This study was limited by the fact that DSM-IV and DSM-5 AUD were classified in largely overlapping rather than independent populations. As a result, comparisons of the clinical profiles of the disorders required partially split sample analyses that reduced the statistical power to discern differences in the profiles. Moreover, many highly relevant aspects of clinical course could not be addressed in this study. For example, questions on age at onset of AUD in the NESARC were asked with respect to the symptoms that defined DSM-IV AUD and could not be extrapolated to the corresponding DSM-5 AUD. Similarly, the questions that as certained chronological clustering of symptoms, necessary to establish a diagnosis for disorders prevalent in first two years of the follow-up interval, could not be extrapolated to the DSM-5 disorders. Because we were unable to create valid measures of DSM-5 AUD for the earlier time period, we were unable to compare the course of AUD (i.e., chronicity, remission, progression to more severe AUD) under the DSM-IV and DSM-5.
