Multi-dimensional prediction of suicidality and non-suicidal self-injury transition in children: from general psychopathological, behavioural, and neurobiological perspectives

IMPORTANCEAccurate prediction of suicide or non-suicidal self-injury (NSSI) among children within a uniform time frame is an essential but challenging task. Furthermore, few studies have comprehensively considered clinical, behavioural, and neurobiological factors to produce multi-dimensional prediction models. OBJECTIVETo examine predictive effects of general psychopathology, behavior inhibition system, and brain signature on childrens suicidality or NSSI transition. DESIGN, SETTING, AND PARTICIPANTSWe adopted a retrospective and longitudinal methodology by utilising the data from the Adolescent Brain Cognitive Development (ABCD) cohort. In total, 9332 individuals aged 9-10 years without any suicidality or non-suicidal self-injury (NSSI) history at baseline were included in our analyses. Then, four subgroups were generated based on whether they had developed suicide ideation (Healthy control [HC]-SI), NSSI (HC-NSSI) or suicide attempt (HC-SA) in a year, while the remaining group was considered a control group (HC-HC). MAIN OUTCOMES AND MEASURESParticipants suicidal behaviors and non-suicidal self-injury behaviors were assessed with the Kiddle Schedule for Affective Disorders and Schizophrenia. Meanwhile, general psychopathology (i.e., p-factor) was calculated based on scores of Child Behavior Checklist, behavioral inhibition system (BIS) was assessed though BIS/BAS scale, and the brain morphometrics were also collected though sMRI. Multinomial logistic regression models were used for assessing the predictive effects of general psychopathology, behavioral inhibition system, and whole-brain cortical area on childrens STB and NSSI transition. RESULTSAs a result, we found higher general psychopathology in baseline predicted higher NSSI (1.52 [1.28-1.80]), SI (OR=1.34 [95%CI 1.17-1.53]) and SA (2.05 [1.34-3.14]) risk in a year. From a behavioural perspective, higher BIS sensitivity predicted higher SI (2.05 [1.61, 2.61], and NSSI (1.68 [1.24, 2.28]) in a year. From a neurobiological perspective, abnormalities in the cortical area of the superior insula, inferior frontal area, superior temporal area, and superior precentral area were all shown to be associated with childrens NSSI, SI and SA in the future. CONCLUSIONS AND RELEVANCEThis study is the first to look at the predictive factors for the different transitions of NSSI and suicidal behaviour from the biopsychosocial framework. Our findings offered empirical evidence on the predictive effect of baseline general psychopathology, BIS sensitivity and biological marker on childrens suicidality or NSSI in a year, providing early biomarkers for all types of transition. In this case, the early identification of those factors may facilitate the development of early prevention or intervention that could potentially alleviate more relevant public health issues. Key PointsO_ST_ABSQuestionC_ST_ABSCould general psychopathology, behavior inhibition system, and brain signature predict suicidality or NSSI transition in children? FindingsIn a longitudinal observational study (9332 children), higher general psychopathology at baseline predict higher risk of suicidality and NSSI transition in a year. Meanwhile, higher BIS sensitivity also predict higher risk of suicidality and NSSI transition. To note, abnormalities in the cortical area of the superior insula, inferior frontal area, superior temporal area, and superior precentral area were all shown to be associated with childrens suicidality and NSSI transition. MeaningThe early identification of biopsychosocial factors associated with suicidality or NSSI transition in children could facilitate early prevention.