Different measures of alcohol use as predictors of DSM-5 alcohol use disorder among adolescents - A cohort study from Sweden

Background and aims: This study addresses a significant gap in existing research by investigating the longitudinal relationship between various measures of alcohol use and the development of alcohol use disorders (AUDs) in a cohort of Swedish adolescents. Methods: A prospective longitudinal survey was conducted on 3999 adolescents in Sweden who were in 9th grade in 2017 and were followed up in 2019. Baseline assessments included lifetime alcohol use, recent use (past 30 days), risky drinking (AUDIT -C), and heavy episodic drinking (HED). Follow-up assessments comprised eleven items measuring DSM-5 AUD criteria. The study explores prospective associations between these diverse alcohol use measures and the occurrence of AUD, while also calculating population attributable fractions (PAF). Findings: The proportion of alcohol consumers who met the criteria for AUD at follow-up was 31.8%. All baseline measures of alcohol use exhibited associations with subsequent AUD. Notably, the HED group demonstrated the highest prevalence of AUD at 51.4% (p<.001). However, when calculating PAFs, any lifetime alcohol use emerged as the most substantial contributor, accounting for 10.8% of all subsequent AUD cases. Conclusions: This study underscores that alcohol use during mid -adolescence heightens the risk of developing AUD in late adolescence. Among the various measures, heavy episodic drinking presents the highest risk for later AUD. From a public health perspective, preventing any alcohol use emerges as the most effective strategy to mitigate the population -level burden of disease of AUD.