Drug use is often a social activity that occurs in an environment with other users (Fountain & Korf, 2007). In his classic study How to become a marihuana user, Becker (1963) argues that continued drug use is typically the result of social learning. In addition, Zinberg (1984) states that the effects of drugs experienced by users are influenced by three intertwined factors (drug, set, and setting), the latter including informal social control. The relevance of this social control lies in minimizing the harm of drug use, thus resulting in controlled intoxicant use. Violating the appropriate informal rules and norms regarding drug use within a group of users leads to social sanctions, and violators might eventually be excluded from the group. Alternatively, rituals and elements of social settings prevent uncontrolled drug use like dependence (Dunlap, Johnson, Sifaweck, & Benoit, 2005; Goode, 1999; Zinberg, 1984) and this explains why many cannabis users carefully choose when, where and with whom to use the drug (Reinarman & Cohen, 2007). Consequently, processes of social inclusion and exclusion may play an important role in drug using careers (cf. Vervaeke, van Deursen, & Korf, 2008).
In a wider context, Glueck & Glueck (1950) argue that, in line with the saying "birds of a feather flock together," once deviant, juveniles are more likely to associate with deviant peers. In his classic theory of differential association, Sutherland (1947) stated that deviant behavior is learned through association and interaction with other delinquents, especially in small, informal peer groups. Akers (1998), who further explored and refined Sutherland's theory, found that the probability of frequent substance use increased when individuals in their social networks were more often exposed to favorable than to unfavorable definitions of use, including cannabis use.
A crucial question in the current study is the extent to which frequent cannabis use and cannabis dependence is an important unifying factor in peer networks. Is cannabis the "feather" that makes users flock together? In line with Zinberg' s theory, it could be argued that cannabis dependence indicates less controlled use, and thus a violation of the social norms of drug use, which then leads to social exclusion from social networks of frequent but not dependent users. The question then is, whether cannabis dependent users become socially isolated (exclusion), or tend to congregate in social networks of dependent users (differential inclusion). Alternatively, it could be argued that frequent cannabis use by itself is already a violation of the social norms of controlled use, and therefore dependence will not lead to social exclusion from frequent but nondependent users. The question then arises of which other factors might bond social networks of frequent cannabis users. The main purpose of the current study is to explore the role of social exclusion and inclusion, by analyzing social networks within the total sample of 600 frequent cannabis users, who were recruited through respondent-driven sampling, by exploring the role of cannabis use - cannabis dependence in particular - and sociodemographic variables in peer associations in high-risk and dependent cannabis users. The relevance of these issues lies mainly in methodological strategies and insights for sampling in future research.
Internationally, high rates (11-13%) of cannabis dependence have been found in cohort studies among young adults (Boden, Fergusson, & Horwood, 2006; Noack, Höfler, Gründler, Schulz, & Paul, 2009) but longitudinal research on risk and protective factors for cannabis dependence - especially as regards the transition from regular use to dependence - is sparse. A major problem in previous studies targeting cannabis dependence is that general population surveys observe only small numbers of subjects with a diagnosis of cannabis dependence. For example, in the most recent German general population survey on substance use, only 113 of 7979 (1. …