Abstract

Depression is often assumed to be a heterogeneous condition that varies in symptom presentation. For instance, dual-diagnosed patients are more likely to experience changes in sleep and appetites along with apathy and anxious mood than those without substance use disorder (SUD) symptoms. If a substance worsens or relieves specific symptoms, symptom profiles may differ depending on the alcohol use status. The high prevalence of alcohol use disorders in the Korean population may greatly improve our understanding of its etiology. It is both timely and necessary to evaluate the BDI-II as a severity-rating tool across this population. This study compared the symptomatic presentation between patients with major depressive disorder (MDD) patients and comorbid alcohol abuse, and those without. However, the factor structure in a clinical sample of diagnosed patients has not yet been reported. Clinical interviews were conducted to screen for depression and other comorbid disorders. Measurement invariance was tested in Beck et al. (1996)’s SA-C model and Buckley et al. (2001)’s C-A-S model. Both models indicated a reasonable fit to the data; however, a two-factor model was selected for parsimony. The results indicated strict measurement invariance across groups, implying that the BDI-II could provide a consistent structure across patients with MDD regardless of alcohol use problems.