INTERACTIONS BETWEEN MALARIA AND IRON OR VITAMIN A STATUS IN RURAL ZAMBIAN CHILDREN

Background and objectives Evidence regarding the bidirectional interactions between iron or vitamin A status and malaria remains inconclusive. We assessed the longitudinal associations between iron (defined by ferritin) or vitamin A (defined by retinol) and malaria, assessed 6 months later by microscopy. Additionally, the changes in ferritin, soluble transferrin receptor (sTfR) and retinol during malaria and/or inflammation (AGP >1 g/L) were estimated. Design We included 1024 children, 4-8 years old from Mkushi District, Zambia, participating in a 6-month trial, designed to evaluate the efficacy of a provitamin A intervention. We analyzed baseline (August, 2013) and endline (March, 2014) survey data, collected in the low and high malaria transmission seasons respectively. We estimated incidence rate ratios (IRR) comparing endline malaria risk across baseline iron or vitamin A status using, modified Poisson regression, controlling for inflammatory changes in the indicators of iron and vitamin A at baseline. Results Inflammation alone was associated with changes of up to 66% in ferritin, up to 12% in sTfR, and up to -13% in serum retinol relative to their respective reference groups. Malaria with inflammation was associated with changes of up to 280% in ferritin, up to 40% in sTfR and up to -36% in retinol. After controlling for baseline inflammation, the IRR, comparing the low (0.7-1.05 µmol/L) and adequate (>1.05 µmol/L) vitamin A groups to the deficient group (75 µg/L) ferritin groups, relative to the deficient group (