Iron Bioavailability from Milk: FAP Delivers 2.2× Higher Absorption than Ferric Pyrophosphate in Adults
In a double-blind, randomized crossover clinical trial in 38 young women, geometric mean iron absorption from ferrous ammonium phosphate (FAP)-fortified full-cream milk powder was 7.4%, compared to 10.4% for ferrous sulfate (FeSO₄, the water-soluble reference) and only 3.3% for ferric pyrophosphate (FPP). FAP absorption was 2.24-fold higher than FPP (p < 0.0001). Relative bioavailability (RBV) versus FeSO₄ was 0.71 for FAP and 0.32 for FPP [1]. This establishes FAP as the only organoleptically neutral iron compound that retains clinically meaningful absorption in a difficult-to-fortify dairy matrix where soluble FeSO₄ causes unacceptable discoloration and rancidity.
| Evidence Dimension | Geometric mean fractional iron absorption (%), erythrocyte incorporation dual stable isotope (⁵⁷Fe, ⁵⁸Fe) |
|---|---|
| Target Compound Data | FAP: 7.4% (95% CI not fully reported; RBV = 0.71 vs FeSO₄) |
| Comparator Or Baseline | FeSO₄: 10.4%; FPP (ferric pyrophosphate): 3.3% (RBV = 0.32 vs FeSO₄) |
| Quantified Difference | FAP vs FPP: 2.24-fold higher absorption (p < 0.0001); FAP vs FeSO₄: 1.41-fold lower (p = 0.0002) |
| Conditions | Double-blind randomized crossover design; n = 38 young women; full-cream milk powder vehicle; erythrocyte incorporation dual stable isotope technique (⁵⁷Fe, ⁵⁸Fe); single oral dose |
Why This Matters
For procurement in food fortification programs, FAP uniquely bridges the gap between high-bioavailability but organoleptically destructive FeSO₄ and organoleptically inert but poorly bioavailable FPP, providing a 2.2× bioavailability advantage over the incumbent insoluble iron alternative in milk-based products.
- [1] Walczyk T, Kastenmayer P, Storcksdieck Genannt Bonsmann S, Zeder C, Grathwohl D, Hurrell RF. Ferrous ammonium phosphate (FeNH₄PO₄) as a new food fortificant: iron bioavailability compared to ferrous sulfate and ferric pyrophosphate from an instant milk drink. European Journal of Nutrition 2013; 52(4): 1361–1368. View Source
