The committee's opinion, based on the findings presented above and the information cited in the references to this report, is that the measures now in place are successfully addressing the problem of iron deficiency anemia among infants and young children. The preliminary finding that the prevalence of iron deficiency anemia is somewhat higher among children in one Hispanic sub-group, if confirmed, suggests a time lag in effectively reaching groups made up of recent immigrant populations. In the case of women, the prevalence of iron deficiency anemia persists at a level of 4 to 10 percent, and better information is needed to know why this is so. There is a need, in general, to have better data on specific population groups to define meaningful cutoff points for iron deficiency anemia in people of different racial and ethnic groups and to target interventions more effectively. These needs are further explained in the section on recommendations for research at the end of the committee's report.

A recent article by Salonen and colleagues (1992) has renewed interest in the role of iron in chronic disease. In the study of Salonen and colleagues, the central question was relatively straightforward: Is excess body iron, as indicated by the plasma ferritin concentration, a significant positive risk factor for myocardial infarction? The committee determined that it would be necessary to review the role of iron in relation to chronic disease (cancer, atherosclerosis, and neurodegenerative disorders) postulated as resulting from excess iron through iron-catalyzed, biologically undesirable reactions. The information considered by the committee in pursuit of this question is contained in the paper by committee member John L. Beard, which appears in Appendix C to this report.

The committee determined that data are insufficient for a satisfactory test of the hypothesis of Salonen et al. (1992) in the U.S. population or to link excess iron intake to an increased risk of other chronic diseases. A follow-up analysis of NHANES I data (Sempos et al., under review) found no increased relative risk of myocardial infarction associated with excessive dietary iron intake or high levels of transferrin saturation, hemoglobin, or hematocrit (serum ferritin concentration were not measured) in people of all age, gender, and ethnic groups that were studied. Other analyses (Cooper et al., 1993; Daviglus et al., 1993; Dunn et al., 1993; Rimm et al., 1993; Stampfer et al., 1993) reached essentially the same negative conclusion.

In addition to the potential direct effects of high dietary iron intakes, large doses of supplemental iron may have an effect on the levels of other minerals (i.e., zinc, manganese, and copper) in plasma. Although there is a substantial research base on the interaction between iron and other minerals, it remains uncertain whether recommendations for the use of supplemental iron for the prevention of iron deficiency anemia may have a significant effect on other