Alcohol Problems as a Mediator of the Association Between Abuse and Violence
Several analyses of data collected by Widom and her colleagues, comparing substantiated abuse victims to a matched control group, suggest that alcohol abuse may mediate the association between abuse/neglect and later violent arrests— fully for females, but only partially for males (Widom, Schuck, & White, 2006). In another analysis, alcohol abuse also partially mediated the relationship between early abuse and neglect and later intimate partner violence (White & Widom, 2003). Widom’s most recent statements about the association between child abuse and alcohol abuse affirm that the association has best been demonstrated in women, and her statements about the mediating effect of alcohol remain tentative (Widom & Hiller-Sturmhofel, 2014).
Drug Problems as a Mediator of the Association Between Abuse and Violence
In Widom’s data, the evidence that drug use mediates the association between abuse and violence is even less consistent. This is due, in part, to the fact that only the abused girls in her sample were more likely to abuse drugs later in life (Widom & White, 1997).
In other data sets, the effects of child abuse have been dramatically attenuated when analysts have controlled for substance abuse. Ehrensaft et al. found that abuse was no longer associated with partner violence in a multivariate model, accounting for “emerging adult substance use disorder” (Ehrensaft, Cohen, Brown, Smailes, Ehen, & Johnson, 2003). Appleyard et al. found that the association between maternal experience of physical abuse and offspring victimization was fully mediated by substance problems, but the association between sexual abuse and offspring victimization was only partially explained by maternal substance abuse (Appleyard, Berlin, Rosanbalm, & Dodge, 2011). We should remind ourselves, however, that this type of model may mask indirect effects where abuse victimization causes substance abuse, which in turn causes violent behavior.