Sex-segregated analyses were performed for both clinical and population samples separately. First, multinomial regression analysis was used to investigate the association between symptoms of psychological ill health (response variable) and exposure to multiple forms of violence (model 1) as well as violence by multiple perpetrators (model 2). “No symptoms of psychological ill health” was used as the reference category for the response variable. In the male clinical sample, experiences of multiple forms of violence including sexual violence as well as violence by family and partner perpetrators were only reported by a few respondents. For this reason these categories had to be collapsed and the explanatory variables were restricted to include information about whether respondents reported one or more forms of violence (model 1) and one or more kinds of perpetrators (model 2). All models were adjusted for age-group, educational level, civil state and current occupation.