無症候性個人における鉄欠乏検出を強化する従来の指標と網状赤血球指標の組み合わせ

Iron deficiency (ID) is the most common micronutrient deficiency globally, often underdiagnosed due to nonspecific symptoms and limitations of standard laboratory tests. Reticulocyte-derived indices from modern hematology analyzers offer enhanced early detection of iron-restricted erythropoiesis.

We studied 103 self-reported asymptomatic healthy adults who underwent complete blood counts and reticulocyte analysis on XN-2000 analyzers (Sysmex Corp., Kobe, Japan). Detailed biochemical iron profile and vitamin B12/folate testing were done. Diagnostic performance of hematologic parameters was assessed and combinatorial logistic regression models and discriminant indices were developed.

Participants' median age was 28 years (range 22-79); 69.9% (n = 72) were women. 54.4% (n = 56) had ID and 21.4% (n = 22) were anemic. Latent ID was detected in 34.9% (36/103). B12 and folate deficiencies were present in 41% and 40%, respectively - extensively overlapping with ID. Key individual indices for ID included RETHe, RETY, and RET-RBC-Y, each with area under the ROC curve (AUC) >85%. A comprehensive 21-variable logistic regression model yielded AUC 93.7% (95% C.I. 88.9-98.5), sensitivity 89.1%, and specificity 70.2%. A pared-down 3-variable model achieved AUC 90.1% (95% C.I. 84.1-96.1), sensitivity 87.5%, and specificity 70.9%. Of the two heuristic composite indices tested, one - [RDW-CV × PLT × 10⁵] / [RET-He × RET-RBC-Y × IRF-Y × RET-Y] - showed AUC 89.8% (95% C.I. 83.7-95.9), sensitivity 87%, and specificity 72.3%.

In a cohort with high rates of nutritional deficiency, cost-effective equations combining conventional and advanced reticulocyte indices demonstrated strong diagnostic utility for screening ID, with potential for broader application in resource-limited settings.