JOHANNESBURG, 6 July 2010 (PlusNews) - A new study has challenged widely held assumptions about income level in relation to HIV, finding that neither wealth nor poverty are reliable predictors of HIV infection in Africa.

Previously, the argument that poverty drove HIV epidemics was supported by the World Bank and UNAIDS, as well as less reliable authorities like former South African President Thabo Mbeki, who told the International AIDS Conference in Durban in 2000 that the disease was a partner with "poverty, suffering, social disadvantage and inequity".

More recent research suggests that the reality is far more complex. For example, Botswana and South Africa, described as two of the wealthiest countries on the continent, also have among the highest rates of HIV infection.

Nevertheless, the idea that poverty fuels the spread of HIV has persisted as "a very dominant narrative", according to Justin Parkhurst of the London School of Hygiene and Tropical Medicine.

Parkhurst analyzed and compared data on HIV and wealth from demographic and health surveys in 12 sub-Saharan African countries with generalized epidemics (national prevalence rates higher than 1 percent); his findings are published in the July issue of the Bulletin of the World Health Organization.

He noted that in lower-income countries HIV prevalence tended to rise in tandem with wealth - in Uganda and Cote d'Ivoire, for example, women in the highest income bracket had the highest HIV prevalence.

In countries with a per capita gross domestic product higher than US$2,000, the link between wealth and prevalence was less clear.

Parkhurst also found that the relationship between wealth and HIV changed over time. A survey was conducted In Tanzania in 2003, and another in 2008; in the intervening five-year period, HIV prevalence declined among women in higher income brackets and rose among those in the lower income groups. Among men, prevalence stayed the same in the poorest group but was lower in all other groups, with the biggest declines in the highest income groups.

"HIV spreads through sexual behaviours, and these are social behaviours that change over time and are responsive to outside influences," Parkhurst told IRIN/PlusNews. He compared the way HIV affected different social groups with the way tobacco use and obesity once affected mainly the rich, but were now bigger problems among the poor.

Wealthier people were often harder hit early in an HIV epidemic, probably because of their broader social and sexual networks. "Over time, the wealthy tend to be more educated [about HIV risk] and more likely to think about their future health," said Parkhurst.

However, these trends are by no means universal and the patterns for men and women differ. In Swaziland, for example, which has the highest HIV prevalence of all the countries Parkhurst looked at, there was little evidence of a link between household wealth and individual prevalence.

Know your epidemic

Parkhurst's findings have implications for one-size-fits-all prevention campaigns that do not take into account the complex and changing ways in which wealth, education level and gender can affect risk-taking behaviours.

"We need to educate people [about HIV] in a way that's relevant to their context," he said. "It's about letting local actors to find out what's going to work best. If we try to work out the solution from London ... it's unlikely to work."

Parkhurst said "bottom-up" HIV prevention initiatives targeting the specific lifestyles and risk behaviours of a community were more likely to work. This approach is already catching on, with UNAIDS urging countries to "know your epidemic" and design prevention programmes accordingly.

"Health practitioners know they have to diagnose a problem before they can treat it," he said. "I think the international community is starting to recognize the importance of addressing structural drivers of HIV, not just broadly, but to look at the specifics for specific communities."