food-based recommendations for the generally healthy U.S. and Canadian populations. That was not the task of this committee for whom the focus has been the quantitative nutrient requirements and upper levels of intake.
Currently, the mainstays of DRI development are the EAR, and the Tolerable Upper Intake Level, or UL (also referred to at times as Upper Levels of Intake). The RDA is to be derived from the EAR and reflects an estimate of an intake that meets the needs of 97.5 percent of the population’s requirements. It is not a target intended to be met by all individuals, and intakes below the RDA cannot be assumed to be inadequate because the RDA by definition exceeds the actual requirements of all but 2 to 3 percent of the population. The Adequate Intake (AI) was originally incorporated into the framework to address the inevitable uncertainties associated with specifying requirements for infants, given the challenges in obtaining sufficient information for this group, but has expanded to include use when available data for any life stage group are too limited to establish a requirement. The AI is the subject of some debate, given that it does not appear to readily “fit” into the probability assumptions for DRI use (Taylor, 2008). There are also other reference values, as described in other IOM documents (IOM, 2006), but as these are not relevant to this report, they are not described here.
The EAR is the average daily nutrient intake level that is estimated to meet the nutrient needs of half of the healthy individuals in a life stage or gender group. Although the term “average” is used, the EAR is actually an estimated median requirement (IOM, 2006). Therefore, by definition, the EAR exceeds the needs of half of the population and is less than the needs of the other half (Taylor, 2008).
The 1994 to 2004 DRI process placed emphasis on the distribution of requirements for a population, rather than focusing on a single value constructed to “cover” the great majority of the population, as had been the case in earlier efforts (Taylor, 2008). This, along with the development of newer methodologies for assessing and planning adequate intakes for groups, made the EAR a central reference value, along with the UL. The 10 years of DRI development moved the process from a black-and-white cutoff in the form of an RDA to consideration of a probability model. Doing so made it clear that there is a distribution of requirements in the population (Taylor, 2008).
The EAR itself presents little controversy as an expressed reference value. Beyond the question of how to handle EAR estimation in the face of limited data, most of the issues that surround EAR development are