Despite the FDA approval of naltrexone and the anticipated approval of acamprosate for the treatment of alcohol dependence, other pharmacological options are needed, particularly those with demonstrated acceptability to prescribers, if medications are to play an important role in alcoholism treatment. Selective serotonin reuptake inhibitors (SSRI' s) are the most widely prescribed psychotropic medications in the world. A variety of SSRI's have been evaluated as potential adjuncts to the psychosocial treatment of alcohol dependence. Many studies have shown positive effects of these medications in the treatment of alcohol dependence, though a substantial number of studies have failed to provide evidence of efficacy. In a patient-treatment matching study, we found a significant interactive effect of alcoholic subtype and fiuoxetine on drinking outcomes. Using a similar approach, Pettinati et al. observed an interaction of alcoholic subtype with sertraline, with a significant advantage for sertraline among later-onset, Type A alcoholics. Johnson et al., in a placebo-controlled study of the 5-HT_ antagonist ondansetron, found a dose-response relationship between the medication and reduced drinking in early-onset, but not late-onset alcoholics. Together, these findings help to explain variable findings in the literature on the use of serotonergic medications to treat alcohol dependence and provide the basis for evaluating a differential therapeutic approach using these medications. The present proposal, which has been revised in response to comments by reviewers, is to conduct a prospective, randomized, double-blind test of the hypothesis that sertraline is efficacious among late-onset alcoholics (i.e., those with alcoholism onset >25 years), but is not superior to placebo among early-onset alcoholics (i.e., those with alcoholism onset <25 years). This hypothesis will be evaluated in a 14-week treatment trial in which a total of 160 alcohol-dependent outpatients (80 early-onset alcoholics and 80 late-onset alcoholics) will receive either sertraline (to a maximum of 200 mg/day) or matching placebo. Urn randomization will serve to balance the two patient groups in relation to medication assignment. Daily process measures of positive and negative events, mood, desire to drink, and drinking frequency and quantity will be used to provide insight into the mechanisms by which sertraline may exert its effects in early-onset alcoholics. A 6-month post-treatment follow-up period will evaluate the duration of medication effects. Finally, for both theoretical and clinical reasons, the study will aim to refine the approach to matching by comparing the utility of a univariate (i.e., age of onset) approach to subtyping with that of a cluster analytic (i.e., Type A/B) approach. The study has important implications for the treatment of alcohol dependence, both because of the potential advantage of sertraline over placebo among early-onset alcoholics, but also because it involves a medication that is familiar to a very large number of practitioners. Consequently, demonstration of sertraline's efficacy in alcoholism can serve as a bridge for these practitioners to address alcohol use disorders in practice, an enduring problem in the effort to promote alcoholism identification and intervention by physicians.
