The association between alcohol use and intimate partner violence: linear effect, threshold effect, or both?

The present study examined whether the association between alcohol and male-to-female intimate partner violence (IPV) is most meaningfully described as a linear relationship, a threshold effect, or both. Men in two nationally representative samples, the National Family Violence Survey (NFVS) and the National Survey of Families and Households (NSFH), were divided into similar drinking groups based on quantity and frequency of alcohol consumption. Analyses of the association between IPV and drinking category revealed that although linear associations between drinking classification and IPV were significant in both samples, the associated effect sizes were very small. Further, only heavy drinkers and binge drinkers were major contributors to the significant chi-squares. Overall, the results revealed both linear and threshold effects, and suggest that distinctions among drinking patterns may be more important than incremental increases in quantity or frequency in conceptualizing alcohol as a risk factor for IPV.