A person living with HIV will transmit the virus, on average, once out of every 900 acts of unprotected heterosexual sex. At least, this is the estimate based on a study of mixed-status couples in Africa, published on Jan. 11 in the Journal of Infectious Diseases.
The exact risk of HIV transmission during a specific activity has always been a difficult number to pin down, and some experts were quick to note that the "1 in 900" figure should not be regarded as carved in stone. "The study provides a number that people may quote as to the risk of infection per episode of intercourse, but this ranged considerably," said David Wohl, M.D., one of the experts in TheBody.com's Safe Sex and HIV Prevention Forum. He noted that the risk "was influenced by the viral load of the infected person, the circumcision status, and the presence of STIs [sexually transmitted infections] in the uninfected partner."
Researchers from the University of Washington reached their "1 in 900" conclusion after analyzing data from 3,297 mixed-status couples enrolled in the study, which followed them for a maximum of 2 years. At enrollment, the participants were tested for STIs; men were examined to determine circumcision status; and the negative partners were tested for genital herpes (HSV-2). They were also given HIV prevention counseling and free condoms.
Each month, the positive participants self-reported the number of times they had sex with their partners and whether or not they used a condom. In addition, their HIV viral loads were measured at 0, 3, 6, 12 and 24 months.
The negative partners were tested for HIV every three months. If positive, the researchers used genetic testing to confirm whether or not the negative partners contracted HIV from their study partner.
At the end of the two-year study period, 86 of the negative partners analyzed tested positive.
Viral load played a key role in predicting HIV transmission. The higher the viral load, the higher the chances of transmission. The researchers estimated that with every log10 increase of viral load (e.g. going from 1,000 copies/ml to 10,000 copies/ml, or from 10,000 copies/ml to 100,000 copies/ml), the per-act risk of transmission increased by 2.9 times. This reaffirmed previous studies, which found lower viral loads decreased transmission.
In the study, men were found to be about twice as likely to transmit HIV to women, but the researchers stated this may mostly be due to higher viral loads in the male participants.
The study found that male circumcision reduced transmission by 47 percent, while condoms reduced transmission by 78 percent, also reaffirming previous studies.
"This study helps to confirm the role that several key, modifiable factors play in HIV transmission, especially viral load and condom use," lead researcher James P. Hughes, Ph.D., told TheBody.com. "None of the factors we identified are really new -- others have raised their role in HIV transmission before -- but this study offers greater precision in determining the effects."
When asked if the study suggested lower effectiveness for condoms than we're generally used to, Hughes stated, "Unfortunately, the data by itself can't answer that question, so I can only speculate. However, my very strong bias is that the true effectiveness of condoms is much higher than 78%, and the apparent reduced effectiveness we are seeing is largely due to self-reporting discrepancies."
The findings that an HIV-positive person's viral load affects his or her HIV transmission risk appear to support the now-famous "treatment as prevention" study findings released last year, but Dr. Hughes urged caution when interpreting the results. "While these results, along with the recently published HPTN 052 results, do suggest that transmission of HIV can be significantly reduced by lowering viral load with antiretroviral treatment, that isn't the only consideration in starting treatment early," he noted. "The START study will consider the effect of starting treatment early on the HIV-infected person's health."
Similarly, the results don't suggest that an undetectable viral load makes unprotected sex OK. "I hope that this study does not lead people with a low or undetectable viral load to stop using condoms," Dr. Hughes said. "Even if an individual's viral load is undetectable at the time of measurement, it is possible to have transient increases in viral load, and this could lead to transmission if a condom is not used."
Dr. Wohl added a warning against applying these study results to same-sex couples or people who have anal sex. "This study looked at heterosexual couples in Africa. It is a leap to think about this in terms of the risk with anal sex, which was likely uncommon in this study cohort," he said. "Overall, these results will be most useful for us to talk more about the need to reduce people's viral load in case they do not use condoms, which worked great to prevent infection, but are hardly loved."
Lisa Hightow-Weidman, M.D., another expert in TheBody.com's forum on safe sex and HIV prevention, concurred. "I encourage all my patients to use condoms, but I also talk with them about this research," she said. "If they decide, within the context of a mutually monogamous, serodiscordant relationship in which the pos partner is on meds and has been undetectable for a while (I encourage at least 6 months), to not use condoms, I give them the facts. Risk is low -- not zero, but low."
Warren Tong is the research editor for TheBody.com and TheBodyPRO.com.
Follow Warren on Twitter: @WarrenAtTheBody.
Copyright © 2012 Remedy Health Media, LLC. All rights reserved.