Reticulocyte indices in a canine model of nutritional iron deficiency

Background: Reticulocyte indices, especially reticulocyte hemoglobin content (CHretic), have shown promise as markers of iron deficiency (ID), but there have been no prospective investigations of reticulocyte indices in experimental models of ID. Objectives: The objective of this study was to compare reticulocyte indices with conventional hematologic and biochemical indices as markers of ID in dogs. Methods: Iron deficiency was induced in 7 dogs by feeding an iron-deficient diet, and corrected by restoring dietary iron and by giving iron parenterally. Blood samples were collected at weekly intervals. Results of hematologic and biochemical tests were compared using t-tests and receiver operator characteristic (ROC) curve analysis. Results: Comparing mean values on days 0 and 35, by which time hemoglobin concentration decreased to <= 90% of baseline in all dogs, % Macro(retic), % Hypo(retic), % Low CHretic, and % High CHretic, differed by greater than 3-fold, whereas no conventional hematologic or biochemical indices differed by as much as 2-fold. Comparing conventional hematologic and reticulocyte indices by ROC curve analysis using 4 different biochemical diagnostic criteria of ID, CHretic, % Hypo(retic), % Low CHretic, and % High CHretic had higher areas under the curve (AUC) than either MCVconv or MCHCconv according to all 4 criteria, and % Macro(retic) and MCVretic had higher AUC values according to 3 of 4 criteria (differences were not always statistically significant). Conclusions: Results of this study support the value of reticulocyte indices in the diagnosis and monitoring of ID in dogs. Species similarities in the pathophysiology and hematologic manifestations of ID suggest these findings also may be relevant to ID in people.