Micronutrients Associated With Anemia in School-age Children and Adolescents 2005-2018: Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) Project

Abstract

Background: School-age children and adolescents may be at risk of anemia through demands on micronutrients required for growth and maturation.

Objectives: This multicountry analysis examined the burden of anemia in children aged 5-19 y by sex and age category and associations with micronutrient deficiencies, inflammation, and BMI.

Methods: Children aged 5-19 y from surveys in the Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) Project were included with hemoglobin, ≥1 micronutrient (iron, vitamin A, folate, vitamin B₁₂, or zinc) and inflammation biomarker, and n > 100 per survey. Factors with bivariate relationships with anemia (P < 0.1) were included in multivariable modified Poisson regression models to examine the attributable burden of anemia.

Results: This analysis included 54,534 children from 17 surveys in 16 countries (16 surveys for 15-19 y; 9 surveys for 10-14 y; 8 surveys for 5-9 y). Median overall anemia prevalence was 16% (range: 5% in Ecuador, United Kingdom, and United States to 59% in Côte d'Ivoire) with the highest burden in 15-19-y-old females (24%). In most surveys, anemia prevalence did not differ by sex for children aged 5-14 y, and median anemia prevalence was lower in children aged 10-14 y (7%) than in those aged 5-9 y (9%) or 15-19 y (22%). In most surveys, higher anemia prevalence was associated (P < 0.05) with iron deficiency (15%) [prevalence ratio (PR): 1.6-14.2; 5-9 y, 4/7 surveys; 10-14 y, 6/6 surveys; 15-19 y, 13/14 surveys), vitamin A deficiency (2%) (PR: 1.8-3.0; 5-9 y, 2/2 surveys; 10-14 y, 2/3 surveys; 15-19 y, 2/3 surveys), and inflammation (13%) (PR: 1.4-2.4: 5-9 y, 4/4 surveys; 10-14 y, 2/4 surveys; 15-19 y, 6/8 surveys). Folate, vitamin B₁₂, zinc, and BMI had weak, variable associations with anemia.

Conclusions: Iron deficiency and vitamin A deficiency are consistently associated with anemia in school-age children and adolescents, whereas inflammation and other micronutrients had context-dependent associations. This research underscores the importance of examining multiple micronutrients associated with anemia in the context of factors such as country, age, and sex.

Keywords: Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA); adolescents; anemia; body mass index (BMI); global; iron deficiency (ID); micronutrients; nutrition; school-age children; youth.

FIGURE 1

Flow chart of school-age children and adolescents aged 5–19 y included in analysis of inflammation and micronutrients associated with anemia: BRINDA Project. AGP, α-1-acid glycoprotein; BRINDA, Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia; CRP, C-reactive protein; N, number of data sets; n, number of children; RBC, red blood cell; RBP, retinol-binding protein; sTfR, soluble transferrin receptor.

FIGURE 2

Severity of anemia by sex in children aged (A) 5–9 y, (B) 10–14 y, and (C) 15–19 y in the BRINDA Project. Severity of anemia was assessed using WHO cutoff values for altitude and smoking-adjusted hemoglobin [severe: <80 g/L (all); moderate: ≥80 g/L and <110 g/L (all); mild: ≥110 g/L and <115 g/L (children <12 y), ≥110 g/L and <120 g/L (children 12–14 y or females ≥15 y), ≥110 g/L and <130 g/L (males ≥15 y)] [28]. Only 2 surveys of 15–19-y-old children assessed hemoglobin in males. Sample size for each survey and age group is presented in Table 1, as all children included in analysis had complete measures for hemoglobin and inflammation. BRINDA, Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia.

FIGURE 3

Burden of anemia attributable to iron deficiency in children aged (A) 5–9 y, (B) 10–14 y, and (C) 15–19 y in the BRINDA Project. BRINDA inflammation-adjusted iron deficiency [11,36,38] was assessed using ferritin <15 μmol/L or sTfR >8.3 mg/L when ferritin was unavailable. Population attributable fraction (PAF) was estimated using prevalence ratios [22] from Table 3 and presented in this figure when P < 0.05 and estimates were stable. Supplemental Table 4 shows PAF for all micronutrients and inflammation. Sample size for each PAF model is presented in Table 3, and surveys are listed in order of country GDP. BRINDA, Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia; GDP, gross domestic product; sTfR, soluble transferrin receptor.