Iron Status and Associated Malaria Risk Among African Children

Abstract

Background: It remains unclear whether improving iron status increases malaria risk, and few studies have looked at the effect of host iron status on subsequent malaria infection. We therefore aimed to determine whether a child's iron status influences their subsequent risk of malaria infection in sub-Saharan Africa.

Methods: We assayed iron and inflammatory biomarkers from community-based cohorts of 1309 Kenyan and 1374 Ugandan children aged 0-7 years and conducted prospective surveillance for episodes of malaria. Poisson regression models were fitted to determine the effect of iron status on the incidence rate ratio (IRR) of malaria using longitudinal data covering a period of 6 months. Models were adjusted for age, sex, parasitemia, inflammation, and study site.

Results: At baseline, the prevalence of iron deficiency (ID) was 36.9% and 34.6% in Kenyan and Ugandan children, respectively. ID anemia (IDA) affected 23.6% of Kenyan and 17.6% of Ugandan children. Malaria risk was lower in children with ID (IRR, 0.7; 95% confidence interval [CI], 0.6, 0.8; P < .001) and IDA (IRR, 0.7; 95% CI, 0.6, 0.9; P = .006). Low transferrin saturation (<10%) was similarly associated with lower malaria risk (IRR, 0.8; 95% CI, 0.6, 0.9; P = .016). However, variation in hepcidin, soluble transferrin receptors (sTfR), and hemoglobin/anemia was not associated with altered malaria risk.

Conclusions: ID appears to protect against malaria infection in African children when defined using ferritin and transferrin saturation, but not when defined by hepcidin, sTfR, or hemoglobin. Additional research is required to determine causality.

Clinical trials registration: ISRCTN32849447.

Keywords: African children; iron deficiency; iron status; malaria risk.

Figure 1.

The effect of iron status on subsequent malaria. A, Scatter plots of iron biomarkers stratified by no subsequent malaria or one or more subsequent malaria episodes. Horizontal line indicates mean while vertical line indicates standard deviation. P value was derived from Poisson regression model. B, Adjusted incidence rate ratios for the effect of iron biomarkers on subsequent malaria episodes. Circle marker indicates overall, diamond Kenya and square Uganda. Labels indicate incidence rate ratio and 95% confidence intervals. Poisson regression models were adjusted for age, sex, parasitemia, inflammation, length of follow-up and study site. Maximum length of follow-up was 6 months. Abbreviations: sTfR, soluble transferrin receptor; TSAT, transferrin saturation.

Figure 2.

Kaplan-Meier curves of time to first malaria episode according to (A) iron deficiency (ID) defined by low ferritin, (B) ID defined by transferrin saturation (<10%), (C ) ID anemia, and (D) anemia. P values were derived from log-rank tests for equality of survivor functions. Abbreviations: ID, iron deficiency; IDA, iron deficiency anemia; TSAT, transferrin saturation.

Figure 3.

Metaanalysis of observational studies examining the relationship between iron deficiency (ID) and malaria risk. Study-specific estimates and their relative contribution (percentage weight and sample size) to overall estimates are shown. Definitions of ID varied by study: Nyakeriga et al 2004 [11], ferritin <12 µg/L plus transferrin saturation <10%; Jonker et al 2012 [12], ferritin <30 µg/L; Gwamaka et al 2012 [13], ferritin <30 µg/L if C-reactive protein (CRP) <8.2 mg/L or ferritin <70 µg/L if CRP >8.2 mg/L; Barffour et al 2017 [14], ferritin <12 µg/L in children aged <5 years or <15 µg/L in children aged ≥5 years; and current study, ferritin <12 µg/L or <30 µg/L if CRP >5 mg/L in children aged <5 years or <15 µg/L in children aged ≥5 years. Abbreviations: CI, confidence interval; IRR, incidence rate ratio.