Adolescent Nonsuicidal Self-Injury and Suicidality: A Latent Class Analysis and Associations with Clinical Characteristics in an At-Risk Cohort

Nonsuicidal self-injury (NSSI) is frequently encountered in adolescents, but its predictive value for suicidality or other clinical characteristics is challenging due to its heterogeneous nature. This study used latent class analysis to identify subgroups of NSSI and compared these on sociodemographic characteristics, adverse outcomes and protective factors. The study included 966 high-risk adolescents, Mage 14.9 y, SD 0.9 y, 51.8% female. Four classes emerged: (1) "Low NSSI-Low suicidality", (2) "Moderate NSSI-Low suicidality", (3) "Moderate NSSI-High suicidality", and (4) "High NSSI-High suicidality". Girls predominated in the high suicidality classes. Generally, Class 4 had the poorest outcomes: more internalizing and externalizing problems, less social support from friends and families and worst self-esteem. These findings emphasize the need for interventions tailored to specific phenotypes of adolescents engaging in NSSI.

Self-reported problem behavior in young children with and without a DSM-disorder in the general population

BACKGROUND

Problem behavior of young children is generally not assessed with structured child interviews. This paper examined how information about problem behavior, obtained by structured interviews with six-year-old children, relates to DSM-disorders obtained from parents and to treatment referral.

METHODS

In a population-based cohort, caregivers of 1084 young children (mean age 6.7 years) were interviewed with the DSM-based Diagnostic Interview Schedule-Young Child version (DISC-YC), and they scored the Child Behavior Checklist (CBCL). Children themselves were interviewed about problem behavior using the semi-structured Berkeley Puppet Interview (BPI). Information regarding treatment referral to mental health services was obtained by parent-reported questionnaire when children were on average eight years old.

RESULTS

DSM-disorders and CBCL problems in the clinical range were cross-sectionally associated with higher levels of child self-reported problems. Associations were strongest in the externalizing domain (e.g. DISC-YC externalizing disorders with BPI externalizing scores: F(1, 416)=19.39, P<0.001; DISC-YC internalizing disorders with BPI internalizing scores: F(1, 312)=3.75, P=0.054). Moreover, higher BPI internalizing and externalizing problem scores predicted treatment referral two years later.

CONCLUSIONS

We conclude that systematically interviewing preschool and young elementary school-aged children should be an integral part of child assessment. This approach may contribute to a better understanding of child development and may predict future problems.