Hazards of Stigma: The Sexual and Physical Abuse of Gay, Lesbian, and Bisexual Adolescents in the United States and Canada Elizabeth M. Saewyc, Carol L. Skay, Sandra L. Pettingell, Elizabeth A. Reis, Linda Bearinger, Michael Resnick, Aileen Murphy, and Leigh Combs Some studies suggest lesbian, gay, and bisexual (LGB) teens are at higher risk than peers for violence at home, in school, and in the community. That can bring them into the child welfare system or services for runaway and homeless teens. This study compared self-reported experiences of sexual and physical abuse based on sexual orientation and gender in seven population- based surveys of youth. The authors used d- and age- adjusted odds of abuse to compare bisexual to hetero- sexual, mostly heterosexual, and gay and lesbian students. They also provide case studies to illustrate the experiences of such youth. Elizabeth M. Saewyc, PhD, RN, PHN, is Research Director, McCreary Centre Soci- ety, and Associate Professor, School of Nursing, University of British Columbia, Vancouver, BC, Canada; Carol L. Skay, PhD, is a Senior Research Associate, School of Nursing, University of Minnesota, Minneapolis, MI; Sandra L. Pettingell, PhD, is Research Associate, School of Nursing, University of Minnesota; Elizabeth A. Reis, MS, is Public Health Educator, Public Health Seattle-King County, Seattle, WA; Linda H. Bearinger, PhD, MS, RN, is Professor, School of Nursing, University ofMinne- 0009-4021/2006/030195-19 $3.00 © 2006 Child Welfare League of America 195 NOT FOR DISTRIBUTION 196 CHILD WELFARE • Vol. LXXXV, #2 • March/April Case 1. "Troy" was taken from his mother at age 5 be- cause of neglect and abuse. He has moved so often and been in so many foster care facilities that he has lost count. Until he met an openly gay teen at his latest foster home, he had no idea this was even a possibility. Troy really liked this boy; one thing led to another, and they ended up get- ting caught in the midst of a sexual experience, which was very embarrassing. His social worker thought it best the boys be separated, and placed Troy in a home with deeply religious conservative foster parents. Along with other religious rituals that were supposed to eliminate his de- sire for men, his foster parents' church tried exorcism. Negative messages toward sexual minority groups arecommon across North American society: from the constant antigay pejoratives heard in high school hallways, to the bitter protests at school board meetings against bullying policies that specifically include antigay harassment; in sermons at local places of worship and in major statements by the leaders of some religious denominations; in arguments before the courts, as well as debates in the U.S. Congress and Canadian House of Commons; and recently, in the media's wide coverage of cam- paigns against gay marriage, as well as its gratuitous humor in sitcoms and commercials at gay people's expense. In nearly ev- ery social arena, identifying as lesbian, gay, or bisexual (LGB) carries a burden of stigma and societal disapproval. That disap- proval is not always limited to words; research has identified pervasive discrimination experienced by sexual minority adults in U.S. employment, housing, and social interactions nationwide (Mays & Cochran, 2001), as well as hate crimes ranging from prop- erty damage to murder (Federal Bureau of Investigation, 2004). sota; Michael Resnick, PhD, is Professor, Division of Adolescent Health, Department of Pediatrics, University of Minnesota; Aileen Murphy, MA, is Managing Director, McCreary Centre Society; Leigh Combs, AS, CDCP, is Program Coordinator, GLBT Kids: Abuse Intervention, Family and Children's Service, Minneapolis. NOT FOR DISTRIBUTION Saewyc, etal. 197 Health Canada has identified stigma and discrimination based on sexual orientation status as social determinants of health that contribute to health disparities for LGB people (de Bruyn, 2005). Case 2. "Sue," age 15, lives in the suburbs with both par- ents and two younger siblings. She feels that her parents' beliefs are too strict for her, especially since she has a crush on one of her girlfriends. Her parents have no idea about her romantic exploration, but they are concerned her be- havior has changed in the last six months. They also do not know yet that kids at school have found out about Sue and her friend and are harassing the two girls. Sue wants to get away from everyone at this point. Some studies from the U.S. and Canada have documented the anxious, angry, and even violent responses LGB youth expe- rience from their families, at school, and in the community (Reis & Saewyc, 1999). Families can be influenced by the stigmatizing attitudes in their social and cultural environments, and when a son or daughter discloses a gay, lesbian, or bisexual orientation, some families respond with hostility and distress (D'Augelli, Hershberger & Pilkington, 1998; Pilkington & D'Augelli, 1995; Murphy, Sidhu, & Tonkin, 1999; Waldo, Hesson-Mclnnis, & D'Augelli, 1998). One limitation of these studies is they draw on convenience samples of LGB youth and may not represent the general population of sexual minority teens. They do suggest, however, that some of the suspected maltreatment cases child Acknowledgments: These analyses were supported by Grant ROl MH-6258601 from the National Institute of Mental Health, National Institutes of Health (Saewyc, origi- nal PI; Skay, second PI); Grant ROl DA1797901, National Institute of Drug Abuse (Saewyc, original PI; Skay, second PI); and a Michael Smith Foundation for Health Research Career Scholar award (Saewyc, PI). Preliminary results of the study were presented at the 2004 annual meeting of the Society for Adolescent Medicine, St. Louis, Missouri. Thanks to Pamela Hillard, Seattle School District, for access to the Seattle Adolescent Health Surveys of 1995 and 1999, the Minnesota Department of Educa- tion for access to the Minnesota Student Surveys, and Dr Roger Tonkin, McCreary NOT FOR DISTRIBUTION 198 CHILD WELFARE • Vol. LXXXV, #2 • March/April welfare workers encounter will involve LGB youth and unsupportive family members. Social workers also can be influ- enced by community attitudes and respond in ways that are not supportive of LGBTQ youth. Some evidence shows actual disclosure is not the only trigger for family maltreatment toward sexual minority youth. In their study of victimization among LGB college students, Waldo and colleagues (1998) find that those who had disclosed their orien- tation to unsupportive families were not only more likely to ex- perience victimization, but those with higher gender atypicality, which has been correlated with homosexual orientation, also were at risk for victimization, even when they had not disclosed to family. Waldo and colleagues suggest gender-atypical youth of- ten are assumed to be gay, lesbian, or bisexual, and families may not protect such youth from victimization or may even perpe- trate the violence themselves. Case 3. "Joe," age 17, grew up in the rural Midwest. He was sexually abused by his stepbrother, but did not tell anyone. He was harassed throughout his school years, mostly by name calling with derogatory words like "fag." Joe stopped going to school in 10th grade, after being jumped by several boys in the locker room and having his head put in a toilet. He ran away to the city and lived in a squat with other young people he met at a homeless youth program. For money, Joe would stand outside the gay bars at closing time and make a few dollars by giving men oral sex in the parking lot and alleys. A small number of population-based studies also have docu- mented higher prevalence of physical abuse by family members among LGB people compared to heterosexual peers (Horliss, Gochran, & Mays, 2002; Saewyc, Bearinger, Blum, & Resnick, 1999; Saewyc, Skay, Bearinger, Blum, & Resnick, 1998). Like Waldo and colleagues, Horliss and colleagues suggest stigma from gender atypicality or some as-yet unmeasured trait of emerging gay or NOT FOR DISTRIBUTION Saewyc, etal. 199 bisexual orientation may decrease family protection and support for LGB teenagers even before they recognize and self-identify and, thus, may help explain higher risk for maltreatment during childhood and adolescence. Why is it important to identify this disparate risk for violence among stigmatized populations such as LGB teens? Sexual and physical abuse are among the most potent predictors of youth risk behaviors such as substance use and abuse, suicide attempts, running away, and teen pregnancy (Anderson, Teicher, Polcari, & Renshaw, 2002; DeBellis, 2002; Holmes & Slap, 1998; Saewyc, Magee, & Pettingell, 2004). In part, these risk behaviors may be attempts to cope with the stress and trauma from that abuse. In the past decade, several studies have reported a higher preva- lence of these risk behaviors among sexual minority youth, with some of the same surveys also noting a higher risk of harassment and sexual violence for LGB teens (Borowsky, Ireland, & Resnick, 2001; DuRant, Krowchuk, & Sinai, 1998; Hershberger & D'Augelli, 1995; Robin, Brener, Donahue, Hack, & Goodenow, 2002; Rotheram-Borus, Marelich, & Srinivasan, 1999; Russell & Joyner, 2001; Saewyc et al., 1999). While most studies combined LGB teens in a single category, among those surveys that did not, bisexual adolescents or those with both gender attractions appeared to be at higher risk for victimization and reported differing levels of risk behaviors than gay and lesbian peers (Robin et al., 2002; Russell, Franz, & Driscoll, 2001). Case 4. "Colette," age 14, lived with her father and older brother; her mother died when she was 7. She first had sex at age 12, with her best friend from Girl Scouts; eight months later she dated a boy from school and found she liked sex with him, too. Now she's in love with "Jenny," also age 14, and they've been together most of the school year. When she decided to confide in her brother that she's bisexual, she was unprepared for his outrage. He told their father, who kicked her out of the house, and then her NOT FOR DISTRIBUTION 200 CHILD WELFARE • Vol. LXXXV, #2 • March/April brother got a couple of his friends to assault Jenny. Con- vinced Jenny's family would be just as hostile, the two girls ran away. To survive, Jenny panhandles, while Colette works as a stripper and has sex with their land- lord for the basement room they stay in. She has come to the teen clinic four months pregnant. Other population-based studies have documented higher rates of harassment and victimization among LGB youth at school or in the community, rather than family-specific violence (DuRant et al., 1998; Robin, et al., 2002; Russell, Franz & DriscoU, 2001). While two of these studies included sexual abuse in the form of forced intercourse (DuRant et al.; Robin et al.), DuRant's study focused on boys only, and both were limited to the East Coast of the United States. Russell and colleagues used a nationally representative sample to explore both experiencing and witnessing violence among same-gender and both-gender-attracted adolescents, but their study did not include experiences of sexual violence or physi- cal abuse within the family. Thus, a small but growing number of studies, primarily from the eastern United States, and one study from western Canada, have begun to document the increased risk among LBG youth for familial physical abuse, family and other sexual abuse, and harassment and physical violence in the community. If a higher level of stigma and violence directed toward LGB youth is con- sistently documented in surveys across North America, it may help explain higher behavior risks noted among them. The purpose of this study, therefore, is to compare the preva- lence of sexual and physical abuse experienced by bisexual youth with their gay, lesbian, and heterosexual peers in population-based surveys of high school students from the United States and Canada. The authors hypothesized that bisexual boys and girls would re- port greater odds of sexual and physical abuse than heterosexual peers, but not necessarily greater than gay or lesbian peers. NOT FOR DISTRIBUTION Saewyc, et al. 201 Methods As part of a larger study of health and risk behaviors, stigma, and sexual orientation, the authors conducted secondary analyses from seven population-based high school health surveys in the United States and Canada (Table 1). The survey data had been gathered during the 1990s: four were state- or province-wide: the Minnesota Student Surveys of 1992 and 1998 (MSS92 and MSS98) and the Brit- ish Columbia Adolescent Health Surveys of the same years (BC92 and BC98); two from Seattle (SEA95 and SEA99); and the last the first wave of a nationally representative longitudinal study of youth— Add Health. The two surveys each from British Columbia and Min- nesota asked about family physical abuse. All of the surveys assessed sexual abuse experience, although items were worded differently. In the case of the Seattle and Add Health surveys, they were limited to a question about experiencing forced intercourse, which misses youth whose abuse experiences may include only oral sex without genital or anal penetration. This abuse could be just as traumatic but would not be counted with such a question (Saewyc, Pettingell, & Magee, 2003). In addition, while the Add Health survey asked girls if ^ e y had ever been forced to have intercourse, it only asked boys if they had ever forced someone to have intercourse; thus, the authors were limited to sexually experienced girls from that study. BC98 asked about forced intercourse, plus a separate sexual abuse question that was broader and matched the item from BC92. The measures of sexual orientation varied within the different surveys. Add Health asked two questions about the genders of people to whom students were attracted; Seattle and British Co- lumbia had a self-labeling measure that defined the labels on the basis of attraction, with British Columbia including a "mostly het- erosexual" category with heterosexual, bisexual, and gay/lesbian. In contrast, the Minnesota surveys asked two items about genders of sexual partners in the past 12 months. Detailed information about NOT FOR DISTRIBUTION 202 CHILD WELFARE • Vol. LXXXV, #2 • March/April TABLE 1 Datasets and Measures DATASET: Design: BC92 Stratified duster sample (weighted) BC98 Stratified ciuster sampie (weigfited) MSS92 Census of 97% of scfiool districts (subset of ever fiad sex only) MSS98 Census of 92% of scfiooi districts (subset of ever had sex oniy) SEA95 Census of ali students public bigb sciiooi SEA99 Census of aii students pubiic high schooi ADD HEALTH Stratified ciuster sample (nation- wide, giris ever had sex oniy) N: Grade renge: % lemale: Orientation measure; Sexual abuse measure: 237,748 7-12 50.5% Attraction .f iabei Singie abuse item 278,102 7-12 52.7% Attraction * label Singie abuse item.t forced intercourse 24,880 9 & 12 47.7% Gender of sexuai partners past year 2 items: incest and non-famiiy abuse 22,007 9 & 1 2 49.3% Gender of sexuai partners past year 2 items: incest and non-famiiy abuse 7,448 9-12 50.9% Labei Forced intercourse 7,610 9-12 51.4% Labei Forced intercourse 2,228 7-12 100% Attraction Forced intercourse Physical Single item, Singie item. Abused by Abused by None f^one None abuse meastire: physicai abuse by family or anyone physical abuse by famiiy or anyone aduit in househoid or witnessed abuse in family aduit in household or witnessed abuse in famiiy the specific wording of items, percentage of missing and nonresponse, and evaluation of the measures for most of the sur- veys has been reported elsewhere (Saewyc, Bauer, Skay, Bearinger, Resnick, Reis, & Murphy, 2004). Although attraction and behav- ior are not always congruent with orientation self-identity, for brevity throughout this paper the authors will use the conven- tion of bisexual, gay/lesbian, heterosexual, and mostly hetero- sexual for each survey. The authors recognize these labels may not be precise for students who indicate just attraction or recent sexual behavior. NOT FOR DISTRIBUTION Saewyc, et al. 203 All analyses were conducted separately for boys and girls. The authors used cross-tabulations with Pearson's chi square to examine the prevalence of abuse by orientation in each survey, and logistic regression to calculate age-adjusted odds ratios with 95% confidence intervals for risk of abuse for bisexual boys and girls compared to the other groups. Because of the large sample sizes, the authors set alpha to .01 for all analyses, except for the comparisons between bisexual and gay/lesbian groups to avoid detecting statistically significant but not clinically meaningful dif- ferences. For bisexual versus gay/lesbian comparisons, which were smaller samples, alpha was set to .05. Results Sexual Abuse by Orientation The prevalence of sexual abuse based on orientation and related age-adjusted odds ratios for boys and girls in each dataset are re- ported in Table 2. In all surveys except Seattle's, girls were more likely to report abuse than boys, but the differences between ori- entation groups were more marked among boys. Among girls, les- bian and bisexual girls reported the highest prevalence of sexual abuse, with 1 in 4 to nearly half reporting a history of sexual abuse. Heterosexual and mostly heterosexual girls' prevalence of reported sexual abuse ranged from just under 10% to just over 25%. In contrast, the prevalence of sexual abuse or forced inter- course reported by heterosexual and mostly heterosexual boys is well under 10%, but gay and bisexual boys are nearly as likely as bisexual and lesbian girls to report sexual abuse. For most sur- veys, more than 1 in 4 bisexual boys and 1 in 5 gay boys reported sexual abuse. Except for girls in both British Columbia and Minnesota in 1992, bisexual teens of both genders had significantly greater odds of sexual abuse compared to their heterosexual age peers. Among NOT FOR DISTRIBUTION 204 CHILD WELFARE • Vol. LXXXV, #2 • March/April TABLE 2 Prevalence and Age-Act]usted Odds Ratios of Sexual Abuse/Forced Intercourse, by Orientation and Gender Girls, Prevalence ol Sexual Abuse by Orientation (%) BC92 BC98 MSS92 MSS98 SEA95 SEA99 AddHealth Het 21.0 15.4 27.4 20.6 14.2 9.0 19.3 MH 23.1 27.3 - - - - B 23.8 35.9 32.4 39.8 26.3 26.9 32.8 L 43.3 30.2 23.9 39.2 40.7 17.4 27.5 Boys, Prevalence ot Sexual Abuse by Orientation BC92 BC98 MSS92 MSS98 SEA95 SEA99 Het 3.2 3.4 4.8 6.4 5.8 5.4 MH 8.1 8.4 - - - B 14.9 26.9 21.7 27.5 31.3 30.5 G 22.5 26.3 17.7 22.0 17.4 31.6 Girls, Adlusted OR (95% CI) ol Sexual Abuse B vs. Het NS 2.8 (2.6,3.0) t NS 2.3(1.9, 2.9) t 2.1 (1.4, 3.0) t 3.8 (2.7, 5.5) i 2.1 ( 1 . 4 , 3 . 0 ) t B vs. MH NS 1.5(1.4,1.7) t - - • - B v s . L 0.3 (0.2,0.5) * NS NS NS NS NS NS Boys, Adlusted OR (95% Ci) ol Sexual Abuse B vs. Het 5.4 (4.8,6.2) t 10.9(9.7,12.3)* 5.4 (4.6, 6.3) t 5 . 2 ( 4 . 5 , 6 . 1 ) * 7.1(4.5,11.3)* 7.3(4.4,12.0)* B vs. MH 2 . 3 ( 1 . 7 , 2 . 4 ) * 4 . 4 ( 3 . 8 . 5 . 1 ) * - - - - Bvs.G 0.6 (0.5,0.8) t NS NS NS NS NS p<.01; tp<.001; • = not applicable in this survey NS = not significant boys, the age-adjusted odds ratios range from 5 to almost 11, and for girls, most are 2 or higher. That means bisexual boys were up to 10 times as likely and bisexual girls at least twice as likely to report sexual abuse as their heterosexual peers of the same age. Adjusted odds of abuse for mostly heterosexual teens compared to bisexual teens varied between the two years of British Columbia surveys, with higher odds ratios for bisexual boys in both years but only in 1998 for bisexual girls. Differences were generally not sigruficant between gay/lesbian and bisexual teens, but in BC1992, gay or les- bian teens had higher odds of sexual abuse for both boys and girls. NOT FOR DISTRIBUTION Saewyc, et al. 205 Physical Abuse by Family Members In the four surveys that assessed physical abuse, girls in all ori- entation groups report higher prevalence of physical abuse than their male peers (Table 3). Bisexual and lesbian girls also report a higher prevalence of physical abuse than heterosexual girls in the more recent surveys and, in BC98, also higher than mostly heterosexual girls. The age-adjusted odds of physical abuse for bisexual girls was twice as likely when compared to heterosexual peers in the 1998 surveys but not significant for the 1992 surveys. In the British Columbia surveys, bisexual girls had significantly lower odds of abuse than lesbian peers, but the differences were not significant in Minnesota surveys. In each survey, gay and bisexual boys also report a higher preva- lence of physical abuse, with nearly 1 in 5 up to 1 in 3 reporting abuse, compared to 1 in 8 heterosexual boys. The age-adjusted odds of experiencing physical abuse among boys showed bisexual youth have at least twice the odds of abuse compared to their hetero- sexual peers for all datasets, with odds ranging from 2.0 to 2.8. However, comparisons with gay peers were mixed: In the two ear- lier surveys, the differences were not significant, but in BC98, bi- sexual boys had lower odds of physical abuse, while in MSS98, they had nearly twice the odds of abuse as their gay peers. Thus, while most surveys had no significant differences in odds of sexual abuse between bisexual teens and their gay or lesbian peers, differences in physical abuse occurred in half the surveys. In all but one instance that had significant differences between the two sexual minority groups, gay or lesbian teens had higher odds of reporting physical abuse. Discussion Patterns across these seven surveys from various regions of the United States and Canada, in both earlier and later years of the NOT FOR DISTRIBUTION 206 CHILD WELFARE • Vol. LXXXV, #2 • March/April T A B L E 3 Prevalence and Age-Adjusted Odds Ratios of Physical Abuse, by Orientation and Gender Girls, Prevalence of Physical Abuse by Orientation Dataset: BC92 BC98 MSS92 MSS98 Het 24.0 19.6 23.0 19.1 MH 28.4 28.1 - - B 21.0 37.8 29.9 35.1 L 26.6 40.8 28.9 43.1 Boys, Prevalence ol Physical Abuse by Orientation Oataset: BC92 BC98 MSS92 MSS98 Het 14.1 12.2 12.2 12.3 MH 22.1 25.5 - B 39.2 27.3 23.5 28.5 G 25.2 33.5 17.1 16.2 Girls, Adjusted OR (95% CI) of Physical Abuse B vs. Het 0.8 (0.7, 0.9) t 2.3 (2.1, 2.5) t NS 2.0(1.6, 2.5) t B vs. MH 0.6 (0.6, 0.7) i 1.6(1.5,1.7) t - - Bvs. L 0.7 (0.5, 0.9) * 0.8 (0.6, 0.9) * NS NS Boys, Adiusted OR (95% CI) of Physloai Abuse B vs. Het 2.0(1.8, 2.3) t 2.8(2.5, 3.1) t 2.1 (1.8, 2.4) t 2.4(2.1,2.8)1 B vs. MH 1.2(1.0,1.3)® NS - Bvs.G 0.5 (0.4, 0.6) t 0.8 (0.7, 0.9) ® NS 1.8(1.2,2.8)® p<.05: p<.01; t p<.001; - = not applicable in this survey: NS = not significant 1990s, with different sampling methods, different wording of abuse items, and even different measures of sexual orientation, are strikingly similar. Although the actual proportion of teens reporting abuse varies from survey to survey, sexual minority teens in nearly all surveys still are significantly more likely to report sexual and physical abuse than their same-age heterosexual counterparts. In most surveys, bisexual teens of either gender were more than twice as likely to report abuse compared to hetero- sexual teens, with the differences markedly stronger for boys. In contrast, only a few surveys had significant differences in abuse histories between bisexual and gay or lesbian teens, and no con- sistent pattern of differences by gender, type of abuse, or region of survey. Of concern, when two surveys are from the same region but from different years, the prevalence of either type of abuse for NOT FOR DISTRIBUTION Saewyc, et al. 207 bisexual teens appears to be higher in later surveys compared to earlier ones, and their odds of abuse versus heterosexual peers are also greater (except for sexual abuse for bisexual boys in Min- nesota, which is unchanged from 1992 to 1998). Technically, three points in time are required to document a trend, but these data suggest the disparities in abuse between bisexual and hetero- sexual teens was worsening, not improving, as the past decade progressed. Further tracking of abuse and orientation in surveys after 2000 must take place to determine whether this pattern con- tinues to worsen. A number of issues should be considered when putting these findings in context. Beginning at puberty, sexual identity is a de- velopmental task of adolescence, influenced by the societal and cultural contexts in which young people make sense of their physi- cal, emotional, and cognitive changes. Sexual orientation, as one component of sexual identity, unfolds over this period of develop- ment, and the processes begin earlier for some teens than for oth- ers (Rosario, Meyer-Bahlburg, Hunter, & Exner, 1996). But all the surveys reported here are cross-sectional snapshots in time; some of the teens who did not identify as gay, lesbian, or bisexual may have done so at a later point in high school or even in early adult- hood. Thus, the findings primarily pertain to adolescents who be- gan to recognize their orientation and self-identify, even if they have not disclosed it. In addition, some of the measures of sexual orientation in these school-based surveys deal with only one di- mension of orientation—either attraction or behavior rather than self-label—and sometimes, may underestimate the number of LGB adolescents (Saewyc, Bauer, et al., 2004). The similarity of the find- ings across different surveys with different measures, however, lends support to the stability of these results. Another complication in this picture is the timing of physical and sexual abuse: before or after self-identification, during child- hood or during adolescence? For what percentage of teens was the abuse a direct, clear consequence of disclosing their orientation, or NOT FOR DISTRIBUTION 208 CHILD WELFARE • Vol. LXXXV, #2 • March/April a response to a teen's subtle behavioral cues to orientation out any overt disclosure? Because the surveys are cross-sectional, they cannot disentangle the complexity of timing and determine causality, such as a teen being abused because of her lesbian or bisexual status. Physical and sexual abuse can occur across a lifespan, but peak prevalence of maltreatment in the United States appears to be during adolescence (Snyder & Sickmund, 2000)— concurrent with the years of developing sexual orientation—and not limited to either early or later adolescence (Saewyc et al., 2003). That makes it impossible to determine which most likely came first: coming out or being abused. Sexual and physical abuse, however, clearly are not the cause of developing a gay, lesbian, or bisexual orientation, as earlier researchers sometimes considered (Simari & Baskin, 1982). In the seven surveys as in other popula- tion-based ones, the majority of adolescents who identify as gay, lesbian, or bisexual do not report any abuse, and the overwhelm- ing majority of adolescents who report sexual or physical abuse identify as heterosexual, because they outnumber LGB students so greatly. Although the stigma may lead to less protection for LGB teens, or a perception that they are more deserving of abuse, a potential alternate explanation for the higher prevalence of abuse among teens exists. Physical and sexual abuse also carries a burden of shame and stigma (Finkelhor & Browne, 1985). Sexual abuse, even more than physical abuse, carries a potent stigma, what Goffman (1968) describes as "spoiled identity" in his pioneering work on stigma: Society presents clear messages that those who have been sexually violated somehow are tainted or "damaged," no longer innocent and pure. As a result of their lost "innocence," victims of sexual assault or exploitation suffer a further loss of status and worth. For most teens, the likely costs of acknowledging a stigma- tized orientation may be too risky to recognize same-gender at- tractions, so they may suppress that knowledge. In contrast, abused teens, having already suffered loss of status and self-worth NOT FOR DISTRIBUTION Saewyc, et al. 209 owing to abuse, may be more open to recognizing and disclosing their same-gender or both-gender attractions because they have less to lose socially by admitting another stigmatized status. They still may risk further victimization after disclosure, and an addi- tive effect of multiple types of stigma may occur, which further increases their odds of substance abuse and other risky coping behaviors. This additive effect was suggested by a different analy- sis of the four cohorts from the Pacific Northwest, where the au- thors found an interaction between sexual minority status and abuse that predicted higher involvement in HIV risk behaviors such as rv drug use and risky sexual behaviors (Saewyc, Richens, Skay, Reis, Poon, & Murphy, in press). Regardless of the timing or reason for the abuse, the higher prevalence of abuse among LBG youth suggests they will be dis- proportionately found among runaway, homeless, juvenile jus- tice, and foster care populations, which a number of studies have already confirmed (Frankowski & the American Academy of Pe- diatrics Committee on Adolescence, 2004; Murphy, Poon, & Weigel, 2001). Child welfare and other youth services need to be aware of the added vulnerability of LGB youth and be prepared to provide orientation-sensitive and appropriate care among those populations. Education for social workers, foster parents, and others involved in child protection work should include training on the development of sexual orientation and gender identity among children and adolescents, identification and fostering of healthy environments for sexual minority teens, and boundary issues related to personal and religious values. The routine assessment for experiences of physical and sexual violence among all adolescents is necessary, and LGB youth are no exception. Professionals who work with the families of LGB youth in crisis should help them respond to sexual orientation disclosure in helpful, not harmful, ways. The American Psychi- atric Association, the American Psychological Association, the American Academy of Pediatrics, the National Association of NOT FOR DISTRIBUTION 210 CHILD WELFARE • Vol. LXXXV, #2 • March/April Social Workers, and other professional organizations have de- nounced as unethical therapy aimed at trying to change sexual orientation with children and adolescents, because the methods have the potential to cause great harm, and little credible evi- dence for their effectiveness exists (American Psychiatric Asso- ciation, 2000). Families need encouragement and resources that help them support their sexual minority child, rather than mes- sages further stigmatizing them. For families and youth concerned about how their religious community views sexual orientation, it may be useful to provide information about safe, open, and af- firming religious institutions within their local area. Individual and community interventions intended to prevent child abuse and family violence should incorporate messages that emotional pressure or corporal punishment will not change a child's most natural gender expression or sexual orientation. Orgaruzations and caseworkers must raise awareness about the unacceptable level of violence directed at LGB teens, whether in families or in the community, but they must do it in ways that work toward reducing that risk rather than unintentionally rein- forcing the expectation of violence from the people youth should look to for love and support. Schools, as a key institution for posi- tive youth development, have an important role in supporting LGB youth, as well as linking teens to supportive community resources when needed. Effective prevention of the violence and enacted stigma target- ing sexual minority teens will require further societal efforts to re- duce the stigma of LGB orientation. The Canadian Charter of Rights aftirms individuals' right to protection from discrimination based on characteristics such as race, sex, religion, or age; it does not mention orientation, but the courts have since ruled it also prohib- its discrimination based on other characteristics, including sexual orientation (de Bruyn, 2004). In the United States, several cities and states have incorporated similar language in their equal op- portunity and antidiscrimination laws. Legal efforts, however, must NOT FOR DISTRIBUTION Saewyc, et al. 211 be augmented with advocacy and interventions to increase respect for diversity and reduce community acceptance of violence toward those marginalized. When society truly values and nurtures all children, LGB teens will no longer face the added hazards from stigma in their families, foster care, school, and their community. References American Psychiatric Association (2000). Therapies focused on attempts to change sexual orientation (reparative or conversion therapies). Position statement. APA Document Reference No. 20001, May 2000. Accessed on April 21,2006, at www.psych.org/edu/ other_res / lib_archives / archives / 200001 .pdf. Anderson, C. M., Teicher, M. H., Polcari, A., & Renshaw, P. F. (2002). Abnormal T2 relaxation time in the cerebellar vermis of adults sexually abused in childhood: Potential role of the vermis in stress-enhanced risk for drug abuse. Psychoneuroendocrinology 27,231-244. Borowsky, I. W., Ireland, M., & Resnick, M. D, (2001). Adolescent suicide attempts: Risks and protectors. Pediatrics, 107, 485-493. D'Augelli, A. R., Hershberger, S. L., & Pilkington, N. W. (1998). Lesbian, gay, and bi- sexual youth and their families: Disclosure of sexual orientation and its consequences. American Journal of Orthopsychiatry, 68(3), 361-371. De Bellis, M. D. (2002). Developmental traumatology: A contributory mechanism for alcohol and substance use disorders. Psychoneuroendocrinology, 27,155-170. de Bruyn, T. (2004). A plan of action for Canada to reduce HIV/AIDS-related stigma and dis- crimination. Canadian HIV/AIDS Legal Network. Accessed March 2005 at www.aidslaw.ca. DuRant, R. H., Krowchuk, D. P., & Sinai, S. H. (1998). Victimization, use of violence, and drug use at school among male adolescents who engage in same-sex sexual behav- ior. Journal of Pediatrics, 133(1), 113-118. Federal Bureau of Investigations. (2004). Hate crime statistics 2003. Washington, DC: Fed- eral Bureau of Investigations, U.S. Department of Justice. Accessed April 4, 2006, at www.fbi.gov/ucr/ucr.htmffhate. Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry, 55, 530-540. NOT FOR DISTRIBUTION 212 CHILD WELFARE • Vol. LXXXV, #2 • March/April Frankowski, B.L., & the American Academy of Pediatrics Committee on Adolescence. (2004). Sexual orientation and adolescents. Pediatrics, 113,1827-1832. Accessed April 4, 2006, at www.pediatrics.org/cgi/content/full/113/6/1827. Goffman, E. (1968). Management of spoiled identity. In E. Rubington and M.S. Weinberg, (Eds.), Deviance: The interactionist perspective. Text and readings in the sociology of devi- ance (pp. 344-348). New York: The MacMillan Company. Hershberger, S. L., & D'Augelli, A. R. (1995). The impact of victimization on the mental health and suicidality of lesbian, gay, and bisexual youths. Developmental Psychology, 31, 65-74. Holmes, W.C. & Slap, G.B. (1998). Sexual abuse of boys: Definition, prevalence, correlates, sequelae, and management. Journal of the American Medical Association, 280,1855-1862. Horliss, H. L., Cochran, S. D., & Mays, V. M. (2002). Reports of parental maltreatment during childhood in a United States population-based survey of homosexual, bi- sexual, and heterosexual adults. Child Abuse & Neglect, 26,1165-1178. Mays, V. M., & Cochran, S. D. (2001). Mental health correlates of perceived discrimina- tion among lesbian, gay, and bisexual adults in the United States. American Journal of Public Health, 91,1869-1876. Murphy, A., Poon, C , Weigel, M. (2001). No place to call home: A profile of street youth in British Columbia. Burnaby, BC, Canada: The McCreary Centre Society Accessed April 4, 2006, at www.mcs.bc.ca. Murphy, A., Sidhu, A., & Tonkin, R. (1999). Being out: Lesbian, gay, bisexual & transgender youth in BC, an adolescent health survey. Burnaby, BC, Canada: The McCreary Centre Society. Assessed April 4, 2006, at www.mcs.bc.ca. Pilkington, N. W., & D'Augelli, A. R. (1995). Victimization of gay, lesbian, and bisexual youth in community settings. Developmental Psychology, 23, 33-56. Reis E., & Saewyc, E. (1999). 83,000 Youth: Selected findings of eight population-based studies as they pertain to anti-gay harassment and the safety and well-being of sexual minority students. Safe Schools Coalition of Washington, May 1999. Accessed April 4, 2006, www.safeschoolscoalition.org. Robin, L., Brener, N. D., Donahue, S. R, Hack, T., Hale, K., & Goodenow, C. (2002). Asso- ciations between health risk behaviors and opposite-, same-, and both-sex sexual partners in representative samples of Vermont and Massachusetts high school stu- dents. Archives ofPediatric and Adolescent Medicine, 156, 349-355. NOT FOR DISTRIBUTION Saewyc, et al. 213 Rosario, M., Meyer-Bahlburg, H.R, Hunter, J., & Exner, T.M. (1996). The psycho-sexual development of urban lesbian, gay, and bisexual youths. Journal of Sex Research, 33, 113-126. Rotheram-Borus, M. J., Marelich, W. D., & Srinivasan, S. (1999). HIV risk among homo- sexual, bisexual, and heterosexual male and female youths. Archives of Sexual Behav- ior, 28,159-177. Russell, S. X, Franz, B. T, & Driscoll, A. K. (2001) Same-sex romantic attraction and experiences of violence in adolescence. American Journal of Public Health, 91,903-906. Russell, S. T., & Joyner, K. (2001). Adolescent sexual orientation and suicide risk: Evi- dence from a national study. American Journal of Public Health, 91,1276-1281. Saewyc, E. M., Bauer, G. R., Skay, C. L., Bearinger, L. H., Resnick, M. D., Reis, E., Murphy, A. (2004). Measuring sexual orientation in adolescent health surveys: Evaluation of eight school-based surveys. Journal of Adolescent Health, 35, 345e.l-e.l6. Saewyc, E. M., Bearinger, L. H., Blum, R. W., & Resnick, M. D. (1999). Sexual intercourse, abuse, and pregnancy among adolescent women: Does sexual orientation make a difference? Family Planning Perspectives, 31(3), 127-131. Saewyc, E. M., Magee, L. L., & Pettingell, S. (2004). Teenage pregnancy and associated risk behaviors among sexually abused adolescents. Perspective on Sexual and Repro- ductive Health, 36(3), 98-105. Saewyc, E. M., Pettingell, S., & Magee, L. L. (2003). The prevalence of sexual abuse among adolescents in school. Journal of School Nursing, 19, 266-272. Saewyc, E. M., Skay, C. L., Bearinger, L. H., Blum, R. W., & Resnick, M. D. (1998). Sexual orientation, sexual behaviors, and pregnancy among American Indian adolescents. Journal of Adolescent Health, 23(4), 238-247. Snyder, H., & Sickmund, M. (2000). Sexual assault of young children as reported to law en- forcement: Victim, incident, and offender characteristics. July 2000. National Centre for Juvenile Justice. Accessed April 4, 2006, at www.unh.edu/ccrc/factsheet.htmW3. Waldo, C. R., Hesson-Mclnnis, M. S., & D'Augelli, A. R. (1998). Antecedents and conse- quences of victimization of lesbian, gay, and bisexual young people: A structural model comparing rural university and urban samples. American Journal of Commu- nity Psychology, 26, 307-334. NOT FOR DISTRIBUTION NOT FOR DISTRIBUTION