Abstract Background Iron deficiency (ID) is the most common nutrient deficiency in the world and iron deficiency anemia (IDA) is estimated to affect 1.24 billion people. Serum ferritin is an iron storage protein commonly used as a biomarker for ID and IDA. The ferritin lower limit of normal (LLN) defined in many clinical laboratories is lower than evidence-based cutoffs and may not detect ID and IDA with adequate sensitivity, particularly when sex-based reference intervals (RIs) are used. This may be due in part to the use of asymptomatic, iron-depleted individuals for RI studies. Recent efforts have been made to increase the LLN for ferritin to at least 30 ng/mL and re-evaluate sex based RIs to improve diagnosis of ID and IDA. In this study, we aimed to investigate LLN ferritin cutoffs for ID and IDA in our patient population. Methods Patient data from June 2022-May 2023 were collected with approval from the Institutional Review Board (IRB), including age, sex, and values for ferritin, iron, transferrin saturation, and hemoglobin. Patients were classified as iron replete (IR), ID, or IDA based on measurements for iron, transferrin saturation, and hemoglobin. IR patients were defined as having iron and transferrin saturation values within current RIs, while those below the cutoff were grouped as ID. ID and IDA were further classified based on hemoglobin values. A range of ferritin concentrations between 1 and 100 ng/mL were used to calculate true and false positive rates (TPR and FPR) and diagnostic accuracy (DA). Receiver operator characteristic (ROC) curve and J index analysis were used to define optimal ferritin cutoffs for ID and IDA. Results Data from 26,455 patients (65% female) were used in this study. Conclusions These findings highlight the need to re-examine ferritin RIs in the context of ID/IDA, especially for those patients assigned female at birth.
Read full abstract