Conceptual and methodological issues regarding the epidemiology of iron deficiency and their implications for studies of the functional consequences of iron deficiency.

This paper reviews the measures of iron status (hemoglobin, hematocrit, mean corpuscular hemoglobin concentration, mean cell volume, free erythrocyte protoporphyrin, serum iron, transferrin saturation, and serum ferritin) that are potentially available for inclusion in field studies of the relationship between iron and mental performance. The characteristics of these measures (sensitivity to iron status, specificity to iron, and diurnal and day-to-day variability) are reviewed and the implications of choice of variable for the design, analysis, and interpretation of studies are discussed. Brief consideration is given to the question of confounding variables and to sources of both false-positive and false-negative conclusions. The explicit message of the paper is that there is no perfect choice of measure of iron status but, given explicit definition of the research question, there are preferred choices that can most effectively combine the choice of variable and the design of the study.