This study must be interpreted with several limitations in mind. Foremost is the fact that the sample was selected for its high risk for AUD and therefore cannot be generalized to the larger population. However, the examination of the context of remission in a high-risk sample extends similar findings in population-based and clinical samples. The lack of recency ages for individual symptoms at baseline necessitated basing current AUD status at baseline on a single question about age at most recent symptom. Finally, the sample was primarily white with little power to detect ethnic differences.
