This laboratory study measured iron and zinc in biofortified and conventional pearl millet, wheat, sweet potato and pulses grown for an Indian feeding trial, then compared iron uptake from single crops and meal-like combinations with an in vitro digestion and Caco-2 cell ferritin assay. Biofortified pearl millet and red lentil carried roughly twice the iron of their controls. Adding sweet potato or sweet potato plus pulse raised uptake more than iron concentration alone predicted.
Key numbers
- Iron, µg/g, mean ± SD, n = 3 (Table 1, p. 4): pearl millet biofortified 77.6 ± 1.7, control 39.1 ± 0.6; wheat biofortified 61.2 ± 8.1, control 47.5 ± 0.3; sweet potato biofortified 6.7 ± 0.1, control 5.7 ± 0.3 (wet weight); pulse (red lentil) 62.1 ± 0.6, control pulse (pigeon pea) 30.2 ± 0.4.
- Sweet potato dry matter proportion: 0.20 (biofortified orange) and 0.31 (control white) (Table 1 footnote).
- Single-crop Caco-2 iron uptake, ng ferritin per mg cell protein: pearl millet 5.01 ± 1.66 versus 2.17 ± 0.96 (p = 0.036); pulse 9.58 ± 1.61 versus 4.26 ± 0.61 (p = 0.0011); wheat 4.64 ± 1.13 versus 5.37 ± 1.13 (p = 0.47) (Table 2, p. 4).
- Within-crop iron differences: 38.5 µg/g for pearl millet and 46.0 µg/g for pulses (Results, p. 4; the 46.0 figure does not equal the Table 1 difference of 31.9 µg/g and is reported as printed).
Methods (brief)
Crops were cooked as planned for the trial (or by standard laboratory protocol), freeze-dried and ground. Minerals were measured in triplicate by ICP-AES after nitric and perchloric acid digestion of 0.5 g dried, ground sample. Seven crop combinations representing clusters of planned meals were digested with pepsin then pancreatin-bile and applied to Caco-2 monolayers; ferritin normalized to cell protein was the outcome.
Evidence fitness
Supports crop-variety iron concentrations for a single lot of each Indian crop and relative in vitro uptake comparisons. The Caco-2 ferritin response is a screening proxy, not measured human absorption. Grain and pulse concentrations are not labeled with a moisture basis in the table; only sweet potato is stated as wet weight.
Limitations
One sample lot per crop and variety, triplicate analysis only, crop proportions rather than prepared recipes, and no human data. The work was supported by public agricultural research and a HarvestPlus grant, which also supplied the crops; one author disclosed equity in a nutritional-status diagnostics company.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.