|Home | About | Journals | Submit | Contact Us | Français|
To estimate how many heterosexual and gay/bisexual men self-define abusive childhood sexual experiences (CSEs) to be childhood sexual abuse (CSA) and to assess whether CSA self-definition is associated with risky behavioral and psychiatric outcomes in adulthood.
In Philadelphia County, 197 (66%) of 298 recruited men participated in a telephone survey. They were screened for CSEs and then asked if they self-defined abusive CSEs to be CSA; they also were asked about risk behavior histories and posttraumatic stress disorder (PTSD) and depression symptoms.
Of 43 (22%) participants with abusive CSEs, 35% did not and 65% did self-define abusive CSEs to be CSA (“Non-Definers” and “Definers,” respectively). Heterosexual and gay/bisexual subgroups’ CSA self-definition rates did not significantly differ. When self-definition subgroups were compared to those without CSEs (“No-CSEs”), Non-Definers had lower perceived parental care (p = .007) and fewer siblings (p = .03), Definers had more Hispanics and fewer African Americans (p = .04), and No-CSEs had fewer gay/bisexual men (p = .002) and fewer reports of physical abuse histories (p = .02) than comparison groups. Non-Definers reported more sex under the influence (p = .001) and a higher mean number of all lifetime sex partners (p = .004) as well as (only) female sex partners (p = .05). More Non-Definers than Definers reported having experienced penetrative sex as part of their CSA (83% versus 35%, p = .006). Different explanations about self-definition were provided by subgroups.
Many men with abusive CSEs do not self-define these CSEs to be CSA, though not in a way that differs by sexual identity. The process by which men self-define their abusive CSEs to be CSA or not appears to be associated not only with self-explanations that differ by self-definition subgroup, but also with behavioral outcomes that impart risk to Non-Definers.
Many men and women with childhood-documented histories of sexual abuse do not subsequently report a childhood sexual abuse (CSA) history when asked in adulthood (Widom & Morris, 1997). Numerous factors likely figure into this non-reporting, including forgetfulness, repressed memories, unwillingness to disclose one’s sensitive history to unknown researchers, and not defining one’s abusive childhood sexual experiences (CSEs) to be CSA even though others would. This latter possibility has been a subject of increased interest, particularly for men in whom CSA non-reporting appears more frequent than in women (Widom & Morris, 1997).
Some posit that the cognitive appraisal of CSA events as abusive or not plays a different or more complicated role for men than for women (Fondacaro, Holt, & Powell, 1999). Emerging data not only suggest 1) that not all men with abusive CSEs cognitively appraise their abusive CSEs to have been CSA, but also 2) that the cognitive appraisal of abusive CSEs as CSA or not appears to drive whether CSA is associated with poor outcomes (Carballo-Dieguez & Dolezal, 1995; Fondacaro, Holt, & Powell, 1999; Stander, Olson, & Merrill, 2002; Stanley, Bartholomew, & Oram, 2004; Steever, Follette, & Naugle, 2001). Some argue that this latter dynamic is particularly at play for gay/bisexual males, hypothesizing that the social constraints limiting young gay/bisexual males’ ability to have sexual encounters with peers drive them to have sexual encounters with older gay/bisexual man that are, by definition, CSA, but that are on the whole more of a benefit than a liability to their development (e.g., the benefit of affirming their sexual identity in a self-empowering way that encourages self-acceptance overwhelms any negative influences that experiencing an abusive CSE may cause) (Stanley, Bartholomew, & Oram, 2004; Steever, Follette, & Naugle, 2001).
Of the CSA self-definition publications noted above, only Dolezal & Carballo-Dieguez (2002) and Stanley, Bartholomew, & Oram (2004) studied gay/bisexual men (n = 100 [all Latino] and n = 50, respectively). The sexual identity of the other three samples were not characterized, though the samples were from populations in which heterosexual identity likely predominated. Fondacaro, Holt, and Powell (Fondacaro, Holt, & Powell, 1999) studied predominantly Caucasian inmates; Steever, Follette, and Naugle (2001) studied a convenience sample of undergraduate men; and Stander, Olson and Merrill (2002) studied predominantly Caucasian Navy recruits (n = 86, 40, and 615, respectively).
Of men from the gay/bisexual samples, 38% to 59% self-defined abusive CSEs to be CSA, though only 18% reported that they had initially perceived the abusive CSEs to be CSA at the time of the experience(s). Similarly, 15% to 59% of men from the likely-heterosexual samples self-defined abusive CSEs to be CSA. Stander et al. simultaneously reported on a female subsample with similar CSE histories; they self-defined abusive CSEs to be CSA three times more often than the male subsample.
Variables associated with gay/bisexual men self-defining their abusive CSEs to be CSA were younger age at the time of the abusive CSE(s) and a larger victim-perpetrator age difference (Dolezal & Carballo-Dieguez, 2002; Stanley, Bartholomew, & Oram, 2004). There was no reported difference in perpetrator age by self-definition subgroups, suggesting that larger victim-perpetrator age differences in definers (versus non-definers) were driven by boys’ younger ages at the time of abusive CSEs (Stanley, Bartholomew, & Oram, 2004). Definers also reported having been physically forced, physically hurt, threatened, and/or emotionally hurt during abusive CSEs more often than non-definers (Dolezal & Carballo-Dieguez, 2002).
These same variables (operating with the same directionality) were found to be associated with CSA self-definition in likely-heterosexual samples (Stander, Olson, & Merrill, 2002; Steever, Follette, & Naugle, 2001). Stander et al. and Fondacaro et al. also reported that definers’ versus non-definers’ abusive CSEs were more likely to have occurred with a male perpetrator and/or with an immediate/extended family member, and to have involved penetration (Fondacaro, Holt, & Powell, 1999; Stander, Olson, & Merrill, 2002).
Studies of gay/bisexual samples revealed definers to have reported more alcohol use, unprotected anal sex, and male sex partners than non-definers (Dolezal & Carballo-Dieguez, 2002). Definers also revealed more interpersonal problems (e.g., problems with expressiveness) than non-definers, and these problems in definers appeared to drive the differences found in comparisons of those with CSA versus no-CSA histories (e.g., non-definers’ interpersonal problems were not significantly different from those with no-CSA) (Stanley, Bartholomew, & Oram, 2004).
Similarly, studies of likely-heterosexual samples revealed definers to have higher lifetime and current PTSD symptom scores and distress and to report that the overall impact of abusive CSEs on their lives was more negative than (it was for) non-definers; definers also were significantly more likely to have participated in mental health treatment than non-definers (Fondacaro, Holt, & Powell, 1999). Unlike non-definers from gay/bisexual samples, however, likely-heterosexual non-definers had higher levels of lifetime alcohol abuse/dependence than definers.
Similarities across (sexual identity) sample types, then, appear stronger than differences, although some differences are suggested. Likely-heterosexual samples studied to date, though, have not been highly generalizable (e.g., inmates, undergraduates, Navy recruits). The current, hypothesis-generating study sought to assess CSA self-definition, whether it differed by sexual identity, and whether it was associated with behavioral and psychiatric outcomes, in a probability sample of men recruited from urban areas that are predominantly heterosexual and where the outcomes of risk behavior (e.g., HIV infection) are prevalent.
After the study was approved by the University of Pennsylvania Institutional Review Board’s Committee on Studies Involving Human Beings, potential participants were recruited by random digit dialing (RDD). RDD employed a “Waksberg modification,” whereby sampling from an enumeration of the Working Residential Hundred Blocks (WRHB) of active exchanges occurred within a sampling area (Waksberg, 1978). Sampling areas were Philadelphia County zip code areas with high AIDS incidence (City of Philadelphia Department of Public Health, 2000). Non-working and non-residential numbers were replaced by other RDD numbers selected in the same WRHB stratum, as were ineligible households (e.g., no adult in the household, language barriers). Once a working number was obtained, 10 callbacks were made to speak to a household member 18 years, at which time a short screening interview was completed.
One hundred men were recruited using the noted probability sampling into each of three age groups: 18–29; 30–39; and 40–49 years. Contact information for men who agreed to participate in a study of how “childhood experiences have affected adult men’s health and well being” was obtained, as was age, race/ethnicity, and educational attainment information. A research firm with expertise in telephone surveys on sensitive subjects completed the screening and all interviewing.
A study packet was mailed to potential participants. It described the study and principal investigator, offered a $15.00 incentive for study participation, and included the Institutional Review Board-approved consent form. A stamped, addressed envelope was provided in the packet so that after complete description of the study to participants, written informed consent was obtained and returned to the principal investigator. Men were called back 2 weeks later and a full telephone interview was administered to those agreeing to participate. For those men who had not returned a signed informed consent, verbal consent was obtained---after providing a complete (verbal in addition to the prior written) description of the study---before proceeding with the interview.
Standard sociodemographic questions (e.g., age, race/ethnicity) were asked first in the interview. However, more sensitive sociodemographic questions were asked later in the interview after rapport with the participant had been established. Sexual identity, for example, was asked about more than half-way through the interview and done so by asking, “If you had to choose one of the following terms to describe yourself, which one would you select? Homosexual/gay; Bisexual; Heterosexual/straight; Don’t know/Confused; or Don’t want to answer.”
Information about parenting history was obtained with the Parental Bonding Instrument (PBI) (Parker, 1990; Parker, Tupling, & Brown, 1979). The PBI allows scoring of two parenting dimensions. The care score (8 items) indicates parental warmth and understanding at one extreme compared with parental rejection and withdrawal at the other. The protection score (12 items) indicates parental over-control at one extreme and complete autonomy at the other. The Cronbach’s alphas for these dimensions in this sample were .83 and .75, respectively. Childhood socioeconomic status (CSES) was assessed by asking participants what their father’s and mother’s occupations where when they were growing up. Responses were used to estimate likely earnings per year per parent as forward-discounted to sex- and state-specific 2000 Census “Earnings by Detailed Occupation: 1999” tables (United States Census Bureau, 2002). Earnings for variations on “military service” were set at an E-7 enlistee level who had 14 years of service (discounted to FY1999) (Powers, 2005); earnings for variations on “housewife” were set to zero; no stated occupation and occupations described as a variation on “civil servant” were set to missing. Total earnings by total number in household (obtained from response to question “How many brothers/sisters did you have while you were growing up?”) was used to characterize CSES as being # versus > two times the poverty line, or “low CSES” versus “high CSES,” respectively (Department of Health and Human Services Office of the Secretary, 1999).
Finkelhor’s four “funneling” questions then were used to screen for CSA (Finkelhor, Hotaling, Lewis, & Smith, 1990). Each question began with “When you were a child or teenager (before 18 years of age), can you remember having any experience you would now consider sexual abuse . . .” and ended with: 1) “…like someone trying or succeeding in having any kind of sexual intercourse with you, or anything like that?”; 2) “…involving someone touching you, or grabbing you, or kissing you, or rubbing up against your body either in a public place or private---or anything like that?”; 3) “…involving someone taking nude photographs of you, or someone exhibiting parts of their body to you, or someone performing some sex act in your presence---or anything like that?”; 4) “…involving oral sex or sodomy---or anything like that?” An affirmative response to any of these questions led to additional questions that detailed who the potential perpetrator was and how old s/he was at the time of the experience(s) that had led to a “yes” response; what occurred at the initial experience, as well as any subsequent experience(s) that might have occurred with the potential perpetrator; how old the respondent was at the initial experience (and at the final experience if more than one event had occurred); how many experiences with this potential perpetrator had occurred altogether (if there was more than one event); whether there were other people than the primary (potential) perpetrator involved in the experience(s); and whether the potential perpetrator(s) ever used force (such as restraint or a weapon), or any other form of coercion or threat. The respondent was asked if they self-defined the experience(s) to be sexual abuse with the question, “Would you define the experience(s) you experienced as ‘sexual abuse’?”
Those who self-defined their CSEs to be CSA were asked, “Why do you think, in your own words, you define the experience(s) that occurred as being sexual abuse?” Those who did not self-define their CSEs to be CSA were asked, “Why do you think you don’t define the experience(s) that occurred as being sexual abuse?” Possible options were provided prior to offering these respondents the option to explain in their own words. The options were: “It’s too embarrassing to think of myself as having been abused;” “I don’t think the person(s) who initiated the experience(s) meant harm;” “I would have to confront the person(s) who did this to me then;” “I think ‘sexual abuse’ is an excuse people use to excuse their own failures;” “I just haven’t thought about it like that;” “It felt good, and if it felt good then it can’t be abuse;” “My body responded, and if your body responds, it can’t be abuse;” and “I just can’t admit it to myself.”
Later in the interview, the respondent was asked, “How old were you the first time you had any type of sexual experience with another person willingly?” If---earlier in the interview---they had answered the sexual abuse questions affirmatively, this question was introduced with the phrase, “Not counting the experience(s) you mentioned in the last section.” After this question they were asked with whom they had had their first willing sexual experience, who/how old this person was, and what occurred.
The investigator-applied definition for CSA was any sexual experience before 18 years of age (even those characterized as “willing”) in which: 1) a power differential existed between a victim and perpetrator wherein the perpetrator was 5 years older than a victim <13 years, was 10 years older than a victim 13–17 years, or was an authority figure (e.g., teacher); 2) coercion was reported to have occurred; or 3) penetration (e.g., oral, anal, vaginal) of victim or perpetrator (by victim) occurred when victim was prepubertal (#11 years) and perpetrator was postpubertal (>12 years) as well as at least 2 years older than the victim (Needleman, 1996; Tanner & Davies, 1985). Three participants were included even though they did not know the age of their perpetrator; all reported experiences (at ages 6, 15, and 15 years) with a stranger who they described as an “adult.”
Frequency of sex that participants had had as an adult under the influence of a substance was assessed by asking, “How often do you have sex while under the influence of alcohol or another drug (such as marijuana or cocaine)?” Response options were “None of the time,” “A little of the time,” “Some of the time,” “Most of the time,” and “All of the time.” Sexually transmitted disease history was assessed by asking, “Have you ever been diagnosed as having any of the following sexually transmitted infections---gonorrhea, syphilis, chlamydia, genital herpes, genital warts, trichomoniasis, or any other sexually transmitted infection?” Response options were “yes” and “no.” Number of lifetime sexual partners was assessed by asking, “How many sex partners would you say you’ve had in your life since you turned 18 years old (when you were no longer a teenager)?” Respondents were asked to enumerate from this total number the number of females versus males.
PTSD symptoms were assessed with the Posttraumatic Stress Diagnostic Scale (PDS); a score threshold of >10 (symptom severity more than “mild”) was used to define likely PTSD (Foa, Riggs, Dancu, & Rothbaum, 1993). The PDS was modified to begin, “I will read a list of problems that people sometimes have after experiencing a traumatic event. Listen to each stated problem carefully and choose the description that best describes how often that problem has bothered you in the past month. Rate each problem with respect to any traumatic event you have experienced that continues to bother you in some way.” Depressive symptoms were assessed with the Center for Epidemiologic Studies---Depression scale (CES-D); a score threshold of >16 was used to define likely depression (Radloff, 1977; Roberts, 1980; Stallones, Marx, & Garrity, 1990). The Cronbach’s alphas for these instruments in this sample were .91 and .88, respectively.
For continuous variables, two-group comparisons were performed using 2-tailed t-tests and three-group comparisons were performed using analysis of variance (ANOVA). Pairwise comparisons using Sheffé means comparison tests were completed if the overall ANOVA test was statistically significant. For categorical variables, Chi-square methods were performed for comparisons. An alpha-level of .05 was employed. Data were managed and analyzed using SPSS 12.0 for Windows ©SPSS Inc., 1989–2003 (Chicago, IL).
A total of 298 men were recruited; 197 (66%) participated. Participants did not differ from non-participants on the three variables assessed at screening: age (p = .78); race (p = .88); and educational attainment (p = .06).
Table 1 shows the characteristics of the men and their families of origin. The sample, overall, was predominantly minority, 7% were gay/bisexual, and education and income distributions were relatively equal across assessed categories. Most (73%) perceived their parents to have been caring, and most (88%) had siblings. Half of the participants perceived their parents to have been (over)protective. Half reported a childhood socioeconomic status that was # two times the poverty line. Half had been physically abused.
Forty-four of 197 participants (22%) responded affirmatively to at least one of the four CSA screening questions: 29 (15%), 35 (18%), 16 (8%), and 13 (7%) men answered “yes” to questions 1–4, respectively (see Methods for question order). Thirty-nine (89%) of these met our CSA criteria. Four who answered “no” to all screening questions reported an age differential with the partner of their first “willing” sexual experience that met the criterion for CSA; all abusive partners were adult women. According to the investigator-applied definition of CSA, then, CSA prevalence in this sample was 43/197 (22%).
Of the 43 men with CSA determined by application of the investigator’s definition (“abusive CSEs”), 15 (35%) did not and 28 (65%) did self-define their histories to be CSA (“Non-Definers” and “Definers,” respectively). Gay/bisexual men were not significantly more or less likely than heterosexual men to self-define abusive CSEs as CSA: four of eight (50%) gay/bisexual men with abusive CSEs were Definers; and 24/35 (69%) heterosexual men with abusive CSEs were Definers (p = .32).
Table 1 reports comparisons in subgroup composition for Non-Definers, Definers, and those who did not report abusive CSEs (“No-CSEs”). There were significant differences across subgroups by race/ethnicity (p = .04), sexual identity (p = .002), parental caring (p = .007), having had siblings (p = .03), and having been physically abused (p = .02). Significance of parental care and sibling findings appeared to be driven by differences between No-CSEs and Non-Definers. Significance of the race/ethnicity finding appeared to be driven by differences between No-CSEs and Definers. Significance of sexual identity and physical abuse findings appeared to be driven by differences between the Non-Definer/Definer subgroups and No-CSEs.
Table 2 reports comparisons of Non-Definers’ and Definers’ abuse characteristics. The only significant difference identified was whether abusive CSE(s) had involved oral, vaginal, and/or anal penetration. A majority (83%) of Non-Definers had experienced penetration, whereas only a minority (35%) of Definers had (p = .006).
When abuse characteristics were stratified by sexual identity, gay/bisexual Definers were found to have been fondled, whereas Non-Definers were not (100% versus 0%, respectively; p = .03); and gay/bisexual Non-Definers were found to have been fellated, whereas Definers were not (100% versus 0%, respectively; p = .03). Such significant differences did not exist for heterosexual Definers versus Non-Definers (fondling, p = .91; oral sex, p = .94; vaginal/anal, p = .13; and any penetration, p = .08). Incidentally, no gay/bisexual or heterosexual Non-Definers reported force/coercion; only Definers in each sexual identity subgroup had reported force/coercion (though numbers were quite small).
Table 3 reports comparisons among all subgroups’ rates of risky behavior and psychiatric symptoms. There were significant differences for the proportion having sex under the influence (p = .001) and the mean total number of all lifetime sex partners (p = .004) as well as (only the) female sex partners (p = .05). Significance of the sex-under-influence finding appeared to be driven by differences across all three subgroups. Post-hoc comparisons indicated that mean number of all sex partners and mean number of female sex partners were greater for Non-Definers than for those with no CSEs (p = .005 and p = .05, respectively).
When comparisons of outcome differences were stratified by sexual identity, the only statistically significant findings found were for heterosexual men. Heterosexual Non-Definers reported a STD history at higher rates than Definers and No-CSEs (55% versus 18% versus 20%, respectively; p = .02). Heterosexual Non-Definers also reported a higher mean total number of overall sex partners, female sex partners, and male sex partners than Definers and No-CSEs (p = .02, p = .02, and p < .001, respectively). Heterosexual Definers reported more PTSD symptoms than Non-Definers and No-CSEs (11.0 versus 5.7 versus 6.0, respectively; p = .04).
Given that some studies have conjectured that gay/bisexual men may experience abusive CSEs to have been beneficial to their sexual development (and, thus, not self-define them to be CSA), the Non-Definer subgroup of this study was examined to identify men meeting this possible scenario. Only two of these 15 men were gay/bisexual men who identified an unrelated adult male as the perpetrator of a post-pubertal (>11 years old) abusive CSE. When outcomes (listed in Table 3) were assessed for this small subgroup and compared with the rest of the Non-Definers, two significant differences were identified (in a direction possibly suggesting a not-positive or at least not-neutral impact of the abusive CSEs for these two men): more sex under the influence (p = .01) and greater PTSD severity (p = .01).
Table 4 provides a possible taxonomy of categories for explanations given by Non-Definers and Definers for why they did not or did self-define their abusive CSEs to be CSA, respectively. Non-Definers’ explanations could be organized into fewer categories than could those of Definers. Three Non-Definer explanations were perpetrator-related, one was related to self, and one was about the experience, whereas three Definer explanations were perpetrator-related, four were related to self, one was about the age difference between the perpetrator and self, and one was about the experience. Only one category---about consent---was shared by both subgroups, though the explanations that were a part of this category went in opposite directions according to which subgroup men were members. Non-Definers reported that they did not define their abusive CSEs to be CSA because they had consented to the experience(s), while Definers reported that they defined their abusive CSEs to be CSA because they had not consented to the experience(s).
Almost a quarter of this community-based, probability sample of men reported CSEs meeting investigator-imposed CSA criteria that are routinely used in the sexual abuse literature. Nearly two-thirds of these men with investigator-defined CSA histories self-defined their abusive CSEs to be CSA themselves. This rate of CSA self-definition did not significantly differ by sexual identity. Behavioral but not psychiatric outcomes differed across the two self-definition and No-CSE subgroups, when comparisons were completed for the whole sample. Non-Definers had the highest proportion of men who reported having had sex under the influence and they also had the highest number of lifetime sex partners; the No-CSE subgroup had the lowest proportion of men who reported having had sex under the influence, and they had the lowest number of lifetime sex partners. When outcome comparisons were stratified by sexual identity, however, results indicated that the heterosexual subsample had additional outcome differences across subgroups that were both behavioral and psychiatric: Non-Definers reported STD histories at higher rates than the Definer and No-CSE subgroups, and Definers reported higher PTSD symptom levels than the Non-Definer and No-CSE subgroups.
Though the rate of CSA self-definition appeared to be high, the questions used to capture participants’ abuse histories included the words “sexual abuse” and, thus, a bias toward identifying a higher rate of self-definition existed for this study. Within that context, then, it was somewhat surprising to find so many men who responded affirmatively to case-finding questions containing the words “sexual abuse” and then stated, in response to questions asked only five to 10 minutes later, that they did not consider themselves to have been sexually abused. This may suggest that, despite what initially appeared to be a high rate of self-definition in this study, barriers to self-defining abusive CSEs as CSA still remain for men. Furthermore, given that self-definition rates were not significantly different for heterosexual and gay/bisexual men, these barriers would seem to be at play regardless of a man’s sexual identity. Some of these barriers, as well as the means by which some men avoid them, may be represented by the taxonomy of explanations about abusive CSEs that men provided in this study, explanations that differ substantially by whether men were Non-Definers or Definers.
This study is the first to explore CSA self-definition in a probability sample of community-based, predominantly-heterosexual adult men who are more representative of the general population than prison inmates, college undergraduates, and Navy recruits, samples that have been used in prior studies of heterosexual men’s CSA self-definition. This study also is the first to compare CSA self-definition in heterosexual men to that in gay/bisexual men recruited from the same underlying population. That rates of CSA self-definition are similar across sexual identity subgroups for this study’s sample (mirroring similar findings from the less generalizable samples described earlier in the manuscript) appears to counter claims by those who articulate an exceptionalism argument for gay/bisexual males wherein their abusive CSEs are seen as potentially being more of a benefit to them than a liability (Stanley, Bartholomew, & Oram, 2004; Steever, Follette, & Naugle, 2001). When coupled with the finding that all Non-Definers were significantly more likely to engage in sexual risk behavior than the Definer and No-CSE subgroups, findings suggest that regardless of one’s sexual identity there are substantial liabilities not only in experiencing CSA, but also in not defining it as such.
Prior studies have consistently shown that younger age at abuse, larger age difference between victim and perpetrator, and the abuse characteristics of force and coercion are associated with CSA self-definition (Dolezal & Carballo-Dieguez, 2002; Stander, Olson, & Merrill, 2002; Stanley, Bartholomew, & Oram, 2004; Steever, Follette, & Naugle, 2001). Heterosexual samples have also indicated that they are more likely to self-define abusive CSEs to be CSA when the perpetrator is male or a family member, and when intercourse is involved (Fondacaro, Holt, & Powell, 1999; Stander, Olson, & Merrill, 2002). Though none of these variables---except penetration---was associated with CSA self-definition in the quantitative analyses of this study, even when analyses were stratified by sexual identity, Definers did note young age of victim, victim-perpetrator age difference, and perpetrator being a family member as reasons for their self-definition in qualitative explanations. These mixed methods findings suggested similarity between this study’s findings and others’---a good external validity check.
PTSD symptom severity was the only psychiatric outcome that differed across self-definition subgroups (and then only in a statistically significant way for the heterosexual subsample). The finding---that PTSD symptom severity was highest in Definers---also replicated findings reported previously, providing some additional external validation for study results (Fondacaro, Holt, & Powell, 1999).
Other findings from this study, however, differed from those reported previously. For example, Non-Definers were found to have the highest rates of sex-under-the-influence and the highest number of lifetime sexual partners. Though this may be similar to the Fondacaro et al. report (about a likely-heterosexual sample) in which Non-Definers had more alcohol problems than Definers, it appears to run counter to the report by Dolezal & Carballo-Dieguez (from a gay/bisexual sample) that noted the reverse: Definers had more alcohol problems and lifetime sexual partners (and risky sex) than Non-Definers (Dolezal & Carballo-Dieguez, 2002; Fondacaro, Holt, & Powell, 1999). This may indicate a need to be cautious in drawing firm conclusions from this study, or it may suggest that having used a more generalizable sample for this hypothesis-generating study was beneficial in that it indicated need for future study of self-definition in larger and even more generalizable samples.
Non-Definers appeared to be more likely to have low-care parents, and not to have had siblings. Self-defining abusive CSEs to be CSA may be difficult in the context of uncaring parents who may not have provided clear boundaries for interpersonal contact, and/or in the context of no feedback from siblings who might otherwise confirm perceptions about abusive CSEs about which one otherwise might have been uncertain. These possibilities, however, are entirely conjectural and require future confirmation.
This study also described open-ended explanations for CSA self-definition. The emerging taxonomy indicates that one core determinant of self-definition common to both groups is the issue of consent and/or willingness. Unlike Definers, Non-Definers’ consent-related comments indicated an inaccurate belief that children and adolescents without the cognitive ability to provide consent to sexual relations with an adult are able to do so (at least insofar as their own experience was concerned), a dynamic that has also been reported by Stanley, Bartholomew, and Oram.(2004) When coupled with the finding that Non-Definers’ had higher rates of penetrative CSEs, one might speculate that CSA self-definitions of abuse could be complicated by experiences of physical pleasure during abusive CSEs. In fact, Okami reported that a majority (91%) of males who described positive reactions to their CSA histories also recalled the experiences to have been physically pleasurable (Okami, 1991). It is possible that Non-Definers in this study were more likely to have had an erection, if not experienced ejaculation, during abusive CSEs. Such experiences may lead to self-perceptions of complicity, which may lead to self-perceptions of willingness and/or consent.
Men with CSA histories who do not define their abusive CSEs to be CSA also may be engaged in avoidant coping: it may be easier to claim sexual agency than grapple with having been a victim; and/or it may diminish the perceived need to label as a perpetrator someone about whom one has caring feelings and/or who may be a family provider and/or personal mentor. Fondacaro et al. hypothesize that men who employ such coping strategies may be more likely to abuse alcohol or to ignore emotions, both of which result in greater distress (Fondacaro, Holt, & Powell, 1999). Whether some men do not self-define abusive CSEs to have been CSA because of a perception of complicity or because of avoidant coping or because of other reasons altogether, concluding from this study’s findings and these hypotheses that Non-Definers must be made to perceive their abusive CSEs as CSA is not an obvious deduction. Rather, potential interventions aimed at avoiding/reversing apparent negative outcomes for Non-Definers perhaps can simply address sexual histories and sexual scripts, as well as self-perceptions of, satisfaction with, and potential outcomes related to these histories/scripts, without labeling histories/scripts CSA and CSA-related.
Three study limitations, in addition to those that have already been noted above, warrant specific notice. First, even though study participants did not differ from non-participants on the three variables assessed during the initial telephone screening, no further information about the reasons for non-participation are known; as a result, unmeasured bias may exist. Second, the study sample was non-affluent, urban, and largely minority. Thus, results may not be able to be generalized beyond these groups. In addition, exclusion of potential participants on the basis of language barriers further limits findings to those who speak English, a not-insubstantial concern given the growing population of non-English speakers in the U.S. Third, the study was hampered by small numbers, resulting in many near-significant findings about which no conjectural claims are possible. This highlights the need for future studies to recruit larger samples to achieve subgroups with CSA histories that are more numerous; this, in turn, will lead to Non-Definer and Definer subgroups that are large enough to provide adequate power for comparisons across numerous potentially-important variables. These future studies should be powered to adjust for these other potentially explanatory variables and potential confounders/effect modifiers so as to assess better the possibility that self-definition subgroup differences are driven by other factors than self-definition (e.g., race/ethnicity, sexual identity, parental care, sibling number, abuse characteristics).
In these future studies, a more accurate assessment of self-definition also should occur. This would be done best by collecting information on all participants’ CSEs---without labeling them to be anything other than sexual experiences---some time after which participants would be asked if they believe themselves to have a CSA history. Responses to this CSA question could be compared to the previously enumerated CSEs that have been subjected to an externally-applied, widely-accepted CSA definition.
The potential weakness of the “sex under the influence” measure is another limitation. Although it and variations have been used in numerous studies (Celentano et al., 2006), it is possible using this approach for a person who has sex one time per week and who always does so while under the influence to be defined inaccurately as higher risk than another person who has sex 30 times per month and does so while high “a little of the time” (say five of those 30 times). Future studies should assess this variable better.
The emerging taxonomy of why participants of this study did or did not self-define their abusive CSEs to be CSA also indicates that future research into CSA self-definition should be approached with more in-depth mixed methods.(Tashakkori & Teddlie, 1998) Knowing how men’s un(der)explored internal experiences contribute to the apparent difficulties in self-defining abusive CSEs to be CSA appear to be critical. Until such data become available, not only researchers, but also mental health professionals should attend more closely to the personal restrictions boys and men, and their surrounding culture, may impose on CSA self-definitions. These cultural impositions include not only those that might suggest boys cannot be victims, but also moves by some to relax CSA definitions for gay/bisexual youth to the point of considering child-with-adult sexual interactions to be socially normative. Attending to these cultural impositions will mean that researchers and clinicians directly ask men with abusive CSEs whether they self-define them to be CSA, which may provide an opportunity to identify in these men varied internal conflicts about labeling that have clinical relevance
This work was funded by a grant from the National Institute of Drug Abuse (DA015635) and a Research Career Development award from the Health Services Research & Development Service of the Department of Veterans Affairs.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.