Objectives: The aim of this study was to examine the association between current suicidal or violent behavior and deficits of specific neurocognitive variables in child and adolescent inpatient samples diagnosed with mood disorder.

Methods: A retrospective review of the charts of mood disorder patients hospitalized at Samsung Medical Center between April 2004 and April 2015 was conducted. Child and adolescent patients aged between 10 and 18 years old and those who finished neurocognitive function testing during their hospitalization were included. Among them patients whose full scale IQ was between 85 and 115 were selected (N=111). Participants were first divided into two age-groups?group Y (10≤age≤15, N=54) and group O (16≤age≤18, N=57)?because neurocognitive function test tools were different according to age [Wechsler Intelligence Scale for Children (WISC) for 10 to 15-year-old patients, Wechsler Adult Intelligence Scale (WAIS) for 16 to 18-year-old patients]. They were then divided according to their suicidal or violent behavior?non suicidal/violent group (NG), suicidal group (SG), violent group (VG), and both suicidal/violent group (BG). The Child Behavior Checklist (CBCL) was checked for measurement of participants’ behavior and the Gordon Diagnostic System was checked for measurement of their attention efficiency. Kruskal-Wallis Test and Tukey test was used to determine the differences in neurocognitive function between groups.

Results: O-SG patients showed lower scores on the comprehension subscale of WAIS-III than O-NG patients (χ²=8.454, p=.015). O-VG patients showed lower scores on the block design subscales of WAIS than O-SG patients (χ²=7.496, p=.024). Y-VG patients showed higher scores in aggressive behavior, externalizing problems, and total problems scores of CBCL.

Conclusion: This study showed relationship between specific neurocognitive deficits and suicidal or violent behavior. These relationships were significant in relatively older adolescents.