Elsevier

Child Abuse & Neglect

Volume 25, Issue 4, April 2001, Pages 557-584
Child Abuse & Neglect

Understanding childhood sexual abuse as a predictor of sexual risk-taking among men who have sex with men: The Urban Men’s Health Study1,

https://doi.org/10.1016/S0145-2134(01)00226-5Get rights and content

Abstract

Objective: The prevalence and characteristics of childhood sexual abuse (CSA) among men who have sex with men (MSM), and links with sexual risk are explored. A model linking CSA and sexual risk among MSM is proposed.

Method: A telephone probability sample of urban MSM (n = 2881) was recruited and interviewed between November 1996 and February 1998. The interview covered numerous health issues, including history of sexual victimization.

Results: One-fifth reported CSA, primarily by non-family perpetrators. Initial CSA experiences are characterized by high levels of force (43% involved physical force/weapons), and penetrative sex (78%; 46% reported attempted or actual anal intercourse). Such men are more likely than never-coerced men to engage in high risk sex (unprotected anal intercourse with a non-primary partner or with a serodiscordant male). In multivariate analyses, the effect of childhood sexual coercion on sexual risk is mediated by substance use, patterns of sexual contacts, and partner violence, but not by adult sexual revictimization or by depression.

Conclusions: Findings are interpreted within the context of social learning theory and prior research on sexual risk-taking. The high risk for CSA among MSM, which can predispose such men to patterns of HIV sexual risk, warrants new approaches in HIV prevention.

Résumé

Objectif: Explorer la fréquence et les caractéristiques des abus sexuels subis dans l’enfance (CSA) chez des hommes qui ont des relation sexuelles avec des hommes (MSM) et leurs liens avec les comportements à risque. Un modèle est proposé reliant CSA et comportements à risques chez ces hommes.

Méthode: Un échantillon probabiliste urbain d’hommes MSM a été recruté et interviewé par téléphone (n = 2881) entre novembre 1996 et février 1998. L’entretien couvrait de nombreux sujets concernant la santé, y compris les antécédents d’abus sexuels.

Résultats: Un cinquième des abus sexuels avaient eu une origine essentiellement extra-familiale. Les expériences initiales sont caractérisées par un haut niveau de violence (43% comportaient force physique/armes), et par la pénétration sexuelle (78%; 46% ont fait état de tentatives ou de réalisation de rapports anaux). Ces hommes sont, plus que ceux qui n’ont pas subi de violences, susceptibles de s’engager dans des pratiques à haut risque (rapport anal avec un partenaire qui n’est pas le partenaire principal ou avec un homme serodiscordant). Dans les analyses à plusieurs variables, l’effet de la violence sexuelle subie durant l’enfance est influencé par l’usage de stupéfiants, les patterns de contacts sexuels et la violence du partenaire, mais pas par la revictimisation ou par la dépression.

Conclusion: Les résultats sont interprétés dans le contexte de la théorie de l’apprentissage social et d’une recherche précédente sur les conduites sexuelles à risque. Le haut risque présenté de trouver des hommes qui ont été victimes d’abus sexuels dans leur enfance (CSA) parmi les hommes qui ont des rapports sexuels avec des hommes (MSM), ce qui les prédispose à des conduites à risque VIH, justifie de nouvelles approches dans la prévention du SIDA.

Resumen

Objetivo: El objetivo es explorar la prevalencia y las caracterı́sticas del abuso sexual infantil entre varones que tienen relaciones sexuales con varones y conocer las conexiones con las conductas sexuales de riesgo. Se propone un modelo para conectar el abuso sexual infantil con las conductas de riesgo sexual entre varones que tienen relaciones sexuales con varones.

Método: Una muestra probabilı́stica de sujetos varones que tienen relaciones sexuales con varones (n = 2881) fue seleccionada por teléfono y entrevistada entre Noviembre de 1996 y Febrero de 1998. La entrevista cubrió numerosas cuestiones de salud, incluyendo la historia de victimización sexual.

Resultados: Uno de cada cinco sujetos notificó episodios de abuso sexual infantil fundamentalmente de tipo extrafamiliar. Las experiencias iniciales de abuso sexual infantil se caracterizan por elevados niveles de utilización de la fuerza fı́sica o la amenaza con armas de fuego (43% de los casos) y de intento (78%) o consumación (46%) de penetración anal. Estos varones son más proclives que los que nunca han sido forzados a implicarse en actos sexuales de alto riesgo (relación sexual anal sin protección con pareja no conocida o con un varón serodiscordante). En los análisis multivariados se observa que el abuso sexual infantil ejerce su influencia en la conducta sexual de riesgo a través del efecto mediador de abuso de drogas, los patrones de contacto sexual y la violencia de pareja, pero no a través de la revictimización sexual adulta o la depresión.

Conclusiones: Los resultados son interpretados en el contexto de la teorı́a del aprendizaje social y de las investigaciones precedentes sobre conductas sexuales de riesgo. El alto riesgo de abuso sexual infantil entre varones que tienen relaciones sexuales con varones, que puede predisponer a tales varones a patrones sexuales comportamentales de riesgo para el HIV, merece nuevas aproximaciones en la prevención del sı́ndrome de inmunodeficiencia adquirida.

Introduction

Prior studies, utilizing primarily opportunistic samples, have found prevalence levels of childhood sexual abuse (CSA; defined here as prior to age 18) among adult men who have sex with men (MSM) that are close to levels reported by adult heterosexual women (25–39%; Doll et al 1992, Finkelhor 1994, Jinich et al 1998). The current study extends this earlier work in estimating CSA prevalence in a large probability-based sample of adult MSM in four large urban centers of the United States (New York, Los Angeles, San Francisco, and Chicago). In addition, we provide data describing characteristics of the CSA experiences of MSM.

As with heterosexual women (e.g., Choi, Binson, Adelson, & Catania, 1998), the observed relationship between CSA and adult sexual behavior has led investigators to consider CSA’s role in the transmission of Human Immunodeficiency Virus (HIV) among MSM. High HIV prevalence (>20%; Catania et al., 1998) and incidence rates (consistently 1–2% over the past decade; Catania et al., in press) among MSM in the United States underscore the continuing health threat HIV poses to MSM. To address continuing HIV risk, research on antecedents—which has focused on proximal causes of sexual risk (see reviews: Hospers and Kok 1995, Flowers et al 1997)—has expanded to include such developmental antecedents as CSA (e.g., see the Spring 1998 issue of AIDS and Behavior).

Prior studies have shown that a history of CSA is associated with HIV prevalence and sexual risk behavior among MSM Bartholow et al 1994, Carballo-Dieguez and Dolezal 1995, Doll et al 1992, Holmes 1997, Jinich et al 1998, Lenderking et al 1997. These findings, however, are based on samples (e.g., patients at clinics for sexually-transmitted diseases, street youth, those seeking HIV testing) for which estimates of sexual risk behaviors and other pertinent associated behaviors (e.g., substance use) may be inflated. We extend this prior work by examining the CSA/sexual risk behavior relationship and potential mediators of that relationship in a representative sample of MSM.

This research is guided by a social learning model of the CSA/sexual risk behavior relationship that we have developed specific to MSM. Models based upon samples of women may be overly influenced by gender-specific patterns and roles (e.g., compliance, passivity, internalized patterns of distress) that are less common among men. Furthermore, gender differences have been found in important details of childhood sexual coercive experiences, how these relationships are understood, and their long-term impact upon adult interpersonal and sexual functioning (Holmes & Slap, 1998). CSA has been related to higher levels of eroticism, lower levels of sexual anxiety, and more sexualized behaviors among male victims of CSA than among female victims Feiring et al 1999, Watkins and Bentovim 1992. Males who have been sexually abused are more apt than their female counterparts to exhibit aggressive, hostile behavior (Watkins & Bentovim, 1992), and to victimize others, possibly due to identifying with the aggressor Becker et al 1987, Burgess et al 1988, Carmen et al 1984, Groth 1979, Stevenson and Gajarsky 1991.

This model integrates prior research findings on the overlap between specific long-term sequelae of CSA and observed correlates of HIV risk behavior (albeit primarily with heterosexual adult women; e.g., Fergusson et al 1996, Fergusson and Horwood 1998, Miller 1999, Whitmire et al 1999) with other empirical findings and theoretical formulations Browne and Finkelhor 1986, Catania and Paul 1999, Hoier et al 1992. The proposed model hypothesizes relationships between CSA and emotional, cognitive, and interpersonal outcomes that impact directly or indirectly upon two key components of HIV sexual risk-taking models Catania et al 1990, Fisher and Fisher 1992: (a) the appraisal of potential risk, and (b) the capacity to enact behaviors necessary to reduce risk. Among the proposed mediators of this CSA/risk behavior process are motivational factors (e.g., social motives, including need for social acceptance; sexual feelings; sexual impulse control; affective states and emotions, such as interpersonal anger, depression, and anxiety), coping strategies (e.g., over-reliance on escape-avoidance coping), risk appraisal processes (e.g., inattention to danger cues), interpersonal regulatory abilities (e.g., low sexual assertiveness), and sexual scripts (e.g., established patterns of sexual passivity or aggression). The severity of these long-term outcomes of CSA-related trauma may vary from minimal symptomatology to such diagnoses as post-traumatic stress disorder or various personality disorders.

Our mediational model proposes a variety of pathways whereby key behavioral, cognitive, and emotional patterns are related to sexual risk-taking, and is framed by the following theoretical considerations with respect to CSA. Consistent with Hoier and her colleagues (1992), we conceptualize CSA (particularly severe CSA) as a painful experience on many levels (e.g., physiological, psychological, emotional), which arouses flight/fight responses under conditions that deter physical escape or effective avoidance, and which may also punish any efforts by the victim to change or evade the situation. The overgeneralization of the loss of control experienced during the traumatic event can effectively result in “learned helplessness” Abramson et al 1978, Seligman 1971, powerlessness, and low self-efficacy, associated with poor interpersonal regulation of adult sexual relationships. In extreme situations, learned helplessness may lead to depressive and self-destructive tendencies Boudewyn and Liem 1995, Briere and Runtz 1987, Browne and Finkelhor 1986. These tendencies indirectly deter enactment of health-promoting sexual practices by influencing key belief systems such as risk perceptions (e.g., being “unconcerned” about health risks). Poor adult interpersonal regulation (and consequent poor enactment of sexual health practices) may also stem from the fact that some CSA perpetrators use control techniques that denigrate the victim and lead victims to blame themselves for being abused. These conditions may result in the development of a negative self-concept that, in turn, impairs appropriate assertiveness in adult sexual relationships (e.g., the ability to refuse sexually aggressive partners).

Importantly, CSA differs from other forms of traumatization in that such early experiences may directly shape a victim’s subsequent sexuality and sexual relationships. “Traumatic sexualization” may lead to aversive, anxious/hostile reactions to sex, as well as poor impulse control. In addition, subjective positive aspects of these childhood sexual experiences (e.g., obtaining attention, nurturance) may reinforce the use of sexual behavior as a mechanism of relief/soothing, or a “commodity” to meet non-sexual interpersonal needs. For example, clinical studies of adults CSA survivors have described sexual “promiscuity” as a compulsive behavior motivated by anxiety and other disturbing emotions, which is reinforced by outcomes such as orgasm-related tension reduction (e.g., see Timms & Connors, 1992). Compulsive sexual patterns Dimock 1988, Krug 1989 are one explanation for CSA victims reporting higher numbers of sexual partners than do the non-victimized (for MSM, see Jinich et al., 1998). Numbers of sexual partnerships may also be elevated due to CSA-related outcomes that disrupt the development of long-term relationships (e.g., impaired interpersonal trust and attachment capacities), leading to a pattern of short-term sexual liaisons. In the absence of consistent condom use, these higher numbers of sexual partners increase risk for HIV infection. CSA experiences may not only impact the number of sexual partners, but sexual partner selection as well. Adult CSA survivors may be attracted to partners with “high-risk” psychological characteristics that are familiar to the victim and satisfy certain needs that arise from the CSA-perpetrator relationship. (Paradoxically, as Browne & Finkelhor [1986] point out, the betrayal involved in CSA may impair accurate appraisal of later relationships due to an emotional need to perceive a given relationship as trustworthy.) For example, people with a history of CSA are more likely to be attracted to (or to attract) dominating, overly controlling partners. Such partners have less concern about the well-being of others, which in extreme situations may extend to physical brutality, including rape (i.e., adult sexual “revictimization;” see, for example, Muehlenhard, Highby, Lee, Bryan, & Dodrill, 1998).

Furthermore, there may be other long-standing effects of CSA on victims’ adult sexual relationships. As noted previously, CSA perpetrators model sexually aggressive behavior and poor anger control; their victims may learn and subsequently display similar patterns of emotional response and anti-social behavior in their adult sexual relationships. Conversely, perpetrators may reward compliance and passivity in the face of aggression, which reinforces more passive responses in sexual situations by CSA victims. Long-term patterns of sexual aggression or sexual helplessness often result in the establishment of rigid, invariant sexual scripts that promote sexual risk-taking (e.g., through an inability to either insist on healthy sexual practices, or care about the health of others).

Although numerous aspects of CSA experiences may directly influence sexual risk practices, there are also indirect effects mediated by the influence of CSA on coping strategies. In general, CSA victims show an over-reliance on cognitive or behavioral escape-avoidance coping strategies (e.g., substance use, dissociation). This reliance on relatively primitive coping strategies may be a consequence of their learned utility in managing the emotional distress of CSA experiences Chu and Dill 1990, Miller 1999, as well as CSA-related long-term negative outcomes (e.g., anxiety, depression, impaired interpersonal functioning). Psychological or emotional distance is provided (in these coping strategies) by dissociating or compartmentalizing aspects of one’s conscious experience, or by the use of alcohol or drugs to dull dysphoric states. These escape-avoidance coping strategies, however, reduce attention to danger cues that impair accurate risk appraisal, and inhibit self-regulatory processes needed for negotiating and enacting safer sexual practices.

The analysis in this paper is a first effort at verifying some of these proposed relationships. We are constrained by the original purpose of the study (a broad-based general health assessment of MSM) to using proxy variables that suggest, but do not fully test, the theoretical constructs and underlying hypothetical relationships. In the current paper, we examine depression, alcohol/drug use during sex, occurrence of “one-night stands,” abusive relationships, and adult sexual revictimization (coerced sex after age 17). These variables are believed to reflect underlying motivational, cognitive, and interpersonal factors that we have hypothesized to mediate the link between CSA and sexual risk-taking.

Depression, as noted, is a common outcome of CSA (Browne & Finkelhor, 1986), and negative affective states have been noted as a correlate of sexual risk behavior among gay men (Marks, Bingman, & Duval, 1998). Depression is linked to motivational deficits and “learned helplessness” beliefs; the subsequent sense of low self-efficacy may negatively influence the ability to regulate sexual encounters so as to avoid sexual risk (see de Vroome, de Wit, Stroebe, Sandfort, & van Griensven, 1998). Depression often elicits the use of coping strategies such as substance use that impair accurate assessment of risk by reducing attention to danger cues (Miller, 1999; see below). Finally, depression can precipitate self-destructive impulses that can lead to sexual risk-taking.

Substance use has been noted both as an outcome of CSA Miller 1999, Stein et al 1988, and as a correlate of sexual risk-taking (e.g., Beltran et al 1993, Ekstrand and Coates 1990, Kelly et al 1991, McKirnan et al 1996, Stall et al 1988). As a behavioral escape-avoidance coping strategy, substance use may lead to high-risk sex by undermining interpersonal regulatory abilities that prevent unsafe sex, by reducing perceptions of risk or concerns about safety, or by facilitating sensation-seeking.

CSA victims, relative to non-victims, tend to report higher frequencies of brief sexual relationships in adulthood. (As noted earlier, this is multiply determined by compulsive sexual patterns, the use of sex to meet other interpersonal needs, and interpersonal difficulties.) Among MSM, such frequent sexual partnerships increases the likelihood of having a “one-night stand” (where essential strangers briefly get together for the primary purpose of having sex). Such a scenario increases the likelihood of encountering a partner who demands unsafe sex, is exploitative, or who may be HIV-positive. Difficulties arising in “one-night stand” situations include the failure to make accurate risk appraisals, deficits in the capacity to negotiate sexual interactions, or an emphasis on pleasure at the expense of sexual health, all of which contribute to a failure to enact low-risk sexual activities. Thus, “one-night stands” are not just indicative of certain psychological traits, but a proxy measure for key situational factors.

Having an abusive partner in adulthood (another long-term outcome of CSA) may reflect poor interpersonal regulatory abilities and/or a submissive role in sexual relationships, either of which can lead to a failure to avoid sexual risk. As noted previously, misjudging risks may lead to sexual revictimization in adulthood, which has a high likelihood of involving unsafe sex.

In general, we hypothesized that men with a history of CSA, particularly more severe CSA, will report higher rates of depression, use of substances during sexual encounters, one-night stands, abusive sexual partners, and adult sexual revictimization (relative to non-CSA men). Further, after controlling for other developmental variables (e.g., childhood physical abuse, parental substance abuse) and demographic variations on the dependent variable, we hypothesized that the correlates of CSA will mediate the CSA/sexual risk relationship.

Section snippets

Sample

The data reported here come from the Urban Men’s Health Study, a telephone probability sample of men who have sex with men (MSM) in four US cities (San Francisco, New York, Los Angeles, and Chicago), conducted between November 1996 and February 1998. Data from the Federal Communications Commission and the 1990 US Census indicate that telephone coverage in these four cities is at least 93% Lavrakas 1987, US Bureau of the Census 1995. In addition, research has found that households with

Results

We first examined for mode effects. No differences by mode of interview were found for any self-reported history of sexual victimization (CATI 35.5% vs. T-ACASI 36.6%), childhood sexual abuse (CATI 20.6% vs. T-ACASI 21.0%), or high number (6+ times) of CSA events (CATI 5.0% vs. T-ACASI 4.7%). Thus, prevalence of CSA, despite being sensitive information, was unaffected by the presence of a live telephone interviewer.

The prevalence of CSA was estimated to be 20.6% (95% confidence interval [CI] =

Discussion

Our study confirms and extends prior research indicating high prevalence levels of childhood sexual abuse (CSA) among MSM. Such prevalence levels might be higher if we had elicited data about experiences involving only non-contact sexual victimization (i.e., sexual exposure or exhibitionism). Overall, these men’s CSA experiences were characterized by high levels of penetrative sex, physical force, and perceptions of these events as distressing. A substantial subgroup reported multiple

Acknowledgements

This study would not have been possible without the extensive cooperation of the men who were willing to serve as project informants and participants in the hope that the knowledge gained would be of benefit to other gay/bisexual men. Jesse Canchola and Drs. Judith Moskowitz, and Diane Binson helped to ensure the success of this project by their attention to numerous details on this project since the initiation of preliminary fieldwork. We thank Dr. David Finkelhor for his advice and ideas in

References (79)

  • K Schaaf et al.

    Relationship of childhood sexual, physical, and combined sexual and physical abuse to adult victimization and posttraumatic stress disorder

    Child Abuse & Neglect

    (1998)
  • L Abramson et al.

    Learned helplessness in humanscritique and reformulation

    Journal of Abnormal Psychology

    (1978)
  • J Anderson et al.

    Telephone coverage and measurement of health risk indicatorsdata from the National Health Interview Survey

    American Journal of Public Health

    (1998)
  • W Aquilino

    Interview mode effects in surveys of drug and alcohol usea field experiment

    Public Opinion Quarterly

    (1994)
  • R Baron et al.

    The moderator-mediator variable distinction in social psychological researchconceptual, strategic, and statistical considerations

    Journal of Personality and Social Psychology

    (1986)
  • B Bartholow et al.

    Emotional, behavioral, and HIV risks associated with sexual abuse among adult homosexual and bisexual men

    Child Abuse & Neglect

    (1994)
  • J Becker et al.

    Adolescent sexual offendersdemographics, criminal and sexual histories, and recommendations for reducing future offenses

    Journal of Interpersonal Violence

    (1987)
  • E Beltran et al.

    Predictors of sexual behavior change among men requesting their HIV–1 antibody statusthe Chicago MACS/CCS cohort of homosexual/bisexual men, 1985–1986

    AIDS Education & Prevention

    (1993)
  • D Binson et al.

    Prevalence and social distribution of men who have sex with menUnited States and its urban centers

    The Journal of Sex Research

    (1995)
  • Binson, D., Moskowitz, J., Mills, T., Anderson, K., Paul, J., Stall, R., & Catania, J. (1996). Sampling men who have...
  • J Blair

    A probability sample of gay urban malesThe use of two-phase adaptive sampling

    The Journal of Sex Research

    (1999)
  • S Boney-McCoy et al.

    Is youth victimization related to trauma symptoms and depression after controlling for prior symptoms and family relationships? A longitudinal, prospective study

    Journal of Consulting & Clinical Psychology

    (1996)
  • S Boney-McCoy et al.

    Psychopathology associated with sexual abusea reply to Nash, Neimeyer, Hulsey, and Lambert (1998)

    Journal of Consulting & Clinical Psychology

    (1998)
  • A Boudewyn et al.

    Childhood sexual abuse as a precursor to depression and self-destructive behavior in adulthood

    Journal of Traumatic Stress

    (1995)
  • J Briere et al.

    Post sexual abuse traumadata and implications for clinical practice

    Journal of Interpersonal Violence

    (1987)
  • A Browne et al.

    Impact of child sexual abusea review of the literature

    Psychological Bulletin

    (1986)
  • A Burgess et al.

    Serial rapists and their victimsreenactment and repetition

    Annals of the New York Academy of Science

    (1988)
  • Cannell, C., Oksenberg, L., Kalton, G., Bischoping, K., & Fowler, F. (1989). New techniques for pretesting survey...
  • E Carmen et al.

    Victims of violence and psychiatric illness

    American Journal of Psychiatry

    (1984)
  • J Catania et al.

    Effects of interviewer gender, interviewer choice, and item wording on response to questions concerning sexual behavior

    Public Opinion Quarterly

    (1996)
  • J Catania et al.

    Methodological research on sexual behavior in the AIDS era

    Annual Review of Sex Research

    (1995)
  • J Catania et al.

    Towards an understanding of risk behavioran AIDS risk-reduction model (ARRM)

    Health Education Quarterly

    (1990)
  • Catania, J., & Paul, J. (1999). Sexual development and mental health among men who have sex with men. Paper presented...
  • Catania, J., Osmond, D., Stall, R., Pollack, L., Paul, J., Canchola, J., Blower, S., Porco, T., Binson, D., Mills, T.,...
  • Catania, J., Morin, S., Canchola, J., Pollack, L., Chang, J., & Coates, T. . Discordance between prevention financing...
  • K.-H Choi et al.

    Sexual harassment, sexual coercion, and HIV risk among US adults 18–49 years

    AIDS & Behavior

    (1998)
  • J Chu et al.

    Dissociative symptoms in relation to childhood physical and sexual abuse

    American Journal of Psychiatry

    (1990)
  • R Coates et al.

    Validity of sexual histories in a prospective study of male sexual contacts of men with AIDS or an AIDS-related condition

    American Journal of Epidemiology

    (1988)
  • E de Vroome et al.

    Sexual behavior and depression among HIV-positive gay men

    AIDS & Behavior

    (1998)
  • Cited by (243)

    • Association between childhood sexual abuse and patterns of HIV risky sexual behaviors among men who have sex with men in China: A latent class analysis

      2021, Child Abuse and Neglect
      Citation Excerpt :

      This finding can provide important evidence for future behavioral intervention on the sexual minority group. The current study indicated that the CSA (22.6%) is prevalent among MSM, which is higher than Chinese men (9.1%) and consistent with previous Western studies (Lloyd & Operario, 2012; Ma, 2018; Paul et al., 2001). But a study in urban north China found that the prevalence of CSA was only 5% among urban Chinese MSM (Pan et al., 2017).

    View all citing articles on Scopus

    Primary support for this study was provided by NIMH/NIA Grant No. MH54320 (PI—Joseph Catania). Supplemental support came from NIMH Grant No. MH42459 (The Center for AIDS Prevention Studies, PI—Thomas Coates), NIAAA Grant No. AA 10194 (PI—Ronald Stall), and the Centers for Disease Control and Prevention’s Division(s) of HIV/AIDS Prevention.

    1

    The random digit dial (RDD) sample frame was constructed by the Survey Research Center (SRC) of the University of Maryland (UMD), in collaboration with Dr. Graham Kelton at Westat and University of California investigators.

    View full text