Estimating Vitamin A Absorption Using Compartmental Modeling of Plasma Retinyl Ester Kinetics in Humans

Practical and accurate methods for quantifying vitamin A (VA) absorption in humans are needed.

The objectives were to determine VA absorption in humans by applying compartmental modeling to postprandial plasma retinyl ester (RE) kinetic data obtained using a carefully designed controlled feeding protocol.

After collection of baseline fasting blood, 3 healthy young adult women ingested 1.91 μmol unlabeled VA and a breakfast of 145 g fat-/VA-free yogurt mixed with 7 g canola oil. Blood was collected every 30 min for 8 h, with subjects consuming snacks of 2 g canola oil mixed with 18 g yogurt every 30 min until 4 h, followed by a VA-free lunch at 4.5 h. Plasma RE were analyzed by high-performance liquid chromatography and data (corrected for baseline RE) were analyzed using Simulation, Analysis and Modeling software.

Plasma RE was above baseline at 30 min postdosing, reaching a maximum at 90, 150, and 180 min in the 3 subjects. Plasma RE then declined at different rates, almost reaching baseline values by 8 h. To fit these data, we developed a complex, physiologically realistic compartmental model with 12 compartments and (initially) 10 adjustable parameters. After evaluating the uniqueness of initial estimates for VA absorption, we determined final model-predicted values of 72%, 81%, and 78%, respectively, for the 3 women.

We conclude that VA absorption efficiency can be estimated in humans by applying compartmental modeling to postprandial plasma RE kinetic data obtained under a controlled feeding protocol and with the collection of a sufficient number of serial blood samples. Additional work is needed to refine this approach so that one can confidently identify a unique value for VA absorption in human subjects.