Taking antiretroviral therapy (ART) did not raise the likelihood of sex without condoms in a cross-sectional study of 3178 men and women in the United Kingdom. ART was associated with much lower prevalence of sex that poses a risk of HIV transmission, according to the study.
Research including HPTN052 and the PARTNER study provides strong evidence that starting ART greatly lowers the risk of HIV transmission to sex partners. Findings like these heighten the importance of understanding how starting ART and perception of viral load affect a person's sexual behavior. Researchers who conducted the UK study, ASTRA, note that the impact of ART on HIV transmission could be undermined if ART use or perceived suppressed viral load lead to more condomless sex (CLS).
To address these issues, ASTRA recruited HIV-diagnosed adults seen at one of eight HIV outpatient clinics in the UK from February 2011 to December 2012. Participants completed a questionnaire about sociodemographic factors and sexual behavior, including questions about current ART use and perceived viral load. The researchers grouped participants into three viral load categories (≤50 copies, >50 copies, don't know) and three ART categories (not on ART, on ART with undetectable self-reported viral load, on ART with detectable self-reported viral load). The questionnaire also established whether participants had condomless sex with an HIV-serodifferent partner (CLS-D) and whether CLS-D participants ran a risk of HIV transmission (CLS-D-HIV-risk), which meant CLS-D and not on ART or CLS-D with a clinic-recorded viral load >50 copies/mL.
The analysis focused on 3178 adults diagnosed with HIV, including 2189 men who have sex with men (MSM), 360 heterosexual men and 629 women. Proportions on ART were 87% of MSM, 93% of heterosexual men and 89% of women. Respective proportions with self-reported undetectable viral load were 84%, 62% and 71%.
Among MSM, 38% reported CLS and 15% reported CLS-D. Among 2163 MSM with self-reported ART/viral load, those on ART were less likely to report CLS (37% versus 48%) and CLS-D (14% versus 19%) in the past three months (P < .05 for both). CLS-D prevalence was significantly higher among MSM with an undetectable self-reported viral load (15% versus 10%, P = .014). Among MSM not on ART, CLS-D prevalence was higher than for both on-ART groups (19%). After statistical adjustment for sociodemographic and HIV-related factors, MSM on ART with detectable self-reported viral load had a significantly lower CLS-D prevalence than MSM on ART with undetectable viral load (adjusted prevalence ratio [aPR] 0.66, 95% confidence interval [CI] 0.45 to 0.95). But MSM not on ART had CLS-D prevalence similar to that of MSM on ART with an undetectable viral load (aPR 1.08, 95% CI 0.78 to 1.49).
Among heterosexual men, 23% reported CLS and 6% reported CLS-D. Among women, 21% reported CLS and 11% CLS-D. Among 351 heterosexual men with self-reported ART/viral load, there was no significant difference between those on ART versus not on ART in prevalence of CLS (24% versus 17%) or CLS-D (6% versus 4%). Similarly, women on ART did not differ significantly from those not on ART in prevalence of CLS (21% versus 24%) or CLS-D (11% versus 9%). Adjusted analysis of heterosexual men and women combined found that self-reported ART/viral load was not associated with CLS-D.
CLS-D-HIV-risk prevalence stood at 3% in all ASTRA participants and was highest among people not on ART (16%), lower for those on ART with detectable self-reported viral load (4%) and lowest for those on ART with undetectable self-reported viral load (0.6%). Among MSM, respective proportions of CLS-D-HIV-risk prevalence were 4%, 19%, 6% and 0.8%. Among heterosexuals, respective rates were 2%, 8%, 2.5% and 0.3%.
These last findings, the authors note, mean prevalence of CLS-D-HIV-risk was much lower in people on ART than not on ART and "extremely low" (below 1%) in people on ART with an undetectable self-reported viral load. They observe that among MSM on ART, CLS-D prevalence was higher among those reporting an undetectable viral load than among those who did not. This finding could mean "that some MSM were choosing not to use a condom with HIV-serodifferent partners because of knowledge of very low infectiousness based on their perceived viral load." Although it appeared that perceived undetectable viral load among MSM influenced their condom use, "any such effect was modest at the time of the study, and would not undermine the effect of early ART on HIV/STI [sexually transmitted infection] transmission."