Alcohol consumption questions revealed that proportionally fewer older than younger subjects were currently drinking, in line with other studies (Vaillant, 2003) and supportive of the first part of Hypothesis 3. The prevalence of current drinking among COGA subjects aged 65+ (46%) was somewhat lower than that reported by same-aged population-sample subjects in NESARC-III (55.2%; Grant et al., 2017), CBHSC (56%; Center for Behavioral Health Statistics and Quality, 2015), and the National Health Interview Surveys (approximately 57%; Breslow et al., 2017). These differences may be due to several factors. First, a number of our historically alcohol dependent subjects may have adhered to the widespread treatment goal of total abstinence, which leads to a more stable recovery pattern than moderate drinking (Vaillant, 2003). Second, almost one-quarter of our target sample had died since their baseline interview, thereby selectively removing some of the most severely alcohol dependent individuals. Finally, 40% yet to be located subjects had a relatively severe addiction history, further contributing to the underrepresentation of highly affected individuals. Although fewer COGA subjects reported current drinking, a dramatically higher proportion of those
