Gay and bisexual men's age-discrepant childhood sexual experiences.

Despite a proliferation of research on the impact of child sexual abuse (CSA), minimal effort has been directed toward understanding gay men's experiences of CSA. According to the best available prevalence estimates, 27% of women and 14% of men in both college and community samples have experienced CSA (Rind, Tromovitch, & Bauserman, 1998). Initial investigations indicate that CSA rates in gay and bisexual men may be at least as high as prevalence rates found among women. For example, Doll et al. (1992) found that 31% of gay and bisexual men attending sexually transmitted disease (STD) clinics reported sexual encounters before age 16 with an older or more powerful partner. Similarly, Lenderking et al. (1997) found that 36% of the gay and bisexual men participating in a health clinic study reported a childhood sexual encounter before age 16 with an older person.

Although relatively high rates of CSA among gay men have been consistently reported, little attention has been paid to the nature of these experiences. Considerable research indicates that boys' and girls' experiences of CSA differ with regard to the nature of the sexual activity, perceptions, and effects (e.g., Dhaliwal, Gauzas, Antonowicz, & Ross, 1996; Rind et al., 1998). Evidence that gay men report higher rates of CSA than heterosexual men (Coxell, King, Mezey, & Gordon, 1999) and the suggestion that their sexual encounters are more likely to be with older male partners indicate that gay and heterosexual men's experiences of CSA also may systematically differ (West & Woodhouse, 1993). Further, the development of young gay men's sexuality differs from that of heterosexual men, given the lack of social support for their sexual orientation and the lesser availability of same-age sexual partners (Fisher & Akman, 2002; Savin-Williams, 1998). Thus, it cannot be assumed that the study of CSA in women and heterosexual men is relevant to the experiences of gay men, leaving researchers and clinicians alike with little information to guide their work. To address this gap in the CSA research, we examined gay and bisexual men's experiences of CSA and the impact of those experiences.

This research incorporated the conventional definition of CSA and a modified definition of CSA to evaluate which definition best accounted for problems in adjustment. CSA is typically defined as a sexual interaction between a child or adolescent and a person who is at least 5 years older (Rind et al., 1998). This age-based definition maps onto moral beliefs and the American legal criterion of CSA (Kilpatrick, 1987; Nelson & Oliver, 1998; Okami, 1991). However, the age-based definition aggregates all children's sexual activity with older persons and thus masks the degree to which experiences can vary. For example, incestuous and coercive sexual activity that involves a young child is not distinguished from consensual sexual activity between an adolescent and an unrelated adult. Focusing solely on whether an age-discrepant sexual activity occurred, as directed by the age-based criterion, diverts attention away from examining the nature and context of the sexual activity. However, the characteristics of a sexual encounter appear to better explain adjustment than the occurrence of a sexual encounter. Notably, the relationship with the perpetrator and the degree of coercion are related to intensity of negative effects (e.g., Constantine, 1981). For gay men, negative perceptions of their sexual experiences are related to age at the time of experience and to the presence of coercion (Dolezal & Carballo-Dieguez, 2002; Doll et al., 1992). Thus, nonsexual factors such as assent and age of the child and the older person may better explain children's adjustment to CSA than simply whether age-discrepant sexual activity occurred.

The age-based definition of CSA is based on an implicit assumption that USA invariably leads to harm, an assumption that has limited empirical support. A recent meta-analysis revealed that only a small minority of men (unweighted mean = 9%) and women (unweighted mean = 13%) with CSA histories report lasting negative effects (Rind et al., 1998). Although on average people with CSA experiences are slightly less well-adjusted than those without such experiences, CSA accounts for less than 1% of the variance in adjustment (Bauserman & Rind, 1997; Rind et al., 1998). Furthermore, young gay men who described an age-discrepant sexual experience between the ages of 12 and 17 were found to have equally well-developed self-esteem and sexual identity as young gay men without such experiences (Rind, 2001).

Although it is commonly assumed thai CSA is invariably an intensely negative experience, research consistently shows considerable variability in retrospective perceptions of CSA experiences. Moreover, these perceptions predict associated outcomes. For example, individuals who described negative childhood sexual experiences were more likely to report difficulties (such as decreased sexual satisfaction, poorer general functioning, and suicidal ideation) than individuals who described positive or neutral experiences (Okami, 1991). In addition, among inmates (Fondacaro, Holt, & Powell, 1999) and among Latino men who have sex with men (Dolezal & Carballo-Dieguez, 2002), cognitively appraising an age-based CSA experience as abusive was found to be associated with worse psychological outcomes than not appraising the experience as abusive. Among college men, those who labeled their childhood sexual experiences with older persons as abusive reported psychological distress at levels nearly double those reported by men who had age-based CSA experiences but who did not label these experiences as abusive (Steever, Follette, & Naugle, 2001).

Not surprisingly, evaluating a sexual encounter with an older person negatively appears to be related to the degree of coercion. For males, sexual activity involving force or coercion is more likely to be interpreted as negative compared to noncoercive sexual activity (Constantine, 1981; Okami, 1991). Further, when boys' sexual experiences with adults are coercive, boys report a similar degree of adjustment problems as do girls (Rind et al., 1998). Coxell and associates (Coxell et al., 1999) found that slightly less than half of men's experiences labeled as CSA by age-based criteria were coercive and perceived as abusive. Whereas coercive experiences were predominately with men and experienced at a mean age of 11 years, the majority of noncoercive experiences were with women at a mean age of 14 years. However, this sample was overwhelmingly heterosexual. It has been suggested that boys may tend to be open and active in their sexual encounters with older females, perhaps because actively seeking sex is consistent with the male sex role (Fondacaro et al., 1999; Nelson & Oliver, 1998). However. this interpretation assumes that the sexual orientation of the youth is heterosexual. Given that the majority of boys who have sexual encounters with older females report a positive experience (Fromuth & Burkhart, 1987: West & Woodhouse, 1993), some homosexual boys may perceive their sexual experiences with older men as positive (cf. Rind, 2001).

All previous studies examining CSA and associated outcomes in gay and bisexual men have relied on convenience samples. For example, Doll et al. (1992) studied a sample of gay and bisexual men attending STD clinics in large U.S. cities, and Rind (2001) examined Savin-Williams' (1998) sample of volunteers from university classes, local establishments, and gay newsletters and listserves. To extend the generalizability of previous findings, we examined CSA in gay and bisexual men drawn from a randomly selected community sample.

By defining CSA solely on an age-related basis, differences in experiences that affect outcome are relegated to a position of secondary importance. Therefore, we adopted a perception-based definition of CSA (CSA-P) which refers to age-discrepant sexual experiences perceived as negative, coercive, and/or abusive. In contrast, the term childhood sexual experience (CSE) refers to age-discrepant sexual experiences that were perceived positively or neutrally from a childhood perspective, that did not involve coercion, and that were not perceived as abusive. To evaluate the predictive validity of the age-based and perception-based definitions of CSA, we examined the associations between these experiences and three areas of adult adjustment: self-esteem, hopelessness, and interpersonal problems. We predicted that perception-based experiences would be more strongly associated with maladjustment than age-based CSA experiences.

METHOD

The present study was part of the West End Relationships Project (WERP), which examined gay men's relationship experiences. For the first phase of WERP, 300 gay and bisexual men living in the West End of Vancouver, British Columbia, Canada, were selected through a random digit dialing procedure and administered a 15-minute telephone survey. The overall response rate for the first phase of the study, calculated as the number of interviews completed divided by the number of known eligible respondents, was 49%. For the second phase, 192 gay and bisexual men (64%) from the original sample of 300 completed an in-person interview and a series of questionnaires. Phase 2 participants received a $20 honorarium.

Participants

The sample consisted of 192 gay and bisexual men who participated in the second phase of the WERP study. Canadian census data do not include sexual orientation and thus we cannot assess the representativeness of this sample. However, there were no notable differences between the current sample and the Phase 1 sample of randomly selected men on standard demographic variables, including age, ethnicity, income, and education. Participants ranged in age from 20 to 71 years with a mean of 39 years (SD = 9.37). Sexual orientation was identified on a 7-point continuum ranging from 1 = exclusively gay to 7 = exclusively heterosexual. Eighty-one percent identified as exclusively gay, 15% identified as predominately gay, 3% identified as somewhat more gay than heterosexual, 1% identified as bisexual, and one participant identified as somewhat more heterosexual than gay. Participants' self-reported ethnicity was British Commonwealth (46%), other European (28%), other Canadian (13%), French Canadian (5%), and other (9%). The distribution of personal income in Canadian dollars was as follows: less than $20,000 (17%), $20,000 to $29,999 (18%), $30,000 to $39,999 (24%), $40,000 to $49,999 (16%), and over $50,000 (24%).

Measures

CSA Measures

History of Attachments Interview (Henderson, 1998). The questions relevant to this study were embedded in the History of Attachments Interview. This semistructured interview consists of two sections, the first focusing on family relationships and the second focusing on close friendships and romantic partnerships. Interviews took 2 to 2 1/2 hours to complete and were conducted at the WERP office in Vancouver's West End. Within the family section, we asked participants, "As a child or adolescent, did you have any sexual contact with an adult or older person?" Participants who responded affirmatively were asked to describe their experiences. Further probing concerned the participant's age at time of the sexual encounter, the relationship with and age of the older person, and the participant's perceptions of the experience.

Coding. Two trained coders listened to the family section of the taped interviews and independently coded the 50 interviews that included descriptions of sexual experiences occurring before the age of 17 with someone at least 5 years older (an age-based CSA experience). We assessed reliability of ratings on 31 of the 50 interviews (62%) containing descriptions of age-based CSA. However. due to variability in the degree of detail provided by participants, the ns fluctuate across variables. Interrater agreement was assessed using percent agreement and Cohen's kappa for categorical variables, and Pearson's correlation coefficient for continuous variables.

Childhood sexual encounters with older persons were the focus of the coding. If a participant reported more than one CSA experience, we coded the earliest experience because of its possible influence on later sexual encounters. If a participant described more than one incident within the focal relationship, we coded the most extreme sexual act because we assumed it had the strongest impact on the participant's experience. We considered contact sexual encounters (e.g., fondling or intercourse) and non-contact sexual encounters (e.g., exposure to pornography or exhibitionism). However, all men described contact sexual experiences and all but one man reported experiences involving genital contact.

The relationship between the older person and youth was categorized as parent, sibling, extended family, acquaintance, family friend, stranger, or community member ([kappa] = .95, 96% agreement, n = 25). We rated the length of time over which the sexual experiences occurred according to the following scale: once, twice, a few times, 1-8 weeks, 2-4 months, 5-12 months, and more than 1 year. Interrater reliability for the duration of the sexual activity was r(26) = .97. The presence of coercion in participants' description of the sexual activity was coded on a yes-no basis ([kappa] = 1.00, 100% agreement, n = 29), and participants' childhood perceptions of the sexual experience as abusive were coded on a yes-no basis ([kappa] = .87, 97% agreement, n = 30). Participants' current perceptions of the sexual experience as abusive were also coded on a yes-no basis ([kappa] = .76, 90% agreement, n = 31), while participants' perceptions of the valence of the experience were rated as either negative or positive/neutral ([kappa] = .86, 93% agreement, n = 28). Perceptions judged as negative included sexual experiences that were described as disturbing, traumatic, or negative in any way. We included experiences that were described as involving both positive and negative aspects in the negative category.

Participants' sexual experiences before age 17 with someone at least 5 years older were categorized as age-based CSA. We further coded those experiences identified as age-based CSA according to participants' perceptions of the experience. Perception-based CSA consisted of experiences that were coercive, were perceived as negative, and/or were perceived as abusive from either a child or an adult perspective. In contrast, a childhood sexual experience (CSE) consisted of age-discrepant sexual activity that was not described as abusive or coercive and that was described positively or neutrally ([kappa] = .80, 90% agreement, n = 31).

Adjustment Measures

Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965). The RSES is a 10-item sell-report measure designed to assess self-acceptance and self-worth. Items were scored on a 4-point Likert-type scale ranging from 1 (strongly disagree) to 4 (strongly agree), with higher total scores indicating higher self-esteem. Cronbach's alpha for the total score was .87.

Beck Hopelessness Scale (BHS; Beck, Weissman, Lester, & Trexler, 1974). This scale measures respondents' negative expectations about self and future prospects. Hopelessness is proposed to be a core component of depression (Beck et al., 1974). The BHS is a 20-item, true-false, self-report scale. An example item is "Things just won't work out for me the way I want them to." The total score had a Cronbach's alpha of .94.

Inventory of Interpersonal Problems (IIP; Horowitiz, Rosenberg, Baer, Ureno, & Villasenor, 1988). The IIP is a 64-item self-report inventory assessing interpersonal problems. A set of eight scales, composed of 8 items each, assess interpersonal problems in the areas of dominance, competitiveness, coldness, introversion, passivity, exploitability, nurturance, and expressiveness. For example, a high nurturance score indicates difficulties with being overly nurturing in relationships. Each item is rated on a 0 (not at all) to 4 (extremely) Likert scale. The coefficients for the eight scales ranged from .72 to .88 with a mean of .81. Cronbach's alpha for the total score was .95.

RESULTS

Prevalence and Nature of Childhood Sexual Abuse

Fifty of the 192 men (26%) reported sexual contact with someone at least 5 years older before the age of 17. On average, the men were 10.10 years of age (SD = 4.45) at the time of sexual contact with a range of 2 to 16 years. The age of the older person ranged from 11 to 60 years with a mean of 24.61 (SD = 8.70). The age difference between the youth and older person averaged 14.25 years (SD = 7.71) and ranged from 5 to 45 years difference. The vast majority of older sexual partners were men (92%) with only 4 (8%) of the 48 men who indicated the sex of the older person reporting that it was a woman. Most commonly, the sexual contact occurred only once (n = 21,48%), although the duration of sexual contact ranged up to 12 years. Ten of the 45 men (22%) who described the duration of sexual contact reported that the sexual contact lasted for more than 1 year.

Forty-five men indicated they had a relationship with the older person (see Table 1). Fifteen men (33%) reported that the older person was a member of their community (e.g., shopkeeper, leader of a youth organization, or babysitter). Family friends (n = 7, 16%) and strangers (n = 7, 16%) were the next most commonly mentioned relations. Sexual contact with a member of the extended family (e.g., cousins and uncles) was reported by five men (11%). Another five men (11%) described the relationship as an acquaintanceship. Four men (9%) reported sexual relations with older brothers, and two men reported sexual relations with fathers (4%).

Table 2 presents participants' perceptions of their CSA experiences. The majority of men did not describe the experience as coercive, with only 6 of the 47 men (13%) indicating that they were forced to some degree to engage in sexual activity. Several men reported that they engaged in the sexual activity consciously and willingly. For example, some men reported going to an area that gay men were known to frequent with the intention of having a sexual encounter. However, in several men's experiences, they were too young to understand sex and as a result engaged in the experience unwittingly or in the spirit of exploration. For example, one man reported that he did not think that anything was wrong about his sexual encounters with his older brother until his parents reacted strongly when he cheerfully told them, "Bro kissed my dinky."

The majority of men (n = 41, 82%) reported that at the time of sexual contact they did not perceive the experience as abusive. Of the 48 men who described their childhood perception of the sexual experience, 20 (42%) perceived the encounter negatively. However, the degree of negativity varied and included descriptions such as "the worst thing in my life" to mixed perceptions acknowledging good and bad aspects of the experience. Twenty-eight men (58%) described that as a child they perceived the experience positively or neutrally. The degree of positivity also varied and ranged from "wonderful, affirming, and hot" to more neutral and nonnegative descriptions of the sexual contact. From an adult perspective, the majority of men (31 out of 50; 62%) did not define their experience as abusive. In addition to the 9 men who continuously perceived their CSA experiences as abusive, 10 men changed their characterization of the experience from nonabusive to abusive. Of the 10 men who came to perceive their experience as abusive, many indicated that as a child they "did not understand" the experience and that they are now viewing it from an "adult perspective."

Based on men's descriptions, we placed their experiences into the categories of perception-based CSA (CSA-P) or childhood sexual experience (CSE). Sexual encounters that (a) were perceived as negative, and/or (b) involved coercion, and/or (c) were perceived as abusive from a childhood or adult perspective were labeled as perception-based CSA (CSA-P). Sexual encounters that the men experienced as (a) positive or neutral, (b) nonabusive, and (c) noncoercive were placed into the CSE category. We did not include one man's age-discrepant sexual experience (he performed oral sex on his father while his father was passed out) in either perception-based CSA or CSE. This man perceived the experience as negative and therefore it could not be categorized as CSE. However, it was not categorized as CSA-P because under the circumstances the older adult could not be considered abusive; in fact, the participant is more accurately considered the abusive partner.

Twenty-four of the 49 men (49%) described age-based CSA experiences that were also rated as a perception-based CSA experiences. Twenty-five men (51%) related that their age-based CSA experiences were neither abusive, negative, nor coercive and thus we coded the experiences as CSE. Notably, 9 of the 10 men who originally did not define their experience as abusive but eventually came to perceive it as abusive reported coercion (4 men) and/or a negative perception (7 men) and thus would have been included in the perception-based CSA category regardless of their current perceptions of abuse.

We compared the nature of the sexual experiences considered perception-based CSA and CSE. We found men in the CSE category to be older at the time of sexual contact (M = 11.64 years, SD = 3.74) than men whose experiences were coded as perception-based CSA (M = 8.17 years, SD = 4.44), t(46) = 2.93, p < .01. Men in the CSE category also reported a smaller age difference between them and the older persons (M = 11.88 years, SD = 4.28) than did men in the perception-based CSA category (M = 16.68 years, SD = 9.88), t(45) = 2.11, p < .05. In addition, a trend suggested that the relationships between the youth and older persons were different for the perception-based CSA group than for the CSE group (see Table 1), [chi square](6) = 11.81, p < .07. Unfortunately, the ns in the relationship categories were too small to allow for statistical comparisons and thus these differences can only be commented on. Notably, only in the CSE category was the older person a stranger (n = 7). We found no differences between the groups in the age of older person, t(46) = -0.71, ns, or the duration of sexual contact, 1(42) = -0.70, ns. Given that the vast majority of older persons were men, comparisons were not conducted on the gender of the older person.

Current Adjustment

Table 3 presents the means of the adjustment variables for the CSA and non-CSA groups. First, we compared the age-based CSA group (the CSE and CSA-P groups combined) and the non-CSA group on adult adjustment. Three of 11 planned comparisons on measures of current adjustment were significant: Participants with age-based CSA histories reported lower self-esteem, t(190) = -1.87, p < .05, more problems with expressiveness, t(190) = 2.60, p < .01, and more problems with nurturance, t(190) = 2.60, p < .01, than those without CSA histories. The effect sizes for the RSE (self-esteem) and the IIP subscales of expressiveness and nurturance were small to medium (d = .35, .42, .42).

Next, we compared the perception-based CSA group and the non-CSA group to assess whether participants with a perception-based CSA experience reported greater maladjustment than participants who did not report an age-discrepant sexual experience. The results were largely similar to the comparisons between the age-based CSA and non-CSA groups. We found significant group differences for reports of problems regarding nurturance, t(28) = 2.33, p < .05, d = .61, and expressiveness, t(164) = 3.34, p < .001, d = .73, and for the mean level of interpersonal problems, t(28) = 1.84, p < .05, d = .49, indicating that the perception-based CSA group reported more distress than the non-CSA group. In contrast, there were no significant differences in adjustment between those who had CSE experiences and those who did not have an age-discrepant sexual experience.

To assess whether perceptions of age-discrepant childhood sexual experiences were associated with maladjustment, we compared the perception-based CSA and the CSE groups (see Table 3). There were differences between the groups on four indicators of interpersonal problems--competitiveness, t(47) = 2.25, p < .05; coldness, t(47) = 2.08, p < .05; expressiveness, t(47) = 1.95, p < .05; and mean level of interpersonal problems, t(47) = 1.91, p < .05--indicating that the perception-based CSA group reported more adjustment problems than the CSE group. The effect size for competitiveness was d = .67, for coldness was d = .65, for expressiveness was d = .61, and for mean level of interpersonal problems was d = .60.

Regression analyses tested whether the category of perception-based CSA significantly predicted adjustment beyond the variability accounted for by the category of age-based CSA (see Table 4). For these regression analyses, age-based CSA was entered first followed by perception-based CSA in predicting indicators of adjustment. We conducted these regression analyses on only the adjustment measures that were significantly associated with age-based CSA and/or perception-based CSA (i.e., self-esteem, competitiveness, coldness, nurturance, expressiveness, and mean level of interpersonal problems). Four of the six analyses indicated that perception-based CSA predicted adjustment to a significant degree, even alter the variance associated with age-based CSA was removed.

DISCUSSION

Twenty-six percent of the gay and bisexual men forming this sample reported a sexual encounter when they were 16 years or younger with a person at least 5 years their senior. This prevalence rate of age-based CSA is consistent with rates found in previous samples of gay and bisexual men (Cameron & Cameron, 1995; Doll et al., 1992; Lenderking et al., 1997). However, the age-based criterion aggregates a wide variety of experiences into one category and, therefore, this prevalence rate is somewhat misleading. Only half of our participants experienced these events as negative as a child, in sharp contrast to the current conceptualization of CSA as uniformly negative (cf. Rind et al., 1998). Although these experiences are typically defined as abusive by professionals such as researchers, clinicians, and lawmakers, only 38% of the participants agreed with this definition. Further, only 13% of these experiences involved coercion.

Reconceptualizing CSA according to a perception-based definition that focused on childhood sexual encounters with older persons that were coercive or perceived as abusive or negative cut the prevalence rate to 12.5%. The other half of the age-based CSA experiences described by participants were not coercive, were not considered abusive, and were experienced as positive or neutral. Therefore, we categorized these latter experiences as childhood sexual experiences (CSE) rather than CSA. The percentage of age-based CSA experiences we categorized as CSE (52%) echoes Coxell et al.'s (1999) finding that 59% of men's sexual experiences defined by British law as childhood sexual abuse were consensual. Also, Savin-Williams (1998) found that gay men's descriptions of their age-based CSA experiences were evenly split, half positive and half negative.

Several findings in this study support using the modified definition of CSA based on perceptions of the sexual experience instead of the current age-based definition. Comparisons of experiences categorized as perception-based CSA and CSE indicate important differences in the nature of the sexual encounters. Participants with perception-based CSA experiences were younger at the time of sexual contact, reported a larger age diFference between themselves and the older persons, and may have had different relationships with the older person. Similarly, Steever et al. (2001) found that men who labeled their age-discrepant sexual experience as abusive were younger at the time and were more likely to have sexual experiences with men and with Family members than were men who did not label their age-discrepant sexual experience as abusive. Doll and colleagues (1992) also found that child age and age difference were associated with negative perceptions of childhood sexual experiences in a sample of gay men. However, Rind (2001) found that for adolescent gay and bisexual boys, a greater age difference was not associated with a negative reaction and instead was associated with consent.

Given that participants who reported positive and noncoercive childhood sexual experiences were generally older, they may have been better able to understand and consent to sexual activity. Several men described an awareness of their sexual orientation and actively seeking out a sexual contact with another male. Men with a childhood sexual experience often described an encounter with a stranger or someone familiar to them that was pleasurable and exciting because it was their first sexual experience consistent with their sexual orientation. For example, one man, at 14 years old, researched the location of gay bars and met a man with whom he had a few sexual encounters and with whom he has maintained a friendship for nearly 20 years. Several of the sexual encounters described, particularly those in the acquaintance and stranger relationship categories, occurred in the context of developing romantic and sexual relationships in which the only apparent distinctive Feature was the age difference between partners. One man fondly recalled his "first relationship with a man" and opposed the label of abuse. Stories such as these are in sharp contrast to sexual encounters that the participants engaged in before they had developed a sense of their own sexuality or in which they felt forced or manipulated.

Evidence that perception-based CSA experiences were more strongly associated with adjustment problems than were age-based CSA experiences supports the proposition that the perception-based definition more accurately reflects childhood sexual abuse in gay and bisexual men. Although adjustment difficulties were associated with the age-based criterion, these results were largely a function of the perception-based CSA group's reported difficulties. In contrast, no differences in adjustment were found between participants with CSE histories and participants who did not report an age-based CSA experience. Additionally, the perception-based definition predicted maladjustment in four areas of interpersonal difficulties over and above that predicted by the age-based criterion. Given this pattern of results, the age-based definition may limit our ability to detect maladjustment reported by those with histories of negative and possibly more traumatic experiences (cf. Steever et al., 2001).

The reports of distress from men who experienced negative encounters should not be minimized. Participants with negative sexual experiences reported difficulties with competitiveness, coldness, expressiveness, and general interpersonal problems to a greater degree than participants with nonnegative, noncoercive sexual experiences. Examining the items associated with these scales reveals that participants with perception-based CSA experiences are ambivalent about relationships. Difficulties with expressiveness indicate attention seeking, excessively disclosing personal information, and pursuing knowledge of others' personal information. These behaviors serve to establish contact and a sense of intimacy with others in a maladaptive and possibly enmeshed way. In contrast, reports of competitiveness and coldness indicate difficulties trusting, supporting, and being close to others. These behaviors serve to establish and maintain a protective distance and possibly indicate an avoidance of closeness in relationships. This pattern of interpersonal problems is consistent with the conflicting drives for close relationships and for protective distance from others commonly found among trauma victims (Herman, 1992).

Although perception-based CSA experiences were associated with some adjustment difficulties, the magnitude of the associations was small. For the six outcome measures associated with CSA, only 2% to 5% of the variance in adjustment was accounted for by both age-based and perception-based CSA. These findings are consistent with Rind et al.'s (1998) finding that age-based CSA explained less than 1% of the adjustment variance in college samples. Therefore, neither age-based nor perception-based CSA can be assumed to produce lasting and pervasive negative effects in gay and bisexual men. However, some men in this sample reported lasting negative effects from their childhood sexual encounters, including one man who said "It ruined my whole life." Future qualitative explorations of the stories of gay men reporting such lasting effects could help to shed light on the factors determining when age-discrepant sexual experiences are likely to be associated with harmful outcomes.

The development of a scientific understanding of childhood sexual abuse is at a crossroads. On one hand, these findings indicate that a perception-based definition of CSA consisting of experiences perceived as negative. coercive, and/or abusive more accurately predicts deleterious outcomes than an age-based definition in gay and bisexual men. These results echo previous studies of primarily heterosexual samples which indicate that perceptions of the experience predict adjustment (e.g., Okami, 1991; Rind et al., 1998; Steever et al., 2001). On the other hand, Western societal values denounce children's age-discrepant sexual experiences as inappropriate on the basis that children are not emotionally or cognitively mature enough to comprehend the complexity of sexual relations. Although these assumptions may be accurate, empirical evidence indicates that age-discrepant childhood sexual experiences are not necessarily harmful (e.g., Constantine, 1981; Rind et al., 1998; Steever et al., 2001). Therefore, it must be acknowledged that a violation of social norms, which is the basis for the age-based definition, does not necessarily result in harm. A definition of CSA based on social norm violations is further problematic for same-sex relations because same-sex sexual activity is considered a social norm violation by many.

However, research revealing an association between child age and perception of the age-discrepant sexual experience suggests that the assumption regarding children's vulnerability to harm based on their cognitive and emotional maturity may have merit. The present study found that experiences perceived as negative or abusive were more likely to have occurred at a younger age, whereas experiences perceived as positive and not abusive were more likely to have occurred at an older age (cf. Coxell et al., 1999; Steever et al., 2001). Consistent with this finding, Rind et al. (1998) suggest that children's and adolescents' sexual encounters be distinguished, since adolescents are more likely than children to recognize sexual desires and to understand the implications of these desires and how to decline sexual advances.

Although child age appears to be an important factor, it is not a reliable indicator of whether a childhood sexual experience will have lasting negative effects (e.g., Bauserman & Rind, 1997; Beitchman et al., 1992; Constantine, 1981). Children may perceive an experience as negative or positive at any age. In this study, there was no younger age at which participants reported only abusive, negative, or coercive experiences. Nor was there a clear age boundary at which participants stopped reporting abusive, negative, or coercive experiences. Thus, perception-based CSA experiences and CSE overlapped considerably on child age. These findings imply that it is not important only that the child engaged in sexual activity with an older person before an appropriate age, but also that the nature of the experience is an important indicator of potential for harm.

The distinction between age and perception in defining CSA has particular relevance for gay men. Some in the gay community believe that some sexual experiences involving mature adolescents and older partners may be beneficial (e.g., Sandfort, 1983; Savin-Williams, 1998). Several arguments can be made supporting this position. These sexual experiences may provide these adolescents with the opportunity to explore their sexuality and feel affirmed by the gay community. Gay youth often speak of feeling different from their childhood peers and unaccepted by the dominant culture. It may be less threatening for young gay males to seek out an older gay male than to risk rejection and possible humiliation from making sexual advances toward a peer (cf. Savin-Williams, 1998). A sexual advance toward a peer may be dangerous for a gay youth if it is responded to with physical aggression, outing to the larger group of peers, and/or social rejection (Fisher & Akman, 2002). Therefore, the assumptions of the heterosexual community may not apply to gay youth when it is that very community which does not allow gay youth an outlet to explore their sexuality.

The randomly selected community sample is a considerable strength of this study. Unfortunately, we cannot establish the representativeness of the sample because Canadian census data does not include sexual orientation. At best, these findings can only be generalized to men who live in an urban area with a large gay community (25% of men in the area) and who identify as gay or bisexual. It is important that these findings be replicated in broader samples, including gay men living in rural areas.

This study relied on retrospective reports about sexual behavior, which may be prone to distortions (Abramson, 1990). Some clinicians have suggested that all childhood age-discrepant sexual experiences perceived neutrally or positively are actually a refraining of a traumatic experience to facilitate coping (e.g., Russell, 1986). However, in the current study only one participant's evaluation of his sexual experience seemed inconsistent with the details of the event. Despite his positive perception, the presence of coercion in his description resulted in his experience being coded as perception-based CSA. In addition, some stories involved a seemingly natural progression of sexual relations with the only apparent abnormality being the age difference between partners. Although there was considerable internal consistency in participants' reports, it is possible that the details of events were changed or refrained.

Given that childhood sexual experiences were not the primary focus of the original study, the coders were required to make judgments based on limited material and, at times, could not code certain variables due to lack of information. Moreover, consistent with Rind's (2001) study of young gay men's age-discrepant sexual relations, there was limited variability in the sexual acts described: Almost all experiences involved genital contact. It is unclear if this lack of variability was a methodological problem or an accurate reflection of men's experiences. Perhaps the gay and bisexual men in this sample did not perceive milder experiences such as exhibitionism as sexual and therefore tended not to report such experiences.

The response of guilt to sexual encounters as a child presented a challenge in coding the interviews. Several men spoke of feeling guilty about having a sexual experience with another male. Because guilt is a negative emotion, such feelings were coded as indicating a negative perception. However, in some cases the men's guilt appeared to result from family or community disapproval of homosexual relations rather than their own discomfort with the sexual acts with an older person. For example, one man stopped his sexual encounter because he thought that others believed homosexuality was wrong. He spoke of feeling guilty as a result of "someone else's self-imposed guilt" and of feeling confused about his sexual orientation. Another man stated that his "guilt came from enjoying it." Both of these men, as boys, wished that they could have talked to someone about their experience but did not feel comfortable revealing their same-sex attraction. Thus, the most damaging aspect of their sexual encounter may have been that they did not feel safe talking about their sexual orientation.

In conclusion, the standard convention of defining age-based childhood sexual abuse as uniformly negative, harmful, and coercive may not accurately represent gay and bisexual men's sexual experiences. Combining perception-based CSA experience with noncoercive, nonnegative, nonabusive experiences, as the age-based definition does, presents a misleading picture of childhood sexual abuse. An age-based CSA definition inflates prevalence rates of childhood sexual abuse and inaccurately suggests that the maladjustment associated with perception-based CSA experiences applies to all childhood age-discrepant sexual encounters. In contrast, these results suggest that gay men with histories of nonnegative, noncoercive childhood sexual experiences with older people are as well adjusted as those without histories of age-discrepant childhood sexual experiences. However, both definitions of CSA account for only a very small proportion of the variance in adult adjustment problems. Contrary to popular belief, negative outcomes do not inevitably follow from gay and bisexual men's childhood age-discrepant sexual encounters.

Table 1. Number of Reported Relationship Between Youth
and Older Sexual Partner

                     Perception-
Relationship          based CSA    CSE   Total

Intrafamilial
  Sibling                 2         2     4
  Parent                  1         0     2 (a)
  Extended family         4         1     5
Extrafamilial
  Acquaintance            2         3     5
  Family friend           3         4     7
  Stranger                0         7     7
  Community member       10         5    15
Total                    22        22    45

Note. CSA = childhood sexual abuse. CSE = childhood sexual experience.

(a) One participant not included in either the perception-based CSA or
CSE categories.

Table 2. Participants' Perceptions of CSA Experiences

                             Yes           No

Evidence of coercion       6 (12.8%)   41 (87.2%)
Perception of abuse
  At the time              9 (18.0%)   41 (82.0%)
  Now                     19 (38.0%)   31 (62.0%)
Perception of valence
  Negative at the time    20 (41.7%)   28 (58.3%)

Table 3. Means (SD) of Current Adjustment Measures Among CSA Categories

                                        Category means

                                    Non-CSA          CSE
Measures of adjustment             (n = 142)       (n = 25)

Self-esteem (RSE)                 3.28 (0.50)     3.12 (0.72)
Hopelessness (BHS)                2.04 (0.73)     2.22 (0.81)
Autocraticness                    0.82 (0.55)     0.77 (0.55)
Competitiveness                   0.82 (0.59)     0.61 (0.52)
Coldness                          1.04 (0.72)     0.77 (0.72)
Introversion                      1.33 (0.83)     1.12 (0.80)
Passiveness                       1.40 (0.75)     1.42 (0.95)
Exploitability                    1.28 (0.68)     1.32 (0.70)
Nurturance                        1.22 (0.68)     1.38 (0.69)
Expressiveness                    0.91 (0.60)     0.99 (0.59)

Mean of interpersonal problems    1.10 (0.49)     1.05 (0.49)

                                  Category means

                                       CSA-P
Measures of adjustment               (n = 24)

Self-esteem (RSE)                   3.06 (0.61)
Hopelessness (BHS)                  2.26 (0.91)
Autocraticness                      0.93 (0.53)
Competitiveness                     1.02 (0.73)
Coldness                            1.25 (0.90)
Introversion                        1.48 (0.85)
Passiveness                         1.63 (1.03)
Exploitability                      1.51 (0.86)
Nurturance                          1.66 (0.88)
Expressiveness                      1.38 (0.78)

Mean of interpersonal problems      1.36 (0.64)

Note. CSA = childhood sexual abuse. CSE = childhood sexual experience.
CSA-P = perception-based childhood sexual abuse.

Table 4. Hierarchical Multiple Regression Analyses for the
Prediction of Adjustment by Age-Based CSA and
Perception-Based CSA

Adjustment measure                  Step      B
Self-esteem
  Age-based CSA                       1     -0.19 *
  Perception-based CSA                2     -0.07
Competitiveness
  Age-based CSA                       1     -0.01
  Perception-based CSA                2      0.41 *
Coldness
  Age-based CSA                       1     -0.03
  Perception-based CSA                2      0.49 *
Nurturance
  Age-based CSA                       1      0.29 *
  Perception-based CSA                2      0.29
Expressiveness
  Age-based CSA                       1      0.27 *
  Perception-based CSA                2      0.39 *
Mean of interpersonal problems
  Age-based CSA                       1      0.09
  Perception-based CSA                2      0.31 *

Adjustment measure                  [beta]     [R.sup.2] change
Self-esteem
  Age-based CSA                     -0.15 *          0.02 *
  Perception-based CSA              -0.04            0.00
Competitiveness
  Age-based CSA                     -0.01            0.00
  Perception-based CSA               0.22 *          0.03 *
Coldness
  Age-based CSA                     -0.02            0.00
  Perception-based CSA               0.21 *          0.03 *
Nurturance
  Age-based CSA                      0.18 *          0.03 *
  Perception-based CSA               0.13            0.01
Expressiveness
  Age-based CSA                      0.18 *          0.03 *
  Perception-based CSA               0.20 *          0.02 *
Mean of interpersonal problems
  Age-based CSA                      0.08            0.01
  Perception-based CSA               0.20 *          0.02 *

* p < .05.

REFERENCES

Abramson, P. R. (1990). Sexual science: Emerging discipline or oxymoron? The Journal of Sex Research, 27, 147-165.

Bauserman, R., & Rind, B. (1997). Psychological correlates of male child and adolescent sexual experiences with adults: A review of the nonclinical literature. Archives of Sexual Behavior, 26, 105-141.

Beck, A. T., Weissman, A., Lester, D., & Trexler, L. (1974). The measurement of pessimism: The Hopelessness Scale. Journal of Consulting & Clinical Psychology, 42. 861-865.

Beitchman, J. H., Zucker, K. J., Hood, J. E., daCosta, G. A., Akman, D., & Cassavia. E. (1992). A review of the long-term effects of child sexual abuse. Child Abuse and Neglect, 16, 101-118.

Cameron. P., & Cameron. K. (1995). Does incest cause homosexuality? Psychological Reports, 76, 611-621.

Constantine, L. L. (1981). The effects of early sexual experience: A review and synthesis of research. In L. L. Constantine & F. M. Martinson (Eds.), Children and sex. Boston: Little, Brown.

Coxell, A., King. M., Mezey, G., & Gordon. D. (1999). Lifetime prevalence, characteristics, and associated problems of non-consensual sex in men: Cross sectional survey. British Medical Journal, 318, 846-850.

Dhaliwal, G. K., Gauzas, L., Antonowicz, D. H., & Ross, R. R. (1996). Adult male survivors of childhood sexual abuse: Prevalence, sexual abuse characteristics, and long-term effects. Clinical Psychology Review, 16, 619-639.

Dolezal, C., & Carballo-Dieguez, A. (2002). Childhood sexual experiences and the perception of abuse among Latino men who have sex with men. The Journal of Sex Research, 39, 165-173.

Doll, L. S., Joy, D., Bartholow, B. N., Harrison, J. S., Bolan, G., Douglas, J. M., et al. (1992). Self-reported childhood and adolescent sexual abuse among adult homosexual and bisexual men. Child Abuse & Neglect, 16, 855-864.

Fisher, B., & Akman, J. S. (2002). Normal development in sexual minority youth. Review of Psychiatry, 21, 1-16.

Fondacaro, K. M., Holt, J. C., & Powell, T. A. (1999). Psychological impact of childhood sexual abuse on male inmates: The importance of perception. Child Abuse & Neglect, 23, 361-369.

Fromuth, M. E., & Burkhart, B. R. (1987). Childhood sexual victimization among college men: Definitional and methodological issues. Violence and Victims, 2(4), 241-253.

Henderson, A. J. Z. (1998). It takes two to tango: An attachment perspective exploring women's and men's relationship aggression. Unpublished doctoral thesis, Simon Fraser University, Burnaby, British Columbia, Canada.

Herman, J. L. (1992). Trauma and recovery. New York: Basic Books.

Horowitz, L. M., Rosenberg, S. E., Baer, B. A., Ureno, G., & Villasenor, V. S. (1988). Inventory of interpersonal problems: Psychometric properties and clinical applications. Journal of Consulting and Clinical Psychology, 56, 885-892.

Kilpatrick, A. C. (1987). Childhood sexual experiences: Problems and issues in studying long-range effects. The Journal of Sex Research, 23, 173-196.

Lenderking, W. R., Wold, C., Mayer, K. H., Goldstein, R., Losina, E., & Seage, G. R., III. (1997). Childhood sexual abuse among homosexual men: Prevalence and association with unsafe sex. Journal of General Internal Medicine, 12, 250-253.

Nelson, A., & Oliver, P. (1998). Gender and the construction of consent in child-adult sexual contact: Beyond gender neutrality and male monopoly. Gender & Society, 12, 554-577.

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

Rind, B. (2001). Gay and bisexual adolescent boys' sexual experiences with men: An empirical examination of psychological correlates in a nonclinical sample. Archives of Sexual Behavior, 30, 345-368.

Rind B., Tromovitch P., & Bauserman, R. (1998). A meta-analytic examination of assumed properties of child sexual abuse using college samples. Psychological Bulletin, 24, 22-53.

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

Russell, D. (1986). The secret trauma: Incest in the lives of girls and women. New York: Basic Books.

Sandfort, T. G. M. (1983). Pedophile relationships in the Netherlands: Alternative lifestyle for children. Alternative Lifestyle, 5, 164-183.

Savin-Williams, R. C. (1998). "... And then I became gay": Young men's stories. New York: Routledge.

Steever, E. E., Follette, V. M., & Naugle, A. E. (2001). The correlates of male adults' perceptions of their early sexual experiences. Journal of Traumatic Stress, 14, 189-204.

West, D., & Woodhouse, T. (1993). Sexual encounters between boys and adults. In C. Li, D. West, & T. Woodhouse (Eds.), Children's sexual encounters with adults (pp. 3-137). New York: Prometheus.

Manuscript accepted May 25, 2004

Jessica L. Stanley, Kim Bartholomew, and Doug Oram

Simon Fraser University

Address correspondence to Jessica Stanley or Kim Bartholomew, Department of Psychology, Simon Fraser University, 8888 University Drive, Burnaby, BC, Canada, V5A 1S6; e-mail: [e-mail] or [e-mail].