Do social networks explain 12-step sponsorship effects? A prospective lagged mediation analysis

Sponsorship is a basic and important part of the 12-step approach to recovery from substance abuse (Alcoholics Anonymous, 2005) and research has shown that having a sponsor is associated with increased involvement in 12-step programs and improved outcomes (Bond, Kaskutas, & Weisner, 2003; Tonigan & Rice, 2010). However, little is known about how sponsorship improves outcomes. Given research demonstrating bivariate associations between sponsorship and social support for abstinence (Majer, Jason, Ferrari, Venable, & Olson, 2002), we hypothesized that the association between having a sponsor and increased abstinence outcomes would be explained by increases in one's abstinence-based social network. Prospective fully lagged mediational analyses did not support this hypothesis and these results ran counter to findings of five previous studies (cf. Groh, Jason, & Keys, 2008). A review of these studies showed that researchers often used cross-sectional or partially lagged methods to test mediation and the mediational effect of the social network was small in magnitude. Results suggest that the prospective association between sponsorship and abstinence is not explained by increases in the abstinence-based social network and demonstrate the need for future studies to use rigorous and time-lagged methods to test social support for abstinence as a mediator of the effects of 12-step involvement.