South African preschool children habitually consuming sheep liver and exposed to vitamin A supplementation and fortification have hypervitaminotic A liver stores: a cohort study

Background: In some regions, multiple vitamin A (VA) interventions occur in the same target groups, which may lead to excessive stores. Retinol isotope dilution (RID) is a more sensitive technique than serum retinol to measure VA status.

Objective: We evaluated VA status before and after a high-dose supplement in preschool children living in a region in South Africa with habitual liver consumption and exposed to VA supplementation and fortification.

Methods: After baseline blood samples, subjects (46.7 ± 8.4 mo; n = 94) were administered 1.0 μmol [14,15]-13C2-retinyl acetate to estimate total liver retinol reserves by RID with a follow-up 14-d blood sample. Liver intake was assessed with a frequency questionnaire. In line with current practice, a routine 200,000 IU VA capsule was administered after the RID test. RID was repeated 1 mo later. Serum retinyl esters were evaluated using ultra-performance liquid chromatography.

Results: At baseline, 63.6% of these children had hypervitaminosis A defined as total liver retinol reserves ≥1.0 μmol/g liver, which increased to 71.6% after supplementation (1.13 ± 0.43 to 1.29 ± 0.46 μmol/g; P < 0.001). Total serum VA as retinyl esters was elevated in 4.8% and 6.1% of children before and after supplementation. The odds of having hypervitaminosis A at baseline were higher in children consuming liver ≥1/mo (ratio 3.70 [95% CI: 1.08, 12.6]) and in children receiving 2 (4.28 [1.03, 17.9]) or 3 (6.45 [0.64, 65.41]) supplements in the past 12 mo. Total body stores decreased after the supplement in children in the highest quartile at baseline compared with children with lower stores, who showed an increase (P = 0.007).

Conclusions: In children, such as this cohort in South Africa, with adequate VA intake through diet, and overlapping VA fortification and supplementation, preschool VA capsule distribution should be re-evaluated. This trial was registered at https://clinicaltrials.gov/ct2/show/NCT02915731 as NCT02915731.