Redefining Gay Male Anal Intercourse Behaviors: Implications for HIV Prevention and Research

Journal of Sex Research, Nov, 1998 by Donald C. Barrett, Gail Bolan, John M. Jr. Douglas

Research on factors (e.g., age) that influence gay male involvement in HIV risk behaviors usually examines the relationships between such factors and involvement in unprotected anal intercourse (UAI). Frequently the relationship between predictors and risk uses, as a classification of risk, whether or not the subjects report any unprotected anal intercourse (Paul, Stall, Crosby, Barrett, & Midanik, 1994; Peterson et al., 1992). A second means of classifying groups men into those who report UAI outside of a monogamous relationship and those who report no UAI or report UAI only within a monogamous relationship (Kelly et al., 1995; Stall et al., 1992). When such categories are used to identify the correlates of HIV risk-taking for prevention purposes, there is an implicit assumption of social, psychological, and behavioral similarity within groups of men who take risks and within groups of men who are safer.

Research by Paul et al. (1994) suggests that this assumption of similarity may not be valid. In their research on gay men attending a drug treatment program, they found that the implicit or explicit intent of the subject or his partner to withdraw was cited as a reason for having UAI by approximately 50% of the men reporting UAI. Withdrawing before ejaculation may thus represent a common compromise between using condoms or not having UAI. If this compromise represents decisions regarding sexual safety, then men who have made the decision to withdraw during anal intercourse may share social, personality, or behavioral characteristics with other safer men.

Similarly, the assumption of commonality between safer men who have protected anal intercourse and safer men who do not have anal intercourse is questionable. It is well known that condoms are not 100% reliable; thus, men whose anal intercourse is always protected may represent a group of low to moderate risk-takers who share some characteristics with men who have UAI. Men who have no anal intercourse may thus represent a relatively unique group that places a very strong emphasis on sexual safety, avoiding all anal intercourse.

Recent writings by Johnston (1995) and by Odets (1995) further suggest the value of examining how we classify anal intercourse. As these writers suggest, anal penetration is likely to be symbolically meaningful and closely linked to intimacy needs for some men. Thus, men who engage in UAI to ejaculation may rank the intimacy value of anal penetration well above concerns about infection, while men who engage in UAI with withdrawal, men who have protected anal intercourse only, and men who do not have any anal penetration may represent decreasing rankings of the intimacy value of anal penetration and increasing levels of concern for infection. Prevention classifications that dichotomize sexual activity may be missing these important variations in risk-taking.

Whether variations in anal intercourse behaviors represent differing priorities regarding infection is important for both HIV prevention and research. Prevention efforts that treat all men who have UAI the same may be ineffective if a large portion of that group has already made their decision about sexual safety. Similarly, assumptions about the resources and skills necessary for safer behavior may be incorrect if "safe" does not represent a common group among gay men. For research, if dichotomous classifications are incorrect then estimates of the relative importance of risk factors may be incorrectly specified, and may result in the misdirection of research efforts.

The purpose of this research is to examine the value to HIV prevention of defining more detailed groupings of risk behaviors. Thus, this research focuses on three issues: 1) the ability to identify distinct groups of men based on their anal intercourse behaviors, 2) the differences between such groups that are relevant for prevention, and 3) the effect of such a distinction on research. The classification scheme used here divides the men into four groups: 1) men who report that they or their partner ejaculated into the receptive partner's rectum at least once during UAI (UAI to ejaculation group), 2) men who report that they and their partners consistently withdrew during UAI without ejaculating into the receptive partner's rectum (UAI with consistent withdrawal group), 3) men who report having protected anal intercourse and no UAI (protected anal only group), and 4) men who report having no anal intercourse.

Another way of categorizing that has been suggested is to define groups by whether the subject was in the insertive or receptive role. While such a distinction seems relevant for analyzing risks of HIV infection, such a distinction is less relevant to prevention. Key to the overall message of HIV prevention is one of shared responsibility and shared risk; thus, condom messages stress protecting oneself and others from potential infection. This message is reinforced by stressing that insertive partners are also at risk during UAI. Therefore, from a prevention perspective risk-taking is not related to sexual role, and a distinction between insertive and receptive roles is inappropriate. Besides the basic characteristics of prevention messages, there is a practical consideration. As will be shown in the results, the percent of men in the sample engaging strictly in either receptive anal intercourse or strictly in insertive anal intercourse is relatively small, thus reducing the ability to analyze such differences.