Objective: To measure associations of childhood sexual abuse (CSA) with sexual orientation, behaviors, and attractions and HIV/sexually transmitted infection (STI) incidence in a nationally representative sample of men and women.
Methods: Data from the 2004–2005 Wave 2 of the National Epidemiologic Survey on Alcohol and Related Conditions were analyzed, including frequencies of CSA and HIV/STI incidence for 5 subgroups defined by sexual orientation based on identity and behaviors and attraction to the same sex or opposite sex.
Results: Overall, 14.9% of women and 5.2% of men reported CSA. Among women, bisexuals, lesbians, and heterosexuals with same-sex partners had 5.3 times, 3.4 times, and 2.9 times the odds, respectively, for CSA occurring sometimes/more frequently (vs. never) compared with heterosexuals not having same-sex partners or attractions. Among men, bisexuals, gay men, and heterosexuals with same-sex partners had 12.8 times, 9.5 times, and 7.9 times the odds, respectively, for CSA. Men and women sometimes or frequently abused had significant increases in odds for HIV/STI incidence compared with those not abused. Among women, sexual minorities had 3.8 times the odds and heterosexuals had 2.8 times the odds, whereas among men, sexual minorities had 4.2 times odds and heterosexuals had 1.5 times odds.
Conclusions: Extraordinarily high rates of CSA were observed for sexual minorities, and sexual minorities were more likely to have incident HIV or STIs, in this U.S. population survey. Identifying the impact of CSA among heterosexuals and sexual minorities in the US is a crucial first step in examining the sequelae of CSA, including the potential mediators of mental health and substance abuse disorders in the relationship between CSA and sexual risk taking.
Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, PA.
Correspondence to: Thersa Sweet, PhD, MPH, Department of Epidemiology and Biostatistics, Mail Stop 1033, Bellet Building, 6th Floor, 1505 Race Street, Philadelphia, PA 19145 (e-mail: firstname.lastname@example.org).
Supported by the National Institutes of Health.
Presented at the 3rd North American Congress of Epidemiology, June 21–24, 2011, Montreal, Quebec, Canada.
The authors have no conflicts of interest to disclose.
Received July 7, 2011
Accepted October 28, 2011
Previous research indicates that lesbian, gay, and bisexual (LGB) individuals have a higher risk of childhood sexual abuse (CSA) than their heterosexual counterparts. Numerous studies have shown high prevalence rates of self-reported CSA, among LGB individuals, ranging from about 15% to nearly 50%, depending on a number of factors including definitions of both sexual orientation and CSA.1–11
Although these previous studies have been invaluable in identifying higher rates of CSA among LGB individuals, there are some limitations. Many of the previous studies have not had a large enough sample size to analyze the differences among LGB persons, in particular, the difference between gay/lesbian and bisexual individuals. Most studies have also only contained data on men or women, not both. Furthermore, many of the studies recruited study participants from convenience samples through sexually transmitted infection (STI) clinics, LGBT Pride events, or via self-selection in response to advertisements in LGB media and have not utilized heterosexual subgroups as referents. In the absence of heterosexual reference groups, comparisons between heterosexual and LGB populations were not made, or statistics from previous work describing the prevalence of CSA among heterosexuals were adapted for comparison.
There is general consensus that all people, regardless of sexual orientation who have been sexually abused as children, are much more likely to engage in high-risk sexual behaviors as adults and adolescents12,13. High-risk sexual behavior puts a person at increased risk for acquisition of STIs, including HIV. CSA has also been shown to increase the risk of alcohol dependence14–16 and illegal drug use,16,17 which may mediate or modify the relationship between CSA and subsequent risky sexual behaviors.18,19 However, the majority of these studies have only been done with women. Those that have focused on men have not included data on sexual orientation or have primarily targeted self-identified homosexual or bisexual men.
Our current population-based study of sexuality and CSA utilizes data from the national epidemiologic survey on alcohol and related conditions (NESARC), a large longitudinal survey designed to analyze the mental health status and substance use patterns of individuals living in the United States.20,21 Wave 2 of the study contained specific questions about sexual identity, behaviors and attractions, and detailed questions about previous CSA, and HIV, and STI incidence.
Use of data from NESARC, along with survey sampling weights, enabled us to estimate US population-level rates of LGB sexual identity, rates of CSA, and HIV and STI incidence for these groups in comparison with heterosexuals. We theorized that men and women who identify as LGB or have same-sex attraction and/or behavior would report higher prevalences of CSA compared with heterosexuals who report neither same-sex partners nor attraction. However, this should in no way be interpreted that CSA is a cause of developing a LGB sexual orientation, rather LGB individuals as children are at a higher risk of experiencing CSA. We also expect that sexual minority men and women in the United States will have higher rates of HIV/STI incidence than their heterosexual peers among those who have experienced CSA and those who have not.
We utilized the data from Wave 2 of NESARC. NESARC is a nation-wide household survey administered by NIAAA. Data were collected from individuals in computer assisted face-to-face interviews conducted in participants' homes in 2 waves; wave 1 from 2001 to 2002 and wave 2 from 2004 to 2005. The population comprised civilian, noninstitutionalized adults living in the United States. The NESARC study design involved multistage sampling, with stratification and clustering of the study population. Additionally, blacks and Hispanics were oversampled as were young adults in the age group 18–24. The NESARC study design is further described elsewhere.20–22
Of the 39,959 individuals who completed wave 1 of NESARC and were eligible for Wave 2, 34,653 participated in wave 2, for a response rate of 86.7%. The response rate for wave 1 was 81%, giving an overall response rate of 70.2% for wave 2.
To be eligible for inclusion in the current study, people had to provide information on sexual orientation, attraction, and behaviors with men and/or women. Additionally, people needed to answer questions concerning CSA. If participants provided no answer to 1 or more of the 4 questions concerning CSA, yet provided a response of “sometimes” or greater to any of the remaining questions, they were included in the study. Thirty-three thousand nine hundred two (97.8% of wave 2) were included in the present study. The Drexel University Institutional Review Board approved the current study.
Data Included for Analyses
Childhood Sexual Abuse
CSA was defined as sexual experiences with an adult or any other person younger than 18 years when the individual did not want the sexual experience or was too young to know what was happening. Specifically, survey items measured occurrences of the following: (1) being touched or fondled in a sexual way; (2) being forced to touch another in a sexual way; (3) sexual intercourse; or, (4) intercourse was attempted. For each of these, frequencies included the following: never, almost never, sometimes, fairly often, very often, or unknown. We collapsed the CSA data for all analyses into 3 groups: never, almost never, and frequent/sometimes.
HIV/STI Incidence was measured using the responses to questions about HIV or STI infections diagnosed and confirmed by a health professional in the 12 months before participation in wave 2.
Sexual orientation was classified based on self-identity (heterosexual, gay or lesbian, bisexual, or not sure), sexual behavior (had sex with only males, only females, both, or never had sex), and sexual attraction (only attracted to females or only attracted to males, mostly attracted to females or mostly attracted to males, or equally attracted to females and males). From these data fields, we classified sexuality into 5 groups as follows: gay/lesbian, bisexual, heterosexual with no same-sex sexual partners or attraction, heterosexual with some same-sex attraction but no same-sex partners, and heterosexual with at least some same-sex sex partners. To examine the association between HIV/STI incidence and CSA, we further collapsed sexual orientation into 2 groups. The referent group consisted of heterosexual by identity and behavior. The sexual minority group included gay/lesbian, bisexual, and heterosexual with some same-sex behavior. This was done to preserve statistical power as the rates of HIV/STI incidence are low in the United States.
Demographic data collected during wave 2 included race/ethnicity (white, non-Hispanic; black, non-Hispanic; American Indian/Alaska Native, non-Hispanic; Asian/Native Hawaiian/other Pacific Islander, non-Hispanic; or Hispanic, any race), age at the time of the interview (20–30, 31–40, 41–50, 51–60, >60), region (Northeast, Midwest, South, West), educational level (high school or less, some college, completed 4-year degree, or at least some grad school), annual household income (<$15,000, $15,000–24,999, $25,000–34,999, $35,000–49,999, $>50,000), and marriage status (married/living together, never, divorced/separated, or widowed).
All data were weighted by NESARC wave 2 weights, and all analyses were performed using survey procedures with SAS 9.1 software, which accounted for the complex study design. “We specifically used both proc surveyfreq and proc surveylogistic with the strata, cluster, and weight variables, defined by the NESARC dataset”.
To identify potential confounders in the relationship between sexual orientation and CSA, we conducted contingency table analysis of demographic variables by CSA, sexual orientation, and HIV/STI incidence, separately, for men and women. Rao–Scott χ2 tests were utilized to test for association, compensating for the survey design.
To determine the magnitudes of the association between CSA and sexual orientation, CSA was regressed on sexual orientation and all potential confounding demographic variables using the never group of CSA and heterosexual with no same-sex behaviors or attractions as the referents. Men were analyzed separately from women. Final models were constructed using backward elimination of potential confounding variables, which did not alter the regression coefficients of the full model in 2 or more categories by at least 10%. The same strategy was utilized to determine the association between HIV/STI incidence and CSA, regressing HIV/STI incidence on CSA.
Self-Reported Sexual Identity and Demographic Correlates
All proportions are weighted results. Among women, 0.6% identified as lesbian, 0.8% identified as bisexual, 1.5% identified as heterosexual, but with some same-sex partners, 4.6% identified as heterosexual, with some same-sex attraction, and 92.5% identified as heterosexual, with no same-sex partners or attraction (Table 1).
For men, 1.1% identified as homosexual, 0.4% identified as bisexual, 2.4% identified as heterosexual with some same-sex partners, 1.8% identified as heterosexual with some same-sex attraction, and 94.3% identified as heterosexual with no same-sex partners or attraction (Table 2).
All associations of demographics with sexual identity were statistically significant (P < 0.0001) (Tables 1, 2) due to the very large sample size of the NESARC survey. However, only age at the time of the interview was a confounding variable in the relationship between CSA and sexual orientation, for women. Lesbians, bisexuals, and women who identified as heterosexual with at least some same-sex partners were younger (79.2%, 83.5%, 68.7%, respectively, <51 years of age) compared with heterosexual women with no same-sex attractions or partners (57% <51 years of age).
There were no demographic variables found to be confounding in the association between CSA and sexual orientation for men at the highest level of CSA, yet for men almost never abused compared with those never abused, age, ethnicity/race, education level, and marriage status were all confounders. Similar associations of demographics with sexual identity were seen for men (Table 2).
Among all women, 7.2% reported CSA having occurred sometimes or frequently and 7.7% reported CSA having occurred almost never (Table 3). A much larger proportion sexual minority women reported CSA compared with heterosexual women, with 43.5% of bisexuals, 38.1% of lesbians, and 28% of heterosexuals with same-sex partners reporting any CSA compared with 14.2% of heterosexual women reporting any CSA (Fig. 1).
Fewer men reported any CSA compared with women as follows: 1.8% reported having been abused sometimes or more frequently and 3.5% reported being abused almost never (Table 4). Although the overall rates of CSA were higher for women than for men, the trends among the different sexual orientation groups were similar, with sexual minority men reporting more CSA than heterosexual men, as 19% bisexuals, 18.6% homosexuals, 19.4% heterosexual with same-sex partners, and 4.6% heterosexuals reported any CSA (Fig. 1).
The relative odds of CSA were markedly higher for bisexual, lesbian, and heterosexual women with some same-sex partners and/or attractions compared with heterosexual women with neither (Table 5). Bisexual women had the highest relative odds of being sometimes or more frequently abused when compared with the heterosexual referent group, with 5.3 times odds. Similarly lesbians, heterosexual women with same-sex partners, and heterosexual women with same-sex attraction had 3.4 times, 2.9 times, and 1.6 times the odds of reporting sometimes or more frequent abuse compared with heterosexual women without same sex partners or attraction. Women almost never abused had similar trends in odds, although the magnitudes were smaller (Table 5).
The adjusted associations of sexual identity with reported CSA were higher for men than was reported for women (Table 5). For men, none of the covariates were confounding for the model comparing sometimes or frequent CSA to never, whereas 4 covariates were included in the final model comparing almost never CSA to none as follows: age, race, education level, and marriage status. Bisexual men had the highest relative odds of being sometimes or more frequently sexually abused, with 12.8-times odds compared with the heterosexual referent group. Similarly, gay and heterosexual men with same-sex partners had 9.5 times and 7.9 times the odds of reporting sometimes or more frequent abused compared with heterosexual men with no same-sex partners or attraction.
Nearly all associations of demographics with HIV/STI incidence were statistically significant. The only demographic that did not vary among those with incidence HIV/STI and those without was education in heterosexual women and was not evaluated as a potential confounder. For heterosexual women, age and marriage status were confounders; for sexual minority women, age, race, and region were confounders. For heterosexual men, age and income were confounders; for sexual minority men, race, region, income, education, and marriage status were confounders (Table 6).
Women who were sexually abused as children had an increased risk of HIV/STI incidence compared with those who were not abused for all levels of abuse (Table 6). Heterosexual women almost never abused were 1.6 times as likely to have incident HIV or STI, whereas those sometimes or frequently abused were 2.8 times as likely compared with those never abused. Sexual minority women almost never abused were 7.1 times as likely and those sometimes/frequently abused were 3.8 times as likely to have incident HIV or STI compared with those never abused.
Both heterosexual men and sexual minority men who were almost never abused did not have a significant increase in risk of HIV/STI incidence. However, heterosexual men sometimes or frequently abused had 1.5-fold risk [odds ratio (OR) = 1.54, 95% confidence interval (CI) = 1.34 to 1.78] and sexual minority men had 4.2-fold risk (OR = 4.23, 95% CI = 2.36 to 7.59).
This is perhaps the first nationally representative study to examine the relationship between sexual orientation and CSA in both men and women, using contemporaneous heterosexual populations as referent groups. We found that individuals who identify themselves as gay, bisexual, or heterosexual with some same-sex partners have sharply increased risks of experiencing CSA than heterosexuals who report neither same-sex partners nor attraction. This is in agreement with previous studies which have shown high prevalence rates of CSA among gay and bisexual individuals1–7,9–11,23,24. Many of these studies have used convenience samples of LGB individuals without a heterosexual referent group. The large diverse NESARC study population enabled us to study differences in risk of CSA by sex and sexual orientation. With the use of survey statistical methods, these results are applicable to the general US population and adjusted for confounding demographic variables to reveal less biased estimates of risk.
What is particularly striking is that of all subgroups, men and women who identified as bisexual had the highest odds for frequent or sometimes CSA compared with the referent heterosexual group. In comparison with gay men and lesbians, the odds for having been sometimes or frequently sexually abused for bisexual men and women were 1.3 times and 1.6 times, respectively, the odds observed for gays or lesbians. Our estimates of CSA among gay and bisexual men are similar to previous work from Arreola et al25,26 and higher than previous studies examining abuse in lesbians and bisexual women.4
Additionally, heterosexual men and women who reported same-sex partners had much higher rates of CSA compared with heterosexuals having no same-sex partners or attraction; in our study, the risks of CSA occurring sometimes or more frequently in these subgroups of men and women were roughly equivalent to what we estimated for gay men and lesbians. This group of individuals has only recently been examined in terms of sexual risk-taking and CSA because most studies characterize people based on either sexual behavior or sexual identity, not both. Yet, a high degree of discordance between sexual behavior and sexual identity has previously been reported.27,28 The importance of including this group of self identified heterosexuals with same-sex sexual practices is underscored by the fact that in previous studies, self identified heterosexual men who had sex with men were more likely to have had STIs compared with heterosexual men having no same-sex partners28,29 and were less likely to use condoms during sex than homosexual men.28,30 Among women, studies have found that “mostly heterosexual” women are more likely to engage in risky sexual behaviors than heterosexual women31,32.
The data presented here must be analyzed cautiously so that CSA is not identified as causally related to individuals becoming gay/lesbian or bisexual as adults. A more likely explanation is that as children some of these individuals may have declared their sexual preference or exhibited subtle behavioral cues that identified them as more vulnerable and thus targets for sexual abuse by predators. In fact, previous studies have reported that many LGB adults remember being “gender atypical” as children and reported physical and sexual abuse at the hands of peers and family members because of this difference.33,34
Men and women sometimes or frequently abused had significant increases in risk for HIV/STI incidence compared with those not abused. However, men almost never abused did not have an increase in risk of incident HIV/STI, although women almost never abused did. One possible reason for this discrepancy is that boys and men are less likely to report CSA than girls or women, and this misclassification could have reduced the risk of HIV/STI incidence among men who were abused. Underreporting of CSA by males in particular has been described in a recent report which found that boys take longer to disclose CSA at the time of the abuse than girls and, as adults, men wait much longer to discuss the experience than women.35
Although the coding of sexual abuse did not differentiate between the different forms of abuse, this grouping was necessary to provide adequate levels of exposure for analysis. The effect of including less severe forms of sexual abuse would be to underestimate effect size (if men with less severe abuse were more like those who were not abused) because the abuse group would be heterogeneous with respect to more and less severe forms of abuse.
Although sexual minority women almost never abused had an increased risk over heterosexual women, those sometimes or frequently abused did not. It seems that for women sometimes or frequently abused, sexual minority status does not add to the risk of HIV/STI incidence, yet the same is not true for sexual minority men, among which, those who were sometimes or frequently abused have a higher risk than heterosexual men. Further research should be performed to determine the different effects of CSA on men and women, looking at potential mediators in the pathway between CSA and HIV risk, which may be different for men and women.
One limitation to this study was underreporting of being gay, lesbian, or bisexual. For the National Survey of Family Growth, administered in 2002, questions concerning sexual behavior were self-entered by the respondent into a computer. The rates of homosexuality and bisexuality for both women (1.3% and 2.8%, respectively) and men (2.3% and 1.8%, respectively)36 were higher than the current study. However, in the current study, 68% of the individuals who did not give information on sexual orientation also did not give information on CSA, and another 31% reported that they had never been sexually abused (data not shown). So, although the rates of minority sexual orientation and CSA could have been lower than accepted national numbers, the results showing the overall impact of sexual orientation on CSA are likely to be underestimates of the actual risks in these populations.
Another potential source of misclassification is that sexual minorities may be have been somewhat more likely to report child sexual abuse as they already feel like “outsiders” from the mainstream population and therefore are more comfortable describing uncomfortable events that occurred in their past. However, one of the highest risk groups was heterosexuals with some same-sex partners. These individuals do not consider themselves a sexual minority, yet they have a high risk of CSA. Misclassification of sexual orientation also could have occurred if abuse by a same-sex person during childhood was described as a sexual experience. However, we would not expect such a strong association between CSA and sexual minorities by such misclassification.
A limitation to the analyses is the use of ORs. The SAS surveylogistic procedure contains statements to specify the design strata, clusters, and weights, which are necessary to correctly estimate the variances of the estimators. Because ORs can overestimate relative risks, we compared the crude ORs to the relative risks for the weighted frequencies. We found that for men, the relative risk estimates were not significantly different from the ORs as they all fell within the 95% CIs of the ORs and for women, the ORs overestimated the relative risks by 28%, on average. The use of the SAS survey procedures enabled us to determine the association between CSA and sexual orientation using this data, allowing for control of confounding. HIV/STI incidence is rare in all 4 populations, ranging from 0.5% to 3.5%. As would be expected, the weighted ORs were virtually identical to the estimates of the relative risks, therefore, we are confident that the disease ORs for HIV/STI are not overestimates of the relative risks.
This national survey adds to important findings reported in previous studies. With the large sample size, we are able to perform separate analyses by sex and utilized detailed definitions of sexual orientation, involving sexual identity, attraction, and practices. Our data show that CSA is reported more often by sexual minorities, which places them at an increased risk of HIV/STI. Identifying the impact of CSA among both heterosexuals and sexual minorities in the United States is a crucial first step in examining the sequelae of CSA, including the potential mediators of mental health and substance abuse disorders in the relationship between CSA and sexual risk taking.
We would like to thank Dr Randall Sell for his valuable insights on the measurement of sexuality and Dr Jana Mossey (both at the Drexel University School of Public Health) for her assistance in working with NESARC data.
1. Austin SB, Jun HJ, Jackson B, et al.. Disparities in child abuse victimization in lesbian, bisexual, and heterosexual women in the Nurses' Health Study II. J Womens Health (Larchmt). 2008;17:597–606.
2. Balsam KF, Rothblum ED, Beauchaine TP. Victimization over the life span: a comparison of lesbian, gay, bisexual, and heterosexual siblings. J Consult Clin Psychol. 2005;73:477–487.
3. Brennan DJ, Hellerstedt WL, Ross MW, et al.. History of childhood sexual abuse and HIV risk behaviors in homosexual and bisexual men. Am J Public Health. 2007;97:1107–1112.
4. Corliss HL, Cochran SD, Mays VM, et al.. Age of minority sexual orientation development and risk of childhood maltreatment and suicide attempts in women. Am J Orthopsychiatry. 2009;79:511–521.
5. Doll LS, Joy D, Bartholow BN, et al.. Self-reported childhood and adolescent sexual abuse among adult homosexual bisexual men. Child Abuse Negl. 1992;16:855–864.
6. Lenderking WR, Wold C, Mayer KH, et al.. Childhood sexual abuse among homosexual men. Prevalence and association with unsafe sex. J Gen Intern Med. 1997;12:250–253.
7. Paul JP, Catania J, Pollack L, et al.. Understanding childhood sexual abuse as a predictor of sexual risk-taking among men who have sex with men: the Urban Men's Health Study. Child Abuse Negl. 2001;25:557–584.
8. Samuel J, Paul JP, Stall R, et al.. Childhood sexual abuse and HIV risk-taking behavior among gay and bisexual men. AIDS Behav. 1998;2:41–51.
9. Tjaden P, Thoennes N, Allison CJ. Comparing violence over the life span in samples of same-sex and opposite-sex cohabitants. Violence Vict. 1999;14:413–425.
10. Tomeo ME, Templer DI, Anderson S, et al.. Comparative data of childhood and adolescence molestation in heterosexual and homosexual persons. Arch Sex Behav. 2001;30:535–541.
11. Welles SL, Baker AC, Miner MH, et al.. History of childhood sexual abuse and unsafe anal intercourse in a 6-city study of HIV-positive men who have sex with men. Am J Public Health. 2009;99:1079–1086.
12. Senn TE, Carey MP, Vanable PA. Childhood and adolescent sexual abuse and subsequent sexual risk behavior: evidence from controlled studies, methodological critique, and suggestions for research. Clin Psychol Rev. 2008;28:711–735.
13. Greenberg JB. Childhood sexual abuse and sexually transmitted diseases in adults: a review of and implications for STD/HIV programmes. Int J STD AIDS. 2001;12:777–783.
14. Dinwiddie S, Heath AC, Dunne MP, et al.. Early sexual abuse and lifetime psychopathology: a co-twin-control study. Psychol Med. 2000;30:41–52.
15. Nelson EC, Heath AC, Madden PA, et al.. Association between self-reported childhood sexual abuse and adverse psychosocial outcomes: results from a twin study. Arch Gen Psychiatry. 2002;59:139–145.
16. Kendler KS, Bulik CM, Silberg J, et al.. Childhood sexual abuse and adult psychiatric and substance use disorders in women: an epidemiological and cotwin control analysis. Arch Gen Psychiatry. 2000;57:953–959.
17. Nelson EC, Heath AC, Lynskey MT, et al.. Childhood sexual abuse and risks for licit and illicit drug-related outcomes: a twin study. Psychol Med. 2006;36:1473–1483.
18. NIMH Multisite HIV Prevention Trial Group. A test of factors mediating the relationship between unwanted sexual activity during childhood and risky sexual practices among women enrolled in the NIMH Multisite HIV Prevention Trial. Women Health. 2001;33:163–180.
19. Miller M. A model to explain the relationship between sexual abuse and HIV risk among women. AIDS Care. 1999;11:3–20.
20. Grant BF, Moore TC, Shepard J, et al.. Source and Accuracy Statement: Wave 1 National Epidemiologic Survey on Alcohol and Related Conditions NESARC. Bethesda, MD: NIAAA; 2004.
21. Grant BF, Dawson DA. Introduction to the national epidemiologic survey on alcohol and related conditions. Alcohol Res Health. 2006;29:74–78.
22. Grant BF, Kaplan KD. Source and Accuracy Statement for the Wave 2 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Rockville, MD: NIAAA; 2005.
23. Jinich S, Paul JP, Stall RD, et al.. Childhood sexual abuse and HIV risk-taking behavior among gay and bisexual men. AIDS Behav. 1998;2:41–51.
24. Saewyc EM, Skay CL, Pettingell SL, et al.. Hazards of stigma: the sexual and physical abuse of gay, lesbian, and bisexual adolescents in the United States and Canada. Child Welfare. 2006;85:195–213.
25. Arreola SG, Neilands TB, Diaz R. Childhood sexual abuse and the sociocultural context of sexual risk among adult Latino gay and bisexual men. Am J Public Health. 2009;99(suppl 2)S432–S438.
26. Arreola S, Neilands T, Pollack L, et al.. Childhood sexual experiences and adult health sequelae among gay and bisexual men: defining childhood sexual abuse. J Sex Res. 2008;45:246–252.
27. Ross MW, Essien EJ, Williams ML, et al.. Concordance between sexual behavior and sexual identity in street outreach samples of four racial/ethnic groups. Sex Transm Dis. 2003;30:110–113.
28. Pathela P, Hajat A, Schillinger J, et al.. Discordance between sexual behavior and self-reported sexual identity: a population-based survey of New York City men. Ann Intern Med. 2006;145:416–425.
29. Zellner JA, Martinez-Donate AP, Sanudo F, et al.. The interaction of sexual identity with sexual behavior and its influence on HIV risk among latino men: results of a community survey in northern San Diego County, California. Am J Public Health. 2009;99:125–132.
30. Centers for Disease Control and Prevention (CDC). Condom use and sexual identity among men who have sex with men–Dallas, 1991. MMWR Morb Mortal Wkly Rep. 1993;42:7, 13–14.
31. Austin SB, Roberts AL, Corliss HL, et al.. Sexual violence victimization history and sexual risk indicators in a community-based urban cohort of mostly heterosexual and heterosexual young women. Am J Public Health. 2008;98:1015–1020.
32. Corliss HL, Austin SB, Roberts AL, et al.. Sexual risk in mostly heterosexual young women: influence of social support and caregiver mental health. J Womens Health (Larchmt). 2009;18:2005–2010.
33. D'Augelli AR, Grossman AH, Starks MT. Childhood gender atypicality, victimization, and PTSD among lesbian, gay, and bisexual youth. J Interpers Violence. 2006;21:1462–1482.
34. Waldo CR, Hesson-McInnis MS, D'Augelli AR. Antecedents and consequences of victimization of lesbian, gay, and bisexual young people: a structural model comparing rural university and urban samples. Am J Community Psychol. 1998;26:307–334.
35. O'Leary PJ, Barber J. Gender differences in silencing following childhood sexual abuse. J Child Sex Abus. 2008;17:133–143.
36. Mosher WD, Chandra A, Jones J. Sexual behavior and selected health measures: men and women 15–44 years of age, United States, 2002. Adv Data. 2005;362:1–55.
Keywords:© 2012 Lippincott Williams & Wilkins, Inc.
childhood sexual abuse; HIV risk; NESARC; sexual orientation