Testing whether multi-level factors protect poly-victimised children against psychopathology in early adulthood: a longitudinal cohort study.

Aims: Exposure to multiple forms of victimisation in childhood (often referred to as poly-victimisation) has lifelong adverse effects, including an elevated risk of early-adulthood psychopathology. However, not all poly-victimised children develop mental health difficulties and identifying what protects them could inform preventive interventions. The present study investigated whether individual-, family- and/or community-level factors were associated with lower levels of general psychopathology at age 18, among children exposed to poly-victimisation. Additionally, it examined whether these factors were specific to poly-victimised children or also associated with fewer mental health difficulties in young adults regardless of whether they had been poly-victimised.

Methods: We used data from the Environmental Risk (E-Risk) Longitudinal Twin Study, a population-representative cohort of 2,232 children born in 1994-1995 across England and Wales and followed to 18 years of age (with 93% retention, n = 2,066). Poly-victimisation (i.e., exposure to two or more of physical abuse, sexual abuse, emotional abuse and neglect, physical neglect, bullying by peers, and domestic violence) and nine putative protective factors (intelligence quotient, executive functioning, temperament, maternal and sibling warmth, atmosphere at home, maternal monitoring, neighbourhood social cohesion, and presence of a supportive adult) were measured prospectively between ages 5 and 12 years from interviews with mothers and children, surveys of neighbours, child-protection referrals, and researchers' observations. Early-adulthood psychopathology was assessed in interviews with each twin at age 18 and used to construct a latent factor of general psychopathology.

Results: Approximately a third (n = 720) of participants were prospectively defined as exposed to poly-victimisation (53% male). Poly-victimised children had greater levels of general psychopathology at age 18 than non-poly-victimised children (adjusted [adj.] β = 4.80; 95% confidence interval [95% CI] 3.13, 6.47). Presence of a supportive adult was the only factor robustly associated with lower levels of general psychopathology among poly-victimised children (adj.β = -0.61; 95% CI -0.99, -0.23). However, this association was also evident in the whole sample regardless of poly-victimisation exposure (adj.β = -0.52; 95% CI -0.81, -0.24) and no significant interaction was observed between the presence of a supportive adult and poly-victimisation in relation to age-18 general psychopathology.

Conclusions: Having at least one adult to turn to for support was found to be associated with less psychopathology in early adulthood among both poly-victimised and non-poly-victimised children. This suggests that strategies to promote better availability and utilisation of supportive adults should be implemented universally. However, it may be beneficial to target these interventions at poly-victimised children, given their higher burden of psychopathology in early adulthood.