Child maltreatment and adult violent offending: population-based twin study addressing the ‘cycle of violence’ hypothesis

Abstract

Associations between child maltreatment and adult violence, often termed the 'cycle of violence', are well documented. However, the nature of such links after appropriate control for confounding remains uncertain. We aimed to determine whether child maltreatment causes adult violent offending or whether suggested links are due to genetic or family environment confounding. A total of 18 083 20- to 47-year-old twins from the Swedish population-based Study of Twin Adults: Genes and Environment (STAGE) participated. We linked information on self-reported child maltreatment with national register data on convictions for adult crime. We used a case-control design to elucidate associations among unrelated individuals and also conducted within-discordant twin pair analyses to estimate the influence of familial confounding on this association. The odds ratio (OR), adjusted for age, sex and education, for violent offending in maltreated children grown up versus unrelated controls was 1.98 [95% confidence interval (CI) 1.52-2.57]. However, the association decreased to 1.18 (95% CI 0.62-2.25) when maltreated children were compared to their non-maltreated twins, suggesting substantial confounding by genetic or family environmental factors (within-twin OR<1.98) and a weak, non-significant causal effect (within-twin OR>1.00). Familial confounding was also pronounced for the association between child maltreatment and any offending. Childhood maltreatment was found to be a weak causal risk factor for adult violent offending; hence, reducing maltreatment might decrease violent crime less than previously expected. Instead, considerable familial confounding of the link between child maltreatment and adult violent offending suggests that prevention strategies need to address overlapping genetic and/or family environmental liability for abusive and violent behavior.