ABSTRACT
Objectives: Ferritin is considered the most suitable indicator of low iron stores in population-based surveys. Ferritin is an acute-phase protein (APP) and studies find this can impact on its use and interpretation. This study explores the relationship between ferritin and other APPs in an effort to identify approaches to account for this relationship in the assessment of iron status in young children. Methods: Using data from the Biomarkers Reflecting Inflammation and Nutrition Determinants of Anemia (BRINDA) project, the association between ferritin and specific APPs, C-reactive protein (CRP) and α1-acid-glyco-protein (AGP), was explored. The prevalence of low-iron stores were compared using approaches to account for inflammation: exclude individuals with inflammation, use a higher ferritin cutoff, internally generated correction factors or correction factors generated via meta-analysis, and regression methods. Results: The dataset includes approximately 30,000 children (6-59 months) and encompasses all 6 WHO geographic regions. Based on preliminary findings from 11 countries, the prevalences of CRP (14-40%) and AGP (21-61%) were high. The Kendall tau correlations between ferritin and CRP or AGP ranged from -0.02 to 0.42. There was considerable variation in the prevalence of unadjusted low iron stores (0.3-32%) and in the percentage change induced by adjusting for CRP and AGP (0-82%) Conclusions: Ignoring inflammation generally resulted in underestimates of low iron stores in young children. These findings should inform efforts to develop a harmonized approach to the use of ferritin as a biomarker of iron status, aiding in the interpretation of the effectiveness of interventions.