Mean or median intake seldom, if ever, can be used to assess nutrient adequacy of group diets. In the past, nutrient intake data have frequently been evaluated by comparing mean intakes with RDAs. In particular, studies that found mean intakes equal to or exceeding the RDA often concluded that group diets were adequate and conformed to recognized nutritional standards. However, this is inappropriate because the prevalence of inadequacy depends on the shape and variation of the usual intake distribution, not on mean intake. Indeed, for most nutrients, group mean intake must exceed the RDA for there to be an acceptably low prevalence of inadequate intakes. Moreover, the greater the variability in usual intake relative to the variability in requirement, the greater the mean usual intake must be relative to the RDA to ensure that only a small proportion of the group has inadequate intake. If group mean intake equals the RDA, there will be a substantial proportion of the group with usual intake less than requirement. Chapter 4 provides more detail on issues related to comparing mean intakes to the DRIs. Even stronger caution is needed when comparing group mean intakes with the EAR. If mean intake equals the EAR, it is likely that a very high proportion of the population will have inadequate usual intake. In fact, roughly half of the population is expected to have intakes less than their requirement (except for energy).

When the AI represents the group mean intake of an apparently healthy group (or groups) of people, similar groups with mean intakes at or above the AI can be assumed to have a low prevalence of inadequate intakes for the defined criteria of nutritional status. For AIs that were either experimentally derived or developed from a combination of experimental and intake data, a similar assessment can be made, but with less confidence. Each AI is described in terms of its derivation and selected criterion of adequacy in the individual nutrient panel reports (IOM, 1997, 1998b, 2000). When mean intakes of groups are below the AI it is not possible to make any assumptions about the extent of intake inadequacy. It is not appropriate to try to estimate an EAR from an AI.