Preventing Adolescent Alcohol Use: Processes and Outcomes of a Community-Based Intervention in Trelleborg

The aim of this thesis is to investigate the potential of applying a community-based intervention, implemented in the Swedish city of Trelleborg 1999-2002, to reduce adolescent alcohol use. In order to do so, we had four specific aims. First, we studied the processes, both in terms of action and institu-tionalization, within the intervention. This would improve our understanding of the quantitative analyses. Second, we investigated the effects of the intervention, based on a number of indicators related to the interven-tion objectives. Third, we investigated which risk factors were significant for adolescent alcohol use in Trelleborg. Fourth, we exam-ined whether the changes of the intervention's target indicators resulted in less alcohol-related accidents and violence among the Trelleborg youth. In the process evaluation (Paper I), we carried out a qualitative data analysis, apply-ing an iterative explanation-building process, using interviews, public records, meeting minutes and evaluation seminars. In Papers II-IV, we analyzed school survey data, apply-ing logistic regression modelling and other statistical analyses. The process evaluation showed that the community intervention did not adequately institutionalize alcohol and drug prevention within the city administration, mainly due to numerous un-planned events and that the structure was individual oriented. Neverthe-less, the community mobilization and the raise in public awareness regarding the alcohol issue are factors which could have contributed to the impact of the project. The effect evaluation (Papers II-IV) showed that the community-based interven-tion in Trelleborg is likely to have contributed to a reduction in adolescent alcohol consump-tion. When comparing the data, for a popula-tion 15 and 16 years old, from the first school survey in 1999 with the one carried out in 2003, there is a decrease in the proportion of alcohol consumers (decreased by 21.8%), monthly heavy episodic drinking (38.2%), excessive drinking (36.3%), alcohol-related accidents (38.5%), and alcohol-related vio-lence (50.0%). The data analyses show that these changes are independent of significant factors on community, group and individual level. Availability of alcohol was also reduced in the target group during the intervention, but this change was not sustained after the project completion. In addition, the analyses supported the hypothesis that the decrease in alcohol-related accidents and violence was related to the reduction in adolescent alcohol consumption. The findings in the thesis support the as-sumption that by implementing a city policy program, including a plan of action, a mu-nicipality can reduce adolescent alcohol use, and subsequently alcohol-related harm. How-ever, in order to institutionalize the interven-tions and the alcohol issue there has to be a balance of internal and external input throughout the project. It seems that the implementation process in terms of institu-tionalization and the effects of the interven-tion do not have to be correlated, i.e., a weak institutionalization does not necessarily result in weak effects