Personality and mental health as mediators linking childhood maltreatment to intimate partner violence victimization: a Mendelian randomization?direction of causation twin study

Pezzoli, Patrizia, Barkhuizen, Wikus, Oginni, Olakunle, Pingault, Jean-Baptiste, McCrory, Eamon and Viding, Essi 2026. Personality and mental health as mediators linking childhood maltreatment to intimate partner violence victimization: a Mendelian randomization?direction of causation twin study. The Lancet Regional Health Europe 66 , 101653. 10.1016/j.lanepe.2026.101653

Preview

PDF - Published Version Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

Background Childhood maltreatment elevates risk of later intimate partner violence (IPV) victimization, both major psychiatric risk factors. Personality and mental health have been proposed as mediators, but these pathways may be confounded by genetic and environmental influences shared across maltreatment, IPV victimization, and intermediary phenotypes. Here, we examined how maltreatment confers additional IPV risk beyond shared etiological liabilities. Methods We analyzed data from the UK's Twins Early Development Study (N = 11,342, aged 22) and genome wide association study summary statistics. We used structural equation modeling (SEM), twin modeling, and Linkage Disequilibrium Score Regression to estimate phenotypic and etiological associations between maltreatment, IPV victimization, and eighteen intermediary phenotypes. In a genotyped subsample (n = 8406), we used Mendelian Randomization-Direction of Causation (MR-DoC) to examine causal mediation beyond genetic and environmental confounding. Lived-experience input informed design and interpretation. Findings Childhood maltreatment and IPV victimization showed a moderate phenotypic association (β = 0.23), small-to-moderate heritability (h2twin = 17–31%; h2SNP = 1–4%), and strong genetic correlations with each other (rg = 0.52–0.82) and with most intermediary phenotypes (rg = 0.16–0.84). Two latent factors capturing shared variance across all intermediary phenotypes partially mediated the effect of maltreatment on IPV victimization in SEM (total indirect effect ∑ab = 0.11, 48% mediated, direct effect c’ = 0.12, padj<0.001). In MR-DoC, which further adjusts for confounding, we also observed mediation, and no significant direct effect remained (∑ab = 0.09, 65% mediated, c’ = 0.05, padj = 0.161). Three mediators were robust across models examining individual mediators (separately or simultaneously): low well-being, conduct problems, and aggression (ab = 0.01–0.05). Interpretation Maltreatment elevates risk of IPV victimization through its impacts on personality and mental health development, beyond risk conferred by shared genetic and environmental liabilities. Preventing IPV victimization among young people exposed to maltreatment requires greater understanding of how these psychological impacts operate within social transactional contexts, and developing targeted strategies that address them.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Schools > Medicine
Publisher:	Elsevier
ISSN:	2666-7762
Date of First Compliant Deposit:	7 April 2026
Date of Acceptance:	5 March 2026
Last Modified:	02 Jun 2026 10:47
URI:	https://orca.cardiff.ac.uk/id/eprint/186173

Actions (repository staff only)

Edit Item