Diagnosis of iron deficiency / Elisabeth Rybo.

Tese (doutor) - Göteborg University, 1984

Haematological variables were studied in a randomised population sample of 372 38-year-old women resident in Gothenburg, Sweden during the period 1968-69. The participation rate was 91.4 %. The study was reformed to evaluate the laboratory diagnosis of iron deficiency. The haematological investigation included determination of haemoglobin concentration, erythrocyte and reticulocyte counts, serum iron concentration, total iron binding capacity, packed cell volume, sternal bone marrow aspiration with estimation of stainable iron in the reticuloendothelial cells, sideroblast count, iron absorption test with 59Fe in a whole-body counter and determination of menstrual blood loss. Serum ferritin concentrations were later determined on samples which had been frozen for 10 years. Eighty-nine women with no or only traces of reticuloendothelial iron in the bone marrow completed a double-blind supplementation trial, in which iron and placebo therapy were compared. Determination of stainable iron in the reticuloendothelial cells of the bone marrcow was used as a direct method for the evaluation of iron stores.In 31.5% of the women there was no reticulendothelial iron and in 14.3 % there were only traces. Of the indirect methods compared with the bone marrow method the serum ferritin concentration and iron absorption gave the highest correlations. A high correlation was also observed between menstrual blood loss and the content of reticuloendothelial iron. An arbitrarily chosen scale for the estimation of the amount of stainable reticuloendothelial iron in the bone marrow was used for these correlation analyses. The differences in mean corpuscular haemoglobin, transferrin saturation, sideroblast count, iron absorption, and menstrual blood loss were all highly statistically significant when the series of women with no or only traces of reticuloendothelial iron in the bone marrow were compared to those with clearly visible or abundant amounts of stainable iron. There were no significant differences concerning these variables between women who had no traces of stainable iron and those who had only traces of stainable iron. Increased iron absorption was found in women with a lack of stainable iron in the reticuloendothelial cells compared to women with stainable iron, irrespective of sideroblast count. The results from the iron supplementation study indicates that the haemoglobin concentration values of many of the women, although being with in the normal range, were suboptimal for the individual woman. However, we considered it not possible to ascertain which women had haematologically suboptimal values who wore responders to iron therapy, as a comparison between the effects of iron and placebo could not be made for the individual subject. According to multivariate analysis, determination of serum ferritin concentration was the most valuable indirect method for the diagnosis of depleted iron stores, and we are of the opinion that serum ferritin determination may replace bone marrow examination in the diagnosis of iron defi- ciency in women with or without anaemia who are otherwise healthy.