Assessment of seasonal and climatic effects on the incidence and species composition of malaria by using GIS methods

Malaria is highly dependent on climate and environmental factors. This thesis incorporates environmental and climatic factors into mathematical and geographic information system (GIS) models in order to assess the feasibility of an early warning system in a strongly seasonally transmitted region in Iran. It also measures Plasmodium spp interactions through meta­ analysis, modelling, and further analysis of a large epidemiological dataset. The first part of the thesis assesses the feasibility of malaria prediction models based on ground climate and remote sensing data. Predicted values were typically extrapolated from the previous month’s data; adding ground climate data can improve these predictions by around ten percent. Predictive variables for these models are readily available in the field, so an improvement of even a few percent makes them feasible. However, more ground climate data are needed for prediction at finer than district spatial scales. The second part of the thesis measures interactions between malaria species. A systematic literature review and meta-analysis assessed the heterogeneity of interaction terms between malaria species. Mathematical models assessed the effects of within-population heterogeneity in infection risks. Finally, data from a large epidemiological study in a highly malaria- endemic area (Garki, West Africa) were analysed cross-sectionally and longitudinally. Random-effect meta-analysis produced a summary OR between P. falciparum and P. vivax of less than one (0.6, 95% Cl 0.46-0.8). The very 2 wide range of ORs seen between studies (0.02 to 10.9) could be explained partly by species prevalence and the temporal span of studies. Mathematical models indicated that within-population heterogeneity in infection risks may, by itself, explain ORs as great as ten or more. Longitudinal analysis of the Garki data produced lower ORs than those from cross-sectional analysis. P. falciparum had suppressive effects on the other species. In addition, Plasmodium spp interactions highly depend on subject age and the temporal and spatial distribution of species. In conclusion, heterogeneity in infection risks, due to heterogeneity either in acquired immunity or in exposure risk, is the most important factor on interactions between Plasmodium spp. Finally, it seems that species-specific models would improve the predictions due to the different impacts of climate on the transmission of species and the interaction between them