Many digital interventions for unhealthy alcohol use are based on personalized normative feedback (PNF) and personalized feedback on risks for health (PFR). The hypothesis is that PNF and PFR affect drinkers' perceptions of drinking norms and risks, resulting in changes in drinking behaviors. This study is a follow-up mediation analysis of the primary and secondary outcomes of a randomized controlled trial testing the effect of a smartphone-based intervention to reduce alcohol use.
This study aimed to investigate whether perceptions of drinking norms and risks mediated the effects of a smartphone-based intervention to reduce alcohol use.
A total of 1770 students from 4 higher education institutions in Switzerland (mean age 22.35, SD 3.07 years) who screened positive for unhealthy alcohol use were randomized to receive access to a smartphone app or to the no-intervention control condition. The smartphone app provided PNF and PFR. Outcomes were drinking volume (DV) in standard drinks per week and the number of heavy drinking days (HDDs) assessed at baseline and 6 months. Mediators were perceived drinking norms and perceived risks for health measured at baseline and 3 months. Parallel mediation analyses and moderated mediation analyses were conducted to test whether (1) the intervention effect was indirectly related to lower DV and HDDs at 6 months (adjusting for baseline values) through perceived drinking norms and perceived risks for health at 3 months (adjusting for baseline values) and (2) the indirect effects through perceived drinking norms differed between participants who overestimated or who did not overestimate other people's drinking at baseline.
The intervention's total effects were significant (DV: b=-0.85, 95% bootstrap CI -1.49 to -0.25; HDD: b=-0.44, 95% bootstrap CI -0.72 to -0.16), indicating less drinking at 6 months in the intervention group than in the control group. The direct effects (ie, controlling for mediators) were significant though smaller (DV: b=-0.73, 95% bootstrap CI -1.33 to -0.16; HDD: b=-0.39, 95% bootstrap CI -0.66 to -0.12). For DV, the indirect effect was significant through perceived drinking norms (b=-0.12, 95% bootstrap CI -0.25 to -0.03). The indirect effects through perceived risk (for DV and HDD) and perceived drinking norms (for HDD) were not significant. Results of moderated mediation analyses showed that the indirect effects through perceived drinking norms were significant among participants overestimating other people's drinking (DV: b=-0.17, 95% bootstrap CI -0.32 to -0.05; HDD: b=-0.08, 95% bootstrap CI -0.15 to -0.01) but not significant among those not overestimating.
Perceived drinking norms, but not perceived risks, partially mediated the intervention's effect on alcohol use, confirming one of its hypothesized mechanisms of action. These findings lend support to using normative feedback interventions to discourage unhealthy alcohol use.
ISRCTN Registry 10007691; https://doi.org/10.1186/ISRCTN10007691....

Smartphone-Based Intervention Targeting Norms and Risk Perception Among University Students with Unhealthy Alcohol Use: Secondary Mediation Analysis of a Randomized Controlled Trial

10.2196/55541