Tuesday, February 7, 2012

Another Look At Sexual (and Non-Sexual?) HIV Risk

A recently published article finds that, out of 1834 African participants who belong to various groups thought to face high risk of being infected with HIV, incidence (the rate of new infections per year) was highest among men who had sex with men (MSM). In Western countries, MSM often have the highest rates of HIV transmission, followed by intravenous drug users (IDU). But because such a huge proportion of transmission in African countries occurs among heterosexuals who don't fall into any of the above risk groups, MSM generally account for a relatively small percentage of those infected with HIV there, as do IDUs.

It was also found that "[p]aying for sex was inversely associated with HIV infection". In Western countries, sex workers are not especially likely to be infected unless they are also IDUs. But in African countries extraordinary prevalence rates have been reported for this group, even where they are not reported to have other risks. Prevalence rates of 70 or 80% have been reported, prompting some to wonder if the women might have faced one or more non-sexual risk not examined by those doing the reporting. The alarmingly high rates reported in the 1980s and 1990s dropped rapidly, often long before the country involved made any attempts to reduce infection rates; but it's good to hear that the risks this group faces may be receding.

The study aimed to identify risk populations for HIV prevention trials. If those taking part in the trial do not face much risk it will be difficult for a trial to show an effect. But the results should also be of interest to those whose job consists of making claims such as the one about 80-90% of HIV transmission in African countries resulting from heterosexual sex. Where HIV appears to be transmitted rapidly and heterosexual sex is found to be the only significant risk, there is then a problem of explaining what is so risky about heterosexual sex between Africans, something that has eluded those working in the field so far. Some of the elusiveness may stem from the fact that heterosexual sex was assumed, rather than found to be the only risk.

One of the most surprising findings was that no HIV infections were found in over 300 women in Nairobi who identified themselves as sex workers. The authors speculate that this may be due to condom use, fewer HIV positive clients, more clients on antiretroviral drugs, etc. But hopefully the finding will prompt some reanalysis of some of the eyewatering claims made about numbers of clients per day and other speculation about unsafe sex mentioned (ad nauseum) in the past. Much of it seemed to be created to fit the rates of HIV infection found rather than to investigate if such behavior really existed or if it was common enough in high HIV prevalence areas to explain transmission rates; little effort seems to have been made to establish if sex workers faced other, non-sexual HIV risks.

The finding that pregnancy rates were higher than expected may suggest that women, even sex workers, were not using condoms particularly consistently. Extremely high rates of sexually transmitted infections in one area, and fairly high in the others, also suggest that condom use campaigns may not yet have had much impact among some high risk groups (and those thought to be high risk). If the inverse association between paying for sex and HIV prevalence among MSM is an indication that condom campaigns can work when properly targeted, that is certainly an important finding, as is the one that MSM perceive anal sex with men to be lower risk than sex with women. But some well publicized claims about female sexual behavior may have supported that perception.

One of the worrying things about this paper is that the participants are drawn from groups thought to be at high risk. However, the bulk of HIV transmission in countries where modes of transmission studies have been carried out occurs among groups of people who are not thought to be at high risk. In other words, this research excludes most of the HIV positive population in countries like Uganda and Kenya, where relatively low risk sex is said to account for over 60% of the total.

Finally, my attention has been drawn to the appendix, which I am currently unable to access, which shows that those who have received an injection in the last three months are nearly five times more likely to be infected with HIV. Those involved in clinical trials would appear to have very good reason to start looking more closely at non-sexual risks.


No comments: