sex practices may underlie, or even predict, changes in the practice of UAI among gay men.During the early 1990s, outbreaks of hepatitis A among gay men were reported in Australia 20,21 and internationally. 19,[22][23][24] Hepatitis A virus (HAV) is mainly transmitted through the faeco-oral route. More frequent oral-anal contact during sex (rimming) may have put homosexual men at increased risk of acquiring HAV. 25 In this study, we examined changes in sex practices with casual partners among gay men in Sydney. As well as the practice of anal intercourse and the use of condoms, we also examined fisting, watersports, rimming, sadomasochism and fellatio. Whereas some of these have been recorded in some, mainly cross-sectional, studies at different times throughout the period of the HIV epidemic, the data reported here have been collected relatively consistently throughout this period from similar studies of gay men.
Methods
Participants and procedureBaseline data from four studies of gay men in Sydney were reanalysed. These studies were: 1986/87 data from the Social Aspects 32,33 open cohort of HIV-negative gay men. Each study recruited a broadly based, although nonrandom, sample of men who had sex with at least one man during the previous five years or who identified as homosexual. In each study, the men were interviewed face to face using detailed questionnaires, which included compatible questions on a range of sex practices. Men were recruited into each study using similar methods through diverse and broadly similar gay community sources. The most recent of these studies -HIM and pH -are specifically of HIV-negative and HIV-positive men, respectively. HIV-positive men have previously been shown to demonstrate important differences in sexual behaviour to other homosexual men. 34 We have therefore analysed the data separately for HIV-positive men and other men.
QuestionnaireThe interview schedules were based on items used successfully across all four studies. 32 Many items, including those regarding sexual relationships and sex practices with men, were asked in similar ways in each study. In all four studies, men were asked about their specific sex practices, separately for casual and regular partners, during the previous six months. Most sex practices were also asked with respect to both the receptive and insertive positions. For both oral and anal intercourse, they were also asked whether ejaculation inside the mouth or rectum (respectively) occurred, and, for anal intercourse, whether a condom was used.
Data analysisThe data were analysed using frequencies and summary statistics. Chi-square tests were conducted to examine the significance of univariate relationships across the time periods, including Mantel-Haenszel trend tests for linearity. As there were minor differences in the age profiles of these samples, we also analysed these trends separately adjusted for age.
ResultsAltogether, 535 men completed a baseline questionnaire for SAPA during 1986and 1987, 1,147 for SMASH between 1993, 237 for pH between ...