Controlled drinking : a viable treatment goal in alcohol use disorder?

Alcohol use disorders (AUD) has one of the largest treatment gaps among psychiatric diagnoses, with a treatment coverage of 10 to 20 %. One of the key contributing factors is the lack of treatments aiming for controlled drinking (CD) instead of abstinence. Although a large number of trials investigated CD as the outcome, there are major limitations to the existing studies, such as small sample sizes, non-adequate control conditions, and heterogenous definitions of CD outcomes. Further, very few studies have investigated clinically relevant predictors of outcomes specifically in CD in sufficiently large sample sizes. Efficacy studies have been the main focus in research on CD and very few studies have investigated patient perspectives on the treatments offered for a CD goal. Lastly, there are few validated clinically relevant measures for the assessment and evaluation of impaired control over alcohol consumption, which is predictive of outcome in treatment for controlled drinking. The primary aim of the thesis was to investigate if a CD goal was viable in a treatment seeking sample of individuals of patients with AUD. This aim was broken down in the four following studies. Study I was a randomized controlled superiority trial including 250 individuals with alcohol use disorder. We hypothesized that Behavioral Self-Control Training (BSCT), a five-session cognitive and behavioral treatment, would be superior to Motivational Enhancement Therapy (MET), which was a four-session treatment based on Motivational Interviewing in reducing weekly alcohol consumption. Linear mixed models were used to analyze primary and secondary outcomes alongside with a Bayes factor analysis for the primary outcome. No differences were identified between groups for the primary outcome of mean weekly alcohol consumption at 26 weeks (primary endpoint). The secondary outcome proportion of weeks with hazardous drinking defined in line with the former Swedish low-risk drinking recommendations of >9/>14 weekly standard drinks was found to be statistically different between groups. Study II was a prospective cohort study, in which we investigated differences at 52 weeks post inclusion between BSCT and MET for both alcohol consumption and related consequences as well as predictors of treatment outcome. Linear and logistic mixed regression models were used for the outcomes at 52 weeks, and linear and logistic regression models for the predictor analyses. BSCT was superior to MET for the chang