Background and Aims The relationship between violence and problem gambling in general population samples is
under-researched and requires further attention to inform treatment and prevention efforts.We investigated the relationship
between gambling problems and violence among men and sought to determine if the link can be accounted for by
mental disorders, alcohol and drug dependence and impulsivity. Design A cross-sectional survey. Setting A UK
representative general population survey conducted in 2009. Participants A total of 3025 UK men aged
18–64 years. Measurements Binary logistic regression was used to examine relationships. Outcomemeasures included
gambling behaviour and self-reports of violence. Covariates included alcohol and drug dependence, mental illness, impulsivity
and socio-demography. Findings Problem gambling and probable pathological gambling were associated with increased
odds of the perpetration of violence [adjusted odd ratios (AOR) = 3.09, confidence interval (CI) = 1.90–5.00 and
4.09, CI = 2.76–6.30, respectively] and a range of other behaviours, such as using aweapon (AORs = 4.93, CI = 2.52–9.63
and 6.33, CI = 3.52–11.38) and the perpetration of intimate partner violence (AOR = 9.80, CI =2.45–39.04). The
results were attenuated when adjusted for comorbid mental illness and impulsivity, but remained statistically significant.
Alcohol and drug dependence had the most impact; relationships were most attenuated when they added into
the models, with the latter having the largest effect. Conclusions Among men in the United Kingdom, self-reports
of problem/pathological gambling remain predictive of a range of measures of violent behaviour after adjusting for alcohol
and drug dependence, comorbid mental disorder and impulsivity; of the covariates, alcohol and drug dependence
have the greatest effect in attenuating the gambling-violence association.