Mediating and moderating variables in the prediction of self-harm in young people: A systematic review of prospective longitudinal studies.

Abstract

Listen

Translate article

Self-harm is widespread amongst young people. A growing body of research has explored factors that predict self-harm in young people, however, a systematic review of mediators and moderators of those factors has not yet been offered. This review aims to fill this gap by synthesising research about mediators and moderators of factors that prospectively predict self-harm in young people. A systematic review of research trials published up until 2018 was undertaken. Electronic databases Scopus (Elsevier), CINAHL, PsychINFO (EBSCO) and Medline were searched. Included studies utilised prospective longitudinal designs with participants aged 25 years or younger and self-harm outcome measures with published or reported psychometric properties. The aim of the review was to identify mediators and moderators of factors that predict self-harm in young people. Of the 25 studies that met inclusion criteria, 22 reported at least one positive finding of a mediator or moderator. Specifically, 15 significant mediators and 20 significant moderators were identified in relation to a broad range of predictors of self-harm. Predictors were classified as adverse childhood experiences and parenting factors, psychological and psychiatric factors, social factors and intrapersonal factors. A number of potentially modifiable mediators and moderators were identified including interpersonal difficulties, impulsivity, self-esteem and self-compassion. Gender was the most commonly reported moderator. Included studies were assessed as limited by the heterogeneity of the mediators and moderators assessed, and by methodological factors including study durations, population characteristics, and the definition and assessment of self-harm. In addition, replication research was limited. Therefore it was difficult to integrate results and draw firm conclusions. This review allowed us to explore diverse relationships between factors predictive of self-harm in young people and to identify a number of potentially modifiable mediators and moderators. Our findings have important implications for future research and treatment efforts as the identification of mediators and moderators is demonstrated to assist in identifying high risk individuals as well informing potential targets for treatment.