Research Project
10077808
Alcoholics Anonymous (AA) is one of the most popular resources for dealing with alcohol-related problems, and 12-step therapy (TS), based upon AA doctrine and practice, is one of the prevailing alcohol treatment approaches in the United States. Two large multisite trials, one high in internal validity (PMRG, 1997) and the second high in external validity (VA Substance Abuse Treatment Study; Quimette, 1997) came to the same conclusion, TS was equally effective as more research supported therapies, and may actually be superior when total abstinence is the treatment goal. A primary objective of TS is to facilitate AA affiliation and strong evidence suggests that this aim is a major factor accounting for the effectiveness of TS (e.g., Tonigan, 2005). High priority has therefore been assigned to the investigation of what actually occurs in AA, with a special focus on identifying prescribed AA behaviors and processes that are predictive of drinking reduction. The guiding assumption of these efforts is that the key to improve TS is to first understand AA better. To this end, this study will generate, for the first time, a comprehensive and definitive process model of AA-related behavior change. This objective will be realized through the highly innovative use of EMA data collection among early AA affiliates. Using real-time daily data, aim 1 will determine if four MOBC identified by AA researchers (gains in social support, increased abstinence self-efficacy, spiritual practices, and negative urgency) mediate the linkage between three types of AA prescribed behaviors and drinking outcome. Noteworthy, these analyses will include the first rigorous testing of six of seven of Kazdin's (2007) criteria to confirm (or reject) that these four statistical mediators are MOBC. Aim 2 will investigate whether the actions of the AA active ingredients on mediators (a path) and the actions of the mediators (b path) are constant over time or, alternatively, if there are critical periods of influence. Last, aim 3 will determine if the four MOBC operate differently across distinct subpopulations. To achieve study aims, we propose a two-group randomized longitudinal study (N = 190). In one group (n = 130) we will collect 6-months of continuous EMA data, allowing us to examine near real-time associations between AA active ingredients in three domains, four MOBC, and drinking. In tandem, we will also conduct in-person interviews at baseline, 3, and 6-months. Assessment reactivity is a concern, especially so because this will be the first study to use EMA in addition to in-person interviews in AA research. We will therefore include a traditional fixed assessment group (n = 60) also interviewed at baseline, 3, and 6-months to identify potential measurement biases introduced in our innovative approach. Achievement of study aims will generate the first empirically validated AA process model that will inform TS with critical information for improving treatment outcomes.
