In contrast to the high level of concordance between DSM-IV alcohol dependence and DSM-5 severe AUD, (80.5%) there was a considerably lower level of concordance between DSM-IV alcohol abuse and DSM-5 moderate AUD (58%). Discrepancies reflect both the criteria upon which the two disorders are based – which include core characteristics of AUD such as tolerance, withdrawal, craving and impaired control for moderate AUD but not abuse – and the requirement of two positive criteria for moderate AUD compared to one positive criterion for abuse. When individuals with DSM-5 moderate AUD were compared to those with DSM-IV abuse, there were striking reductions in the proportions of Whites and males and a striking increase in the proportion of low-income individuals, reflecting gender, race/ethnic and income disparities (Harford et al., 2009; Kahler and Strong, 2006; Keyes and Hasin, 2008; Saha et al., 2006) in the endorsement of hazardous use, which was sufficient in itself to establish a diagnosis of DSM-IV abuse but not DSM-5 moderate AUD. These findings closely mirrored those of Agrawal et al. (2011), reflecting the fact that all the
