Gender differences in characteristics of violent and sexual victimization in patients with psychosis: a cross-sectional study

In the current study, we combined data from three separate studies. Baseline data from the pilot study [17] and the multi-centre randomized controlled trial [18] of the BEATVIC project were used. The BEATVIC project aimed to gain more insight in victimization of individuals with psychosis and to develop and investigate the efficacy of a body-oriented preventive intervention. Additional data were extracted from a large ongoing Dutch observational cohort study, the Pharmacotherapy Monitoring and Outcome Survey (study 3; PHAMOUS [19];. For PHAMOUS, patients with psychotic disorders from four mental health institutions in the northern Netherlands are assessed yearly using Routine Outcome Monitoring (ROM), which is part of regular clinical practice. For a sub-study within this monitor, the VICTROM study, a victimization questionnaire was added to the screening. The protocols of the three aforementioned studies were approved by the medical ethical board of the University Medical Center Groningen and conducted in accordance with the Helsinki Declaration guidelines. For each study, participants provided written informed consent to participate and consent for processing personal data for research purposes. Before giving consent, participants were informed that a potential disadvantage of participating in the study could be that questions might be experienced as difficult or confronting.

In the current study, individuals with a diagnosis in the psychotic spectrum according to DSM-IV [20] or DSM 5 [21] who were in care at a mental health institution, were included in the analyses. Participants had to be 18 years of age or older and be able to provide informed consent. For BEATVIC [18] exclusion criteria consisted of severe psychotic symptoms, substance dependence, co-morbid neurological disorder or personality disorder, IQ below 70 or pregnancy. For PHAMOUS [19] no further exclusion criteria were utilised.

Analyses were performed in SPSS 22.0 [24]. Statistical significance was set at p < .05. To examine gender differences in demographic and clinical characteristics chi-squared test (in case of categorical variables) and independent t-tests (in case of continuous variables) were used. To examine gender differences in victimisation (threat, physical and sexual) in the past year or past 5 years, chi-squared tests were computed. Patients that had been a victim of threats, physical or sexual violence in the past 5 years were included in subsequent analyses. Three binary logistic regression models were run for threat, physical and sexual victimization incidents, with gender as the dependent variable and location, perpetrator and disclosure of the crime as independent variables. The models were run twice for each victimization type, to alternate the reference category for predictors with more than two categories (perpetrator and disclosure). Logistic regression models were corrected for significant confounders, which were identified in the descriptive analysis of demographic and clinical characteristics.

A total of 482 participants were included in this study, of which 105 (21.8%) were from the BEATVIC RCT study, 24 (5%) were part of the BEATVIC pilot study and 353 (73.2%) were part of the PHAMOUS-VICTROM study. Of the 482 participants, 170 (35.3%) were women. Demographic information per gender on the entire study sample can be found in Table 1. Women were significantly older than men. In comparison to women, a significantly greater proportion of men were diagnosed with schizophrenia as opposed to another psychotic disorder. There was no significant difference in gender for housing but compared to men, women were more likely to have no work. Further analyses were corrected for age, diagnoses and employment.

In the past year, there were no differences between men and women in violent victimization (e.g., physical assault, threat with violence or with a weapon) and sexual victimization (e.g., forced sexual penetration, sexual touch without consent or sexual harassment) (see Table 2). However, over the past 5 years, men were more likely to have experienced threat with violence and were less likely to have experienced sexual victimization as opposed to women. For victimization of physical assault no significant differences between men and women in the past 5 years were found.

The current study demonstrated that men diagnosed with a psychotic disorder were more likely to have experienced threats of physical violence than women, whereas women with a psychotic disorder were more likely to have experienced sexual harassment and assault in the past 5 years. Assessment of victimization in the past year shows a similar, yet non-significant trend. Overall, these findings are in line with patterns identified in the general population [7, 9]. One could argue that whereas studies from the general population conclude that there is a difference in prevalence for violent and sexual victimization between men and women, this difference might be smaller in people with a psychiatric disorder. It is possible that as victimization rates increase in individuals with a psychiatric disorder [1], the difference in prevalence of victimization types between men and women also fades and becomes less prominent.

With regard to physical violence, women more often reported being victimized at home while men more often reported being victimized in the streets or elsewhere. This finding is also in line with studies examining victimization in the general population [7, 9]. Previous research found that men with a mental disorder more often used street drugs or alcohol when using violence themselves and women are more often violent at home (Robbins et al., 2003). Similarly, the use of street drugs might play a role in victimization of men in the streets. For threats with violence, a similar yet non-significant pattern of findings was found. As we suspected these findings on location (higher risk at home) to be (partly) correlated with the findings on the characteristics of the perpetrator (higher risks of being assaulted by partner than by a stranger) we performed a post-hoc test. This extra analysis demonstrated that location holds significantly different risks for men and women when added as a single predictor, but that this effect disappears (and is thus explained by) when type of perpetrator is added to the model. Other studies that did find a significant effect of location did not include multiple characteristics in the same model [7, 24] which is likely to explain the inconsistency between our study and previous research. Our findings indicate that location and perpetrator are closely linked violent victimization characteristics, and should be taken into account when tailoring personalized interventions.

For sexual assault and harassment, there were no significant differences in location between men in women. However, these results need to be interpreted with caution as only 3.6% of the men reported sexual victimization. Although it was hypothesized that women would be more likely victimized at home, there was a higher proportion of women (though not significant) who reported being sexually assaulted and/or harassed outside of their home. However, this pattern is in fact similar to the findings of a recent Dutch population survey, reporting that most victims (regardless of gender) were sexually assaulted or harassed outside the house, for example, in the streets or in a pub by someone from outside the domestic sphere [16]. It appears that on the basis of previous studies, that did not differentiate between victimization types; one would conclude that women are more likely to be victimized at home, where in fact this conclusion does not hold up for all types of victimization (i.e. violent victimization vs sexual victimization). More in depth research is needed to learn more about how characteristics of sexual victimization differ from characteristics of violent victimization or overall victimization.

Regarding the type of perpetrator, men are more prone to be threatened by strangers whereas for women the risk is higher to be threatened by a person more familiar; these findings in our sample of people with a psychotic disorder are similar to findings in the general population [4, 5] and in patients with dual diagnosis [6]. Descriptive statistics showed similar results for actual physical violence although in formal testing this did not reach significance.

For sexual harassment and assault, no significant gender differences were found in type of perpetrator. Again, this seems to imply that characteristics of sexual victimization differ from violent victimization.

As expected there were no gender differences in disclosing the crime for actual violence and sexual assault or harassment. For threats with victimization, compared to men, women were less likely to disclose threat of physical violence to their partner, a friend or family than to disclose to a professional or the police, or to not disclose at all. Although this is not in line with our hypothesis that men and women do not differ in disclosure, our findings are likely explained by the fact that the partner, a friend or a family member in many cases was the perpetrator. This is also consistent with a previous study (CBS, 2015 [15]; that found no gender differences in reporting to the police in the general population, as in our sample men and women were just as likely to report to the police or a professional (45.2% vs 52.5%).

The current study has some limitations. First, victimization was examined with a self-report questionnaire, which is subject to memory bias. This could lead to an underreporting or overreporting of past incidences of victimization. However, a study on self-reports of violent victimisation in severe mental illness, showed that trauma history can be reliably assessed for research purposes [25]. Second, few victimization incidents were reported for the past year, and thus it was only possible to conduct formal analyses on victimization incidents of the past 5 years. Additionally, due to the low number of incidents some analyses may have been underpowered, yielding potentially biased estimates. Third, participants in this study were receiving treatment and were relatively stable in terms of symptoms at the time of assessment. It is likely that incidence rates of victimisation are higher in individuals with a psychotic disorder who are currently not in treatment. An important strength of the current study is the amount of detailed information, which enabled us to investigate gender differences in characteristics of different types of victimization in patients with a psychotic disorder for the first time.