(P01-012-23) Determining Vitamin A Absorption Efficiency in Humans: Development of an Expanded Compartmental Model and Results of Pilot Studies Using an Updated Experimental Approach

Objectives: Objectives were to use a new, more physiological model for vitamin A (VA) absorption to analyze pilot data on plasma retinyl ester (RE) kinetics after ingestion of a test dose of unlabeled VA and a fat-containing breakfast followed by several fat-containing snacks to facilitate “washing out” of absorbed VA from enterocytes while minimizing interference due to appearance of hepatic VLDL RE in plasma.

Methods: A blood sample (~4 mL) was collected from a fasting young adult woman via an antecubital vein catheter. She then ingested a capsule containing 1.91 µmol VA and fat-/VA-free yogurt mixed with 7 g canola oil. Blood was collected every 30 min from 1–4 h and then hourly until 8 h, with subject consuming 2 g canola oil at 30, 60, 90 and 120 min. Plasma samples were analyzed for RE by HPLC. After correcting RE concentrations for fasting RE, data were analyzed using Simulation, Analysis and Modeling software (WinSAAM).

Results: We modified the “front end” of our existing VA model to include 3 compartments representing movement into the absorptive area of the small intestine, followed by 2 compartments for VA absorption (one for earlier “duodenal” absorption and the other for “jejunal” absorption, which was also the exit site for non-absorbed VA); these 2 compartments each had output to a delay component. VA from both delays entered a chylomicron RE compartment and remnant RE were subsequently cleared from circulation. The new model with 7 adjustable parameters, including 1 for absorption efficiency, provided a good fit to the data. RE peaked at 90 min, followed by a plateau from 90 min–4 h, and then a rapid decline back to baseline at 8 h. Data were compatible with a model-predicted VA absorption of 76%.

Conclusions: We conclude that the added fat feedings were a source of chylomicron triglycerides which allowed all absorbed VA to be carried into plasma in 4–6 h without the apparent appearance of VLDL RE to complicate the modeling. Additional testing of this protocol (currently in progress) may provide reliable information on VA absorption which is needed to accurately estimate VA stores by retinol isotope dilution and for improved modeling of whole-body VA kinetics.

Funding Sources: Funded by the International Atomic Energy Agency. JA-P received a fellowship from the Mexican National Research and Technology Council.