Despite preventative and treatment efforts, alcohol use disorders remain prevalent among adolescents, and available treatment outcome research among this group indicates that relapse rates are high. In this study, subjects will be adolescents diagnosed with an alcohol use disorder (n=100) who are participants in the Pittsburgh Adolescent Alcohol Research Center (PAARC) and their therapists. Subjects' attachment to parents and peers will be assessed and used to predict relapse following treatment. The extent to which adolescents are susceptible to peer influence, their ability to regulate negative affect, and the strength of the therapeutic alliance formed during treatment are each proposed as potential mediators in the relationship between attachment and relapse. A baseline evaluation, including structured interview for DSM-IV diagnoses and a measure of security of attachment to parents and peers, will be conducted within two weeks of outpatient treatment initiation. Adolescents' susceptibility to peer influence, their ability to regulate negative affect, and their therapeutic alliance formed with an outpatient therapist will be assessed after one month of outpatient treatment. Monthly telephone interviews assessing alcohol and drug use will then be conducted for one year post-baseline in order to create relapse event histories. Attachment security will be used to predict the time to lapse (return to alcohol use) and relapse (alcohol use-related consequences) in event history analyses using Cox proportional hazard model. Multiple regression analyses will be used to test the ability of attachment security to predict adolescents' use-related consequences and the frequency of their alcohol consumption. Finally, multiple regression analyses will be used to test the proposed mediators. If the mediational model is supported, mechanisms of influence will have been identified which can be targeted in order to enhance treatment strategies and reduce relapse rates among adolescents.
