Caution is warranted in interpreting the study results because of several limitations. Most importantly, despite the fact that the concordance between the AUDADIS-IV and K-CIDI 2.1 DSM-IV diagnoses and criteria were fair to good, they were two different diagnostic instruments. Hence, the observed disparities between the two countries might reflect differences between the two interview schedules. Another reflects the insufficient numbers of cases in the Korean sample for several psychiatric disorders, due either to low prevalences in the Korean general population or to a substantially smaller sample drawn in Korea than in the United States. As a result, several comorbid associations could not be fully examined. In addition, as a cross-sectional study the current investigation could not address causality, or underlying mechanisms, of the SUD-mental disorder relationship. Longitudinal studies are needed to elucidate the progression and processes associated with comorbidity. Future research should also investigate the temporal sequencing of comorbid disorders. Alternatively, to examine the prospective associations between mental disorders and SUD one may examine several different transition points, e.g., from use to abuse, and from abuse to dependence. This
