Red blood cell homeostasis in children and adults with and without asymptomatic malaria infection in Burkina Faso

Abstract

Asymptomatic malaria infections may affect red blood cell (RBC) homeostasis. Reports indicate a role for chronic hemolysis and splenomegaly, however, the underlying processes are incompletely understood. New hematology analysers provide parameters for a more comprehensive analysis of RBC hemostasis. Complete blood counts were analysed in subjects from all age groups (n = 1118) living in a malaria hyperendemic area and cytokines and iron biomarkers were also measured. Subjects were divided into age groups (<2 years, 2-4, 5-14 and ≥15 years old) and clinical categories (smear-negative healthy subjects, asymptomatic malaria and clinical malaria). We found that hemoglobin levels were similar in smear-negative healthy children and asymptomatic malaria children but significantly lower in clinical malaria with a maximum difference of 2.2 g/dl in children <2 years decreasing to 0.1 g/dl in those aged ≥15 years. Delta-He, presenting different hemoglobinization of reticulocytes and RBC, levels were lower in asymptomatic and clinial malaria, indicating a recent effect of malaria on erythropoiesis. Reticulocyte counts and reticulocyte production index (RPI), indicating the erythropoietic capacity of the bone marrow, were higher in young children with malaria compared to smear-negative subjects. A negative correlation between reticulocyte counts and Hb levels was found in asymptomatic malaria (ρ = -0.32, p<0.001) unlike in clinical malaria (ρ = -0.008, p = 0.92). Free-Hb levels, indicating hemolysis, were only higher in clinical malaria. Phagocytozing monocytes, indicating erythophagocytosis, were highest in clinical malaria, followed by asymptomatic malaria and smear-negative subjects. Circulating cytokines and iron biomarkers (hepcidin, ferritin) showed similar patterns. Pro/anti-inflammatory (IL-6/IL-10) ratio was higher in clinical than asymptomatic malaria. Cytokine production capacity of ex-vivo whole blood stimulation with LPS was lower in children with asymptomatic malaria compared to smear-negative healthy children. Bone marrow response can compensate the increased red blood cell loss in asymptomatic malaria, unlike in clinical malaria, possibly because of limited level and length of inflammation. Trial registration: Prospective diagnostic study: ClinicalTrials.gov identifier: NCT02669823. Explorative cross-sectional field study: ClinicalTrials.gov identifier: NCT03176719.

Conflict of interest statement

All authors report no potential conflicts. This work was supported by SYSMEX Europe GmbH. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Figures

Fig 1. Flow diagram study subjects.

Diagram represents how the participants included in the analysis were selected from the two studies as described in the methodology section. Only participants with documented malaria result were considered.

Fig 2. Hemoglobin (Hb), hematocrit (HCT), red blood cells count (RBC), reticulocyte count (RET#) and percentage (RET%) Reticulocyte Production Index (RPI), Immature Reticulocyte Fraction count (IRF#), nucleated RBC count (NRBC#) and Delta-He per clinical group and age category.

Plots display the status of each hematology parameter per age category. In each age category, participants are divided regarding the health status according to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patients with fever are represented by “Clinical malaria”. Whiskers bottom and top limits are 5 th and 95 th percentiles respectively; (): continuous line is used for comparison between clinical status within the same age category and each clinical status was compared with all the other status; (—): dotted line is used for comparison between age category in the “No malaria” group; *: p value with statistically significant difference (p<0.05) between clinical status within the same age category; ▪: p value with statiscally significant difference (p<0.05) between age category in the “No malaria” group; Mann-Whitney U test was used for comparison between groups; yrs: years; #:absolute count; %: percentage count. No malaria Asymptomatic malaria Clinical malaria.

Fig 3. Iron biomarkers, soluble Transferrin Receptor (sTfR), ferritin and hepcidin per clinical group and age category.

Plots display the level of iron biomarker per age category. In each age category, participants are divided regarding the health status according to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patients with fever are represented by “Clinical malaria”. Whiskers bottom and top limits are 5 th and 95 th percentiles respectively; (): continuous line is used for comparison between clinical status within the same age category and each clinical status was compared with all the other status; (—): dotted line is used for comparison between age category in the “No malaria” group; *: p value with statistically significant difference (p<0.05) between clinical status within the same age category; ▪: p value with statistically significant difference (p<0.05) between age category in the “No malaria” group; Mann-Whitney U test was used for comparison between groups; yrs: years. No malaria Asymptomatic malaria Clinical malaria.

Fig 4. Correlation between hemoglobin (Hb) and reticulocytes absolute count (RET#) in no malaria, asymptomatic malaria and clinical malaria cases.

Correlation between the hemoglobin level and reticulocytes absolute count in three groups classified according to the health status and based to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patients with fever are represented by “Clinical malaria”. Spearman test was used to assess the correlation. The line represents the nonlinear fit regression line between the variables.

Fig 5

A. Circulating cytokines levels (IL-6, IFN-ϒ, IL-10 and TNF-α) per clinical group and age category and Ratio IL-6/IL-10 in malaria infected subjects according to age groups. In each age category, participants are divided regarding the health status according to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patients with fever are represented by “Clinical malaria”. Whiskers bottom and top limits are 5th and 95th percentiles respectively; (): continuous line is used for comparison between clinical status within the same age category and each clinical status was compared with all the other status; (—): dotted line is used for comparison between age category in the “No malaria” group; *: p value with statistically significant difference (p<0.05) between clinical status within the same age category; ▪: p value with statiscally significant difference (p<0.05) between age category in the “No malaria” group; Mann-Whitney U test was used for comparison between groups; yrs: years. No malaria Asymptomatic malaria Clinical malaria. B. Correlation between IL-6 levels and parasite density (PD) in asymptomatic malaria and clinical malaria cases as previously classified. Spearman test was used to assess the correlation. The line represents the nonlinear fit regression line between the variables. C. Ex-vivo cytokines production (IFN-ϒ, TNF-α, IL-1β, IL-6 and IL-10) after whole blood stimulation with Lipopolysaccharide (LPS) in asymptomatic children less than 5 years old, that were microscopically malaria positive or negative. Whiskers bottom and top limits are 5th and 95th percentiles respectively; (): continuous line is used for comparison between clinical status *: p value with statistically significant difference (p<0.05) between clinical groups; Mann-Whitney U test was used for comparison between groups. No malaria Asymptomatic malaria.

Fig 6

6. A. The proportion of activated monocytes (RE-MONO%) over total number of monocytes and phagocytozing monocytes count (Phago-MONO) per clinical status and age category. In each age category, participants are divided regarding the health status according to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patienst with fever are represented by “Clinical malaria”. Whiskers bottom and top limits are 5 th and 95 th percentiles respectively; (): continuous line is used for comparison between clinical status within the same age category and each clinical status was compared with all the other status; *: p value with statistically significant difference (p<0.05); yrs: years; Mann-Whitney U test was used for comparison between groups. No malaria Asymptomatic malaria Clinical malaria. B. Correlation between Activated Monocytes count (RE-MONO#), haemoglobin (Hb) and ciculating TNF-α levels in “Asymptomatic malaria” and “Clinical malaria” as previously classified. Spearman test was used to assess the correlation. The lines represents the nonlinear fit regression line between the variables for each clinical status and as indicated in the legend (Red for “Clinical malaria” and Black for “Asymptomatic malaria”).

Fig 7

A. Free-hemoglobin levels (g/L) per clinical status and age category. In each age category, participants are divided regarding the health status according to the case definitions whereby healthy smear-negative subjects are represented by “No malaria”, smear-positive asymptomatic infections represented by “Asymptomatic malaria” and smear-positive patienst with fever are represented by “Clinical malaria”. Whiskers bottom and top limits are 5 th and 95 th percentiles respectively; (): continuous line is used for comparison between clinical status within the same age category and each clinical status was compared with all the other status; *: p value with statistically significant difference (p<0.05); yrs: years; Mann-Whitney U test was used for comparison between groups. No malaria Asymptomatic malaria Clinical malaria. B. The correlation between the free-haemoglobin (Free-Hb) and the parasite density (PD) in “Asymptomatic malaria” and “Clinical malaria” as previously classified. Spearman test was used to assess the correlation. The line represents the nonlinear fit regression line between the variables.