OBJECTIVE: This study evaluates the association of serum retinol, hepcidin levels, and anemia in children. METHODS: This cross-sectional study included 312 children, ages 6 to 59 mo, from Rio de Janeiro, Brazil. The association between hepcidin and retinol levels, hematologic parameters, and body mass index (BMI) was analyzed using a generalized linear model with and without adjustment for C-reactive protein (CRP) level. Logistic regression analysis was used to test anemia as an outcome and serum retinol level as a predictive variable using the odds ratio (OR) function. RESULTS: Anemia was present in 14.6% of the children, 5.8% presented iron deficiency anemia, and 9.6% had vitamin A deficiency. The increase in serum retinol levels reduced the chances of anemia (OR = 0.13; confidence interval = 0.29-0.59). When CRP level was not adjusted for in the multiple regression analyses, retinol, ferritin levels, and BMI/age were predictors of serum hepcidin levels (β = -3.36, 0.14, 1.02, respectively; P = 0.032). Accordingly, serum retinol levels were inversely associated with CRP levels (β = -0.025 and P < 0.001). CONCLUSIONS: The association between serum retinol and hepcidin levels in children ages 6 to 59 mo seems to be dependent on inflammation. Taken together, the results reinforce the need for the development of further studies to better understand the relationship between vitamin A and anemia of inflammation.
OBJECTIVE: This study evaluates the association of serum retinol, hepcidin levels, and anemia in children. METHODS: This cross-sectional study included 312 children, ages 6 to 59 mo, from Rio de Janeiro, Brazil. The association between hepcidin and retinol levels, hematologic parameters, and body mass index (BMI) was analyzed using a generalized linear model with and without adjustment for C-reactive protein (CRP) level. Logistic regression analysis was used to test anemia as an outcome and serum retinol level as a predictive variable using the odds ratio (OR) function. RESULTS: Anemia was present in 14.6% of the children, 5.8% presented iron deficiency anemia, and 9.6% had vitamin A deficiency. The increase in serum retinol levels reduced the chances of anemia (OR = 0.13; confidence interval = 0.29-0.59). When CRP level was not adjusted for in the multiple regression analyses, retinol, ferritin levels, and BMI/age were predictors of serum hepcidin levels (β = -3.36, 0.14, 1.02, respectively; P = 0.032). Accordingly, serum retinol levels were inversely associated with CRP levels (β = -0.025 and P < 0.001). CONCLUSIONS: The association between serum retinol and hepcidin levels in children ages 6 to 59 mo seems to be dependent on inflammation. Taken together, the results reinforce the need for the development of further studies to better understand the relationship between vitamin A and anemia of inflammation.