Requirement (EAR), Adequate Intake (AI), and Tolerable Upper Intake Level (UL), but not the Recommended Dietary Allowance (RDA). The methods used in estimating usual intakes and applying the different types of reference values are described below and in Appendix G.
The usual intake of a nutrient is an individual’s long-term average intake of that nutrient (NRC, 1986). Usual intake must be estimated; it cannot be observed, because day-to-day intakes vary considerably. The Iowa State University method (Nusser et al., 1996) is the commonly used and accepted approach for estimating the usual intakes by population groups. This method estimates the distribution of usual intakes by using a single 24-hour recall for all members of the group and a second 24-hour recall for some proportion of the group. For NHANES 2003–2004, a second 24-hour recall was collected for all persons in the sample. The personal computer version of the Software for Intake Distribution Estimation (PC-SIDE; ISU, 1997) was used to estimate (1) usual nutrient intake distributions and (2) the proportion of children and adults with usual intakes above or below the defined cutoff values.
The Dietary Reference Intakes (DRIs) released by the IOM replaced the previously used RDAs (NRC, 1989) as authoritative reference values. The DRIs provide several types of reference values for use in the assessment and planning of diets of groups, including the EAR, the AI for nutrients without an EAR, the UL, and the Acceptable Macronutrient Distribution Range (AMDR). Definitions of these DRIs are given in Box 3-1.
Contrary to earlier practice, all DRIs except the RDAs are useful for the assessment of nutrient intakes. The RDAs are inappropriate for the assessment of the nutrient intakes of groups because the percentage of individuals with usual intakes below the RDA does not estimate the percentage of individuals with inadequate intakes. The prevalence of inadequate intakes can be estimated by comparing usual intakes in the group to the entire distribution of requirements in the same group. This approach is known as the full probability approach and was proposed by the National Research Council (1986). The EAR cut-point method, a shortcut of the full probability approach, may also be used to obtain such an estimate (Beaton, 1994; Carriquiry, 1999; IOM, 2000). Estimation of the prevalence of nutrient inadequacy in a group by determining the proportion of individuals with