Received: 29 June 2016 Revised: 7 January 2017 Accepted: 14 January 2017 DOI: 10.1002/jcop.21885 ARTICLE Impact of victimization, community climate, and community size on the mental health of sexual and gender minority youth Megan S. Paceley1 Jacob Goffnett2 1 University of Kansas School of Social Welfare Megan Gandy-Guedes3 Abstract 2 University of Illinois at Urbana-Champaign Sexual and gender minority (SGM) youth are at risk of stigma and School of Social Work 3 University of Oklahoma Anne and Henry Zarrow victimization, which may lead to increased mental health symptoms, School of Social Work including depression, anxiety, and stress. The role of the commu- Correspondence Email: mpaceley@ku.edu nity in exacerbating or mitigating the frequency of victimization and This study was funded by The Williams Institute and the University of Illinois at Urbana-Champaign Graduate College. Using a minority stress framework, this study examined the associa- mental health symptoms among SGM youth has been understudied. tion between physical and nonphysical anti-SGM victimization, per- The first author thanks her dissertation committee for their assistance with this study: Janet Liechty, Ramona F. Oswald, Benjamin Lough, Jennifer Greene, and Shelley Craig. Thank you also to Amanda Hwu and Hortencia Arizpe for their assistance in data collection and analysis. ceived community climate, and community size and their effects on depression, anxiety, and stress among SGM youth. An online survey was completed by 296 SGM-identified youth (aged 14–18 years). Results revealed that nonphysical forms of victimization were associated with depression, anxiety, and stress above and beyond physical victimization and community variables. Both types of victimization were more common in communities that participants perceived to be hostile and small towns. Practice implications include ways to reduce victimization and improve mental health via community climate and community interventions. 1 INTRODUCTION Sexual and gender minority (SGM) youth grow up in a society that marginalizes and stigmatizes their SGM identities (Birkett, Russell, & Corliss, 2014). For some SGM youth, this stigma is reproduced within the communities in which they live in the form of anti-SGM victimization. Victimization may account for the disparate mental health problems experienced by SGM youth compared to heterosexual and cisgender youth (whose gender identity and assigned sex are congruent; Meyer, 2003), including depression and suicide (Burton, Marshal, Chisolm, Sucato, & Friedman, 2013; Meyer, 2003), anxiety (Remafedi, 2007; Rosario, Schrimshaw, & Hunter, 2006), and stress (Mustanski, Garofalo, & Emerson, 2010; Roberts, Rosario, Corliss, Koenen, & Austin, 2012). Finding ways to mitigate the effect of victimization on mental health is critical to improving the well-being of SGM youth. Some researchers have suggested an association between mental health and the social climate of schools (Birkett, Espelage, & Koenig, 2009) and communities (Hatzenbuehler, 2011); however, research has yet to examine the J. Community Psychol. 2017; 00: 1–14 wileyonlineliberary.com/journal/jcop c 2017 Wiley Periodicals, Inc.  1 PACELEY ET AL . 2 role of perceived community climate in exacerbating or mitigating the frequency of different types of victimization and associated mental health outcomes among SGM youth. In addition, the size of one’s community may affect SGM youth’s experiences of victimization and mental health. This study aimed to account for this gap in the research by exploring the association between two types of anti-SGM victimization (physical and nonphysical), perceived community climate, and community size on mental health symptoms among SGM youth. 1.1 Anti-SGM victimization Broadly, anti-SGM victimization can include a variety of physical and nonphysical acts (Oswald & Culton, 2003). Physical forms of victimization may include experiencing loss of or harm to personal property and physical and sexual assaults; nonphysical forms include overhearing stigmatizing anti-SGM rhetoric, verbal insults and threats, and anti-SGM discrimination. Research suggests SGM youth endure greater amounts of both forms of victimization in their schools (Poteat, Aragon, Espelage, & Koenig, 2009; Robinson & Espelage, 2011) and homes (Friedman, Marshal, Guadamuz, Wei, Wong, Saewyc, & Stall, 2011) compared to heterosexual and cisgender youth. Bullying has received heightened attention in recent decades, illuminating the prevalence of anti-SGM victimization in schools. In their biennial National School Climate Survey, the Gay, Lesbian and Straight Education Network sampled 7,800 SGM middle and high school students; 85% reported being verbally harassed within the past year and 65% reported frequently overhearing homophobic remarks (Kosciw, Greytak, Palmer, & Boesen, 2014). This is consistent with other empirical work in which SGM students reported overhearing homophobic comments from students (Kosciw & Diaz, 2006; Kosciw, Diaz, & Greytak, 2008) and experiencing verbal abuse, threats of violence, and physical assault (D’Augelli, Grossman, & Starks, 2006; Friedman, Koeske, Silvestre, Korr, & Sites, 2006). SGM youth have also reported that school staff, teachers, and administrators have overlooked harassment from students and participated in harassment themselves (Birkett et al., 2009; Kosciw et al., 2014). While a majority of anti-SGM victimization research has focused on experiences within schools, experiences of victimization are not confined to any one context; SGM youth may encounter similar forms of victimization at home within their families. SGM youth’s disclosure of their sexual or gender identity may elicit rejection from family. Ryan, Huebner, Diaz, and Sanchez (2009) retrospectively measured family rejection among sexual minorities and found a mean of 20.91 negative family reactions (range = 0-51) toward their adolescents’ sexual minority status. Further, they found negative family reactions in adolescence were associated with negative health problems in adulthood. For gender minority youth, parents may discourage gender nonconforming behaviors through punishment and reparative counseling (D’Augelli et al., 2006). 1.2 SGM youth and mental health Anti-SGM victimization contributes to disproportionately higher rates of mental health symptoms among SGM youth compared to heterosexual and cisgender youth (Burton, et al., 2013; Meyer, 2003; Robinson, Espelage, & Rivers, 2013; Russell, Ryan, Toomey, Diaz, & Sanchez, 2011). Mental health indicators associated with anti-SGM victimization include depression and suicide (Birkett et al., 2009; Burton et al., 2013), anxiety (Remafedi, 2007; Rosario et al., 2006), and stress (Mustanski et al., 2010; Roberts et al., 2012). In a meta-analysis of 24 studies, Marshal et al. (2011) found that sexual minority youth have higher rates of depression and suicidality compared to heterosexual youth. Although Marshal et al.’s (2011) analysis did not incorporate predictors of depression and suicidality, they postulated that victimization is likely an antecedent. Toomey, Ryan, Diaz, Card, and Russel (2010) found SGM-specific victimization fully mediated the relationship between gender nonconformity and depression in young adulthood. Suicidality is not contained to adolescence. Over their lifetimes, gay and bisexual men are four times more likely to attempt suicide than heterosexual men; lesbian and bisexual women are nearly twice as likely as heterosexual women to attempt suicide (King et al., 2008). Disparities in the lifetime prevalence of anxiety for sexual minorities have also been documented. Gay men are over two times more likely than heterosexual men and lesbian women are 1.6 times more likely than heterosexual women to experience anxiety (Meyer, 2003). It is likely PACELEY ET AL . 3 that these disparities exist because of the stigma and victimization experienced by SGM individuals. For example, Woodford, Paceley, Kulick, and Hong (2015) found a positive association between hearing anti-SGM messages, which pervade society, and anxiety symptoms among sexual minority young adults. Not only have mental health conditions been associated with current anti-SGM victimization, but also research with retrospective samples has shown an association between present mental health conditions and past victimization, particularly with posttraumatic stress symptoms (D’Augelli et al., 2006). 1.3 Minority stress theory The prevalence of mental health conditions among SGM people was once thought to be inherent with an SGM identity and consequently the SGM identity was pathologized (Meyer, 2003). The pathologizing of SGM identities further promoted stigma within American society; however, research has helped refute the idea that identifying as SGM is a condition that requires treatment and researchers have since sought to explain the disparities in mental health conditions. Minority stress theory (MST) was formulated through a meta-analysis of research studies connecting mental health outcomes to sexual minority-based victimization. Meyer (2003) posited that sexual minorities face additional stressors in society because of their stigmatized identities; these stressors account for disproportionate mental health problems. In 2012, Hendricks and Testa proposed ways to expand MST to include stressors relevant to gender minorities (e.g., high rates of sexual and physical violence); their adaptation suggests that MST is an applicable model for understanding stressors and subsequent adverse outcomes present among gender minorities. MST renders a spectrum of stressors ranging from distal to proximal. Distal stressors are those found in society and can include nonphysical acts, such as institutional discrimination and overhearing inflammatory rhetoric, or physical acts, such as physical and sexual assault. Proximal stressors are those situated within the individual, such as internalized homophobia/transphobia, concealment of one’s SGM identity, and anticipation of rejection. Burton et al. (2013) used a 6-month longitudinal study to test the validity of MST in regards to depression and suicide. Among their sample of 199 adolescents, 55 identified as sexual minorities, and they found higher levels of sexual minority-specific victimization, depressive symptoms, and suicidality among this group compared to their heterosexual sample. In 2015, Mereish and Poteat tested mediational pathways among MST and found that proximal stressors mediate the relationship between distal stressors and adverse health outcomes. Although Mereish and Poteat’s (2015) research is limited by its crosssectional nature, it helps us to understand how stigma moves from distal places outside the individual through the individual’s proximal appraisal systems, resulting in adverse mental health. MST acknowledges that not all minority stressors result in adverse outcomes as SGM individuals often build resilience through developing a strong SGM identity and connecting to the larger SGM community (Meyer, 2003). MST emphasizes that minority stress is “socially based … it stems from social processes, institutions, and structures beyond the individual” (Meyer, 2003, p. 676). These structures may help create a climate that is nurturing or neglectful of SGM identities and are found in distal positions within an individual’s ecology, such as the community in which SGM youth are embedded. 1.4 Community Two variables that may affect SGM youth’s experiences with distal stressors, and subsequent mental health, are community climate and community size. For example, climate and size may determine the amount and type of SGM-specific resources available; research has shown that involvement in local SGM resources can buffer the effects of anti-SGM victimization (LeBeau & Jellison, 2009; Ramirez-Valles, Fergus, Reisen, Poppen, & Zea, 2005). 1.4.1 Community climate Community climate is “the level of community support” (Oswald, Cuthbertson, Lazarevic, & Goldberg, 2010, p. 215) for SGM individuals. Community climate can be measured objectively by evaluating structural components (e.g., legal rights, workplace and school policies) that support SGM individuals within a community, state, or region, or PACELEY ET AL . 4 subjectively by asking SGM individuals to rate their perceptions of community supportiveness (perceived community climate). Oswald et al. (2010) have shown objective and perceived measures of community climate to be positively correlated with each other; they postulated that when objective factors within a community exist (such as the presence of other SGM people, SGM-affirming religious institutions, and political support for SGM people), they create a supportive environment that facilitates connections among SGM members of a community. These connections produce a social web of affirming individuals able to buffer the effects of minority stress. Research on school climate supports this proposition–a positive school climate is associated with greater well-being among SGM youth (Birkett et al., 2009; Espelage, Aragon, Birkett, & Koenig, 2008). Research on community climate among sexual minority youth suggests that an objective climate supportive of SGM individuals is associated with fewer symptoms of alcohol abuse and fewer sexual partners (Olson, Cadge, & Harrison, 2006), while living in a nonsupportive climate places sexual minority youth at greater risk of attempting suicide (Hatzenbuehler, 2011). Additionally, SGM individuals who perceive their climate as hostile are at an increased risk of anxiety and stress (Woodford et al., 2015). Although these studies have found an association between climate and well-being, they have not examined the role of perceived climate alongside different types of victimization in the mental health of SGM youth. 1.4.2 Community size The size of SGM youth’s community may also affect their experiences. Colloquially, rural communities have been depicted as hostile toward SGM people (Gray, 2007; Kazyak, 2011; Wienke & Hill, 2013). Swank, Fahs, and Frost (2013) found sexual minorities living in rural areas were exposed to more stigma and discrimination compared to those living in urban areas. Research on rural SGM youth suggests rural communities may place youth at a greater risk of victimization and poor mental health outcomes compared to urban SGM youth: Rural SGM youth overhear more homophobic language in school (Palmer, Kosciw, & Bartkiewicz, 2012) and report more suicidal thoughts (Poon & Saewyc, 2009). Additionally, rural SGM youth report greater affective distress than rural heterosexual and cisgender youth (Cohn & Leake, 2012). Research also suggests that support for pro-sexual minority policies (e.g., same-sex adoptions) differs between urban and rural communities, with people in urban communities expressing more support (Strange & Kazyak, 2015). Much less research has compared the experiences of gender minorities in rural areas. Some research on SGM adults has contested the conception that rural communities are inherently hostile. Factors associated with small towns, such as traditional small town values and deeper connections to other SGM individuals because of population size, may help nurture positive well-being for rural SGM individuals (Kazyak, 2011; Oswald & Culton, 2003). Indeed, some research suggests that rural sexual minorities report significantly more happiness and greater health than urban sexual minorities (Wienke & Hill, 2013), although this study did not specifically measure community climate. Gray’s (2007) ethnographic work on SGM youth in rural Kentucky depicted youth as developing their sexual and gender identities in different but not inferior ways to SGM youth in urban areas. For example, youth created their own queer resources in established public spaces and used the Internet to cultivate their identity and bring queer culture to their local communities. These findings suggest that the size and climate of SGM youth’s communities may be a complex factor in promoting or inhibiting stigma or well-being. There is a need to better understand the mechanisms that affect mental health symptoms among SGM youth. Understanding the ways in which different types of distal stressors, such as physical and nonphysical victimization, are associated with mental health conditions under varying community contexts is critical to expanding MST and developing interventions to promote the well-being of SGM youth. Given this research and current gaps in the literature, this study examined the association between physical and nonphysical anti-SGM victimization, perceived community climate, and community size on depression, anxiety, and stress among SGM youth. We hypothesized that increased levels of physical and nonphysical victimization, smaller community size, and greater perceptions of a hostile community climate would be associated with increased depression, anxiety, and stress among SGM youth while adjusting for other predictors and controls. Additionally, we hypothesized that higher levels of physical and nonphysical victimization would occur in small towns and hostile communities. PACELEY ET AL . 2 5 METHOD An online survey was developed to measure victimization, mental health symptoms, perceived community climate, and community size among SGM youth. The survey was pilot tested with a subgroup of adolescents to ensure comprehension of questions. Data were collected between January 2014 and June 2014. 2.1 Sampling and participants Youth between 14 and 18 years of age living in one Midwestern state who identified as sexual or gender minorities or as questioning their sexual orientation or gender identity were eligible to participate. Participants were recruited via fliers posted in SGM and non-SGM spaces and organizations (e.g., SGM community centers, libraries, coffee shops), social worker or teacher referrals, and online advertisements on social media (Facebook, Twitter, Tumblr). The final analytic sample included 296 SGM youth. See Table 1 for participant demographics. 2.2 Procedures Participants completed the survey online via Qualtrics after reading an informed consent/assent form and answering questions about their eligibility on the first page of the survey. Because of the inherent risk involved in asking SGM youth to disclose their identity, the university ethics board granted a waiver of parental consent to obtain parental permission (Taylor, 2008). The survey took 20–40 minutes to complete and did not ask for identifying information until the survey was completed and submitted. At that time, participants could elect to enter a drawing to win 1 of 10 $20 gift cards. The university’s institutional review board approved all procedures. 2.3 Measures The survey used existing measures to assess participants’ demographics, experiences with anti-SGM victimization, mental health symptoms, and communities. 2.3.1 Demographics Participants were asked to indicate their current age, race/ethnicity (African-American/Black, Asian/Pacific Islander, Hispanic/Latino/a, Native American, White, multiracial, other-specify), gender identity (male, female, transgender, genderqueer, questioning, other-specify), and sexual identity (bisexual, gay, lesbian, pansexual, queer, questioning, straight, other-specify). For each item except age, participants could indicate more than one category. 2.3.2 Anti-SGM victimization Victimization was assessed using a measure from the Rainbow Illinois study (Oswald & Holman, 2013). Participants were asked how frequently over the past year they had experienced the following types of victimization based on their SGM identity: overheard anti-SGM comments, shunned, teased, threatened, pushed, punched, sexually assaulted, asked to leave, kicked out of the house, refused service, damaged property, outed, and followed. Answer options included never, once, a few times, monthly, weekly, or daily. Consistent with original research using this measure (Paceley, Oswald, & Hardesty, 2014), we aggregated the frequency of victimization into three categories: never (0), once to a few times (1), and monthly or more (2). The items were grouped into subcategories based on the type of victimization: physical or nonphysical. Items included in the physical subscale were pushed, punched, sexually assaulted, or property damage. All other items were placed in the nonphysical subscale, which included overheard anti-SGM comments, shunned, asked to leave, threatened, kicked out of home, refused service, outed, or teased. The scores were computed by summing each item. The mean, standard deviations, and range for this scale are available in Table 1. PACELEY ET AL . 6 TA B L E 1 Demographics of Survey Participants (N=296) and Descriptive Results of Independent Variables Demographic n % Age 14 29 9.8% 15 47 15.9% 16 78 26.4% 17 92 31.1% 18 49 16.6% Missinga 1 Race/ethnicity White, non-Hispanic 209 70.6% Non-White 87 29.5 Asian 7 2.4% Black / African American 20 6.8% Hispanic/Latino 28 9.5% Multiracial 31 10.5% Missinga 1 Gender identity Male Female 57 19.3% 181 61.1% Trans & gender variant 44 14.9% Gender questioning 13 4.4% Sexual orientation Lesbian or gay 75 25.3% Bisexual 87 29.4% Pansexual or queer 81 27.4% Questioning 39 13.2% Other 11 3.7% Missinga 3 – County size Medium/large metro 125 42.2% Small metro 93 31.4% Nonmetro 78 Independent variable 26.4% Mean (SD) Range 1.75 (.77) 1.00–6.25 Age Physical victimization Nonphysical victimization 2.66 (.05) 1.00–5.75 DASS Depression subscale 19.92 (9.35) 0–42 DASS Anxiety subscale 16.55 (8.59) 0–42 DASS Stress subscale 20.89 (8.02) 0–42 Community climate 2.07 (.61) 1–3 Note. DASS = Depression, Anxiety, and Stress Scale. a These questions were skipped. However, participants completed eligibility questions pertaining to age and sexual orientation and therefore their survey remained in the analytic sample. PACELEY ET AL . 2.3.3 7 Mental health Mental health symptoms were assessed using the Depression, Anxiety, and Stress Scale, short-version (DASS-21; Lovibond & Lovibond, 1995). The DASS-21 includes three subscales with seven questions each to measure depression, anxiety, and stress. The depression subscale (a = .88) included questions such as “I couldn’t seem to experience any positive feelings at all” and “I felt that I had nothing to look forward to.” The anxiety subscale (a = .81) included questions such as “I was worried about situations in which I might panic and make a fool of myself” and “I felt I was close to panic.” The stress subscale (a = .82) included items such as “I found it hard to wind down” and “I tended to over-react to situations.” Participants indicated how often they experienced each symptom over the past week using a 4-point Likert scale ranging from 0 (did not apply to me at all) to 3 (all of the time). Each subscale was scored by summing items (see mean, standard deviation, and range in Table 1) and then scores were labeled based on the scheme used in the DASS manual, which entailed categories of none, mild, moderate, severe, or extremely severe (Lovibond & Lovibond, 1995). The scores were then recoded to a dichotomous variable to fit the assumptions of a logistic regression analysis. Each scale was dummy coded 0 (none to mild symptoms) or 1 (moderate, severe, or extremely severe). 2.3.4 Community climate Perceived community climate was measured by asking, “What is the climate toward lesbian, gay, bisexual, transgender, queer, and questioning (LGBTQ) people where you live?” (Oswald & Holman, 2013). Answer options included hostile (0), tolerant (1), or supportive (2). For the purposes of creating predictor variables, scores on perceived climate were recoded as a binary: hostile (0) or tolerant/supportive (1). 2.3.5 Community size Participants provided their zip code or town name. County size was calculated as medium/large metropolitan (metro county population ≥250,000), small metropolitan (metro county population < 250,000), or nonmetropolitan based on categorizations in the National Center for Health Statistics (2013) urban–rural classification scheme. Medium/large metropolitan and small metropolitan were each dummy-coded variables, and the reference category was nonmetropolitan. Additionally, participants indicated how long they had lived in their community (more than a year, 6 months to 1 year, or less than 6 months). Because victimization was measured as occurring within the past year, only participants living in their community for more than a year were included in the analysis. 2.4 Data analysis Data were analyzed using Stata/IC (version 14.1). First, missing data were analyzed. Anti-SGM victimization, depression, anxiety, and stress variables were found to be missing at random and thus were replaced using multiple imputation. Imputations went through 10 iterations and then were pooled and aggregated before analysis. There were no problems with multicollinearity or outliers. Next, the data were analyzed with three separate binary logistic regressions, using the dummy-coded variable of each subscale of the DASS-21 as the dependent variable for each (depression, anxiety, and stress). The independent variables were as follows: age; race (dummy coded as binary for White/non-White); gender (dummy coded to represent categories of male, female, transgender, and questioning as the reference category); sexual orientation (dummy coded to represent categories of lesbian/gay, bisexual, queer, and questioning/other as the reference category); community size (dummy coded to represent large/medium metropolitan area, small metropolitan area, and nonmetropolitan area as the reference category); perceived community climate (dummy coded as binary for hostile or supportive/tolerant); physical victimization; and nonphysical victimization. The data were also analyzed using a Kruskal-Wallis test to determine if there was a difference among groups by perceived community climate and community size on the mean score of the two victimization subscales. The nonparametric Kruskal-Wallis test was used to account for the fact that data were not normally distributed. PACELEY ET AL . 8 3 RESULTS Each model in the logistic regression analyses had a good fit based on the Hosmer-Lemeshow chi-square test (depression c2 = 1.355, degree of freedom [df] = 8, p = 0.995; anxiety c2 = 1.087, df = 8, p = 0.998; stress c2 = 8.324, df = 8, p = 0.403). The regressions predicted each respective dependent variable for each model with an accuracy level of 78.5% for depression, 82.6% for anxiety, and 67.6% for stress. The Wald criterion indicated that in all three models, the only statistically significant predictor of clinically significant DASS-21 scores was nonphysical victimization, except that in the stress model, small metropolitan size was also a significant predictor (see Table 2). Based on the odds ratios, respondents who had higher nonphysical victimization scores were 2.13 times more likely to score within the clinically significant range on depression, 2.37 times more likely on anxiety, and 2.01 times more likely on stress. In addition, those who lived in a small metropolitan area were 2.66 times more likely to score in the clinically significant range for stress. A Kruskal-Wallis test was used to determine whether there were differences between categories of community size for scores on the victimization subscales, and whether there were differences between categories of perceived community climate for scores on the victimization subscales. There was a statistically significant difference between nonphysical victimization and community size, H(2) = 15.64, p<.001, with a mean rank of 131.12 for medium/large metropolitan, 145.74 for small metropolitan, and 179.65 for nonmetropolitan, which indicates that nonphysical victimization scores were highest for those in a nonmetropolitan community size. There was a statistically significant difference between physical victimization and perceived community climate, H(2) = 36.74, p<.001, with a mean rank of 216.70 for hostile, 140.78 for tolerant, and 122.44 for supportive, which indicates that the lowest level of physical victimization was reported by those in a community climate perceived as supportive. There was a statistically significant difference between nonphysical victimization and perceived community climate, H(2) = 51.31, p<.001, with a mean rank of 216.00 for hostile, 149.84 for tolerant, and 97.52 for supportive, which indicates that the highest level of nonphysical victimization scores were among those in a community climate perceived as hostile. There was no significant difference in physical victimization among levels of community size, H(2) = 4.20, p = .123. 4 DISCUSSION The findings from this study provide insight on the ways in which different types of distal stressors are associated with mental health among SGM youth living within varying community sizes and community climates. Findings revealed that both forms of victimization were more common in nonmetropolitan communities and communities perceived as hostile. Only nonphysical victimization, however, was associated with depression, anxiety, and stress among SGM youth above and beyond physical victimization, demographics, and community-level factors. Other findings suggest that community size and perceived climate were not associated with mental health symptoms, with one exception: SGM youth in small metropolitan counties reported greater stress than SGM youth in other community sizes. 4.1 Distal stressors: Physical and nonphysical victimization The association between anti-SGM victimization and poor mental health outcomes has been well documented in the empirical literature (Burton et al., 2013; Meyer, 2003; Remafedi, 2007; Woodford, Kulick, Sinco, & Hong, 2014). These studies frequently collapse various types of anti-SGM victimization into a single victimization variable, with a few exceptions (see Dragowski, Halkitis, Grossman, & D’Augelli, 2011; Woodford et al., 2014). The current study adds to this literature base as well as MST by discerning between two types of distal stressors–physical and nonphysical victimization–and examining their effects on mental health under varying community conditions. The finding that there were no associations between physical victimization and mental health, when nonphysical victimization and community factors were considered, was surprising. This is contrary to research examining physical victimization and mental health. For example, Dragowski et al. (2011) found an association between both verbal and PACELEY ET AL . TA B L E 2 Logistic Regression Results for Three Models of Dependent Variables: DASS-21 Subscales Depression, Anxiety, and Stress Depression Predictors Age B SE Anxiety Wald Exp(B) 95% CI B Stress SE Wald Exp(B) 95% CI B SE Wald Exp(B) 95% CI .074 .134 .303 1.076 [.828, 1.399] −.167 .145 1.323 .846 [.637, 1.125] −.021 .113 .033 .980 [.785, 1.223] Race (1=White) −.008 .351 .001 .992 [.499, 1.974] −.386 .361 1.138 .680 [.335, 1.381] −.266 .290 .846 .766 [.434, 1.352] Gender (1=male) −.207 .795 .068 .813 [.171, 3.860] −1.554 1.156 1.807 .211 [.022, 2.038] −.241 .701 .118 .786 [.199, 3.103] .639 .725 .777 1.894 [.458, 7.840] −.760 1.101 .477 .467 [.054, 4.043] .159 .639 .062 1.173 [.335, 4.101] Gender (1=female) Gender (1=trans) .008 .810 .000 1.008 [.206, 4.934] −1.926 1.154 2.785 .146 [.015, 1.399] −.118 .709 .028 .889 [.222, 3.565] Sex. ornt. (1=lesbian,gay) −.339 .515 .434 .712 [.260, 1.953] .220 .529 .173 1.246 [.442, 3.515] −.097 .408 .057 .907 [.408, 2.017] Sex. ornt. (1=bisexual) −.007 .506 .000 .993 [.369, 2.675] .294 .502 .341 1.341 [.501, 3.591] .309 .388 .634 1.362 [.636, 2.916] Sex. ornt. (1=queer) −.072 .509 .020 .931 [.343, 2.525] .147 .508 .084 1.159 [.429, 3.134] .373 .396 .887 1.453 [.668, 3.159] Comm. size (1=med,large) −.419 .414 1.021 .658 [.292, 1.482] −.066 .443 .022 .936 [.393, 2.230] .561 .332 2.852 1.752 [.914, 3.360] 2.662 [1.312, 5.363] .740 [.325, 1.684] .237 .460 .266 1.267 [.515, 3.120] −.079 .467 .029 .924 [.370, 2.307] .979 .357 ** 7.498 −.566 .666 .723 .568 [.154, 2.093] .383 .573 .447 1.467 [.477, 4.513] −.301 .420 .514 Physical victimization .531 .394 1.811 1.700 [.785, 3.682] .462 .410 1.269 1.587 [.711, 3.542] −.048 .285 Nonphysical victimization .756 .328 5.302 2.130 [1.119, 4.052] .866 .339 6.518 2.378 [1.223, 4.624] .700 .263 −1.245 1.307 .907 .288 −.075 1.518 .002 .928 −1.594 1.048 Comm. size (1=small) Climate (1=tolrnt.,supprtv.) Constant * ** ** .027 .954 [.546, 1.668] 7.059 2.014 [1.203, 3.370] 2.316 .203 Note. DASS = Depression, Anxiety, and Stress Scale; SE = standard error; CI = confidence interval; Sex. ornt = Sexual orientation; omt = ; comm. = ; tolrnt = ; supportv = . Reference category for gender and for sexual orientation is questioning/other, and for community size is nonmetro. For each dependent variable, 1=moderate, severe, or extremely severe symptoms. *p<.05. **p<.01. 9 PACELEY ET AL . 10 physical victimization and posttraumatic stress symptoms among SGM youth. Alternatively, Woodford et al. (2014) found that more subtle forms of anti-SGM victimization, such as microaggressions, were more influential on mental health than blatant forms of victimization. The lack of significant findings between physical victimization and mental health in this study may be explained by a number of factors, but warrants further research to be more fully understand. It may be that because nonphysical forms of victimization, such as being teased, threatened, or shunned from peers, happen on an ongoing basis, rather than a single occurrence, its effect on mental health is more consistent over a 1-year period than a single act of physical victimization. Birkett, Newcombe, and Mustanski (2015) assessed victimization (physical and nonphysical combined into one composite variable) among SGM youth and found that higher levels of victimization were associated with higher levels of psychological distress. This finding may also be due to the low number of youth reporting physical victimization. Finally, because we did not measure proximal stressors, it may be that these types of stressors would have mediated the relationship between physical victimization and mental health, similar to Mereish and Poteat’s (2015) findings. These findings should not discount the effect of physical victimization on SGM youth, as additional research is needed to better understand the differences in the effect of physical and nonphysical victimization within varying community contexts. Consistent with prior research, the findings of this study do provide support for the relationship between nonphysical forms of victimization and poor mental health among SGM youth. Some nonphysical forms of victimization may be considered microaggressions: “brief and commonplace daily verbal, behavioral, or environmental indignities, whether intentional or unintentional, that communicate hostile, derogatory, or negative slights and insults toward members of oppressed groups” (Nadal, 2008, p. 23). Because microaggressions are commonplace within society, they become built into the institutions and rhetoric that make up our communities. Although microaggressions may be considered less serious than physical forms of victimization, research suggests that overhearing anti-SGM comments, for example, is associated with poor well-being and decreased feelings of acceptance (Woodford, Howell, Silverchanz, & Yu, 2012; Woodford et al., 2014). The findings of this study are consistent with this: Nonphysical victimization was associated with symptoms of depression, anxiety, and stress, even when physical victimization and community variables were considered. 4.2 Community Community-level variables are generally not included in MST (Meyer, 2003); however, conceptually it follows that community climate and size would influence the experiences of victimization and mental health of its members. Findings suggested that, generally, community size and perceived community climate were not associated with mental health outcomes, except for increased stress among SGM youth living in small metropolitan communities. It may be that SGM youth in small metropolitan communities have more visibility surrounding their SGM identity than SGM youth in nonmetropolitan communities, thus increasing their stress around being outed or victimized, yet they have less access to supportive resources to cope with stress than youth in medium/large metropolitan communities. This finding warrants further investigation and promotes the study of community size as a continuum rather than strictly examining rural/urban differences. Additionally, perceived community climate did not affect mental health above and beyond that of nonphysical victimization. This differs from the work of Hatzenbuehler (2011) and Woodford et al. (2015) who found associations between objective community climate and mental health among SGM youth. However, our findings are consistent with Craig and McInroy (2013), who found that the social environment of SGM youth of color did not affect their suicide risk. One possible explanation for these discrepancies may lie in how community climate is measured (perceived vs. objective measures; Oswald et al., 2010) and the influence of experiences of victimization on how one perceives the climate of their community. Including victimization and climate as separate variables in this study revealed that the effect of nonphysical victimization on mental health was significant beyond that of perceived climate. It may be that community climate affects mental health via increased victimization or that one’s perceived climate is directly related to their experiences of victimization, although future studies are needed to test these hypotheses. PACELEY ET AL . 4.3 11 Strengths and limitations This study has multiple strengths. First, it includes a diverse sample of SGM youth across a range of community sizes. The survey included established measures of victimization and mental health with good reliability. This study is not without limitations, however. Because of its cross-sectional nature, we are able to illustrate only associations rather than effects of victimization and community. Additionally, the findings related to perceived climate and victimization experiences may be confounded. For example, youth who experienced greater amounts of physical and nonphysical victimization may have been more likely to identify their communities as hostile. We are unable to determine from these findings if a hostile community led to more victimization or if youth who experience victimization perceive their communities as more hostile. Additionally, only one item was used to assess perceived climate. Using a single item may not capture various facets and nuances comprising perceptions of community climate. Last, objective climate was not included in these analyses. 4.4 Implications This study has implications for practice, policy, theory, and research. First, when working with sexual and gender minorities, it is important to attend to the structural inequalities contributing to mental health disparities (LaSala, 2006). Our sample found nonphysical victimization, which is often built into the structure of society and communities, to be more distressing than physical victimization. Although this should not discount the potential effect of physical victimization, it does emphasize the importance of validating the effect of nonphysical victimization on SGM youth. It may be that nondiscrimination and antibullying policies are interpreted as relating more to physical victimization rather than nonphysical forms of microaggressions and victimization. Social workers and helping professionals can help advocate for policy-level changes and trainings on nondiscrimination and victimization among this population. Because of the limitations of our research design, it is difficult to determine if a hostile climate causes victimization or if youth report a hostile climate based on their experiences of victimization. While in need of further research, a hostile or supportive community climate may predict the amount and type victimization faced by SGM youth. Antidiscrimination ordinances or policies inclusive of SGM individuals may be conducive to a more supportive environment and, in turn, improved mental health. Although community size is not a variable that can be changed, interventions may be developed to enhance community climate toward SGM people. In addition to adding supportive policies to a school or community, the inclusion of SGM-specific resources may contribute to a supportive climate and access to supportive resources for SGM youth. Davis, Saltzburg, and Locke (2010) asked 20 SGM youth in a metropolitan area to identify and rank specific community-based needs; the youth, predominantly ethnic minorities, reported the need for SGM role models, social support, and community support and involvement as highly important. Future research should attend to the role of community when studying SGM youth, mental health, and victimization. Community climate can be measured objectively and subjectively using a variety of measures while community size should be treated as a continuum rather than as a binary variable. In our study, there were differences in the amount of stress SGM youth experienced in small metropolitan communities, yet this category of community size is often excluded in other research because it is categorized as “urban.” Additionally, future research should attempt to understand the frequency and effect of different types of victimization on SGM youth rather than aggregate them into a single variable. Finally, this study has important implications for MST. Distal stressors make up a wide range of stressors and may need to be considered individually as well as collectively. Additionally, while distal stressors are included in the minority stress model, social context and community, where distal stressors can occur, are often neglected. It is worthwhile to include these variables into the minority stress model to better understand the ways in which community exacerbates or mitigates the effect of minority stressors. PACELEY ET AL . 12 4.5 Conclusion This study aimed to understand the effect of two broad types of victimization and community-level variables on the mental health of SGM youth. Understanding the ways in which minority stressors affect the mental health and wellbeing of SGM youth and the factors that may exacerbate or mitigate these outcomes are critical to developing interventions to promote support and well-being. The findings of this study contribute to expanding minority stress theory and our understanding of how SGM youth’s community may affect their well-being. REFERENCES Birkett, M., Espelage, D. L., & Koenig, B. (2009). LGB and questioning students in schools: The moderating effects of homophobic bullying and school climate on negative outcomes. Journal of Youth & Adolescence, 38, 989–1000. doi:10.1007/s10964-008-9389-1 Birkett, M., Newcombe, M. E., & Mustanski, B. (2015). Does it get better? A longitudinal analysis of psychological distress and victimization in lesbian, gay, bisexual, transgender, and questioning youth. Journal of Adolescent Health, 56, 280–285. doi:10.1016/j.jadohealth.2014.10.275 Birkett, M., Russell, S. T., & Corliss, H. L. (2014). Sexual-orientation disparities in school: The mediational role of indicators of victimization in achievement and truancy because of feeling unsafe. American Journal of Public Health, 104, 1124–1128. doi:10.2105/AJPH.2013.301785 Burton, C., Marshal, M., Chisolm, D., Sucato, G., & Friedman, M. (2013). Sexual minority-related victimization as a mediator of mental health disparities in sexual minority youth: A longitudinal analysis. Journal of Youth & Adolescence, 42, 394–402. doi:10.1007/s10964-012-9901-5 Cohn, T. J., & Leake, V. S. (2012). Affective distress among adolescents who endorse same-sex sexual attraction: Urban versus rural differences and the role of protective factors. Journal of Gay & Lesbian Mental Health, 16, 291–305. doi:10.1080/19359705.2012.690931 Craig, S. L., & McInroy, L. (2013). The relationship of cumulative stressors, chronic illness, and abuse to the self-reported suicide risk of Black and Hispanic sexual minority youth. Journal of Community Psychology, 41, 783–798. doi:10.1002/jcop. 21570 D’Augelli, A. R., Grossman, A. H., & Starks, M. T. (2006). Childhood gender atypicality, victimization, and PTSD among lesbian, gay, and bisexual youth. Journal of Interpersonal Violence, 21, 462–1482. doi:10.1177/0886260506293482 Davis, T. S., Saltzburg, S., & Locke, C. R. (2010). Assessing community needs of sexual minority youths: Modeling concept mapping for service planning. Journal of Gay & Lesbian Social Services, 22, 226–249. doi:10.1080/10538720903426354 Dragowski, E. A., Halkitis, P. N., Grossman, A., & D’Augelli, A. R. (2011). Sexual orientation victimization and posttraumatic stress symptoms among lesbian, gay, and bisexual youth. Journal of Gay & Lesbian Social Services, 23, 226–249. doi:10.1080/10538720.2010.541028 Espelage, D. L., Aragon, S. R., Birkett, M., & Koenig, B. W. (2008). Homophobic teasing, psychological outcomes, and sexual orientation among high school students: What influence do parents and schools have? School Psychology Review, 37, 202– 216. Friedman, M. S., Koeske, G. F., Silvestre, A. J., Korr, W. S., & Sites, E. W. (2006). The impact of gender-role nonconforming behavior, bullying, and social support on suicidality among gay male youth. Journal of Adolescent Health, 38(5), 621– 623. Friedman, M. S., Marshal, M. P., Guadamuz, T. E., Wei, C., Wong, C. F., Saewyc, E. M., & Stall, R. (2011). A meta-analysis of disparities in childhood sexual abuse, parental physical abuse, and peer victimization among sexual minority and sexual nonminority individuals. American Journal of Public Health, 101(8), 1481–1494. Gray, M. L. (2007). From websites to Wal-Mart: Youth, identity work, and the queering of boundary publics in small town, USA. American Studies, 48(2), 49–59. Hatzenbuehler, M. L. (2011). The social environment and suicide attempts in lesbian, gay, and bisexual youth. Pediatrics, 127, 896–903. doi:10.1542/peds.2010-3020 Hendricks, M. L., & Testa, R. J. (2012). A conceptual framework for clinical work with transgender and gender nonconforming clients: An adaptation of the Minority Stress Model. Professional Psychology: Research and Practice, 43, 460–467. doi:10.1037/a0029597 Kazyak, E. (2011). Disrupting cultural selves: Constructing gay and lesbian identities in rural locales. Qualitative Sociology, 34, 561–581. doi:10.1007/s11133-011-9205-1 PACELEY ET AL . 13 King, M., Semlyen, J., Tai, S. S., Killaspy, H., Osborn, D., Popelyuk, D., & Nazareth, I. (2008). A systematic review of mental disorder, suicide, and deliberate self harm in lesbian, gay, and bisexual people. BMC Psychiatry, 8(1), 1–17. doi:10.1186/1471-244X-8-70 Kosciw, J. G., & Diaz, E. M. (2006). 2005 National School Climate Survey: The experiences of lesbian, gay, bisexual, and transgender youth in our nation’s schools. New York: Gay, Lesbian and Straight Education Network. Kosciw, J. G., Diaz, E. M., & Greytak, E. A. (2008). 2007 National School Climate Survey: The experiences of lesbian, gay, bisexual and transgender youth in our nation’s schools. New York: Gay, Lesbian and Straight Education Network. Kosciw, J. G., Greytak, E. A., Palmer, N. A., & Boesen, M. J. (2014). The 2013 National School Climate Survey: The experiences of lesbian, gay, bisexual and transgender youth in our nation’s schools. New York: GLSEN. LaSala, M. C. (2006). Cognitive and environmental interventions for gay males: Addressing stigma and its consequences. Families in Society, 87(2), 181–189. doi:10.1606/1044-3894.3511 LeBeau, R. T., & Jellison, W. A. (2009). Why get involved? Exploring gay and bisexual men’s experience of the gay community. Journal of Homosexuality, 56, 56–76. doi:10.1080/00918360802551522 Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales (2nd ed.) Sydney: Psychology Foundation. Marshal, M. P., Dietz, L. J., Friedman, M. S., Stall, R., Smith, H. A., McGinley, J. … Brent, D. A. (2011). Suicidality and depression disparities between sexual minority and heterosexual youth: A meta-analytic review. Journal of Adolescent Health, 49, 115– 123. Mereish, E. H., & Poteat, V. P. (2015). A relational model of sexual minority mental and physical health: The negative effects of shame on relationships, loneliness, and health. Journal of Counseling Psychology, 62(3), 425. Meyer, I. H. (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129, 674–697. doi:10.1037/0033-2909.129.5.674 Mustanski, B. S., Garofalo, R., & Emerson, E. M. (2010). Mental health disorders, psychological distress, and suicidality in a diverse sample of lesbian, gay, bisexual, and transgender youths. American Journal of Public Health, 100(12), 2426– 2432. Nadal, K. L. (2008). Preventing racial, ethnic, gender, sexual minority, disability, and religious microaggressions: Recommendations for promoting positive mental health. Prevention in Counseling Psychology: Theory, Research, Practice and Training, 2(1), 22–27. National Center for Health Statistics. (2014). 2013 NCHS urban-rural classification scheme for counties. Washington, DC: US Department of Health and Human Services. http://doi.org/10.1037/0033-2909.129.5.674 Olson, L., Cadge, W., & Harrison, J. (2006). Religion and public opinion about same-sex marriage. Social Science Quartlery, 87, 340–360. doi:10.1111/j.1540-6237.2006.00384.x Oswald, R. F., & Culton, L. S. (2003). Under the rainbow: Rural gay life and its relevance for family providers. Family Relations, 52(1), 72–81. Oswald, R. F., Cuthbertson, C., Lazarevic, V., & Goldberg, A. E. (2010). New developments in the field: Measuring community climate. Journal of GLBT Family Studies, 6(2), 214–228. doi:10.1080/15504281003709230 Oswald, R. F., & Holman, E. G. (2013). Rainbow Illinois: How downstate LGBT communities have changed (2000-2011). Department of Human and Community Development, University of Illinois at Urbana-Champaign. Paceley, M. S., Oswald, R. F., & Hardesty, J. (2014). Predicting involvement in nonmetropolitan LGBTQ community centers: Proximity? Generativity? Minority stress? Social location? Journal of Homosexuality, 61, 1481–1500. doi:10/1080/00918369.2014.928582 Palmer, N. A., Kosciw, J. G., & Bartkiewicz, M. J. (2012). Strengths and silences: The experiences of lesbian, gay, bisexual and transgender students in rural and small town schools. New York: GLSEN. Poon, C. S., & Saewyc, E. M. (2009). Out yonder: Sexual-minority adolescents in rural communities in British Columbia. American Journal of Public Health, 99(1), 118–124. doi:10.2105/AJPH.2007.122945 Poteat, V. P., Aragon, S. R., Espelage, D. L., & Koenig, B. W. (2009). Psychosocial concerns of sexual minority youth: Complexity and caution in group differences. Journal of Consulting and Clinical Psychology, 77(1), 196–201. doi:10.1037/a0014158 Roberts, A. L., Rosario, M., Corliss, H. L., Koenen, K. C., & Austin, S. B. (2012). Childhood gender nonconformity: A risk indicator for childhood abuse and posttraumatic stress in youth. Pediatrics, 129(3), 410–417. Robinson, J. P., & Espelage, D. L. (2011). Inequities in educational and psychological outcomes between LGBTQ and straight students in middle and high school. Educational Researcher, 40(7), 315–330. doi:10.3102/0013189X11422112 Robinson, J. P., Espelage, D. L., & Rivers, I. (2013). Developmental trends in peer victimization and emotional distress in LGB and heterosexual youth. Pediatrics, 131(3), 423–430. PACELEY ET AL . 14 Ramirez-Valles, J., Fergus, S., Reisen, C. A., Poppen, P. J., & Zea, M. C. (2005). Confronting stigma: Community involvement and psychological well-being among HIV-positive Latino gay men. Hispanic Journal of Behavioral Sciences, 27(1), 101–119. doi:10.1177/0739986304270232 Remafedi, G. (2007). Lesbian, gay, bisexual, and transgender youths: Who smokes, and why? Nicotine & Tobacco Research, 9(Suppl 1), S65–S71. doi:10.1080/14622200601083491 Rosario, M., Schrimshaw, E. W., & Hunter, J. (2006). A model of sexual risk behaviors among gay and bisexual men: Longitudinal associations of mental health, substance abuse, sexual abuse, and the coming out process. AIDS Education & Prevention, 18(5), 444–460. doi:10.1521/aeap.2006.18.5.444 Russell, S. T., Ryan, C., Toomey, R. B., Diaz, R. M., & Sanchez, J. (2011). Lesbian, gay, bisexual, and transgender adolescent school victimization: Implications for young adult health and adjustment. Journal of School Health, 81(5), 223–230. Ryan, C., Huebner, D., Diaz, R. M., & Sanches, J. (2009). Family rejection as a predictor of negative health outcomes in White and Latino lesbian, gay, and bisexual young adults. Pediatrics, 123(1), 346–352. doi:10.1542/peds.2007-3524 Stange, M., & Kazyak, E. (2016). Examining the Nuance in public opinion of pro-lgb policies in a “red state.” Sexuality Research and Social Policy, 13(2), 142–157. Swank, E., Fahs, B., & Frost, D. M. (2013). Region, social identities, and disclosure practices as predictors of heterosexist discrimination against sexual minorities in the United States. Sociological Inquiry, 83(2), 238–258. doi:10.1111/soin.12004 Taylor, C. G. (2008). Counterproductive effects of parental consent in research involving LGBTTIQ youth: International research ethics and a study of transgender and two-spirit community in Canada. Journal of LGBT Youth, 5, 34–56. doi:10.1080/19361650802162177 Toomey, R. B., Ryan, C., Diaz, R. M., Card, N. A., & Russell, S. T. (2010). Gender-nonconforming lesbian, gay, bisexual, and transgender youth: School victimization and young adult psychosocial adjustment. Developmental Psychology, 46(6), 1580. Wienke, C., & Hill, G. J. (2013). Does place of residence matter? Rural–urban differences and the wellbeing of gay men and lesbians. Journal of Homosexuality, 60, 1256–1279. doi:10.1080/00918369.2013.806166 Woodford, M. R., Han, Y., Craig, S., Lim, C., & Matney, M. M. (2014). Discrimination and mental health among sexual minority college students: The type and form of discrimination does matter. Journal of Gay & Lesbian Mental Health, 18, 142–163. doi:10.1080/19359705.2013.833882 Woodford, M. R., Howell, M. L, Silverchanz, P., & Yu, L. (2012). “That’s so gay!”: Examining the covariates of hearing this expression among gay, lesbian, and bisexual college students. Journal of American College Health, 60, 429–434. doi:10.1080/07448481.2012.673519 Woodford, M. R., Paceley, M. S., Kulick, A., & Hong, J. S. (2015). The LGBQ social climate matters: Policies, protests, and placards and psychological well-being among LGBQ emerging adults. Journal of Gay & Lesbian Social Services, 27, 116–141. doi:10.1080/10538720.2015.990334 How to cite this article: Paceley MS, Goffnett J, & Gandy-Guedes M. Impact of victimization, community climate, and community size on the mental health of sexual and gender minority youth. J Community Psychol. 2017;00:00–00. doi:10.1002/jcop.21885