Controlled Drinking: More Than Just a Controversy

Michael E. Saladin; Elizabeth J. Santa Ana


Curr Opin Psychiatry. 2004;17(3) 

In This Article


Of the CD strategies we reviewed, BSCT has the most extensive history of evaluation in controlled clinical trials. Results of studies conducted in the past 3 years echo those conducted over the preceding 25 years; BSCT is a highly efficacious treatment for alcohol-related problems. Recent studies suggest that naltrexone may boost the efficacy of BSCT, especially in heavier drinkers, and that computer-based versions of BSCT may substantially increase the accessibility and cost-effectiveness without sacrificing efficacy when delivered by a therapist. More generally, the Internet accessible DCU intervention and the on-line forums of Moderation Management have the potential to greatly increased access to moderation-oriented approaches to problematic alcohol use. The Sobells' community-level self-change intervention, delivered via mail, represents yet another attempt to increase accessibility to moderation-friendly services.

MOCE is one of the newest developments in CD. All three extant controlled trials involving MOCE have employed BSCT as a comparison treatment. Overall, both treatments were found to produce substantial reductions in drinking over a 6-8-month follow-up. Although few differential treatment effects were identified, it appears that MOCE might yield greater reductions in alcohol consumption when delivered in a group format. Since most treatment services in the US are delivered in a group format, this finding could have considerable practical significance by guiding decision making about optimal treatment type-format combinations. Another potential differential treatment effect was that BSCT might be more effective than MOCE with heavier drinkers. This, together with the observation that naltrexone may enhance BSCT's effectiveness with heavy drinkers, suggests that naltrexone combined with BSCT, rather than MOCE, might be the optimal treatment combination for individuals who have a severe drinking problem and are seeking a moderation goal. It is also noteworthy that one of the MOCE studies identified a strong association between drinking behavior at 6-month follow-up and a self-efficacy measure assessing an individual's ability to control their drinking obtained 6 months earlier. One potential interpretation of this association is that a CD treatment will be effective to the extent that it enhances an individual's confidence in their ability to control drinking.

The severity of dependence hypothesis[93] states that the more problematic a person's alcohol use is, the less likely they will be able to control their drinking. While this hypothesis has garnered support in previous reviews of the literature,[94] Walters' more recent review[29] has questioned the strength of this conclusion. His review of 17 controlled clinical trials of BSCT found no evidence of an association between severity of drinking problem and controlled drinking. Moreover, the present review of the BSCT and MOCE literature failed to identify one study confirming this association. In fact, all three of the reviewed controlled trials comparing MOCE with BSCT specifically reported no relationship between severity of alcohol use and drinking outcome. As already noted, the best overall predictor of outcome was a measure of drinking-related self-efficacy. It might be the case that understanding drinking outcomes in controlled-drinking studies, both short and long term, might be better served by monitoring changes in efficacy over the course of treatment rather than focusing on static pretreatment factors like drinking severity.

GSC is a relatively new moderation-oriented intervention that has received considerable empirical validation. A recent study examining the contribution of social support to GSC reported significant changes in drinking outcome, none of which were uniquely associated with social support. Another study reported drinking outcomes that non-significantly favored GSC (four sessions) over a single session of advice combined with a self-help manual. Importantly, study participants preferred GSC to the briefer intervention. The last of three recent studies was unique in that it compared the effects of two community level mail-out, GSC-based interventions for problem drinking. Significant reductions in drinking-related outcomes at 1-year follow-up were reported for both treatments. One of the greatest strengths of this study was that it showed that an efficacious problem drinking intervention could be delivered inexpensively to a large number of individuals. A logical extension of this community-based study would be to examine the effectiveness of an Internet/PC-based version of GSC that problem drinkers could access on-line.

The problem of alcohol misuse in adolescents and college students is nearly epidemic in proportion. The application of harm reduction approaches to school-age children, adolescents and college students have been shown to significantly impact both drinking behavior and alcohol-related problems. In high-risk college students, it appears that the impact of brief interventions on drinking behavior per se may diminish over extended periods of time (i.e. 4 years). While 'booster' sessions every year or two might enhance maintenance of reduced drinking behavior, it would be interesting to determine if low-risk college peers could be trained to aid in the delivery of booster sessions (i.e. developing a buddy system). Not only could this substantially reduce the cost of delivering the booster sessions, but it also might enhance the effectiveness of the program because college peers who are role models for moderate drinking might be perceived as more credible, trustworthy and understanding than individuals from outside the student body. As already noted, the Internet/PC-based DCU intervention has the potential to greatly enhance the cost effectiveness and availability of controlled drinking. Once the efficacy of DCU has been firmly established, it might be potentially valuable to determine if a modified version of the application could be used to reduce alcohol misuse on US college campuses.

We are encouraged by some important developments in the last few years that have promise with regard to enhancing the utility and accessibility of this general class of interventions. While the overall tone of the present review is one of optimism, the stormy past of the controlled drinking treatment approach continues to have a disruptive effect on the social, political and economic factors that impact contemporary treatment research and service delivery. Searles[95] has noted that challenges to conventional wisdom, like controlled drinking, tend to attract harsh opposition that is remarkably persistent even in the face of compelling scientific evidence. He refers to this type of opposition as a 'kind of scientific fascism'. Both the scientific community and society in general appear to have an affinity for this kind of 'fascism', and this affinity has not abated with the passage of time. To illustrate, a 1998 meta-analytic study published in a respected psychological journal[96] attracted a litany of criticism including the first-ever, unanimous congressional vote to condemn the primary findings of a scientific publication.[97] The article questioned the strength of association between self-reported history of child sexual assault and psychopathology. Among other things, the authors were accused of trying to normalize pedophilia. Regardless of the belief/value being challenged, it is likely that the machinery of science will have to work overtime to ensure that reason and truth are not obscured by an impassioned 'herd mentality'. At least in the case of controlled drinking, the persistent efforts of clinical science have resulted in a significant expansion of treatment options for persons who misuse alcohol.