The association of abstinence in one family member with abstinence in another stands in contrast to a host of null findings regarding familial influences on remission from other studies in population-based (11, 13), high-risk, and clinical samples (12, 14, 51, 52) using a variety of definitions of remission. The current analyses used an explicit abstinent and non-abstinent remission phenotype, distinct from AUDs and consistent with the idea that the distribution of risks for development of, and for remission from, AUDs may not lie on the same continuum (15). Our results suggest that there may be genetic or familial environmental influences on abstinent remission and demonstrate that departing from the more common risk-factor-to-remission comparisons within families may indeed prove useful. When remission is the target phenotype, remission in all family members should be measured explicitly, rather than measuring it as an outcome only in target subjects but not in their relatives. This will facilitate the examination of potentially heritable characteristics underpinning abstinent outcomes, such as social responsiveness, that may increase the likelihood of remission, as well as the investigation of family
