Background
Vitamin A deficiency can cause anemia as the nutrient is essential for hematopoiesis, mobilization of iron store and immunity. Nevertheless, clinical trials endeavored to evaluate the effect of Vitamin A Supplementation (VAS) on hemoglobin concluded inconsistently. Accordingly, the objective of the current study is to assess the effect of single high dose VAS on the hemoglobin status of children aged 6–59 months.

Methods
The study was conducted based on the data of Ethiopian Demographic Health Survey 2011. The data from 2397 children aged 6–59 months who received a single dose of 30 or 60 mg of VAS (depending on age) in the preceding 6 months were matched with similar number children who did not receive the supplement in the reference period. The matching was based on propensity scores generated from potential confounders. Distributions of hemoglobin concentration and risks of anemia were compared between the groups using paired t-test, matched Relative Risk (RR) and standardized mean difference.

Result
The supplemented and non-supplemented groups were homogeneous in pertinent socio-demographic variables. Compared to propensity score matched non-supplemented children, those who received vitamin A had a 1.50 (95% CI: 0.30-2.70) g/l higher hemoglobin concentration (P = 0.014). In the supplemented and non-supplemented groups, the prevalences of anemia were 46.4% and 53.9%, respectively. VAS was associated with a 9% reduction in the risk of anemia (RR = 0.91 (95% CI: 0.86-0.96)). Stratified analysis based on household wealth status indicated that the association between VAS and hemoglobin status was restricted to children from the poor households (RR = 0.74 (95% CI: 0.61-0.90)). Effect size estimates among all children (Cohen’s d = 0.07) and children from poor households (d = 2.0) were modest.

Conclusion
Single high dose VAS among Ethiopian children 6–59 months of age was associated with a modest increase in hemoglobin and decrease in risk of anemia. Household wealth status may modify the apparent association between VAS and hemoglobin status.