A Path Model of Discrimination, Social Integration, Social Support, and Substance Use for Asian American Adults Open Access
Downloadable ContentDownload PDF
The current research was a secondary analysis of the cross-sectional 2002-2003 National Latino and Asian American Study (NLAAS) (Alegria, Jackson, Kessler, & Takeuchi, 2007). Multi-stage area probably sampling with core and high-density sub-samples yielded a final weighted response rate of 65.60% and an adequate sample size of 2,095. Participants were Asian Americans ages 18 or older in households in the 48 contiguous states, Hawaii, and District of Columbia. Guided by theory, ordinal logistic and logistic regressions, moderation and mediation testing (Baron & Kenny, 1986), and path analyses were conducted using Stata Version 10.0 (StataCorp, 2007a), which enabled adjustment for complex sampling, and LISREL Version 8.80 Student Edition (Jöreskog & Sörbom, 2008). The dependent variables were level of drinking (ordinal with abstaining/non-risky drinking/risky drinking) and drug use (dichotomous with drug use/no drug use). Friend confidant support had a positive relationship with level drinking; it provided complete moderation between everyday discrimination and level of drinking for females. Rates of past year non-risky drinking, risky drinking, and drug use decreased across males, females with high friend confidant support, and females with low friend confidant support. Level of drinking had a positive association with drug use for all three groups. Everyday discrimination had positive associations with level of drinking and drug use for males and females with high friend confidant support; level of drinking mediated the relationship between everyday discrimination and drug use. Family cultural conflict had positive associations with level of drinking and drug use for males and females with low friend confidant support; level of drinking mediated the relationship between family cultural conflict and drug use. Family pride had a negative association with drug use for females with low friend confidant support. Ethnic identity had a negative association with drug use for males and negative effect on level of drinking but not drug use for females with high friend confidant support. Participants were more likely to engage in non-risky or risky drinking and drug use if they were Filipino, higher generation, younger, and never married. Future research should further examine protective factors against substance use, such as acculturation and ethnic identity.