Iron Status of Kenyan Pregnant Women after Adjusting for Inflammation Using BRINDA Regression Analysis and Other Correction Methods

Serum ferritin concentration is the preferred biomarker to assess population iron status in the absence of inflammation. Interpretation of this biomarker is complicated in populations with a high burden of infection, however, because inflammation increases serum ferritin concentration independently of iron status. We aimed to compare estimates of iron status of Kenyan pregnant women, with circulating ferritin concentrations adjusted for inflammation using newly proposed methods by the BRINDA project, or using previously proposed adjustment methods. We re-analyzed data from pregnant Kenyan women living in a rural area where malaria is highly endemic (n = 470) or in an urban area (n = 402). As proposed by the BRINDA group, we adjusted individual ferritin concentration by internal regression for circulating concentrations of C-reactive protein (CRP) and α1-acid glycoprotein (AGP). Other adjustment methods comprised: (a) arithmetic correction factors based on CRP or AGP; (b) exclusion of subjects with inflammation (CRP >5 mg/L or AGP >1 g/L); and (c) higher ferritin cut-off value (<30 μg/L). We additionally adjusted for Plasmodium infection as appropriate. Lastly, we assessed iron status without adjustment for inflammation. All correction methods increased prevalence of iron deficiency compared to the unadjusted estimates. This increase was more pronounced with the internal regression correction method. The iron deficiency prevalence estimate increased from 53% to 87% in rural Kisumu study and from 30% to 41% in the urban Nairobi study after adjusting for inflammation (CRP and AGP) using the BRINDA internal regression method. When we corrected for both inflammation and Plasmodium infection using the regression correction, it resulted in lower prevalence estimates compared to uninfected women. Application of linear regression methods to adjust circulating ferritin concentration for inflammation leads to markedly decreased point estimates for ferritin concentration and increased estimates for the prevalence of iron deficiency in pregnancy