Is Emerging Adult Status Associated with Substance Use: A Meta-Analysis

Background and Purpose. Much debate exists surrounding Arnett’s theory of emerging adulthood in terms of its breadth and application. Researchers have attempted to capture dimensions of emerging adulthood (e.g., experimentation, negativity/instability, other-focus, self-focus, and feeling in-between) through self-report assessment, using variations of the Inventory of the Dimensions of Emerging Adulthood or IDEA. Results from studies investigating this relationship have been mixed. We conducted a meta-analysis on the association between substance use and the IDEA outlined by Arnett (2000). Methods. Data were extracted to calculate correlational associations with substance use as well as typical moderators found in the literature. We implemented mixed effects meta-regression to assess the association between substance use and IDEA subscales as well as the full IDEA scale. We fit random-effects models, which produce more conservative estimates by estimating both the error of the effect size as well as the between-study variance. We evaluated the possibility of publication bias by examining funnel plot asymmetry using the nonparametric trim and fill approach based on Duval and Tweedie as well as the precision-effect estimate with standard error (PEESE) method. Twelve studies were meta-analyzed. Results. Across all 12 studies we estimated effect sizes for 13,223 participants. The mean age was 21.8 years (SD = 2.76) with an average of 45.4% (SD = 28.7) female participants. Samples primarily consisted of White (53.4%, SD = 29.6) and employed (50.8%, SD = 21.2) participants. The average percentage of participants currently attending college was 30.0% (SD = 32.7%). Most studies were conducted in the United States (67%). We found no association between substance use and self or other-focus. We did find small associations (range: ρ = -.03 - .15; d = .06 - .30) between the IDEA scores such as identity exploration, experimentation/possibilities, negativity/instability, and feeling in-between and substance use. We found higher severity (dependence diagnosis) of participants yielded larger associations across all dimensions (ρ = .16), and proportion of college students to be a subscale-specific moderator (experimentation, negativity/instability, other-focus, self-focus, and feeling in-between). Alcohol use outcomes also provided larger subscale-specific associations (experimentation, negativity/instability, other-focus, self-focus). Conclusions and Implications. The dimensions of emerging adulthood may be less effective in predicting substance use among non-college samples and those studies focusing on drug use. Further research should prioritize exploring variation in the transition to emerging adulthood among non-college samples and the longitudinal associations between IDEA and substance use. Important contributions include the modest association between IDEA and substance use as well as specific participant characteristics that amplify or mitigate the association between IDEA and substance use.