|
Article Excerpt
AS THE FIELD OF ALCOHOLISM TREATMENT research moves beyond simple efficacy studies and the search for matching effects, increased attention has focused on the elucidation of mechanisms of change (Longabaugh et al., 2005). Although the 12-step approach is highly prevalent in the United States, and its efficacy compares favorably to other empirically validated forms of addiction treatment (Moos et al., 1999; Ouimette et al., 1998; Project MATCH [Matching Alcoholism Treatments to Client Heterogeneity] Research Group, 1997, 1998), there has been surprisingly little work examining mechanisms of 12-step recovery. The existing research indicates that increase in self-efficacy to sustain abstinence is a strong causal mechanism explaining later drinking reduction (Connors et al., 2001; Morgenstern et al., 1997). A previous analysis of Project MATCH data found relatively small, and marginally significant, bivariate correlations between Alcoholics Anonymous (AA) attendance and subsequent self-efficacy (DiClemente et al., 2001). Somewhat stronger relationships were found in structural equation modeling (SEM) analyses of both outpatient and aftercare samples, however, along with evidence that abstinence self-efficacy was a significant mediator of the effect of AA participation on distal percent days abstinent (Connors et al., 2001).
Even less is known about how mechanisms of change may differ between distinct subgroups of alcoholics. Numerous typologies of alcoholism have been proposed, with the goal of reliably identifying homogeneous types with discriminative, predictive, and construct validity (Epstein et al., 2002; Penick et al., 1999). These typologies are clinically useful to the extent that they predict the course of illness and response to various treatments. The A/B typology (Babor et al., 1992) has been studied extensively and was used in Project MATCH as a possible matching variable. Type-A alcoholics are characterized by later onset; fewer childhood behavioral problems; and lower severity in terms of physical dependence, alcohol-related problems, and psychopathology. Type-B alcoholics have more childhood behavioral problems; stronger family history of alcoholism; earlier onset of alcohol-related problems; and greater severity with respect to alcohol dependence, concomitant drug use, treatment history, psychopathology, and life stressors. This typology has been shown to be predictive of outcome, with worse outcomes in the more severe Type-B alcoholics (Babor et al., 1992; Yoshino and Kato, 1996). Type-A versus Type-B status has been found to predict differential response to selective serotonin reuptake inhibitor antidepressants (Dundon et al., 2004; Kranzler et al., 1996; Pettinati et al., 2000). The simpler age-of-onset typology (based on a cutpoint, usually 25 years, for age of onset of alcohol dependence) has also been found to predict differential response to ondansetron (Johnson et al., 2000).
Alcoholism severity (which overlaps with many typologies, including that of Type A/B) is a positive predictor of affiliation and engagement with 12-step programs (Tonigan et al., 1996). Although an earlier study found a significant matching effect of typology (Type-B alcoholics having better outcomes with cognitive-behavioral coping-skills treatment and worse outcomes with interactional therapy; Litt et al., 1992), Project MATCH failed to replicate this finding, and typology was found to have no significant treatment-matching effects (Litt and Babor, 2001). In their analysis of typological effects in Project MATCH, Litt and Babor found cognitive change, including changes in both self-efficacy and motivation, to be a strong predictor of substance-use outcomes (with the exception of Type-B alcoholics in the outpatient sample). AA attendance was not included in their analysis.
Given the systematic differences between types with respect to dependence severity, consequences, psychopathology, and genetic loading, it is not reasonable to assume that mechanisms of change are constant across types. If Type-B alcoholics respond better to cognitive-behavioral than to interactional therapy, this would suggest the relatively greater importance of skill acquisition and self-efficacy in this more severe group, perhaps due to its members' greater loss of control. If, however, Type-B alcoholics are more likely to participate actively in AA but no more likely to benefit from 12-step-oriented treatment, one explanation could be that self-efficacy has a weaker effect in the Type-Bs. It is also possible that the relationships between typology, AA attendance, and self-efficacy could vary by treatment assignment. One of the proximal goals of twelve-step facilitation (TSF) is attendance of AA meetings. Thus, we might expect that assignment to TSF and AA attendance would interact positively to predict increased self-efficacy, because the positive meaning ("success") of AA attendance should be greater for those receiving TSF.
In a sample of 115 substance-abuse treatment participants (including 32 with drug but no alcohol abuse), Morgenstern et al. (1998) found lower baseline self-efficacy in Type-B substance abusers. Lower self-efficacy during treatment did not appear to account for poorer outcomes in Type-Bs, however....
|
