Child abuse and suicidality in the context of the Interpersonal Psychological Theory of Suicide: A network analysis.

Abstract

While there is evidence for an association of child abuse with suicidality in the course of life, the underlying mechanisms remain unclear. The Interpersonal Psychological Theory of Suicide (IPTS) provides a theoretical framework to investigate this relationship. The present study examines how different subtypes of child abuse are related to suicidal ideation and to attempts in the context of the IPTS. 146 psychiatric inpatients (M=37.9years, 62% female) with an acute suicidal crisis (n=71) or a recent suicide attempt (n=74) were examined at baseline (T0) and six (T1) months later. We measured emotional, physical, and sexual abuse, the constructs of the IPTS (thwarted belongingness, perceived burdensomeness, and capability for suicide) and suicidal ideation as well as suicide attempts. Using the statistics program R, a network analysis of all named constructs was conducted. Centrality measures were computed. Emotional abuse was the most central kind of abuse in the network and had a direct relationship with suicide attempts and an indirect relationship with suicidal ideation via perceived burdensomeness. Physical and sexual abuse showed no significant relations with the different constructs of the IPTS. The major limitation of this study was the modest sample size which reduced the number of variables able to be included in the network. Regarding child abuse, the results underline that emotional abuse plays a central role in this network and may be important for suicide risk assessment. Future research should address this topic in a larger sample. Emotional abuse was the most central kind of abuse in this network analysis. Sexual abuse was the only kind of abuse with a direct relation to suicidal ideation. Capability for suicide had just a marginal position in the network analysis. Early interventions addressing the effects of child abuse are recommended. Replications in larger samples and with more relevant variables are needed.