| Abstract: | Prevalences of anemia were estimated by two methods for 742 Black and 3,074 White nonpregnant women of childbearing age drawn from a large probability sample of the United States civilian noninstitutionalized population (NHANES I). One method defines the prevalence of anemia as the proportion of women with hemoglobin levels below a 12 g/dl "cut-off". The second method defines the prevalence of anemia as the proportion of women whose hemoglobin values are shifted downwards relative to a distribution of hemoglobin values of non-anemic women. Estimates produced by both methods suggest a higher prevalence of anemia in Black than in White women. Estimates produced by the "cut-off" method, however, are higher than those from the "distribution" method for both racial groups, probably because the "cut-off" method results in large overestimates in populations where anemia prevalence is low. The "distribution" method is further used to estimate the contribution of iron deficiency to anemia. Essentially all anemia in White women and a high proportion of anemia in Black women is associated with iron deficiency in the US civilian noninstitutionalized population. Iron supplementation trials are needed in order to define the magnitude of the problem accurately and plan appropriate public health programs. |
