data for each individual in the sample is indefensible from a statistical viewpoint if the objective is to estimate prevalence of inadequacy.
Results from the same analyses applied to phosphorus intakes are shown in Figure 4-13. For phosphorus, prevalence of inadequacy estimates computed from the one-day and the adjusted intake distributions are 25 and 11 percent, respectively.
In these two cases (where the means of the intake distributions are greater than the EAR), the bias in the prevalence estimate that results from not removing the day-to-day variability in intakes leads to an overestimation of the proportion of individuals in the group whose intakes are inadequate. This is not always so; if the mean of the usual intake distribution is less than the EAR, using the one-day distribution to estimate prevalence may result in underestimation.
Should the Recommended Dietary Allowance (RDA) be used to assess the proportion of individuals in a group who are at risk of nutrient inadequacy? No. Estimating prevalence of nutrient inadequacy in a group by computing the proportion in the group with intakes below the RDA always leads to an overestimation of the true prevalence of inadequacy. |
By definition, the RDA is the intake level that exceeds the requirements of a large proportion of individuals in the group. In fact, when requirements in the population are distributed as normal random variables, the RDA exceeds the requirement of more than 97 percent of all individuals in the group.
As indicated previously in this chapter, the proportion of individuals in a group with nutrient intakes below their requirements can be estimated by using the Estimated Average Requirement (EAR) cut-point method (calculating the proportion of individuals in the group with intakes below the EAR). Examples were presented in which the cut-point method was shown to perform well. That is, when populations were simulated for which both nutrient intakes and requirements were known, approximately the same prevalence estimates resulted either by counting the actual number of individuals with nutrient intakes below their requirements or the number of individuals with intakes less than the EAR.