The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth

Social Science & Medicine xxx (2015) 1e7

Contents lists available at ScienceDirect

Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed

The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth* Fredrick Butcher a, *, Joseph D. Galanek b, Jeff M. Kretschmar a, Daniel J. Flannery a a Begun Center for Violence Prevention Research & Education, Jack, Joseph and Morton Mandel School of Applied Social Sciences, Case Western Reserve University, Cleveland, OH, United States b SciMetrika, LLC, Atlanta, GA, United States

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 December 2014 Received in revised form 2 October 2015 Accepted 6 October 2015 Available online xxx

Previous research has demonstrated that exposure to violence (ETV) is a serious concern across the north-south socioeconomic divide. While studies have found that social support is a protective factor for youth exposed to violence and trauma, little is known about the impact of trauma symptoms on forming and maintaining social relationships which are key to accessing a vital social resource that fosters resilience in youth experiencing trauma symptomatology. Building on previous models that examine the impact of neighborhoods on exposure to violence and trauma, the current study examines the impact of neighborhood disorganization on ETV among youth and ETV's effects on trauma symptoms and social relationships. Data were collected on 2242 juvenile justice-involved youth with behavioral health issues in 11 urban and rural counties in the Midwestern United States. Using structural equation modeling (SEM), our data demonstrated that living in highly disorganized neighborhoods was associated with higher levels of ETV and that ETV was positively associated with trauma symptoms. Mediational analysis showed that trauma symptoms strongly mediated the effect of ETV on social relationships. Freely estimating structural paths by gender revealed that hypothesized associations between these variables were stronger for females than males. Findings here highlight the need to provide trauma-informed care to help youth to build and maintain social relationships. Identification and treatment of trauma symptoms that is culturally informed is a critical first step in ensuring that identified protective factors in local contexts, such as social relations and social support, have opportunities to minimize the impact of ETV among youth across northern and southern nations. © 2015 Elsevier Ltd. All rights reserved.

Keywords: United States Violence exposure Trauma Resilience Social relationships Structural equation modeling

Youth exposure to community violence is recognized as a significant global public health issue (Finkelhor et al., 2013; O'Donnell et al., 2011; Seedat et al., 2004; World Health Organization [WHO], 2008; Zinzow et al., 2009). For example, in a representative sample of 3164 U.S. youth, 37.8% of adolescents witnessed assaults with weapons; sexual assaults; robberies; threats with weapons; or physical assaults (Zinzow et al., 2009). Research on community

* This research was supported in part by grants from the Ohio Department of Youth Services and the Ohio Department of Mental Health & Addiction Services. * Corresponding author. Begun Center for Violence Prevention Research & Education, Jack, Joseph and Morton Mandel School of Applied Social Sciences, Case Western Reserve University, 11402 Bellflower Road, Cleveland, OH 44106-7167, United States. E-mail address: [email protected] (F. Butcher).

exposure to violence (ETV) has been primarily conducted in the US (O'Donnell et al., 2011). This community ETV, either through victimization or witnessing, is disproportionately distributed through populations within nations, based on socioeconomic and demographic factors (Buka et al., 2001; Haj-Yahia et al., 2011; Snyder and Sickmund, 2006; Wolf et al., 2014). Childhood ETV is linked with negative outcomes including trauma symptomatology (Buka et al., 2001; Finkelhor et al., 2007; Singer et al., 1995). These findings support the WHO's assertion that violence prevention is a global health priority (Krug et al., 2002). The current study applies a social determinants of health framework by examining the effect of neighborhood ETV on trauma at the individual level. Specifically, we propose a statistical model that depicts the impact of neighborhood disorganization on neighborhood ETV, trauma symptoms, and social relationships.

http://dx.doi.org/10.1016/j.socscimed.2015.10.013 0277-9536/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

2

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7

Poor urban minority U.S. males are disproportionately exposed to community violence compared to their female peers (Stein et al., 2003). This trend has also been identified, for example, in Arab youth in Israel; males were more likely to be exposed to witnessing assaults with weapons or witnessing beatings than females (HajYahia et al., 2011). While studies have generally found that males report higher ETV in the neighborhood (Ceballo et al., 2001; Singer et al., 1995), females who are exposed to violence exhibit higher levels of trauma symptoms as a result (Chen, 2010; Cooley-Quille et al., 2001). Similar to youth in poor urban areas, research on youth involved in the juvenile justice system in the US have found high levels of ETV and trauma symptomatology in comparison with the general population (Cauffman et al., 1998; Ford et al., 2008; Kretschmar et al., 2014). One study found that 67% of a sample of detained youth reported at least two and up to 16 separate incidents of traumatic events in their lifetime (Ford et al., 2008). As with community samples, research on juvenile justice youth have found strong associations between traumatic exposure and behavioral health issues such as depression, anxiety, suicidal ideation, and substance use (Kerig et al., 2009). Among juvenile justice-involved youth, females report particularly high levels of trauma symptoms (Cauffman et al., 1998; Kerig et al., 2009). An ecological model of violence prevention takes into account not only individual level characteristics (such as gender), but also social networks (of family and peers) and community contexts (such as poverty), to identify appropriate targets for intervention. For example, youths residing in poor urban communities across the north-south socioeconomic divide are at increased risk of ETV, both as witnesses and as victims (Buka et al., 2001; Rhodes et al., 2012; United Nations [UN], 2003). Social disorganization theory identifies that neighborhood factors such as resource deprivation, family disruption, and residential mobility contribute to diminished social controls and a community's inability to maintain common values through informal and formal social networks (Kawachi et al., 1999; Sampson and Groves, 1989). While research in the past has examined the effect of social disorganization on neighborhood level crime, recent research has examined the implications of social disorganization on individual level consequences including ETV (Chauhan et al., 2009; Gibson et al., 2009) and mental health issues such as depression and anxiety (Stockdale et al., 2007; Xue et al., 2005). This perspective parallels a social determinants of health framework in which epidemiological methods investigate the association between neighborhood characteristics, such as high levels of poverty and crime, and health (e.g. Kawachi et al., 1999). Recent research using structural equation modeling has found that the relationship between neighborhood disorganization and trauma is mediated by ETV and that the impact of ETV on trauma is further mediated by social support (Turner et al., 2013). Along these general lines, Sampson (2012) argued that community level interventions may mitigate the impact of these neighborhood effects on individuals by enhancing social relationships or peer mentorship for youth. However, these interventions do not take into account the potential for these identified neighborhood effects to increase risks to experience affective disturbances, such as trauma symptoms, which in turn affect the youth's ability to build and maintain social relationships. Social support, dependent on an individuals' social relationships within social networks, has been demonstrated in both US and international samples to be a protective factor for trauma symptomatology among youth exposed to violence (Berkman et al., 2000; Betancourt et al., 2012; Kennedy et al., 2010; Salami, 2010; Turner et al., 2013). However, social support and mental health symptoms may have a reciprocal relationship. Disclosure of trauma

symptoms, for example, may consequently affect social networks' responses to individuals' symptoms, or use of coping strategies. For example, avoidance can be a reaction to non-supportive familial or peer social networks (Guay et al., 2006). Trauma symptoms such as anger or dissociation, may impede possibilities of supportive relationships, which may further erode perceived or enacted social support (Guay et al., 2006). Research on US war veterans shows that PTSD symptoms moderate social support and diminish the protective factors of such support (Bertram and Dartt, 2009; Jakupcak et al., 2010). Youth in contexts of urban disorganization may also only have access to social networks whose own abilities to provide social support are likewise diminished due to persistent community stressors (Bertram and Dartt, 2009). Although social support has been examined as a key protective factor for trauma symptoms resulting from ETV, there has been no investigation into the complex relationships between neighborhoods, ETV, trauma, and social relationships among at-risk youth. 1. Current study Existing research has generally found that childhood ETV is associated with increased trauma symptomatology (Finkelhor et al., 2007; Singer et al., 1995). Further, recent research has extended the concept of neighborhood disorganization to test individual level outcomes such as ETV and associated trauma symptoms (Stockdale et al., 2007; Turner et al., 2013). Studies have also shown gender differences in the prevalence of neighborhood ETV and differential effects of ETV on trauma symptomatology (Chen, 2010). In our study we attempt to integrate the findings of social disorganization studies and the recent research on ETV and trauma to map the pathways from these neighborhood effects to individual trauma symptoms. We theorize that these symptoms may negatively impact the ability to initiate and maintain a significant protective factor, social relationships. While researchers have conceptualized social support as a protective factor that may mediate the role between ETV and trauma symptoms (Turner et al., 2013; Kennedy et al., 2010), there has been little research testing the role of trauma symptoms on social relationships. Trauma may in fact be a mediating variable that affects a youth's ability to build and maintain social relationships. Conceptualizing the model in this way may have a significant impact on treating youth exposed to violence. To address these limitations in the literature, we propose a theoretical model that tests the associations already found in the existing literature and extend the model by examining the impact of trauma symptoms on social relationships. We hypothesize that the direct effect of neighborhood disorganization on trauma is mediated by ETV. Additionally, we hypothesize that the effect of ETV on social relationships is mediated by trauma symptoms. In addition to testing the hypothesized model, we propose to test whether the hypothesized model is equivalent across gender. 2. Method 2.1. Sample This sample consists of juvenile justice-involved youth with behavioral health issues participating in the Behavioral Health Juvenile Justice (BHJJ) initiative, a U.S. community based program in 11 urban and rural counties in the state of Ohio, that diverts youth from the criminal justice system into mental health treatment programs (Kretschmar et al., 2014). Youth must have a history of juvenile justice involvement, at least one DSM-IV diagnosis, and be between the ages of 10 and 18. Additional optional eligibility criteria include: behavioral impairment, substance abuse diagnosis, criminal behavior, history of trauma or domestic violence, and a

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7

history of multi-system involvement. Community caseworkers collected de-identified data between 2006 and 2013 during an intake interview for all youth enrolled in the program. Initially, 2532 youth agreed to complete the intake interview with 2242 given all four measures used in the current study. The sample was comprised of 58.3% (n ¼ 1307) males and 41.7% (n ¼ 935) females with an average age of 15.60 years (SD ¼ 1.51). Almost all youth self-identified as Caucasian (53.2%, n ¼ 1192), African American (38.9%, n ¼ 874), or Multi-racial (5.9%, n ¼ 133). We therefore collapsed race into two categories: White and Nonwhite youth. Ethical approval for study protocols and research was obtained from the Institutional Review Board of Case Western Reserve University. 2.2. Measures 2.2.1. Neighborhood disorganization Neighborhood level information was collected for each youth using data from the 2010 US Census at the zip code level. To measure neighborhood disorganization, we used a similar approach to Tewksbury et al. (2010) and Shoff and Yang (2012) who used US Census data to construct a measure of neighborhood level poverty, family disruption, and residential mobility. The following observed measures made up the latent construct: percentage of female headed households, those above 25 years of age with less than a high school diploma or equivalency, renter occupied housing units, unemployed, and families receiving public assistance (a ¼ .81). While neighborhood disorganization was measured at the neighborhood level, small cluster sizes prohibited the modeling of neighborhood disorganization as a level-2 variable. 2.2.2. Recent exposure to violence The Recent Exposure to Violence (REVS) scale is a self-report questionnaire that measures children's ETV in several locations including the school, neighborhood, and at home (Singer et al., 1995). Data on the REVS have been reported in a number of community and at-risk samples and has demonstrated good psychometric properties (Butcher et al., 2014; Singer et al., 1995; Van Dulmen et al., 2008). For the current study, we examined six items that measured frequency of youth witnessing or being victimized by threats, assaults, and physical beatings in the neighborhood (a ¼ .69). Responses on the REVS ranged from 0 (never) to 5 (almost every day). 2.2.3. Trauma Symptom Checklist for Children e alternative The alternative version of the Trauma Symptom Checklist for Children (TSCC-A) is a 44 item self-report measure of trauma symptoms designed to measure psychological trauma in children between the ages of 8 and 16 years (Briere, 1996). Researchers have also administered the TSCC-A to youth 17 years of age (e.g. Butcher et al., 2014; Finkelhor et al., 2007; Kretschmar et al., 2014; Singer et al., 1995). Responses on the TSCC-A range between 0 (never) to 3 (almost all of the time) and comprise five subscales of trauma symptoms: anger, anxiety, depression, dissociation, and posttraumatic stress. Psychometric properties of the TSCC-A have been well established in both community and at-risk samples (Butcher et al., 2014; Lanktree et al., 2008). While the TSCC is comprised of five subscales, a total symptoms scale has demonstrated good psychometric properties (Butcher et al., 2015). In the interest of a parsimonious model, we utilized this total symptoms scale (a ¼ .90). 2.2.4. Ohio scales The Ohio Scales include measures of problem severity and everyday functioning to assess clinical outcomes for children

3

receiving treatment for severe emotional and behavioral disorders (Ogles et al., 2001). To measure positive social relationships, we included three items from the Ohio Scales that measures relationships with friends, family, and adults outside the family (a ¼ .56). Responses range between 0 (never) and 4 (doing very well). The measure been used previously in juvenile justice samples (Colwell et al., 2012). 2.3. Analysis plan The hypothesized model described above was tested using Structural Equation Modeling (SEM). We first examined a configural model tested on the entire sample. We then tested whether the model was invariant for both gender groups. All proposed models were estimated using the Maximum Likelihood method in Mplus n and Muthe n, 2011). Using an iterative process, version 7.11 (Muthe modification indices were examined to identify potential misspecification and used to respecify the model when theoretically and statistically appropriate. Several commonly used goodness of fit indices were considered when evaluating models. Indices reported in the following sections include a c2 test, comparative fit index (CFI), standardized root mean square residual (SRMR), and the root mean square error of approximation (RMSEA). A significant c2 value indicates that the hypothesized model adequately fitting the data was an unlikely event. Larger samples, however, tend to result in an inflated c2 statistic. We considered a value greater than .95 for the CFI, a SRMR value less than .08, and a RMSEA value less than .06 as indicative of an adequately fitting model (Hu and Bentler, 1999). A RMSEA value of up to .08 has been used as an indicator of adequate model fit (Wang et al., 2012). Missing data accounted for up to 7% of the variables in the analysis and were imputed using the EM algorithm. The EM algorithm is a particularly robust estimation method for missing data in large samples (Allison, 2001). To examine mediating effects, we estimated indirect effects using the bootstrapping method with 5000 samples (Bollen and Stine, 1990). We examined whether the hypothesized structural model was invariant across gender groups utilizing the same criteria to assess model fit. We considered whether the model was invariant across groups when factor loadings, intercepts, factor means, and structural paths were constrained across groups. 3. Results Descriptive statistics for each of the observed variables included in the proposed model and bivariate correlations of study variables are presented in the supplemental appendices (Appendix A and B). Prior to estimating the full structural model, we first estimated measurement models to test whether the factor structure of the measurement instruments included in the model fit the data. Measurement models indicated five residual covariances that were previously unspecified as a possible source for model misfit. The measurement model showed adequate fit to the data, c2(79) ¼ 808.15, CFI ¼ .96, SRMR ¼ .05, RMSEA ¼ .06, once these modifications were made. We used these modifications applied to the measurement model and added structural paths to estimate the hypothesized full structural model on the total sample (see Fig. 1). The effect of demographic variables (age and race) are accounted for in the structural model but not shown in Fig. 1. Fit indices suggested that the full structural model adequately fit the data, c2(104) ¼ 939.11, CFI ¼ .96, SRMR ¼ .05, RMSEA ¼ .06. Fig. 1 shows results of the structural model including standardized factor loadings and structural paths. Neighborhood disorganization was positively associated with ETV in the neighborhood, b ¼ .25, SE ¼ .03 but negatively associated with trauma symptoms, b ¼ .20, SE ¼ .02. Neighborhood ETV was positively

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

4

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7

Fig. 1. Factor Loadings and Coefficients for Structural Model. Note. Standardized coefficients are presented with dashed line indicating a non-significant path. Model c2(104) ¼ 939.11, CFI ¼ .96, SRMR ¼ .05, RMSEA ¼ .06. *p < .05. FHH ¼ Female headed household, Grad ¼ Percentage with less than a high school diploma, Ren ¼ Percentage of renter occupied housing units, Une ¼ Percentage Unemployed, Pub ¼ Percentage of families receiving public assistance, ANG ¼ Anger, ANX ¼ Anxiety, DEP ¼ Depression, DIS ¼ Dissociation, PTS ¼ Posttraumatic Stress.

associated with trauma symptoms, b ¼ .29, SE ¼ .03, while trauma symptoms were negatively associated with social relationships, b ¼ .52, SE ¼ .02. The path between trauma symptoms and social relationships was not statistically significant. Pathways shown in Fig. 1 conceptualize two mediators: the association between neighborhood disorganization and trauma symptoms mediated by ETV, as well as the mediating role of trauma symptoms on the association between ETV and social relationships. The indirect effect of neighborhood disorganization on trauma through ETV was significant b ¼ .07, SE ¼ .01, p < .001. However, given that all three direct paths were statistically significant, ETV only partially mediated the relationship between neighborhood disorganization and trauma symptoms. The indirect effect of ETV on social relationships was significant, b ¼ .15, SE ¼ .02, p < .001. Because the direct path from ETV to social relationships was not statistically significant, these data provide evidence that trauma is a strong mediator of the association between ETV and social relationships. 3.1. Group invariance Prior to assessing group invariance, measurement and structural models were each tested separately for males and females to identify any potential modifications that are group specific. We examined several models that tested different assumptions of invariance. While the data exhibited poor model fit when all paths

were constrained, a model that constrained factor loadings while freely estimating structural paths and intercepts across groups adequately fit the data, c2(225) ¼ 1265.35, CFI ¼ .95, SRMR ¼ .05, RMSEA ¼ .06 suggesting partial group invariance. Results of this model, including gender specific structural estimates are presented in Fig. 2. In comparison with males, all statistically significant associations were stronger for females. Similar to the overall sample, the indirect effect of neighborhood disorganization on trauma through ETV was significant for both males (b ¼ .06, SE ¼ .01, p < .001) and females (b ¼ .09, SE ¼ .02, p < .001) suggesting partial mediation. Indirect effects of ETV on social relationships through trauma were statistically significant for both males (b ¼ .15, SE ¼ .02, p < .001) and females (b ¼ .23, SE ¼ .03, p < .001). 4. Discussion This study examined relationships between neighborhood disorganization, exposure to violence, trauma symptoms, and social relationships. We proposed a SEM that conceptualized an association between neighborhood disorganization and trauma symptoms mediated by ETV and an association between ETV and social relationships mediated by trauma symptoms. Studies have conceptualized social relationships as being a protective factor for the impact of ETV on trauma (Turner et al., 2013). However, past research has shown that ETV can negatively affect a child's

Fig. 2. Structural Paths by Gender. Standardized coefficients presented with dashed line indicating a non-significant path. Model c2(225) ¼ 1265.35, CFI ¼ .95, SRMR ¼ .05, RMSEA ¼ .06. Underlined coefficients for females. *p < .05.

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7

prosocial skills (Holmes et al., 2014) and may, in turn, affect their ability to form social relationships. We therefore hypothesized that trauma symptoms may in fact mediate the relationship between neighborhood ETV and social relationships. Consistent with past research, results indicated that higher levels of neighborhood disorganization were associated with higher levels of neighborhood ETV (e.g. Chauhan et al., 2009; Gibson et al., 2009) and higher levels of ETV were associated with higher levels of trauma (e.g. Finkelhor et al., 2007; Singer et al., 1995). The data demonstrated that neighborhood disorganization was associated with trauma symptoms indirectly through neighborhood ETV. Neighborhood disorganization in conjunction with individual level exposure to violence in the neighborhood has a measurable effect on trauma symptoms. While the direct relationship between neighborhood disorganization and trauma symptoms was negative, the data demonstrated that youth are affected by neighborhood disorganization through elevated levels of ETV, similar to the model presented by Turner et al. (2013). In addition to this finding, the data showed that trauma symptoms are a strong mediator of the association between ETV and social relationships. In our sample of children exposed to violence, trauma symptoms are an important factor affecting their ability to form social relationships. Further, we found that while the factor loadings did not differ by gender, the best fitting model allowed all structural paths and intercepts to be freely estimated. While the significance of all structural paths were similar for both males and females, the direct effects of ETV on trauma and trauma on social relationships were stronger for females. Additionally, the indirect effect of ETV on social relationships through trauma symptoms was stronger for females. Past research has also found that the negative effects of ETV are stronger for females than males (Chen, 2010; Holmes et al., 2014). 4.1. Limitations Several limitations of this current study are of note. First, the sample described in the study is a sample of juvenile justice involved youth with behavioral health issues in the United States. Youth in this sample are identified by a juvenile court as having behavioral health issues and are at risk for further delinquent behavior. Results of this study may not be generalizable to community samples. However, the present sample is made up of a large sample of youth, with many living in high levels of poverty. Some of the neighborhood level social problems that these youth face may be similar for youth in low and middle income southern nations. Second, in estimating levels of neighborhood disorganization, the smallest available unit of analysis in the data was at the zip code level. Zip codes may cover areas and populations that are diverse and may be only a reasonable approximation of neighborhood characteristics (Stockdale et al., 2007). For example, zip codes do not take into account local constructs of community, which may be grounded in cultural identity (Engstrom et al., 2013). Third, the variables that formed the latent construct, social relationships had low internal consistency. While this may be due to social relationships consisting of few variables, the measurement model demonstrated good overall fit for the latent variables. In instances where items demonstrate weak internal consistency, the items may yet provide accurate estimates of the latent construct when there is sufficient variability and a confirmatory analysis is used (Little et al., 1999). Finally, with the exception of the neighborhood level variables, all other measures were youth reports. There are methodological issues including malingering and minimizing that affect studies assessing self-report data of sensitive issues such as trauma and ETV, particularly in settings where an interviewer is present

5

(Tourangeau and Yan, 2007). It is important to use caution when interpreting these results. 4.2. Policy implications Given these limitations, the current study demonstrated the associations between neighborhood disorganization, ETV, trauma symptoms, and social relationships in a US population of at-risk youth. Data presented here build on similar models in the literature by conceptualizing trauma as an impediment to initiating and maintaining social relationships. Researchers have generally found support for theoretical models conceptualizing the quality of social relationships and support as a protective factor between exposure to a traumatic event and trauma symptoms (e.g. Hobfoll et al., 2006; Turner et al., 2013). However, social relationships and trauma may have a reciprocal relationship where trauma affects a youth's ability to form relationships and may therefore diminish opportunities for social support. Conceptualizing the model in this way helps us to understand that first addressing trauma symptoms can provide opportunities to enhance and enable violence exposed youths' ability to access this critical protective factor. Current emphasis in the US on trauma-informed care is consistent with this perspective (Ko et al., 2008). Several theories inform our conceptualization and understanding of the impact of violence exposure and how to treat trauma affected youth. Social disorganization theory has evolved from focusing on the impact of neighborhood level stressors on crime rates (Sampson and Groves, 1989) to the impact of the same neighborhood level stressors to individual outcomes of violence exposure and associated mental health outcomes (Stockdale et al., 2007). This perspective can explain the mechanisms by which neighborhood contexts affect individual level mental health outcomes. The data presented here enhances the theoretical model that neighborhoods matter by demonstrating that one mechanism in which neighborhoods affect mental health outcomes is individual level ETV. Further, Hobfoll's Conservation of Resources theory informs the way that we view the treatment of youth affected by trauma. Hobfoll argues that limiting the loss of resources, such as supportive relationships, is a critical first step in creating a recovery environment for trauma treatment (Hobfoll, 1998). Once the resource loss has been limited, identifying and creating resources to facilitate individuals' resilience to stressors is critical. These resources can include both material resources, such as safe housing, as well as social resources such as social support (Hobfoll, 1989, 1998). Implementation of resources should focus on the multiple ecological domains in which youth encounter risks. These domains include: the individual, social, and community (Sameroff et al., 2003). Taking these perspectives into account, treatment of youth exposed to violence should focus on risk and resilience in multiple ecological domains. Pro-social interventions for youth such as mentoring can facilitate the creation of social resources at the individual level to build and enhance resilience to stressors. Several studies have demonstrated effectiveness in building social support for youth experiencing trauma symptoms (e.g. Britner et al., 2006; Munson and McMillen, 2008). Further, Sampson (2012) argues that interventions at the community level that focus on stabilizing neighborhoods and providing incentives for mixed-income housing can have a measurable effect on neighborhood level crime. Interventions at the neighborhood level can ultimately affect individual level outcomes by reducing the risk for ETV and associated trauma symptoms. While interventions at multiple levels can have a measurable effect on resource building, data presented here suggest that without addressing underlying symptoms of trauma, youth can have difficulty engaging in and maintaining the social

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

6

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7

relationships that are necessary to access a key resource: social support. These findings highlight the need to incorporate traumafocused screening and intervention into resilience building interventions. A starting point initiating trauma informed care is appropriate screening of youth that come to the attention of social welfare and care-giving systems. Trauma may be expressed differentially across cultural groups, however, which may make it challenging to identify these symptoms. For example, expressions of symptoms, or idioms of distress, vary due to culturally specific interpretations and expressions of suffering (Hollan, 2004). Modifications in trauma screening is feasible, based on identification of the cultural characteristics of expressions of distress that are applicable to local populations (Hinton et al., 2013). The TSCC has been validated as a reliable trauma screening instrument in Chinese youth samples, suggesting that cross-culturally, trauma symptoms may be identified through appropriate screening measures (Li et al., 2009). Accounting for local contexts' constraints and strengths is the next step in implementing trauma informed care (Feierman et al., 2010). Implementation efforts should necessarily take into account local socio-political realities and the dynamic interplay of gender, age, social role and status within these contexts, as well as local relationships of power (Lewis-Fernandez and Kleinman, 1995). For example, cultural acceptability of interventions may inhibit an intervention's implementation as significantly as lack of systemic capacity for systems of care (Patel et al., 2011). Building community capacity to collectively engage with social problems, such as youth exposure to violence and subsequent mental health symptoms, recognizes that local explanatory models of trauma have salience to treatment models, and these models can inform interventions that target structural determinants of violence exposure (Petersen et al., 2012). While data presented here are discussed in the context of its implications on a global approach to violence prevention, these data on at-risk youth in the US should be replicated in both northern and southern nations. We intend for the findings in the current study to advocate for trauma-informed interventions that leverage local resources for enhancing social resources. While the current study may not be generalizable across every sample of atrisk youth, particularly in trauma symptoms' measurement, our model can inform local treatment systems' conceptualization of interventions for youth affected by trauma. Future research should focus on the applicability of northern nations' trauma screening instruments to culturally distinct groups to identify youth who have been affected by ETV. Additionally, continued research on implementation and incorporation of trauma informed care within local treatment systems can provide guidance on how best to leverage capacity in low-resource communities in addressing the treatment needs of youth exposed to violence. Globally, research has documented the importance of social support as a protective factor in children exposed to terrorism in Israel (Hobfoll et al., 2006) and children exposed to political violence in Indonesia (Tol et al., 2010). Our study highlights the importance of traumainformed treatment approaches so that youth can enhance their resilience to trauma by leveraging the benefits of social relationships. Taking into account local contexts' cultures and infrastructure is a necessary and concurrent step in implementing such systems of care.

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.socscimed.2015.10.013.

References Allison, P.D., 2001. Missing data. Sage, Thousand Oaks, CA. Berkman, L.F., Glass, T., Brissette, I., Seeman, T.E., 2000. From social integration to health: Durkheim in the new millennium. Soc. Sci. Med. 51 (6), 843e857. http:// dx.doi.org/10.1016/S0277-9536(00)00065-4. Bertram, R.M., Dartt, J.L., 2009. Post traumatic stress disorder: a diagnosis for youth from violent, impoverished communities. J. Child Fam. Stud. 18 (3), 294e302. http://dx.doi.org/10.1007/s10826-008-9229-7. Betancourt, T.S., Salhi, C., Buka, S., Leaning, J., Dunn, G., Earls, F., 2012. Connectedness, social support and internalizing emotional and behavioural problems in adolescents displaced by the Chechen conflict. Disasters 36 (4), 635e655. http://dx.doi.org/10.1111/j.1467-7717.2012.01280.x. Bollen, K.A., Stine, R., 1990. Direct and indirect effects: classical and bootstrap estimates of variability. Sociol. Methodol. 20, 115e140. Briere, J., 1996. Trauma Symptoms Checklist for Children (TSCC): Professional Manual. Psychological Assessment Resources, Odessa, FL. Britner, P.A., Balcazar, F.E., Blechman, E.A., Blinn-Pike, L., Larose, S., 2006. Mentoring special youth populations. J. Community Psychol. 34 (6), 747e763. http:// dx.doi.org/10.1002/jcop.20127. Buka, S.L., Stichick, T.L., Birdthistle, I., Earls, F.J., 2001. Youth exposure to violence: prevalence, risks, and consequences. Am. J. Orthopsychiatry 71 (3), 298. http:// dx.doi.org/10.1037/0002-9432.71.3.298. Butcher, F., Kretschmar, J.M., Lin, Y., Flannery, D.J., Singer, M.I., 2014. Analysis of the validity scales in the Trauma Symptom Checklist for children. Res. Soc. Work Pract. 24 (6), 695e704. http://dx.doi.org/10.1177/1049731513516803. Butcher, F., Kretschmar, J.M., Singer, M.I., Flannery, D.J., 2015. Confirmatory factor analysis of the Trauma Symptom Checklist for children in an at-risk sample of youth. J. Soc. Soc. Work Res. 6 (2), 251e268. http://dx.doi.org/10.1086/681545. Cauffman, E., Feldman, S., Watherman, J., Steiner, H., 1998. Posttraumatic stress disorder among female juvenile offenders. J. Am. Acad. Child Adolesc. Psychiatry 37, 1209e1216. http://dx.doi.org/10.1097/00004583-199811000-00022. Ceballo, R., Dahl, T.A., Aretakis, M.T., Ramirez, C., 2001. Inner-city children's exposure to community violence: how much do parents know? J. Marriage Fam. 63 (4), 927e940. Chauhan, P., Reppucci, N.D., Turkheimer, E.N., 2009. Racial differences in the associations of neighborhood disadvantage, exposure to violence, and criminal recidivism among female juvenile offenders. Behav. Sci. Law 27 (4), 531e552. http://dx.doi.org/10.1002/bsl.868. Chen, W., 2010. Exposure to community violence and adolescents' internalizing behaviors among African American and Asian American adolescents. J. Youth Adolesc. 39 (4), 403e413. http://dx.doi.org/10.1007/s10964-009-9427-7. Colwell, B., Villareal, S.F., Espinosa, E.M., 2012. Preliminary outcomes of a preadjudication diversion initiative for juvenile justice involved youth with mental health needs in Texas. Crim. Justice Behav. 39 (4), 447e460. http:// dx.doi.org/10.1177/0093854811433571. Cooley-Quille, M., Boyd, R.C., Frantz, E., Walsh, J., 2001. Emotional and behavioral impact of exposure to community violence in inner-city adolescents. J. Clin. Child Adolesc. Psychol. 30 (2), 199e206. http://dx.doi.org/10.1207/ S15374424JCCP3002_7. Engstrom, R., Ofiesh, C., Rain, D., Jewell, H., Weeks, J., 2013. Defining neighborhood boundaries for urban health research in developing countries: a case study of Accra, Ghana. J. Maps 9 (1), 36e42. http://dx.doi.org/10.1080/ 17445647.2013.765366. Feierman, S., Kleinman, A., Stewart, K., Farmer, P., Das, V., 2010. Anthropology, knowledge-flows and global health. Glob. Public Health 5 (2), 122e128. Finkelhor, D., Ormrod, R.K., Turner, H.A., 2007. Poly-victimization: a neglected component in child victimization. Child Abuse Negl. 31, 7e26. http://dx.doi.org/ 10.1016/j.chiabu.2006.06.008. Finkelhor, D., Turner, H.A., Shattuck, A., Hamby, S.L., 2013. Violence, crime, and abuse exposure in a national sample of children and youth: an update. JAMA Pediatr. 167, 614e621. http://dx.doi.org/10.1001/jamapediatrics.2013.42. Ford, J.D., Hartman, J.K., Hawke, J., Chapman, J., 2008. Traumatic victimization, posttraumatic stress disorder, suicidal ideation, and substance abuse risk among juvenile justice-involved youths. J. Child Adolesc. Trauma 1, 75e92. http://dx.doi.org/10.1080/19361520801934456. Gibson, C.L., Morris, S.Z., Beaver, K.M., 2009. Secondary exposure to violence during childhood and adolescence: does neighborhood context matter? Justice Q. 26 (1), 30e57. http://dx.doi.org/10.1080/07418820802119968. Guay, S., Billette, V., Marchand, A., 2006. Exploring the links between posttraumatic stress disorder and social support: processes and potential research avenues. J. Trauma. Stress 19 (3), 327e338. http://dx.doi.org/10.1002/jts.20124. Haj-Yahia, M.M., Leshem, B., Guterman, N., 2011. Exposure to community violence among Arab youth Israel: rates and characteristics. J. Community Psychol. 39 (2), 136e151. http://dx.doi.org/10.1002/jcop.20423. Hinton, D.E., Kredlow, M.A., Pich, V., Bui, E., Hofmann, S.G., 2013. The relationship of PTSD to key somatic complaints and cultural syndromes among Cambodian refugees attending a psychiatric clinic: the Cambodian Somatic Symptom and Syndrome Inventory (CSSI). Transcult. Psychiatry 50 (3), 347e370, 1363461513481187. Hobfoll, S.E., 1989. Conservation of resources: a new attempt at conceptualizing stress. Am. Psychol. 44 (3), 513e524. http://dx.doi.org/10.1037/0003066x.44.3.513. Hobfoll, S.E., 1998. Stress, Culture, and Community: The Psychology and Philosophy

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013

F. Butcher et al. / Social Science & Medicine xxx (2015) 1e7 of Stress. Plenum Press, New York. Hobfoll, S.E., Canetti-Nisim, D., Johnson, R.J., 2006. Exposure to terrorism, stressrelated mental health symptoms, and defensive coping among Jews and Arabs in Israel. J. Consult. Clin. Psychol. 74 (2), 207e218. http://dx.doi.org/ 10.1037/0022-006X.74.2.207. Holmes, M.R., Voith, L.A., Gromske, A.N., 2014. Lasting effect of intimate partner violence exposure during preschool on aggressive behavior and prosocial skills. J. Interpers. Violence. http://dx.doi.org/10.1177/0886260514552441. Advance online publication. Hu, L., Bentler, P.M., 1999. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct. Equ. Model. Multidiscip. J. 6, 1e55. Hollan, D., 2004. Self systems, cultural idioms of distress, and the psycho-bodily consequences of childhood suffering. Transcult. Psychiatry 41 (1), 62e79. http://dx.doi.org/10.1177/1363461504041354. Jakupcak, M., Vannoy, S., Imel, Z., Cook, J.W., Fontana, A., Rosenheck, R., McFall, M., 2010. Does PTSD moderate the relationship between social support and suicide risk in Iraq and Afghanistan War Veterans seeking mental health treatment? Depress. Anxiety 27 (11), 1001e1005. http://dx.doi.org/10.1002/da.20722. Kawachi, I., Kennedy, B.P., Wilkinson, R.G., 1999. Crime: social disorganization and relative deprivation. Soc. Sci. Med. 48 (6), 719e731. http://dx.doi.org/10.1016/ S0277-9536(98)00400-6. Kennedy, A.C., Bybee, D., Sullivan, C.M., Greeson, M., 2010. The impact of family and community violence on children's depression trajectories: examining the interactions of violence exposure, family social support, and gender. J. Fam. Psychol. 24 (2), 197e207. http://dx.doi.org/10.1037/a0018787. Kerig, P.K., Ward, R.M., Vanderzee, K.L., Moeddel, M.A., 2009. Posttraumatic stress as a mediator of the relationship between trauma and mental health problems among juvenile delinquents. J. Youth Adolesc. 38 (9), 1214e1225. http:// dx.doi.org/10.1007/s10964-008-9332-5. Ko, S.J., Ford, J.D., Kassam-Adams, N., Berkowitz, S.J., Wilson, C., Wong, M., Layne, C.M., 2008. Creating trauma-informed systems: child welfare, education, first responders, health care, juvenile justice. Prof. Psychol. Res. Pract. 39, 396e404. http://dx.doi.org/10.1037/0735-7028.39.4.396. Kretschmar, J.M., Butcher, F., Flannery, D.J., Singer, M.I., 2014. Diverting juvenile justice-involved youth with behavioral health issues from detention: preliminary findings from Ohio's Behavioral Health Juvenile Justice (BHJJ) initiative. Crim. Justice Policy Rev. http://dx.doi.org/10.1177/0887403414560885. Advance online publication. Krug, E.G., Mercy, J.A., Dahlberg, L.L., Zwi, A.B., 2002. The world report on violence and health. The Lancet 360 (9339), 1083e1088. http://dx.doi.org/10.1016/ S0140-6736(02)11133-0. Lanktree, C.B., Gilbert, A.M., Briere, J., Taylor, N., Chen, K., Maida, C.A., Saltzman, W.R., 2008. Multi-informant assessment of maltreated children: convergent and discriminant validity of the TSCC and TSCYC. Child Abuse Negl. 32, 621e625. http://dx.doi.org/10.1016/j.chiabu.2007.10.003. Lewis-Fernandez, R., Kleinman, A., 1995. Cultural psychiatry: theoretical, clinical, and research issues. Psychiatr. Clin. N. Am. 18 (3), 433e448. Li, X., Fang, X., Stanton, B., Zhao, G., Lin, X., Zhao, J., Chen, X., 2009. Psychometric evaluation of the Trauma Symptoms Checklist for Children (TSCC) among children affected by HIV/AIDS in China. AIDS Care 21 (3), 261e270. http:// dx.doi.org/10.1080/09540120802195119. Little, T.D., Lindenberger, U., Nesselroade, J.R., 1999. On selecting indicators for multivariate measurement and modeling with latent variables: when “good” indicators are bad and “bad” indicators are good. Psychol. Methods 4 (2), 192e211. http://dx.doi.org/10.1037/1082-989.X4.2.192. Munson, M., McMillen, J.C., 2008. Natural mentoring and psychosocial outcomes among older youth transitioning from foster care. Child. Youth Serv. 31 (1), 104e111. http://dx.doi.org/10.1016/j.childyouth.2008.06.003. n, L.K., Muthe n, B.O., 2011. Mplus user's guide, sixth ed. Muthe n & Muthe n, Muthe Los Angeles, CA. O'Donnell, D.A., Roberts, W.C., Schwab-Stone, M.E., 2011. Community violence exposure and post-traumatic stress reactions among Gambian youth: the moderating role of positive school climate. Soc. Psychiatry Psychiatr. Epidemiol. 46 (1), 59e67. http://dx.doi.org/10.1007/s00127-009-0162-x. Ogles, B.M., Melendez, G., Davis, D.C., Lunnen, K.M., 2001. The Ohio scales: practical outcome assessment. J. Child Fam. Stud. 10, 199e212. http://dx.doi.org/10.1023/ A:1016651508801. Patel, V., Chowdhary, N., Rahman, A., Verdeli, H., 2011. Improving access to psychological treatments: lessons from developing countries. Behav. Res. Ther. 49 (9), 523e528. Petersen, I., Baillie, K., Bhana, A., 2012. Understanding the benefits and challenges of community engagement in the development of community mental health services for common mental disorders: lessons from a case study in a rural South African subdistrict site. Transcult. Psychiatry 49 (3e4), 418e437. http://

7

dx.doi.org/10.1177/1363461512448375. Rhodes, T., Wagner, K., Strathdee, S., Shannon, K., Davidson, P., Bourgois, P., 2012. Structural violence and structural vulnerability within the risk environment: theoretical and methodological perspectives for a social epidemiology of HIV risk among injection drug users and sex workers. In: O'Campo, P., Dunn, J.R. (Eds.), Rethinking Social Epidemiology. Springer, Netherlands, pp. 205e230. http://dx.doi.org/10.1007/978-94-007-2138-10. Salami, S.O., 2010. Moderating effects of resilience, self-esteem and social support on adolescents' reactions to violence. Asian Soc. Sci. 6 (12), 101e110. Sameroff, A., Gutman, L.M., Peck, S.C., 2003. Adaptation among youth facing multiple risks: prospective research findings. In: Luthar, S.S. (Ed.), Resilience and Vulnerability: Adaptation in the Context of Childhood Adversities. Cambridge University Press, New York, pp. 364e391. Sampson, R.J., 2012. Neighborhood inequality, violence, and the social infrastructure of the American city. In: Tate, W.F. (Ed.), Research on Schools, Neighborhoods, and Communities: Toward Civic Responsibility. Rowman & Littlefield Publishers, New York NY, pp. 11e28. Sampson, R.J., Groves, W.B., 1989. Community structure and crime: testing socialdisorganization theory. Am. J. Sociol. 94 (4), 774e802. Seedat, S., Nyamai, C., Njenga, F., Vythilingum, B., Stein, D.J., 2004. Trauma exposure and post-traumatic stress symptoms in urban African schools: survey in CapeTown and Nairobi. Br. J. Psychiatry 184, 169e175. http://dx.doi.org/10.1192/ bjp.184.2.169. Shoff, C., Yang, T., 2012. Untangling the associations among distrust, race, and neighborhood social environment: a social disorganization perspective. Soc. Sci. Med. 74 (9), 1342e1352. http://dx.doi.org/10.1016/j.socscimed.2012.01.012. Singer, M.I., Anglin, T.M., Song, L.Y., Lunghofer, L., 1995. Adolescents' exposure to violence and associated symptoms of psychological trauma. JAMA J. Am. Med. Assoc. 273, 477e482. http://dx.doi.org/10.1001/jama.1995.03520300051036. Snyder, H.N., Sickmund, M., 2006. Juvenile Offenders and Victims: 2006 National Report. US Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention, Washington, DC. Stein, B.D., Jaycox, L.H., Kataoka, S., Rhodes, H.J., Vestal, K.D., 2003. Prevalence of child and adolescent exposure to community violence. Clin. Child Fam. Psychol. Rev. 6 (4), 247e264. http://dx.doi.org/10.1023/B: CCFP.0000006292.61072.d2. Stockdale, S.E., Wells, K.B., Tang, L., Belin, T.R., Zhang, L., Sherbourne, C.D., 2007. The importance of social context: neighborhood stressors, stress-buffering mechanisms, and alcohol, drug, and mental health disorders. Soc. Sci. Med. 65 (9), 1867e1881. http://dx.doi.org/10.1016/j.socscimed.2007.05.045. Tewksbury, R., Mustaine, E.E., Covington, M., 2010. Offender presence, available victims, social disorganization and sex offense rates. Am. J. Crim. Justice 35 (1e2), 1e14. http://dx.doi.org/10.1007/s12103-010-9070-6. Tol, W.A., Komproe, I.H., Jordans, M.J.D., Gross, A.L., Susanty, D., Macy, R.D., de Jong, J.T.V.M., 2010. Mediators and moderators of a psychosocial intervention for children affected by political violence. J. Consult. Clin. Psychol. 78 (6), 818e828. http://dx.doi.org/10.1037/a0021348. Tourangeau, R., Yan, T., 2007. Sensitive questions in surveys. Psychol. Bull. 133, 859e883. http://dx.doi.org/10.1037/0033-2909.133.5.859. Turner, H.A., Shattuck, A., Hamby, S., Finkelhor, D., 2013. Community disorder, victimization exposure, and mental health in a national sample of youth. J. Health Soc. Behav. 54 (2), 258e275. United Nations, 2003. World Youth Report. Retrieved from: http://www.un.org/esa/ socdev/unyin/documents/worldyouthreport.pdf. Van Dulmen, M.,H.,M., Belliston, L.,M., Flannery, D.J., Singer, M., 2008. Confirmatory factor analysis of the recent exposure to violence scale. Child. Sch. 30, 93e102. http://dx.doi.org/10.1093/cs/30.2.93. Wang, B., Li, X., Barnett, D., Zhao, G., Zhao, J., Stanton, B., 2012. Risk and protective factors for depression symptoms among children affected by HIV/AIDS in rural China: a structural equation modeling analysis. Soc. Sci. Med. 74, 1435e1443. http://dx.doi.org/10.1016/j.socscimed.2012.01.007. Wolf, A., Gray, R., Fazel, S., 2014. Violence as a public health problem: an ecological study of 169 counties. Soc. Sci. Med. 104, 220e227. http://dx.doi.org/10.1016/ j.socscimed.2013.12.006. World Health Organization, 2008. Preventing Violence and Reducing its Impact: How Development Agencies Can Help. Retrieved from: http://whqlibdoc.who. int/publications/2008/9789241596589_eng.pdf. Xue, Y., Leventhal, T., Brooks-Gunn, J., Earls, F.J., 2005. Neighborhood residence and mental health problems of 5- to 11-year-olds. Arch. General Psychiatry 62 (5), 554e563. http://dx.doi.org/10.1001/archpsyc.62.5.554. Zinzow, H.M., Ruggiero, K.J., Resnick, H., Hanson, R., Smith, D., Saunders, B., Kilpatrick, D., 2009. Prevalence and mental health correlates of witnessed parental and community violence in a national sample of adolescents. J. Child Psychol. Psychiatry 50 (4), 441e450. http://dx.doi.org/10.1111/jcpp.2009.50.issue-4.

Please cite this article in press as: Butcher, F., et al., The impact of neighborhood disorganization on neighborhood exposure to violence, trauma symptoms, and social relationships among at-risk youth, Social Science & Medicine (2015), http://dx.doi.org/10.1016/j.socscimed.2015.10.013