A Comparative Demographic and Sexual Profile of Older Homosexually Active Men Author(s): Paul Van de Ven, Pamela Rodden, June Crawford and Susan Kippax Source: The Journal of Sex Research, Vol. 34, No. 4 (1997), pp. 349-360 Published by: Taylor & Francis, Ltd. Stable URL: http://www.jstor.org/stable/3813477 Accessed: 25-09-2015 03:27 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Sex Research. http://www.jstor.org This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions The Journal of Sex Research Vol. 34, No. 4, 1997 pp. 349-360 A Comparative Demographic and Sexual Profile of Older Homosexually Active Men Paul Van de Ven Pamela Rodden June Crawford Susan Kippax National Centrein HIV Social Research School of Behavioural Sciences Macquarie University Our analysis draws a comparativeprofile of older homosexuallyactive men. For an Australian national telephonesur- vey (ProjectMale-Call), 2,583 homosexuallyactive men were interviewed.Questions about demographics,types of sex- ual partners, attachment to gay community,HIVIAIDS, and sexual practices were asked. About 10%(n = 256) of the Male-Call men wereover 49 years. These older men were likely to live alone (52.7%o), to be or have been married (62.9%1c), to have children (56.4%o), and to have lived at their present address for more than five years (67.5%o). Relatively few (12.4%)lived in gay areas, but a significant number (29.21%o) lived in rural regions. They weregenerally less likely than younger men to have disclosed their sexual orientation (p < .00005). Although their attachment to gay communitywas quite strong, it was less than younger men's in terms of social attachment (p < .00001), cultural involvement(p < .001), and sexual involvement(p < .00005). As a group, they were less likely to have been testedfor HIV antibodies (p < .0005). Older men had as many male and female sexual partners in the past six months as did younger men. They had a nar- rowerrange of anal (p < .00005) and oral /tactile (p < .001) homosexualpractices, but differencesin oral /tactile reper- toire were not significant after controlling for other differences between younger and older men. There were no significant age differencesin rates of condom use during anal intercoursewith regular or casual male partners; how- ever, the older men were morelikely to have no anal intercoursewith casual partners (p < .005). Weconcludedthat older homosexuallyactive men are fairly closely attached to gay community.They are sexually active, albeit with a less ex- tensive range of anal practices than their younger counterparts.Although they are generally as safe in sexual conduct as youngermen, education campaigns targetingolder men would benefitfrom using a varietyof metropolitanand rural, mainstream and gay media to improvesafe sex understanding and encourageHIV antibody testing. ver the past two decades or so, compare younger and older men's this stereotypical picture. Weinberg much has been written about the sexual identities, attachments to gay and Williams (1974), who explored sexual identities and behaviours of community, sexual relations and homosexual adaptations in the United homosexually active men. The litera- practices, HIV testing, and contact States, the Netherlands, and Den- ture has flourished especially since with the epidemic. (Throughout this mark, reported that older (over 45 the advent of HIV/AIDS and the at- article, we refer to gay community years) homosexual men were higher tendant concerns to understand the rather than the gay community to on some measures of psychological social, psychological, and behavioural reflect a diversity of communities and well-being than were younger homo- dimensions of gay men's lives. Various the fact that gay community is not all sexual men. Their data did not sup- cohorts such as young gay men, gay of a piece.) We were interested to see port the image of the older homosexual men from language backgrounds if older men differed from younger man as lacking in self acceptance or other than English, and men who men, particularly as age differences being anxious, depressed, lonely, and have sex with both men and women might signal the importance of cul- unhappy. However, they found that have been the focus of numerous and tural norms and social processes in older homosexual men attended gay detailed research studies. From a sex- shaping who homosexual men are ual practices viewpoint, one group, and what they do. however, has been almost entirely As other researchers have pointed We gratefully acknowledge those who neglected. There is no comprehensive out (Bennett & Thompson, 1980; made the study possible: the participants; account of the social and sexual lives staff and volunteers of various AIDS councils Berger, 1980; Berger & Kelly, 1986; and organisations;and those who promoted, of older homosexually active men, a Friend, 1987), the stereotype of the recruited, and interviewed for the survey. situation that we sought to redress older homosexual man has been one Correspondencemay be addressed to Paul by comparisons between younger of disengagement from homosexual Van de Ven, Ph.D., National Centre in HIV and older homosexually active men. community, loneliness, rejection, de- Social Research, School of Behavioural Sci- Specifically, we analysed data from pression, and unhappiness. Several ences, Macquarie University, Sydney, NSW an Australian nationwide sample to 2109, Australia. E-mail: paul.vandeven@ empirical studies have contradicted mq.edu.au. 349 This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 350 Older Homosexually Active Men venues less often, were more likely mon in pre-liberation times. Role laws that restricted the emotional to live alone, reported homosexual playing in relationships was highly and sexual expression of male homo- contacts less frequently, and, in the atypical among younger and older sexuality, and the development of United States, were lower in overall homosexual men alike. As a result of gay community.As several accounts social involvement in homosexual intense social pressures in the past, in Wotherspoon's(1986) collection of communitythan were youngerhomo- many older homosexual men had autobiographical essays exemplify, sexual men. been or currentlywere married. Con- these historical events have pro- In a later study involving Aus- tinuing sexual interest and activity duced an apparent generational shift tralian men, Bennett and Thompson were the norm for older homosexual for gay men that has made it some- (1980) found no evidence that older men. Moreover,integrationinto a local what easier for those growingup with (over 45 years) homosexual men dis- homosexual community was associ- homosexual desires to express and engaged, or were forcedto disengage, ated with psychosocialadaptationfor be more open about their sexual ori- from the homosocial aspects of gay older gay men. entation. community.In comparisonwith their It has been suggested that coming Whereas the literature provides younger counterparts, older homo- to terms with homosexuality may useful insights into some social and sexual men were as involved in the facilitate adjusting to the aging psychological aspects of older homo- homosexual world, reported a simi- process in a number of ways (Friend, sexually active men's lives, there is a lar incidence of visiting gay venues, 1987, 1989; Lee, 1987; McDougall, conspicuouslack of detail about their perceived their popularity among 1993). Successful management of the sexual relations and practices. In the other homosexuals to be as high, and process of coming out in a heterosex- era of HIV and AIDS, this is a serious were as satisfied with their sexual ist world possibly provides the indi- omission. To remedy this deficiency, orientation. Older men were less vidual with coping mechanisms that we analysed data collectedin 1992 as likely than those who were 26 to 35 may generalise to other crises in later partof ProjectMale-Call,an Australian years old to share accommodation life. Gender-roleflexibility may allow national telephone survey of men who with lovers and more likely to live older homosexual men to develop have sex with men (Kippax,Crawford, alone. Reflecting a possible genera- ways of taking care of themselves Rodden, & Benton, 1994). A 42-page tional difference,the older men were that feel comfortableand appropriate. interview schedule was used. It cov- significantlymore concernedthan the Homosexual men often exchange the ered a range of issues related to sex- younger men about the exposure of family of birth supportsthat were lost ual behaviour (nature of sexual their homosexuality. when they came out with a broader relationships, sexual history, sexual Berger (1980) studied 112 homo- surrogate family and community practices with men and with women, sexual men between the ages of 41 network.Otherstrengthsmay include safe sex strategies, attachment to gay and 77. Whereas38%of the men lived learning to fend for oneself from an community,degreeof contactwith the alone, the majorityeither lived with a early age, increased personal auton- epidemic,HIV test result) and a num- sexual partner(43%)or with friendsor omy, and learning to live with a stig- ber of demographicvariables. (Some members of their family of birth matised identity throughout life. countries use safe, and others use (19%).Far frombeing isolated and re- These competencies and reconstruc- safer, as the term to describe risk-re- jected by youngermen, most men had tions tend to be associated with duction strategies with reference to many friends, were sexually active, healthy psychological adaptation by HIV transmission, and the same ar- and sustained close relationshipswith older gay men (Friend, 1989; Mc- guments are used for and against ei- friends and sexual partners. Dougall, 1993). ther term. As used in this article, it is Through a comprehensive review The current generation of older Australian policy to referto safe sex.) of the literature, Berger and Kelly homosexually active men has lived Whereasour focus was on older (over (1986) discredited some popularmis- through remarkable changes in per- 49 years) homosexually active men, conceptions and provided a realistic ceptions and attitudes toward homo- we realise that people don't change picture of aspects of the lives of older sexuality, not least the breaking of suddenlyat age 50. Forthis reason,we gay men. They found that few older the "wall of silence" that once pre- tested for a numberof developmental gay men were loners. In fact, older cluded any public discussion of homo- changesoverfive age groups:less than gay men had more options in rela- sexuality (Wotherspoon, 1986). The 25, 25-29, 30-39, 40-49, and over 49 tionships than did their heterosexual past three decades also have wit- years. However,for expositional pur- counterparts and revealed a variety nessed the sexual revolution of the poses,we describemost resultsby talk- of living arrangements accordingly. 1960s and 1970s, the establishment ing about older men versus younger Older homosexual men in a relation- of homophile organisations concur- men. ship were not more likely to play a rent with the Stonewall Riots in Based on limited although fairly role along the lines of the active/pas- New York in 1969, the dawn of gay consistent previousfindings, we test- sive dichotomy assumed to be com- political activism, alterations to the ed a number of specific hypotheses This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions  Van de Ven, Rodden, Crawford,and Kippax 351 about older (over 49 years) men's, cluded sections of the organised gay chance to read through the list, indi- comparedwith their younger counter- community (radio, venues, gyms, cating generallyclear senses of sexual parts', sexual relations and practices: businesses, publications); places of identity. Oldermen are morelikely to live alone sexual contact within, outside, and Men were asked abouttheir regular rather than with a sexual partner marginal to organised gay communi- and casual male partners. The use of (Hypothesis 1). Older men are more ties (gay brothels, sex shops, beats, the term casual partnership did not likely to have been or to be married saunas);health centres frequentedby imply any judgment about the degree (Hypothesis 2). Older men are less gay men; and pornography outlets. of seriousness of the sexual action likely to have disclosed their homo- Strategies for contacting potential between a man and his casual partner, sexual orientation to others (Hypoth- respondentsincludedstandardadver- nor of the moral worth of casual as esis 3). Older men are as attached to tising that emphasisedthe importance opposed to regular relationships. It gay communityas younger men (Hy- of research information, advertising was meant merely to denote an occa- pothesis 4). Older men are likely to with some sexual titillation, cardsand sional sexual partner as distinct from have had the same number of male fliers, stickers and posters, notices in a regular or steady partner in a com- partners in the past six months as Personals columns, and articles and mitted ongoing relationship. younger men (Hypothesis 5). interviews in the media. Gay identity disclosure was mea- Withouta bodyof evidenceon older sured with a seven-item scale con- Measures homosexually active men's sexual structed from questions about people practices,we posed the researchques- The interview schedule (available the respondent had told about his tion: Do older homosexually active from the first author) consisted of 42 homosexual practice (mother,father, men engage in the same sexual prac- pages of questionsbased on those used otherrelatives, straight friends,work- tices as their younger counterparts successfully in the Social Aspects of mates, neighbours, anyone else). The (ResearchQuestion 1)?In the absence the Prevention of AIDS study and range of possible scores was 0 to 7, of previous data pertaining specifi- the follow-up telephone survey, Sus- with a higher scoreindicatinggreater cally to older homosexually active taining Safe Sex (Kippax, Connell, disclosure. Cronbach'salpha was .82. men's behaviours in relation to HIV/ Dowsett, & Crawford, 1993). The The mean score for the total sample AIDS, we also posed three additional schedule included a number of demo- was 2.85, with a standard deviation research questions. Are older homo- graphic variables (age category as of 2.25. sexually active men, compared with employed in the Australian Census, Social attachment was measured their younger counterparts, as likely place of birth, educationallevel, occu- with a scale that included eight items to have been tested for HIV antibod- pation, income bracket, place of resi- about the amount of free time spent ies (Research Question 2)? Are older dence, heterosexual marriage) and with gay men, number of gay friends, men as responsive to safe sex cam- items used to constructmilieu or con- and where the respondent went with paigns, as measured by condom use text variables (sexual identity and gay friends (gay bars, discos, parties, with regular and casual male part- disclosure; gay community attach- pool/beach, meetings/organisations). ners (Research Question 3)? Are ment as measured by social attach- The range of possible scores was 0 to older men as likely to have had sim- ment, cultural involvement, and 13, with a higher score indicating a ilar levels of contact with the epi- sexual involvement scales). These greater degree of social attachment demic (Research Question 4)? two sets of variables may be distin- to gay community. Cronbach'salpha Method guished from the other variables of was .86. The mean score for the total interest (outcome measures), which sample was 7.32, with a standard included sexual practice, number of deviation of 3.63. Participants sexual partners, frequency of sex, Cultural involvement was mea- The analysis was based on tele- condomuse, HIV antibodystatus, and sured with a scale that contained phone interviews with 2,583 homo- degree of contact with the epidemic. three items: membership in gay sexually active men, of whom 2,580 Measures and scales that are not organisations, perception of belong- provided age data and thus could be self-explanatory are described next. ing to gay community,and readership included in the analysis. Men were For sexual identity, men were of gay press. The range of possible included in the survey according to asked if they thought of themselves scores was 0 to 3, with a higher score the criterion that they had had sex as any of the following, in this order: indicating a greater cultural involve- with at least one other man during heterosexual, straight, bisexual, gay, ment in gay community. Cronbach's the five years prior to interview. homosexual, camp, queer, other (par- alpha was .62. The mean scorefor the There is no way of enumerating ticipant specified), and don't know/ total sample was 1.86, with a stan- the populationof homosexuallyactive unsure. It was common practice for dard deviation of 1.02. men. Participants were necessarily the men to nominate their sexual Sexual involvement measured de- volunteers. Recruitment sources in- identity before the interviewer had a gree of immersion in the sexual ele- This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 352 Older Homosexually Active Men ments of gay communitywith a scale indicating engagement in a larger Male interviewers did not raise the that contained 16 items: 14 about number of anal practices with male issue of the sex of the interviewer, where respondents went to look for partners. Cronbach'salpha was .80. but female interviewersalways asked male sexual partners (e.g., bars, The overall mean score was 3.78, whetherthe callerwouldprefera male saunas, cruising areas, sex clubs), 1 with a standard deviation of 2.49. interviewer. about the number of male sexual The Oral/Tactile Practice Scale Results partners in the past 6 months, and 1 contained six items about wet and concerning frequency of sex with dry kissing, insertive and receptive The men were categorisedinto one male casual partners in the previous oral-genital sex, mutual masturba- of five age groups: less than 25, month. The range of possible scores tion, and sensuous touching. The 25-29, 30-39, 40-49, and over 49 was 14 to 35, with a higher score in- range of possible scores was 0 to 6, years. (Equal intervals of 10 years dicating greater sexual involvement with a higher score indicating en- produced highly unequal ns.) Uni- in gay community-looking for sexual gagement in a larger number of variate differences between the men partners in a greater range of sexual oral/tactile practices with male part- in the different age groups were in- sites, having a greaternumberof male ners. Cronbach'salpha was .85. The vestigated by Chi-square tests of as- partners,and having more male part- overall mean score was 5.20, with a sociationfor categoricalvariables and ners per month. Cronbach's alpha standard deviation of 1.53. by analyses of variance (ANOVAs) was .80. The mean score for the total fornumericalvariables.In accordwith Procedure the Bonferroni principle, alpha was sample was 20.14, with a standard deviation of 3.91. Project Male-Call was a national set at .001 for these comparisonsbe- Contact with the epidemic was telephone survey of gay and homo- cause of the large number of statisti- measured with a scale that contained sexually active men in Australia. Re- cal tests applied to the data. Where three items: knowing a person who cruitment,includingadvertising,took feasible, linear trends across the age is seropositive or living with AIDS, place in all states and territories, all groups were tested by the Mantel- knowinga personwho has died follow- capital cities, and urban and rural Haenszel statistic (categorical data) ing AIDS, and having been involved areas. National, regional, and local or F-ratio for linear trend in ANOVA in caring for someonewith AIDS. The recruitmentstrategies were used. The (numerical data). A multivariate range of possible scores was 0 to 3, survey was conductedso as not to co- analysis of covariance (MANCOVA) with a higher scoreindicatinggreater incide with the majorholiday periods was also conducted to see whether contact.Cronbach'salphawas .68. The of Easter and Christmas. there were any age group differences mean score for the total sample was To safeguard respondents' ano- on the sexual practices with men 1.33, with a standard deviation of nymity, attract a wide cross-section scales (Anal Practiceand Oral/Tactile 1.12. of homosexually active men, gener- Practice) after controllingfor a num- A number of sexual practice indi- ate an atmosphere in which the men ber of demographicand milieu vari- cators and measures was employed. could speak openly and honestly ables. Data were collectedon the number of about their sexual practices and life Demographics male and female partners,the nature situations, and minimise costs, tele- of men's sexual relationships(regular phone rather than face-to-faceinter- There were 256 older (over 49 or casual), condomuse, and, concern- views were used. For May and June years) men in the Male-Call sample, ing male partners,the frequencyof 10 1992, eight 008 (free of charge to representing just under 10% of the sexual behaviours (sensuous touch- caller)telephonelines were connected sample of 2,583 men. In each of the at MacquarieUniversity. Trainedin- other age bands of less than 25, ing, kissing, masturbation, oral-gen- ital sex with and without semen terviewers informed callers that the 25-29, 30-39, and 40-49 years of exchange, finger fucking, anal inter- survey would take around half to age, there were 529, 512, 767, and course with and without ejaculation, three quarters of an hour and asked 516 men, respectively. Compared the caller if he had that time avail- with Australian Census data, the rimming, and fisting) with regular and casual partners. Sexual practice able at present. Most callers were in- sample underrepresented men over with men was also described with terviewed immediately. Others chose 49 years of age (Census = 30.3%). reference to Anal Practice and Oral/ to call back. Moreover, there was an overrepre- Most interviewerswere men. Some sentation of tertiary educated and Tactile Practice Scales. The Anal Practice Scale contained female interviewers and one trans- professionalmen in the whole sample eight items about insertive and re- sexual interviewer were also em- compared with Australian averages (Kippax et al., 1994). ceptive anal intercourse with and ployed. Some participantschose to be More than three quarters of the without ejaculation, rimming, and interviewed by a female interviewer, older men (fewer than the younger fingering. The range of possible others insisted on a male interviewer, scores was 0 to 8, with a higher score but most did not express a preference. men) had been born in Australia or This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions  Van de Ven, Rodden, Crawford,and Kippax 353 New Zealand (78.1%);a proportion- Table 1 ately high number in the United Gay Identity Disclosure Scale by Age Kingdom or elsewhere in Europe Under 25 25-29 30-39 40-49 50 or over (19.5%);and relatively few in other n = 492 n = 479 n = 714 n = 472 n = 222 countries (in Asia, Africa, Oceania, M 2.99 3.25 2.97 2.57 1.88 or North, Central, or South America; SD 2.23 2.21 2.29 2.24 1.90 2.3%).This represented a significant F(4, 2374) = 17.56, p < .00005 differencebetween the age groups in country of birth, X2(16df, n = 2577) = 76.27, p < .00005. Most older men were in paid em- higher education;and the remainder Milieu ployment, either full time (37.5%), (16.8%) had some other post-sec- part time (8.2%), or self-employed Sexual identity and disclosure. ondary qualifications, X2(12df, n = When asked to describe their sexual (16.4%), although a considerable 2575) = 104.49, p < .00005. number (29.7%)-more than for any More than half the older men identity, 60.9%of the older men iden- age other group-were receiving so- tified as gay or homosexualand 28.5% cial security payments or were out of (52.7%), a greater proportion than as bisexual, proportions similar to for any other age group, lived alone. other men over 30 years. Younger the workforce, X2(24 df, n = 2577) = Other older men, more so than those 740.80, p < .00005. Older men were men, particularly those under 30 under 30 years, lived with a female about as likely to have had no occu- sexual partner(19.9%)or male sexual years, were more likely to embrace pation (33.3%; a similar proportion the termgay rather than homosexual, partner(13.7%).Few in this age group to those under 25 years) as they lived with platonic friends, either gay X2(24 df, n = 2577) = 72.28, p < were to have been employed in man- .00005. Only 3.9% of the older men or straight, or natal family members, identified as heterosexual, a similar agerial or professional occupations X2(28 df, n = 2578) = 459.10, p < (30.6%;less than men 30-49 years); .00005. By Mantel-Haenszel test, proportion to the other age groups. 9.5%of them, less than for the other The older men were by far the least there was an increasingtrendforolder age groups, were employed in para- men to live alone (p < .00005), and likely to have disclosed their homo- professional or clerical occupations, sexual orientation to others (see correspondingdecreasing trends for Table 1). By F-ratio for linear trend, X2(16 df, n = 2567) = 272.14, p < men to live with natal familymembers .00005. Their representation in the there was a significant decreasing other occupational groups of trade/ (p < .00005) or friends (p < .00005), linear trend for Gay Identity Disclo- manual and sales/service was simi- confirmingHypothesis 1. sure (p < .000005). Moreover,there lar to that of their younger counter- Relatively few older men lived in was a significant quadratic relation- predominantly gay areas (12.4%), ship between Gay Identity Disclosure parts. whereas there was an overrepresen- The distribution of older men's in- and age (p < .000005), resulting from tation of older men in rural areas 25- to 29-year-old men having dis- comes was more even than for any other age group, X2(12df, n = 2515) (29.2%), X2(8 df, n = 2548) = 51.31, closed their identity more often than = 369.09, p < .00005, with approxi- p < .00005. By Mantel-Haenszel test, a significant trend indicated that the younger or older men (partly sup- mately a quarter of them falling into older the men were, the more likely porting Hypothesis 3). each of the four income categories: Gay community attachment. At- 25.3% (a higher proportion than all they were to live in rural areas (p < tachment to gay community,whether but the under 25 years age group) .00005). measured in terms of social attach- Older men were more likely than earned less than $15,000 per year, their younger counterparts to be or ment, cultural involvement,or sexual and another 25.3% earned more have been married (62.9%),x2(4df, n involvement, was less for older men than $40,000 (with 23 of these men than for younger men (see Table 2). = 2572) = 487.96, p < .00005 (con- having earned more than $60,000); By F-ratiofor linear trend, there were 21.2% earned $15,001-$26,000, and firming Hypothesis 2). Correspond- significant decreasing linear trends, 28.2% earned $26,001-$40,000. (At ingly, the older respondents were with age, in social (p < .000005), cul- more likely to have children (56.4%), tural (p < .0001), and sexual involve- the time, US$1 was approximately x2(4 df, n = 2544) = 454.68, p < ment (p < .005) in gay community. equivalent to Aust$1.30.) About a .00005. Older men were more likely third (32.8%)of the older men, more For sexual involvement alone, there to have lived at their present ad- was a significant quadratic relation- than for the younger men, had edu- dress for more than five years cation up to Year 10 only; about a ship between involvement and age (67.5%), X2(16 df, n = 2569) = 389.05, quarter (23.4%) had completed the (p < .000005), indicating peak sexual p < .00005. involvement in gay community for equivalent of Year 11 or Year 12; more than a quarter (27.0%) had those in the 25-29 age groupand less This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 354 Older Homosexually Active Men Table 2 In terms of female partners, 42 of Gay Community Attachment Scales by Age the 256 older men (16.4%)had never Under 25 25-29 30-39 40-49 50 or over had sex with a woman, whereas 50 Social Attachmenta n = 529 n = 512 n = 764 n = 516 n = 256 men (19.5%)had had sex with a sole M 8.11 8.13 7.22 6.61 5.79 female partner. A further 97 men SD 3.49 3.48 3.67 3.64 3.25 (37.9%) had had between 2 and 10 Cultural Involvementb n = 529 n = 512 n = 766 n = 516 n = 256 female partners ever. The older men M 1.91 1.98 1.84 1.80 1.67 (including those 40-49 years of age) SD 0.98 1.04 1.02 1.04 0.96 were likely to have had more female Sexual Involvementc n = 510 n = 501 n = 746 n = 503 n = 245 partners in their lifetime than their M 19.84 20.56 20.40 20.16 19.08 younger counterparts, particularly SD 3.93 3.95 4.08 3.66 3.51 so when compared with those under aF(4, 2572) = 30.44, p < .00005 30 years, X2(24df, n = 2579) = 133.42, bF(4, 2574) = 4.99, p < .001 p < .0005. However, the older men CF(4, 2500) = 7.65, p < .00005 were no more or less likely than the men in the younger age groups to involvement for those in younger or had caslual sex only. They, as well as have had sex with a woman in the older groups. men under 25 years, were more likely six months prior to interview: 29.0% When the scores on the social at- than th.e other age groups to be in a of the entire sample, and 28.5%of the tachment scale were used to divide monogaimous relationship and less older men, had, X2(16 df, n = 2576) = the sample into two groups, gay com- likely t,o be in a regular plus casual 23.83, p = .09. munity attached (GCA;social attach- relatiorlship. Compared with men under 30 ment > 4) versus non-gay community Numibersof partners/frequency of years, older men were more likely to attached (NGCA; social attachment sex. Nott unexpectedly,older men had have had sex with their regular part- <4), 62.9% of the older men (signifi- more rrlale sexual partners in their ner 1-5 times per month (39.3% of cantly less than for younger groups) lifetime than younger men, X2(32df, those with regular partners) and less were classified as GCA, and 37.1% n = 25773) = 251.09, p < .00005, but likely to have had sex more than 15 were NGCA, Z2(4df, n = 2580) = 61.37, there wvereno significant age differ- times (19.7%), X2(16 df, n = 1157) = p < .00005 (contraryto Hypothesis 4). ences iin the number of male sexual 52.71,p <.00005. Nevertheless, older Older men were the least likely to partnerrs in the six months prior to men had sex as frequently as those have named the gay press (11.8%)and intervie:w, X2(20df, n = 2578) = 28.39, aged 40-49 years, with 26.5% hav- cards/posters(1.2%)as the source of p = .10 4(confirmingHypothesis 5). Al- ing had sex 6-15 times per month. informationabout the current survey. most thireequarters of the older men There were no age differences in the Conversely,they were the most likely had eitlher 1 (28.5%)or between 2-10 frequencyof sex with casual partners, to have named mainstream papers (44.9%) partners during the preced- and almost half (48.9%)of the older (11.4%) and mail order video and ing 6 mionths.For the older men, the men had sex with a casual partner other catalogues(36.2%),X2(44df, n = modal rrangefor number of male sex- one to five times per month, X2(16 df, 2571) = 158.77,p < .00005.About 1 in ual pa.rtners ever was 101-500 n = 1947)= 11.37, p = .79. 10 older men (10.2%)named a local (21.6%); 2.7% had had sex with 1 Sexualpractice.Data were collected newspaper, a rate similar to men in partner only;and between 10.2%and on a rangeof sexual practices.The per- other age groups. 15.7%rreportedhaving had sex with centages of older men who engaged the nunnber of partners indicated by in each sexual practice with regular Sexual Relations and Practices eah of the following ranges: 2-10, each and casual male partners are given Sexual relations. As shown in 11-20, 21-50, 51-100, 501-1000, or in "50 or over" columns of Tables 4 Table 3, about half of the older men > 1000. and 5, respectively. Most behaviours were practised more frequently with Table 3 regular than with casual partners. Sexual Relations at Time of Interview by Age (Percentages) Sensuous touching, mutual mastur- Under 25 25-29 30-39 40-49 50 or over bation, oral-genital sex (without Partnership ejaculation), and kissing were the n = 527 n = 512 n = 765 n = 516 n = 255 most common practices. About half None 11.4 9.0 10.1 4.3 7.1 Monogamous 24.3 17.6 17.4 14.7 21.6 the older men engaged in anal inter- Regular plus casual 15.0 20.1 20.7 24.0 16.5 course with their regular partners, Several regular 2.5 3.5 3.5 4.8 6.3 approximately one third with their Casual only 46.9 49.8 48.4 52.1 48.6 casual partners. x2(16 df, n = 2575) = 56.38, p < .00005 When there were significant age- This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions  Van de Ven, Rodden, Crawford,and Kippax 355 Table 4 Percentagesof Men in Each Age GroupWhoEngaged Occasionallyor Oftenin Sexual Practices with Regular Male Partners in the Six MonthsPrior to Interview Under 25 25-29 30-39 40-49 50 or over X2 P n = 234 n = 224 n = 327 n = 228 n = 116 Sensuous touching 98.3 98.2 99.1 98.3 99.1 1.361 .85 Deep/wet kissing 96.6 96.9 92.0 88.6 82.8 31.857 <.00005 Dry kissing 94.4 97.3 90.8 89.5 83.6 23.709 <.0005 Mutual masturbation 96.1 96.9 96.0 90.8 90.5 14.815 .005 Oral-genital no ejaculation Insertive (being sucked) 89.7 90.2 91.1 77.6 86.1 26.644 <.00005 Receptive (sucking) 95.7 94.6 92.9 85.1 87.8 23.640 <.0005 Oral-genital with ejaculation Insertive 45.9 45.1 44.9 37.7 44.8 4.216 .38 Receptive 42.3 39.3 40.0 38.2 37.1 1.255 .87 Finger fucking Insertive 58.5 75.9 74.8 68.4 53.4 34.245 <.00005 Receptive 65.0 73.7 71.7 58.3 46.6 35.703 <.00005 Anal intercoursea Insertive 56.8 61.2 54.7 57.9 52.6 3.252 .52 Receptive 62.4 55.8 53.7 50.9 53.4 7.097 .13 Anal with withdrawala Insertive 39.5 46.0 39.0 28.9 19.0 31.471 <.00005 Receptive 43.6 41.1 37.7 26.3 18.1 34.235 <.00005 Rimming Insertive 50.0 55.4 52.3 46.5 34.5 15.411 .004 Receptive 62.4 61.6 52.6 46.1 32.8 38.471 <.00005 Fisting Insertive 7.7 10.3 7.4 7.9 4.3 3.942 .41 Receptive 10.3 7.6 8.0 8.8 6.9 1.682 .79 aWithor without condoms Table 5 Percentagesof Men in Each Age GroupWhoEngaged Occasionallyor Often in Sexual Practices with Casual Male Partners in the Six MonthsPrior to Interview Under 25 25-29 30-39 40-49 50 or over xz p n = 364 n = 394 n = 560 n = 401 n = 171 Sensuous touching 96.4 96.2 96.4 95.5 95.3 0.867 .93 Deep/wet kissing 87.0 84.4 77.8 71.6 71.5 40.730 <.00005 Dry kissing 86.2 82.4 76.5 75.1 68.6 30.349 <.00005 Mutual masturbation 92.3 94.1 94.4 94.3 87.8 11.196 .02 Oral-genital no ejaculation Insertive (being sucked) 92.3 88.8 89.6 81.0 78.5 37.141 <.00005 Receptive (sucking) 90.9 92.6 89.4 82.3 75.6 47.556 <.00005 Oral-genital with ejaculation Insertive 46.0 38.5 40.4 36.2 51.7 16.818 .002 Receptive 30.7 24.7 23.3 21.2 25.6 10.161 .04 Finger fucking Insertive 61.6 63.3 60.6 58.9 48.8 11.362 .02 Receptive 65.7 62.2 61.1 55.1 48.8 18.782 <.001 Anal intercoursea Insertive 43.1 46.2 44.7 40.6 32.0 11.690 .02 Receptive 45.3 36.3 34.5 32.9 33.9 15.950 .003 Anal with withdrawala Insertive 32.7 33.8 24.9 20.9 14.5 37.826 <.00005 Receptive 31.3 30.6 22.0 19.2 12.9 37.785 <.00005 Rimming Insertive 40.7 40.4 37.6 27.9 26.2 26.016 <.00005 Receptive 56.4 54.1 52.3 40.4 36.0 37.621 <.00005 Fisting Insertive 8.3 10.5 10.9 8.0 3.5 10.684 .03 Receptive 8.8 8.7 7.9 5.2 2.3 11.594 .02 aWithor without condoms This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 356 Older Homosexually Active Men related differences for particular Table 6 sexual practices, the older men were Sexual Practice Scales by Age generally the least likely, or among Under 25 25-29 30-39 40-49 50 or over those least likely, to have engaged in Anala the practice. Several differences be- n = 529 n = 512 n = 767 n = 515 n = 256 tween the age groups emerged in sex- M 4.00 4.25 3.83 3.48 2.83 ual practices with regular partners SD 2.50 2.54 2.50 2.38 2.26 (see Table4). Oldermen were the least OraVTactileb likely to have engaged in insertive n = 529 n = 512 n = 767 n = 516 n = 256 and receptive anal intercourse with M 5.30 5.39 5.10 5.16 4.95 withdrawal. They were also the least SD 1.54 1.40 1.68 1.38 1.56 likely to have engaged in wet and dry aF(4, 2574) = 17.22, p < .00005 kissing, to have engaged in receptive bF(4, 2575) = 5.24, p < .001 rimming, and to have had insertive Table 7 and receptive finger fucking. Like their 40- to 49-year-oldcounterparts, Mode of Anal Intercourseby Age (Percentages) older men were less likely to have Partner and had insertive and receptive oral-gen- Mode Under 25 25-29 30-39 40-49 50 or over ital sex (without ejaculation) than With Regular Partnersa n = 190 n = 183 n = 252 n = 178 n = 84 men under the age of 40. Insertive only 12.6 18.0 18.3 27.0 19.0 Receptive only 17.4 10.4 13.9 16.9 21.4 By Mantel-Haenszeltest, there was Both 70.0 71.6 67.9 56.2 59.5 a decreasinglinear trend with age for most sexual practices with regular With Casual Partnersb n = 238 n = 248 n = 341 n = 241 n = 90 Insertive only 17.2 27.8 31.1 34.9 31.1 partners: wet (p < .00005) and dry Receptive only 18.5 10.1 14.4 19.5 25.6 kissing (p < .00005); mutual mastur- Both 64.3 62.1 54.5 45.6 43.3 bation (p < .005); insertive (p < .005) With Either Regular or and receptive (p < .001) oral-genital Casual Partnersc n = 373 n = 366 n = 517 n = 360 n = 156 sex without ejaculation;receptivefin- Insertive only 14.5 20.5 25.0 30.8 25.6 ger fucking (p < .001); receptive anal Receptive only 16.6 9.0 13.2 16.4 22.4 intercourse (p < .05); insertive (p < Both 68.9 70.5 61.9 52.8 51.9 .00005)and receptive(p < .00005)anal Note: Only includes men who engaged in anal int intercourse with withdrawal;and in- aX2(8 df, n = 887) = 20.96, p = .007 bX2(8 df, n = 1158) = 40.77, p < .00005 sertive (p < .01) and receptive (p < CX2(8df, n = 1772) = 55.15, p < .00005 .00005) rimming. Withcasual partners,the oldermen were less likely to have had insertive receptive (p < .00005) finger fucking; Anal Practice, there also was a sig- and receptive anal intercourse with insertive (p < .05) and receptive (p < nificant quadratic relationship (p < withdrawal, to have engaged in dry .001) anal intercourse;insertive (p < .0001) that indicated a peak range of kissing, to have had insertive and .00005)and receptive(p < .00005)anal anal practices for 25- to 29-year-old receptiveoral-genitalsex withoutejac- intercoursewith withdrawal;insertive men and fewer anal practices for ulation, and to have engaged in re- (p < .00005) and receptive(p < .00005) younger or older men. ceptive rimming and receptive finger rimming; and receptive fisting (p < Mode of anal intercourse. Of the fucking. Togetherwith the men aged .005). There were no increasinglinear older men, 156 (60.9%)had had anal 40-49 years, they were less likely than trends for any sexual practices with intercourse during the 6 months younger men to have engaged in wet either regular or casual partners. prior to interview: 84 (32.8%) with kissing and insertive rimming (see There was confirmation of these regular partners and 90 (35.2%)with Table 5). patternsof sexual experienceon scales casual partners. Of the men who en- By Mantel-Haenszeltest, there was that measured the range of specific gaged in anal intercourse, some age a decreasinglinear trend with age for sexual practices with male partners. differences were found for mode of most sexual practices with casual The older men had the lowest scores intercourse(see Table7). Withregular partners: wet (p < .00005) and dry on both the Anal Practice and Oral/ partners, there were no significant kissing (p < .00005); insertive (p < Tactile Practice scales (see Table 6). differences between the age groups. .00005) and receptive (p < .00005) By F-ratio for linear trend, there was With casual partners,oldermen were oral-genital sex without ejaculation; a decreasing trend with age for both the most likely to have had receptive receptive oral-genital sex with ejacu- Anal (p < .000005) and Oral/Tactile intercourse exclusively; those under lation (p < .05);insertive (p < .01) and Practices (p < .0005). In the case of 25 years were the least likely to have This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions  Van de Ven, Rodden, Crawford, and Kippax 357 had insertive intercourse exclusively; Table 8 those over 40 years were less likely Anal Intercourseand CondomUse with Regular and Casual Male Partners byAge (Percentages) than younger men to have had both Partner and Practice Under 25 25-29 30-39 40-49 50 or over insertive and receptive intercourse. Regular Partnersa n = 232 n = 223 n = 329 n = 229 n = 117 This pattern of men over 40 years No anal intercourse 18.1 17.9 23.4 22.3 28.2 being less likely to have had both in- Always condom 45.3 43.0 36.5 32.3 25.6 sertive and receptive intercourse was Sometimes unprotected 36.6 39.0 40.1 45.4 46.2 duplicated when intercourse with Only Those Who Had Anal Intercourse with Regular Partnersb both regular and casual partners was n = 190 n = 183 n = 252 n = 178 n = 84 taken into account. Always condom 55.3 52.5 47.6 41.6 35.7 Condom use. To answer Research Sometimes unprotected 44.7 47.5 52.4 58.4 64.3 Question 3, we examined condom use Casual Partnersc n = 370 n = 395 n = 578 n = 428 n = 189 with regular and casual partners. No anal intercourse 35.7 35.2 41.0 43.7 52.4 With regular partners, there were no Always condom 45.7 49.9 43.3 43.7 30.7 Sometimes unprotected 18.6 12.9 15.7 12.6 16.9 significant age group differences (after Bonferroni adjustment) in con- Only Those Who Had Anal Intercourse with Casual Partnersd dom use during anal intercourse (see n = 238 n = 248 n = 341 n = 241 n = 90 Table 8). Of those who had anal in- Always condom 71.0 79.4 73.3 77.6 64.4 Sometimes unprotected 29.0 20.6 26.7 22.4 35.6 tercourse with regular partners, more aX2(8df, n = 1130) = 20.47, p = .009 than 40% of the younger and older bx2(4 df, n = 887) = 13.51, p = .009 men sometimes did not use condoms. cX2(8 df, n = 1960) = 28.32, p < .0005 In the case of casual partners, there dx2(4 df, n = 1158) = 10.91, p = .03 were significant differences between the age groups. Older men were sig- Table 9 nificantly more likely than the HIV Status by Age (Percentages) younger men to exclude anal inter- Test Status Under 25 25-29 30-39 40-49 50 or over course from their sexual repertoire n = 529 n = 512 n = 767 n = 516 n = 256 with casual partners. However, for No test 25.7 22.9 21.9 27.9 36.3 younger and older men who had anal Negative 70.5 68.4 70.7 64.0 60.5 intercourse with casual partners, there Positive 3.8 8.8 7.4 8.1 3.1 were no significant differences (after X2(8 df, n = 2580) = 40.12, p < .00005 Bonferroni adjustment) between age Table 10 groups in condom use. Test status and contact with the Contact with the Epidemic Scale by Age epidemic. Older men were the least Under 25 25-29 30-39 40-49 50 or over likely to have been tested for HIV n = 529 n = 512 n = 767 n = 516 n = 256 antibodies, X2(4df, n = 2576) = 30.70, M 1.06 1.38 1.47 1.36 1.30 p < .00005 (answering Research Ques- SD 1.09 1.13 1.12 1.12 1.14 tion 2 in the negative). There were F(4, 2575) = 11.17, p < .00005 eight HIV-positive older men, a small- er proportion than for the age groups between 25 and 49 (see Table 9). sharply up to the 30-39 age group conducted on the Anal Practice and Contact with the epidemic (the sub- and thereafter declined slightly with Oral/Tactile Practice scales to inves- ject of Research Question 4) varied increasing age. tigate whether there were any age significantly with age; however, the group differences in range of sexual Multivariate Perspective older men generally had a similar practices after adjustment for other degree of contact to the other men Our findings that anal and oral/tac- variables. Age and region were treated over 25 years (see Table 10). The older tile practices declined with age were as factors alongside the following co- men's Contact with the Epidemic new. To answer our first research variates: social attachment, cultural Scale score of 1.30 indicated moder- question about older men's sexual involvement, occupation, living situ- ate involvement with people living practices, it was important to know ation, length of time at current ad- with HIV/AIDS. By F-ratio for trend, whether these differences could be dress, country of birth, marital status, there was a significant quadratic re- explained by the other differences that sexual identity, and antibody status. lationship between Contact with the were observed between younger and There was no significant age x region Epidemic and age (p < .000005), re- older men. Thus, a multivariate analy- interaction, F(16, 5018) = 0.82, p = .66. flecting the fact that contact increased sis of covariance (MANCOVA) was After adjustment for all covariates, This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 358 Older Homosexually Active Men including region, the analysis yielded were more likely to have children. partners in a six-month period as significant age-related differences in These findings are in accord with younger men. the scales of sexual practice, F(8, earlier evidence reported by Berger We turn now to our first research 5018) = 4.46, p < .00005 (Pillai's cri- and Kelly (1986) and Lee (1987). question that asked if older homo- terion). Confirming our univariate Older men differed somewhat from sexually active men engage in the results, differences on the Anal Prac- younger men in terms of self-defined same sexual practices as their younger tices Scale were significant, F(4, sexual identity, with older men more counterparts. Some older men engaged 2509) = 6.68, p < .00005. However, likely to describe themselves as homo- in an extensive range of sexual prac- there were no significant differences sexual rather than gay. Moreover, tices with regular and casual male on the Oral/Tactile Practices Scale and largely consonant with Hypoth- partners, but as a group older men after adjustment for the covariates, esis 3 and the findings of Bennett tended to engage in the practices less F(4, 2509) = 1.43, p = .22. and Thompson (1980), older men were frequently than their younger coun- less likely to have disclosed their terparts. For most sexual practices, a Discussion sexual orientation to others. It would fairly consistent pattern indicated a Using data from an Australian appear that the more open and sup- decreasing linear trend in frequency nationwide telephone survey (Project portive post-Stonewall climate facili- with age. The finding that this trend Male-Call: Kippax et al., 1994), we set tates younger gay men's divulging applied to most practices is important. out to develop a profile of older (over their sexual identity to a broad range It suggests a general downward trend 49 years) homosexually active men. of family members, friends, work col- for a whole range of sexual practices Our focus was on the neglected area leagues, and acquaintances, particu- rather than practice substitution with of such older men's sexual practices larly after the age of 25. This accords age. and features that differentiate older with the generational changes exem- Older men as a group scored sig- men from their younger counterparts. plified in various accounts by gay nificantly lower than younger men in Based on limited although fairly men in Wotherspoon's (1986) collec- terms of extent of anal and oral/tac- consistent past findings (Bennett & tion of autobiographies. tile sexual repertoire. Differences be- Thompson, 1980; Berger & Kelly, 1986; Contrary to Hypothesis 4 and some tween the age groups in range of oral/ Weinberg & Williams, 1974), we gen- previous findings (Bennett & Thomp- tactile practices were not significant erated five hypotheses about older son, 1980), older men's attachment after controlling for demographic men's living arrangements, gay com- to gay community (measured in terms and milieu factors. Separate analyses munity involvement, and sexual prac- of social attachment, cultural involve- revealed that differences in oral/tac- tices. In the absence of previous data, ment, or sexual involvement) was less tile practices were largely accounted we also posed three research ques- than for their younger counterparts. for by levels of gay community in- tions about older men's sexual prac- These results, though, are partly in volvement, with men not attached to tices and their testing, responses, line with Weinberg and Williams' gay community less likely to have an and contact in relation to HIV/AIDS. (1974) finding that older homosexual extensive oral/tactile repertoire. How- In contrast with the younger men, men were lower in overall social in- ever, differences in anal practices the older men in the sample were volvement than younger homosexual were actually age based and could not more ethnically homogeneous and, men. Nonetheless, the data indicated be attributed to other factors such as as is to be expected, more likely to be that most older men maintained mod- region of residence or degree of attach- out of the workforce. Comparatively, erately strong social links with gay ment to gay community. This raises older men were more stable in terms community, a finding that augurs well the interesting possibility that the anal of residence. They tended not to live for their psychosocial adaptation in practice difference may be genera- in predominantly gay areas, whereas older age (Berger & Kelly, 1986). tional, a product of the cultural climate they were overrepresented in rural Consistent with Hypothesis 5 and in which the men were socialised. areas. Consistent with Hypothesis 1 previous evidence (Berger, 1980), older Proportionately, older men engaged and earlier Australian and interna- men had as many different male part- in more anal intercourse with regu- tional data (Bennett & Thompson, ners in a six-month period as younger lar than casual partners. However, 1980; Weinberg & Williams, 1974), men. For approximately half of the regardless of partner type, older men older men were more likely to live older group, these partnerships were who routinely engaged in anal inter- alone than with a male sexual part- casual encounters exclusively. About course were as likely to be the in- ner, although more older than younger one in five older men was in a monog- sertive as the receptive partner. A men lived with a female sexual part- amous regular relationship, and as majority of the older men who en- ner. In line with Hypothesis 2, older many older men were in regular gaged in anal intercourse were versa- men were more likely to be or have relationships that did not preclude tile with regular partners, switching been married. Correspondingly, and liaisons with other male sexual part- between insertive and receptive roles. perhaps because of a time factor, they ners. Older men had as many female This corroborates earlier evidence This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions  Van de Ven, Rodden, Crawford,and Kippax 359 documented by Berger and Kelly encounters as inherently less safe. ports, and have not accommodated (1986). Only about one in five older An alternate explanation is that a well to the aging process, and whose men who engaged in anal intercourse smaller proportionof the older men, needs should not be diminished. in a regular relationship was exclu- belongingto the pre-AIDSgeneration, Despite the sampling limitations, sively insertive and a similar propor- were comfortablewith and adept at the Project Male-Call data and the tion exclusively receptive. In contrast using condoms.If either of these latter findings presented here paint an in- with regular partners, fewer older possibilities is correct, it would indi- teresting picture of older Australian men were versatile, and more of them cate that safe sex campaigns, which homosexually active men. Our find- were exclusivelyinsertive or receptive are often targeted explicitly or im- ings are generally in agreement with with their casual partners. plicitly-the latter by virtue of their the limitedpreviousresearchevidence When anal intercourse with both media and youthful iconography-at concerningolderhomosexualmen. As types of partners (either regular or younger gay men, are not having as a group, older homosexually active casual) was taken into account, older much impact among older homosexu- men are sexually motivated, have men (includingthose in the 40-49 age ally active men. Furtherworkis need- opportunitiesto satisfy sexual desires, group) were significantly less likely ed to unpackthe reasons behind older and are involved in gay community. than younger men to have engaged men'sless frequentengagementin anal They are committed to a homosexual in both insertive and receptive inter- intercourse with casual partners. orientation, although they are more course. Engagement in both modes of In terms of having known people likely than post-Stonewall genera- anal intercourse was much more who died followingAIDS and having tions to have accommodatedto pre- usual for men under 30 years. known or having cared for seroposi- vailing heterosexist ideologythrough Turningto our three researchques- tive people, older men had similar limited disclosure of their sexual tions connected with HIV/AIDS, are levels of contact with the epidemic as identity and through straight mar- older homosexual men as likely as the men in the 25-29 and 40-49 age riage and family life. their younger counterparts to have brackets.The youngest men have had This analysis points to some spe- been tested for HIV,as responsive to least and the 30- to 39-year-oldmen cific issues for safe sex campaigns safe sex campaigns (as assessed by greatest contact with the epidemic. aimed at older homosexually active condomuse), and as close to the HIV An important consideration is to men. In the HIV/AIDSera, oldermen epidemic? Older men were signifi- be heeded in the interpretation and are more likely than their younger cantly less likely to have been tested generalisation of our findings: The counterparts to have no anal inter- for HIV antibodies.Regardless,of the sample was not and could not be course rather than condom-protected men who engaged in anal intercourse, drawn randomly. The Project Male- anal intercourse with casual part- older men were no less likely than Call men were recruited through di- ners. Such a strategy (be it deliberate younger men always to use condoms verse sources and included men from or otherwise) is quite safe so long as during anal intercourse with their widely different regional, social, and it is sustainable.Almost half the older regularor casual partners.In the case ethnic backgrounds.Nevertheless,the men sometimes do not use condoms of casual partnerships, there was a representativeness of the sample to for anal intercoursewith regularmale conspicuous difference between the the total population of homosexually partners. This may or may not be a age groups, with more than half the active Australian men is not known problem, depending on whether the older men not engaging in anal inter- because of the lack of a sampling men know their HIV status and have course with their casual male part- frame for this group.There is a possi- negotiateda reliable agreementabout ners. This may be due to a number of bility that the sample was more rep- safe sex outside their partnership. In factors, such as the availability of resentative of the better informed the absenceof HIV antibodytesting- condoms(if condomsare less available and motivated segment of the homo- a more common characteristic of the to oldermen in rural areas), economic sexual population and more repre- older men-such sexual negotiation constraints (if older men are paying sentative of men affiliated with gay to exclude HIV from the relationship for sex more often than younger men community.Certainly,comparedwith is meaningless. It is therefore impor- -anal being more expensive than Australian Census data, the sample tant to extend existing campaigns to oral sex), and sites of sexual activity underrepresentedoldermen and over- improve older men's safe sex under- (if older men are having more sex in represented tertiary educated and standings and to encourageHIV anti- publicplaces,makinganal intercourse professionally employed men. Repli- body testing. Given that older men less possible). cation of our findings from a house- were more likely to have been recruit- On the other hand, this finding hold-based sampling frame would be ed through the mainstream rather may point to a possible age difference useful. Such replicationis important, than gay press, those running gay in response to safe sex campaigns. as there may be older homosexually men's education programs would be One interpretation is that older men active men who are socially isolated, advised to use a variety of media, erroneously perceived casual sexual have few community or cultural sup- metropolitan and rural, and not just This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions 360 Older Homosexually Active Men gay avenues, to ensure that all con- Friend, R. A. (1987). The individual and social Lee, J. A. (1987). What can homosexualaging stituents are reached. psychology of aging: Clinical implications studies contribute to theories of aging? forlesbians and gay men. Journal of Homo- Journal of Homosexuality,13, 43-71. References sexuality,14, 307-331. McDougall, G. J. (1993). Therapeutic issues Friend, R. A. (1989). Older lesbian and gay with gay and lesbian elders. Clinical Bennett, K. C., & Thompson,N. L. (1980). So- people: Responding to homophobia.Mar- Gerontologist,14(1), 45-57. cial and psychological functioning of the riage and Family Review, 14, 241-263. Weinberg, M. S., & Williams, C. J. (1974). ageing male homosexual. British Journal Kippax, S., Connell, R. W., Dowsett, G. W., & Male homosexuals: Their problems and of Psychiatry,137, 361-370. Crawford,J. (1993). Sustaining safe sex: adaptations. New York:OxfordUniversity Berger,R. M. (1980). Psychologicaladaptation Gay communities respond to AIDS. Lon- Press. of the older homosexual male. Journal of don:Falmer Press. Wotherspoon,G. (Ed.). (1986).Being different: Homosexuality,5, 161-175. Kippax, S., Crawford,J., Rodden, P., & Ben- Nine gay men remember.Sydney: Hale & Berger, R. M., & Kelly, J. J. (1986). Working ton, K. (1994).Reporton ProjectMale-Call: Iremonger. with homosexuals of the older population. National telephonesurveyof men who have Social Casework:The Journal of Contem- sex with men. Canberra:Australian Gov- Manuscript accepted March 5, 1997 porary Social Work,67, 203-210. ernment Publishing Service. This content downloaded from 149.171.67.164 on Fri, 25 Sep 2015 03:27:17 UTC All use subject to JSTOR Terms and Conditions