Abstract
PURPOSE: Post-traumatic stress (PTS) symptoms are highly comorbid with substance use (i.e., alcohol, tobacco, and cannabis). Few studies have investigated potential individual-, household-, and neighborhood-level socioeconomic effect modifiers of this comorbidity in longitudinal analyses. We aim to examine interactions between this multi-level environment and PTS symptoms on future substance use behaviors. METHODS: Data were drawn from the Advancing Understanding of RecOvery afteR traumA (AURORA) study, including 2943 individuals who presented to the emergency department (ED) within 72 h of a traumatic event. Frequency of tobacco, alcohol, cannabis use, and PTS symptoms were reported at 6 timepoints. Mixed effect Poisson models, clustered by state, were used to generate incidence rate ratios (IRRs) substance use, both cross-sectionally and prospectively. Moderation analysis of PTS and substance use, stratified by household income and area deprivation index (ADI), was conducted using mixed effect models and parallel process growth curves. RESULTS: Significant associations were observed between PTS with tobacco, alcohol, and cannabis use frequency cross-sectionally, and for tobacco and alcohol and PTS exposure prospectively. Lower income (P < 0.001) and higher deprivation (P < 0.001) were associated with tobacco use, while higher income (P < 0.001) and less deprivation (P = 0.01) were associated with increased alcohol use. We found modest modification by household income for alcohol and tobacco, and little evidence of modification by neighborhood ADI. CONCLUSIONS: Household income had greater evidence of effect modification for substance use, compared to neighborhood-level ADI. Our findings demonstrate that household indicators of socioeconomic status likely modify the relationship between PTS and substance use.