Infection by Helicobacter pylori cytotoxin-A-associated antigen-positive strains is associated with iron deficiency anemia in a longitudinal birth cohort in Brazil

Globally, anemia affects 40.0% of the child population.¹ In developing countries, anemia is a major public health problem in infants and young children. Iron is the most common cause of single-nutrient deficiency¹ due to persistent negative iron balance in association with rapid child growth that leads to high iron requirements. In parallel, in developing countries, Helicobacter pylori infection prevalence rates are as high as 70.0-90.0%, and the infection is mainly acquired early in childhood.^1,2 This lends considerable support to the hypothesis that H. pylori infection plays a role in anemia in infants / young children in low resource settings. However, the mechanism by which the infection contributes to anemia remains to be elucidated.

Gastric acidity is essential for iron absorption by the reduction and solubilisation of the non-heme iron, the main form of iron in the dietary intake, and consequently for the iron duodenal absorption.

H. pylori cagA that encodes the major H. pylori virulence factor, cytotoxin-associated gene A (CagA), is a component of the cag pathogenicity island (cag PAI). cag PAI genes are mechanistically involved in the repression of the transcription of H⁺K⁺a (H⁺K⁺ alpha-subunit) of the parietal cell H⁺K⁺-ATPAse leading to the hypochlorydria that impairs iron absorption.³ Additionally, CagA is able to alter the polarity of the transferrin / transferrin receptor-iron uptake system, which allows the bacterium to shuttle iron across the epithelium.^4-6

Notably, the H. pylori virulent proteins, CagA and vacuolating cytotoxin (Vac) A, may act in concert to alter the host transferrin trafficking allowing the bacterium to use the cell surface as a replicative niche,⁵ which may explain why CagA⁺ strains colonise the stomach at higher density.

Since there are no studies evaluating CagA-positive H. pylori infection and anemia in infants and young children, we investigated whether infection with CagA-positive H. pylori strains is associated with anemia in a longitudinal birth cohort of children living in a poor community in Brazil.

The prospective birth cohort study was made up of 123 children from randomly selected healthy mothers who had attended the antenatal follow-ups, living in a poor urban community in Fortaleza, Ceará, in northeastern Brazil. We evaluated the influence of H. pylori infection upon the risk of iron deficiency (ID) and iron deficiency anemia (IDA) in young children. Exclusions included preterm infants, low birth weight (<2500 g), severe disease that required hospitalization or chronic severe illness, among them congenital hemolytic anemia. The study was approved by the National Ethics Committee on Research of the Health Ministry of Brazil / European Union Ethics Committee.

Blood samples were obtained at 6-month intervals from six until 36/40 months of age for iron and CagA status assessment. ¹³C-urea breath test (¹³C-UBT) and stool antigen test were performed for H. pylori diagnosis.⁷ Children were considered H. pylori-positive when both tests were positive and H. pylori-negative when both tests were negative. A questionnaire with demographic and clinical data (illness, antibiotic use) was administered at 3-month intervals.

CagA IgG serological status was determined by a validated immunoassay (cut-off value=7.5 units),⁸ using recombinant CagA protein (kindly provided by Dr G. del Guidice, Novartis Vaccines, Siena, Italy).

Hemoglobin and hematocrit were determined in an automated electronic counter (Sysmex XT 1800i; Sysmex Corporation, Kobe, Japan). Serum concentrations of ferritin, iron and total iron binding capacity (TIBC) were assayed as previously described,⁹ as well as transferrin saturation (TSat).

Data were analyzed by the public domain statistical software R version 4.0.2 (www.r-project.org)¹⁰ and the SPSS statistical software package version 20.0 (SPSS Inc., Chicago, IL, USA). Two-tailed Student t test or Mann-Whitney U test, as well as x²-test with Yates’ correction or Fisher’s exact test was employed as indicated. Log-rank test was assessed to estimate the Kaplan-Meier survival. P≤0.05 was considered statistically significant.

Notably, to avoid maternal antibody interference, only children ≥11 months of age were followed at different intervals at 18, 24, 30 and 36-40 months - mean of evaluations of 4.68 times along the time, which gave us the opportunity to identify whether ID/IDA followed the acquisition of CagA-positive H. pylori infection. The cut-offs for diagnosing anemia was hemoglobin <105.0 g/L (children aged 6-26 months) and <115.0 g/L (children aged 24-36 months). Anemia was defined as IDA when TSat <15% or serum ferritin levels ≤7 ng/mL or serum iron levels and <40 μg/dL.

Of the 102 children (54 girls) older than 11 months who neither had other infections nor had been previously treated with iron, 30 (29.4%) had IDA and 17 (16.7%) had ID, in at least 2 evaluations; 55 (53.9%) did not have ID/IDA in all evaluations. In the group of IDA, 23 (76.7%) and 7 (23.3%) children were CagA positive and CagA-negative, respectively. A similar result was observed in the ID group (13 [76.5%] and 4 [23.5%], respectively, with no difference between groups [two-tailed Fisher test, P=1.0]). In the ID/IDA negative children, 11 (30.9%) were CagA positive and 44 (69.1%) were CagA-negative; these values became significantly different when compared with either ID (P<0.001, two-tailed Fisher’s exact test) or IDA (Odds Ratio [OR] = 13.14, 95%Confidence Interval [CI] = 4.49-38.45, P<0.001, two tailed x²-test).

Because the data were longitudinal, they were analyzed by regression models of generalized estimating equation (GEE).¹¹ To assess the fit of the models, quality indications of adjustment of the models in the adjustment phase were evaluated by the area under the ROC curve. The quality measure was ≥0.72 in all GEE models.

A logistic marginal regression was adjusted to evaluate the categorical dependent variable CagA status. CagA-positive H. pylori infection was significantly more frequent in children with ID/IDA than in those without anemia (Table 1).

Next, to explain the categories in function of the time and of each individual variable, other GEE models were used to identify the predictors that were independently affected by CagA. In the univariate analysis, male gender, hemoglobin, iron and ferritin values were selected for the multivariate analysis. In the initial multivariate model, predictors with P values ≤0.20 were included in the final model. Hemoglobin and iron values remained independent and negatively associated with CagA IgG seropositive status (Table 2).

To demonstrate that the association between anemia and CagA seropositive status was not due to an overlapping between CagA seropositive status, we used GEE models to evaluate H. pylori-positive and H. pylori-negative children in the group of CagA-negative children. H. pylori infection was not associated with anemia in this group confirming absence of overlapping between CagA and H. pylori-positive status. However, H. pylori infection was associated with age, which was expected because in childhood the prevalence of infection increases as children grow older (Online Supplementary Table S1).

When the hemoglobin concentration was below the reference values and the results were confirmed in the subsequent evaluation, the child was treated with oral ferrous sulphate (3 mg/kg/dose, maximum, 60 mg/dose, twice daily). Among 30 IDA children, 2 recovered from IDA after they had spontaneously cleared the infection (negative ¹³C-UBT and stool antigen test, and CagA IgG serum reversion) and remained without H. pylori infection and IDA in further evaluations. Among 28 children who were treated, 23 returned to be re-evaluated. The levels of hemoglobin, iron and ferritin increased to normal values in 19 (82.61%) children. Therapeutic failure was observed in 4 (17.39%) children; 3 of them remained CagA IgG-positive until 40 months of age despite receiving iron replacement therapy. Figure 1 shows the Kaplan-Meier curves demonstrating that the time required to achieve increased ferritin and iron levels as well as IDA cure in the children infected with CagA-positive H. pylori strains (50% of probability of IDA cure after around 16 months) is longer (P<0.001) than that observed in the CagA-negative children (i.e., around six months).

The role of CagA in iron acquisition should be considered in the light of the strong association between anemia and CagA-positive status we observed. Because gastric acidity is essential for the reduction and solubilisation of non-heme dietary iron, reduced non-heme iron absorption may be attributable to the gastric hypochlorhydria that occurs in the early phase of H. pylori infection.^12-14

We speculate that CagA may impair the activity of the endogenous alpha-subunit (H⁺,K⁺α) of the gastric H⁺,K⁺-ATPase, the parietal cell enzyme that mediates acid secretion³ and it increases iron uptake via transferrin endocytosis, as well as lysosomal iron thorough augmented expression of H-ferritin. Although in this study the CagA status was evaluated by serology, which would be considered one limitation, the rigorous experimental protocol and the data analysis strengthen the results. The test has been previously validated for children using a culture as a gold standard, and the agreement between ELISA and Western blotting in detecting IgG anti-CagA was 100%.⁸

Table 1.Association between iron deficiency and iron deficiency anemia and HP⁺/CagA⁺ status and HP⁺/CagA^- * status in young children.

Table 2.Variables associated with serum CagA IgG positive status in infants and young children with iron deficiency and iron deficiency anemia.

Figure 1.Estimation of iron deficiency anemia cure over time in the H. pylori CagA-positive children compared with CagA-negative children by Kaplan-Meier.

To the best of our knowledge, this is the first study to demonstrate that infection with CagA positive H. pylori strains is an independent risk factor for IDA in infants and young children. In conclusion, young children infected with CagA-positive H. pylori strains are at increased risk of developing iron deficiency anemia. Further exploration of our findings towards a mechanistic understanding of the underlying pathogenesis of CagA-associated IDA is warranted.

Footnotes

• Received April 19, 2024
• Accepted August 5, 2024

Correspondence

D.M. DE MAGALHÃES QUEIROZ - dqueiroz@medicina.ufmg.br

Disclosures

No conflicts of interest to disclose.

Funding

Acknowledgments

We thank Dr G. del Giudice, Novartis Vaccines, Siena, Italy, who kindly provided recombinant CagA protein.