HIV in Lesotho: The Unpopular Questions Are in Most Need of Answers

Yesterday I blogged about HIV in Lesotho and the finding that clusters of high HIV prevalence for women and clusters of high HIV prevalence for men do not overlap very much. I wrote to the corresponding author of the finding, Myong-Hyun Go, to ask if the extent to which HIV is transmitted sexually might thereby be in question. His reply was a definite no, but without any argument or evidence.

And that's pretty much how the HIV industry works. The view that HIV is mainly transmitted heterosexually in African countries is sometimes called the 'behavioral paradigm'. And it is not really amenable to evidence, for or against. It is, let's be frank, a prejudice. So it is worth looking a bit more closely at some figures for HIV in Lesotho.

The study in question uses data from the 2004 Lesotho Demographic and Health Survey (DHS), but as the 2009 survey is available, I'll be referring to it instead. After all, the authors expect their work to influence health policy, so it would need to apply equally to both early and later data.

In common with many other African countries, women usually report very low levels of multiple sex partners, with less than 10% having more than one partner in the previous 12 months and a mean number of lifetime partners never exceeding 3. And while HIV prevalence differs a lot spatially, levels of multiple sex partners and mean number of lifetime partners does not.

Men usually report far higher levels of multiple partnership, double or even triple the levels for women. Their mean number of sexual partners is also up to three times higher. Such differences are equally pronounced among young people. Yet far more women than men are HIV positive in most age groups and in most areas. This is also the case in other African countries.

HIV prevalence by sexual behavior figures show that orthodox thinking about age at first intercourse is in need of review. For women, those who first have sex at less than 16 years old have the same HIV prevalence as those who start at 20 years or over. For men, prevalence actually rises with age, with those starting early having lower prevalence. Prevalence for men and women who use condoms is about the same as it is for those who don't.

But the most remarkable thing about the figures is not that HIV prevalence does or does not correlate with sexual behavior; in some respects it does and in some it does not. What is remarkable is that rates of sexual behavior considered to be unsafe are no higher than other countries with far lower rates of HIV, countries in Africa and elsewhere. People in Lesotho are at higher risk than almost anywhere else in the world, but no research has shown why this is so.

The study authors also claim that their findings provide a 'concrete basis' for male circumcision. Well, there are countries where male circumcision status may be correlated with HIV prevalence. In some countries, circumcision appears to be protective, for example, Kenya. But Lesotho is one of the countries where the opposite may be the case. Of course, these correlations may point to something else entirely.

About half of all men in Lesotho are circumcised. But while uncircumcised men are a bit more likely to be HIV positive in some age groups, overall, circumcised men are far more likely to be HIV positive. It could be argued that many circumcised men were infected before being circumcised, but there is then the problem of explaining how they were infected with a virus that can have as little as a 1 in one thousand chance of being transmitted during heterosexual intercourse.

I would argue that both the spatial data created by Myong-Hyun Go and Co and the DHS data show that Lesotho has extraordinarily high HIV prevalence figures and both sets of data suggest that a substantial proportion of HIV may be transmitted non-sexually. But the disturbing thing about each set of data is that neither really shed light on the possible extent of non-sexual and sexual transmission. Without knowing this, the way forward is no clearer than it was before the data was collected.