An in vitro digestion/Caco-2 cell culture system accurately predicts the effects of ascorbic acid and polyphenolic compounds on iron bioavailability in humans.

We developed a rapid in vitro digestion/Caco-2 cell culture model for assessing relative bioavailabilities of iron in foods and meals. The objective of the present study was to determine how closely our Caco-2 model reflects the human response. Meals described in published reports of studies on effects of varying levels of ascorbic acid (AA) and tannic acid (TA) on iron absorption by human subjects were carefully replicated. Iron absorption ratios (iron absorption from meals containing AA or TA divided by iron absorption from identical meals without these enhancers or inhibitors) were determined using the Caco-2 model. Ferritin formation by the Caco-2 cells was used as an indicator of iron absorption. Response patterns of effects of AA and TA on absorption ratios (AR) calculated from Caco-2 and human data were very similar: AA increased ARs in a dose-response manner and TAs decreased AR. The natural logs of the ARs determined in Caco-2 and human studies were correlated: R = 0.935 (P = 0.012) and 0.927 (P = 0.007) for AA and TA, respectively. When results from meals with AA and TA were pooled, a linear relation between the natural logs of ARs from Caco-2 and human studies was observed (R = 0.968, P < 0.001). We conclude that our Caco-2 model accurately predicts the human response to AA and TA in the meals we tested. If future work reproduces the precision and accuracy shown in this paper for predicting iron bioavailability to humans, then the implications for saving time and resources in iron bioavailability measurements are considerable.