Experts have long assumed that the violence, wide-scale rape and refugee crises are the inevitable by-products of war that fuel HIV/AIDS epidemics, but an analysis of HIV prevalence surveys from seven sub-Saharan African countries with similar recent histories found no evidence that higher HIV infection rates accompany conflict.

The study over the past five years in Democratic Republic of Congo (DRC), southern Sudan, Rwanda, Uganda, Sierra Leone, Somalia and Burundi, was published on Friday in The Lancet medical journal.

In conflict-affected parts of Burundi, Rwanda and Uganda, HIV prevalence actually decreased at similar rates as in unaffected areas. Contrary to the stigmatising belief that refugees from war zones bring HIV to their host countries, nine of the 12 refugee camps surveyed had lower rates of HIV infection than the surrounding populations; only one camp had a higher HIV prevalence.

Even the use of mass rape as a weapon of war in Rwanda, Sierra Leone and DRC did not appear to have raised their overall HIV infection rates.

Lead researcher Dr Paul Spiegel, a senior HIV/AIDS technical officer with the United Nations High Commission for Refugees (UNHCR), said previous studies claiming that conflict exacerbated HIV infections were probably based on poor survey methods and biased interpretation of data.

"It's very natural to think that HIV is going to increase, so I think people jumped to conclusions. The hardest part is looking at the individuals and then looking at the broad epidemiological affects," he told IRIN/PlusNews.

He added that conventional wisdom did not take into account some of war's protective effects against the spread of HIV, such as the difficulty of moving between rural areas and higher prevalence urban areas, and the breakdown of commercial sex networks.

Reacting to the Lancet study, Gopa Kumar Nair, HIV and AIDS advisor to Save the Children, a UK-based non-governmental organisation that assists children in need, stood by the findings of a 2002 report by his organisation, which argued that conflict could increase young people's vulnerability to HIV.

"These are not just interpretations of data, these are based on experiences in the field," he said. "We're seeing orphans, violence, protection systems failing, and no access to HIV/AIDS information and services, and we're sure these are conditions that would facilitate HIV transmission."

Kumar Nair expressed concern that the Lancet study's conclusions were based on scarce data drawn from a handful of countries. "Our experience is that getting [HIV prevalence] data from conflict-affected populations is a nightmare," he said.

Acknowledging the general lack of reliable HIV prevalence figures from countries experiencing armed combat situations, Spiegel said there was enough data from the seven countries surveyed for the study to draw conclusions, "but we have to be careful not to extrapolate that to every situation."

What was not in doubt, commented Dr Tom Ellman, HIV advisor to medical humanitarian organisation, Medecins Sans Frontieres, was the lack of care and support available to people affected by HIV/AIDS in circumstances of war.

"If we are talking about the gap between need and provision, the gap is likely to be greater in conflict-affected settings because they're neglected by international donor agencies - they're often the last people to get aid," he said.

"These are extremely poor settings and they need more help and support," agreed Kumar Nair. "I'm a bit concerned donors might think HIV is not a problem, which would be a disaster."

Spiegel emphasised that the study's findings should motivate HIV prevention efforts in such situations. "We need to provide these HIV interventions during conflicts, because if we can ensure people are educated about risk factors immediately post-conflict, it'll be easier for us to deal with."