Mathematic predictive models for calculating copper, iron and zinc dialysability in infant formulas

Differences in the dialysability of a mineral element from infant formulas of the same type were detected in a previous study. As these may be due to the effects that different levels of the components of the formulas could have on dialysability, we attempted to establish mathematical models to predict the dialysability of Cu, Fe and Zn from infant formulas according to their ascorbic, citric, or selected amino acid contents. A simple linear regression was applied between the ascorbic acid, citric acid and amino acid contents and the Cu, Fe and Zn dialysability of 18 powdered infant formulas of different types. Significant correlations (P<0.05) were obtained between the dialysability of Fe and ascorbic acid, citric acid, cysteine (Cys), glycine (Gly) and lysine (Lys) contents: between Cu and Cys, arginine (Arg) and tyrosine (Tyr) contents; and between Zn and Cys, Gly, Lys and Arg contents. A multiple regression was applied in those cases in which a significant correlation (P<0.05) was obtained. A joint effect of ascorbic and either citric acid or Cys on Fe dialysability was observed. The effect of Arg on Zn dialysability and the joint effect of Tyr and Cys on Cu dialysability must also be mentioned.