The aim of this work was to assess the association between childhood SAM and long-term hematological disorders through the analysis of blood counts of a group of adults with a history of SAM living in the eastern DRC.
This investigation was a sub-study of a larger survey (Lwiro Cohort Study) that attempts to contribute to understanding the long-term effects of SAM during childhood in LICs.
Our results suggest that adults with a history of SAM have a higher WBC count overall with a neutrophilic predominance. In addition, they have a higher platelet count than non-exposed individuals. However, they have no particular risk of anemia.
As regards the WBC line, our data suggest that former SAMs have a high risk of hyperleukocytosis in adulthood and have high mean WBCs with a neutrophilic predominance. Indeed, studies of the hematological profile of children in the acute phase of SAM show almost systematic hyperleukocytosis compared to non-SAM children (19-21,43-47). All these studies ascribed this hyperleukocytosis to the high susceptibility to infections that malnourished children develop. These children have less activation of macrophagic cells linked to a low expression of GM-CSF (Granulocyte-Macrophage Colony-Stimulating Factor) and M-CSF (Macrophage Colony-Stimulating Factor), resulting in a stunted antibacterial immune response (25). Indeed, it has been shown that SAM leads to atrophy of lymphoid organs (thymic and splenic), which blunts the antigenic response (5,48,49). In addition, a study conducted in Kenya demonstrated an improvement in phagocytic function following treatment of SAM according to WHO guidelines, although lymphocyte counts and free radical production remained lower than controls (50).
Furthermore, high proportion of thinness in exposed adults compared to unexposed in this study suggest a prolonged effect of SAM on nonlinear growth, as observed in other studies of this cohort (38,39). This may have contributed to persistent impairment of cellular immunity or other defense functions, such as phagocytosis and humoral-mediated immunity. In Zimbabwe, a study in an 18-month-old population with a high prevalence of stunting suggests that stunted children have a dysfunctional LPS (liposaccharide)-mediated response to Escherichia coli endotoxin compared to children without malnutrition. This would predispose them to infections because of the significant decline in antibacterial immune function (48).
In addition, children with SAM develop environmental enteropathies with altered intestinal mucosa and barrier function that favor systemic microbial translocation from the gut with substantial bacterial sepsis (51). A study carried out in Chile, in children recovering from SAM, revealed persistent deterioration of the intestinal mucosa although there was an improvement on electron microscopy of the border brush. Intestinal atrophy exposes to microbial translocation from the gut microbiota to the rest of body (52). It has even been described that despite nutritional rehabilitation with normalization of anthropometric parameters, these children kept very high levels of inflammatory cytokines related to LPS activation of gram-negative bacteria from the gastrointestinal tract, even when there were no clinical signs of gut infection (50-53).
The neutrophilia in this context would also be partly secondary to neutrophilic demargination and mobilization of the medullary neutrophils pool induced by calciprotein produced by injured endothelial cells in response to LPS (51). Following correction of anthropometric parameters, there remained markers of damage to the intestinal microbiota and loss of integrity of the intestinal mucosa, with impaired micronutrients absorption (48,51,52). This could also contribute to higher susceptibility to infections from the gut but also to risk of relapse in former malnourished children due to micronutrient deficiency. Hence these children in acute post-nutritional rehabilitation follow-up have a high mortality rate linked to infectious diseases (50,54,55). This was suggested in a study conducted on the Lwiro cohort, in which the first cause of death during the first 5 years following nutritional rehabilitation were linked to infectious diseases, which were also responsible for a significant rate of relapse (39). Finally, micronutrient deficiency associated with pattern of infections promotes chronic oxidative stress, which has deleterious effects on several organs and growth (56,57).
The main cohort of the Lwiro study showed a correlation between SAM and the occurrence of visceral obesity, metabolic syndrome, and carbohydrate homeostasis disorder in adulthood (38).
These diseases are among the chronic NCDs that are frequently associated with a chronic subclinical low-grade inflammation component (56). The metabolic profile of children with SAM differed from that of healthy children and did not normalize after growth resumption, as observed in several studies conducted in LICs (52,56,57). Apart from the infectious context, joint responses between neutrophils and platelets during chronic inflammatory diseases could contribute to the combination of elevated neutrophil count and platelet counts reported here.
A literature review published in January 2022 examined the involvement of neutrophils in chronic inflammatory diseases (58). It describes neutrophilic infiltration of adipose tissue in obesity-related diseases. This results in neutrophil activation with secretion of neutrophil elastase and myeloperoxidase. Activated neutrophils not only demarginate and mobilize from their marrow pool, but also induce activation of other mononuclear cells, namely monocytes and macrophages, by degranulation (58). Angèle Gros and coll. demonstrated that platelets and neutrophils are involved in the development of chronic inflammatory diseases (59). Several experimental and human models described in this review suggest platelet receptors involvement in platelet and leukocyte recruitment, but also in vascular permeability, via formation of platelet-neutrophil or platelet-macrophage complexes, present in chronic inflammatory diseases such as rheumatoid arthritis, or in ischemic stroke and atherosclerosis in general. These complexes are linked to the expression of platelet receptors as well as to cytokines secreted by mononuclear cells, mainly neutrophils. Furthermore, in mouse models of T2 Diabetes Mellitus, platelet counts are elevated upon secretion of S100A8/A9 by neutrophils, which increase liver production of thrombopoietin (60). In addition, atherosclerotic patients have neutrophil-platelet complexes that promote inflammatory response based on neutrophil migration and activation (61). Thus, a trend to elevated neutrophilia and platelets could be viewed in the Lwiro cohort as markers of inflammation in the context of metabolic syndrome and central obesity, two chronic non-communicable inflammatory diseases.
Nevertheless, our results do not meet the laboratory model hypothesis of the impact of SAM on hematopoiesis. The latter, based on murine and porcine models, suggests that undernutrition would alter not only cell cycling of hematopoietic stem cells but also their microenvironment. This is not supported by the results obtained in our study. In fact, Aparecida et al. and Borelli et al. have shown that SAM leads to a reduction of hematopoietic stem cells and pluripotent cells not by apoptosis but by cell cycle arrest (24,25,44). This is explained by a marked reduction in the expression of genes coding for cell differentiation and maturation (44). Thus, pluripotent cells lose their capacity for self-renewal and restoration of hematopoiesis. These different laboratory models do not, however, incriminate pluripotent cells alone in the overall malnutrition-related marrow hypoplasia.
In terms of the RBCs line (RBC, Hb, Wintrobe’s constant and risk of anemia), no statistically significant difference was observed between exposed and non-exposed. This lack of difference could result from the benefit of nutritional rehabilitation, which would have provided all necessary nutrients for subsequent RBC production, such as amino-acids, iron and vitamins.
This observation was highlighted in a study conducted in Palestine on the evaluation of the impact of a humanitarian project to combat SAM (62). In this study, the prevalence of anemia in a group of malnourished people was almost halved from 30.1% to 18.8% following food supplementation (62). Another study conducted in Burkina Faso, also on the evaluation of the impact of food supplementation, noted an 8% decrease in the prevalence of anemia after nutritional intervention (63,64).
Nevertheless, these results are surprising, to say the least, given that a recent study carried out in the same region in children under 5 years of age reported a high prevalence of chronic iron and zinc deficiencies (65). One would expect changes in RBC line secondary to such deficiencies, especially since about half of the exposed cases had an unsatisfactory dietary situation, characterized by a traditional diet low in protein and other essential micronutrients (little meat and fish, low in eggs and dairy products, low in oils and poor vegetable diversification). It would also be assumed that, given the proportion of the population that was thin among exposed people, there would be an adaptation to the drop-in hemoglobin levels, but this was not the case. A study carried out in the Gambia with the aim of establishing hematological reference values for West African populations noted that RBC values increased with age and were not influenced by nutritional status (10).
Finally, in contrast to unexposed, exposed adults had a high platelet count. We would readily attribute it to the chronic inflammatory disease model. However, it cannot be ruled out that it was also linked to martial deficiency. The mechanism involved in thrombocytosis in iron-deficient subjects remains unclear. Recent studies have shown that an iron-deficient environment preferentially diverts megakaryocytic-erythroid progenitors (MEPs) towards commitment to the megakaryocytic lineage, in addition to altering their metabolism, attenuating extracellular signal-regulated kinase (ERK) signaling and decreasing MEPs proliferation (66). Also, a tendency for elevated platelets in exposed cases could be explained in part by inflammation due to a prevalence of infections but also to the acquisition of a metabolic syndrome phenotype (38).
Regarding SAM subtypes, our results suggest that former marasmus-exposed were three times more likely to develop hyperleukocytosis in adulthood than other malnourished children. Bone marrow analyses from malnourished children in Turkey showed systematic hyperleukocytosis, unrelated to type of malnutrition (18). Only in those with kwashiorkor was a lower leukocyte count described (18). This human model suggestive of increased proliferative activity in the granular lineage does not corroborate the data from laboratory models already discussed, but is closer to the results obtained on this cohort, albeit with some hindsight. Furthermore, it was also shown that patients with a history of marasmus were more likely to develop NCDs later than those with kwashiorkor (67). As stated above, hyperleukocytosis could be linked to low-grade chronic inflammation (58,59,68). However, we did not find any data from studies that analyzed bone marrow tissue of former malnourished people that would explain the long-term impact of SAM on the hematopoietic process.
Our study has some obvious limitations:
First, there is a survival bias: only those who survived to adulthood (of 30 years) and were still present in local villages two to three decades after the SAM episode were studied. Nevertheless, there is no obvious reason to believe that the association between nutritional status of children hospitalized for SAM and hematological profile would have been different among those lost to follow-up, as the characteristics at hospital admission did not differ between the lost and traced subjects (39).
Secondly, the sample size of our study was not very large. Some differences could have become statistically significant with a larger sample size.
Thirdly, we did not have data on bone marrow, spleen and liver status nor on other markers of inflammation and infection. Indeed, changes in these parameters could have contributed to the hematological disturbances independently of SAM.
Fourthly, we did not have serum iron status and transferrin saturation data, which are related to platelet aggregation parameters, nor had measures of micronutrients involved in hematopoiesis.
Fifth, it is questionable whether all non-exposed people recruited were healthy. Although they did not experience kwashiorkor nor were treated for SAM, it is possible that some of them had some degree of malnutrition related to poor socio-economic conditions of the region, which did not result in hospitalization. This permanent unfavorable situation in which the two groups evolved may have contributed substantially to attenuating possible differences in many biological markers under study. Sixthly, the design of our study was unable to establish causality and it is therefore difficult to separate mechanisms due to SAM per se from mechanisms due to the persistent effects of the early childhood environment or the persistence of a precarious environment in which the subject continued to live on the observed abnormalities.