Does Antisocial Behavior in Early Childhood Influence Rates of IPV and Associated Risks in Urban Adolescents and Young Adults?

Prior research has identified a link between antisocial behavior and risk for intimate partner violence (IPV). However, many studies have not examined the influence of early antisocial behavior in high-risk diverse samples, nor have they examined risk for both IPV behaviors and associated attitudinal risks such as views on normative violence and perceived sanctions for committing violence. The study sample is drawn from a randomized control trial testing whether a family-based parenting intervention (Building Blocks; Brotman et al., 2005; 2008) reduced the risk for conduct disorder in an at-risk sample. Ninety-nine preschoolers were randomly assigned to the intervention or control condition. At 10-year follow-up (Wave 1), 45 preschoolers and 43 of their siblings who were living at home during the intervention completed a telephone-based assessment of their romantic relationship development, including measures of physical and psychological dating violence. In Wave 2, we recruited 103 participants, and in Wave 3, 109 participants. Attrition analyses revealed no differences among females recruited from the intervention or control groups; however, differences among males limited further analyses. The current study focuses on a subsample of 70 females at Wave 3, who were a mean age of 19.2 (SD = 4.8, range = 12 – 30).

We examined the influence of early antisocial behavior (measured by social competence, aggression, and defiance) at mean age 5 on IPV perpetration (relational, verbal/emotional, and physical) and associated attitudinal variables at mean age 19. The attitudinal variables construct consisted of three variables including perceived normative violence from boys to girls, and from girls to boys, and perceptions of sanctions for perpetrating IPV. Structural equation modeling results show that antisocial behavior predicts IPV perpetration (estimate = .63 (.12), p <.01), but does not predict the associated attitudinal variables. However, IPV perpetration is highly correlated with the attitudinal variables (r = .59, p < .01). The model was a good fit to the data, χ²(39) = 48.2, p = .15, CFI = .95, TLI = .93, RMSEA = .06 (.00 - .11, p = .38). Implications of different findings for IPV and associated attitudinal variables will be discussed.