Diagnostic validity and reliability of a Korean version of the Parent and Adolescent General Behavior Inventories.

OBJECTIVES: The purpose of this study was to evaluate the validity and reliability of a Korean version of the Parent General Behavior Inventory-10-item Mania Scale (P-GBI-10M) and the Adolescent General Behavior Inventory (A-GBI) for bipolar and depressive disorder in youths.

METHODS: Ninety-two subjects with mood disorder and their parents were recruited from September 2011 to June 2013 through the Department of Psychiatry at the Asan Medical Center, Seoul, Korea. In addition, 125 community participants were recruited through two middle schools and one high school in Seoul. The parents of subjects completed the Parent-version Mood Disorder Questionnaire (P-MDQ), P-GBI-10M and Attention-deficit/hyperactivity disorder Rating Scale (ARS). Adolescents complete the 76-item A-GBI, Beck Depression Inventory (BDI), and Adolescent version of the Mood Disorder Questionnaire (A-MDQ).

RESULTS: Different profiles were evident between the clinic-referred group and the community control, including different P-GBI-10M (t=3.07, p=0.003), A-GBI Depressive (t=4.99, p<0.001), Hypomanic/Biphasic subscales (t=3.17, p=0.002), and BDI (t=4.76, p<0.001) scores. The A-GBI Depressive subscale score (t=3.02, p=0.003), BDI score (t=2.12, p=0.037) and A-GBI Hypomanic/Biphasic subscale score (t=2.71, p=0.008) were significantly different between patients with bipolar disorder and those with depressive disorder. Of the 73 items of the Depressive and Hypomanic/Biphasic subscales of the A-GBI, eight discriminated between bipolar and depressive disorder. Furthermore, A-GBI Depressive subscale scores were significantly correlated with BDI (r=0.81, p<0.001), A-GBI Hypomanic/Biphasic subscale (r=0.88, p<0.001), A-MDQ (r=0.58, p<0.001), P-MDQ (r=0.22, p=0.005), and ARS (r=0.26, p<0.001) scores. Cronbach's α of the A-GBI was 0.98.

CONCLUSION: The Korean version of the Parent and Adolescent General Behavior Inventories showed excellent internal consistency, fair-to-good construct, and discriminant validity.