Are There Gender, Racial, or Religious Denominational Differences in Religiosity's Effect on Alcohol Use and Binge Drinking Among Youth in the U.S.? a Propensity Score Weighting Approach

 

Background and Purpose: Alcohol use/misuse is a prevalent health issue among youth and may lead to adverse consequences. Religiosity has been identified as a protective factor against alcohol use/misuse among youth. Identifying moderators in the religiosity-alcohol relationship has important implications for intervention development. If moderation effects are present, preventions/interventions should be adapted to specific youth subgroups rather than taking a “one-size-fits-all” approach. Theories and previous research suggest that gender, race, and religious denomination may moderate the religiosity-alcohol relationship. However, extant studies suffer from various limitations that prevent more informative conclusions to be drawn. This study focuses on two research questions: (a) Is the effect of religiosity on alcohol use and binge drinking among American youth conditional on gender, race, and/or religious denomination? and (b) In the event that gender, racial, or denominational differences exist, what are the average treatment effects (ATEs) in the sub-populations (i.e., male and female, White and non-White, Protestant and non-Protestant)?

Methods: This study overcame previous studies’ limitations by using longitudinal data from the National Study of Youth and Religion, a nationally representative telephone survey, and robust analytical methods (N=1,969). This study used the propensity score weighting method to control for a large number of confounders. Propensity score weighting approach allows observational studies to be designed similar to randomized experiments by reducing systematic differences in observed covariates between treated (i.e., high religiosity group) and control subjects (i.e., low religiosity group). Propensity score weights were estimated nonparametrically using Generalized Boosted Models, a machine learning method that affords advantages over logistic regression.

Results: Findings indicate that gender and religious denomination might not moderate the religiosity-alcohol relationship, whereas racial difference was present. Among White youth, religiosity was found to have a buffering effect against alcohol use (average treatment effect [ATE]= -0.57, CI_.95 = -0.83, -0.32) and binge drinking (ATE= 0.54, CI_.95 = 0.38, 0.71). However, among non-White youth, religiosity was not found to have an effect on alcohol use (ATE= 0.08, CI_.95 = -0.31, 0.47) or binge drinking (ATE= 1.07, CI_.95 = 0.68, 1.64).

Conclusions and Implications: Findings about the absence of gender or denominational differences in the religiosity-alcohol relationship suggest that preventions/interventions of youth alcohol involvement that are religiously based or sensitive may not need to adapt their efforts based on youth’s gender or religious denominations. In light of the finding that racial difference was present in the religiosity-alcohol relationship and religiosity had a buffering effect against alcohol involvement among White youth but not among non-White youth, religiously based interventions may be more effective for White youth than non-White youth. In addition, when addressing alcohol use/misuse issue among White American youth, religious and faith-based organizations, youth religious group leaders, and clergy should be included in the prevention/intervention efforts.