As described earlier in this chapter, trying to compare an individual 's intake to his or her requirement for a nutrient is difficult for two main reasons: (1) one needs to know an individual's requirement; and (2) one needs to know an individual's long-term usual intake of the nutrient. Neither the individual's requirement nor the usual intake of an individual is known.
Appendix B presents in detail a proposed approach, summarized below, to address this issue, recognizing that nutrient requirement and usual intake are not observable for a given individual. This approach is based on the following assumptions:
The EAR is the best estimate of an individual's requirement.
There is person-to-person variation in requirements. The standard deviation of the requirement is an indicator of how much the individual 's requirement for a nutrient can deviate from the median requirement (EAR) in the population.
Mean observed intake of an individual is the best estimate of an individual's usual intake.
There is day-to-day variation in intake for an individual. The within-person standard deviation of intakes is an indicator of how much observed intake may deviate from usual intake.
Inferences about the adequacy of an individual's diet can be made by looking at the difference between observed intake and the median requirement. That is, D is the difference between the mean observed intake for an individual and the median requirement (EAR, called r for simplicity) for the life stage and gender group to which the individual belongs,
D = − r.
If the difference D is large and positive, that is, if observed intake is much greater than the median requirement, then it is likely that an individual 's intake is adequate. Conversely, if the difference D is large and negative, that is, observed intake is much less than the median requirement, then it is likely that an individual's intake is not adequate. In between, there is considerable uncertainty about the adequacy of the individual's intake.
The obvious question then, concerns how large D would have to be before it could be concluded with some degree of assurance that the individual's unobservable usual intake exceeds the individual's