Contrary to earlier report by Lehmann and Rapper [7], our results show that, the prevalence of the AS carrier state was highest in Eastern Uganda, followed by Bundibugyo (see Figure 1-map of Uganda, and Figure 2 for distribution of results). The fact that the first survey on sickle cell disease (SCD) in Uganda was done over 60 years ago, and no similar surveys have been conducted since then; it is not surprising that SCD has changed its dynamics. This is because the inherited nature of the SCD trait predicts likely changes in the prevalence and distribution of the same depending on the prevailing biological (malaria interventions) and social factors (marriage patterns). Indeed, contrary to an earlier report of 45% prevalence of the AS trait in Bundibugyo, we found the current AS prevalence to be 13.4%; changes that we have appropriated to the possible earlier adoption of intermarriage-avoidance in Bundibugyo resulting from a similarly earlier knowledge of AS prevalence [7]. Specifically, although the difference in the prevalence of AS between Eastern Uganda and Bundibugyo in the west was statistically insignificant, the prevalence of AS in Bundibygyo (13.4%) was found to be much lower than estimated by Lehmann and Raper (45%), see Tables 1 &2 for details. Several hypotheses may be used to explain this finding. The same should, however, be cautiously employed here since huge heterogeneity within the spatial distribution of genetic factors have been found in many studies (even over relatively very small areas) and the small number of subjects sampled in our study makes such comparisons inaccurate. First, basing on the first survey of 1949, which was 60 years ago, the Baamba was one of the exclusively preserved tribes in Uganda which practiced high level of consanguinity. However, due to the movement of the people, the Baamba may have intermarried into other tribes leading to sickle cell gene admixture or dilution. Secondly, though it is possible that improved malaria interventions may have selected for increased numbers of haemoglobin AA and decreased numbers of AS [15], it is unlikely, because the heterozygous HbAS trait has been observed to have beneficial effects in protecting against severe forms P. falciparum infection. On the other hand the prevalence of SCT in Mbale/Sironko was slightly lower than the 20-28% estimated by Lehmann) and Rapper [7]. It is possible that due to improved malaria intervention programmes in Uganda, some of the people with AS may have been denied resistance against malaria infection as already explained above.
A number of notable difference between our and Lehmann and Raper's study [7] are observable. For instance, while our finding of 3% AS in Mbarara/Ntungamo was similar to the 1-5% reported by Lehmann, it is clear that the prevalence of AS in these districts had remained low. The most likely reason for this observation was as cited above. Improved malaria interventions could have denied persons with AS protection against malaria thus keeping their numbers low [15]. Secondly, it is possible that the level of intermarriage between these communities and other tribes could still be very low and as a result, emergence of new cases of sickle cell disease due to gene admixture may have been curtailed. On the other hand, in as much as the prevalence of SCT in Bundibugyo was lower than 45% reported by Lehmann, it is possible that the prevalence of sickle cell trait among Baamba adult population is still high. This hypothesis is supported by the fact that the study found high prevalence of SS among the children which could be reflective of a high prevalence of AS in the adult population. This finding is further supported by the fact that our study sampled children while Lehmann study sampled adults. The observed prevalence of SS in all these districts was actually much lower than expected [16]. In absence of cost effective screening interventions [17] and therapeutic measures for sickle cell disease within this setting such as bone marrow transplantation and gene therapy [18], the obvious question to ask regards what may have happened to the missing SS progeny. It is possible that many of these children could have succumbed to the disease before celebrating their fifth birthday because of the absence of comprehensive sickle cell screening and management programmes in these districts. Serjeant and Ndugwa alluded to this in their advocacy paper [9]. This hypothesis appeared to have been supported by the fact that all the children who were detected with SS in Mbale/Sironko in Eastern and Bundibugyo in the West were less than 4 years old; suggesting that ss cases are hardly detectable at 5 years since many of these children could possibly have died before their fifth birthday. This may probably further explains why the data on median survival of persons with sickle cell anemia in many developing countries including Uganda is scarce. The fact that no children with SS were detected in Mbarara/Ntungamo does not mean that these children do not exist. The probable reason for this observation was that the number of AS persons may have been kept further low in these districts by natural law of selection as new effective malaria intervention progammes have been put in place This, therefore, could have in turn kept the number of SS low and therefore influencing the detectability of SS cases. Another reason for this could be that the delectability power (precision) used was probably not sensitive enough such that, if this study was to be repeated today using a more sensitive power, children with SS would have been detected in Mbarara/Ntungamo and more children with SS would have been detected in Bundibugyo in the West and Mbale/Sironko in the East. Lastly, there are several shortcomings in our study. First, although we used Hb electrophoresis as our 'gold standard' because of its affordability and availability, better gold standards such as automated capillary Hb electrophoresis are preferred. (Sebia Parc Technlonogique, Leonard de Vinci. CP 8010 Lisses-91008. EVRY Cedex-France). Secondly, in as much as the recruitment of the study participants was randomized, the selection of the study districts was based on convenience sampling and did not therefore give equal chance to the rest of the districts in Eastern and Western Uganda to be represented. So the current results cannot be generalized as representative of the whole Eastern or Western Uganda. Thirdly, as observed above (results section), the statistical difference in the prevalence of AS between Mbarara/Ntungamo in the West and both Mbale/Sironko in the East and Bundibugyo in the West was highly significant (<0.001); and this may affect our study versus real figures on ground.