Childhood sexual experiences and the perception of abuse among Latino men who have sex with men.

This paper is based on interviews with men who have had childhood sexual experiences with an older partner (CSEOP). At the time of the interview, some of these men felt that their experiences were childhood sexual abuse (CSA) and some did not. The first goal was to determine what aspects of their experiences differentiate these two groups (i.e., what experiences are associated with a perception of sexual abuse among men who have had CSEOP). This should broaden the understanding of CSA and contribute to a greater consensus regarding the definition of CSA. The second goal was to explore possible negative correlates (in adulthood) associated with CSEOP and the perception of having been the victim of childhood sexual abuse.

The Perception of Sexual Abuse Among Men who Have had CSEOP

In the last several years, many studies and doctoral dissertations have focused on childhood sexual abuse. Several recent review articles and meta-analyses have attempted to summarize and synthesize these studies' findings (Holmes & Slap, 1998; Jumper, 1995; Neumann, Houskamp, Pollock, & Briere, 1996; Rind & Tromovitch, 1997; Rind, Tromovitch, & Bauserman, 1998). Yet there is still a general lack of consensus among researchers as to how to define CSA. Definitions of CSA typically involve several criteria: age of the child (most set the maximum age at some point between 13 and 18); age of partner (sometimes based on an absolute age, other times on the age difference between child and partner; furthermore, sometimes this age differential varies based on age of child); relationship between child and partner (e.g., the partner must be an authority figure such as a parent, caregiver, or teacher); the sexual acts involved in the experience (criteria range from noncontact behaviors, such as physical exposure or sexual talk, to penetrative sex); the reaction of the child (many definitions ask about "something that you did not want to do," while others do not consider the volition of the child); and the extent of coercion involved. Given that this is a partial list of differences among definitions, it is clear that any comparison of CSA prevalence across a variety of studies is nearly meaningless unless some similarity in definition is established. For example, Fromuth and Burkhart (1987) demonstrated that prevalence rates varied from 4% to 24% in their samples of college males, simply by varying the definition of CSA. Similarly, Jumper (1995) found little association between self-esteem and CSA in college samples, compared to noncollege samples, but points out that all four college samples in the meta-analysis included non-contact behaviors in their definition of CSA, indicating that the differences in the degree of associations are possibly due to differences in the severity of the experiences considered CSA in the various studies, rather than actual sample differences. A greater consensus regarding how CSA is defined would allow for cross-study comparisons that are currently difficult, given the broad variability in operationalizing CSA.

Surprisingly, the individual's own perception of whether her/his experiences constitute sexual abuse is typically not considered in the various definitions outlined above. Freire (1990) emphasizes the importance of engaging victims of injustice in dialogue regarding their oppression. He writes, "Attempting to liberate the oppressed without their reflective participation in the act of liberation is to treat them as objects which must be saved from a burning building," and "It is necessary to trust in the oppressed and in their ability to reason" (Freire, 1990, pp. 52-53). Researchers and theorists should take into consideration the perspective of those who have experienced the abuse or sexual events with which they are concerned. In addition, understanding the factors associated with a perception of abuse among people who have had CSEOP could contribute to efforts toward a greater consensus regarding the definition of CSA. Knowledge of such accounts will also raise awareness regarding the variety of childhood sexual experiences and thereby inform those who are interested in CSA, whether for scientific, clinical, or personal reasons.

Adulthood Correlates of Childhood Sexual Experiences With an Older Partner

Another topic of continuing debate is the issue of harmful repercussions throughout life for those who experience CSA. Although it is implicit in most research that CSA is a traumatic experience for the victim, followed by long-term suffering and negative consequences, this assumption has been challenged in two meta-analyses conducted by Rind et al. (1997, 1998). They conclude that negative psychological consequences of CSA are typically neither pervasive nor intense. They also note that negative reactions are less likely for males who have experienced CSA compared to females. They even convincingly challenge a contemporary meta-analysis (Jumper, 1995) that reported relatively large effect sizes for the relationship between CSA and negative psychological consequences (it seems that the effect size calculations of the latter report were often erroneous; see Rind et al., 1998, p. 26). Rind et al. (1998) found that even when negative correlates of CSA were found, the effect was greatly attenuated if there was any statistical adjustment for family environment covariates. Rind's conclusions provoked strong reactions (Goode, 1999) because critics felt they gave "pedophiles a green flag" and trivialized the impact of CSA experiences. Politicians; activists, and media personalities spoke out condemning Rind's articles. Nevertheless, some studies have found that sexual experiences in childhood with older partners are sometimes seen as positive experiences by the children themselves (Okami, 1991: males and females who reported sexual contact, as defined by the participant, prior to age 16 with another person 5 or more years older; Sandfort, 1984: boys in a pedophiliac relationship with an adult male in which sexual contact occurred), so Rind's conclusions are not without precedent.

The current study assessed self-esteem, substance use, sexual dysfunction, and sexual-risk behavior and their association with CSA and the perception of sexual abuse. Analyses also considered the role of childhood family environment as a confounding factor.

Study Background

The current study is based on a sample of Latino men who have sex with men (MSM) in New York City. The study was designed to compare the sexual behavior and correlates of HIV-risk behavior among four ethnic groups. In the early 1990s, one author of this paper (Carballo-Dieguez) conducted several in-depth interviews with Puerto Rican MSM in preparation for a study within that population regarding determinants of HIV-related sexual risk behavior. The interviews included questions about psychosexual development, and some of the men mentioned early sexual experiences with adults, with various emotional reactions. Further exploration of these issues seemed warranted. In the study of Puerto Rican MSM that followed, questions regarding childhood sexual experiences were included as potential predictors of adult sexual risk behavior. Early sexual experiences with older partners were found to be quite common, although many felt they were willing participants and said they were not physically or emotionally hurt (Carballo-Dieguez & Dolezal, 1995). In the current study of Latino MSM, this section of the interview was expanded to learn more about these men's experiences and their current perception of the events as abuse or not.

METHOD

This study was designed to compare the sexual behaviors and correlates of sexual risk among four ethnic groups of Latino MSM. A convenience sample of Latin American men was recruited in New York City (predominantly in 1996) from a variety of sites (bars, dance clubs, parades, AIDS service organizations, community centers, public parks, etc.). To qualify for the study, men had to (a) be 18 or older; (b) be of Colombian, Dominican, Mexican, or Puerto Rican descent; and (c) have had sex with a male partner at least once in the last 12 months and at least 10 times in their lifetime. These four ethnic groups were chosen because they were among the largest Latino groups in New York City and they were located in distinct geographical localities that were judged conducive to successful recruitment. The criterion regarding sexual behavior was to exclude men who were not recently sexually active with men and also those who may have had only rare, isolated sexual experiences with men, such as in prison.

The assessment included an interviewer-administered section and a self-administered section. Interviews were audio taped and monitored for quality control. Assessments were available in Spanish or English, according to the participants' stated preferences (about two thirds were conducted in Spanish). Participants were paid $10 per hour for their participation. The study design and all study materials were approved by New York State Psychiatric Institute's IRB.

All participants signed consent forms that outlined the nature of the study and guaranteed confidentiality. The consent form signed by the men stated "This is an exploratory study to learn more about Hispanic men who have sex with men, about their knowledge concerning HIV disease and AIDS, and their sexual behavior." Although the men knew they would be asked questions of a sexual nature, the various aspects of sexuality involved (current sexual behavior, attitudes/intentions/beliefs about sexual practices, condom use, social norms, how pleasurable various sexual practices were, commercial sex involvement, drug and alcohol use during sexual activity, and early childhood sexual experiences) were not specified. While there was no attempt to deceive the participants about the themes of the interview, the men were not recruited based on any reference to CSA. This is important because it establishes that this is not in any way a clinical sample (e.g., men from clinical settings offering therapeutic services to victims of CSA) or a sample who volunteered (or refused to volunteer) because the study included an assessment of childhood sexual experiences. Sample bias is therefore less of a concern than it would be if the men were recruited based on CSA-related criteria.

The interview included a section that asked about childhood sexual experiences.-Our criteria for CSEOP specified that the participant had to have had (a) sexual contact, defined as an occasion in which the respondent's mouth, anus, or genitals were in contact with those of another person (kissing on the lips not included); (b) contact prior to age 13; and (c) contact with a partner at least 4 years older than the participant. These criteria were chosen based on a review of the literature available at the time (1995). Thirteen was chosen as the age cut-off point, since it was felt that males are more likely to be actively seeking sexual experiences as they go through puberty. A 4-year age gap was specified in order to exclude sex with peers, although it is acknowledged that it is possible for someone to be the victim of coercive, .violent, and abusive sexual experiences at the hands of same age, Or even younger, individuals. No criterion regarding force or volition was included precisely because we wanted to capture a broader range of sexual experiences and not only the more extreme situations in which an individual was forced to do something that he did not want to do. Those who met the above criteria were asked to provide extensive details of the event(s), and the participants' current perceptions of the experiences were elicited. All of the details used in this report were elicited via close-ended questions with the exception of one open-ended, qualitative question regarding why participants did or did not think what happened to them was sexual abuse. These qualitative responses were reviewed by both authors and broadly categorized into common themes for descriptive purposes.

Current sexual risk behavior (last 12 months) was assessed using the Sexual Practices Assessment Schedule (Carballo-Dieguez, Dolezal, Nieves-Rosa, & Diaz, 1995). This interviewer-administered section asked about various sexual practices and frequency of condom use separately by partner gender and type (lover, one-night-stand, other). Included in our analyses (as indicators of HIV transmission risk) are the frequency of unprotected anal sex with male partners and the number of male sex partners. Due to skewed distributions, these values were log-transformed prior to statistical group comparisons. Alcohol and drug use was assessed for the past 12 months via an interviewer-administered set of questions about 15 substance-use categories (some categories covered multiple substances, e.g., marijuana/hashish or barbiturates/tranquilizers). Participants responded to questions about frequency of use using a 7-point scale from 0 = never to 6 = more than once a day. Drug use was measured as the mean response to all nonalcohol substance categories. Childhood family environment was assessed by 7 questions regarding the occurrence and severity of problems within the family when participants were growing up (e.g., substance use, conflict, violence, mental health problems, etc.). The 5-point response scale ranged from not at all to extremely concerned about these problems, and a Family Environment score was calculated as the mean of these items.

Two measures included in our analyses came from a self-administered questionnaire completed after the interview portion of the assessment. These are the widely used Rosenberg Self-Esteem Scale (Rosenberg, 1965) and sexual functioning. The sexual functioning score is the mean of 3 questions regarding sexual drive, frequency of erotic thoughts, and satisfaction with sex life on 7-point scales. These items are from the Brief Sexual Functioning Questionnaire (Reynolds et al., 1988).

Descriptive statistics are presented on various aspects of the participants' experiences. Mann-Whitney tests were used for 2-group comparisons on continuous variables; chi-square tests were used for 2-group comparisons on dichotomous variables. ANOVAs were used for 3-group comparisons which were followed by post hoc 2-group contrasts using Tukey's adjustment for multiple comparisons. ANACOVAs paralleled the ANOVAs but adjusted for childhood family environment.

RESULTS

The full study sample consisted of 307 men (80 Colombian, 80 Dominican, 80 Puerto Rican, and 67Mexican). The mean age of the men was 31 (range = 18-54), they had on average 13.6 years of education (range = 3-20 years), and their mean annual income was $16,236 (range = $0-$95,000). (For other manuscripts based on this sample see Carballo-Dieguez et al., 2000; Dolezal, Carballo-Dieguez, Nieves-Rosa, & Diaz, 2000; Nieves-Rosa, Carballo-Dieguez, & Dolezal, 2000.)

Table 1 shows the breakdown of the sample in terms of CSEOP. Forty-one percent of the 307 men did not meet criteria because they did not have sexual contact prior to age 13. Therefore, the majority of the men did have sexual contact at an early age. An additional 26% (of the total sample of 307) did not meet criteria for CSEOP, because although they had sexual contact prior to age 13, their partner was not 4 years older than them at the time. If any partner was 4 years older and had sexual contact with the participant prior to his 13th birthday, the participant met criteria (i.e., questions were not limited to first sexual experience). One hundred men met our criteria for CSEOP. These 100 did not differ from the remainder of the sample in ethnic composition, age, or income. They did, however, have significantly less education: on average, one year less than the 207 men who did not meet CSEOP criteria.

Characteristics of Childhood Sexual Experiences With an Older Partner

Table 2 shows the characteristics of CSEOP for this sample. The median age of the first such experience was 8.5 and the partner was, on .average, 9 years older. For 80% of the men there was more than one occasion of sexual contact with that partner, ranging up to 6000 sexual occasions over 20 years. The vast majority of the older partners were male. A third of the partners were relatives (20 reported the partner was a cousin, 7 an uncle, 3 a brother, 2 a father, 1 a step-father, and 1 an aunt). A substantial number of the men were forced, threatened, and/or physically hurt during the experience, and slightly over half felt that they were emotionally hurt. Positive, negative, and indifferent reactions were fairly equally common, especially at the timeof the experience. For 32 men their reaction at the time of the event(s) was different from their current view. Over half (n = 17) of those whose opinion changed were initially indifferent. Of these, 14 currently have a negative view, while only 3 have adopted a more positive view. Only 4 men had a positive view at the time of the experiences and have changed to an indifferent (n = 3) or negative (n = 3) view. Of the 11 men who changed from an initially negative view, 4 were currently indifferent and 7 were currently positive. Very few (14%) told anyone about the experience. Over half of the men currently considered their experiences to be sexual abuse.

Table 3 shows the prevalence of various sexual behaviors involved in these sexual experiences. Fondling was the most common behavior. Over half of the participants with a male older partner fellated him and 40% were penetrated (anal sex) by the male partner. Only two men reported having penetrated their partner anally, one with a male partner, one with a female partner.

Aspects of CSEOP That are Associated With a Perception of Sexual Abuse

As was reported in Table 2, 59% of these men considered their experiences to be sexual abuse. These 59 men were compared to the 41 who did not believe that their experience was sexual abuse. The two groups did not differ with regard to the age of the older partner, whether the partner was male or female, whether the partner was a relative, the number of occasions involved, the length of the "relationship," or whether they told anyone about the experience. Since the 100 men with CSEOP differed from the remaining 207 men in the study on their level of education, we also assessed whether education was associated with the perception of abuse and found it was not. Table 4 shows results for five variables that did distinguish the two groups of men. Those who felt they were sexually abused were approximately 2 years younger when the first sexual contact took place. They were also much more likely to report that they were forced, threatened, or hurt emotionally or physically. The two groups were also compared on the sexual behaviors reported on Table 3 (results not shown). The first 11 behaviors listed regarding a male older partner were assessed; other behaviors were too infrequent for statistical tests. Two differences were found: Those who did not consider their experiences to be sexual abuse were more likely to report that they kissed (/44% vs. 20%) and exposed themselves to their older partner (50% vs. 22%).

Each participant was asked why they did or did not consider the event(s) to be sexual abuse. For those who did consider it sexual abuse, over half of the men referred to their age (e.g., "I was a child," "I was too young," and "A child doesn't know what he is doing"). The next most common response had to do with volition (e.g., "It was done without my consent," "It was against my will," and "I was forced to do things"). The men who did not consider the event to be abuse also frequently referred to their volition, with approximately two thirds stating that it was consensual and that they were not forced into the situation. Several felt that they had actually initiated the experience (e.g., "It was my initiative," "I was the one who went out for it," and "I exposed myself in front of him and provoked him." More common responses simply stated that they agreed to the encounter(s) (e.g., "I agreed to everything," "I was consenting," and "I was curious, I wanted to do it"). One participant was 12 and had a 17-year-old male partner. They had sexual contact 5 times over the course of a year. The participant was penetrated orally and anally. He said he was not coerced but was emotionally and physically hurt. Yet he did not consider it sexual abuse because "I agreed to it. I insisted more than the other person for this to take place." Another participant was 10 when he had sexual contact on 20 occasions over 3 months with a 25-year-old male neighbor. The events involved mutual masturbation and oral sex. The participant did not feel coerced or hurt and did not feel it was sexual abuse "because I seduced the neighbor."

The four ethnic groups were compared regarding the prevalence of CSEOP and the perception of abuse. The four groups did not differ in the prevalence of CSEOP, with roughly a third of each group reporting an early sexual experience with an older partner. Among those who did report such an experience, there were significant differences in the proportion of each group that felt their experiences were sexual abuse (chi-square p = .021). Seventy-six percent of the Colombian men felt they had been sexually abused, as well as 62% of the Mexican men and 64% of the Puerto Rican men. However, only 35% of the Dominican men felt they were abused. When the experiences of the Dominican men were compared to the other groups, it became clear that their experiences were less likely to contain elements that were associated with a perception of abuse in this sample. They were older when the event(s) began (ANOVA p = .040), were less likely to have been forced (chi-square p = .047), and less likely to have been threatened (chi-square p = .071). They also had the lowest prevalence of being emotionally hurt and physically hurt although these differences were not significant. So, it appears that the experiences of the Dominican men were less severe, and that--rather than a more lenient attitude toward abusive experiences--is the reason fewer of them considered their experiences sexual abuse.

Adulthood Correlates of Childhood Sexual Experiences With an Older Partner

Table 5 shows how three groups differ on several psychological and behavioral measures. The three groups are (a) those who did not have sex prior to thirteen with a partner at least four years older than them, (b) those who did have such a sexual experience (CSEOP) but do not consider it to be sexual abuse, and (c) those who had the experience and feel that they were the victims of sexual abuse.

The three groups differed regarding their frequency of alcohol use (marginally, p = .051), frequency of unprotected anal sex (p = .042), number of male sex partners (p = .031), and childhood family environment (p = .018). Post hoc 2-group comparisons with a Tukey adjustment for missing data showed that in all four cases the only significant 2group comparisons were between the extreme groups (i.e., those who did not have CSEOP and those who did and consider it abuse; all p < .05). When childhood family environment is included as a covariate in the ANACOVAs, the three groups continue to differ on alcohol use (from marginal without adjustment to p = .045 after adjustment), while group differences in frequency of unprotected anal sex become marginal (p = .068), as do differences in the number of male sex partners (p = .052). These are very subtle changes that happen to shift back and forth across the conventional p value of .05. The results indicate that group differences are quite modest, and adjusting for childhood family environment does not have much of an effect.

DISCUSSION

Prevalence of CSEOP Among Latino MSM

It is clear that early sexual contact is common among these men. Fifty-nine percent had had some sexual/genital contact prior to their 13th birthday. In the majority of those cases, they had a partner who was at least 4 years older than they were. Since a third of the sample reported childhood sexual experiences with an older partner, these experiences cannot be considered rare or isolated occurrences for this population. Since these rates were found among a convenience sample, the representativeness of the findings is an issue to question. To address this issue we compared our rates to data collected during 1997 and 1998 in a Center for AIDS Prevention Studies project called the Urban Men's Health Study (UMHS; Paul, Catania, Pollack, & Stall, 2001). This study involved telephone interviews with a probability sample of men who have sex with men in four cities, including New York. Seventy-one Latino men in New York City responded to questions about early sexual activity. Age of participant and age of partner allowed us to identify the proportion of the men who had sex prior to age 13 with a partner that was at least 4 years older. Nine of the men (12.7%) reported such experiences. However, their questions were specifically about sexual coercion ("Have you ever been forced or frightened by someone into doing something sexually that you did not want to do?"). When we identified the men in our sample who said they had been forced, they represented 12.7% of our sample as well (39 of 307). These remarkably identical rates indicate that our sample is probably not atypical of Latino MSM with regard to CSEOP, at least those in New York, which enhances the generalizability of our findings. Note that not all of the nine UMHS men were from one of our four ethnic groups; however, most were (three were Puerto Rican, two Mexican, two South American [which includes Colombia], one Central American, and one European).

This comparison of our data and the UMHS data again emphasizes the impact of varying definitions of childhood sexual abuse. Their rate of 12.7% is substantially lower than the 32.6% of our sample that was identified as having sex prior to age 13 with a partner 4 years older. This is because their assessment only collected data from men who were forced into doing something sexually that they did not want to do. While this is a legitimate and widely used criterion for childhood sexual abuse, there are a great many other early sexual experiences that are never detected or assessed. These other experiences may often be relevant to the interests of the research being conducted. Furthermore, the more strict criterion may give the impression that a relatively small proportion of a sample engaged in sexual activity with an older partner and that all such activity involves coercion, both of which are not supported by our data.

Although we see some indication that the rates of early sexual experiences in our sample are not atypical of Latino MSM in New York City, it is still difficult to compare our findings to other samples, largely due to the wide range of definitions used. Holmes and Slap (1998) summarized 166 studies that reported data on male childhood sexual abuse in 149 samples and found that prevalence rates ranged from 4% to 76%, depending on the definition used and the sample involved. Although the rates of CSEOP found in our sample seem higher than rates of sexual abuse reported in most studies involving males, it seems of limited value at this point to attempt to establish whether these rates are, for example, higher than among White, heterosexual male college samples but lower than among serial rapists. What is clear, however, is that early childhood sexual experiences with older men or women are quite common in the histories of Latino MSM.

The Perception of Sexual Abuse Among Men With CSEOP

The experiences of these men varied widely, from a single occasion to ongoing relationships. In the vast majority of the cases the older partner was a male. Studies have shown that men who had been victimized by males are much more likely to identify as gay or bisexual than their peers who were not abused (Holmes & Slap, 1998). However, no causal direction has been established. It may be that effeminate boys or boys exhibiting gender role nonconformity are more likely to be pursued by older men. The behavior reported by these men commonly involved mutual fondling, although slightly more than half of the children fellated the older partner. Anal sex was also not uncommon, with over a third of the men reporting having been penetrated.

Several aspects of their experiences did not distinguish those who felt sexually abused from those who did not. The age of the older partner, whether the partner was a male or a relative, the extent of the relationship in terms of number of occasions or length of time, and the sexual behavior engaged in (excepting kissing and the child exposing himself) did not differ for the two groups. However, those who felt they were abused were much more likely to have been coerced, both by threats and physical force. Those who felt abused also felt they were hurt physically and emotionally by the experiences more so than those who did not feel abused. Age of the child at the time of the experiences was also associated with the perception of abuse; the men who felt abused had the experiences approximately 2 years earlier than those who did not. These two themes--young age and the child's willingness to participate--were also commonly expressed in the qualitative responses to why or why not the respondents felt their experiences constituted sexual abuse.

Of course, the participants' perception of sexual abuse does not determine whether the experiences were actually childhood sexual abuse. Some men reported having been forced and hurt by the experiences but still did not think they were sexually abused. Others did not report any coercion and were not hurt emotionally or physically, but still felt that they had been abused. Indeed, many people would consider all of these experiences childhood sexual abuse, feeling that any child younger than age 13 is not capable of engaging in consensual sex. Nevertheless, there is a substantial amount of sexual activity at a young age with older partners that is not perceived to be abusive by the men who experienced it. For this sample of men, a perception of abuse is associated with coercion and the age of the child. This, suggests that these factors should be central to any definition of childhood sexual abuse and assessed in research on this topic. However, similar research is necessary with other populations (heterosexuals, women, other ethnic groups) to determine whether similar patterns exist across samples.

Adulthood Correlates of Childhood Sexual Experiences With an Older Partner

Regarding the issue of harm, the ability to distinguish between experiences that are perceived to be abusive from those that are not was useful in this study. For all of the psychological/behavioral variables we were able to assess, there was a continuum evident: Those who had CSEOP but did not consider the experiences to be sexual abuse in every case had scores that were between those who considered themselves abused and those who did not have CSEOP. It appears that more traumatic experiences are associated with more negative outcomes (it could be debated whether higher numbers of male sex partners is a negative outcome; we have considered it as such only because of the greater potential for HIV transmission).

Those who felt they were the victims of abuse (i.e., those who were more likely to have been forced and hurt by the experiences) had higher rates of unprotected anal sex (replicating our previous report involving Puerto Rican MSM: Carballo-Dieguez & Dolezal, 1995), had more male sex partners, and had more alcohol use in adulthood. Rates were not as high for those who did not feel that they had been abused; this group did not significantly differ from either other group. Lowest rates were reported by those who did not have an early sexual experience with an older partner. The severity of the experiences (as reflected in the perception of abuse) was related to the extent of negative scores for all of the psychological/behavioral variables assessed, though usually not significantly. This indicates that methodologically it is meaningful to collect data in a manner that allows for the experiences that make up this continuum to be distinguished. Although CSA may be most appropriately defined in terms of coercion and the child's age, assessing early sexual experiences that are not considered coercive may provide important information as well. With larger samples, the men with CSEOP who did not consider themselves sexually abused would likely have been statistically distinguishable on negative outcome variables from the other two groups assessed.

The relevance of considering childhood family environment is seen in the fact that those who felt they were abused had the most problematic family environment scores in this sample. The association between this score and the perception of abuse was actually stronger than that found for any of the psychological/behavioral variables assessed. Family environment did not, however, prove to have a very large effect as an explanatory covariate. Differences between unadjusted and adjusted results were quite subtle.

Limitations

In addition to being a convenience sample, a primary limitation to the generalizability of these findings is the uniqueness of the sample. The participants are ethnic minorities and many are immigrants (only 27% were born in the US). Their sexual orientation as gay and bisexual also imposes a minority status on them. In terms of traditional research on CSA, even the fact that they are males puts them in a minority category. Rind et al. (1998) found that women typically have more negative reactions to childhood sexual experiences compared to men. It is likely that a sample of Latina women with experiences similar to our sample would have been much more likely to consider such experiences to be sexual abuse. Similarly, a sample of heterosexual Latino men may have had quite different responses, particularly if the older partner were male. In that case, there may have been higher reports of coercion, threat, negative attitudes, and the perception of sexual abuse. But such comparisons are difficult and speculative, and, furthermore, we could also speculate about similarities. Heterosexual men who had sexual experiences in childhood with an older female may be just as likely to feel that they sought the experience and that it was consensual. It is feasible that those men also would not consider their experiences sexual abuse. In any case, the uniqueness of the sample must be considered along with the findings.

Another limitation stems from the fact that the primary focus of the parent study was on sexual risk behavior and differences among ethnic groups of Latino MSM, not on CSA. If the study had been designed to determine the correlates of CSA, additional measures would have been appropriate (e.g., measures of psychological distress, depression, posttraumatic stress disorder, etc.). We had some measures that were relevant, but a more definitive examination of the negative consequences of CSEOP would have required more extensive testing.

Finally, regarding 'cultural norms about sexuality, research typically is based on heterosexual samples and it is unclear how relevant these reports are to an urban sample of MSM. In fact, negative attitudes toward homosexuality are extremely common among Latinos. For example, in one survey of Latinos in nine U.S. states (Marin & Gomez, 1997), less than 20% agreed that sex between men is acceptable and almost two thirds said they believed homosexuality was distasteful. We are unable, therefore, to address the issue of how cultural/ethnic norms may influence the responses found here. We see no evidence in-the literature or in personal experience that these ethnic groups condone or encourage early childhood sexual experiences with older partners and do not feel that ethnic norms are the reason for such high rates of CSEOP, but there is a lack of research available to support this contention.

Conclusion

These data show that not all childhood sexual experiences with older partners are perceived as negative or are associated with damaging repercussions. This finding is, of course, not presented to condone these practices. In fact, the majority of the men who had such experiences were either forced or threatened or were physically or emotionally hurt (65% reported at least one of these). There were also indications that these experiences are associated with negative outcomes in adulthood, in particular with alcohol use and sexual behavior. Victims and clinicians are very familiar with the devastating effects that can be the result of the sexual abuse of children (Gonsiorek, Bera, & LeTourneau, 1994; Lew, 1990). However, our findings, consistent with the modest effect sizes for correlates of CSA reported by Rind et al. (1997, 1998), reflect that the negative consequences of early sexual experiences are not always pervasive or severe (at least regarding the limited measures available in this study), especially if the experiences were perceived as consensual.

Although this sample is unique, the findings suggest methodological approaches that deserve reiteration. First, the individual's perception of sexual abuse seems to be a useful thing to assess in such studies. We were able to identify certain factors that were associated with this perception, which should contribute to efforts at developing a definition of CSA. Also, distinguishing between those who did and did not feel they were abused was relevant in assessing the subsequent correlates of abuse. Second, the components of early sexual experiences (age, age/characteristics of partner, sexual behavior, responses to the event, coercion, etc.) should be assessed separately, rather than in a more global way. It may be easier to simply ask "Did you have genital contact before you were 13 with a person at least 4 years older than you that you did not want to have?," but there are limitations to such an approach. If the components are assessed separately, prevalence rates can be calculated using various criteria, making it easier to compare rates with other samples that use different definitions. This is illustrated in our ability to compare our sample with the UMHS. Third, rather than conceptualizing CSA as a dichotomous variable (i.e., contrasting those who do or do not meet some criteria for CSA), it seems useful to operationalize CSA as a continuum, with some indication of the severity of the experiences. In this report, distinguishing between those who believe they were abused and those who do not provided a basic indication of the severity of the experience that was relevant to the correlates of CSEOP. Those who felt they were abused had the most negative scores, while those who did not have CSEOP had the most positive, with the nonabused falling in the middle. Attempts to operationalize CSA as a continuum, reflecting the great variety of experiences, should prove useful in subsequent research.

Table 1. Sample Breakdown Regarding Childhood Sexual
Experiences With an Older Partner (CSEOP)

Total Sample                                        307    (100%)
Did not meet criteria for CSEOP because ...
   No sexual contact prior to 13th birthday         126    (41% of 307)
   Had sexual contact prior to 13th birthday but
   not with a partner who was 4 years older          81    (26% of 307)
Met criteria for CSEOP                              100    (33% of 307)

Note. CSEOP criteria = Sexual contact (participant's mouth, anus, or
genitals were in contact with those of another person; kissing on lips
not included) prior to 13th birthday with a partner 4 or more years
older than participant.

Table 2. Characteristics of Childhood Sexual Experiences
With an Older Partner (CSEOP)

                                                  Median (Range)     %

Age of participant at first occasion (years)       8.5 (3-12)
Age of partner (years)                            17.5 (9-60)
Number of occasions with this partner              4.0 (1-6000)
Length of sexual relationship with this partner
(months)                                           2.0 (0.03-240)
Partner was male                                                    91
Partner was a relative                                              34
Participant was physically forced                                   39
Participant was physically hurt                                     34
Participant was threatened                                          32
Participant was emotionally hurt                                    52
Reaction at the time
   Positive                                                         32
   Indifferent                                                      33
   Negative                                                         34
Current view of experience
   Positive                                                         35
   Indifferent                                                      27
   Negative                                                         38
Participant told someone what happened                              14
Participant considers it sexual abuse                               59

Note. N = 100.

Table 3. Sexual Behaviors Involved in Childhood Sexual
Experiences With an Older Partner (CSEOP)

                         Male Older Partner    Female Older Partner
                             N(% of 91)            N(% of 9)

OP fondled C                  74 (81%)             9 (100%)
OP exposed self               64 (70%)              8 (89%)
C masturbated OP              52 (57%)              4 (44%)
C fellated OP                 52 (57%)                NA
C fondled OP                  51 (56%)              6 (67%)
Anal sex, C receptive         36 (40%)                NA
OP masturbated C              32 (35%)              5 (56%)
C exposed self                30 (33%)              7 (78%)
Deep kissing                  27 (30%)              5 (56%)
OP fellated C                 23 (25%)              2 (22%)
OP rimmed C                   11 (12%)              1 (11%)
C rimmed OP                    1 (1%)               1 (11%)
Anal sex, C insertive          1 (1%)               1 (11%)
Vaginal sex                      NA                 6 (67%)
Cunnilingus                      NA                 2 (22%)

Note. OP = Older partner; C = Child (study participant).

Table 4. Significant Differences Between Men who Consider Their
Experiences Sexual Abuse and Men who Do Not

                                                     No, Not Abuse
                                                       (N = 41)

Median age of participant at first occasion (yrs.)      10.0
Participant was physically forced (N/%)                  4 (10%)
Participant was physically hurt (N/%)                    6 (18%)
Participant was threatened (N/%)                         2 (6%)
Participant was emotionally hurt (N / %)                10 (19%)

                                                      Yes, Abuse
                                                       (N = 59)

Median age of participant at first occasion (yrs.)       8.0
Participant was physically forced (N/%)                 35 (90%)
Participant was physically hurt (N/%)                   28 (82%)
Participant was threatened (N/%)                        30 (94%)
Participant was emotionally hurt (N / %)                42 (81%)

                                                       sig. (a)

Median age of participant at first occasion (yrs.)       .001
Participant was physically forced (N/%)                  .000
Participant was physically hurt (N/%)                    .001
Participant was threatened (N/%)                         .000
Participant was emotionally hurt (N / %)                 .000

* Mann-Whitney U test for continuous variable and Chi-square tests
for dichotomous variables.

Table 5. The Association Between Psychological/Behavioral
Variables and Childhood Sexual Experiences With an Older
Partner (CSEOP) and the Perception of Sexual Abuse

                                              CSEOP

                                None        Not Abuse      Abuse
                              (N = 207)     (N = 41)      (N = 59)

Self-Esteem (a)                  2.23          2.12         2.l0
Alcohol use (a)                  1.97          2.12         2.46
Drug use (a)                     0.21          0.24         0.26
Unprotected anal sex (a,b)      21.8          27.2         30.4
Male sex partners (a,b)         16.8          32.0         82.5
Sexual functioning (a)           3.97          3.89         4.06
Family environment (a)           1.00          1.25         1.33

                                ANOVA      ANACOVA (c)
                                sig.          sig.

Self-Esteem (a)                 .107          .264
Alcohol use (a)                 .051          .045
Drug use (a)                    .547          .625
Unprotected anal sex (a,b)      .042          .068
Male sex partners (a,b)         .031          .052
Sexual functioning (a)          .723          .403
Family environment (a)          .018           NA

Note. N = 307. CSEOP: None = did not have sexual contact prior to
age 13 with a partner that was at least 4 years older; Not Abuse = did
have CSEOP but does not consider the experiences to be sexual abuse;
Abuse = did have CSEOP and does consider the experiences to be sexual
abuse. (a) Higher means indicate higher self-esteem, more substance
use, more unprotected sex in past year, more male sex partners in
past year, better sexual functioning, and more problems in childhood
family environment, (b) Mean is of raw values which were
log-transformed prior to statistical test due to skewness.
(c) Test of differences adjusting for childhood family environment.

The authors wish to thank the men who gave of their time to participate in this study for sharing such personal information about their lives. We also wish to thank Luis Nieves Rosa, the project coordinator, and his staff, especially Francisco Diaz. Finally, thanks to Heino Meyer-Bahlburg for his comments in the preparation of this manuscript.

REFERENCES

Carballo-Dieguez, A., & Dolezal, C. (1995). Association between history of childhood sexual abuse and adult HIV-risk sexual behavior in Puerto Rican men who have sex with men. Child Abuse and Neglect, 19, 595-605.

Carballo-Dieguez, A., Dolezal, C., Nieves-Rosa, L., & Diaz, E (1995). Sexual Practices Assessment Schedule-LMSM Version (SPAS-MSM). Available from the authors.

Carballo-Dieguez, A., Stein, Z., Saez, H., Dolezal, C., Nieves-Rosa, L., & Diaz, F. (2000). Frequent use of lubricants for anal sex among men who have sex with men: The prevention potential of a microbicidal gel. American Journal of Public Health, 90, 1117-1121.

Dolezal, C., Carballo-Dieguez, A., Nieves-Rosa, L., & Diaz, F. (2000). Substance use and sexual risk behavior: Understanding their association among four ethnic groups of Latino men who have sex with men. Journal of Substance Abuse, 12, 1-14.

Freire, Paulo. (1990). Pedagogy of the oppressed. New York: Continuum.

Fromuth, M. E., & Burkhart, B. R. (1987). Childhood sexual victimization among college men: Definitional and methodological issues. Violence Victims, 2, 241-253.

Gonsiorek, J. C., Bera, W. H., & LeTourneau, D. (1994). Male sexual abuse. Thousand Oaks, CA: Sage Publications, Inc.

Goode, E. (1999, June 13). Study on child sex abuse provokes a political furor. New York Times, p. 1-33.

Holmes, W. C., & Slap, G. B. (1998). Sexual abuse of boys: Definition, prevalence, correlates, sequelae, and management [review]. The Journal of the American Medical Association, 280, 1855-1862.

Jumper, S. A. (1995). A meta-analysis of the relationship of child sexual abuse to adult psychological adjustment. Child Abuse and Neglect, 19, 715-728.

Lew, M. (1990). Victims no longer: Men recovering from incest and other sexual child abuse. New York: Harper Collins Publishers, Inc.

Marin, B. V., & Gomez, C. A. (1997). Latino culture and sex: Implications for HIV prevention. In J. G. Garcia & M. C. Zea (Eds.), Psychological interventions and research with Latino populations (pp. 73-93). Boston: Allyn and Bacon.

Neumann, D. A., Houskamp, B. M., Pollock, V. E., & Briere, J. (1996). The long-term sequelae of childhood sexual abuse in women: A meta-analytic review. Child Maltreatment, 1, 6-16.

Nieves-Rosa, L., Carballo-Dieguez, A., & Dolezal, C. (2000). Domestic abuse and HIV-risk behavior in Latin American men who have sex with men in New York City. The Journal of Lesbian and Gay Social Services: Issues on Practice, Polio, and Research, 11, 77-90.

Okami, P. (1991). Sell-reports of "positive" childhood and adolescent sexual contacts with older persons: An exploratory study. Archives of Sexual Behavior, 20, 437-457.

Paul, J., Catania, J., Pollack, L., & Stall, R. (2001). 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 and Neglect, 25, 557-584.

Reynolds, C. F., III, Frank, E., Thase, M. E., Houck, P. R., Jennings, J. R., Howell, J. R., et al. (1988). Assessment of sexual function in depressed, impotent, and healthy men: Factor analysis of a brief sexual function questionnaire for men. Psychiatry Research, 24, 231-250.

Rind, B., & Tromovitch, P. (1997). A meta-analytic review of findings from national samples on psychological correlates of child sexual abuse. The Journal of Sex Research, 34, 237-255.

Rind, B., Tromovitch, P., & Bauserman, R. (1998). A meta-analytic examination of assumed properties of child sexual abuse using college samples. Psychological Bulletin, 124, 22-53.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Sandfort, T. G. M. (1984). Sex in pedophiliac relationships: An empirical investigation among a nonrepresentative group of boys. The Journal of Sex Research, 20, 123-142.

Manuscript accepted January 9, 2002

Address correspondence to Curtis Dolezal, New York State Psychiatric Institute, 1051 Riverside Drive, Unit #15, New York, NY 10032; e-mail: [e-mail].

Curtis Dolezal and Alex Carballo-Dieguez New York State Psychiatric Institute and Columbia University