A diagnostic approach incorporating a predictive model based on simple parameters for differentiating dimorphic anemia from nutritional macrocytic or microcytic anemia.

Abstract

The red blood cell (RBC) indices of dimorphic anemia (DA) may present as macrocytic, microcytic, or normocytic, leading to missed diagnoses when relying solely on empirical judgment of these indices. This is particularly problematic in underdeveloped regions where tests for iron, vitamin B12, and folate status are unavailable. We retrospectively analyzed patients diagnosed with iron deficiency anemia (IDA) and megaloblastic anemia (MA) at our institution. We compared patient characteristics and complete blood count (CBC) parameters between IDA, MA, and DA, and subsequently developed a diagnostic model. About 142 DA, 288 IDA, and 60 MA patients were included. DA showed predominant microcytosis, with minor normocytic and rare macrocytic cases, enabling clear differentiation from MA by mean corpuscular volume (MCV) but not IDA. To distinguish DA from IDA, we analyzed patient characteristics and CBC parameters. By univariate analysis and logistic regression, sex, RBC, and WBC were identified as independent predictors. A predictive model assigned scores: 3.5 points for male sex, 1 point for RBC <3.36 × 10¹²/L, and 2 points for WBC > 5.87 × 10⁹/was built and internally validated. A flowchart for the diagnosis of nutritional anemia was proposed, with a focus on the rational measurement of iron, folate, and vitamin B12 to discover DA. Our algorithm provides a clinical model for DA prediction using simple parameters, enabling targeted guidance for the measurement of iron, vitamin B12, and folate. Noting the single-center limitation and sample constraints, this diagnostic approach requires validation in diverse populations and regions to potentially derive corresponding new models.