Religious involvement and racial disparities in opioid use disorder between 2004–2005 and 2012–2013: Results from the National Epidemiologic Survey on Alcohol and Related Conditions

BACKGROUND: Psychosocial factors have rarely been studied to understand racial differences in opioid use disorders (OUD). We investigate religious involvement and Black-White differences in OUD risk between 2004–05 and 2012–13. METHODS: We use Non-Hispanic Black and White adults from the National Epidemiologic Survey on Alcohol and Related Conditions (wave 2, N = 26,661 and NESARC-III, N = 26,960) (NESARC). We conducted survey-weighted logistic regression to examine whether race moderates the association between religious involvement and lifetime DSM-IV and −5 OUD and whether those differences change (i.e., are modified) by time, adjusted for covariates such as age, education, and urbanicity. Religious involvement measures were service attendance, social interaction, and subjective religiosity/spirituality. RESULTS: The prevalence of lifetime DSM-IV (3.82 vs 1.66) and DSM-5 (2.49 vs 1.32) OUD in NESARC-III was higher among White compared to Black respondents. Never attending services declined for both races over time. Race moderated the association between service attendance (F(4,65) = 14.9, p = 0.000), social interaction (F(4,65) = 34.4, p = 0.000) and subjective religiosity/spirituality (F(2,65) = 7.03, p = 0.000) on DSM-IV OUD in wave 2 and using DSM-5 OUD in NESARC-III (F(1,113) = 2.79, p = 0.066). Race differences in religion and DSM-IV OUD risk was modified by time (i.e., survey year) (all p < 0.000). For instance, higher service attendance was associated with lower DSM-IV risk for Black respondents in wave 2 but higher risk in NESARC-III. There were no changes in regression slopes among White respondents. CONCLUSIONS: Religious involvement may be important for prevention and treatment practices that respond to racial differences in risk of OUD. Replicate studies should examine other religious factors and specific types of opioids.