specifically identified in the nutrient reports (e.g., for folate in the prevention of neural tube defects [IOM, 1998b]), there is no currently established benefit to healthy individuals associated with ingestion of nutrients in amounts exceeding the RDA or AI.

The UL is based on an evaluation conducted using the methodology for risk assessment of the adverse effects of nutrients (IOM, 1998a). The need to establish ULs grew out of the increasingly common practice of fortification of foods with nutrients and the increased use of dietary supplements. For some nutrients, data may not be sufficient for developing a UL. This indicates the need for caution in consuming high intakes and should not be interpreted as meaning that high intakes pose no risk of adverse effects.

Unless otherwise stated, all values given for EARs, RDAs, AIs, and ULs represent the total quantity of the nutrient or food component to be supplied by foods (including nutrients added to foods) and by nutrients ingested as supplements. These values are also based on usual or continuing intakes. The DRIs apply to the apparently healthy population. RDAs and AIs are not expected to replete individuals who are already malnourished, nor are they intended to be adequate for those who may have increased requirements because of certain disease states. Appropriate goals for intake should be provided to those with greatly increased nutrient requirements. Although the RDA or AI may serve as the basis for such guidance, qualified medical and nutrition personnel should make necessary adaptations for specific situations.

In general, both values are intended to cover the needs of nearly all members of a life stage group. For both RDAs and AIs, values for children and adolescents may be extrapolated from adult values if no other usable data are available. However, there is much less certainty about an AI value in comparison to an RDA value.

The RDA is based on specific knowledge of the requirement and assumptions about its distribution and is set to meet the requirements of almost all (97 to 98 percent) of the population. In contrast, the AI is an experimentally derived or observed mean intake that appears to maintain a specific criterion of adequacy in a group of apparently healthy people. Therefore, by definition, the RDA incorporates only the estimated variability in requirements, where-