The problem with using the EAR cut-point method for food energy can best be illustrated by considering an admittedly extreme example of both a perfect correlation between individual intakes and requirements and mean intake equal to the average requirement. Because each individual in the group has a usual intake equal to his or her requirement, the prevalence of inadequacy is zero. However, because one-half of the group has usual intakes less than the average requirement and one-half has usual intakes exceeding the average requirement, the cut-point method would estimate that 50 percent of the group is at risk of inadequate intakes when, in fact, the prevalence of inadequacy is zero.
Therefore, to assess energy adequacy, information other than intakes could be used, such as body weight for height, body mass index, or other anthropometric measures.
Situations in which nutrient intakes and requirements may be related to a third variable (e.g., energy and thiamin, body weight and protein) have not been well studied.
A good example of an asymmetrical requirement distribution is iron requirements in menstruating women. The iron requirement includes the need to replace urine, fecal, and dermal iron losses, and this aspect of the requirement does appear to be symmetrically distributed in the population (FAO/WHO, 1988). For menstruating women, iron lost in menstrual flow varies considerably—the mean loss averaged over 1 month has been estimated at 0.5 mg/day but about 5 percent of women have losses averaging more than 1.4 mg/day (FAO/WHO, 1988; Hallberg et al., 1966). This means that the distribution of iron requirements in women is skewed—there are more women with needs 25 percent or more above the mean, for example, than with needs 25 percent or more below the mean. In this case, the mean requirement is different from the median requirement (or EAR) in the group.
Figure 4-10 illustrates this situation, which is modeled after the information about iron requirements in women given in the FAO/ WHO report of 1988. The median requirement (EAR) is 10 mg but the distribution of requirements is not symmetrical around the