Background Recent international meta-analyses show that children and adults with disabilities are more likely to experience interpersonal violence than those without disabilities. People with disabilities are a heterogeneous group, and further work is required to determine the variations in the experience of violence by impairment type. It is also important to assess the extent to which any between-group differences may be explained by socio-economic situation.

Methods We analysed the 2015 Australian Bureau of Statistics Survey on Personal Safety of more than 21,000 adults. We used population-weighted, age-adjusted, logistic regression to estimate the odds of violence in the last 12 months by impairment type and gender. We stratified by five forms of violence (physical, sexual and intimate partner, emotional abuse and stalking/harassment) and for specific impairment types (physical, intellectual, psychological and sensory). We also investigated potential effect modification by poverty status.

Results Results for women showed that those with sensory and physical impairments fared the best, with a two-fold increase in odds across all five forms of violence. Women with intellectual and psychological impairments fared worse for all forms of violence. There was a three-fold increase in the odds for emotional abuse and stalking/harassment. There was an even greater increase in the odds for sexual, physical and partner violence. Women with intellectual impairments had between a five and seven-fold increase in the odds of these forms of violence, while women with psychological impairments had been a four and six-fold increase. All results were statistically significant at p=0.05.

Results for men showed much less variation by impairment type. There was a statistically significant two-fold increase in odds of physical violence for those with sensory, intellectual and psychological impairments. There were no significant differences between men with impairments compared to men without disabilities for sexual violence, partner violence and emotional abuse. There was evidence that stalking/harassment was worse for men with psychological impairments.

Some of these associations were moderated by poverty status, with the increased odds of exposure to violence among people with disabilities being greater for people living in poverty.

Conclusion These results show there are important variations in the experience of violence for people with disability depending on their impairment type, gender and socio-economic status. There is a clear need to develop interventions that are targeted to the particular circumstances and needs of these high-risk groups.