Biofortified yellow cassava and vitamin A status of Kenyan children: a randomized controlled trial uri icon


  • Background: Whereas conventional white cassava roots are devoid of provitamin A, biofortified yellow varieties are naturally rich in beta-carotene, the primary provitamin A carotenoid.
  • Conclusions: In our study population, consumption of yellow cassava led to modest gains in serum retinol concentration and a large increase in beta-carotene concentration. It can be an efficacious, new approach to improve vitamin A status.
  • Design: We randomly allocated 342 children aged 5-13 y to receive daily, 6 d/wk, for 18.5 wk 1) white cassava and placebo supplement (control group), 2) provitamin A-rich cassava (mean content: 1460 mu g beta-carotene/d) and placebo supplement (yellow cassava group), and 3) white cassava and beta-carotene supplement (1053 mu g/d; beta-carotene supplement group). The primary outcome was serum retinol concentration; prespecified secondary outcomes were hemoglobin concentration and serum concentrations of beta-carotene, retinol-binding protein, and prealbumin. Groups were compared by using ANCOVA, adjusting for inflammation, baseline serum concentrations of retinol and beta-carotene, and stratified design.
  • Objective: We assessed the effect of consuming yellow cassava on serum retinol concentration in Kenyan schoolchildren with marginal vitamin A status.
  • Results: The baseline prevalence of serum retinol concentration <0.7 mu mol/L and inflammation was 27% and 24%, respectively. For children in the control, yellow cassava, and beta-carotene supplement groups, the mean daily intake of cassava was 378, 371, and 378 g, respectively, and the total daily supply of provitamin A and vitamin A from diet and supplements was equivalent to 22, 220, and 175 mu g retinol, respectively. Both yellow cassava and beta-carotene supplementation increased serum retinol concentration by 0.04 mu mol/L (95% CI: 0.00, 0.07 mu mol/L); correspondingly, serum beta-carotene concentration increased by 524% (448%, 608%) and 166% (134%, 202%). We found no effect on hemoglobin concentration or serum concentrations of retinol-binding protein and prealbumin.

publication date

  • 2016
  • 2016
  • 2016