Alcoholism: Why and How People Change
Most alcohol treatment researchers and clinicians agree that treatment is effective (Miller, Walters, & Bennett, 2001). Yet, very little is known about the mechanisms by which treatment is effective (Morgenstern & Longabaugh, 2000), and even less is known about how individuals change without treatment (Matzger, Kaskutas, & Weisner, 2005; Sobell, Ellingstad, & Sobell, 2000).
Matzger et al. (2005) addressed the question of why people change through conducting a quantitative analysis of ‘reasons for drinking less’ in a sample of 659 problem drinkers, including 239 adults who did not receive treatment in the past 12 months and 420 adults who received some form of public or private alcohol treatment. Problem drinking was defined as affirmative responses to two of the following: one binge episode at lease once a month, alcohol-related social consequences, and one or more alcohol dependence symptoms. At 1, 3, and 5 years following a baseline interview, the respondents who indicated they were drinking ‘a lot less’ were provided with a list of potential reasons for why they would be drinking less. In both samples (treatment and no treatment), logistic regression analyses predicting sustained remission from problem drinking showed that the odds of remitting were highest in individuals who identified ‘hitting rock bottom’, traumatic events, and spiritual/religious experiences as reasons for cutting down. In the no-treatment sample, individuals who reported ‘receiving a spouse's warning about their drinking’ and those who reported weighing pros and cons of drinking had lower odds of being non-problem drinkers at follow-up. In the treatment sample, individuals who reported receiving a warning from their doctor and weighing the pros and cons of drinking had lower odds of being in remittance. The results from Matzger et al. (2005) suggest that ‘quantum changes’ (Miller & C’de Baca, 2001) greatly increase the odds of an individual remitting from problem drinking, and interventions performed by family members or medical doctors are negatively related to positive outcomes.
Witkiewitz (2005) and Witkiewitz and Masyn (2004) conducted a thorough analysis of how people change following treatment using latent growth mixture modeling, an analytic strategy that estimates common patterns in individual trajectories. Three hundred and ninety-five individuals were assessed monthly on measures of drinking frequency and quantity for 12 months following treatment. The results supported a model with four common drinking trajectories following an initial lapse. The most common outcome trajectory (64% of the sample) was characterized by an initial lapse followed by a return to abstinence or moderate drinking. Only 12% of the sample reported a stable heavy drinking pattern following the initial lapse. Individuals with the heaviest drinking trajectories were unique from all other drinkers in that they had significantly lower scores on measures of coping and self-efficacy and significantly higher scores on measures of negative affect and distal risks.
Interestingly, many individuals in the Witkiewitz (2005) analysis, particularly those with higher scores on measures of negative affect and distal risk, had very turbulent drinking trajectories characterised by nearly 100% abstinence in one month, followed by nearly 0% abstinence in the next month. This pattern of results and the analyses described by Matzger et al. (2005) point to a relationship between risk factors and drinking behavior that is highly nonlinear. In the Matzger analysis, the participants identified ‘hitting rock bottom’, traumatic events, and quantum changes (Miller & C’de Baca, 2001) as the reasons for remittance from problem drinking.