Variation in serum retinol concentrations overtime in Zambian children during an intervention study (624.27)

Serum retinol concentrations are often used to evaluate intervention programs. However, they are influenced by many factors including inflammation, co‐micronutrient deficiencies, and to a small extent, recent dietary intake. Further, during healthy states, serum retinol concentrations are homeostatically controlled and maintained at a constant level over a broad range of liver retinol concentrations. This study evaluated the variations in serum retinol concentrations during an intensive intervention study. The main study design was such that 4 serum retinol concentrations were successfully determined in the same Zambian children (n = 127) over a 5‐month period. After a baseline blood, children (age 5.8 +/‐ 1.0 y; n = 140) were randomized to treatments of white maize and placebo, biofortified orange maize and placebo, and white maize with a daily 400 ug retinol equivalents supplement. A follow‐up blood was two weeks later before the 15‐week intervention. Finally, blood samples were obtained at 18.5 and 20.5 weeks after recruitment. The range of serum retinol concentrations by time was 0.92 +/‐ 0.26 (time 4) to 1.07 +/‐ 0.29 umol/L (time 2). The coefficient of variation was 6.5% for the mean serum retinol over the four times. The child’s individual CV, on the other hand, was 17 +/‐ 7% over the four time periods. Therefore, the variation by individual child was much higher than the mean by time. Moreover, only 9/127 of the subjects showed mean serum retinol values <0.7 umol/L; yet, 44/127 had at least 1 time‐point where their serum retinol concentration was below or equal to that cutoff. An individual’s high retinol concentration in one time‐point may outweigh the others. Further, investigation will be done by serum volume analyzed. Grant Funding Source : Supported by HarvestPlus, UW‐Madison Global Health and NIH