The DRIs for calcium and vitamin D established in 1997 (IOM, 1997) also relied on bone health as the indicator in setting reference values for adequacy. However, the 1997 report established an AI for all life stage groups; no EARs or RDAs were specified. Newer data plus an integration of data have allowed the estimation of EARs and RDAs for all life stages except infants. Quantitative comparisons between AIs and EARs and RDAs are not appropriate.

In 1997, AIs were established for calcium in lieu of EARs and RDAs as a result of uncertainties associated with balance studies, lack of concordance between observational and experimental data, and lack of longitudinal data to verify the relationship between calcium intake, calcium retention and bone loss (IOM, 1997). In the past 10 years, newer evidence on skeletal health has emerged from a combination of large-scale randomized trials and calcium balance studies as described in Chapter 4. Further, there are now data relative to a number of life stage groups, and these help to avoid reliance on extrapolating or scaling data from one life stage to another unstudied life stage.

In the case of vitamin D, the 1997 report concluded that there were inadequate data available for EARs and RDAs as a result of uncertainties about sun exposure, the vitamin D content of the diet, and vitamin D stores (IOM, 1997). In the intervening years data have emerged that allow a requirement distribution to be simulated for vitamin D, which, in turn, has been found to be concordant with other available data. This analysis unexpectedly indicated that the dose–response relationship regarding median requirements is not significantly affected by age. Further, several newer studies can be used to elucidate the contributions made by sun exposure and to help separate total intake contributions from contributions stemming from cutaneous synthesis. Strides have been made in estimating the vitamin D content of foods as well as the amounts of vitamin D consumed by the U.S. and Canadian populations.

Despite new data since the earlier Institute of Medicine (IOM) report (IOM, 1997), there remain a number of uncertainties that have caused challenges in estimating DRI values for calcium and vitamin D. Notable among these is the absence of intervention trials that study dose–response relationships for the nutrients. Rather, most of the evidence is derived from a single dose that is often relatively high. Further, some studies fail to specify information about the background diet and hence the total level of intake is lacking. When this is the case, the mean population requirement may be below the dose used in the study, but cannot be further specified. In addition, there is the common practice of designing studies to examine calcium and vitamin D in combination, thereby precluding the ability to discern the effects of each nutrient alone, which is of interest when establishing a reference value for a nutrient.