Power Point Help

Journal of Interpersonal Violence 2014, Vol. 29(6) 987 –1005

© The Author(s) 2013 Reprints and permissions:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260513506054

jiv.sagepub.com

Article

Trajectories of Recurring Victimization Among People With Major Mental Disorders

Brent Teasdale, PhD,1 Leah E. Daigle, PhD,1 and Ellen Ballard, MS1

Abstract Relatively little is known about the violent victimization experiences of people with major mental disorders. Moreover, to date, no studies have examined recurring violent victimization experiences of people with major mental disorders. Using a risk heterogeneity framework commonly used in the study of recurring victimization, the current study examines the extent of recurring victimization among people with Diagnostic and Statistical Manual of Mental Disorders (DSM) Axis I mental disorders and trajectories of recurring violent victimization (n = 262), across four waves of data collected during a 1-year longitudinal study. Multilevel logistic regression analyses tested disorder, time, and time by disorder cross-level interactions predicting recurring victimization. Results suggest that recurring violent victimization is not uncommon among mentally disordered victims of violence, with 64% of victims experiencing a recurring victimization at a later point in time. However, trajectories of recurring violent victimization are not uniform across types of mental illness. Indeed, individuals diagnosed with a substance abuse disorder or major depression show significantly declining trajectories across the follow-up period whereas individuals diagnosed with a manic disorder or a schizophrenia spectrum disorder have flat trajectories of recurring violent victimization across the study period. Results of tests

1Georgia State University, Atlanta, USA

Corresponding Author: Brent Teasdale, Department of Criminal Justice and Criminology, Georgia State University, P.O. Box 4018, Atlanta, GA 30303-4018, USA. Email: [email protected]

506054 JIV29610.1177/0886260513506054Journal of Interpersonal ViolenceTeasdale et al. research-article2013

988 Journal of Interpersonal Violence 29(6)

for cross-level interactions between disorder type and time demonstrate that individuals with a major depression or substance abuse/dependence diagnosis are significantly different from those with a schizophrenia spectrum diagnosis in their trajectories of recurring victimization.

Keywords recurring violent victimization, mental disorders, psychiatric symptomatology, risk heterogeneity

Recurring Victimization and Mental Illness

Although a great deal is known about the extent and correlates of violence perpetration by people with major mental disorders, relatively little is known about the violent victimization experiences of people with mental disorders. Indeed, little empirical research has focused on violence committed against people with mental illnesses. The studies that do exist suggest that people with mental illness are more likely to be victimized than their nondisordered counterparts (Goodman et al., 2001; Hiday, Swanson, Swartz, Borum, & Wagner, 1999, 2001, 2002; Lehman & Linn, 1984; Marley & Buila, 2001; Silver, Arseneault, Langley, Caspi, & Moffitt, 2005; Teplin, McClelland, Abram, & Weiner, 2005; Walsh et al., 2003). In fact, prevalence rates of vio- lent victimization for individuals with mental disorders range from a rate approximately 2.5 times (Hiday et al., 1999, 2001; Walsh et al., 2003) to four times greater (Teplin et al., 2005) than the rate of violent victimization found in the general population (for a review, see Maniglio, 2009.)

In studying the risk factors for the violent victimization of people with major mental disorders, studies have emphasized that the rate of victimiza- tion differs by diagnostic category (Hiday et al., 1999; Silver et al., 2005). In an early study, Chuang, Williams, and Dalby (1987) found that schizophrenic patients experienced an increased risk of violent victimization. In a study examining the association between specific clinical diagnoses and violent victimization, Silver et al. (2005) found that individuals with alcohol depen- dence disorders were almost twice as likely to be victims of a completed physical assault, and persons with marijuana dependence disorders had a lit- tle more than 2 times the odds of an attempted assault when compared with people without mental disorders.

Other research has examined symptomatology (Brekke, Prindle, Bae, & Long, 2001; Hiday et al., 2002; Teasdale, 2009; Walsh et al., 2003), comorbid personality disorders (Walsh et al., 2003), comorbid substance abuse (Brekke et al., 2001; Goodman et al., 2001; Hiday et al., 1999, 2002; Silver et al.,

Teasdale et al. 989

2005; Teasdale, 2009; Walsh et al., 2003), history of abuse (Goodman et al., 2001; Walsh et al., 2003), and demographic characteristics (Hiday et al., 1999, 2002; Silver, 2002; Silver et al., 2005; Teasdale, 2009; Teplin et al., 2005; Walsh et al., 2003) as risk factors for violent victimization of persons with major mental disorders. One possible explanation for the link between psychiatric symptomatology and victimization has relied on the concept of conflicted social relationships (Silver, 2002; Teasdale, 2009). Teasdale (2009) asserted that individuals, especially those who are symptomatic, may moti- vate others to offend against them by failures to obey interaction rituals (Felson, 1992) or through engaging in conflicted social relationships (Silver, 2002), increasing the likelihood of being victimized. That is, behaviors that are inappropriate for the situation may provoke attack or signal that an indi- vidual is an easy target. This reasoning falls in line with Felson’s (1992) social interactionist theory of violence which proposes that aggression is an instrumental attempt to socially control rule violators (see also Silver, 2002; Silver et al., 2005; Teasdale, 2009).

In spite of the recent advances in the knowledge base surrounding vio- lence against people with mental disorders, lacking is an investigation into the recurring victimization experiences among this population. This gap is particularly surprising given the attention that recurring victimization has garnered in the victimization literature and the high risk of violent victimiza- tion that persons with mental disorders face. Research on recurring victim- ization has revealed a startling finding—a small percentage of victims experience more than one victimization event, even over a relatively short period of time (Fisher, Daigle, & Cullen, 2010; Gidycz, Hanson, & Layman, 1995; Lauritsen & Davis Quinet, 1995; Mele, 2009; Pease, 1998; Sagovsky & Johnson, 2007). For example, results from the British Crime Survey revealed that of those who experienced victimization during the previous 12 months, 28% experienced more than one incident. Recurring victimization was greatest for victims of domestic violence—44% of domestic violence victims indicated that they had experienced more than one incident during the previous 12 months. Another finding from this body of research was that victims who experienced more than one incident experienced a dispropor- tionate share of all victimizations. In Daigle, Fisher, and Cullen’s (2008) study of college women, just more than 7% of the college women experi- enced more than one sexual victimization incident, but these women experi- enced more than 72% of all of the sexual victimization incidents occurring during a single academic year.

To understand why some people are not victimized at all, some are victim- ized a single time, and still others are victimized time and again, two para- digms have been proffered—risk heterogeneity and state dependence (Farrell,

990 Journal of Interpersonal Violence 29(6)

Phillips, & Pease, 1995; Pease, 1998). The risk heterogeneity paradigm sug- gests that individuals’ risk factors for victimization will place them at risk of an initial victimization and, if left unchanged, will keep them at risk of sub- sequent victimization. Individuals who consistently engage in risky behav- iors, behave in bizarre and annoying ways (see, for example, Felson, 1992; Silver, 2002), or possess other risk characteristics, such as psychiatric symp- tomatology, have higher rates of victimization across time, compared with individuals who do not have a risky behavioral profile, motivate others to offend against them, or possess other attributes that put them at risk of vic- timization. Paralleling this development, the state-dependence paradigm sug- gests that the victimization event itself impacts future victimization risk. It is how the victim and the offender respond during and after the initial victimiza- tion incident that either increases or decreases the probability of recurring victimization. For example, an offender who successfully burglarizes a home learns how to enter the home, what items are available to steal, and the layout of the home. The burglar also learns that he can steal from the home without being arrested. These pieces of information will serve to impact that home’s chances of future burglary.

In addition to establishing the extent to which recurring victimization occurs, evidence for both state dependence and risk heterogeneity has been established in the existing empirical research on recurring victimization (Fisher et al., 2010; Lauritsen & Davis Quinet, 1995; Tseloni & Pease, 1998). Although not explicit tests of state dependence or risk heterogeneity, research has found that household (Mukherjee & Carcach, 1998; Osborn & Tseloni, 1998; Tseloni, 2000) and individual characteristics (Gabor & Mata, 2004; Lasley & Rosenbaum, 1988; Mukherjee & Carcach, 1998; Outlaw, Ruback, & Britt, 2002; Tseloni, 2000) are related to recurring victimization. These household and individual characteristics are conceived of as risk heterogene- ity factors that place individuals at risk of subsequent victimization.

Although there appears to be evidence that individuals with mental disor- ders are at increased risk of recurring victimization, no studies have exam- ined the prevalence or correlates of recurring victimization in this population. In a sample of 172 individuals treated for schizophrenia or schizoaffective disorder, Brekke et al. (2001) found that 65 individuals reported a total of 118 separate victimizations. The vast majority (91%) of these victimization expe- riences were violent victimizations. Similarly, in a sample of 265 individuals with major mental disorders, Marley and Buila (2001) found that, during their lifetimes, women reported experiencing an average of 10.33 types of crimes and men reported experiencing an average of 8.80 types of crimes. These numbers highlight that recurring victimization is occurring to victims with mental disorders. Beyond these studies that merely identified that

Teasdale et al. 991

recurring victimization is occurring, recurring victimization has not been examined in the extant literature on victimization of persons with mental disorders. This is a surprising omission given the developing literature that shows that victimized persons are at an increased risk of experiencing a sub- sequent victimization. This pattern may indeed hold true for mentally disor- dered individuals. If so, prevention and intervention efforts would need to be specifically designed and implemented to intervene in the victimization- recurring victimization nexus for this population.

To fill this gap, in the current study, we investigate whether mentally dis- ordered individuals are at risk of experiencing recurring victimization and whether diagnosis correlates with trajectory of recurring victimization. In doing so, we borrow from the risk heterogeneity paradigm to suggest that the trajectories of recurring victimization for people with Diagnostic and Statistical Manual of Mental Disorders (DSM) Axis I mental disorders vary depending on the specific disorder (schizophrenia spectrum disorders, manic spectrum disorders, major depression, or substance abuse/dependence disor- ders). Based on the risk heterogeneity perspective, it is unlikely that all indi- viduals experience the same probability of recurring victimization across time. Moreover, based on the literature that establishes variations in victim- ization by disorder type, we expect variations in the trajectories of recurring victimization to vary by disorder type.1 We examine the prevalence of recur- ring victimization and the shape and correlates of recurring victimization tra- jectories for persons diagnosed with mental illnesses using longitudinal data from the MacArthur Violence Risk Assessment Study (MacRisk).

Method

To examine recurring victimization among people with major mental disor- ders and to test the ideas presented above, we analyze data from the MacArthur Violence Risk Assessment Study (MacRisk). The MacRisk study is a longi- tudinal study of individuals released from inpatient psychiatric hospitals at three sites (Worcester, Massachusetts; Pittsburgh, Pennsylvania; and Kansas City, Missouri). The data collection involved a baseline survey in the hospital and five follow-up waves of data collection in the community, after the indi- vidual was released from the hospital. These five follow-up waves were spaced approximately 10 weeks apart and covered the 1 year following release from the hospital. The five community-based follow-up data collec- tion efforts were interview based and provided the information on victimiza- tion examined for the current project. Data collection for the project began in 1992 and concluded in 1995. Subjects were continuously enrolled (n = 1,136), so the 1-year data collection period ended in 1993 for some subjects,

992 Journal of Interpersonal Violence 29(6)

while for others enrolled in 1994, data collection concluded in 1995. (For additional detail regarding the MacRisk study, see Monahan et al., 2001; Silver, 2002; Silver, Mulvey, & Monahan, 1999; Steadman et al., 1998; Teasdale, 2009; Teasdale, Silver, & Monahan, 2006.)

Sampling

To be considered eligible for the study, psychiatric hospital patients were screened for inclusion criteria. Only English-speaking, White and African American individuals (and Latinos at the Worcester site) were included in the MacRisk study. Individuals were required to have a primary chart diagnosis of substance abuse/dependence disorder, schizophrenia spectrum disorder, manic spectrum disorder, or major depression. Participants had to be between 18 and 40 years of age. Patients admitted through criminal proceedings were excluded. Patients who had been hospitalized for longer than 20 weeks were also excluded. The MacRisk study used a stratified random sampling design. Eligible individuals were stratified by gender, race, and age, and a simple random sample within each of the strata was conducted until a quota for each stratum was met. Informed consent to participate was obtained from 71% (n = 1,203) of the subjects approached to participate in the study, and 94% (n = 1,136) of those who consented subsequently completed a baseline interview. Of those who completed a baseline interview, 84% (n = 951) completed at least one community follow-up and 72% (n = 818) completed three or more follow-up interviews. Institutional review board approval for the study was obtained at Policy Research Associates and at each of the three research sites (Worcester, Massachusetts; Pittsburgh, Pennsylvania; and Kansas City, Missouri).

Measures

Violent victimization was based on subject self-reports to the following items: (a) “Since the last interview, has anyone hit you with a fist or object or beaten you up?” (b) “Since the last interview, has anyone threatened you with a knife or gun or other lethal weapon?” (c) “Since the last interview, has any- one used a knife or fired a gun at you?” (d) “Since the last interview, has anyone thrown something at you?” (e) “Since the last interview, has anyone slapped you?” and (f) “Since the last interview, has anyone kicked, bitten, or choked you?” At each community follow-up, subjects who responded yes to any of the above questions were coded 1 and subjects who responded no to all the above questions were coded 0. We selected only individuals who were coded as 1 on the first community follow-up victimization measure for the

Teasdale et al. 993

current study (n = 262). Thus, we are able to track the recurring victimization experiences of this cohort of victims (t = 1) from the second through fifth community follow-up interviews.

The dependent variable for the current project was whether or not an indi- vidual experiences a recurring violent victimization incident at the second through the fifth waves of data collection, given that we selected only victims at the first follow-up. For each follow-up wave, recurring victimization is coded as 1 if the respondent answered yes to any of the violent victimization questions and is coded 0 if they responded no to all of the violent victimiza- tion questions for that wave.

In addition to the recurring victimization measure, we also utilized diag- nostic categories as our primary independent variables of interest. A baseline interview of each participant was conducted by a trained clinical interviewer (in the psychiatric hospital); participants were coded based on Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev.; DSM-III-R; American Psychiatric Association, 1987) criteria and divided into four pri- mary diagnostic categories: (a) major depression, (b) schizophrenia spectrum disorders, (c) manic spectrum disorders, and (d) substance abuse/dependence disorders. Although some diagnostic criteria for these disorders may have been revised in subsequent revisions of the DSM, use of this edition in this study was unavoidable. DSM-III-R was the most current edition at the time of data collection. Diagnoses made by research staff were checked against chart diagnoses and disagreements were resolved through meetings between research and clinical staff.

Time is also a substantive variable of interest in these analyses. We coded time as linear and quadratic to allow for the decelerating curve we found in the graphs of the observed data (see Figure 1). Time (both linear and qua- dratic) by diagnosis cross-level interactions were also included to test the idea that the trajectories of recurring victimization vary by diagnostic group- ings. That is, the cross-level interactions tested the hypothesis that the time effect on recurring victimization (the time trend in recurring victimization) is not the same for all disorders.

In addition to the substantive variables of interest, the analyses also con- trol for psychiatric symptomatology by including scores on the Brief Psychiatric Rating Scale (BPRS). The BPRS is an 18-item scale on which clinically trained interviewers rate the participants from 1 (mild) to 7 (very severe) on a variety of dimensions of psychiatric disorders (see Teasdale, 2009, for a complete listing of the 18 BPRS items utilized in the current study). The assessment of BPRS was made at each follow-up interview. It is expected that changes in symptomatology will be reflected in participants’ BPRS score. That is, fluctuations in BPRS may also reflect changes such as

994 Journal of Interpersonal Violence 29(6)

remission. Although diagnosis is assessed once at the baseline interview, stabil- ity of diagnosis across all waves of data collection is assumed and controlled for by symptomatology. Changes in BPRS score may signal increases in disor- der severity as well as periods of remission. We also controlled for Michigan Alcoholism Screening Test (MAST) scores and Drug Abuse Screening Test (DAST) scores. The MacRisk study used a shortened version of MAST (Pokorny, Miller, & Kaplan, 1972), which is designed to assess self and other perceptions as a problematic drinker, problematic drinking behaviors, and con- sequences of drinking. The DAST (Skinner, 1982) was designed to assess con- sequences of drug use and help-seeking behaviors related to drug use.

Data Analysis

Analyses of longitudinal data require special analytic methods. To take into account the complex nature of the data, we utilized hierarchical binomial regression models, in the current project. These models allow for the nesting of waves of data within individuals and for the dichotomous distribution of

Figure 1. Proportion experiencing recurring victimization across time, by disorder.

Teasdale et al. 995

the outcome variable. These models produce unbiased standard error esti- mates, by taking into account the dependence of the waves of data common to an individual respondent. We follow closely the analysis plan laid out by Horney, Osgood, and Marshall (1995) in their analysis of local life circum- stances data. That is, the hierarchical analysis presented here provides both within-person (Level 1) and between-person (Level 2) models (see also Teasdale, 2009). Variables that change across time (time, BPRS, MAST, and DAST) are included at the within-person level of analysis (Level 1). Variables that are stable across time (diagnosis) are included at the between-person level (Level 2). We also allowed for the cross-level interaction between time and diagnosis. These models are especially well suited for modeling change across time as well as for modeling the shape and correlates of observed vari- able growth trajectories (both foci of the current study).

Results

The sample for this project comprises only those individuals who reported a violent victimization incident during the first community follow-up (n = 262). Table 1 presents the mean (or proportion for dichotomous variables), standard deviation, minimum, and maximum values for study variables. As shown in Table 1, the sample included primarily males (59%). Most of the participants were White (66%) and lived in Kansas City (46%). Almost 32% lived in Pittsburgh, and 22% were from Worcester. The average age of the participants was 29 years (SD = 6.07). Most of the participants were diag- nosed with a depressive disorder (42%). Approximately 27% were diagnosed

Table 1. Descriptive Statistics of Those Experiencing Violent Victimization (N = 262).

M SD Min. to Max.

Male 0.591 0.493 0-1 White 0.664 0.473 0-1 Pittsburgh 0.318 0.466 0-1 Kansas City 0.464 0.500 0-1 Worcester 0.219 0.414 0-1 Schizophrenia 0.146 0.354 0-1 Major depression 0.420 0.494 0-1 Manic 0.124 0.330 0-1 Substance abuse 0.266 0.443 0-1 Age 29.3 6.07 18-40

996 Journal of Interpersonal Violence 29(6)

with a substance abuse disorder, 15% were diagnosed with a schizophrenia spectrum disorder, and 12% were diagnosed with a manic spectrum disorder. Approximately 23% (262 of 1,136) of the total sample were victims at Time 1. Of the victims at Time 1, approximately 64% (168 of 262) were recurring victims at some point during the 1-year follow-up.

At the bivariate level, we examined the observed data for the proportion reporting violent victimization at each wave, given that they were victims at Wave 1. As shown in Table 2, averaged across disorders, there appears to be a large decline in the proportion who experienced recurring victimization as time increased, from 47.9% of victims experiencing recurring victimization at Wave 2 to 25.6% of victims experiencing recurring victimization at Wave 5. The largest decline occurred between Follow-Ups 2 (47.9%) and 3 (31.5%). This apparent decline across time obscures the variation across disorders in the trajectories of recurring victimization. To better represent these time trends, the data are presented graphically in Figure 1. Note, for example, that recurring victimization remained relatively constant over time for individuals diagnosed with manic or schizophrenia spectrum disorders, but declined sharply for individuals with major depression or substance abuse disorders.

Although the descriptive data are interesting and informative, we turn to a multivariate assessment, because this allowed us to produce significance tests for the differences between the trajectories, while adjusting for covari- ates and key data elements (the nested data structure and dichotomous out- comes discussed earlier). As shown in Table 3, results of the multilevel logistic regressions suggest that the disorder categories differed in their inter- cept (Time 2 recurring victimization). Specifically, at Time 2, those diag- nosed with major depression or an alcohol or drug dependence disorder had significantly higher odds of being victimized again, compared with those diagnosed with a schizophrenia spectrum disorder. Those diagnosed with a manic spectrum disorder were not significantly different from those with a schizophrenia diagnosis, at Time 2 (second follow-up wave after release from the hospital). This closely mirrored our observed results (see Figure 1).

Table 2. Proportion Reporting Recurring Violent Victimization by Wave and Diagnosis.

Wave 2 3 4 5

Schizophrenia .297 .344 .344 .278 Depression .525 .255 .284 .256 Mania .321 .308 .250 .292 Substance abuse .569 .364 .294 .222 Total .479 .315 .304 .256

Teasdale et al. 997

There were significant within-person effects of BPRS and MAST scores on the experience of recurring victimization. Specifically, waves in which an individual had higher scores on the BPRS and MAST, the odds of a recurring victimization experience were significantly increased, compared with waves in which that individual had a lower score on the BPRS or MAST. That is, those with more problematic use of alcohol and those with higher symptom severity scores were more at risk of recurring victimization. Specifically, the odds of recurring victimization were multiplied by 1.03 for each one-unit increase on the BPRS and the odds of recurring victimization were multiplied by 1.24 for each one-unit increase on the MAST.

Finally, we estimated wave by disorder interaction effects predicting recurring victimization. Here, we found that compared with those diagnosed with schizophrenia spectrum disorders, those with major depression or alco- hol or drug abuse dependence disorders had significantly declining slopes (see also Figure 1). In addition, compared with those with a schizophrenia spectrum disorder, those with a major depression diagnosis had a significant

Table 3. Multilevel Logistic Regression Predicting Recurring Violent Victimization.

b SE OR

Level 2 Intercept −2.35 1.64 0.10 Depression 5.79** 2.01 326.57 Mania 2.45 2.59 11.55 Alcohol/drug dependence 4.80* 2.07 121.08 Level 1 Wave 1.25 0.97 3.49 Wave2 −0.20 0.13 0.82 BPRS 0.03** 0.01 1.03 MAST 0.22** 0.08 1.24 DAST 0.12 0.17 1.13 Cross-level interactions Wave × Depression −3.58** 1.21 0.03 Wave × Mania −1.70 1.57 0.18 Wave × Alcohol/drug −2.64* 1.26 0.07 Wave2 × Depression 0.49** 0.17 1.64 Wave2 × Mania 0.25 0.22 1.29 Wave2 × Alcohol/drug 0.34† 0.18 1.40

Note. OR = odds ratio; BPRS = Brief Psychiatric Ratings Scores; MAST = Michigan Alcoholism Screening Test; DAST = Drug Abuse Screening Test. †p < .10. *p < .05. **p < .01.

998 Journal of Interpersonal Violence 29(6)

interaction with the wave squared term. This finding finding indicated that the time course of recurring victimization declined significantly and then that decline decelerated. That is, the curve declined sharply initially but then flat- tened out in later waves. This flattening of the curve was significant for those diagnosed with major depression. Although not significant at the p < .05 level, the wave squared by alcohol/drug dependence disorder interaction was marginally significant (p =.054).2 This indicated that the sharp decline in recurring victimization in the early waves tended to flatten out in the later waves. This paralleled the depression trajectory (see also Figure 1).

Discussion

The results of the current study suggest that individuals diagnosed with major mental illnesses are indeed at risk of recurring victimization, but that recur- ring victimization risk varies by disorder type. For example, of the 262 vic- tims, 64% were recurring victims, but only 40% of individuals diagnosed with schizophrenia spectrum disorder and 67% of those diagnosed with a manic disorder were recurring victims across the study period. Not only the prevalence of recurring victimization, but also the time course of recurring victimization risk varied by disorder type. Individuals diagnosed with major depression or substance abuse/dependence disorders are most at risk of recur- ring victimization in the time period immediately following the initial victim- ization. Their risk declines sharply following a recurrence and then flattens out. This finding that individuals are at increased risk of recurring victimiza- tion in the time period immediately following their initial victimization cor- responds to findings in the repeat sexual victimization literature (Daigle et al., 2008), repeat intimate partner violence victimization literature (Mele, 2009), and the repeat burglary literature (Polvi, Looman, Humphries, & Pease, 1990, 1991; Townsley, Homel, & Chaseling, 2000). Persons with these disorders may be most at risk of recurring victimization because they engage in risky lifestyles (e.g., drug and alcohol use) and lack capable guard- ianship. Indeed, our current findings suggest that for each one-unit increase in MAST scores, the odds of recurrent victimization are multiplied by 24%. Recurring victimization risk may decline across time because multiple fac- tors stabilize as time from hospitalization increases (see, for example, Teasdale, 2009).

Individuals diagnosed with manic or schizophrenia spectrum disorders had lower initial risk of recurring victimization but their risk remained flat across the 1-year study period. One possible explanation for this stable risk of recurring victimization relies on the nature of these disorders. It is possible that major depressive episodes and substance dependence disorders both

Teasdale et al. 999

stabilize more quickly after release from an inpatient hospitalization, but schizophrenia and manic spectrum disorders take significantly longer than our 1-year follow-up to stabilize. Thus, features of the disorders may account for the varying trajectories of recurring victimization. In addition, both alco- hol abuse (as measured by MAST scores) and psychiatric symptomatology (as measured by BPRS scores) are significant within-person correlates of recurring victimization risk. That is, during times in an individual’s life when she or he is more symptomatic or having more alcohol abuse problems, that individual is at a greater risk of a recurring victimization event. This within- person approach helps build on the current explanations of and methods for investigating recurring victimization. It is not just time-stable individual- level risks that matter, but also time variant factors that shape a person’s risk of recurring victimization. When these risk factors are high, such as symp- tomatology or alcohol abuse, then risk of victimization is likely also high. This finding is in line with a risk heterogeneity approach—there are underly- ing risk factors for victimization among our sample of individuals. It is also in line with the interactionist perspective offered by Felson. That is, when individuals are highly symptomatic or abusing alcohol, they may be less capable of performing appropriate interaction rituals, resulting in violations of normative expectations and social control attempts that increase the likeli- hood of victimization.

The findings of the current study are not trivial. The recurring victimiza- tion burden for individuals with major mental disorders is significant (rang- ing from 40% to 60% depending on diagnosis). This burden of recurring victimization is associated with significant mental health risks (Finkelhor, Ormrod, & Turner, 2007, 2009). For example, one study found that polyvic- timization was a significant predictor of mental health issues including anger, depression, and anxiety (Finkelhor, Ormrod, Turner, & Hamby, 2005). Another study of child victims found that polyvictimization increased the likelihood of the victims experiencing posttraumatic stress disorder, major depressive disorder, and substance use disorders (Ford, Elhai, Connor, & Frueh, 2010). Based on our findings, individuals with mental disorders are acutely at risk of recurring victimization and potentially its consequences.

These results taken together suggest that individuals released from an inpatient psychiatric hospitalization who are victimized are at a significant risk of recurring victimization and that risk is not uniform across diagnostic categories or across individuals. Those individuals diagnosed with substance abuse or major depression may be at an initially increased risk of recurring victimization (one that declines across time), while individuals diagnosed with schizophrenia or manic spectrum disorders may be stably at risk of recurring victimization during the first year after release from an inpatient

1000 Journal of Interpersonal Violence 29(6)

stay. In addition, individuals with greater symptomatology and alcohol abuse problems are also at an increased risk of recurring victimization. These pat- terns have significant implications for mental health service providers.

Community-based mental health care providers should be attuned to their clients’ risk of victimization, as well as recurring victimization. During times when a client is exhibiting increased symptomatology and alcohol abuse, clinicians should be particularly diligent in managing victimization risks. Doing so is particularly important in that if recurring victimization can be prevented, it will reduce a significant amount of the total victimizations experienced given that recurring victims experience a disproportionate share of all victimizations (Barberet, Fisher, & Taylor, 2004; Daigle et al., 2008; Lauritsen & Davis Quinet, 1995; Pease, 1998). Additional research may assist in the development of risk assessment devices for victimization and recurring victimization that will facilitate clinicians’ assessments of these dynamic (time-varying) risks. We encourage the development of these risk prediction devices.

Clinicians already manage and assess risks of violent behavior; however, the current risk assessment literature has neglected victimization. Increased focus on risk assessment and risk management for victimization and recurring victimization should be prioritized. In the current study, we find that trajecto- ries of recurring victimization vary by diagnosis and recurring victimization experiences vary by symptomatology and alcohol use. Clinicians can be sen- sitized to recurring victimization risks when symptoms or alcohol use are high or for those immediately released from inpatient hospitalizations (particularly for those with depression or substance abuse diagnoses). It is important to highlight the distinction between risk assessment and risk management. Risk assessment tends to be implemented as a one-time cross-sectional picture of risk. This is as opposed to risk management, which unfolds in a longitudinal fashion, as clinicians manage risks in the course of their treating a client. Findings from longitudinal analyses, such as those presented here, may be most informative for risk management, as the time course of the recurring victimization is most relevant to these ongoing managements of risk.

Involuntary outpatient commitments represent one possible tool to pre- vent recurring victimization, for clinicians or family members of persons with major mental disorders. Hiday and colleagues’ (2002) work suggests that involuntary outpatient commitment reduces victimization experiences of individuals with diagnosed mental illness. These civil commitment proceed- ings may also reduce recurring victimization risk of individuals who have been previously victimized. Involuntary outpatient commitment is conceived of as a less restrictive option than forced hospitalization for individuals suf- fering from mental disorders who are at risk of harming themselves or others.

Teasdale et al. 1001

The commitment places the individual in the care of a community-based mental health service provider. The regular treatment and medication regimes provided by an involuntary outpatient commitment may address the risk fac- tors we note for recurring victimization in the current study (e.g., alcohol abuse and psychiatric symptomatology). In addition, other risk factors for recurring victimization may also be addressed (stable living situations, improved social functioning, etc.; Hiday et al., 2002).

Limitations

This study, like any other, is not without limitations. Indeed, although this is one of the largest longitudinal studies of individuals diagnosed with mental ill- nesses, the sample size of victims is rather small (n = 262). A larger sample size would permit the exploration of additional groupings of disorder types. Moreover, the study was limited to a 1-year follow-up. While rates of recurring victimization were quite high during that 1 year (averaging 64%), a longer follow-up period would have extended the trajectories, allowing for the possi- bility of more complex time trends. Because diagnosis categories are based on diagnosis at the baseline interview, diagnosis stability is assumed and con- trolled for by the inclusion of symptomatology measures. In addition, the data did not permit us to tease out the relation of the victim and perpetrator. Finally, the project was limited to a dichotomous indicator irrespective of the frequency of recurring victimization. Additional research may pursue the amount of recurring victimization rather than just its prevalence. Future researchers may wish to expand the follow-up period, measure the actual amount of recurring victimization, and utilize a larger sample size to address these limitations.

Another potential limitation of the data is the age of the data. The data are from the mid-1990s. Multiple revisions to the DSM have occurred since the collection of the data; however, a review of the diagnostic criteria for the disorders included in the current study suggests stability in these diagnoses across DSM revisions. Other societal changes have occurred since the col- lection of these data. For example, the advent of involuntary outpatient com- mitment, crisis intervention team (CIT) training, and mental health courts may have changed the experiences of individuals released from inpatient psychiatric hospitalizations. It is unclear what effects any of these innova- tions would have on the findings presented here, but current data could test whether the relationships found with the MacRisk data still apply in the present context.

In spite of these limitations, the current study is the first (to our knowledge) to study the recurring victimization experiences of individuals with diagnosed mental illnesses. Moreover, this is the first project to establish the variability

1002 Journal of Interpersonal Violence 29(6)

in recurring victimization trajectories by disorder type and the first to explore the correlates of these trajectories. We believe that these are important first steps and encourage the development of risk assessment devices to aid in the monitoring of recurring victimization risk that will assist clinicians in inter- vening to reduce recurring victimization burdens for people with mental dis- orders. We also encourage additional research on the correlates of recurring victimization among disordered populations and additional resources to help clinicians monitor and prevent risks of recurring victimization.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interests with respect to the authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research and/or authorship of this article.

Notes

1. As this study is the first of its kind to investigate recurring victimization by disorder type and there is little guidance in the extant literature on which to base specific hypotheses, we do not specify the direction of the trajectories by disor- der type.

2. Given the small sample size of individuals diagnosed with substance abuse dependence disorders who are also victims at Time 1 (n = 70), power to detect a significant effect is reduced (Cohen, 1988). Consequently, we report the differ- ence as “marginally significant.”

References

American Psychiatric Association. (1987). Diagnostic and statistical manual of men- tal disorders (3rd ed., rev.). Washington, DC: Author.

Barberet, B., Fisher, B. S., & Taylor, H. (2004). University student safety in the East Midlands. London, England: Home Office.

Brekke, J. S., Prindle, C., Bae, S. W., & Long, J. D. (2001). Risks for individuals with schizophrenia who are living in the community. Psychiatric Services, 52, 1358-1366.

Chuang, H. T., Williams, R., & Dalby, J. T. (1987). Criminal behavior among schizo- phrenics. Canadian Journal of Psychiatry, 32, 255-258.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Daigle, L. E., Fisher, B. S., & Cullen, F. T. (2008). The violent and sexual victimization of college women: Is repeat victimization a problem? Journal of Interpersonal Violence, 23, 1296-1313.

Teasdale et al. 1003

Farrell, G., Phillips, C., & Pease, K. (1995). Like taking candy: Why does repeat vic- timization occur? British Journal of Criminology, 35, 384-399.

Felson, R. B. (1992). “Kick’em when they’re down”: Explanations of the relation- ship between stress and interpersonal aggression and violence. Sociological Quarterly, 33, 1-16.

Finkelhor, D., Ormrod, R. K., & Turner, H. A. (2007). Poly-victimization: A neglected component in child victimization. Child Abuse & Neglect, 31, 7-26.

Finkelhor, D., Ormrod, R. K., & Turner, H. A. (2009). Lifetime assessment of poly- victimization in a national sample of children and youth. Child Abuse & Neglect, 33, 403-411.

Finkelhor, D., Ormrod, R. K., Turner, H. A., & Hamby, S. L. (2005). Measuring poly- victimization using the JVQ. Child Abuse & Neglect, 29, 1297-1312.

Fisher, B. S., Daigle, L. E., & Cullen, F. T. (2010). What distinguishes single sexual victims from recurring ones? The role of lifestyle-routine activities and first-inci- dent characteristics. Justice Quarterly, 27, 102-129.

Ford, J. D., Elhai, J. D., Connor, D. F., & Frueh, B. C. (2010). Poly-victimization and risk of posttraumatic, depressive, and substance use disorders and involvement in delinquency in a national sample of adolescents. Journal of Adolescent Health, 46, 545-552.

Gabor, T., & Mata, G. (2004). Victimization and repeat victimization over the life span: A predictive study and implications for policy. International Review of Victimology, 10, 193-221.

Gidycz, C. A., Hanson, K., & Layman, M. J. (1995). A prospective analysis of the relationships among sexual assault experiences: An extension of previous find- ings. Psychology of Women Quarterly, 19, 5-29.

Goodman, L. A., Salyers, M. P., Mueser, K. T., Rosenberg, S. D., Swartz, M., Essock, S. M., . . .Swanson, J. (2001). Recent victimization in women and men with severe mental illness: Prevalence and correlates. Journal of Traumatic Stress, 14, 615-632.

Hiday, V. A., Swanson, J. W., Swartz, M. S., Borum, R., & Wagner, H. R. (1999). Criminal victimization of persons with severe mental illness. Psychiatric Services, 50, 62-68.

Hiday, V. A., Swanson, J. W., Swartz, M. S., Borum, R., & Wagner, H. R. (2001). Victimization: A link between mental illness and violence? International Journal of Law and Psychiatry, 24, 559-572.

Hiday, V. A., Swanson, J. W., Swartz, M. S., Borum, R., & Wagner, H. R. (2002). Impact of outpatient commitment on victimization of people with severe mental illness. American Journal of Psychiatry, 159, 1403-1411.

Horney, J., Osgood, D., & Marshall, I. (1995). Criminal careers in the short-term: Intra-individual variability in crime and its relation to local life circumstances. American Sociological Review, 60, 655-673.

Lasley, J. R., & Rosenbaum, T. J. (1988). Routine activities and multiple personal victimization. Sociology and Social Research, 73, 47-50.

Lauritsen, J., & Davis Quinet, K. (1995). Repeat victimization among adolescents and young adults. Journal of Quantitative Criminology, 11, 143-166.

1004 Journal of Interpersonal Violence 29(6)

Lehman, A. F., & Linn, L. S. (1984). Crimes against discharged mental patients in board and care homes. American Journal of Psychiatry, 141, 271-274.

Maniglio, R. (2009). Severe mental illness and criminal victimization: A systematic review. Acta Psychiatrica Scandinavica, 119, 180-191.

Marley, J. A., & Buila, S. (2001). Crimes against people with mental illness: Types, perpetrators, and influencing factors. Social Work, 46, 115-124.

Mele, M. (2009). The time course of repeat intimate partner violence. Journal of Family Violence, 24, 619-624.

Monahan, J., Steadman, H. J., Silver, E., Appelbaum, P. S., Clark Robbins, P., Mulvey, E. P., . . .Banks, S. (2001). Rethinking risk assessment: The MacArthur study of mental disorder and violence. New York, NY: Oxford University Press.

Mukherjee, S., & Carcach, C. (1998). Repeat victimization in Australia (Australian Institute of Criminology Research and Public Policy Series No. 15). Griffith: Australian Institute of Criminology.

Osborn, D. R., & Tseloni, A. (1998). The distribution of household property crimes. Journal of Quantitative Criminology, 14, 307-330.

Outlaw, M., Ruback, B., & Britt, C. (2002). Repeat and multiple victimizations: The role of individual and contextual factors. Violence and Victims, 17, 187-204.

Pease, K. (1998). Repeat victimization: Taking stock (Crime Detection and Prevention Series Paper 90). London, England: Home Office.

Pokorny, A. D., Miller, B. A., & Kaplan, H. B. (1972). The brief MAST: A short- ened version of the Michigan Alcoholism Screening Test. Journal of American Psychiatry, 129, 342-345.

Polvi, N., Looman, T., Humphries, C., & Pease, K. (1990). Repeat break and enter victimisation: Time course and crime prevention opportunity. Journal of Police Science and Administration, 17, 8-11.

Polvi, N., Looman, T., Humphries, C., & Pease, K. (1991). The time-course of repeat burglary victimization. British Journal of Criminology, 31, 411-414.

Sagovsky, A., & Johnson, S. D. (2007). When does repeat victimisation occur? Australian & New Zealand Journal of Criminology, 40, 1-26.

Silver, E. (2002). Mental disorder and violent victimization: The mediating role of involvement in conflicted social relationships. Criminology, 40, 191-212.

Silver, E., Arseneault, L., Langley, J., Caspi, A., & Moffitt, T. E. (2005). Mental dis- order and violent victimization in a total birth cohort. American Journal of Public Health, 95, 2015-2021.

Silver, E., Mulvey, E. P., & Monahan, J. (1999). Assessing violence risk among dis- charged psychiatric patients: Toward an ecological approach. Law and Human Behavior, 23, 235-253.

Skinner, H. (1982). The Drug Abuse Screening Test. Addictive Behaviors, 7, 363-371. Steadman, H. J., Mulvey, E. P., Monahan, J., Clark Robbins, P., Appelbaum, P. S.,

Grisso, T., . . .Silver, E. (1998). Violence by people discharged from acute psy- chiatric inpatient facilities and by others in the same neighborhoods. Archives of General Psychiatry, 55, 393-401.

Teasdale, B. (2009). Mental disorder and violent victimization. Criminal Justice and Behavior, 36, 513-535.

Teasdale et al. 1005

Teasdale, B., Silver, E., & Monahan, J. (2006). Gender, threat/control-override delu- sions and violence. Law and Human Behavior, 30, 649-658.

Teplin, L. A., McClelland, G. M., Abram, K. M., & Weiner, D. A. (2005). Crime victimization in adults with severe mental illness: Comparison with the National Crime Victimization Survey. Archives of General Psychiatry, 62, 911-921.

Townsley, M., Homel, R., & Chaseling, J. (2000). Repeat burglary victimisation: Spatial and temporal patterns. Australia & New Zealand Journal of Criminology, 33, 37-63.

Tseloni, A. (2000). Personal criminal victimization in the United States: Fixed and ran- dom effects of individual and household characteristics. Journal of Quantitative Criminology, 16, 415-442.

Tseloni, A., & Pease, K. (1998). Nuisance phone calls to women in England and Wales. European Journal on Criminal Policy and Research, 6, 91-111.

Walsh, E., Moran, P., Scott, C., McKenzie, K., Burns, T., Creed, F., . . .Fahy, T. (2003). Prevalence of violent victimisation in severe mental illness. British Journal of Psychiatry, 183, 233-238.

Author Biographies

Brent Teasdale is a graduate of the Pennsylvania State University’s Crime, Law & Justice program, and an associate professor in the Department of Criminal Justice and Criminology at Georgia State University, where he is also the director of graduate studies and the editor of Social Problems Forum. He has published extensively on mental health issues and substance abuse prevention. His recent work has appeared in Social Problems, Prevention Science, Criminal Justice and Behavior, and the American Journal of Criminal Justice.

Leah E. Daigle is an associate professor of criminal justice at Georgia State University. Her most recent research has centered on repeat sexual victimization of college women and the responses that women use during and after being sexually victimized. Her other research interests include the development and continuation of offending over time and gender differences in the antecedents to and consequences of criminal victimization and participation across the life course. She is the coauthor of Unsafe in the Ivory Tower: The Sexual Victimization of College Women, Criminals in the Making: Criminality Across the Life-Course, and the author of Victimology: A Text/Reader. Her research has also appeared in peer-reviewed journals, including Justice Quarterly, Victims and Offenders, the Journal of Quantitative Criminology, and the Journal of Interpersonal Violence.

Ellen Ballard has an MS in criminal justice from Boston University. She is currently a doctoral student in the Department of Criminal Justice and Criminology at Georgia State University. She is interested in research on persons with serious mental illness in the criminal justice system, substance abuse, and stigmatized populations. Her pre- vious work has appeared in Substance Abuse and Rehabilitation.