|Home | About | Journals | Submit | Contact Us | Français|
To test hypothesized relationships among childhood sexual abuse, social discrimination, psychological distress, and HIV-risk among Latino gay and bisexual men in the United States.
Data were from a probability sample of 912 men from Miami, Los Angeles, and New York. Logistic regression and path analyses were used to examine direct and indirect effects of childhood sexual abuse on psychological distress and sexual risk behavior.
The study revealed a 15.8% (95% CI = 12.3–19.2) prevalence of childhood sexual abuse. No sex and consensual sex before age 16 did not differ from each other in predicting any of the outcomes of interest. Forced sex was associated with a significantly increased risk for all outcomes. A path analyses yielded direct effects of childhood sexual abuse and exposure to both child and adult homophobia on psychological distress and indirect effects on risky sexual behavior.
HIV risk patterns among Latino gay and bisexual men are related to childhood sexual abuse and a social context of discrimination that lead to symptoms of psychological distress and participation in risky sexual situations that increase risky sexual behaviors associated with HIV-acquisition.
Latino gay/bisexual men comprise one of the most vulnerable groups in the United States for HIV-infection. Although White men who have sex with men (MSM, of whom most are gay/bisexual) continue to have the highest absolute numbers of HIV-infections, Latino MSM have higher prevalence and incidence rates of HIV.(1–4) In 2004, estimated Latino HIV/AIDS rates (29.5 per 100,000) were 3.3 times higher than for whites (9.0 per 100,000).(4),(5)
Despite intentions to practice safe sex and high levels of knowledge about HIV and safe sex, Latino gay/bisexual men engage in some of the highest levels of risky sexual behavior.(6, 7) Investigators attempting to explain these findings have examined possible determinants of increased risky sexual behavior and found that childhood sexual abuse increases HIV-risk;(8)–(9) however, we know very little about the mechanisms that explain this association.
Prevalence of childhood sexual abuse varies across studies, likely due to discrepant operationalizations: some studies deem all childhood sexual experiences with someone older as abusive; others incorporate the subjective experience of coercion as a criterion. Recent research indicates that MSM who have no sex or consensual sex before age 18 do not differ from each other in psychological outcomes, whereas those who experience forced sex have significantly more psychological distress.(10) Nevertheless, independent of measurement criteria, trends in prevalence estimates of childhood sexual abuse remain consistent. Although childhood sexual abuse is likely under-reported in boys, its prevalence, especially at earlier ages and in severe cases, approaches that of girls.(11)–(12) Furthermore, prevalence of childhood sexual abuse is higher among gay/bisexual men when compared with heterosexual men.(13–15)
As high as childhood sexual abuse prevalence is among gay/bisexual men (17% to 39%),(8),(9),(16)–(17)it is even higher among Latino gay/bisexual men.(8) In a national probability survey of adult MSM, a significantly higher proportion of Latino MSM reported sexual abuse before age 13 (22%) than did non-Latino MSM (11%).(18) Additionally, compared with non-Latinos, Latino boys are more likely to have experienced the following: sexual abuse by an extended family member; more genital fondling; exposure to more sexually abusive behaviors; and more anal abuse.(19),(20) Given the prevalence and severity of childhood sexual abuse among Latino gay/bisexual men and its association with HIV infection, it is essential that we examine the mediators of these associations in the context of Latino gay/bisexual men's lives.
Studies show that childhood sexual abuse significantly predicts negative health outcomes including HIV-risk and mental health issues such as depression, suicidal ideation, and substance abuse among adults.(16)–(17),(21)–(22, 23) A probability sample of Latinogay/bisexual men found a high prevalence of psychological distress during the six months before the interview, including, anxiety, depression, and suicidal ideation. Furthermore, experiences of homophobia strongly predicted psychological distress(24). These findings were specific to Latino gay/bisexual men illustrating the importance of examining the impact of contextual factors such as homophobia experiences, on individually focused outcomes, such as psychological distress. They also provide a foundation for examining the psychological and contextual factors that may explain the relationship between childhood sexual abuse and HIV-risk among Latino gay/bisexual men.
The current study used path analysisto test a model containing hypothesized relationships among self-reported childhood sexual abuse, experiences of homophobia, symptoms of psychological distress, and risk for HIV among Latino gay/bisexual men in the United States. Based on previous studies, we hypothesized that 1) no sex and consensual sex before age 16 would not differ from each other in predicting any of the outcomes of interest, 2) experiences of homophobia would be associated with significantly increased risk for negative psychological, situational and behavioral outcomes, and 3) experiences of forced sex would be associated with significantly increased risk for negative psychological, situational and behavioral outcomes.
Details of the procedures for this study have been described elsewhere(24). In summary, the sampling, recruitment, and interviewing procedures for this study occurred between October 1998 and March 1999 as part of a multisite study of self-identified Latino gay/bisexual men in the United States. A probability sample of 912 men was drawn from men entering social venues (bars, clubs, and weeknight events primarily attended by Latinos and gay men) in the cities of New York (n=309), Miami (n=302), and Los Angeles (n=301). The study protocol was approved by the Institutional Review Board of the University of California, San Francisco.
All measures were based on responses to close-ended surveys individually administered face-to-face.
Consistent with previous findings, we subdivided childhood sexual experiences into three discrete categories: none, consensual and forced sex.(18) Participants were asked if they had had a sexual experience before age 16 with someone five or more years older. If he answered “no,” he was categorized as “none”; if he answered “yes,” he was asked if it was against his will. Those who answered “no” to the latter question were categorized as “consensual,” whereas those who answered “yes,” were categorized as “forced”. For this study, childhood sexual abuse was defined as forced sex (“against your will”) before age 16 with someone five or more years older.
Consistent with previous measures,(24) five items measured respondents' exposure to childhood homophobia (e.g., “As you were growing up, how often did you feel that your homosexuality hurt and embarrassed your family?”); two of these items focused on experiences of verbal harassment and physical assaults in relation to both perceived sexual orientation and gender nonconformity (e.g., “As you were growing up, how often were you hit or beaten up for being homosexual or effeminate?”). Ratings for items were made on a 4-point never to many times scale. The items showed satisfactory internal consistency: α = .65. Six items measured respondents' exposure to adult homophobia (e.g., “As an adult, how often have you had to pretend that you are straight to be accepted?”); two of these items focused on experiences of verbal harassment and physical assaults in relation to both perceived sexual orientation and gender nonconformity (e.g., “As an adult, how often were you hit or beaten up for being homosexual or effeminate?”). The internal consistency was marginally acceptable: α = .59.
This variable was assessed through a 4-item measure; items were rated on a 4-point Likert scale ranging from never (0) to many times (3). Scale items measured symptoms of anxiety, depression, and suicidality during the previous six months (e.g., “In the last six months, how often have you thought of taking your own life?”). Consistent with previous findings (24), reliability analyses showed the scale had good internal consistency (α =.73). Higher scores indicated more psychological distress.
Consistent with previous research,(25) risky sexual situations was made-up of five items including: sex under the influence of drugs or alcohol; sex to escape negative emotions; sex within interpersonal relations of unequal power; situational experiences of sexual dysfunction; and sex with partners who resist condom use (e.g., “In the last 12 months, how often have you been lonely and depressed and had sex in order to feel good?”). Items were endorsed on a 4-point frequency scale ranging from never (0) to many times (3). Analysis of the scale's internal consistency yielded α = .70.
Risky sexual behavior was a categorical variable defined as those who reported having engaged in one or more acts of unprotected insertive or receptive anal intercourse with more than one male partner in the last year (1 = engaged in one or more acts; 0 = did not engage in any such acts).
Demographic characteristics were assessed via proportions, stratified by childhood sexual experience (none; consensual; forced). Proportions of respondents who experienced any vs. no psychological distress or risky sexual situations were computed for each childhood sexual experience group. Additionally, unadjusted odds ratios comparing childhood sexual abuse with non-forced sex groups were generated. Proportions and odds ratios were generated using Stata 9 for Windows. To examine indirect effects of childhood sexual experience and experiences of homophobia on sexual risk behavior by way of psychological distress and risky sexual situations, we used path analysis. Path analysis is a method of simultaneous estimation of multiple regression equations. The simultaneous estimation of more than one regression equation enables the computation of direct effects of each exposure variable on mediating and outcome variables. As well, for models that incorporate mediating variables, the analysis can compute the indirect effects of an exposure variable on an outcome variable by way of one or more mediating variables.
Based on prior population-based research findings,(10),(18),(24–27) we specified a saturated path analyses model in which forced childhood sex, experiences of childhood homophobia and adult homophobia influence the likelihood of subsequent sexual-risk behaviors directly and indirectly via intermediary psychological distress and risky sexual situations variables (Figure 1). All direct and indirect effects were estimated via weighted least-squares with a mean and variance adjustment using Mplus 5.1. Reflecting the probability-based sampling method used to recruit participants, all analyses incorporated case weights, city as a stratification variable, and recruitment venue as a cluster variable. Nativity (born in the US = 1; not born in the US = 2), age in years, and HIV status (0 = negative or unknown; 1 = positive) were included as covariates. For each effect we report the unstandardized regression weight (B), its standard error (S.E.), a Z-statistic that tests the null hypothesis that the estimated parameter is zero, and the standardized regression weight (β).
The prevalence of childhood sexual abuse among Latino gay/bisexual men was 15.8% (95% CI = 12.3%–19.2%). The prevalence of consensual sex was 34.8% (95% CI = 29.5%–40.1%) and no sex before 16 was 49.4% (95% CI = 44.6%–54.2%). The sample was primarily Spanish-speaking, born outside of the US, young (<31), and of modest socioeconomic status (<college-degree, and $100–$500 weekly-income). Childhood sexual experiences did not vary by any of the demographics.
We found no differences in psychological distress outcomes between those who reported no sex compared with consensual sex before age 16. Accordingly, we pooled both groups into a non-childhood sexual abuse group and compared them with a childhood sexual abuse group—comprised of individuals who had experienced forced sex in childhood. The childhood sexual abuse group reported significantly more: 1) anxiety, 2) depression, and 3) suicidality than the non-childhood sexual abuse group (see Table 1).
We found no differences between those who reported none compared with consensual sex before age 16. However, compared with no childhood sexual abuse, those who reported childhood sexual abuse were significantly more likely to report sexual situations involving: 1) drug and or alcohol use, 2) an escape from loneliness or depression, 3) a non-monogamous partner, and 4) difficulty maintaining an erection (Table 2).
Childhood sexual abuse (OR = 2.0; 95%CI = 1.1 – 3.8) significantly predicted increased adult risky sexual behavior in unadjusted logistic regression analyses. As previously reported,(25) we found that psychological distress (OR = 1.7; 95%CI = 1.1 – 2.5); exposure to childhood homophobia (OR = 1.6; 95%CI =1.3 – 2.0); and risky sexual situations (OR = 3.0; 95%CI = 2.0 – 4.3) significantly predicted increased adult risky sexual behavior in unadjusted logistic regression analyses.
Direct effects linking childhood sexual abuse, experiences of childhood homophobia, and experiences of adult homophobia to psychological distress, risky sexual situations, and sexual risk behavior are shown in Table 3. Childhood sexual abuse, exposure to childhood homophobia, and to adult homophobia were positively associated with psychological distress; however, those who experienced none or consensual sex did not differ from each other in levels of psychological distress. Psychological distress, exposure to childhood homophobia, and to adult homophobia were positively associated with participating in risky sexual situations.
Only risky sexual situations were directly linked with risky sexual behavior. This finding was qualified by the presence of several statistically significant indirect effects of exposures and mediators on risky sexual behavior. Childhood sexual abuse (B=.043, S.E =.023, Z=2.02, β=.016), experiences of childhood homophobia (B=.020, S.E.=.008, Z=2.35, β=.013), and experiences of adult homophobia (B=.048, S.E.=.014, Z=3.48, β=.023) were positively related to risky sexual behavior through psychological distress and risky sexual situations. By contrast, respondents who had experienced consensual sex were not significantly different from the no sex group in this indirect effect (B=.004, S.E.=.010, Z=0.45, β=.002). Men with higher levels of psychological distress were more likely to engage in risky sexual situations, which in turn lead to an increased risk of risky sexual behavior (B=.133, S.E.=.034, Z=3.95, β=.092).
A similar set of findings emerged for indirect effects linking childhood sexual abuse and homophobia with risky sexual situations by way of psychological distress. Childhood sexual abuse (B=.473, S.E.=.212, Z=2.23, β=.034), experiences of childhood homophobia (B=.205, S.E.=.081, Z=2.55, β=.029), and experiences of adult homophobia (B=.491, S.E.=.139, Z=3.53, β=.051) were positively associated with participating in risky sexual situations, whereas having non-forced childhood sexual experiences was not (B=.045, S.E.=.101, Z=0.45, β=.004).
Findings indicated that HIV-risk patterns among Latino gay/bisexual men in the United States are related to a history of childhood sexual abuse and a social context of discrimination that result in symptoms of psychological distress and risky sexual situations that increase risky sexual behaviors associated with HIV acquisition. The high prevalence of childhood sexual abuse among Latino gay/bisexual men in the present study is consistent with results of a probability sample study that found Latino MSM (22%) were twice as likely as non-Latino MSM to have a history of childhood sexual abuse.(18)
Using a probability sample from three large US cities, we found further support for the use of both subjective and objective considerations in the definition of childhood sexual abuse(10). There were no differences between Latino gay/bisexual men who reported no sex and those who reported consensual sex before age 16 with someone five or more years older for any of the outcomes under investigation, underscoring the importance of examining both subjective and objective experiences of sexuality in childhood. Consistent with the findings from a national probability sample of MSM,(10) Latino gay/bisexual men who initiated sex consensually before age 16 appeared to be at no greater risk for psychological distress, risky sexual situations, or risky sexual behaviors than those who initiate sex later. By contrast, having experienced childhood sexual abuse (forced sex before age 16 with someone five or more years older), was associated with greater risk for psychological distress, risky sexual situations, and risky sexual behaviors.
Path analysis also indicated no differential direct or indirect effects between no sex and consensual sex before age 16. Nevertheless, childhood sexual abuse was directly associated with an increased risk for psychological distress and indirectly associated with increased risk for Risky Sexual Situations and Risky Sexual Behaviors. Psychological Distress had a direct effect on risky sexual situations, which had a direct effect on risky sexual behaviors. These results suggest that the relationship between childhood sexual abuse and risk for HIV may be mediated by increased psychological distress and risky sexual situations. It may be that for Latino gay/bisexual men, placing themselves in risky sexual situations that increase risk for risky sexual behavior is an immediately gratifying strategy for coping with the psychological distress that comes with having experienced childhood sexual abuse. As Diaz has suggested, when the need for connectedness, affirmation, or self-worth is so great and deeply felt, individuals' usual concern for sexual safety and health seems to be overpowered.(28)
Not surprisingly, experiences of childhood and adult homophobia were directly associated with increased psychological distress, and directly and indirectly associated with risky sexual situations. Whereas the actual occurrence of childhood sexual abuse ends in childhood, a homophobic environment that continues into adulthood can result in ongoing psychological distress that result in maladaptive coping strategies where Latino gay/bisexual men seek out risky sexual situations to distract themselves from these feelings. While using sex as a coping mechanism may be beneficial for reducing unwanted feelings associated with early forced sexual experiences and a pernicious homophobic environment in the absence of meaningful relationships, these very coping strategies may lead to a greater sense of isolation and distress in a cycle that increases risky sexual situations.
The present findings reveal a strong relationship between childhood sexual experiences and patterns of adulthood sexual experiences among Latino gay/bisexual men. Although behavioral interventions to reduce risk behavior may be effective as a way of increasing awareness and reducing risky sexual behaviors among some men, it is becoming increasingly clear that we must develop interventions that address the contextual and psychological factors that influence Latino gay/bisexual men's sexual lives. For instance, many schools and workplaces feature anti-discrimination workshops and classes. These activities encourage children and employees to be mindful of antidiscrimination laws and promote respectful communication among peers, especially regarding topics involving classes of individuals highlighted in antidiscrimination legislation (e.g., racial and ethnic minorities). Sexual orientation should be added to such trainings.
Although using a probability sample of Latino gay/bisexual men in three U.S. cities is a strength of this study, it has limitations including: findings may not generalize to other cities in the U.S. and the model assumes equal correlation structure across the three childhood sexual abuse classification groups. Further, measurement of HIV risk behavior encompasses a one year time-frame including both insertive and receptive partners where a shorter time-frame that differentiated positioning could have captured risk more accurately. The model posited relationships among contemporaneously measured variables, making causal ordering among variables impossible to determine in this cross-sectional sample. Future studies should measure constructs such as exposure to childhood homophobia in samples of youth with the aim to determine if such experiences can be prospectively linked with subsequent sexual risk behavior in adulthood. Conceivably, such studies could investigate similar linkages between childhood sexual abuse and adult HIV-risk behavior, although ethical concerns and reporting requirements make such studies extremely challenging to conduct. It is likely that retrospective investigations such as ours will remain representative of childhood sexual abuse research in the immediate future. Additionally, all responses consisted of self-reports. Although previous research has found good reliability and validity for retrospective reports of childhood sexual abuse(29) and sexual risk behavior,(30, 31) interpretations of temporal or causal relationships are limited by the retrospective and cross-sectional nature of the data.
There is always the possibility of miscategorizing childhood sexual experiences. For example, those who may have been coerced into sexual activity in childhood, but who do not see themselves as having been forced, may have been miscategorized into the consensual group. However, as noted in previous research differentiating forced from consensual childhood sexuality,(10) one would expect miscategorization to have resulted in under-reporting of forced sex and inflated the risk outcomes for the consensual group. Nonetheless, it is worth repeating that individual cases of childhood sexual experiences are not populations. We must pay special attention to the individual's circumstances when determining whether the experience was abusive or not and, most importantly, how to best intervene to improve a sense of worth and safety for Latino gay/bisexual men. The field still grapples with two unresolved theoretical and methodological issues: the possibility of 'objectively' describing a sexual act as consensual; and the validity of asking subjects about the level of consensuality of their experiences. The current findings recommend future research focus on: individual and contextual resiliency factors that may be protective; other sociocultural indicators of both increased and decreased risk outcomes; and uncovering underlying mechanisms that maintain the relationship between childhood sexual abuse and later risk taking behaviors. The literature on childhood sexual abuse among women may offer some direction.
Studies with women have found a strong relationship between childhood sexual abuse and dissociation in cases where the childhood sexual abuse was severe.(32–36) Dissociation is a psychological state or condition in which certain thoughts, emotions, sensations, or memories are separated from the rest of the psyche. Pierre Janet coined the term, emphasizing its role as a defensive mechanism in response to psychological trauma.(37) It may be that the ability to dissociate during childhood sexual abuse is severe.(32–36) Dissociation is a psychological state or condition in which certain thoughts, emotions, sensations, or memories are separated from the rest of the psyche. Pierre Janet coined the term, emphasizing its role as a defensive mechanism in response to psychological trauma.(37) It may be that the ability to dissociate during childhood sexual abuse is adaptive, especially when there is no control over the situation such as is the case when a child is forced into a sexual behavior. However, to the extent that dissociation is generalized to nonthreatening situations (such as consensual sexual encounters in adulthood), the ability to make conscious choices about behavior in non-threatening contexts may be dramatically limited.
The extent to which childhood sexual abuse is associated with dissociation among Latino gay/bisexual men is not yet known. Given the importance of dissociation in the context of childhood sexual abuse found among women and the high level of severity of childhood sexual abuse among Latino gay/bisexual men, dissociation may be one of the mechanisms that accounts for the link between childhood sexual abuse and risky sexual behaviors in adulthood. Future research should examine the role of dissociation among Latino gay/bisexual men.
Finally, we must examine childhood sexual experiences in order to differentiate age appropriate and adaptive sexual development from that which is abusive or coercive. The present findings reveal the need to promote socially contextualized ways to reduce childhood sexual abuse and its sequelae including: increasing awareness through open discourse and education concurrent with laws and policies that make it more difficult to perpetrate sexual abuse of children. For those who have been sexually abused in childhood, interventions must also focus on reducing the psychological distress that appears to override Latino gay/bisexual men's intentions to practice safer sex. For instance, a young Latino gay man who feels unworthy as a result of physical (forced sex) and contextual maltreatment (e.g., homophobia) may not believe that he deserves to be taken seriously and consequently may not be able to garner the resources (psychological and environmental) needed to enact safe sex behaviors consistently. Research must be directed toward interventions that will reduce psychological distress associated with these factors, increase self worth, and create an environment or context that supports and welcomes Latino gay/bisexual men's full engagement in society.
This work was supported by NIH/NIMH Grants P30 MH62246 and MH067127 and NICHD Grant HD045810.