Impact of Acute Dietary Vitamin A Intake on the Measured Variable (Plasma Retinol Specific Activity) Used in Prediction of Total Body Vitamin A Stores by Retinol Isotope Dilution

Retinol isotope dilution (RID) equations are used to estimate vitamin A (VA) status (total body stores; TBS) and the efficacy of VA intervention programs after ingestion of labeled VA. Researchers have been concerned that dietary VA intake between dose administration and blood sampling (i.e., during isotope equilibration) might significantly affect TBS predictions. We recently published equations (Green et al. J Nutr 2016;146:2137–42 and 2407–11) developed from the “Olson equation” (Furr et al. Am J Clin Nutr 1989;49:713–6) that predict TBS from 1/specific activity (SA p ; tracer: tracee) of plasma retinol 4 d after dosing plus several correction factors related to dose absorption and retention, equilibration of tracer between plasma and stores, and VA catabolism during equilibration; we showed that the measured variable, SA p , is the factor in the equation that has the greatest impact on TBS. To address the concern about VA intake after dosing, we used the powerful tool of model‐based compartmental analysis to simulate the effect of restricted VA intake (0 μg/d) vs. recommended intakes (400 μg/d for children and 800 μg/d for adults) on SA p at 4 d, assuming VA fractional catabolic rates of 2.2%/d (children) or 0.5%/d (adults). Intakes of 400 or 800 μg/d, compared with 0, resulted in dilution of 4 d SA p by <5%. Our results indicate that the effect of intake is small because dietary VA input represents <10% of total daily retinol input into plasma. We also further examined the contribution of 1/SA p to TBS predictions derived by model‐based compartmental analysis (WinSAAM) for data from three adult populations (Cifelli et al. J Nutr 2008;138:971–7 and Green et al. J Nutr 2016;146:2129–36). These analyses indicate that, across a wide range of TBS (35 – 2500 μmol), 1/SA p at 4 d and TBS were significantly correlated and the measured variable accounts for the vast majority of the variance in model‐determined TBS. Our results suggest that newly ingested dietary VA is a minor contributor to total daily plasma retinol input compared with VA in readily exchangeable storage pool(s); in other words, VA in exchangeable pools has a greater impact on SA p at 4 d after tracer dosing than VA intake during this period. In conclusion, these data suggest that, at VA intakes near the RDA for children and adults, dietary VA does not dilute plasma retinol specific activity at 4 d after dosing to any substantial degree and thus should have a minimal effect on this variable in the RID equation. Support or Funding Information International Atomic Energy Agency, Bill and Melinda Gates Foundation (149953), and College of Health and Human Development, Penn State University