Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Factors associated with violent victimisation among homeless adults in Sydney, Australia

Factors associated with violent victimisation among homeless adults in Sydney, Australia It has been well‐established internationally that homeless people experience higher levels of violent victimisation, such as physical or sexual assaults, than housed populations. Just over half of the homeless adults interviewed in a large UK study reported having experienced violence in the past year, compared to just 4% of the broader population. A study in the US found that in the previous two months, 18% of a homeless sample had been threatened with a weapon, 16% had been beaten and 6% had been sexually assaulted. A number of risk factors for violent victimisation have been examined. Homeless women are at increased risk of sexual assault compared to homeless men, but gender does not appear to influence risk of other physical victimisation. Poor mental health is strongly indicative of increased risk, with schizophrenia, more severe psychotic symptoms, a history of psychiatric hospitalisation, and general psychological distress all having been associated with violent victimisation of homeless people. The relationship between substance use and the victimisation of homeless people is complex. Simons et al. found substance use to be moderately correlated with victimisation, and Kushel et al. similarly found both alcohol and illicit drug use to be associated with physical assault. Analysing victimisation by gender, Wenzel et al. found that women's victimisation was associated with illicit drug, but not alcohol dependence, with the opposite being found for men. Finally, Lam & Rosenheck noted that days of alcohol intoxication, but not days of drug use, was associated with violence. A shortcoming of these findings is the lack of discrimination between types of illicit drugs (e.g. heroin vs. cannabis), which may have very different effects on an individual's risk of victimisation. Victimisation of homeless people has consequences beyond physical and emotional injury. Although there has been little longitudinal research on homeless populations, it appears that violent victimisation may prolong homelessness, even more so than factors such as an individual's level of social support. Hence, identification of risk factors and preventive strategies for violent victimisation may be an important aspect of reducing chronic homelessness. With the release of the Federal Government's White Paper, The Road Home , homelessness has become a political ‘hot topic’ in Australia. There are approximately 105,000 homeless people in Australia, but very little research has been conducted with this population. A notable exception is the work of Buhrich, Teesson and Hodder, who produced a series of papers documenting the high prevalence of psychiatric and substance use disorders among Sydney's adult homeless. There has been some examination of how family violence precipitates homelessness among women and children, but to our knowledge no studies have examined victimisation of currently homeless men and women in Australia. This study aims to address this gap by describing the prevalence of, and factors associated with, violent victimisation among a sample of homeless adults in inner‐Sydney. Method A cross‐sectional design was used. Data were collected between January 2007 and April 2008 from clients of a shelter for homeless, substance‐using adults in inner‐Sydney. Potential participants were approached by an interviewer, who briefly described the study and invited the client to participate. No incentives for participation were offered. Clients indicating willingness to participate in the study completed informed consent procedures, with the interviewer emphasising that all information was confidential and that the participant was free to terminate the interview at any time. Participants were advised at the beginning of the interview that they could take a break at any time. A standardised interview was administered to each participant. Homelessness was assessed by asking participants how long they had been homeless, and where they slept (rough/outdoors, homeless shelter, boarding house, friends/family, hospital/drug treatment service, prison, squat, other) for most of the previous month. From this latter question, a variable was constructed categorising participants as primary (rough/outdoors, squat), secondary (homeless shelter, friends/family) or tertiary (boarding house, hospital/drug treatment service, prison, other) homeless. Alcohol and other drug use were assessed by asking the participant how many days in the past six months they had used each of heroin, psychostimulants (defined as amphetamine, methamphetamine or cocaine), benzodiazepines, alcohol and cannabis. To simplify analysis of the complex drug use histories reported by participants, a dichotomous variable was constructed for each drug type, with participants classed as either non/occasional users (less than once per week for the previous six months) or regular users (once a week or more for the previous six months) of each drug. Depressive symptoms were assessed using the Depression, Anxiety and Stress Scale (DASS). The DASS provides a score between 0 and 42, with higher scores indicating more severe symptoms. The raw scores were used to construct a variable categorising participants as low‐depression or high‐depression, with the median score used as the cut‐point. Post‐traumatic stress disorder was screened for using the Trauma Screening Questionnaire (TSQ), with a score of greater than five indicative of PTSD. Participants were also asked “have you ever been told by a doctor that you have schizophrenia or a psychotic disorder?” Violent victimisation was assessed by asking the participant “In the past 12 months, has anyone been physically violent toward you? By physically violent, I mean being hit, slapped, pushed, kicked or attacked with a weapon by someone to cause harm.” Data were analysed using SPSS 17. To determine factors associated with violent victimisation, univariate logistic regressions were conducted with past year victimisation (yes/no) as the dependent variable. Factors included in the univariate analyses were socio‐demographic characteristics (age, gender, indigenous status), homelessness variables (length of homelessness, homelessness category), mental health indicators (depressive symptoms, post‐traumatic stress disorder, schizophrenia or other psychotic disorder), and use of heroin, psychostimulants, benzodiazepines, alcohol and cannabis in the previous six months. Factors significant at the p <0.05 level in univariate analyses were entered into a multivariate logistic regression model. Cases with missing data were removed from the analysis. This model was examined and the factor with the largest p value was removed. This process was repeated until a model containing only significant factors was obtained. Results Data were obtained from 106 individuals, representing approximately 25% of clients accessing the shelter during the data collection period. One participant terminated the interview after completing only the demographic questions and was excluded from analysis, giving a sample size of 105. The sample was predominantly male (n=85, 81%) and the mean age was 41.5±11.1 (range 20‐76). Eighteen (17%) participants identified as Aboriginal or Torres Strait Islander. Length of homelessness and category of homelessness in the past month are shown in Table 1 . The majority of participants were categorised as secondary homeless, that is, they had slept at homeless shelters or at the homes of friends or family for most of the previous month. 1 Characteristics of homelessness. Length of homelessness n (%) <1month 22 (21) 1‐6 months 22 (21) 6‐24 months 33 (31.4) >24 months 28 (26.7) Category of homelessness (last month) Primary 30 (28.6) Secondary 59 (56.2) Tertiary 16 (15.2) Participants had complex alcohol and other drug use histories. Polydrug use was the norm, with participants using a median of three drug classes (excluding nicotine) in the previous six months. DASS depression scores suggested high levels of depressive symptoms. Forty‐two per cent of participants fell within the severe/extremely severe depression range, and a further third reported mild/moderate symptoms. The median depression score, for use in the regression analyses, was 18 (inter‐quartile range 10‐32). More than one‐third (n=39, 37.1%) of participants reported a lifetime diagnosis of schizophrenia or other psychotic disorder. Of the 90 participants who completed the TSQ, more than half (n=51, 56.7%) screened positive for PTSD. One hundred and two participants responded to the victimisation question. Of these, 49 (48%) reported that they had been a victim of violence in the previous 12 months. Univariate results for past year victimisation are shown in Table 2 . Being female, having a diagnosis of schizophrenia or other psychotic disorder, current PTSD, higher level of depressive symptoms and regular psychostimulant use were all associated with violent victimisation in the previous 12 months. 2 Past year violent victimisation. Univariate analyses Multivariate model Factor No violent victimisation n (%)* Violent victimisation n (%)* p OR (95% CI) p AOR (95% CI) Age Mean (SD) 43.2 (11.2) 40.6 (10.8) 0.25 0.98 (0.94‐1.02) Gender Male 48 (90.6%) 36 (73.5%) Female 5 (9.4%) 13 (26.5%) 0.03 3.47 (1.13‐10.61) Indigenous status Non‐indigenous 43 (81.1%) 41 (83.7%) Indigenous 10 (18.9%) 8 (16.3%) 0.74 0.84 (0.30‐2.34) Length of homelessness <1month 11 (20.8%) 9 (18.4%) 0.96 referent 1‐6 months 12 (22.6%) 10 (20.4%) 0.98 1.02 (0.30‐3.44) 6‐24 months 17 (32.1%) 16 (32.7%) 0.81 1.15 (0.38‐3.51) >24 months 13 (24.5%) 14 (28.6%) 0.64 1.32 (0.41‐4.20) Homeless category Primary 18 (34.0%) 11 (22.4%) 0.41 referent Secondary 28 (52.8%) 29 (59.2%) 0.26 1.70 (0.68‐4.22) Tertiary 7 (13.2%) 9 (18.4%) 0.24 2.10 (0.61‐7.27) Psychotic disorder or schizophrenia No 41 (77.4%) 23 (46.9%) Yes 12 (22.6%) 26 (53.1%) 0.002 3.86 (1.65‐9.07) 0.02 3.13 (1.24‐7.90) PTSD No 26 (57.8%) 13 (28.9%) Yes 19 (42.2%) 32 (71.1%) 0.007 3.37 (1.40‐8.08) Depressive symptoms Low 33 (63.5%) 19 (40.4%) High 19 (36.5%) 28 (59.6%) 0.02 2.56 (1.14‐5.8) 0.04 2.65 (1.07‐6.59) Heroin use None/occasional 48 (90.6%) 39 (79.6%) Regular 5 (9.4%) 10 (20.4%) 0.13 2.46 (0.78‐7.80) Psychostimulant use None/occasional 48 (90.6%) 31 (63.3%) Regular 5 (9.4%) 17 (34.7%) 0.003 5.27 (1.76‐15.73) 0.008 5.07 (1.53‐16.84) Benzodiazepine use None/occasional 43 (81.1%) 38 (77.6%) Regular 10 (18.9%) 11 (22.4%) 0.66 1.25 (0.48‐3.26) Alcohol use None/occasional 19 (35.8%) 17 (34.7%) Regular 34 (64.2%) 32 (65.3%) 0.90 1.05 (0.47‐2.37) Cannabis use None/occasional 30 (56.6%) 29 (59.2%) Regular 23 (43.4%) 19 (38.8%) 0.70 0.86 (0.39‐1.89) Notes: *n varies between factors due to missing data The five variables significant in the univariate analyses were assessed for collinearity and found to be adequately independent. Three iterations were undertaken to reach the final multivariate model (Hosmer and Lemeshow χ =2.3, df=5, p =0.81), which accurately predicted victimisation in 72% of cases. In this model, regular psychostimulant use was associated with a five‐fold increase in risk of victimisation. Diagnosis of schizophrenia/psychotic disorder and higher level of depressive symptoms were also significantly associated with increased risk of victimisation ( Table 2 ). Discussion This study makes an important contribution to the literature as the first to consider violent victimisation among homeless people in Australia. Just under half of this sample reported past year victimisation. Directly comparable data from a general community sample are not available, but it is worth noting that the recorded assault rate in NSW in 2007 was around 1 per 1,000 persons, a great deal lower than that experienced by this homeless sample. In addition to the generally high level of violence, this study has identified distinct factors that are associated with an increased risk of violent victimisation of homeless persons. Participants with a history of schizophrenia or other psychotic disorder were 3.1 times more likely to be victimised than those without such a history. This aligns with previous research with homeless populations, as well as research focusing specifically on victimisation of individuals with psychotic disorders. For example, in a large Australian sample of individuals with a psychotic disorder, past year prevalence of violent victimisation was 18%. The relationship between psychosis and victimisation is theorised to be a result of the symptoms of psychosis; that is, impaired judgement affects one's ability to identify risk, while symptoms such as talking to oneself and disordered behaviour draw attention to people with psychotic disorders, increasing the likelihood of violence. Given that homeless people spend large amounts of time in public spaces, these effects may be magnified for homeless, psychotic individuals, resulting in even greater risk of violent victimisation. These findings underscore the importance of providing safe, secure accommodation for people with severe mental illnesses. Although substance use in general has previously been associated with victimisation, in this analysis regular psychostimulant use, but not heroin, alcohol, cannabis or benzodiazepine use, was associated with a five‐fold increase in past year violent victimisation. Research on psychostimulant use and violence tends to focus on whether users are perpetrators of violence, with little recognition that psychostimulant use may contribute to victimisation. As with schizophrenia, it is possible that the behavioural effects of psychostimulants – for example, feelings of strength or invincibility, talkativeness and reduced perception of risk – may draw attention to psychostimulant users and increase the likelihood of violence. In keeping with the literature, participants reporting higher levels of depressive symptomatology were two and a half times more likely to report past year victimisation. Perron and colleagues have demonstrated using path analysis that this relationship is mediated by an individual's perceived safety; that is, violent victimisation reduces one's perception of safety that in turn leads to increased severity of depressive symptoms. The relationship between depression, victimisation and homelessness requires further analysis to determine if depressive symptoms such as helplessness and despair contribute to prolonging episodes of homelessness. Although women had a three and a half times greater risk of victimisation than men in the univariate analysis, this effect disappeared in the multivariate analysis. It is possible that gender is an important factor in victimisation of homeless populations, but there was insufficient statistical power to detect this effect. Alternatively, the association between female gender and victimisation may be confounded by a third variable, such as depression, which is more prevalent among women compared to men. The small number of females in the sample precluded analysis of interactions between gender and depression; targeted recruitment of females in future research examining violent victimisation among homeless persons would be useful. Similarly, PTSD failed to retain significance in multivariate analysis, despite being associated with a three‐fold increase in risk of victimisation in univariate analysis. Depression and PTSD frequently co‐occur and a larger sample size may have permitted more sophisticated analysis of how these two conditions interact to predict victimisation. It is important to be mindful of the limitations of this study. Participants were recruited from a shelter specifically for substance users, and although substance use is highly prevalent among homeless persons, this sample may not be representative of the larger homeless population. As a cross‐sectional study, we were unable to identify what effect, if any, victimisation has on homelessness. A longitudinal study designed to determine whether violent victimisation prolongs homelessness is warranted. Finally, the study relied on self‐report measures of substance use, mental disorders and victimisation; hence there may be a degree of under‐reporting of these sensitive topics. In practical terms, this study has demonstrated that clinical staff working with homeless substance users need to be aware of the likelihood of prior victimisation, the potential for future victimisation, and the relationship between victimisation and poor mental health. This population may benefit from assistance in identifying risk situations for victimisation or from learning how to de‐escalate potentially violent situations. Acknowledgements This study was funded by the NSW Department of Health Drug and Alcohol Research Grants Program. NDARC is core‐funded by the Commonwealth Department of Health and Ageing. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Factors associated with violent victimisation among homeless adults in Sydney, Australia

Download PDF
5 pages

Loading next page...
 
/lp/wiley/factors-associated-with-violent-victimisation-among-homeless-adults-in-5VU0uSuJBb

References (27)

Publisher
Wiley
Copyright
© 2009 The Authors. Journal Compilation © 2009 Public Health Association of Australia
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1753-6405.2009.00406.x
pmid
19689595
Publisher site
See Article on Publisher Site

Abstract

It has been well‐established internationally that homeless people experience higher levels of violent victimisation, such as physical or sexual assaults, than housed populations. Just over half of the homeless adults interviewed in a large UK study reported having experienced violence in the past year, compared to just 4% of the broader population. A study in the US found that in the previous two months, 18% of a homeless sample had been threatened with a weapon, 16% had been beaten and 6% had been sexually assaulted. A number of risk factors for violent victimisation have been examined. Homeless women are at increased risk of sexual assault compared to homeless men, but gender does not appear to influence risk of other physical victimisation. Poor mental health is strongly indicative of increased risk, with schizophrenia, more severe psychotic symptoms, a history of psychiatric hospitalisation, and general psychological distress all having been associated with violent victimisation of homeless people. The relationship between substance use and the victimisation of homeless people is complex. Simons et al. found substance use to be moderately correlated with victimisation, and Kushel et al. similarly found both alcohol and illicit drug use to be associated with physical assault. Analysing victimisation by gender, Wenzel et al. found that women's victimisation was associated with illicit drug, but not alcohol dependence, with the opposite being found for men. Finally, Lam & Rosenheck noted that days of alcohol intoxication, but not days of drug use, was associated with violence. A shortcoming of these findings is the lack of discrimination between types of illicit drugs (e.g. heroin vs. cannabis), which may have very different effects on an individual's risk of victimisation. Victimisation of homeless people has consequences beyond physical and emotional injury. Although there has been little longitudinal research on homeless populations, it appears that violent victimisation may prolong homelessness, even more so than factors such as an individual's level of social support. Hence, identification of risk factors and preventive strategies for violent victimisation may be an important aspect of reducing chronic homelessness. With the release of the Federal Government's White Paper, The Road Home , homelessness has become a political ‘hot topic’ in Australia. There are approximately 105,000 homeless people in Australia, but very little research has been conducted with this population. A notable exception is the work of Buhrich, Teesson and Hodder, who produced a series of papers documenting the high prevalence of psychiatric and substance use disorders among Sydney's adult homeless. There has been some examination of how family violence precipitates homelessness among women and children, but to our knowledge no studies have examined victimisation of currently homeless men and women in Australia. This study aims to address this gap by describing the prevalence of, and factors associated with, violent victimisation among a sample of homeless adults in inner‐Sydney. Method A cross‐sectional design was used. Data were collected between January 2007 and April 2008 from clients of a shelter for homeless, substance‐using adults in inner‐Sydney. Potential participants were approached by an interviewer, who briefly described the study and invited the client to participate. No incentives for participation were offered. Clients indicating willingness to participate in the study completed informed consent procedures, with the interviewer emphasising that all information was confidential and that the participant was free to terminate the interview at any time. Participants were advised at the beginning of the interview that they could take a break at any time. A standardised interview was administered to each participant. Homelessness was assessed by asking participants how long they had been homeless, and where they slept (rough/outdoors, homeless shelter, boarding house, friends/family, hospital/drug treatment service, prison, squat, other) for most of the previous month. From this latter question, a variable was constructed categorising participants as primary (rough/outdoors, squat), secondary (homeless shelter, friends/family) or tertiary (boarding house, hospital/drug treatment service, prison, other) homeless. Alcohol and other drug use were assessed by asking the participant how many days in the past six months they had used each of heroin, psychostimulants (defined as amphetamine, methamphetamine or cocaine), benzodiazepines, alcohol and cannabis. To simplify analysis of the complex drug use histories reported by participants, a dichotomous variable was constructed for each drug type, with participants classed as either non/occasional users (less than once per week for the previous six months) or regular users (once a week or more for the previous six months) of each drug. Depressive symptoms were assessed using the Depression, Anxiety and Stress Scale (DASS). The DASS provides a score between 0 and 42, with higher scores indicating more severe symptoms. The raw scores were used to construct a variable categorising participants as low‐depression or high‐depression, with the median score used as the cut‐point. Post‐traumatic stress disorder was screened for using the Trauma Screening Questionnaire (TSQ), with a score of greater than five indicative of PTSD. Participants were also asked “have you ever been told by a doctor that you have schizophrenia or a psychotic disorder?” Violent victimisation was assessed by asking the participant “In the past 12 months, has anyone been physically violent toward you? By physically violent, I mean being hit, slapped, pushed, kicked or attacked with a weapon by someone to cause harm.” Data were analysed using SPSS 17. To determine factors associated with violent victimisation, univariate logistic regressions were conducted with past year victimisation (yes/no) as the dependent variable. Factors included in the univariate analyses were socio‐demographic characteristics (age, gender, indigenous status), homelessness variables (length of homelessness, homelessness category), mental health indicators (depressive symptoms, post‐traumatic stress disorder, schizophrenia or other psychotic disorder), and use of heroin, psychostimulants, benzodiazepines, alcohol and cannabis in the previous six months. Factors significant at the p <0.05 level in univariate analyses were entered into a multivariate logistic regression model. Cases with missing data were removed from the analysis. This model was examined and the factor with the largest p value was removed. This process was repeated until a model containing only significant factors was obtained. Results Data were obtained from 106 individuals, representing approximately 25% of clients accessing the shelter during the data collection period. One participant terminated the interview after completing only the demographic questions and was excluded from analysis, giving a sample size of 105. The sample was predominantly male (n=85, 81%) and the mean age was 41.5±11.1 (range 20‐76). Eighteen (17%) participants identified as Aboriginal or Torres Strait Islander. Length of homelessness and category of homelessness in the past month are shown in Table 1 . The majority of participants were categorised as secondary homeless, that is, they had slept at homeless shelters or at the homes of friends or family for most of the previous month. 1 Characteristics of homelessness. Length of homelessness n (%) <1month 22 (21) 1‐6 months 22 (21) 6‐24 months 33 (31.4) >24 months 28 (26.7) Category of homelessness (last month) Primary 30 (28.6) Secondary 59 (56.2) Tertiary 16 (15.2) Participants had complex alcohol and other drug use histories. Polydrug use was the norm, with participants using a median of three drug classes (excluding nicotine) in the previous six months. DASS depression scores suggested high levels of depressive symptoms. Forty‐two per cent of participants fell within the severe/extremely severe depression range, and a further third reported mild/moderate symptoms. The median depression score, for use in the regression analyses, was 18 (inter‐quartile range 10‐32). More than one‐third (n=39, 37.1%) of participants reported a lifetime diagnosis of schizophrenia or other psychotic disorder. Of the 90 participants who completed the TSQ, more than half (n=51, 56.7%) screened positive for PTSD. One hundred and two participants responded to the victimisation question. Of these, 49 (48%) reported that they had been a victim of violence in the previous 12 months. Univariate results for past year victimisation are shown in Table 2 . Being female, having a diagnosis of schizophrenia or other psychotic disorder, current PTSD, higher level of depressive symptoms and regular psychostimulant use were all associated with violent victimisation in the previous 12 months. 2 Past year violent victimisation. Univariate analyses Multivariate model Factor No violent victimisation n (%)* Violent victimisation n (%)* p OR (95% CI) p AOR (95% CI) Age Mean (SD) 43.2 (11.2) 40.6 (10.8) 0.25 0.98 (0.94‐1.02) Gender Male 48 (90.6%) 36 (73.5%) Female 5 (9.4%) 13 (26.5%) 0.03 3.47 (1.13‐10.61) Indigenous status Non‐indigenous 43 (81.1%) 41 (83.7%) Indigenous 10 (18.9%) 8 (16.3%) 0.74 0.84 (0.30‐2.34) Length of homelessness <1month 11 (20.8%) 9 (18.4%) 0.96 referent 1‐6 months 12 (22.6%) 10 (20.4%) 0.98 1.02 (0.30‐3.44) 6‐24 months 17 (32.1%) 16 (32.7%) 0.81 1.15 (0.38‐3.51) >24 months 13 (24.5%) 14 (28.6%) 0.64 1.32 (0.41‐4.20) Homeless category Primary 18 (34.0%) 11 (22.4%) 0.41 referent Secondary 28 (52.8%) 29 (59.2%) 0.26 1.70 (0.68‐4.22) Tertiary 7 (13.2%) 9 (18.4%) 0.24 2.10 (0.61‐7.27) Psychotic disorder or schizophrenia No 41 (77.4%) 23 (46.9%) Yes 12 (22.6%) 26 (53.1%) 0.002 3.86 (1.65‐9.07) 0.02 3.13 (1.24‐7.90) PTSD No 26 (57.8%) 13 (28.9%) Yes 19 (42.2%) 32 (71.1%) 0.007 3.37 (1.40‐8.08) Depressive symptoms Low 33 (63.5%) 19 (40.4%) High 19 (36.5%) 28 (59.6%) 0.02 2.56 (1.14‐5.8) 0.04 2.65 (1.07‐6.59) Heroin use None/occasional 48 (90.6%) 39 (79.6%) Regular 5 (9.4%) 10 (20.4%) 0.13 2.46 (0.78‐7.80) Psychostimulant use None/occasional 48 (90.6%) 31 (63.3%) Regular 5 (9.4%) 17 (34.7%) 0.003 5.27 (1.76‐15.73) 0.008 5.07 (1.53‐16.84) Benzodiazepine use None/occasional 43 (81.1%) 38 (77.6%) Regular 10 (18.9%) 11 (22.4%) 0.66 1.25 (0.48‐3.26) Alcohol use None/occasional 19 (35.8%) 17 (34.7%) Regular 34 (64.2%) 32 (65.3%) 0.90 1.05 (0.47‐2.37) Cannabis use None/occasional 30 (56.6%) 29 (59.2%) Regular 23 (43.4%) 19 (38.8%) 0.70 0.86 (0.39‐1.89) Notes: *n varies between factors due to missing data The five variables significant in the univariate analyses were assessed for collinearity and found to be adequately independent. Three iterations were undertaken to reach the final multivariate model (Hosmer and Lemeshow χ =2.3, df=5, p =0.81), which accurately predicted victimisation in 72% of cases. In this model, regular psychostimulant use was associated with a five‐fold increase in risk of victimisation. Diagnosis of schizophrenia/psychotic disorder and higher level of depressive symptoms were also significantly associated with increased risk of victimisation ( Table 2 ). Discussion This study makes an important contribution to the literature as the first to consider violent victimisation among homeless people in Australia. Just under half of this sample reported past year victimisation. Directly comparable data from a general community sample are not available, but it is worth noting that the recorded assault rate in NSW in 2007 was around 1 per 1,000 persons, a great deal lower than that experienced by this homeless sample. In addition to the generally high level of violence, this study has identified distinct factors that are associated with an increased risk of violent victimisation of homeless persons. Participants with a history of schizophrenia or other psychotic disorder were 3.1 times more likely to be victimised than those without such a history. This aligns with previous research with homeless populations, as well as research focusing specifically on victimisation of individuals with psychotic disorders. For example, in a large Australian sample of individuals with a psychotic disorder, past year prevalence of violent victimisation was 18%. The relationship between psychosis and victimisation is theorised to be a result of the symptoms of psychosis; that is, impaired judgement affects one's ability to identify risk, while symptoms such as talking to oneself and disordered behaviour draw attention to people with psychotic disorders, increasing the likelihood of violence. Given that homeless people spend large amounts of time in public spaces, these effects may be magnified for homeless, psychotic individuals, resulting in even greater risk of violent victimisation. These findings underscore the importance of providing safe, secure accommodation for people with severe mental illnesses. Although substance use in general has previously been associated with victimisation, in this analysis regular psychostimulant use, but not heroin, alcohol, cannabis or benzodiazepine use, was associated with a five‐fold increase in past year violent victimisation. Research on psychostimulant use and violence tends to focus on whether users are perpetrators of violence, with little recognition that psychostimulant use may contribute to victimisation. As with schizophrenia, it is possible that the behavioural effects of psychostimulants – for example, feelings of strength or invincibility, talkativeness and reduced perception of risk – may draw attention to psychostimulant users and increase the likelihood of violence. In keeping with the literature, participants reporting higher levels of depressive symptomatology were two and a half times more likely to report past year victimisation. Perron and colleagues have demonstrated using path analysis that this relationship is mediated by an individual's perceived safety; that is, violent victimisation reduces one's perception of safety that in turn leads to increased severity of depressive symptoms. The relationship between depression, victimisation and homelessness requires further analysis to determine if depressive symptoms such as helplessness and despair contribute to prolonging episodes of homelessness. Although women had a three and a half times greater risk of victimisation than men in the univariate analysis, this effect disappeared in the multivariate analysis. It is possible that gender is an important factor in victimisation of homeless populations, but there was insufficient statistical power to detect this effect. Alternatively, the association between female gender and victimisation may be confounded by a third variable, such as depression, which is more prevalent among women compared to men. The small number of females in the sample precluded analysis of interactions between gender and depression; targeted recruitment of females in future research examining violent victimisation among homeless persons would be useful. Similarly, PTSD failed to retain significance in multivariate analysis, despite being associated with a three‐fold increase in risk of victimisation in univariate analysis. Depression and PTSD frequently co‐occur and a larger sample size may have permitted more sophisticated analysis of how these two conditions interact to predict victimisation. It is important to be mindful of the limitations of this study. Participants were recruited from a shelter specifically for substance users, and although substance use is highly prevalent among homeless persons, this sample may not be representative of the larger homeless population. As a cross‐sectional study, we were unable to identify what effect, if any, victimisation has on homelessness. A longitudinal study designed to determine whether violent victimisation prolongs homelessness is warranted. Finally, the study relied on self‐report measures of substance use, mental disorders and victimisation; hence there may be a degree of under‐reporting of these sensitive topics. In practical terms, this study has demonstrated that clinical staff working with homeless substance users need to be aware of the likelihood of prior victimisation, the potential for future victimisation, and the relationship between victimisation and poor mental health. This population may benefit from assistance in identifying risk situations for victimisation or from learning how to de‐escalate potentially violent situations. Acknowledgements This study was funded by the NSW Department of Health Drug and Alcohol Research Grants Program. NDARC is core‐funded by the Commonwealth Department of Health and Ageing.

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Aug 1, 2009

There are no references for this article.