2,584 children/adolescents and adults enrolled at specialized ED clinics in Sweden. DSM-IV diagnoses anorexia nervosa, bulimia nervosa, and "not otherwise specified" examples were compared with DSM-5 anorexia, bulimia, and binge ED, as well as atypical anorexia, subthreshold bulimia, and binge eating, purging disorder, and the residual unspecified category. Assessment methods included a semistructured diagnostic interview and self-ratings of ED and psychiatric symptoms.

Results: We studied age-separated diagnostic distributions and explained variance in clinical variables associated with the two systems. Results showed some improvement of diagnostic specification as well as a slight increase in explained variance.

Discussion: Remaining problems with the proposed changes were also highlighted, and possible further refinement is discussed.