Red Cell Phenotype Modifies the Association Between Free Triiodothyronine and Erythrocyte Count: A Cross-Sectional Study of 3673 Adults in a Hemoglobinopathy-Endemic Region of Saudi Arabia

Background/Objectives: Thyroid hormone regulates erythropoiesis, but most population studies have used thyrotropin (TSH) alone and have not examined whether the erythroid association differs by red cell phenotype. We examined this in a population where inherited microcytosis and iron deficiency are both common. Methods: We analyzed 3871 adults of both sexes with paired free triiodothyronine (FT3), free thyroxine, TSH, and complete blood count from three hospitals in Jazan, Saudi Arabia (3673 with FT3 within the reference interval formed the primary analysis). All three analytes were entered simultaneously in models adjusted for age and sex; mean corpuscular volume (MCV) and mean corpuscular hemoglobin (MCH) were pre-specified comparison (negative control) outcomes; red cell discriminant indices were used to assign a normocytic, acquired microcytic, or presumed inherited-trait pattern, which was compared against linked ferritin and transferrin saturation. Results: FT3 was associated with erythrocyte count (standardized β 0.097, 95% confidence interval (CI) 0.066–0.128, p < 0.001) and hemoglobin (+2.7 g/L per pmol/L); TSH was not (p = 0.547). The erythrocyte count association was present in normocytic (β 0.101) and acquired microcytic records (β 0.099) but absent in the presumed trait pattern (β −0.017; interaction p = 0.002) and remained absent under a stricter trait definition, whereas the hemoglobin association did not differ by phenotype (interaction p = 0.639). FT3 was unrelated to iron status, and the association survived adjustment for ferritin and restriction to same-episode iron measurements. In an exploratory comparison, both associations were strongest in iron-replete normocytic patients (hemoglobin +3.6 g/L per pmol/L, p < 0.001) and attenuated in measured iron deficiency. Conclusions: FT3, not TSH, was associated with erythrocyte count and hemoglobin. The count association was absent in the index-defined trait pattern, in which erythrocyte number is already high, and both associations were blunted where iron was limited. These observations are consistent with, but do not demonstrate, a limit on the marrow’s count response, since erythropoietic activity was not measured; they are hypothesis-generating and require replication with confirmed hemoglobin genotype, measured iron status, and reticulocyte or erythropoietin data.