HIV infection is characterised by a state of widespread immune activation which is thought to be responsible for a significant proportion of the immune deficiency that develops during the natural history of the disease . This immune activation is caused by HIV viral proteins , gut-associated bacterial products which translocate into the circulation acting as toll-like receptor ligands following damage to the gut-associated immune defences soon after HIV infection  and other extrinsic infections that occur . Immune activation is measured in vitro by a number of immune activation markers, including cell surface-associated (e.g. CD38, HLA-DR,CD25, CD69) and soluble molecules (e.g. [[beta].sub.2] microglobulin, neopterin) . Africa has the largest burden of HIV worldwide and is where the majority of HIV-associated morbidity and mortality occurs . Markers of immune activation have been studied in HIV-infected Africans in order to investigate a number of questions including the role of endemic co-infections in pathogenesis of African HIV; the possibility of cheaper and simpler viral-load assays; and the mechanisms of immune reconstitution syndromes.
ACTIVATION, WORMS, OTHER ENDEMIC INFECTIONS AND THE PATHOGENESIS OF HIV IN AFRICA
It has been suggested that Africans suffer a more rapid disease progression than European or North American HIV-infected individuals . One hypothesis for this is that infections endemic to Africa, including malaria, tuberculosis and helminths, produce an even greater state of immune activation in Africans which in turn leads to a faster immune decline [4,8]. Markers of activation have been shown in European and North American populations to be of prognostic significance independent of CD4+ T cell count and viral load .
Bentwich and others described the observation that helminth-infected Africans expressed higher levels of immune activation and a dominant Th2 immune profile when compared to European patients and those without helminth infections. Findings included an eosinophilia, high levels of IgE, high Th2 cytokines e.g. IL-4, IL-5, IL-10 and high levels of CD38 and HLA-DR expression amongst CD4+ and CD8+ T cells [10-12]. They therefore hypothesised that helminth infections had a detrimental effect on HIV-infected Africans by increasing immune activation and attenuating the protective Th1 response. This activation fell in immigrant Africans over time and is reduced with anti-helminth treatment . It was proposed that anti-helminth treatment might reduce the immune decline in HIV-infected Africans before antiretroviral therapy (ART). Subsequent studies looking at this potential intervention have only used surrogate markers of prognosis (e.g. viral load) and have yielded mixed results [14,15].
African-based studies have also investigated the role of tuberculosis and malaria-associated immune activation and HIV pathogenesis. HIV-infected Africans with tuberculosis have more immune activation ([[beta].sub.2] microglobulin, CD38 expression) than those without tuberculosis co-infection  and tuberculosis has been shown to be associated with increased HIV viral load  and reduced survival amongst Africans before ART . Episodes of malaria are associated with elevated levels of inflammatory cytokines and immune activation markers (CD38 and HLA-DR)  and studies have shown that the HIV viral load increases [20,21]. It is unclear as to whether malaria affects HIV survival.
Some studies have investigated whether there is a more direct relationship between activation and survival in African HIV. A Rwandan study investigated whether levels of [[beta].sub.2] microglobulin were predictive of prognosis in HIV-infected women. High levels of [[beta].sub.2] microglobulin were found to be associated with a seven-times-greater risk of death. While this was independent of total lymphocyte count, CD4+ and viral load were not investigated . [[beta].sub.2] microglobulin was also found to be predictive of mortality in a Guinea-Bissau-based study of HIV-2, and this was independent of CD4+ T cell count and viral load. Neopterin, another soluble activation marker, was not found to be significantly associated with prognosis . More recently, a Tanzanian study investigated whether C-reactive protein was associated with mortality and found it to be predictive independently of CD4 and viral load [22,24]. There is some evidence that HIV clade is of prognostic significance in African HIV. A number of studies have suggested that patients with clade D progress faster than those with clade A [25,26]. The reason for this remains unexplained although Bousheri et al. have shown in a Ugandan study that those with clade D virus seem to have greater levels of apoptosis in activated (CD38+ HLA-DR+) CD4+ T cells . Sherman et al., in a South African study of vertically infected children, showed that absolute CD38 on CD8+ T lymphocytes was a predictor of mortality at 1 year, independent of CD4 percentage and viral load . There has been no African study looking at the prognostic value of CD38 expression in adults.
Despite all this evidence that endemic co-infections in African HIV-infected individuals may increase immune activation and therefore adversely affect prognosis, there is little evidence that the prognosis of serconverting Africans is any different to Western populations up until the onset of AIDS . Jaffar et al. reviewed natural history studies performed in Africa and concluded...