McSorley ST, Tham A, Jones I, Talwar D, McMillan DC. Regression Correction Equation to Adjust Serum Iron and Ferritin Concentrations Based on C-Reactive Protein and Albumin in Patients Receiving Primary and Secondary Care.
J Nutr 2019;
149:877-883. [PMID:
31050746 DOI:
10.1093/jn/nxz008]
[Citation(s) in RCA: 5] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/14/2018] [Revised: 10/11/2018] [Accepted: 01/17/2019] [Indexed: 11/12/2022] Open
Abstract
BACKGROUND
Systemic inflammation, even at low levels, can greatly interfere with measures of iron status, making diagnosis of iron deficiency difficult. The objective of the present study was to create linear regression correction equations to adjust serum ferritin and iron concentrations based on measurements of the acute-phase proteins C-reactive protein (CRP) and albumin.
METHODS
Data from a cohort (1) of patients (n = 7226) in primary and secondary care who had serum ferritin, iron, CRP, and albumin measured at the same time point were examined. Linear regression coefficients were calculated for CRP and albumin with serum iron and ferritin as the outcome variables. Patients with ferritin <15 µg/L or serum iron <10 µmol/L were categorized as iron deficient. The equation was then applied to a cohort (2) of patients with colorectal cancer who had ferritin and iron measured preoperatively ( n = 356).
RESULTS
In cohort 1 there was a significant difference in the proportions of patients with serum ferritin <15 µg/L and serum iron <10 µmol/L, respectively, when the unadjusted (7% and 55%), adjusted based on CRP alone (13% and 26%), adjusted based on albumin alone (11% and 37%), and adjusted based on both CRP and albumin (24% and 15%) values were compared (both P < 0.001). In cohort 2 there was a significant difference in the proportions of patients with serum ferritin <15 µg/L and serum iron <10 µmol/L, respectively, when the unadjusted (28% and 66%), adjusted based on CRP alone (39% and 57%), adjusted based on albumin alone (39% and 59%), and adjusted based on both CRP and albumin (46% and 44%) values were compared (P < 0.001 and P < 0.004).
CONCLUSIONS
In both cohorts the greatest increase in the proportion of patients meeting definitions of iron deficiency was found when adjustment was made for both CRP and albumin together. Even low levels of inflammation had a significant effect on serum iron and ferritin values.
Collapse