Children living in areas with endemic malaria are usually assumed to have malaria if they are hospitalized with severe anaemia that is preceded by an acute febrile illness. Such patients often have undetectable malaria parasites by microscopy of Giemsa-stained blood films, but their leukocytes commonly contain detectable malaria pigment . It has previously been shown that the diagnosis may be aided by the use of rapid antigen detection tests, which can distinguish patients with malarial anaemia from those with anaemia from other causes , and this was supported in the present study. Thus, more than 90% of the patients with severe anaemia had detectable parasite antigens compared to only one third of those with normal or near normal Hb. Although microscopy is usually considered the gold standard for the diagnosis of malaria, these results suggest that the antigen detection methods, in particular those based on the detection of pLDH, may be superior to microscopy in the diagnosis of malarial anaemia.
In the present study, patients with detectable pLDH, a sign of persistent parasitaemia, were distinguished from those in whom only HRP2 could be detected as a sign of a cleared, recent infection. Only patients with persistent parasitaemia had signs of bone marrow suppression. This corresponds with previous investigations in which release of immature erythrocytes from the bone marrow coincided with the time of parasite clearance [1, 6] and with findings in experimental malaria . In one such study, subpatent P. falciparum infection in vaccinated Aotus monkeys led to severe bone marrow suppression that was rapidly reversed in response to mefloquine induced parasite clearance .
The association between P. falciparum infection and suppression of erythropoiesis has been debated. Menendez et al. found that levels of sTfR were increased in infants with malaria . Similarly, Verhoef found increased levels of sTfR in children with malaria and concluded that there was no suppression of erythropoiesis . However, in these studies it was not indicated if levels of sTfR were increased adequately compared to the degree of anaemia. Furthermore, increased levels of sTfR found in malaria patients might be caused by shedding of receptors from proliferating B-lymphocytes . In the present study, however, there was no correlation between sTfR and lymphocyte count (data not shown). Due to the logistics of the study it was not possible to make reticulocyte counts but the finding that persistent, subpatent parasitaemia was associated with reduced levels of both RDW and sTfR strongly points toward parasite-induced suppression of erythropoiesis. This may explain the fact that healthy school children with so-called asymptomatic P. falciparum infection have reduced Hb and signs of bone marrow suppression . Asymptomatic P. falciparum infections are usually accepted as a necessary evil in order to maintain immunity in individuals, who are at great and constant risk of malarial infections , and this approach is probably the only realistic option in the near future. However, data from the present study imply that the ultimate goal of malaria control ought to be complete parasite eradication due to the detrimental effects of persistent parasitaemia. This is consistent with a previous clinical trial that linked incomplete haematological recovery with lack of parasitological cure . The Hb levels in infants living with high risk of malaria can be improved by impregnated bednets , malaria chemoprophylaxis  and presumptive intermittent treatment . It is likely that this effect is mediated partly by a reduction in the risk of bone marrow suppression. From a clinical point of view this study suggests that in malarial endemic areas, antimalarial treatment should not be withheld from severely anaemic patients presenting with signs compatible with ongoing or recent malaria who have not been treated for malaria, even when parasites are undetectable by microscopy. Controlled clinical trials should be performed to determine whether patients who have already received a full course of antimalarial treatment would benefit from a repeat course, and in particular, if it is important to restrict this treatment to children with a positive pLDH test. Paediatric patients that return with repeated episodes of anaemia are common (ref. 7) and the management of these patients puts strain on the health system. Thus, the cost of improved diagnosis and management of these cases have the potential to benefit both the patients and the health system.
Several studies have shown that the suppressed bone marrow response to anaemia, which is a general feature of malaria, is not caused by insufficient secretion of EPO [1, 25, 26], but this is disputed by other investigations [27–29]. In agreement with the former view, similar levels of EPO in response to low Hb were found in patients with and without malaria. It has been suggested that the bone marrow suppression is a direct effect of TNF-α, which is elevated in malaria . On the other hand, other investigations did not find that inflammatory cytokines to played a role in malarial dyserythropoiesis . In addition, severe malarial anaemia is associated with relatively low TNF-α levels  and signs of systemic inflammation [33, 34] (Awandare et al. unpublished data) compared with cerebral malaria and uncomplicated malaria. It is thus possible that the effect of cytokine perturbations in malarial anaemia is a lack of parasite control that leads to persistence of the infection and thus indirectly causes bone marrow suppression . It has recently been proposed that phagocytosed haemozoin may play a role in the dyserythropoiesis of malaria through induction of 4-hydroxynonenal .