To the Editor Prominent etiologic models that aim to explain the comorbidity of mood disorders and alcohol dependence suggest that alcohol use may be a strategy of self-medicating distressing mood symptoms.1 This is line with studies that identified depressive and anxiety disorders as risk factors of the incidence of alcohol dependence.2 The current study of Crum et al3 provides convincing evidence that self-medication with alcohol indeed plays an important role in the incidence of alcohol dependence. However, we have some concerns regarding their analyses on the persistence of alcohol dependence.