Friday, December 5, 2014
In a paper entitled 'Religious and Cultural Traits in HIV/AIDS Epidemics in Sub-Saharan Africa', the authors conclude that the Islamic faith is protective against HIV. Their conclusions about the role of colonial powers is not quite so clear, except to the extent that former British colonies (FBC) tend to be predominantly Protestant (or non-Catholic) and most of the countries that are predominantly Catholic are former non-British colonies (FNBC).
Making associations between HIV and religion, high prevalence and Christianity, low prevalence and Islam, high prevalence and FBCs, lower prevalence and FNBCs, etc, are very tempting. All the predominantly Muslim countries in Africa have low HIV prevalence, with Guinea-Bissau (3.9%) being the only one with a figure higher than 2% (and it is only 45% Muslim). Prevalence in countries with 90% or more Muslims only reaches a high of 1.1% in Sudan.
All the countries with prevalence above 4% are predominantly Christian; out of these, only four are FNBCs. There are nine countries with over 1 million people living with HIV. Only one is an FNBC (Mozambique) and only one is roughly evenly split into Christians and Muslims (Nigeria). All the highest prevalence figures are in the Christian dominated Southern region, and the four with prevalence below .4% are in the predominantly Muslim North.
But things come apart a bit when you look at countries that are Christian, but not predominantly Protestant. There are six predominantly Catholic countries, all FNBCs, where the highest prevalence figure is 2.9%; all these countries are in Central Africa. Yet, a number of countries made up of between 20% and almost 50% Catholic populations have some of the highest prevalence figures, too.
While Muslims and Catholics (ostensibly) oppose extra-marital sex, homosexuality and various other phenomena, so do Protestants and other non-Catholic Christian churches. Suggesting that such opposition is stronger or more active in countries with lower HIV prevalence risks arguing in a circle.
Some useful generalizations can be made, such as very high prevalence in Southern Africa, very low prevalence in North Africa, mainly low prevalence in West and Central Africa and high prevalence in East Africa. It is also broadly true that most predominantly Christian FBCs are Protestant dominated, rather than Catholic dominated. With the exception of Mozambique, prevalence in all FNBCs is never higher than 5%; but these countries can be predominantly Muslim, Christian, mixed, or Catholic.
There are two major objections to the analysis given or implied in this paper. The first is is that patterns and generalizations that can be made at the regional level, or even at the country level, do not always hold within countries; the second objection is to the assumption that HIV is almost always sexually transmitted.
The authors find some broad correlations but they do not discuss causality. They claim that the populations of countries such as Egypt, Tunisia and Algeria, for example, were protected from HIV because of their Muslim faith and the practices that go with that. But those countries, and others in the North, might have been 'protected' by one of the largest desert areas in the world, the Sahara.
In addition, HIV in those countries is mainly subtype B, which is generally associated with male to male sex (and to a lesser extent injected drug use). Subtype B is rare in other parts of Africa, with the exception of South Africa (where it mainly seems to infect men who have sex with men). HIV epidemics appear to form different patterns across regions and countries. But it also forms different patterns within countries.
High HIV prevalence in the Southern region may be facilitated, to some extent at least, by the well developed infrastructure there, infrastructure that would have been built by the British Colonial power. The same colonial power built far fewer roads or other infrastructure in East Africa, and none at all in Central Africa, where they had very little control.
However, they had control of a number of West African countries, where there is generally a strong infrastructure. Why did HIV not spread around West Africa to the extent it did in Southern Africa? Well developed infrastructure may partly explain variation in HIV prevalence between some countries and some regions, but it doesn't explain enough. There are clearly factors operating within each country that account for some variation in HIV prevalence.
Regarding the second objection, the authors link the Muslim faith with certain moral precepts which they feel protect people from HIV. However, the majority of people in non-Muslim countries were not infected because they engaged in 'immoral' behavior. Even 'official' figures show that the bulk of people infected in many high prevalence countries have only one sexual partner, and most of those partners are HIV negative.
The 'promiscuous African' stereotype can not be used to explain HIV transmission because it is a prejudice, not an empirical fact about people with HIV, or about people from countries with high HIV prevalence. But similarly, the 'non-promiscuous Muslim' is also a stereotype, however positive. If you can not discern a person's sexual behavior from their HIV status, nor discern a person's HIV status from their sexual behavior, the conclusion that being a Muslim is protective against HIV is unwarranted.
Religion and former colonial power may be two important influences in HIV epidemics, but the authors fail to show convincingly how they operate on HIV transmission. Arguing that those and all other relevant factors relate exclusively to indivicual sexual behavior fails to explain the spread of HIV within countries. Heterogeneity between and within African countries suggests that HIV prevalence is not all about sex, and that not all factors operate at the individual level.