Article Text

Download PDFPDF

Original article
Do differences in age between sexual partners affect sexual risk behaviour among Australian gay and bisexual men?
  1. Garrett Prestage1,2,
  2. Fengyi Jin1,
  3. Benjamin Bavinton1,
  4. Stephen Andrew Scott3,
  5. Michael Hurley2
  1. 1Kirby Institute, University of New South Wales, Kensington, New South Wales, Australia
  2. 2Australian Research Centre in Sex Health and Society, La Trobe University, Melbourne, Australia
  3. 3ACON, AIDS Council of NSW, Surry Hills, Australia
  1. Correspondence to Dr Garrett Prestage, Kirby Institute, The University of New South Wales, Kensington, NSW 2052, Australia; gprestage{at}kirby.unsw.edu.au

Abstract

Objective To describe the relationships between HIV-negative Australian gay men's sexual behaviour and differences in age between them and their sex partners.

Design Anonymous computer-assisted internet survey was administered.

Methods We used multinomial logistic regression to examine factors associated with reported differences in partners’ age (10 years older, within 10 years and 10 years younger) among 1476 men.

Results Two-thirds of the men reported their most recent casual sex partners were within 10 years of age to themselves, as were 79.3% of their regular partners. Neither men's own age nor relative differences in age with their partners was associated with the likelihood to engage in unprotected anal intercourse. Among men who reported sex with a regular partner, those with older partners tended to restrict themselves to the receptive position (adjusted ORs (AOR)=2.00; 95% CI 1.02 to 3.92; p=0.044). Among men who reported sex with a casual partner, those with younger partners tended to take the insertive position, both on occasions when a condom was used (AOR=2.42; 95% CI 1.39 to 4.20; p=0.002) and on occasions when a condom was not used (AOR=2.54; 95% CI 1.04 to 6.20; p=0.041).

Conclusions Age differences between gay men and their sex partners make little difference to the likelihood of engaging in sexual risk behaviour overall with either regular or casual partners. Those whose partners are substantially older than themselves are more likely, however, to take the receptive position during anal intercourse. Where risk of HIV infection among younger men is elevated this could be due to differences in sexual position, rather than differences in unprotected anal intercourse per se, during sex with men who are substantially older than themselves.

  • GAY MEN
  • SEXUAL BEHAVIOUR
  • HIV
  • HOMOSEXUALITY

Statistics from Altmetric.com

Introduction

Among gay and bisexual men (GBM), those who report sex with older men have been found to be more likely to engage in unprotected anal intercourse (UAI) or, specifically, receptive UAI.1 ,2 When UAI refers to sex with both regular and casual partners without regard for either sero-concordance between partners or risk reduction practices, it conceals the subtleties of perceived and actual risk.3 ,4 However, the effects of partner age differences may depend on the relationship between partners and the specific sex practices in which they engage.

Although age has been associated with risk of HIV infection among GBM in many industrialised countries, with younger men in particular thought to be at increased risk of HIV infection,5–8 in Australia, the relatively lower incidence of HIV in younger men has inflated relative HIV prevalence in older men.9 Men who were sexually active during the 1980 s and early 1990 s, at the height of the HIV epidemic, were infected at disproportionately high rates. This cohort of men—with its relatively high HIV prevalence—now mostly in their 40s and 50s, disproportionately comprises the bulk of the current category of ‘older partners’. Between 1998 and 2006 in Sydney, Australia, self-reported prevalence of HIV had declined among GBM aged under 30 years but had increased among men aged over 50 years.10 There has been, however, little evidence of age differences in men's likelihood to engage in UAI overall.11 In this context, there may be particular issues around risk behaviour and age-mixing among GBM that should be considered. Elsewhere, sex with older male partners has been associated with increased risk of infection among GBM.2 ,12–15

There has been little investigation of age-mixing as a factor in sexual risk behaviour among Australian GBM and none that directly assesses the age of sex partners.16 ,17 In this paper, we investigated the relationships between differences in age between sex partners and sexual behaviour among Australian GBM to identify whether age difference affected the likelihood to engage in sexual risk behaviour.

Methods

This was a cross-sectional study of Australian GBM conducted in 2009, using an anonymous online survey to investigate their beliefs about HIV and the risks of HIV transmission.18 No compensation for participation was provided. All participants were provided with information about what participation entailed prior to direct access to the survey questionnaire. Ethics approval was obtained from the University of New South Wales and La Trobe University.

Participants

Men were eligible for the study if they were aged 18 or above, gay- or bisexual-identified or had had sex with another man in the previous year, and lived in Australia. Participation in the study was promoted through gay community events and venues, and online via weblinks and banner ads provided on gay community websites, gay dating websites and through gay community online media. Enrolment into the study occurred online resulting in a sample of 2306 respondents in total, of whom 1638 men provided useable complete responses which included their age. There was no difference in demographic or behavioural indicators between these 1638 men and the 668 men who did not complete all questions. Of these 1638 men, 1476 men reported being HIV-negative, including all but six of the 243 men who had never been tested. HIV-positive and HIV-negative men report very different behaviours with respect to condom use and sexual positioning19 and analysis of their sexual risk behaviours should therefore be considered separately. However, the number of HIV-positive men (162) was insufficient to permit detailed analysis of age differences in relation to sexual behaviour. Analyses reported here are of the 1476 HIV-negative men.

Questionnaire and measures

The online questionnaire included demographic items, and questions on sexuality, sexual behaviour and desire, and condom use.18 Men were asked about their HIV testing history and test results. As well as their own age, men were asked what they believed to be the age of their regular partner, and of the most recent (in the previous year) casual sex partner with whom they engaged in protected anal intercourse (PAIC) and the most recent casual sex partner with whom they engaged in UAI (UAIC). All ages were reported in number of years. We calculated actual differences in partners’ age to create categorical variables indicating that partners were 10 years older, within 10 years and 10 years younger than participants. We also included a previously used measure of social engagement with gay men, which consists of two items: proportion of friends who are gay, with responses on a 5-point scale ranging from ‘none’ to ‘all’, and time spent with gay friends with responses on a 4-point scale ranging from ‘none’ to ‘a lot’.20 The Kalichman measure of sexual sensation-seeking, previously found to be strongly associated with sexual risk behaviour, was also included, the mean score for which is derived from 11 items each with responses on a 4-point scale ranging from ‘not at all’ to ‘very much’.21 Questions about sexual behaviour were asked with respect to regular partners (self-defined as an ongoing committed relationship, not necessarily monogamous) and casual partners (with whom participants did not have an ongoing sexual relationship). Separate to the questions about sexual behaviour and sexual role, men were also asked the extent to which they identified with various sexual identities, with responses scored as 0=‘Not at all’, 1=‘Somewhat’ and 2=‘Very much’; these sexual identities included the terms ‘boy’ and ‘daddy’, sexual labels often used by men who prefer younger or older partners, respectively.

Analysis

Quantitative data were analysed with SPSS software. Items considered in our analyses included: age, sexual sensation-seeking scores, identification as a ‘daddy’ or as a ‘boy’, HIV serostatus of sex partners, and sexual behaviour. These items were considered in relation to relative age differences between sex partners. Descriptive statistics were used to compare men who reported sex with partners who were at least 10 years older than themselves with men whose partners were similar in age or younger, and to compare men who reported sex with partners who were at least 10 years younger than themselves with men whose partners were similar in age or older. To estimate statistical associations, we used multinomial logistic regression to examine factors that were associated with reported differences in partners’ age (10 years older, within 10 years and 10 years younger), with those who reported partners of similar age (within 10 years) as the reference group. A multivariate model was developed including factors that were associated with differences in partners’ age with a p value of less than 0.05 in the univariate analysis, and we presented adjusted ORs and 95% CIs.

Results

The mean age among these men was 34.4 years. The majority (n=800, 54.2%) were university educated. Most identified as homosexual (n=1184, 80.2%) or bisexual (n=137, 9.3%); of the remaining 155 men, most identified either strongly (78) or somewhat (49) with the term ‘gay’. Nearly a third of the sample (n=438, 29.7%) reported that ‘most’ of their friends were gay.

Regular partners

Over half (n=822, 55.7%) of the men reported having a regular male partner. Three-quarters (n=592, 72.0%) of men with a regular partner reported having engaged in UAI with that partner (UAIR) in the previous 6 months. Respondents’ own age was not associated with men's likelihood to engage in UAIR, but it was associated with sexual position during UAIR: men who restricted themselves to the insertive position had a mean age of 37.1 years compared with a mean age of 33.5 years among men who only took the receptive position and a mean age of 33.5 years among men who were versatile, that is, they engaged in both insertive and receptive UAIR (p=0.027).

Casual partners

Two-thirds (n=1001, 67.8%) of men reported any sex with casual partners during the previous year. These men reported on up to 18 different sex practices with these partners (mean=7.3; SD=3.1). Younger men reported a slightly narrower sexual repertoire than their older counterparts: men aged under 25 reported a mean of 7.0 practices, compared with men aged 25–44 who reported a mean of 7.3 practices (OR=1.03; 95% CI 0.98 to 1.09; p=0.221), and men aged 45 or more who reported a mean of 7.7 practices (OR=1.09; 95% CI 1.02 to 1.16; p=0.008). Overall, 862 men (58.4%) reported an occasion of anal intercourse with casual partners, including 822 men (55.7%) who reported a sexual encounter involving PAIC, and 346 men (23.4%) who reported an occasion of UAIC. Among the 822 men who reported PAIC in the previous year, 313 (38.0%) also reported a UAIC event. Respondents’ own age was not associated with reporting UAIC overall, but it was marginally associated with sexual position: during both PAIC and UAIC, men who restricted themselves to the insertive position were somewhat older (mean age: PAIC=36.2 years; UAIC=37.7 years) compared with men who only took the receptive position (mean age: PAIC=33.7 years; UAIC=33.9 years) and men who were versatile (mean age: PAIC=35.4 years, p=0.070; UAIC=36.3 years; p=0.033).

Age difference between regular partners

Among the 822 men with a regular male partner, the mean perceived age of these partners was 33.9 years (SD=10.63). The majority of regular partners were similar in age to the respondents, while as many men reported having older partners as reported having younger partners (table 1). Most men aged under 25 and men aged 25–44 had regular partners within 10 years age difference (n=162, 85.3%; and n=398, 84.5%, respectively), but this was true of only 57.1% (92) men aged 45 and above (p<0.001). Most of the remaining men in the older age group (38.5%; 62 men) had partners over 10 years younger than themselves. Relative age difference was not related to either HIV seroconcordance or having engaged in UAIR (table 2). However, men whose partners were at least 10 years older were somewhat more likely to only take the receptive position during UAIR (OR=2.02; 95% CI 1.04 to 3.93; p=0.038), while those whose partners were at least 10 years younger tended to restrict themselves to the insertive position (OR=1.81; 95% CI 0.96 to 3.41; p=0.066). Age, sexual sensation-seeking and identification as a ‘boy’ or as a ‘daddy’ were associated with age differences between regular partners (table 3).

Table 1

Differences in age with regular and casual partners

Table 2

Sexual positioning and differences in partner age

Table 3

Univariate analysis of differences in partner age

In multivariate analysis, only being older was associated with having a younger partner, independent of other factors included in the model, while having an older partner was associated with taking only the receptive position during UAIR, having a higher score on sexual sensation-seeking and being younger (table 4).

Table 4

Results of multinomial regression analysis of differences in partner age

Age difference with most recent PAIC partners

The mean perceived age of PAIC partners was 32.9 years. Similar to regular partners, the majority of PAIC partners were similar in age to the respondents (table 1). Most men aged under 25 and men aged 25–44 reported PAIC partners within 10 years age difference (n=122, 73.1%; and n=381, 78.2%, respectively; p<0.001), but this was true of only 36.5% (61 men) aged 45 and above. Most of the remaining men in the older age group (n=103, 61.7%) reported partners over 10 years younger than themselves. As was the case with regular partners, men who reported PAIC with casual partners who were perceived to be younger than themselves tended to be somewhat more likely to restrict themselves to the insertive position (OR=1.68; 95% CI 0.95 to 2.97; p=0.075) and those whose casual partners were perceived to be older than themselves were more likely to restrict themselves to the receptive position (OR=1.96; 95% CI 1.27 to 3.04; p=0.003; table 2). Age and identification as a ‘boy’ or as a ‘daddy’ were associated with age differences between PAIC partners (table 3).

Perceived age difference was not associated with seroconcordance between PAIC partners. Although 541 men (65.8%) had not discussed HIV serostatus with this partner, most (92.2%; 758) believed he was also HIV-negative, regardless of age difference with that partner.

Men were also asked who initiated the decision to use a condom on this occasion. Only 130 men (15.8%) indicated that this decision was initiated by their partner. Who initiated condom use was not related to differences in age.

In multivariate analysis, having an older partner was associated with being younger, and identifying as a ‘boy’, while having a younger partner was associated with being older, and with taking only the insertive position during PAIC (table 4).

Age difference with most recent UAIC partners

The mean perceived age of UAIC partners was 33.1 years. The majority of UAIC partners were similar in age to the respondents, while twice as many men reported having younger partners as reported having older partners (table 1). Most men aged under 25 and men aged 25–44 reported UAIC partners within 10 years age difference (n=55, 75.3%; and n=142, 75.1%, respectively; p<0.001), but this was true of only 35.7% (30 men) aged 45 and above. Most of the remaining men in the older age group (n=54, 64.3%) reported partners over 10 years younger than themselves. Men who had sex with casual partners who were perceived to be younger than themselves were more likely to restrict themselves to the insertive position (OR=2.12; 95% CI 1.08 to 4.16; p=0.030; table 2). Age and identification as a ‘boy’ or as a ‘daddy’ were associated with age differences between UAIC partners (table 3).

Although 181 men (52.3%) had not discussed HIV serostatus with this partner, most (n=316, 91.3%) believed he was also HIV-negative. Men whose partners were perceived to be at least 10 years older were more likely than other men to believe that their partners were HIV-negative (OR 2.86; 95% CI 0.85 to 9.65; p=0.011).

Only 59 men (17.1%) indicated that the decision not to use a condom was initiated by their partner on this occasion. Who initiated this decision was not related to differences in age.

In multivariate analysis, having an older partner was associated with being younger, while having a younger partner was associated with being older, and with taking only the insertive position during UAIC (table 4).

Discussion

As has previously been found in Australian research among GBM,11 age was not associated with UAI in general, although it was associated with sexual position during anal intercourse. In other previous work, where partners’ age was not directly reported, we have identified that having a partner who is ‘much older’ than oneself was associated with increased risk of HIV infection, even though respondents’ own age was not associated with this risk.16 The data presented here suggest that this elevated risk of infection may not necessarily be due to a greater likelihood to engage in UAI when having sex with older partners, or to greater risk among young men, per se. It may be, instead, that sexual position in general is related to relative age differences among some gay men, and that, when the perceived age difference is substantial, men who are older than their partners tend to take the insertive role more often than their younger partners. This appears to be the case here, regardless of the respondents’ own age. Nonetheless, there was also a tendency for younger men to take the receptive position during UAIC, which has also been reported elsewhere.1 ,2 So, differences in sexual positioning may be due to differences in age between partners but they may also differ according to men's own age. However, even if sexual risk behaviour is similar regardless of differences in age between partners, given that HIV prevalence is substantially higher among older men in general, HIV-negative men who engage in UAI with older partners have a greater chance of doing so with an HIV-positive partner.

Our data do not appear to support any assumptions about ‘age-based power dynamics’ in relation to condom use: decisions about condom use were unrelated to differences in partner age. It is nonetheless possible that many younger men cede this sort of sexual decision-making to their older partners purposefully: the ‘daddy/boy’ role-play fantasy common among some GBM implies that the decisions about the types of sex practices and their contexts will mostly be determined by the ‘daddy’, although the reality may not always conform to that particular stereotype.22 ,23 Decisions about sexual position could have been based on relative power dynamics, with older partners preferring the insertive position. However, this assumes that the insertive position was indeed the preferred position, regardless of relative age, but we cannot know from these data which partner actually made the decision about sexual positioning. The ‘daddy/boy’ fantasy is relational (one requires the other) and sexual position may be due to the agency of the younger partner as much as that of the older partner, and should be explored further.

Given that men with substantially older partners tended to identify with the sexual label of ‘boy’, and those with substantially younger partners tended to identify as ‘daddies’, the relative age of their partners may reflect aspects of some men's own sexualities as a clear sexual preference. For many men, the ‘daddy/boy’ sexual fantasy includes the expectation that the ‘daddy’ would be most likely to ‘top’ (take the insertive position), while the ‘boy’ would ‘bottom’ (take the receptive position), particularly when constructed around concepts of dominant and submissive roles where the younger partner may purposefully seek to be dominated. This is, of course, not always true for all men having sex with men of a substantially different age, but this assumption about sexual positions is most likely common among those who do personally identify with the ‘daddy/boy’ fantasy (and who purposefully select partners much older or younger than themselves to enact that fantasy).

Despite the various possible implications of the ‘daddy/boy’ fantasy, most men did mainly have sex with men who were similar in age, which has been reported elsewhere.24 However, while most men in our sample aged below 45 mainly had sex with partners of a similar age to themselves, men above 45 increasingly tended to report sex with men who were substantially younger than themselves, which has also been observed elsewhere.25–28 This may be due to a general preference for youthful partners overall, or it could reflect the greater availability of younger men as older men ‘retire’ from the active pursuit of new sex partners, or it may also reflect their actual age preferences. Nonetheless, some younger men specifically prefer older partners.27–29

Our analysis incorporated concepts of sexual cultures and identity, in addition to measuring behaviour and age difference directly. Men were unlikely to have concerns about confidentiality in this anonymous, online cross-sectional survey, but, as such, it is not possible to determine any causative relationships in the data. Participants were broadly similar to other samples of Australian gay men.4 ,18 ,20 Despite the broad base and large size of the sample, the main limitations of this study were that, as a volunteer sample, it may not be representative of all homosexually active men in Australia. Also, there may be differences between the Australian context and many other equivalent countries: There has been little evidence of increased rates of HIV infection among younger GBM in Australia, whereas in other countries this has been a significant issue.

These data suggest important contextual insights for health promotion. HIV prevention programmes among GBM should not assume that men only have sex with other similar aged men. For some younger men, there may be a degree of social stigma attached to sex with (much) older men among their peers which may need to be challenged to enable HIV-prevention messages to effectively address men in such situations. Also, the belief that younger GBM lack power during sex with older men may need to be questioned. These data demonstrate that sex with older men does not directly equate to higher risk for younger men. While there may be differences in the patterns of sexual positioning, it should not be assumed that younger men are always less powerful. Finally, while it is important not to reinforce an assumption that HIV mostly affects older GBM, younger men who explicitly desire sex with (much) older men should be aware of the higher background prevalence of HIV among older GBM.

The existence of ‘daddy/boy’ sexual cultures among GBM, and the identification with these labels within this sample, offers potential cues for health promotion. These data suggest that seeking sexual partners where there may be a substantial age difference constitutes an aspect of some GBM's sexual experience that may be marginal to that of many of their peers but is nonetheless important, both to themselves and to gay sexual cultures in general. The marginality of age-differentiated sexual encounters and their potential for experiencing social stigma suggest that health promotion may need to specifically target GBM engaged in age-differentiated sex and to address the values and beliefs associated with such behaviour. ‘Daddy/boy’ sexual cultures can be found among the internet and phone-based portals that accommodate the various fetishes, identities and fantasies in GBM's sexual cultures. These may present opportunities for health promotion with this group. Visual representations in health promotion materials may also need to be adapted to reflect these men's sexual experience. Peer-based health promotion may need to facilitate discussion of age differentiation, and its potential challenges, in sex between GBM.

Conclusions

When there are substantial differences in age, these have little impact on the likelihood to engage in UAI overall, with either regular or casual partners. However, when having sex with partners substantially older than themselves, men are more likely to take the receptive position during anal intercourse. Where younger men are at an increased risk of infection may be due to differences in sexual position in the context of sex with men who are substantially older than themselves rather than increased likelihood to engage in UAI in general. Nonetheless, for many men, both younger and older, when having sex with partners who are substantially different in age to themselves, this choice of partner may be purposeful and reflect aspects of their own sexualities, including the sexual roles they choose to take during sex.

Key messages

  • Gay men tend to have sex with partners of a similar age to themselves.

  • Age differences between gay men and their sex partners make little difference to the likelihood of engaging in sexual risk behaviour overall.

  • When having anal intercourse with partners substantially older than themselves, gay men are more likely to take the receptive position.

  • Increased risk of infection among younger gay men during sex with substantially older men may be due to differences in sexual position.

References

Footnotes

  • Handling editor Jackie A Cassell

  • Contributors GP and MH were the principal investigators. GP designed the data collection instruments and supervised the study implementation. GP analysed the data, and FJ provided advice for statistical analyses. GP wrote the first draft of this manuscript. All authors contributed to and approved the final manuscript.

  • Funding This study was commissioned by and funded by the Departments of Health in the states of NSW, Victoria, South Australia and Western Australia. Indirect support was also provided by the Queensland Health Department. The Kirby Institute is affiliated with the Faculty of Medicine, University of New South Wales. The Australian Research Centre in Sex Health and Society (ARCSHS) is affiliated with the Faculty of Health Sciences, La Trobe University. The Kirby Institute and ARCSHS receive funding from the Commonwealth of Australia Department of Health and Ageing.

  • Competing interests None.

  • Ethics approval University of New South Wales.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.