Levels of Risky Sex Did Not Rise When HIV Therapy Was Initiated in KwaZulu-Natal
Levels of unsafe sexual behavior did not increase when antiretroviral therapy became available in a rural region of KwaZulu-Natal, South Africa, and in some regards residents’ sexual behavior became safer, according to an analysis of seven years of surveillance data.1 For example, from 2005 to 2011, the proportion of adults who reported having used a condom the last time they had sex with their regular partner rose by an average of 2.6 percentage points annually among men and by 4.1 points annually among women; increases were apparent among both HIV-positive respondents and those who were uninfected. Moreover, the proportion of respondents who had had more than one partner in the past year declined, as did the prevalence of concurrency.
Several studies have found that levels of sexual risk behaviors decrease when patients begin antiretroviral therapy. However, evidence has been lacking on whether similar trends occur in the general population; one concern has been that risky behavior may increase if the public believes that the threat posed by HIV is declining (e.g., because treatments are now available). To examine this issue, McGrath and colleagues analyzed surveillance data from Umkhanyakude district, a rural area of KwaZulu-Natal where HIV prevalence is extremely high (29% in 2011) and where data from more than 11,000 households have been collected since 2000 by the Africa Centre for Health and Population Studies.
The current analysis focused on the period from 2005 (around the time the local HIV treatment program was initiated) to 2011. During that time, all residents aged 15 or older were eligible each year to be tested for HIV and to take part in a behavioral survey; they could participate in one, both or neither. A stratified random sample of 10% of nonresident household members was also invited to participate. Respondents provided information on a range of behaviors, including whether they were sexually active, the number of partners they had had in the past year, whether they had had concurrent partners and whether they had used a condom the last time they had had sex with their regular and casual partners. The number of survey participants in a given year varied from 9,400 to 13,300, but generally declined during the study period; participation rates ranged from 26% to 42% among men and from 38% to 54% among women. The researchers used sampling weights to account for nonparticipation, and multiple imputation to adjust for missing responses to specific behavioral questions.
Throughout the study, women were more likely than men to report having ever had sex, reflecting their younger age at sexual debut, but they consistently had had fewer partners and casual partners in the past year and reported lower levels of concurrency. The proportion of participants who said they knew their HIV status rose substantially among both women (from 46% to 82%) and men (from 30% to 55%).
Despite the availability of treatment for HIV, participants’ sexual behavior tended to get safer, not riskier. Notably, the proportion of respondents who reported having used a condom the last time they had sex with their regular partner increased by 2.6 percentage points annually among men and by 4.1 points annually among women. Moreover, the proportion who had had more than one partner in the past year declined, again among both men (by 1.2 points per year) and women (by 0.4 points per year), and the prevalence of concurrency decreased as well (by 0.1–0.6 points annually). Trends were generally similar, though not always statistically significant, in subgroup analyses that categorized respondents according to their age and marital status (30 or older and married, 30 or older and unmarried, or younger than 30) and knowledge of HIV status. No changes occurred, either in the full sample or in most subgroups, in the proportion of respondents who had had a casual partner in the past year, in the proportion who had used a condom the last time they had sex with a casual partner, or in the mean age difference between respondents and their regular partner.
Regardless of gender, the proportion of participants who reported having used a condom the last time they had sex with their regular partner increased among both HIV-positive and HIV-negative respondents, and among those who knew their HIV status and those who did not. HIV-positive women were consistently more likely than their uninfected counterparts to have used a condom at last sex with their regular partner; however, men’s condom use with their regular partner did not differ by HIV status.
Limitations of the study include the high rates of nonparticipation and the possibility that the reductions in risky behavior reflected increased social desirability bias. Nonetheless, the findings reveal “no evidence of an increase in risky sexual behaviour at the population level as access to [antiretroviral therapy] expanded,” the researchers note; moreover, the fact that condom use increased even among individuals who did not know their HIV status suggests that the trend toward higher condom use in the general population was not merely the result of more people learning their status and adjusting their behavior accordingly. While these findings are “welcome news,” continued monitoring of sexual behavior is necessary to ensure that changes in such behavior do not undermine the population-level effects of antiretroviral therapy programs, the authors conclude.—P. Doskoch
1. McGrath N et al., Sexual behaviour in a rural high HIV prevalence South African community: time trends in the antiretroviral treatment era, AIDS, 2013, 27(15):2461–2470.