Sexual risk behaviors among heterosexual HIV serodiscordant couples in the era of post-exposure prevention and viral suppressive therapy

Objectives: To describe awareness and use of antiretroviral treatments, viral load monitoring, and post-exposure prevention; to assess changing concerns about HIV transmission; and to examine the effect of these advances on sexual behavior in HIV-serodiscordant heterosexual couples.

Cited Here...: Cross-sectional analysis of a baseline sample of 104 couples (n = 208 individuals) from the California Partners Study II, an intervention trial for HIV-serodiscordant couples in California. Questions on sexual practices, viral load testing, HIV treatment, post-exposure prevention, and their effect on sexual behaviors, risk taking and transmission concerns were measured at intake.

Cited Here...: Over two-thirds of couple members surveyed reported unprotected sex with their partner in the past 6 months. Among seropositive respondents, 37% were taking protease inhibitor therapy, 92% had undergone viral load testing, and of those, 40% said it had ben undetectable at their most recent test. Most respondents, regardless of serostatus, said that viral load testing and awareness of post-exposure prevention had no effect on their condom use. In addition, perceiving that their partner had an undetectable viral load was associated with having protected sex among seronegative subjects (P < 0.05). Seropositive respondent taking protease inhibitors were 2.4 times less likely to report unprotected sex compared with those not taking protease inhibitors (P = 0.05). However, up to 33% of seropositive and 40% of seronegative respondents acknowledged decreased transmission concerns in the light of the new HIV treatments. In comparison with their seropositive partners, seronegative individuals were more likely to acknowledge increased risk taking and decreased HIV transmission concerns (P < 0.05).

Conclusions: New medical advances were not associated with unprotected sex in HIV-serodiscordant couples. However, new treatment options may decrease concerns about HIV transmission, particularly among seronegative partners. Providers should discuss the effect of antiretroviral treatments on sexual transmission risk with their patients. The inclusion of seronegative partners in counseling interventions may decrease risk taking in serodiscordant couples.

Introduction

In the United States, significant advances have been made in diagnosing and treating HIV infection ^[1]. There are concerns, however, that these medical advances may disinhibit risk behaviors ^[2]. Effective therapies with protease inhibitors can reduce plasma viremia below detectable levels, possibly decreasing the perceived threat of HIV transmission and changing people's attitudes and behaviors. Similarly, the recent availability of non-occupational post-exposure prevention may encourage risk taking among seronegative people ^[3]. Increased laxity towards safer sex practices may be of special concern given new evidence of transmission of multiple-drug-resistant strains of HIV ^[4].

New antiretroviral treatments have decreased risk perception in some gay men and seropositive individuals ^[5-7]. However, it is not clear how these changing perceptions affect risk behavior ^[8-10]. To our knowledge, no data have yet been published on the effect of new treatments among high risk heterosexual individuals. The objectives of this study were to assess: (1) whether participants were aware of and/or used viral load monitoring, antiretroviral treatments and post-exposure prevention; (2) if there was a relationship between awareness and/or use of these medical advances and sexual risk behavior; (3) whether participants reported changes in their degree of concern about HIV transmission because of the availability of new HIV treatments; and (4) if there was a relationship between decreased concerns about transmission and sexual risk behavior.

Methods

Subjects

The California Partners Study II is an ongoing behavioral intervention trial for heterosexual HIV-serodiscordant couples. Risk reduction counseling is combined with behavioral surveys, laboratory testing for HIV, sexually transmitted diseases and pregnancy, and follow-up for 1 year. Enrollment began in December 1996. As of January 1999, 104 eligible couples (i.e., those reporting one or more episodes of vaginal or anal sex in the previous 6 months) had complete baseline information available for analysis. Participants were recruited through a variety of venues, including word of mouth (32%), advertisement/media (28%), clinics (23%), agencies (19%), and direct outreach (8%).

Measures

As part of the baseline visit, study participants were individually administered a 90 min survey by a trained interviewer. Questions covered demographic and relationship information, drug use, sexual practices, HIV and viral load testing, HIV treatments, and their effect on sexual behaviors and transmission concerns.

Sexual risk behavior in the past 6 months

All participants were asked about the amount of condom use with their HIV-serodiscordant partner for both vaginal and anal sex. Unprotected sex was defined for each participant as less than 100% condom use for vaginal or anal sex in the previous 6 months.

Knowledge and use of HIV diagnostics and treatments

HIV-seronegative respondents were asked: (1) if their seropositive partner had undergone viral load testing, and if so, what was the most recent viral load level; (2) if viral load testing by their partner had influenced condom use within the couple; (3) if their seropositive partner was currently taking any antiretroviral therapy; (4) if they had heard of post-exposure prevention; (5) if they had ever taken any medication after a possible exposure to HIV to prevent getting infected; and (6) if awareness of post-exposure prevention had influenced condom use within the couple. Seropositive respondents were asked: (1) if they had undergone viral load testing, and if so, what was the most recent viral load level; (2) if their viral load testing had influenced condom use within the couple; (3) if they were currently taking any antiretroviral, and if so, to list all of their current HIV medications [coded for whether or not they were taking any protease inhibitors]; (4) if they had heard of post-exposure prevention; and (5) if awareness of post-exposure prevention had influenced condom use within the couple.

HIV risk taking and transmission concerns

Participants were asked how strongly they agreed or disagreed with one item measuring perceived risk taking and three items measuring HIV transmission concerns in the light of improved HIV treatments (adapted from ^[5]). Regarding perceived risk taking, seronegative and seropositive partners were asked if new treatments for HIV had influenced them to take more chances when having sex. To measure the effect of new HIV treatments on participants' transmission concerns, seronegative and seropositive partners were asked respectively if new treatments for HIV had influenced them to: (1) have decreased transmission concerns (I am more willing to take a chance of getting infected when having sex with my partner/I am less concerned about passing HIV to my partner when having sex); (2) have decreased infection concerns (I am much less concerned about becoming seropositive myself/I am much less concerned about being seropositive); (3) have decreased infectivity concerns (I am less likely to get infected from my partner who receives new treatments/I am less likely to infect my partner because I am taking the new treatments). This last item was asked only if the seropositive partner was currently taking antiretroviral therapy.

Analysis

Cross-sectional analysis of the baseline sample of 208 respondents (104 couples) was performed using SAS (version 6.12; SAS Institute, Cary, North Carolina, USA). Individual-level variables were used for all the analyses. We used χ² or Fisher's exact tests to assess bivariate associations, Kappa and McNemar tests to assess level of agreement within couples for HIV-transmission concerns. Independent predictors of unprotected sex were examined separately for seronegative and seropositive partners, using logistic regression models.

Results

Demographic and HIV sexual risk behavior information for each partner of the 104 heterosexual HIV-serodiscordant couples are presented in Table 1. A majority of respondents were over 30 years of age, non-white, had a high school degree or less and a history of injecting drug use. Seropositive partners reported a median of 4 years since diagnosis (range 0-17 years). Most couples reported living together and the median length of the relationship was 3 years (range 0-31 years). Over two-thirds of respondents reported unprotected sex in the past 6 months.

Table 1

Use and effect of viral load testing, antiretroviral therapy and post-exposure prevention on sexual risk behavior

Ninety-two percent of seropositive partners reported having their viral load measured; however, only 70% of seronegative individuals reported knowing that their seropositive partner had undergone viral load testing (Table 2). Seropositive subjects who had an undetectable viral load were as likely as those with a higher viral load to have had unprotected sex in the past 6 months. In contrast, seronegative subjects who said that their partner had an undetectable viral load were less likely to have had unprotected sex, compared with those who said their partner had a higher viral load. Among those who reported viral load testing in self or partner, almost all reported that knowing the viral load of the seropositive partner had no effect on condom use in the couple.

Table 2

Only 51% of seropositive partners were currently taking antiretroviral therapy (median number of medications = 3, range 1-5) and 37% were taking protease inhibitors either alone or in combination. Seropositive subjects who currently took antiretroviral therapy were less likely to report unprotected sex in the past 6 months compared with those not taking it. Because improved treatment outcomes are mostly driven by the use of protease inhibitors, we also examined the relationship between taking protease inhibitors and unprotected sex among the seropositive respondents. Seropositive subjects who were taking protease inhibitors were also less likely to report any unprotected sex, compared with those not on protease inhibitors.

Over one-quarter of all respondents had heard about post-exposure prevention; however, only five of the seronegative partners reported taking medications after a possible exposure to HIV. Awareness of post-exposure prevention was not associated with unprotected sex in seronegative or seropositive partners. Finally, of those who had heard about post-exposure prevention, none reported modifying condom use because of its availability (data not shown).

HIV risk taking and transmission concerns in the light of improved HIV treatments

Over one-third of seronegative and 15% of seropositive partners agreed that they had already taken a chance with unprotected sex because of improved HIV treatments. As expected, agreement with the statement about HIV risk taking was associated with having unprotected sex in the past 6 months among both the seronegative and seropositive partners (Table 3).

Table 3

Up to 40% of seronegative and 33% of seropositive subjects agreed that new HIV treatments had changed their HIV transmission concerns in some ways. The association between the three HIV transmission concern items and unprotected sex was examined separately for seronegative and seropositive partners. No significant associations were found (Table 3).

Matched analyses comparing partners' responses to the items measuring HIV risk taking and transmission concerns indicated that in comparison with their seropositive partners, seronegative individuals were more likely to report HIV risk taking in the light of new HIV treatments [34 versus 15%; odds ratio (OR) = 3.1;P = 0.002]. Seronegative subjects were also somewhat more likely to report decreased infectivity concerns compared with their seropositive partners (40 versus 21%; OR = 2.6;P = 0.059). No other significant within-couple differences were found.

Independent predictors of unprotected sex

Among seronegatives respondents, multivariate analyses were not conducted as none of the demographic (age, gender, race, education, income, injecting drug use, geographical location), relationship (living together, relationship length) or treatment variables examined were significantly associated with unprotected sex, with the exception of undetectable viral load in the seropositive partner (see Table 2). Among seropositive respondents, income and relationship length were associated with unprotected sex and were entered into multivariate logistic regression models with the treatment variables, to identify independent predictors of unprotected sex (Perceived risk taking was not entered in the logistic regression models because it had no explanatory value among those who reported no unprotected sex.) (Table 4). Taking protease inhibitors remained significantly associated with protected sex, after controlling for the other significant variable in the model. A similar trend, although not reaching statistical significance, was observed when taking antiretroviral therapy was replaced by taking protease inhibitors in the multivariate model (data not shown).

Table 4

Discussion

Among seropositive respondents, taking protease inhibitors, and being on antiretroviral therapy, were associated with always having protected sex in the past 6 months. Among seronegative individuals, perceiving that their partner had an undetectable viral load was also associated with having consistent protected sex. Furthermore, most HIV-serodiscordant couple members surveyed said that viral load and post-exposure prevention had no effect on their sexual risk behavior. If these findings hold over time, it could indicate that members of HIV-serodiscordant couples aren their sexual risk behavior. If these findings hold over time, it could indicate that members of HIV-serodiscordant couples are concerned about the continued risk of sexual transmission despite undetectable blood viral load and the associated risk of transmitting a resistant virus to the uninfected partner ^[1,4]. Furthermore, seropositive individuals who are willing to take HIV therapy may be more health conscious, and thus perhaps more concerned about keeping their uninfected partner healthy. Access to the latest treatments should include accurate information regarding treatments' effects on viral transmission to help people maintain sexual safety while getting the health benefits of effective therapies.

Despite these reassuring findings, current attitudes in this sample indicated increased risk taking and decreased concerns around HIV transmission among a minority of respondents. Because attitudes can be precursors to behaviors, monitoring this trend should continue, particularly among the seronegative partners. These results also suggest that seronegative and seropositive partners may be differentially influenced by the new medical advances and, as previously reported, that the seronegative partner is often the one struggling with condom use and more willing to take sexual risks ^[10]. The findings also underscore the importance of assessing changes in perception and behavior from each partner's perspective because motivations may differ and risk taking is influenced by both members of a couple.

This study used a convenience sample and thus, has limited generalizability. In addition, the small sample size may lack the power to detect significant associations between attitudes and risk behavior. The cross-sectional design limits causality inferences from the data. Nonetheless, these results are reassuring and suggest that, at present, new medical advances are not associated with increased risk behavior in the majority of HIV-serodiscordant couples surveyed. Use of antiretroviral therapy was lower in this sample compared with seropositive gay men ^[11], and exposure to new treatments may not have reached a level that would cause people to consider HIV/AIDS as a chronic manageable condition. As exposure to and use of new medical advances increase, it is possible that changing attitudes towards the disease will have a greater effect on transmission risk perception and behaviors.

In a changing time during the AIDS epidemic, it is critical to continue to monitor the effect of new medical advances on HIV transmission risk perception and risk taking in seropositive and high-risk seronegative populations. Risk behaviors may eventually increase as better treatments continue to improve the quality of life of seropositive people, and as scientists debate the extent to which antiretroviral therapy may reduce the infectivity of the seropositive partner. Providers should expand the discussion of these advances from health benefits to their effect on the sexual relationship of their patients, and their seronegative partners should be included in such discussions.