ABSTRACT
Anemia is common among Thai females. Thalassemia and iron deficiency are highly prevalent in the Thai population. A therapeutic trial of iron has been used to differentiate between the two conditions, however, no previous study on its usefulness in a Thai population has been reported. Otherwise healthy persons who had complete blood count (CBC) as routine 'check-up' and found to be anemic (Hb < 12 g/dl) at a preventive medicine clinic were tested for hemoglobin typing, serum ferritin, serum iron, and were given oral iron sulfate (120 mg elemental iron per day for at least 2 months) and a repeat CBC on a follow-up visit. Sixty-six individuals, all females, with pre-treatment hemoglobin (Hb) level of 9.5 +/- 1.7 g/dl (mean +/- SD), had complete data for analysis. Final diagnoses were isolated iron deficiency in 23 (34.8%), iron deficient thalassemia traits in 6 (9.1%) and iron-sufficient thalassemia syndromes in 29 (43.9%) anemic subjects. After a therapeutic trial of iron, Hb rose to 12.8 +/- 1.0 g/dl (n = 16, p = 2 x 10(-8)) among the iron deficient group, but not in thalassemia. The authors have identified that the most useful red cell indices that will discriminate between iron deficiency and thalassemia is a combination of red blood cell counts (RBC) > 4.4 x 10(6)/microl and mean corpuscular volume (MCV) < 69 fl. High RBC (> 4.4 x 10(6)/microl) and very low MCV (< 69 fl) is a sensitive (92.9%) and highly specific (100%) criteria for diagnosis of mild thalassemia diseases (Hemoglobin H (HbH), Hemoglobin H-Constant Spring (HbH-CS), and homozygous Hemoglobin E (HbEE)). Conversely, a low RBC (> 4.4 x 10(6)/microl) and/or low to normal MCV (69-85 fl) is highly sensitive (91.3%) but not specific (60%) for the diagnosis of iron deficiency. The authors conclude that a therapeutic trial of iron is useful as a diagnostic test in anemic females except those with high RBC (> 4.4 x 10(6)/microl) and very low MCV (< 69 fl), a subgroup which most likely has thalassemia and are least likely to benefit from iron treatment.