Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arch Sex Behav. Author manuscript; available in PMC 2011 August 1.
Published in final edited form as:
PMCID: PMC2889135

Moderators of Sexual Behavior in Gay Men


We investigated factors that might moderate the association between sexual behavior desires and sexual behavior enactments in gay men. Condom eschewal, number of STIs, HIV serostatus, age, and relationship status were each hypothesized to moderate this association. An Internet survey collected data from 219 self-identifying gay men. Results indicated that sexual behavior desires and enactments were highly correlated, and of the five moderators tested, four varied this association. Condom eschewers had a stronger association between desires and enactments than condom users. Gay men with fewer STIs/STDs (excluding HIV) also had a stronger association between the two variables. HIV serostatus did not exclusively moderate the association. Rather, a three-way interaction was produced such that HIV-positive men with STIs had a stronger association between sexual behavior desires and enactments than HIV-negative men with STIs. Finally, gay men in monogamous relationships were least likely to have their desires associated with enactments. Age was not found to be a significant moderator. Overall, we concluded the moderators representing sexual health and sexual health behaviors were most influential over the enactment of sexual behavior desires.

Keywords: Gay men, Sexual enactment, Sexual desire, HIV, STIs/STDs, Condom use


Gay men are diverse with respect to the sexual behaviors they both desire and enact (Sanderson, 1994). Moreover, gay men differ from other groups in their sexual behavior. Research shows that, on average, gay men have more partners, engage in more risky sexual behavior, and are more likely to seek sexual sensation than other groups, such as heterosexual men and women and lesbians (Bailey, Gaulin, Agyei, & Gladue, 1994; Ekstrand, Stall, Paul, Osmond, & Coates, 1999; Thompson, Yager, & Martin, 1993). Considering that gay men differ from other groups with regard to many sexual activities, one might also expect gay men to be more likely to enact their sexual behavioral desires. Aside from studies on gay male paraphilic behaviors (e.g., Alison, Santtila, Sandnabba, & Nordling, 2001; Sandnabba, Santtila, & Nordling, 1999; Weinberg, Williams, & Calhan, 1994), few studies have examined the relationship between gay male sexual behavior desires and enactments and virtually none have examined moderators of the association between sexual behavior desire and sexual behavior enactment. Little is understood about gay men's sexual inhibition or disinhibition of behaviors when faced with problems such as sexually transmitted infections (STIs), the human immunodeficiency virus (HIV), and sustaining safer sex practices. Thus, we explored possible moderators of the relationship between sexual behavior desires and enactments among gay men. Specifically, we focused on the impact of five potential moderators identified in previous research with gay men: condom eschewal, STIs, HIV, age, and relationship status.

Before the introduction of HIV into the gay community in the 1980s, condom use was extremely low—sharply increasing only after the connection between HIV infection and anal intercourse was conclusively shown (Catania et al., 1991). Many gay men adopted the reinforced ethic of “a condom every time” during these years (Kippax & Race, 2003). However, it has been argued that the introduction of highly active antiretroviral therapies (HAART; DiClemente et al., 2002; Race, 2003), the popularization of “effective” techniques (e.g., coitus interruptus) to prevent HIV transmission (Van de Ven et al., 2002; Wegesin & Meyer-Bahlburg, 2000), and HIV prophylaxis fatigue and rebellion (Crossley, 2004; Gauthier & Forsyth, 1999; Halkitis, Wilton, & Galatowitsch, 2005; Wolitski, 2005) have all contributed to the decrease in condom use. Regardless of explanation, condom eschewal—the voluntary disuse of condoms during anal intercourse—has been associated with a higher valued or more pleasurable sexual experience, drug use before or during sex, and anal intercourse with an increased number of partners (Crosby, Stall, Paul, Barrett, & Midanik, 1996; Ekstrand et al., 1999; Kelly & Kalichman, 1998).

As suggested, gay men are encouraged to wear condoms because they enact behaviors highly associated with STI and HIV infection (Lama & Planelles, 2007; Royce, Sena, Cates, & Cohen, 1997). However, recent research has suggested antithetical responses to the gay male population's high prevalence of STIs and HIV—the bareback movement (Crossley, 2004; Gauthier & Forsyth, 1999; Halkitis et al., 2005; Mansergh et al., 2002; Moskowitz & Roloff, 2007a; Tewksbury, 2003; Wolitski, 2005). In this subculture of gay men, condoms are actively eschewed. For barebackers, this is performed in spite of possible STI/HIV infection and to feel more “sexually liberated” (Crossley, 2004; Halkitis et al., 2005). Additionally, many barebackers believe that becoming HIV-positive is an effective way to increase sexual liberation and reach a “sexual nirvana” because the fear of seroconverting from the enactment of unsafe sexual behaviors is removed from the sexual experience (Gauthier & Forsyth, 1999; Moskowitz & Roloff, 2007a, b). Whether HIV-positive men actually enjoy a sort of “sexual nirvana” remains hotly debated (see Halkitis et al., 2005; Moskowitz & Roloff, 2007b; Parsons & Bimbi, 2007; Wolitski, 2005). It becomes difficult to assess if being HIV-positive allows for the greater enactment of sexual behavior desires; or, alternatively, if being HIV-positive is merely incidental, and it is the factors that contribute to becoming HIV-positive (e.g., condom eschewal) that increase the likelihood for greater sexual behavior enactments. Thus, more research into the association among sexual behavior desires, enactments, and STIs/HIV is needed.

Age has significant social and sexual meaning for gay men (Berger, 1996). Gay men highly value physical attractiveness, and thus younger, fitter, and more attractive individuals become prime choices for sexual selection (Berger, 1996; Grube, 1990). In short, availability exists for young men to have more sex than older men. Research suggests that younger men indeed have more unprotected sex or engage in otherwise risky behaviors than older men and are at a higher risk for STI/HIV contraction (Mansergh & Marks, 1998; McAuliffe et al., 1999; Valleroy et al., 2000; Vincke, Bolton, & Miller, 1997). Research is vague concerning the other sexual behaviors younger versus older men enact. Yet, if a differentiation has been noted between the groups with respect to cautiousness regarding sexual intercourse, it is plausible that younger and older gay men differ with respect to sexual behavior desires and their potential enactments.

Individuals in short and long-term gay relationships sometimes open their relationships either temporarily or indefinitely (Blasband & Peplau, 1985). The literature suggests partners in monogamous relationships report stronger dependency on their partners, more favorable attitudes towards the relationship itself, and lower tension than those in non-monogamous relationships (Kurdek & Schmitt, 1985). However, more recent research suggests that though monogamy may provide those previously mentioned benefits, over time, partner distrust, jealousy, sexual ennui, and anxiety push virtually all homosexual male, romantic relationships towards non-monogamy (Worth, Reid, & McMillan, 2002). The sexual variety, prevention of possessiveness, and the promotion of freedom and egalitarianism all overwhelm the need for exclusivity in many gay male couples (Shernoff, 2006; Yip, 1997). With respect to our current research, if the tendency to open a relationship stems from sexual desire for other partners, then, it may be likely that the association between sexual behavior desires and enactments varies between single, monogamously partnered and non-monogamously partnered gay men.



We posit that gay men who eschew condoms will have a stronger association between their sexual behavior desires and enactments. That is, men who actively increase the sensation of their sexual experience by condom eschewal may be oriented towards fulfilling more of their desires in general; they are pleasure seekers. If, as previous research suggests, they also tend towards having more partners, then pleasure seeking may be compounded by available and willing partners who will engage in the desired behaviors.


We posit that gay men who have been diagnosed with more STIs over the past year (excluding HIV) will have a weaker association between sexual behavior desires and enactments than men who have been diagnosed with fewer STIs. Despite the barebacking trend described in the previous section associated with increased HIV infection, gay men may become more sexually cautious as the number of infections increases. Such men may report decreased behavioral enactments as a corollary.


With respect to HIV, we posit that HIV-positive gay men will have a stronger association between sexual behavior desires and enactments than HIV-negative men. It may be likely that removing the possibility of seroconverting from the enactment of sexual behaviors encourages enactment within already HIV-positive men. Conversely, HIV-negative men may be reluctant to act on certain sexual behaviors that may endanger them for HIV transmission.


We posit that younger gay men will have a stronger association between sexual behavior desires and enactments than older gay men. With a higher availability of sexual partners, young men may have more opportunities to enact sexual behaviors. Older gay men, as Berger (1996) suggests, may have to settle for whomever they can attract, regardless of whether the partner is willing to partake in the enactment of a particular desire.


With regard to relationship status, we posit that single gay men will have the strongest association between sexual behavior desires and enactments, followed by partnered men in non-monogamous relationships, and then partnered men in monogamous relationships. Single men have the active option to pursue anyone willing to satisfy their desires. Similarly, men in partnered, non-monogamous relationships have the options to both pursue anyone outside of the relationship and/or pursue their primary partners. Yet, research suggests partners tenaciously maintain codes of conduct during their encounters with third parties (Kippax et al., 1997) and condoms may be reintroduced to prevent STI/HIV infection of the primary partner (LaSala, 2005). So though partner availability may exist, the non-monogamous partner may not be able to enact some of his desires due to these constraints. Finally, men in monogamous relationships may find that some desires are unfulfilled due to their partner not wanting to enact the behavior.



Participants were solicited to fill out an Internet-based survey. Advertisements with links to the survey were strategically placed around the web. Notices were put on a gay blog page (e.g., to gain participation. Free advertisements were placed on Craigslist, in the “etcetera jobs” section of its employment opportunities pages. These Craigslist ads were placed in Atlanta, Baltimore, Chicago, Ft. Lauderdale, Palm Springs, Toronto, and Vancouver. Advertisements were placed in AOL “m4m” chat rooms, (e.g., “phoenixm4m,” or “south-carolinam4m”). A classified ad was also placed in the weekly publication, Gay Chicago Magazine. Finally, free advertisements were placed on two gay listservs for graduate schools, University of Illinois, Chicago, and University of Michigan. At the end of the survey, participants were invited to click a link in order to receive a $10.00 gift card to a popular coffee chain.

The sample consisted of 219 self-identified gay men who completed the questionnaire. As shown in Table 1, the sample was largely white and tended to live in urban areas. About half the men were single, about a third were in a monogamous relationship, and about a fifth were in a non-monogamous relationship. HIV-positive and HIV-negative men were equally represented in the sample. Though the distributions of some of the demographic variables were unequal, and in some cases skewed, the variables were not significantly related to either the independent or dependent variables. With regard to the moderators, income was correlated with age. This variable was controlled for when testing age as a moderator.

Table 1
Description of the variables


Sexual Behavior Desires and Enactments

To assess sexual behavior desires, participants were presented a list of sexual behaviors and asked to check the behaviors they desired to enact (no = 0, yes = 1). Assessed behaviors included: oral intercourse, anal intercourse, vaginal intercourse, anilingus, fisting, urination, defecation, erotic asphyxiation, domination, submission, voyeurism, exhibitionism, and sexual assault. When applicable, both receptive and insertive forms of the behavior were assessed. A follow-up question was asked that measured which of those same behaviors the participant had actually enacted over the past year (no = 0, yes = 1). We summed the desired sexual acts with good reliability, α = .85, and the enacted sexual acts with similarly good reliability, α = .83. Thus, we created the sexual behavior desires and the sexual behavior enactments scales.

Condom Eschewal

Condom eschewal was measured using a continuum from 0 to 100% of the time using 10% increments. Participants were asked the percent of time they used condoms during receptive and insertive anal intercourse as individual measures. These two scales (for receptive and insertive) were averaged with excellent reliability, α = .93.


Participants were asked to check only STIs they had been diagnosed with in the past year (no = 0, yes = 1): Chlamydia, gonorrhea, pubic crabs or lice, herpes, genital warts or asymptomatic HPV, sexually contracted hepatitis, syphilis, protozoa or fungal infection, and other (from Cates, 1999). The number of reported STIs was summed for each participant.

HIV Serostatus

Individuals self reported either being seronegative, seropositive, or sero-unknown. Sero-unknown men were omitted from the HIV-oriented analyses because of their low representation (n = 11).

Relationship Type

Individuals could report being single, in a monogamous relationship, or in a non-monogamous or open relationship.

Statistical Analysis

The data were analyzed using multivariate analysis of variance (MANOVA), bivariate correlations, and moderated and multiple regression. SPSS 11.0 was used for carrying out the analyses of variance (ANOVA), the bivariate correlations, and for breaking down and interpreting the significant interactions (as defined by Aiken & West, 1991). JMP 5.1 was used for all higher-level analyses (i.e., the moderated regression). Missing data were not found to be a significant problem. However, since the participants had the option to skip questions that did not pertain to them, the sample size varied across some of the analyses. For example, participants could skip questions on condom eschewal during anal intercourse if they had not participated in either the receptive or insertive versions of that behavior.

We used moderated regression in which an interaction term was created by multiplying the given moderator with the independent variable, sexual behavior desires (Aiken & West, 1991). Where the moderators or independent variable were inter-correlated, we controlled for those variables. Variables were all entered on different steps of the regression, with controlled variables entered before the independent variable and moderator variable. The interaction term was always entered last.

When an interaction term proved to be significant, we deconstructed the relationship between behavior desires and enact ments to measure the strength of this association at the different intervals of the moderator. For HIV and relationships status, each was broken down into its nominal variations (e.g., for HIV serostatus: seronegative and seropositive; for relationship status: single, monogamous, and non-monogramous). Dichotomizing continuous variables contributes to the loss of statistical power (Aiken & West, 1991). Thus, to interpret the interaction of continuous moderators, the relationship between behavior desire and enactment was shifted up and down by one SD rather than dichotomizing the variable. That is, all the data were transformed into high and low variations by adding or subtracting one SD to each participant's response. As a result of this transformation, we could calculate the strength of the relationship at different intervals on the continuous variable without losing power or excluding cases. High responses represented falling one SD above the mean, and low represented falling one SD below the mean (as per Aiken & West, 1991). For example, using these statistical methods on age (M = 35.34, SD = 10.33): high age (older men) would test the strength of the relationship between desires and enactments if the mean of the moderator age was shifted up by one SD to 45.67 years old and low age (younger men) would test the strength of the relationship if the mean were shifted down by one SD to 25.01 years old.


Correlations of Key Variables

Regression analysis confirmed a strong and significant association between sexual behavior desires and behavior enactments, F(1, 217) = 170.68, R2 = .44, p < .001. The following will report on whether this association was moderated by condom eschewal, STIs, HIV, age, and relationship status.

As shown in Table 2, both condom eschewal and number of STIs were positively related to an HIV-positive serostatus and age. HIV serostatus and age were related such that an HIV-positive serostatus was associated with increased age. MANOVA revealed that relationship status was associated with age [F(2, 121) = 3.70, R2 = .06, p = .02], and HIV serostatus [F(2, 121) = 7.52, R2 = .11, p < .01]—a Bonferroni-Holm correction showed that younger men were more likely to be in monogamous relationships (p < .02) and HIV-positive men were more likely to be single (p < .01).

Table 2
Correlation matrix of variables


To test the hypothesis that gay men who eschewed condoms would have a stronger association between sexual behavior desires and enactments than men who more readily used them, a model was created using sexual behavior desire, condom eschewal, and the interaction term of sexual behavior desire multiplied by condom eschewal. The model controlled for HIV serostatus and age. The overall model was significant, F(5, 118) = 45.71, R2 = .66, p < .001, and the interaction term added significantly to the fit of the model, t(118) = −3.16, ΔR2 = .03, p < .01, β = −.18.

To interpret the interaction, we broke it into those who often used condoms (shifting the mean to one SD above, or 93.88% of the time) and those who eschewed them (shifting the mean to one SD below, or using them only 11.68% of the time). As Fig. 1 shows, controlling for age and HIV serostatus, condom users showed a weaker association between the variables than condom eschewers, t(118) = 6.93, p < .001, β = .54. Condom eschewers showed a stronger association than condom users, t(118) = 11.13, p < .001, β = .89. The first hypothesis was thus confirmed by this ordinal interaction. Desires and enactments were significantly correlated for both groups, but stronger for condom eschewers.

Fig. 1
Low and high sexual behavior desires predicting sexual behavior enactments by condom eschewers and users. Note, the x-axis represents generalized low (one SD below) and high behavior desires (one SD above the mean), and the y-axis represents the actual ...


To test the hypothesis that STIs (excluding HIV) would reduce the association between behavior desires and enactments, while H3 (that HIV would increase this association), we simultaneously tested a model with both interaction terms included (HIV × desires and STIs × desires). If it were true that HIV serostatus moderated the relationship between desires and enactments, then the interaction should contribute significant variance independent of all the other variables—including the interaction term comprised of STIs and desires. Controlling for age, condom eschewal, and relationship status, we created a model that included number of STIs, HIV serostatus, sexual behavior desires, the interaction term of number of STIs multiplied by desires, and the interaction term of HIV serostatus multiplied by desires. All were entered on different steps of the regression with the HIV interaction term on the last step.

The model was highly significant, F(9, 114) = 28.36, R2 = .69, p < .001. The interaction term containing STIs and desires added significantly to the fit of the model, t(114) = −3.14, ΔR2 = .03, p < .01, β = −.18. However, the interaction term containing HIV and desires was not significant, t(114) = 1.61, ΔR2 < .01, p = .11, β = −.09. Thus, these tests confirmed the second but not the third hypothesis.

To interpret the interaction, we shifted the statistical relationship between desires and enactments by those who had not been affected by STIs (shifting the mean to one SD below, or 0), and those who had been affected by STIs (shifting the mean to one SD above, or 1.34). Though individuals could not have fractions of STIs, for the sake of creating low versus high STI incidence, we merely used the one SD above, 1.34, as a population marker representing the part of the population affected more by STIs. Figure 2 represents this deconstruction. Controlling for age, condom eschewal, monogamy, and HIV, individuals who had contracted fewer or no STIs had a stronger association between the variables, t(115) = 10.61, p < .001, β = .89. Individuals who had developed more STIs had a weaker association than those who had fewer or none, t(115) = 9.21, p < .001, β = .60.

Fig. 2
Low and high sexual behavior desires predicting sexual behavior enactments by number of acquired STIs. Note, the x-axis represents generalized low (one SD below) and high behavior desires (one SD above the mean), and the y-axis represents the actual count ...

Next, we explored whether it was possible that the association between STIs and behavior enactments was moderated by HIV serostatus (i.e., HIV-positive men with and without STIs compared to HIV-negative men with and without STIs). We created a three-way interaction by multiplying HIV serostatus, STIs, and sexual behavior desires and entered it on the last step of the moderated regression analysis. This term accounted for a statistically significant increment of variance, t(112) = 3.31, ΔR2 = .03, p < .01, β = .23. As shown in Fig. 3, among men who were HIV positive, t(27) = 8.00, p < .001, β = .78, or HIV-negative, t(33) = 8.31, p < .001, β = .84, the magnitude of the association between desires and enactments was similar as long as they had no or fewer STIs. However, when STIs affected the health of individuals, HIV serostatus became a significant moderator. No statistically significant association existed between sexual behavior desires and enactments in HIV-negative men infected with more STIs, t(9) < 1, β = .21, but among HIV-positive men infected with more STIs, there remained a statistically significant association between their sexual desires and their sexual enactments, t(33) = 6.23, p < .001, β = .70.

Fig. 3
Low and high sexual behavior desires predicting sexual behavior enactments by HIV serostatus and number of STIs/STDs. Note, the x-axis represents generalized low (one SD below) and high behavior desires (one SD above the mean), and the y-axis represents ...


To test whether younger men had a stronger association between desires and enactments relative to older men, a model containing HIV serostatus, relationship status, condom eschewal, STIs, income, age, sexual behavior desires, and the interaction term of age multiplied by sexual behavior desires was created. Though the model was significant, we were only in terested in the β and ΔR2 of the interaction term, which were not significant, t(114) = −.24, ΔR2 < .01, p = .81, β = −.01. Thus, age was not a moderator of the association between desires and enactments.


To test whether relationships status moderated the association between behavior desires and enactments, a model was constructed of HIV serostatus, age, monogamy, sexual behavior desires, and the interaction term of sexual behavior desires multiplied by relationship status. The model was significant, F(7, 211) = 31.25, R2 = .51, p < .001. However, of the three relationship statuses—single, monogamous, and non-monogamous—only the interaction with desires and the monogamous group statistically differed from the other two, t(211) = −2.64, ΔR2 = .03, p = .01, β = −.14. Men in monogamous relationships were less likely to show an association between the variables, t(59) = 5.73, p < .001, β = .59, as compared to men in non-monogamous relationships, t(34) = 10.27, p < .001, β = .88, or men who were single, t(103) = 8.60, p < .001, β = .62. There was no significant difference between the standardized beta for the single men and the standardized beta for the non-monogamous men.


A strong association was found between gay male sexual behavior desires and behavior enactments. Furthermore, this association was moderated by condom eschewal, STI contraction, and relationships status. HIV serostatus was an insufficient moderator on its own. However, it further defined STIs as a moderator. Age was not found to significantly moderate the association between behavior desires and behavior enactments. Each of these findings contributes to the general understanding of gay male sexual behavior and gives insight into the influential nature of sexual disease, safer sex choices, and relationship agreements.

Specifically, condom eschewers showed a stronger association between sexual behavior desires and enactments relative to condom users in our study. As mentioned in the introduction, some researchers have claimed that gay men who eschew condoms are in search of sexual liberation, exhibition, and freedom (Crossley, 2004; Gauthier & Forsyth, 1999). Our data lend some support to these claims. Men who consistently wear condoms show hesitancy towards risky sex. In our sample, this caution might have been transitive, affecting behavior enactments but not behavior desires. We suggest that a decreased association between one's desires and enactments might also indicate fewer feelings of sexual liberation and freedom, as desires are unfulfilled. Conversely, for condom eschewers, the data showed an extremely strong association between desires and enactments (β ≈ .90). This increased association might suggest greater sexual liberation and freedom among condom eschewers as desires are fulfilled.

STIs and HIV were predicted to moderate the association between desires and enactments. Yet, the hypothesized weaker association was found only with respect to STIs. Ostensibly, it seemed that gay men who contracted STIs recoiled from some sexual behaviors they ultimately desired. Yet, our results showed that men who had STIs and were also HIV-positive continued to show a strong association between behavior desires and enactments. In contrast, that same relationship was non-significant for HIV-negative men. In men whose HIV-positive serostatus was compounded with one or more STIs, their unchanged enactment of desires might be explained by fatalism and apathy towards one's health (as described in previous research into unprotected anal intercourse, see Kalichman, Kelly, Morgan, & Rompa, 1997). For HIV-negative men with STIs, the non-significant association between desires and enactments might have been a reaction to the STI, in which sexual desires were not enacted due to physical infection or fears of contracting future STIs. Further research is needed to understand the role of concomitant HIV infection and other STIs on the enactment of desired sexual behaviors.

It may be that time since diagnosis, adherence to HAART, viral load, CD4 count, and all the rest of the variables endemic of being HIV-positive influence the association between desires and enactments. Yet, the perceptions that HIV-positive men enacted more of their desires and were ultimately sexually freer to behave as they wished were largely unsupported by our data. An argument was presented in the introduction as to whether being HIV-positive encouraged men to enact their desires more, or whether men who enacted their desires more were simply more likely to contract HIV. In the final analysis, the condom eschewal data, the phenomenon of STIs recoiling behavioral enactment, and the negligible effect of HIV itself on enactment all support the latter contention.

Relationship type was only partially shown to influence the strength of sexual behavior desires and enactments. Specifically, men in monogamous relationships were least likely to show an association between their desires and enactments, particularly relative to single men and men in non-monogamous relationships. As a possible explanation, it might have been that the self-imposed, low partner availability affected the enactment of sexual behaviors, where each monogamous partner was limited by his partner's willingness to perform the behaviors. More generally, behavioral enactment of desires does not definitively signify sexual or relationship satisfaction—two variables that have been reported to affect relationship continuity in heterosexuals and homosexuals alike (Kurdek, 1994, 1998). However, the data did show monogamy to stifle the enactment of some sexual behaviors while keeping the desire to perform those same behaviors intact. Reasonably, in such a psychological state, lessened desired behavior enactments might ultimately lead to lessened sexual satisfaction—forcing either a reevaluation of the sort of relationship the partners wish to have or its dissolution.

The strength of all of these conclusions were limited due to our small sample size, which may have restricted the ability to detect small effects (Cohen, 1988). Yet, finding statistically significant results using only 219 men increased the confidence in our hypotheses. Additionally, the sample was recruited using a variety of ways as to ensure an international, diverse, and heterogeneous group of participants. Recruitment for this study had no impact on any of the analyses, even though some of the demographic variables were skewed (e.g., city size, race). Further, the results cannot be generalized to bisexual or heterosexual men. Future studies still might recruit from a broader range of Internet and non-Internet sites as a means to increase the number and diversity of participants.

In addition to these methodological concerns and as a function of using the online survey, it was possible that the measures of sexual behavior desires and enactments might have been affected by the anonymity provided by the Internet. Individuals may have inflated the number of sexual behaviors they actually enacted over the past year. Or alternatively, as suggested by previous research (Moskowitz, Rieger, & Roloff, 2008), by providing so many sexual behavior choices, we may have seeded or introduced sexual behavior desires that would have not been considered through using an open-ended response style. The strength of these behavior desires may thus be weakened. Future studies might correct this limitation by asking participants to self-generate sexual behavior desires.

The moderators of sexual behavior enactment were introduced as a concept under researched. In addition to this current study, there are several ways to increase the scientific understanding of sexual behavior enactments: through examining the enactment of behavior desires in other groups and through the assessment of further moderators. We used gay men exclusively for our study. The sexual behaviors of men who have sex with men aside from those who self-label as gay are only just being understood—that is, men on the “down low,” closeted gay men, etc. (see Martinez & Hosek, 2005; Millett, Malebranche, Mason, & Spikes, 2005; Pathela et al., 2006). The same-sex sexual behaviors of bisexual men, though noted as being enacted less frequently than self-labeling gay men (Stokes, Vanable, & McKirnan, 1997), are usually only discussed as they pertain to STIs/HIV (Kelly et al., 2002; O'Leary & Jones, 2006). The degree to which bisexual men, men on the down low, or closeted men enact their sexual desires (aside from engaging in protected or unprotected anal intercourse) is, as of yet, unknown. Other demographic groups should be assessed as well in future studies. Heterosexual men and women, lesbians, and different ethnic and racial groups might all display different associations between their desires and enactments when compared to each other group and the gay men in our current study.

The five moderators that we examined in this study seemed particular germane to gay men. Future research might find additional moderators that ultimately vary the strength of sexual behavior desires on behavior enactments in this population. Studies might also endeavor to find moderators that affect lesbians, and heterosexual men and women and their respective levels of sexual behavior desires relative to enactments. For example, factors such as religiosity, having children or number of children, and/or marital status may all influence the degree to which heterosexual men and women act on their sexual desires. Finally, with respect to discovering future moderators for different populations, research should explore the compounding effects of moderators on each other, such as was the case with STIs and HIV in our study.

In the final analysis, the identification of factors that encourage or discourage individuals from the enactment of their desires ultimately becomes extremely important—and not merely from perspectives within health psychology or disease prevention. Performing such research, as it did in our study regarding HIV-positive men, may help to dispel commonly held misperceptions about how different individuals sexually behave. It may also help psychologists and relationship therapists to counsel individuals who may be sexually unsatisfied or frustrated, or in romantic partnerships affected by sexual dysfunction. As we show for gay men, the availability of a willing partner is merely the most obvious moderator of sexual behavior desires and their enactments. Psychological, physiological, and social constraints have commensurable influence over the decision and willingness to turn desired sexual behaviors into enacted realities.


Special thanks to David Seal, Ph.D., for his critical feedback. Preparation of this article was supported, in part, by center grant P30-MH52776 from the National Institute of Mental Health (PI: J. A. Kelly) and by NRSA postdoctoral training grant T32-MH19985 (PI: S. Pinkerton).

Contributor Information

David A. Moskowitz, Department of Psychiatry and Behavioral Medicine, Center for AIDS Intervention Research, Medical College of Wisconsin, 2071 North Summit Avenue, Milwaukee, WI 53202, USA, e-mail: ude.wcm@iwoksomd.

Michael E. Roloff, Department of Communication Studies, Northwestern University, Evanston, IL, USA.


  • Aiken LS, West SG. Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications; 1991.
  • Alison L, Santtila P, Sandnabba NK, Nordling N. Sadomasochistically-oriented behavior: Diversity in practice and meaning. Archives of Sexual Behavior. 2001;30:1–12. [PubMed]
  • Bailey JM, Gaulin S, Agyei Y, Gladue BA. Effects of gender and sexual orientation on evolutionarily relevant aspects of human mating psychology. Journal of Personality and Social Psychology. 1994;66:1081–1093. [PubMed]
  • Berger RM. Gay and gray. New York: Harrington Park Press; 1996.
  • Blasband D, Peplau LA. Sexual exclusivity versus openness in gay male couples. Archives of Sexual Behavior. 1985;14:395–412. [PubMed]
  • Catania JA, Coates TJ, Stall R, Bye L, Kegeles SM, Capell F, et al. Changes in condom use among homosexual men in San Francisco. Health Psychology. 1991;10:190–199. [PubMed]
  • Cates W. Estimates of the incidence and prevalence of sexually transmitted diseases in the United States. Sexual Transmitted Diseases. 1999;26:S2–S7. [PubMed]
  • Cohen J. Statistical power analyses for the behavioral sciences. 2nd. Hillsdale, NJ: Lawrence Erlbaum; 1988.
  • Crosby GM, Stall RD, Paul JP, Barrett DC, Midanik LT. Condom use among gay/bisexual male substance abusers using the timeline follow-back method. Addictive Behaviors. 1996;21:249–257. [PubMed]
  • Crossley ML. Making sense of ‘barebacking’: Gay men's narratives, unsafe sex and the ‘resistance habitus’ British Journal of Social Psychology. 2004;43:225–244. [PubMed]
  • DiClemente RJ, Funkhouser E, Wingood G, Fawal H, Holmberg SD, Vermund SH. Protease inhibitor combination therapy and decreased condom us among gay men. Southern Medical Journal. 2002;95:421–425. [PubMed]
  • Ekstrand ML, Stall RD, Paul JP, Osmond DH, Coates TJ. Gay men report high rates of unprotected anal sex with partners of unknown or discordant HIV status. AIDS. 1999;13:1525–1533. [PubMed]
  • Gauthier DK, Forsyth CJ. Bareback sex, bug chasing, and the gift of death. Deviant Behavior. 1999;20:85–100.
  • Grube J. Natives and settlers: An ethnographic note on early interaction of older homosexual men with younger gay liberationists. Journal of Homosexuality. 1990;20:119–135. [PubMed]
  • Halkitis P, Wilton L, Galatowitsch P. What's in a term? How gay and bisexual men understand barebacking. Journal of Gay and Lesbian Psychotherapy. 2005;9:9–34.
  • Kalichman SC, Kelly JA, Morgan M, Rompa D. Fatalism, current life satisfaction, and risk for HIV infection among gay and bisexual men. Journal of Consulting and Clinical Psychology. 1997;65:542–546. [PubMed]
  • Kelly JA, Amirkhanian YA, McAuliffe TL, Granskaya JV, Borodkina OI, Dyatlov RV, et al. HIV risk characteristics and prevention needs in a community sample of bisexual men in St. Petersburg, Russia. AIDS Care. 2002;14:63–76. [PubMed]
  • Kelly JA, Kalichman SC. Reinforcement value of unsafe sex as a predictor of condom use and continued HIV/AIDS risk behavior among gay and bisexual men. Health Psychology. 1998;17:328–335. [PubMed]
  • Kippax S, Noble J, Prestage G, Crawford JM, Campbell D, Baxter D, et al. Sexual negotiation in the AIDS era: Negotiated safety revisited. AIDS. 1997;11:191–197. [PubMed]
  • Kippax S, Race K. Sustaining safe practice: Twenty years on. Social Science & Medicine. 2003;57:1–12. [PubMed]
  • Kurdek LA. Areas of conflict for gay, lesbian, and heterosexual couples: What couples argue about influences relationship satisfaction. Journal of Marriage and the Family. 1994;56:923–934.
  • Kurdek LA. Relationship outcomes and their predictors: Longitudinal evidence from heterosexual married, gay cohabiting, and lesbian cohabiting couples. Journal of Marriage and the Family. 1998;60:553–568.
  • Kurdek LA, Schmitt JP. Relationship quality of gay men in closed or open relationships. Journal of Homosexuality. 1985;12:85–99. [PubMed]
  • Lama J, Planelles V. Host factors influencing susceptibility to HIV infection and AIDS progression. Retrovirology. 2007;4 doi: 10.1186/1742-4690-4-52. [PMC free article] [PubMed] [Cross Ref]
  • LaSala MC. Monogamy of the heart: Extradyadic sex and gay male couples. Journal of Gay and Lesbian Social Services. 2005;17:1–24.
  • Mansergh G, Marks G. Age and risk of HIV infection in men who have sex with men. AIDS. 1998;12:1119–1128. [PubMed]
  • Mansergh G, Marks G, Colfax GN, Guzman R, Rader M, Buchbinder S. ‘Barebacking’ in a diverse sample of men who have sex with men. AIDS. 2002;16:653–659. [PubMed]
  • Martinez J, Hosek SG. An exploration of the down-low identity: Nongay-identified young African-American men who have sex with men. Journal of the National Medical Association. 2005;97:1103–1112. [PMC free article] [PubMed]
  • McAuliffe TL, Kelly JA, Sikkema KJ, Murphy DA, Winett RA, Solomon LJ, et al. Sexual HIV risk behavior levels among young and older gay men outside of AIDS epicenters: Findings of a 16-city sample. AIDS and Behavior. 1999;3:111–119.
  • Millett G, Malebranche D, Mason B, Spikes P. Focusing “down low”: Bisexual black men, HIV risk and heterosexual transmission. Journal of the National Medical Association. 2005;97:525–595. [PMC free article] [PubMed]
  • Moskowitz DA, Rieger G, Roloff ME. Tops, bottoms, and versatiles. Sexual and Relationship Therapy. 2008. in press [Epub ahead of print] Retrieved June 30, 2008, from
  • Moskowitz DA, Roloff ME. The existence of a bug chasing subculture. Culture, Health & Sexuality. 2007a;9:347–358. [PubMed]
  • Moskowitz DA, Roloff ME. The ultimate high: Sexual addiction and the bug chasing phenomenon. Sexual Addiction and Compulsivity. 2007b;14:1–20.
  • O'Leary A, Jones KT. Bisexual men and heterosexual women: How big is the bridge? How can we know? Sexual Transmitted Diseases. 2006;33:594–595. [PubMed]
  • Parsons JT, Bimbi DS. Intentional unprotected anal intercourse among men who have sex with men: Barebacking—from behavior to identity. AIDS and Behavior. 2007;11:177–187. [PubMed]
  • Pathela P, Hajat A, Schillinger J, Blank S, Sell R, Mostashari F. Discordance between sexual behavior and self-reported sexual identity: A population-based survey of New York City men. Annals of Internal Medicine. 2006;145:416–425. [PubMed]
  • Race K. Revaluation of risk among gay men. AIDS Education & Prevention. 2003;15:369–381. [PubMed]
  • Royce RA, Sena A, Cates W, Cohen MS. Sexual transmission of HIV. New England Journal of Medicine. 1997;336:1072–1078. [PubMed]
  • Sanderson T. A-Z of gay sex. London: Other Way Press; 1994.
  • Sandnabba NK, Santtila P, Nordling N. Sexual behavior and social adaptation among sadomasochistically-oriented males. Journal of Sex Research. 1999;36:273–282.
  • Shernoff M. Negotiated nonmonogamy and male couples. Family Process. 2006;45:407–418. [PubMed]
  • Stokes JP, Vanable P, McKirnan DJ. Comparing gay and bisexual men on sexual behavior, condom use and psychosocial variables related to HIV/AIDS. Archives of Sexual Behavior. 1997;16:383–397. [PubMed]
  • Tewksbury R. Bareback sex and the quest for HIV: Assessing the relationship in internet personal advertisements of men who have sex with men. Deviant Behavior. 2003;25:467–482.
  • Thompson JLP, Yager TJ, Martin JL. Estimated condom failure and frequency of condom use among gay men. American Journal of Public Health. 1993;83:1409–1413. [PubMed]
  • Valleroy LA, MacKellar DA, Karon JM, Rosen DH, McFarland W, Shehan DA, et al. HIV prevalence and associated risks in young men who have sex with men. Journal of the American Medical Association. 2000;284:198–204. [PubMed]
  • Van de Ven P, Kippax S, Crawford J, Rawstorne P, Prestage G, Grulich A, et al. In a minority of gay men, sexual risk practices indicates strategic positioning for perceived risk reduction rather than unbridled sex. AIDS Care. 2002;14:471–480. [PubMed]
  • Vincke J, Bolton R, Miller M. Younger versus older gay men: Risks, pleasures and dangers of anal sex. AIDS Care. 1997;9:217–226. [PubMed]
  • Wegesin D, Meyer-Bahlburg HFL. Top/bottom self-label anal sex practices, HIV risk and gender role identity in gay men in New York City. Journal of Psychology and Human Sexuality. 2000;12:43–62.
  • Weinberg MS, Williams CJ, Calhan C. Homosexual foot fetishism. Archives of Sexual Behavior. 1994;23:611–626. [PubMed]
  • Wolitski RJ. The emergence of barebacking among gay and bisexual men in the United States: A public health perspective. Journal of Gay and Lesbian Psychotherapy. 2005;9:9–34.
  • Worth H, Reid A, McMillan K. Somewhere over the rainbow: Love, trust and monogamy in gay relationships. Journal of Sociology. 2002;38:237–253.
  • Yip AKT. Gay male Christian couples and sexual exclusivity. Sociology. 1997;31:289–306.