"Anaemia affects two-third of all preschool-aged children in Africa. Malnutrition, malaria and helminth infections are among the main factors contributing to anaemia in this age group, however their relative contribution across the continent is not well understood. Spatially explicit estimates of anaemia risk are important measures of child morbidity and mortality.The goal of the project is to develop Bayesian geostatistical methodology for very large, non-stationary data and employ it to (i) determine the relative contribution of malaria, helminth infections and malnutrition on anaemia and severe anaemia burden among pre-school children; (ii) obtain spatially explicit estimates of the risk of anaemia, severe anaemia and number of affected children; (iii) characterize co-endemicity patterns of anaemia, malaria, helminth infection and malnutrition; and (iv) quantify the contribution of severe anaemia to child mortality.This research will contribute novel statistical methodologies in (i) spatial analysis of very large non-stationary geostatistical data, (ii) variable selection within a non-stationary model for very large geostatistical data, (iii) modelling disease co-endemicity from spatially misaligned surveys arising from independent regression models and (iv) meta-analyses of heterogeneous large spatial data by coupling geostatistical with mathematical transmission models.Applications of geostatistical methodology will help optimising interventions to combat anaemia by generating (i) the first anaemia risk map and number of affected preschool children across Africa which will guide efficient allocation of nutrient supplements and fortified foods; (ii) anaemia co-endemicity maps and estimates of the relative contribution of anaemia risk factors to design integrated interventions based on local conditions; and (ii) estimates of the anaemia-related mortality across Africa."