The First Model-Based Geostatistical Map of Anaemia

Abstract

Anaemia is the most common blooddisorder globally, is a major underlyingcause of death among children andpregnant women in the developing world,and because it is most prevalent in thepoorest communities, represents a usefulmarker of social and economic marginal-ization [1–3]. The aetiology of anaemia isoften multi-factorial including nutritionaldeficiencies, parasitic and inflammatorydiseases, haemorrhage, and genetic defectsin the molecular structure of the haemo-globin (Hb) [3]. Many of these risk factorsfor anaemia co-exist in communities andaffect individuals in composite ways notadequately understood. In Africa threedominant contributors to anaemia inyoung children are malaria, helminthinfections, and iron-deficient diets. Todate, only national-level estimates ofanaemia burden developed by the WorldHealth Organization are available toguide control [3]. There are limited dataon within-country variations and no de-tailed information on the uncertaintyaround these estimates or on the combi-nation of causal factors important to acountry.The article published in PLoS Medicinethis week by Magalha˜es and Clements [4]explores the use of model-based geostatis-tics to investigate the risks of anaemia andmean Hb concentrations among pre-school-age children (aged 1–4 years)attributable to malnutrition, malaria, andhelminth infections in three West Africancountries (Burkina Faso, Ghana, andMali). The authors develop high resolutionmaps of the geographic distribution ofanaemia accounting for these factors andcompute the estimated number of anaemicchildren. The advantage of using model-based geostatistics to model anaemia isthat it allows the use of the survey datafrom a sample of locations to predictcontinuous surfaces of risk, informed byenvironmental and demographic covari-ates, but in a way that takes into accountthe sample size and spatial characteristic ofthe data to robustly assess uncertainty inthe modelled outputs. The study identifiedmalnutrition as the main driver of anae-mia, followed by malaria and helminthinfections, accounting for a populationattributable fraction (PAF) of almost 60%of all anaemia. The mean Hb concentra-tion increased with age and varied bycountry. The continuous surfaces of anae-mia prevalence and Hb concentrationshow interesting variations within thenational borders that allow for the com-putation of the location and radius of fociof high risks. The estimated case burdenfor 2011 was similar to that provided bythe World Health Organization in its1993–2005 worldwide anaemia preva-lence report [3].