Abstract
Frequency of anaemia was estimated by two methods for 254 menstruating women living in South-eastern Algeria. One method defines the frequency of anaemia as the percentage of women with haemoglobin concentrations below the cut-off point defined by WHO (12 g/dl). The second method defines the frequency of anaemia as the percentage of women whose haemoglobin values are shifted downwards relative to a Gaussian distribution of haemoglobin of nonanaemic women. The conventional cut-off point probably tends to overestimate the true frequency of anaemia: 7% of women with haemoglobin concentration less then 12 g/dl were not found as anaemic using the cumulative frequency method. The contribution of iron deficiency folate deficiency and inflammatory process was estimated using the cumulative frequency distribution after excluding respectively women with biological evidence of iron deficiency (serum ferritin of 12 μg/l or less, transferrin saturation less than 15% and/or MCV less than 80 fl), of folate deficiency (red blood cell folates less than 100 μ/l) and of inflammatory process (C. Reactive Protein more than 12 μ/l, orosomucoid more than 1.4 g/l or wtifte cell counts of more than 10 000/mm3). According to this method Iron deficiency represented the most important cause of anaemia in the context of our sample: iron deficiency contributed to 77% of anaemia. Folate deficiency and inflammatory processes do not in themselves appear to contribute to anaemia.