The current study should be interpreted in light of its limitations. First, identification of high-risk criteria may be influenced by the AUD-enriched family-based study design, though comparisons with prior IRT estimates suggest consistent results.13,14 Notwithstanding, findings here may not generalize to samples more representative of the general population. Second, although the cross-sectional cohort has a wide age distribution, longitudinal data are delimited to younger ages. Third, many DSM-5 criteria assessed using the SSAGA are compound criteria comprising multiple items of varying severity (eg, social/interpersonal problems: objections from others vs loss of friends). Disaggregation of compound criteria into individual items, though not always practical in clinical settings, may help further elucidate symptoms signaling increased risk for development of severe AUD.53
