Mathematical model of iron absorption based on biochemistry of absorption

The study objective was to develop models of heme and nonheme iron absorption based on elementary conceptions of the biochemistry of the absorption processes and to evaluate model validity by fitting the models to published data from healthy adults using nonlinear regression. This approach is in contrast to the numerous published models of iron absorption which rely on simple or multiple linear regression on selected variables with minimal consideration of the actual roles of the variables in the absorption process. When fit to absorption data on heme and nonheme iron administered in the fasting state, the models indicated significant relationships between absorption and amount of iron consumed and serum ferritin concentration; the R 2 values were ≥ 0.73, the parameters adequately estimated and the residuals supported the validity of the models. The fit of the model to nonheme iron absorption from varied meals demonstrated the relation of absorption to dietary iron, phytate and ascorbic acid and serum ferritin. The R 2 was 0.88 and the residuals supported the model, though the parameters were not well‐estimated. In conclusion, the models appear to provide a valid foundation for a biochemically‐based approach to the mathematical modeling of iron absorption. This work was supported by HarvestPlus (#8030, 8057 and 8207) and the National Institutes of Health K24 DK083772.