ARS | Publication request: Algorithms for Estimating Zinc Intake from Whole Diets

Submitted to: National Nutrient Databank Conference
Publication Type: Abstract
Publication Acceptance Date: March 1, 2005
Publication Date: April 1, 2005
Citation: Beisiegel, J.M., Hunt, J.R. 2005. Algorithms for estimating zinc intake from whole diets [abstract]. Presented at National Nutrient Databank Conference, San Diego, CA, April 1, 2005.

Technical Abstract: Objective: Algorithms to predict fractional Zn absorption (FZA) and total Zn absorbed (TZA) were developed from measurements of Zn absorption from whole diets. Methods and Materials: Data included studies measuring Zn absorption of healthy adults from three or more consecutive meals extrinsically labeled with Zn isotope. Diets (n = 28 from 11 studies) fitting these criteria contained 30 - 134 mg protein, 0.17 - 5.1 g phytate, 3.4 - 22 mg Zn, 1.1 - 36 molar ratios of phytate:Zn, 0.12 - 2.1 g Ca, and 5.4 - 27 mg Fe. Dietary variables were normalized to 2,500 kcals to predict FZA (0.14 - 0.49) and TZA (1.7 - 6.0 mg). Results: Individual, logarithmic, and interaction terms were examined for best-fit linear regression models of FZA [Logit FZA = 1.06 - (0.258 * g phytate) - (0.535* ln (mg Zn)) + (0.373 * g Ca) - (0.273 * ln (mg Fe)); R2 = 0.83; p < 0.0001] and TZA [ln (TZA) = 1.16 - (0.192 * g phytate) + (0.0709 * mg Zn) + (0.166 * g Ca) - (0.301 * ln (mg Fe)); R2 = 0.84; p < 0.0001]. Variability in FZA was explained by phytate (62%, partial R-square), Zn (12%), Ca (7%) and Fe (2%). For TZA, zinc explained 43%, phytate 34%, Ca 3%, and Fe 4% of the variation. Protein lent no predictive power to either model. In a previous model, total Zn and phytate:Zn explained 41% of FZA variability from 15 diets (IZiNCG, Food Nutr Bull 2004;25:S94). That model explained 64% of FZA variability from these 28 diets, whereas the new model explained 83%. Significance: Such algorithms enable easy estimation of zinc absorption from whole diets.