This is from Medpundit which has a link to the entire 14-page report.
Some of the highlights of the study:
HIV and STDs: According to the authors' data, African HIV did not follow the pattern of sexually transmitted disease (STD). In Zimbabwe in the 1990s HIV increased by 12% a year, while overall STDs declined by 25% and condom use actually increased among high-risk groups.
Infection rate: HIV spread very fast in many countries in Africa. For the increase to have been all via heterosexual sex, the study claims, it would have to be as easy to get HIV from sex as from a blood transfusion. In fact, HIV is much more difficult than most STDs to transmit via penile-vaginal sex.
Risky sex? Several general behaviour surveys suggest that sexual activity in Africa is not much different from that in North America and Europe. In fact, places with the highest level of risky sexual behaviour, such as Yaounde in Cameroon, have low and stable rates of HIV infection. "Information from the general population shows most HIV in sexually less active adults" .
Children and injections: Many studies report young children infected with HIV with mothers who are not infected. One study in Kinshasa kept track of the injections given to infants under two. In one study, nearly 40% of HIV+ infants had mothers who tested negative. These children averaged 44 injections in their lifetimes compared with only 23 for uninfected children.
Good access to medical care: Countries like Zimbabwe, with the best access to medical care, have the highest rates of HIV transmission. "High rates [of HIV] in South Africa have paralleled aggressive efforts to deliver health care to rural populations".
In the poorest countries people have little or no access to injections and thus are not exposed to dirty needles. In the "richest" countries (actually, those which are less poor) there is a push to deliver health care, much of it by poorly trained non-professionals. As far as HIV is concerned no injections is preferable to injections with dirty needles.
For what it is worth, I find the incidence of HIV in children of uninfected mothers very convincing.