incorporated into the previous discussions focused on establishing reference values for adequacy.
There are often ethical issues associated with conducting clinical trials designed to study the adverse effects of substances that can limit the types of data available for DRI development. For this reason, the derivation of ULs for DRI purposes necessarily relies more heavily on observational data and information derived from animal models than does the approach for the determination of levels of intake for nutritional adequacy. Thus, the emphasis on causality and strength of evidence needed for establishing reference values for adequacy is difficult to apply to the derivation of ULs.
At the outset, it is important to distinguish between the relatively “acute” toxic effects of excess intake and the “chronic” adverse effects of high levels of intake that may manifest in other ways including disease risk. When the ULs for calcium and vitamin D were originally established in 1997, it was noted that the available data were limited relative to adverse outcomes and dose–response relationships (IOM, 1997). In that report, adverse effects from excess intakes of calcium and vitamin D were considered primarily in terms of acute toxicity, which was defined as the condition of hypercalcemia or, in some cases, hypercalciuria with or without hypercalcemia.
The conditions associated with the intoxication syndrome for calcium and vitamin D are informative, but avoiding acute toxicity is not the ideal basis for a UL, a reference value with the larger purpose of public health protection over a life time of chronic intake. Although information concerning chronic excess intakes remains limited, data have emerged recently that may warrant caution about the levels of vitamin D that are consumed and raise questions about the long-term effects of high intakes that are less than those associated with toxicity and that may result in an increase in serum 25-hydroxyvitamin D (25OHD) levels into upper ranges previously considered physiological. Caution may also be warranted in comparing the effects at these high physiological levels of 25OHD achieved through supplementation versus sun exposure, and further research is needed to clarify the relative adverse effects of different sources of vitamin D.
The model developed for UL derivation was summarized in 1998 (IOM, 1998), and it acknowledged that the lack of data would affect the ability to derive precise estimates. Specifically: “Several judgments must be made regarding the uncertainties and thus the uncertainty factor (UF) associated with extrapolating from the observed data to the general population.” Although a number of reports describe the underlying basis for uncertainty factors (Zielhuis and van der Kreek, 1979; Dourson and Stara, 1983), the strength of the evidence supporting the use of a specific UL undoubtedly varies. The summary of the 2007 workshop focused on enhancing DRI development (IOM, 2008) and pointed out the need for uncertainty fac-