A comparative demographic and sexual profile of older homosexually active men.

Over the past two decades or so, much has been written about the sexual identities and behaviours of homosexually active men. The literature has flourished especially since the advent of HIV/AIDS and the attendant concerns to understand the social, psychological, and behavioural dimensions of gay men's lives. Various cohorts such as young gay men, gay men from language backgrounds other than English, and men who have sex with both men and women have been the focus of numerous and detailed research studies. From a sexual practices viewpoint, one group, however, has been almost entirely neglected. There is no comprehensive account of the social and sexual lives of older homosexually active men, a situation that we sought to redress by comparisons between younger and older homosexually active men. Specifically, we analysed data from an Australian nationwide sample to compare younger and older men's sexual identities, attachments to gay community, sexual relations and practices, HIV testing, and contact with the epidemic. (Throughout this article, we refer to gay community rather than the gay community to reflect a diversity of communities and the fact that gay community is not all of a piece.) We were interested to see if older men differed from younger men, particularly as age differences might signal the importance of cultural norms and social processes in shaping who homosexual men are and what they do.

As other researchers have pointed out (Bennett & Thompson, 1980; Berger, 1980; Berger & Kelly, 1986; Friend, 1987), the stereotype of the older homosexual man has been one of disengagement from homosexual community, loneliness, rejection, depression, and unhappiness. Several empirical studies have contradicted this stereotypical picture. Weinberg and Williams (1974), who explored homosexual adaptations in the United States, the Netherlands, and Denmark, reported that older (over 45 years) homosexual men were higher on some measures of psychological well-being than were younger homosexual men. Their data did not support the image of the older homosexual man as lacking in self acceptance or being anxious, depressed, lonely, and unhappy. However, they found that older homosexual men attended gay venues less often, were more likely to live alone, reported homosexual contacts less frequently, and, in the United States, were lower in overall social involvement in homosexual community than were younger homosexual men.

In a later study involving Australian men, Bennett and Thompson (1980) found no evidence that older (over 45 years) homosexual men disengaged, or were forced to disengage, from the homosocial aspects of gay community. In comparison with their younger counterparts, older homosexual men were as involved in the homosexual world, reported a similar incidence of visiting gay venues, perceived their popularity among other homosexuals to be as high, and were as satisfied with their sexual orientation. Older men were less likely than those who were 26 to 35 years old to share accommodation with lovers and more likely to live alone. Reflecting a possible generational difference, the older men were significantly more concerned than the younger men about the exposure of their homosexuality.

Berger (1980) studied 112 homosexual men between the ages of 41 and 77. Whereas 38% of the men lived alone, the majority either lived with a sexual partner (43%) or with friends or members of their family of birth (19%). Far from being isolated and rejected by younger men, most men had many friends, were sexually active, and sustained close relationships with friends and sexual partners.

Through a comprehensive review of the literature, Berger and Kelly (1986) discredited some popular misconceptions and provided a realistic picture of aspects of the lives of older gay men. They found that few older gay men were loners. In fact, older gay men had more options in relationships than did their heterosexual counterparts and revealed a variety of living arrangements accordingly. Older homosexual men in a relationship were not more likely to play a role along the lines of the active/passive dichotomy assumed to be common in pre-liberation times. Role playing in relationships was highly atypical among younger and older homosexual men alike. As a result of intense social pressures in the past, many older homosexual men had been or currently were married. Continuing sexual interest and activity were the norm for older homosexual men. Moreover, integration into a local homosexual community was associated with psychosocial adaptation for older gay men.

It has been suggested that coming to terms with homosexuality may facilitate adjusting to the aging process in a number of ways (Friend, 1987, 1989; Lee, 1987; McDougall, 1993). Successful management of the process of coming out in a heterosexist world possibly proVides,the individual with coping mechanisms that may generalise to other crises in later life. Gender-role flexibility may allow older homosexual men to develop ways of taking care of themselves that feel comfortable and appropriate. Homosexual men often exchange the family of birth supports that were lost when they came out with a broader surrogate family and community network. Other strengths may include learning to fend for oneself from an early age, increased personal autonomy, and learning to live with a stigmatised identity throughout life. These competencies and reconstructions tend to be associated with healthy psychological adaptation by older gay men (Friend, 1989; McDougall, 1993).

The current generation of older homosexually active men has lived through remarkable changes in perceptions and attitudes toward homosexuality, not least the breaking of the "wall of silence" that once precluded any public discussion of homosexuality (Wotherspoon, 1986). The past three decades also have witnessed the sexual revolution of the 1960s and 1970s, the establishment of homophile organisations concurrent with the Stonewall Riots in New York in 1969, the dawn of gay political activism, alterations to the laws that restricted the emotional and sexual expression of male homosexuality, and the development of gay community. As several accounts in Wotherspoon's (1986) collection of autobiographical essays exemplify, these historical events have produced an apparent generational shift for gay men that has made it somewhat easier for those growing up with homosexual desires to express and be more open about their sexual orientation.

Whereas the literature provides useful insights into some social and psychological aspects of older homosexually active men's lives, there is a conspicuous lack of detail about their sexual relations and practices. In the era of HIV and AIDS, this is a serious omission. To remedy this deficiency, we analysed data collected in 1992 as part of Project Male-Call, an Australian national telephone survey of men who have sex with men (Kippax, Crawford, Rodden, & Benton, 1994). A 42-page interview schedule was used. It covered a range of issues related to sexual behaviour (nature of sexual relationships, sexual history, sexual practices with men and with women, safe sex strategies, attachment to gay community, degree of contact with the epidemic, HIV test result) and a number of demographic variables. (Some countries use safe, and others use safer, as the term to describe risk-reduction strategies with reference to HIV transmission, and the same arguments are used for and against either term. As used in this article, it is Australian policy to refer to safe sex.) Whereas our focus was on older (over 49 years) homosexually active men, we realise that people don't change suddenly at age 50. For this reason, we tested for a number of developmental changes over five age groups: less than 25, 25-29, 30-39, 40-49, and over 49 years. However, for expositional purposes, we describe most results by talking about older men versus younger men.

Based on limited although fairly consistent previous findings, we tested a number of specific hypotheses about older (over 49 years) men's, compared with their younger counterparts', sexual relations and practices: Older men are more likely to live alone rather than with a sexual partner (Hypothesis 1). Older men are more likely to have been or to be married (Hypothesis 2). Older men are less likely to have disclosed their homosexual orientation to others (Hypothesis 3). Older men are as attached to gay community as younger men (Hypothesis 4). Older men are likely to have had the same number of male partners in the past six months as younger men (Hypothesis 5).

Without a body of evidence on older homosexually active men's sexual practices, we posed the research question: Do older homosexually active men engage in the same sexual practices as their younger counterparts (Research Question 1)? In the absence of previous data pertaining specifically to older homosexually active men's behaviours in relation to HIV/AIDS, we also posed three additional research questions. Are older homosexually active men, compared with their younger counterparts, as likely to have been tested for HIV antibodies (Research Question 2)? Are older men as responsive to safe sex campaigns, as measured by condom use with regular and casual male partners (Research Question 3)? Are older men as likely to have had similar levels of contact with the epidemic (Research Question 4)?

Method

Participants

The analysis was based on telephone interviews with 2,583 homosexually active men, of whom 2,580 provided age data and thus could be included in the analysis. Men were, included in the survey according to the criterion that they had had sex with at least one other man during the five years prior to interview.

There is no way of enumerating the population of homosexually active men. Participants were necessarily volunteers. Recruitment sources included sections of the organised gay community (radio, venues, gyms, businesses, publications); places of sexual contact within, outside, and marginal to organised gay communities (gay brothels, sex shops, beats, saunas); health centres frequented by gay men; and pornography outlets. Strategies for contacting potential respondents included standard advertising that emphasised the importance of research information, advertising with some sexual titillation, cards and fliers, stickers and posters, notices in Personals columns, and articles and interviews in the media.

Measures

The interview schedule (available from the first author) consisted of 42 pages of questions based on those used successfully in the Social Aspects of the Prevention of AIDS study and the follow-up telephone survey, Sustaining Safe Sex (Kippax, Connell, Dowsett, & Crawford, 1993). The schedule included a number of demographic variables (age category as employed in the Australian Census, place of birth, educational level, occupation, income bracket, place of residence, heterosexual marriage) and items used to construct milieu or context variables (sexual identity and disclosure; gay community attachment as measured by social attachment, cultural involvement, and sexual involvement scales). These two sets of variables may be distinguished from the other variables of interest (outcome measures), which included sexual practice, number of sexual partners, frequency of sex, condom use, HIV antibody status, and degree of contact with the epidemic. Measures and scales that are not self-explanatory are described next.

For sexual identity, men were asked if they thought of themselves as any of the following, in this order: heterosexual, straight, bisexual, gay, homosexual, camp, queer, other (participant specified), and don't know/unsure. It was common practice for the men to nominate their sexual identity before the interviewer had a chance to read through the list, indicating generally clear senses of sexual identity.

Men were asked about their regular and casual male partners. The use of the term casual partnership did not imply any judgment about the degree of seriousness of the sexual action between a man and his casual partner, nor of the moral worth of casual as opposed to regular relationships. It was meant merely to denote an occasional sexual partner as distinct from a regular or steady partner in a committed ongoing relationship.

Gay identity disclosure was measured with a seven-item scale constructed from questions about people the respondent had told about his homosexual practice (mother, father, other relatives, straight friends, work-mates, neighbours, anyone else). The range of possible scores was 0 to 7, with a higher score indicating greater disclosure. Cronbach's alpha was .82. The mean score for the total sample was 2.85, with a standard deviation of 2.25.

Social attachment was measured with a scale that included eight items about the amount of free time spent with gay men, number of gay friends, and where the respondent went with gay friends (gay bars, discos, parties, pool/beach, meetings/organisations). The range of possible scores was 0 to 13, with a higher score indicating a greater degree of social attachment to gay community. Cronbach's alpha was .86. The mean score for the total sample was 7.32, with a standard deviation of 3.63.

Cultural involvement was measured with a scale that contained three items: membership in gay organisations, perception of belonging to gay community, and readership of gay press. The range of possible scores was 0 to 3, with a higher score indicating a greater cultural involvement in gay community. Cronbach's alpha was .62. The mean score for the total sample was 1.86, with a standard deviation of 1.02.

Sexual involvement measured degree of immersion in the sexual elements of gay community with a scale that contained 16 items: 14 about where respondents went to look for male sexual partners (e.g., bars, saunas, cruising areas, sex clubs), 1 about the number of male sexual partners in the past 6 months, and 1 concerning frequency of sex with male casual partners in the previous month. The range of possible scores was 14 to 35, with a higher score indicating greater sexual involvement in gay community--looking for sexual partners in a greater range of sexual sites, having a greater number of male partners, and having more male partners per month. Cronbach's alpha was .80. The mean score for the total sample was 20.14, with a standard deviation of 3.91.

Contact with the epidemic was measured with a scale that contained three items: knowing a person who is seropositive or living with AIDS, knowing a person who has died following AIDS, and having been involved in caring for someone with AIDS. The range of possible scores was 0 to 3, with a higher score indicating greater contact. Cronbach's alpha was .68. The mean score for the total sample was 1.33, with a standard deviation of 1.12.

A number of sexual practice indicators and measures was employed. Data were collected on the number of male and female partners, the nature' of men's sexual relationships (regular or casual), condom use, and, concerning male partners, the frequency of 10 sexual behaviours (sensuous touching, kissing, masturbation, oral-genital sex with and without semen exchange, finger flicking, anal intercourse with and without ejaculation, rimming, and fisting) with regular and casual partners. Sexual practice with men was also described with reference to Anal Practice and Oral/ Tactile Practice Scales.

The Anal Practice Scale contained eight items about insertive and receptive anal intercourse with and without ejaculation, rimming, and fingering. The range of possible scores was 0 to 8, with a higher score indicating engagement in a larger number of anal practices with male partners. Cronbach's alpha was .80. The overall mean score was 3.78, with a standard deviation of 2.49.

The Oral/Tactile Practice Scale contained six items about wet and dry kissing, insertive and receptive oral-genital sex, mutual masturbation, and sensuous touching. The range of possible scores was 0 to 6, with a higher score indicating engagement in a larger number of oral/tactile practices with male partners. Cronbach's alpha was .85. The overall mean score was 5.20, with a standard deviation of 1.53.

Procedure

Project Male-Call was a national telephone survey of gay and homosexually active men in Australia. Recruitment, including advertising, took place in all states and territories, all capital cities, and urban and rural areas. National, regional, and local recruitment strategies were I used. The survey was conducted so as not to coincide with the major holiday periods of Easter and Christmas.

To safeguard respondents' anonymity, attract a wide cross-section of homosexually active men, generate an atmosphere in which the men could speak openly and honestly about their sexual practices and life situations, and minimise costs, telephone rather than face-to-face interviews were used. For May and June 1992, eight 008 (free of charge to caller) telephone lines were connected at Macquarie University. Trained interviewers informed callers that the survey would take around half to three quarters of an hour and asked the caller if he had that time available at present. Most callers were interviewed immediately. Others chose to call back.

Most interviewers were men. Some female interviewers and one transsexual interviewer were also employed. Some participants chose to be interviewed by a female interviewer, others insisted on a male interviewer, but most did not express a preference. Male interviewers did not raise the issue of the sex of the interviewer, but female interviewers always asked whether the caller would prefer a male interviewer.

Results

The men were categorised into one of five age groups: less than 25, 25-29, 30-39, 40-49, and over 49 years. (Equal intervals of 10 years produced highly unequal ns.) Univariate differences between the men in the different age groups were investigated by Chi-square tests of association for categorical variables and by analyses of variance (ANOVAs) for numerical variables. In accord with the Bonferroni principle, alpha was set at .001 for these comparisons because of the large number of statistical tests applied to the data. Where feasible, linear trends across the age groups were tested by the Mantel-Haenszel statistic (categorical data) or F-ratio for linear trend in ANOVA (numerical data). A multivariate analysis of covariance (MANCOVA) was also conducted to see whether there were any age group differences on the sexual practices with men scales (Anal Practice and Oral/Tactile Practice) after controlling for a number of demographic and milieu variables.

Demographics

There were 256 older (over 49 years) men in the Male-Call sample, representing just under 10% of the sample of 2,583 men. In each of the other age bands of less than 25, 25-29, 30-39, and 40-49 years of age, there were 529, 512, 767, and 516 men, respectively. Compared with Australian Census data, the sample underrepresented men over 49 years of age (Census = 30.3%). Moreover, there was an overrepresentation of tertiary educated and professional men in the whole sample compared with Australian averages (Kippax et al., 1994).

More than three quarters of the older men (fewer than the younger men) had been born in Australia or New Zealand (78.1%); a proportionately high number in the United Kingdom or elsewhere in Europe (19.5%); and relatively few in other countries (in Asia, Africa, Oceania, or North, Central, or South America; 2.3%). This represented a significant difference between the age groups in country of birth, [chi square] (16 df, n = 2577) = 76.27, p [is less than] .00005.

Most older men were in paid employment, either full time (37.5%), part time (8.2%), or self-employed (16.4%), although a considerable number (29.7%)--more than for any age other group--were receiving social security payments or were out of the workforce, [chi square] (24 df, n = 2577) = 740.80, p [is less than] .00005. Older men were about as likely to have had no occupation (33.3%; a similar proportion to those under 25 years) as they were to have been employed in managerial or professional occupations (30.6%; less than men 30-49 years); 9.5% of them, less than for the other age groups, were employed in paraprofessional or clerical occupations, [chi square] (16 df, n = 2567) = 272.14, p [is less than] .00005. Their representation in the other occupational groups of trade/manual and sales/service was similar to that of their younger counterparts.

The distribution of older men's incomes was more even than for any other age group, [chi square] (12 df, n = 2515) = 369.09, p [is less than] .00005, with approximately a quarter of them falling into each of the four income categories 25.3% (a higher proportion than all but the under 25 years age group) earned less than $15,000 per year, and another 25.3% earned more than $40,000 (with 23 of these men having earned more than $60,000); 21.2% earned $15,001--$26,000, and 28.2% earned $26,001--$40,000. (At the time, US$1 was approximately equivalent to Aust$1.30.) About a third (32.8%) of the older men, more than for the younger men, had education up to Year 10 only; about a quarter (23.4%) had completed the equivalent of Year 11 or Year 12; more than a quarter (27.0%) had higher-education; and the remainder (16.8%) had some other post-secondary qualifications, [chi square] (12 df, n 2575) = 104.49, p [is less than] .00005.

More than half the older men (52.7%), a greater proportion than for any other age group, lived alone. Other older men, more so than those under 30 years, lived with a female sexual partner (19.9%) or male sexual partner (13.7%). Few in this age group lived with platonic friends, either gay or straight, or natal family members, [chi square] (28 df, n = 2578) = 459.10, p [is less than] .00005. By Mantel-Haenszel test, there was an increasing trend for older men to live alone (p [is less than] .00005), and corresponding decreasing trends for men to live with natal family members (p [is less than] .00005) or friends (p [is less than] .00005), confirming Hypothesis 1.

Relatively few older men lived in predominantly gay areas (12.4%), whereas there was an overrepresentation of older men in rural areas (29.2%), [chi square] (8 df, n = 2548) = 51.31, p [is less than] .00005. By Mantel-Haenszel test, a significant trend indicated that the older the men were, the more likely they were to live in rural areas (p [is less than] .00005).

Older men were more likely than their younger counterparts to be or have been married (62.9%), [chi square] (4 df, n = 2572) = 487.96, p [is less than] .00005 (confirming Hypothesis 2). Correspondingly, the older respondents were more likely to have children (56.4%), [chi square] (4 df, n = 2544) = 454.68, p [is less than] .00005. Older men were more likely to have lived at their present address for more than five years (67.5%), [chi square] (16 df, n = 2569) = 389.05, p [is less than] .00005.

Milieu

Sexual identity and disclosure. When asked to describe their sexual identity, 60.9% of the older men identified as gay or homosexual and 28.5% as bisexual, proportions similar to other men over 30 years. Younger men, particularly those under 30 years, were more likely to embrace the term gay rather than homosexual, [chi square] (24 df, n = 2577) = 72.28, p [is less than] .00005. Only 3.9% of the older men identified as heterosexual, a similar proportion to the other age groups. The older men were by far the least likely to have disclosed their homosexual orientation to others (see Table 1). By F-ratio for linear trend, there was a significant decreasing linear trend for Gay Identity Disclosure (p [is less than] .000005). Moreover, there was a significant quadratic relationship between Gay Identity Disclosure and age (p [is less than] .000005), resulting from 25- to 29-year-old men having disclosed their identity more often than younger or older men (partly supporting Hypothesis 3).

Table 1
Gay Identity Disclosure Scale by Age

        Under 25     25-29        30-39     40-49    50 or over
         n = 492   n = 479      n = 714   n = 472       n = 222
M         2.99       3.25          2.97      2.57          1.88
SD        2.23       2.21          2.29      2.24          1.90

F(4, 2374) = 17.56, p < .00005

Gay community attachment. Attachment to gay community, whether measured in terms of social attachment, cultural involvement, or sexual involvement, was less for older men than for younger men (see Table 2). By F-ratio for linear trend, there were significant decreasing linear trends, with age, in social (p [is less than] .000005), cultural (p [is less than] .0001), and sexual involvement (p [is less than] .005) in gay community. For sexual involvement alone, there was a significant quadratic relationship between involvement and age (p [is less than] .000005), indicating peak sexual involvement in gay community for those in the 25-29 age group and less involvement for those in younger or older groups.

Table 2
Gay Community attachment Scales by Age

                         Under 25           25-29     30-39

Social attachment(a)      n = 529         n = 512   n = 764
M                            8.11            8.13      7.22
SD                           3.49            3.48      3.67

Cultural involvement(b)   n = 529         n = 512   n = 766
M                            1.91            1.98      1.84
SD                           0.98            1.04      1.02

Sexual involvement(c)     n = 510         n = 501   n = 746
M                           19.84           20.56     20.40
SD                           3.93            3.95      4.08

                            40-49    50 or over

Social attachment(a)      n = 516       n = 256
M                            6.61          5.79
SD                           3.64          3.25

Cultural involvement(b)   n = 516       n = 256
M                            1.80          1.67
SD                           1.04          0.96

Sexual involvement(c)     n = 503       n = 245
M                           20.16          19.0
SD                           3.66          3.51

(a) F(4, 2572) = 30.44, p < .00005

(b) F(4 2574) = 4.99, p < .001

(c) F(4 2500) = 7.65, p < .00005

When the scores on the social attachment scale were used to divide the sample into two groups, gay community attached (GCA; social attachment [is greater than] 4) versus non-gay community attached (NGCA; social attachment [is less than or equal to] 4), 62.9% of the older men (significantly less than for younger groups) were classified as GCA, and 37.1% were NGCA, [chi square] (4 df, n = 2580) = 61.37, p [is less than] .00005 (contrary to Hypothesis 4).

Older men were the least likely to have named the gay press (11.8%) and cards/posters (1.2%) as the source of information about the current survey. Conversely, they were the most likely to have named mainstream papers (11.4%) and mail order video and other catalogues (36.2%), [chi square] (44 df, n = 2571) = 158.77, p [is less than] .00005. About 1 in 10 older men (10.2%) named a local newspaper, a rate similar to men in other age groups.

Sexual Relations and Practices

Sexual relations. As shown in Table 3, about half of the older men had casual sex only. They, as well as men under 25 years, were more likely than the other age groups to be in a monogamous relationship and less likely to be in a regular plus casual relationship.

Table 3
Sexual Relations at Time of Interview by Age (Percentages)

Partnership           Under 25    25-29      30-39
                      n = 527    n = 512    n = 765
None                    11.4        9.0       10.1
Monogamous              24.3       17.6       17.4
Regular plus casual     15.0       20.1       20.7
Several regular          2.5        3.5        3.5
Casual only             46.9       49.8       48.4

Partnership              40-49      50 or over
                        n = 516     n = 255
None                      4.3        7.1
Monogamous               14.7       21.6
Regular plus casual      24.0       16.5
Several regular           4.8        6.3
Casual only              52.1       48.6

[chi square] (16 df, n = 2575) = 56.38, p <.00005

Numbers of partners/frequency of sex. Not unexpectedly, older men had more male sexual partners in their lifetime than younger men, [chi square] (32 df, n = 2573) = 251.09, p [is less than] .00005, but there were no significant age differences in the number of male sexual partners in the six months prior to interview, [chi square] (20 df, n = 2578) = 28.39, p = .10 (confirming Hypothesis 5). Almost three quarters of the older men had either 1 (28.5%) or between 2-10 (44.9%) partners during the preceding 6 months. For the older men, the modal range for number of male sexual partners ever was 101-500 (21.6%); 2.7% had had sex with 1 partner only; and between 10.2% and 15.7% reported having had sex with the number of partners indicated by each of the following ranges: 2-10, 11-20, 21-50, 51-100, 501-1000, or [is greater than] 1000.

In terms of female partners, 42 of the 256 older men (16.4%) had never had sex with a woman, whereas 50 men (19.5%) had had sex with a sole female partner. A further 97 men (37.9%) had had between 2 and 10 female partners ever. The older men (including those 40-49 years of age) were likely to have had more female partners in their lifetime than their younger counterparts, particularly so when compared with those under 30 years, [chi square] (24 df, n = 2579) = 133.42, p [is less than] .0005. However, the older men were no more or less likely than the men in the younger age groups to have had sex with a woman in the six months prior to interview: 29.0% of the entire sample, and 28.5% of the older men, had, [chi square] (16 df, n = 2576) = 23.83, p = .09.

Compared with men under 30 years, older men were more likely to have had sex with their regular partner 1-5 times per month (39.3% of those with regular partners) and less likely to have had sex more than 15 times (19.7%), [chi square] (16 df, n = 1157) = 52.71, p [is less than] .00005. Nevertheless, older men had sex as frequently as those aged 40-49 years, with 26.5% having had sex 6-15 times per month. There were no age differences in the frequency of sex with casual partners, and almost half (48.9%) of the older men had sex with a casual partner one to five times per month, [chi square] (16 df, n = 1947) = 11.37, p = .79.

Sexual practice. Data were collected on a range of sexual practices. The percentages of older men who engaged in each sexual practice with regular and casual male partners are given in "50 or over" columns of Tables 4 and 5, respectively. Most behaviours were practised more frequently with regular than with casual partners. Sensuous touching, mutual masturbation, oral-genital sex (without ejaculation), and kissing were the most common practices. About half the older men engaged in anal intercourse with their regular partners, approximately one third with their casual partners.

Table 4 Percentages of Men in Each Age Group Who Engaged Occasionally or Often in Sexual Practices with Regular Male Partners in the Six Months Prior to Interview

                           Under 25    25-29     30-39     40-49
                           n = 234   n = 224   n = 327   n = 228

Sensuous touching             98.3    98.2      99.1       98.3
Deep/wet kissing              96.6    96.9      92.0       88.6
Dry kissing                   94.4    97.3      90.8       89.5
Mutual masturbation           96.1    96.9      96.0       90.8
Oral-genital no
  ejaculation
  Insertive (being sucked)    89.7    90.2      91.1       77.6
  Receptive (sucking)         95.7    94.6      92.9       85.1
Oral-genital with
  ejaculation
  Insertive                   45.9    45.1      44.9       37.7
  Receptive                   42.3    39.3      40.0       38.2
Finger fucking
  Insertive                   58.5    75.9      74.8       68.4
  Receptive                   65.0    73.7      71.7       58.3
Anal intercourse(a)
  Insertive                   56.8    61.2      54.7       57.9
  Receptive                   62.4    55.8      53.7       50.9
Anal with withdrawal(a)
  Insertive                   39.5    46.0      39,0       28.9
  Receptive                   43.6    41.1      37.7       26.3
Rimming
  Insertive                   50.0    55.4      52.3       46.5
  Receptive                   62.4    61.6      52.6       46.1
Fisting
  Insertive                    7.7    10.3      7.4         7.9
  Receptive                   10.3     7.6      8.0         9.8

                             50 or over
                               n = 116   [chi square]      p

Sensuous touching              99.1         1.361           .85
Deep/wet kissing               82.8        31.857       <.00005
Dry kissing                    83.6        23.709        <.0005
Mutual masturbation            90.5        14.815          .005
Oral-genital no
  ejaculation
  Insertive (being sucked)     86.1        26.644       <.00005
  Receptive (sucking)          87.8        23.640        <.0005
Oral-genital with
  ejaculation
  Insertive                    44.8         4.216           .38
  Receptive                    37.1         1.255           .87
Finger fucking
  Insertive                    53.4        34.245       <.00005
  Receptive                    46.6        35.703       <.00005
Anal intercourse(a)
  Insertive                    52.6         3.252           .52
  Receptive                    53.4         7.097           .13
Anal with withdrawal(a)
  Insertive                    19.0        31.471       <.00005
  Receptive                    18.1        34.235       <.00005
Rimming
  Insertive                    34.5        15.411          .004
  Receptive                    32.8        38.471       <.00005
Fisting
  Insertive                     4.3         3.942           .41
  Receptive                     6.9         1.682           .79

(a) With or without condoms

Table 5 Percentages of Men in Each Age Group Who Engaged Occasionally or Often in Sexual Practices with Casual Male Partners in the Six Months Prior to Interview

                           Under 25     25-29     30-39     40-49
                            n = 364   n = 394   n = 560   n = 401

Sensuous touching              96.4      96.2      96.4      95.5
Deep/wet kissing               87.0      84.4      77.8      71.6
Dry kissing                    86.2      82.4      76.5      75.1
Mutual masturbation            92.3      94.1      94.4      94.3
Oral-genital no
  ejaculation
  Insertive (being sucked)     92.3      88.8      89.6      81.0
  Receptive (sucking)          90.9      92.6      89.4      82.3
Oral-genital with
  ejaculation
  Insertive                    46.0      38.5      40.4      36.2
  Receptive                    30.7      24.7      23.3      21.2
Finger fucking
  Insertive                    61.6      63.3      60.6      58.9
  Receptive                    65.7      62.2      61.1      55.1
Anal intercourse(a)
  Insertive                    43.1      46.2      44.7      40.6
  Receptive                    45.3      36.3      34.5      32.9
Anal with Withdrawal(a)
  Insertive                    32.7      33.8      24.9      20.9
  Receptive                    31.3      30.6      22.0      19.2
Rimming
  Insertive                    40.7      40.4      37.6      27.9
  Receptive                    56.4      54.1      52.3      40.4
Fisting
  Insertive                     8.3      10.5      10.9       8.0
  Receptive                     8.8       8.7       7.9       5.2

                                 50 or over
                                 n = 171     [chi square]     p

Sensuous touching                95.3         0.867           .93
Deep/wet kissing                 71.5        40.730       <.00005
Dry kissing                      68.6        30.349       <.00005
Mutual masturbation              87.8        11.196           .02
Oral-genital no ejaculation
  Insertive (being sucked)       78.5        37.141       <.00005
  Receptive (sucking)            75.6        47.556       <.00005
Oral-genital with ejaculation
  Insertive                      51.7        16.818          .002
  Receptive                      25.6        10.161           .04
Finger fucking
  Insertive                      48.8        11.362           .02
  Receptive                      48.8        18.782         <.001
Anal intercourse(a)
  Insertive                      32.0        11.690           .02
  Receptive                      33.9        15.950          .003
Anal with Withdrawal(a)
  Insertive                      14.5        37.826       <.00005
  Receptive                      12.9        37.785       <.00005
Rimming
  Insertive                      26.2        26.016       <.00005
  Receptive                      36.0        37.621       <.00005
Fisting
  Insertive                      3.5         10.684           .03
  Receptive                      2.3         11.594           .02

(a) With or without condoms

When there were significant age-related differences for particular sexual practices, the older men were generally the least likely, or among those least likely, to have engaged in the practice. Several differences between the age groups emerged in sexual practices with regular partners (see Table 4). Older men were the least likely to have engaged in insertive and receptive anal intercourse with withdrawal. They were also the least likely to have engaged in wet and dry kissing, to have engaged in receptive rimming, and to have had insertive and receptive finger fucking. Like their 40- to 49-year-old counterparts, older men were less likely to have had insertive and receptive oral-genital sex (without ejaculation) than men under the age of 40.

By Mantel-Haenszel test, there was a decreasing linear trend with age for most sexual practices with regular partners: wet (p [is less than] .00005) and dry kissing (p [is less than] .00005); mutual masturbation (p [is less than] .005); insertive (p [is less than] .005) and receptive (p [is less than] .001) oral-genital sex without ejaculation; receptive finger fucking (p [is less than] .001); receptive anal intercourse (p [is less than] .05); insertive (p [is less than] .00005) and receptive (p [is less than] .00005) anal intercourse with withdrawal; and insertive (p [is less than] .01) and receptive (p [is less than] .00005) rimming.

With casual partners, the older men were less likely to have had insertive and receptive anal intercourse with withdrawal, to have engaged in dry kissing, to have had insertive and receptive oral-genital sex without ejaculation and to have engaged in receptive rimming and receptive finger fucking. Together with the men aged 40-49 years, they were less likely than younger men to have engaged in wet kissing and insertive rimming (see Table 5).

By Mantel-Haenszel test, there was a decreasing linear trend with age for most sexual practices with casual partners: wet (p [is less than] .00005) and dry kissing (p [is less than] .00005); insertive (p [is less than] .00005) and receptive (p [is less than] .00005) oral-genital sex without ejaculation; receptive oral-genital sex with ejaculation (p [is less than] .05); insertive (p [is less than] .01) and receptive (p [is less than] .00005) finger fucking; insertive (p [is less than] .05) and receptive (p [is less than] .001) anal intercourse; insertive (p [is less than] .00005) and receptive (p [is less than] .00005) anal intercourse with withdrawal; insertive (p [is less than] .00005) and receptive (p [is less than] .00005) rimming; and receptive fisting (p [is less than] .005). There were no increasing linear trends for any sexual practices with either regular or casual partners.

There was confirmation of these patterns of sexual experience on scales that measured the range of specific sexual practices with male partners. The older men had the lowest scores on both the Anal Practice and Oral/Tactile Practice scales (see Table 6). By F-ratio for linear trend, there was a decreasing trend with age for both Anal (p [is less than] .000005) and Oral/Tactile Practices (p [is less than] .0005). In the case of Anal Practice, there also was a significant quadratic relationship (p [is less than] .0001) that indicated a peak range of anal practices for 25- to 29-year-old men and fewer anal practices for younger or older men.

Table 6
Sexual Practice Scales by Age

                Under 25     25-29     30-39    40-49   50 or over
Anal(a)
                 n = 529   n = 512   n = 767   n = 515  n = 256
M                   4.00      4.25      3.83      3.48     2.83
SD                  2.50      2.54      2.50      2.38     2.26
Oral/Tactile(b)
                 n = 529   n = 512   n = 767   n = 516  n = 256
M                   5.30      5.39      5.10      5.16     4.95
SD                  1.54      1.40      1.68      1.38     1.56

(a) F(4, 2574) = 17.22, p < .00005

(b) F(4, 2575) = 5.24, p < .001

Mode of anal intercourse. Of the older men, 156 (60.9%) had had anal intercourse during the 6 months prior to interview: 84 (32.8%) with regular partners and 90 (35.2%) with casual partners. Of the men who engaged in anal intercourse, some age differences were found for mode of intercourse (see Table 7). With regular partners, there were no significant differences between the age groups. With casual partners, older men were the most likely to have had receptive intercourse exclusively; those under 25 years were the least likely to have had insertive intercourse exclusively; those over 40 years were less likely than younger men to have had both insertive and receptive intercourse. This pattern of men over 40 years being less likely to have had both insertive and receptive intercourse was duplicated when intercourse with both regular and casual partners was taken into account.

Table 7
Mode of Anal Intercourse by Age (Percentages)

Partner and
Mode                        Under 25      25-29      30-39

With Regular Partners(a)     n = 190    n = 183    n = 252
  Insertive only                12.6       18.0       18.3
  Receptive only                17.4       10.4       13.9
  Both                          70.0       71.6       67.9

With Casual Partners(b)      n = 238    n = 248    n = 341
  Insertive only                17.2       27.8       31.1
  Receptive only                18.5       10.1       14.4
  Both                          64.3       62.1       54.5
With Either Regular or
Casual Partners'             n = 373    n = 366    n = 517
  Insertive only                14.5       20.5       25.0
  Receptive only                16.6        9.0       13.2
  Both                          68.9       70.5       61.9

Partner and
Mode                         40-49    50 or over

With Regular Partners(a)   n = 178    n = 84
  Insertive only              27.0      19.0
  Receptive only              16.9      21.4
  Both                        56.2      59.5

With Casual Partners(b)    n = 241    n = 90
  Insertive only              34.9      31.1
  Receptive only              19.5      25.6
  Both                        45.6      43.3
With Either Regular or
Casual Partners'           n = 360    n = 156
  Insertive only              30.8       25.6
  Receptive only              16.4       22.4
  Both                        52.8       51.9

Note: Only includes men who engaged in anal intercourse

(a) [chi square] (8 df, n = 887) = 20.96, p = .007

(b) [chi square] (8 df, n = 1158) = 40.77, p < .00005

(c) [chi square] (8 df, n = 1772) = 55.15, p < .00005

Condom use. To answer Research Question 3, we examined condom use with regular and casual partners. With regular partners, there were no significant age group differences (after Bonferroni adjustment) in condom use during anal intercourse (see Table 8). Of those who had anal intercourse with regular partners, more than 40% of the younger and older men sometimes did not use condoms. In the case of casual partners, there were significant differences between the age groups. Older men were significantly more likely than the younger men to exclude anal intercourse from their sexual repertoire with casual partners. However, for younger and older men who had anal intercourse with casual partners, there were no significant differences (after Bonferroni adjustment) between age groups in condom use.

Table 8
Anal Intercourse and Condom Use with Regular and Casual
Male Partners by Age (Percentages)

Partner and Practice   Under 25   25-29   30-39   40-49   50 or over
Regular Partner(a)     n = 232  n = 223  n = 329  n = 229  n = 117

  No anal intercourse     18.1    17.9     23.4    22.3     28.2
  Always condom           45.3    43.0     36.5    32.3     25.6
  Sometimes               36.6    39.0     40.1    45.4     46.2
    unprotected
Only Those Who Had Anal Intercourse with Regular Partners(b)

                       n = 190  n = 183  n = 252  n = 178  n = 84

  Always condom           55.3    52.5     47.6    41.6     35.7
  Sometimes               44.7    47.5     52.4    58.4     64.3
    unprotected

Casual Partners(c)     n = 370  n = 395  n = 578  n = 428  n = 189

  No anal intercourse     35.7    35.2     41.0    43.7     52.4
  Always condom           45.7    49.9     43.3    43.7     30.7
  Sometimes               18.6    12.9     15.7    12.6     16.9
    unprotected

Only Those Who Had Anal Intercourse with Casual Partners(d)

                       n = 238  n = 248  n = 341  n = 241  n = 90

  Always condom           71.0    79.4     73.3    77.6     64.4
  Sometimes               29.0    20.6     26.7    22.4     35.6
    unprotected

(a) [chi square](8 df, n = 1130) = 20.47, p = .009

(b) [chi square](4 df, n = 887) = 13,5 1, p = .009

(c) [chi square](8 df, n = 1960) = 28.32, p < .0005

(d) [chi square](4 df, n = 1158) = 10.91, p = .03

Test status and contact with the epidemic. Older men were the least likely to have been tested for HIV antibodies, [chi square] (4 df, n = 2576) = 30.70, p [is less than] .00005 (answering Research Question 2 in the negative). There were eight HIV-positive older men, a smaller proportion than for the age groups between 25 and 49 (see Table 9). Contact with the epidemic (the subject of Research Question 4) varied significantly with age; however, the older men generally had a similar degree of contact to the other men over 25 years (see Table 10). The older men's Contact with the Epidemic' Scale score of 1.30 indicated moderate involvement with people living with HIV/AIDS. By F-ratio for trend, there was a significant quadratic relationship between Contact with the Epidemic and age (p [is less than] .000005), reflecting the fact that contact increased sharply up to the 30-39 age group and thereafter declined slightly with increasing age.

Table 9
HIV Status by Age (Percentages)

Test Status     Under 25   25-29    30-39     40-49    50 or over
                n = 529   n = 512  n = 767   n = 516    n = 256

No test          25.7      22.9     21.9       27.9       36.3
Negative         70.5      68.4     70.7       64.0       60.5
Positive          3.8       8.8      7.4        8.1        3.1

[chi square] (8 df, n = 2580) = 40.12, p < .00005

Table 10

Contact with the Epidemic Scale by Age

           Under 25    25-29     30-39     40-49    50 or over
           n = 529    n = 512   n = 767   n = 516    n = 256

m           1.06       1.38      1.47      1.36       1.30
SD          1.09       1.13      1.12      1.12       1.14

F(4, 2575) = 11. 17, p < .00005

Multivariate Perspective

Our findings that anal and oral/tactile practices declined with age were new. To answer our first research question about older men's sexual practices, it was important to know whether these differences could be explained by the other differences that were observed between younger and older men. Thus, a multivariate analysis of covariance (MANCOVA) was conducted on the Anal Practice and Oral/Tactile Practice scales to investigate whether there were any age group differences in range of sexual practices after adjustment for other variables. Age and region were treated as factors alongside the following covariates: social attachment, cultural involvement, occupation, living situation, length of time at current address, country of birth, marital status, sexual identity, and antibody status. There was no significant age x region interaction, F(16, 5018) = 0.82, p = .66. After adjustment for all covariates, including region, the analysis yielded significant age-related differences in the scales of sexual practice, F(8, 5018) = 4.46, p [is less than] .00005 (Pillai's criterion). Confirming our univariate results, differences on the Anal Practices Scale were significant, F(4, 2509) = 6.68, p [is less than] .00005. However, there were no significant differences on the Oral/Tactile Practices Scale after adjustment for the covariates, F(4, 2509) = 1.43, p = .22.

Discussion

Using data from an Australian nationwide telephone survey (Project Male-Call: Kippax et al., 1994), we set out to develop a profile of older (over 49 years) homosexually active men. Our focus was on the neglected area of such older men's sexual practices and features that differentiate older men from their younger counterparts. Based on limited although fairly consistent past findings (Bennett & Thompson, 1980; Berger & Kelly, 1986; Weinberg & Williams, 1974), we generated five hypotheses about older men's living arrangements, gay community involvement, and sexual practices. In the absence of previous data, we also posed three research questions about older men's sexual practices and their testing, responses, and contact in relation to HIV/AIDS.

In contrast with the younger men, the older men in the sample were more ethnically homogeneous and, as is to be expected, more likely to be out of the workforce. Comparatively, older men were more stable in terms of residence. They tended not to live in predominantly gay areas, whereas they were overrepresented in rural areas. Consistent with Hypothesis 1 and earlier Australian and international data (Bennett & Thompson, 1980; Weinberg & Williams, 1974), older men were more likely to live alone than with a male sexual partner, although more older than younger men lived with a female sexual partner. In line with Hypothesis 2, older men were more likely to be or have been married. Correspondingly, and perhaps because of a time factor, they were more likely to have children. These findings are in accord with earlier evidence reported by Berger and Kelly (1986) and Lee (1987).

Older men differed somewhat from younger men in terms of self-defined sexual identity, with older men more likely to describe themselves as homosexual rather than gay. Moreover, and largely consonant with Hypothesis 3 and the findings of Bennett and Thompson (1980), older men were less likely to have disclosed their sexual orientation to others. It would appear that the more open and supportive post-Stonewall climate facilitates younger gay men's divulging their sexual identity to a broad range of family members, friends, work colleagues, and acquaintances, particularly after the age of 25. This accords with the generational changes exemplified in various accounts by gay men in Wotherspoon's (1986) collection of autobiographies.

Contrary to Hypothesis 4 and some previous findings (Bennett & Thompson, 1980), older men's attachment to gay community (measured in terms of social attachment, cultural involvement, or sexual involvement) was less than for their younger counterparts. These results, though, are partly in line with Weinberg and Williams' (1974) finding that older homosexual men were lower in overall social involvement than younger homosexual men. Nonetheless, the data indicated that most older men maintained moderately strong social links with gay community, a finding that augurs well for their psychosocial adaptation in older age (Berger & Kelly, 1986).

Consistent with Hypothesis 5 and previous evidence (Berger, 1980), older men had as many different male partners in a six-month period as younger men. For approximately half of the older group, these partnerships were casual encounters exclusively. About one in five older men was in a monogamous regular relationship, and as many older men were in regular relationships that did not preclude liaisons with other male sexual partners. Older men had as many female partners in a six-month period as younger men.

We turn now to our first research question that asked if older homosexually active men engage in the same sexual practices as their younger counterparts. Some older men engaged in an extensive range of sexual practices with regular and casual male partners, but as a group older men tended to engage in the practices less frequently than their younger counterparts. For most sexual practices, a fairly consistent pattern indicated a decreasing linear trend in frequency with age. The finding that this trend applied to most practices is important. It suggests a general downward trend for a whole range of sexual practices rather than practice substitution with age.

Older men as a group scored significantly lower than younger men in terms of extent of anal and oral/tactile sexual repertoire. Differences between the age groups in range of oral/tactile practices were not significant after controlling for demographic and milieu factors. Separate analyses revealed that differences in oral/tactile practices were largely accounted for by levels of gay community involvement, with men not attached to gay community less likely to have an extensive oral/tactile repertoire. However, differences in anal practices were actually age based and could not be attributed to other factors such as region of residence or degree of attachment to gay community. This raises the interesting possibility that the anal practice difference may be generational, a product of the cultural climate in which the men were socialised.

Proportionately, older men engaged in more anal intercourse with regular than casual partners. However, regardless of partner type, older men who routinely engaged in anal intercourse were as likely to be the insertive as the receptive partner. A majority of the older men who engaged in anal intercourse were versatile with regular partners, switching between insertive and receptive roles. This corroborates earlier evidence documented by Berger and Kelly (1986). Only about one in five older men who engaged in anal intercourse in a regular relationship was exclusively insertive and a similar proportion exclusively receptive. In contrast with regular partners, fewer older men were versatile, and more of them were exclusively insertive or receptive with their casual partners.

When anal intercourse with both types of partners (either regular or casual) was taken into account, older men (including those in the 40-49 age group) were significantly less likely than younger men to have engaged in both insertive and receptive intercourse. Engagement in both modes of anal intercourse was much more usual for men under 30 years.

Turning to our three research questions connected with HIV/AIDS, are older homosexual men as likely as their younger counterparts to have been tested for HIV, as responsive to safe sex campaigns (as assessed by condom use), and as close to the HIV epidemic? Older men were significantly less likely to have been tested for HIV antibodies. Regardless, of the men who engaged in anal intercourse, older men were no less likely than younger men always to use condoms during anal intercourse with their regular or casual partners. In the case of casual partnerships, there was a conspicuous difference between the age groups, with more than half the older men not engaging in anal intercourse with their casual male partners. This may be due to a number of factors, such as the availability of condoms (if condoms are less available to older men in rural areas), economic constraints (if older men are paying for sex more often than younger men--anal being more expensive than oral sex), and sites of sexual activity (if older men are having more sex in' public places, making anal intercourse less possible).

On the other hand, this finding may point to a possible age difference in response to safe sex campaigns. One interpretation is that older men erroneously perceived casual sexual encounters as inherently less safe. An alternate explanation is that a smaller proportion of the older men, belonging to the pre-AIDS generation, were comfortable with and adept at using condoms. If either of these latter possibilities is correct, it would indicate that safe sex campaigns, which are often targeted explicitly or implicitly--the latter by virtue of their media a and youthful iconography--at younger gay men, are not having as much impact among older homosexually active men. Further work is needed to unpack the reasons behind older men's less frequent engagement in anal intercourse with casual partners.

In terms of having known people who died following AIDS and having known or having cared for seropositive people, older men had similar levels of contact with the epidemic as the men in the 25-29 and 40-49 age brackets. The youngest men have had least and the 30- to 39-year-old men greatest contact with the epidemic.

An important consideration is to be heeded in the interpretation and generalisation of our findings: The sample was not and could not be drawn randomly. The Project Male Call men were recruited through diverse sources and included men from widely different regional, social, and ethnic backgrounds. Nevertheless, the representativeness of the sample to the total population of homosexually active Australian men is not known because of the lack of a sampling frame for this group. There is a possibility that the sample was more representative of the better informed and motivated segment of the homo sexual population and more representative of men affiliated with gay community. Certainly, compared with Australian Census data, the sample underrepresented older men and over represented tertiary educated and professionally employed men. Replication of our findings from a house hold-based sampling frame would be useful. Such replication is important, as there may be older homosexually active men who are socially isolated, have few community or cultural supports, and have not accommodated well to the aging process, and whose needs should not be diminished.

Despite the sampling limitations, the Project Male-Call data and the findings presented here paint an interesting picture of older Australian homosexually active men. Our findings are generally in agreement with the limited previous research evidence concerning older homosexual men. As a group, older homosexually active men are sexually motivated, have opportunities to satisfy sexual desires, and are involved in gay community. They are committed to a homosexual orientation, although they are more likely than post-Stonewall generations to have accommodated to prevailing heterosexist ideology through limited disclosure of their sexual identity and through straight marriage and family life.

This analysis points to some specific issues for safe sex campaigns aimed at older homosexually active men. In the HIV/AIDS era, older men are more likely than their younger counterparts to have no anal intercourse rather than condom-protected anal intercourse with casual partners. Such a strategy (be it deliberate or otherwise) is quite safe so long as it is sustainable. Almost half the older men sometimes do not use condoms for anal intercourse with regular male partners. This may or may not be a problem, depending on whether the men know their HIV status and have negotiated a reliable agreement about safe sex outside their partnership. In the absence of HIV antibody testing--a more common characteristic of the older men--such sexual negotiation to exclude HIV from the relationship is meaningless. It is therefore important to extend existing campaigns to improve older men's safe sex understandings and to encourage HIV antibody testing. Given that older men were more likely to have been recruited through the mainstream rather than gay press, those running gay men's education programs would be advised to use a variety of media, metropolitan and rural, and not just gay avenues, to ensure that all constituents are reached.

References

Bennett, K. C., & Thompson, N. L. (1980). Social and psychological functioning of the ageing male homosexual. British Journal of Psychiatry, 137, 361-370.

Berger, R. M. (1980). Psychological adaptation of the older homosexual male. Journal of Homosexuality, 5, 161-175.

Berger, R. M., & Kelly, J. J. (1986). Working with homosexuals of the older population. Social Casework: The Journal of Contemporary Social Work, 67, 203-210.

Friend, R. A. (1987). The individual and social psychology of aging: Clinical implications for lesbians and gay men. Journal of Homosexuality, 14, 307-331.

Friend, R. A. (1989). Older lesbian and gay people: Responding to homophobia. Marriage and Family Review, 14, 241-263.

Kippax, S., Connell, R. W., Dowsett, G. W., & Crawford, J. (1993). Sustaining safe sex: Gay communities respond to AIDS. London: Falmer Press.

Kippax, S., Crawford, J., Rodden, P., & Benton, K. (1994). Report on Project Male-Call: National telephone survey of men who have sex with men. Canberra: Australian Government Publishing Service.

Lee, J. A. (1987). What can homosexual aging studies contribute to theories of aging? Journal of Homosexuality, 13, 43-71.

McDougall, G. J. (1993). Therapeutic issues with gay and lesbian elders. Clinical Gerontologist, 14(1), 45-57.

Weinberg, M. S., & Williams, C. J. (1974). Male homosexuals: Their problems and adaptations. New York: Oxford University Press.

Wotherspoon, G. (Ed.). (1986). Being different: Nine gay men remember. Sydney: Hale & Iremonger.

We gratefully acknowledge those who made the study possible: the participants; staff and volunteers of various AIDS councils and organisations; and those who promoted, recruited, and interviewed for the survey. Correspondence may be addressed to Paul Van de Ven, Ph.D., National Centre in HIV Social Research, School of Behavioural Sciences, Macquarie University, Sydney, NSW 2109, Australia. E-mail: paul.vandeven@ mq.edu.au.

Manuscript accepted March 5, 1997