Helicobacter pylori Infection Is Associated With Low Serum Ferritin Levels in Israeli Arab Children—A Seroepidemiologic Study

Abstract

We examined the association between Helicobacter pylori seropositivity and low ferritin levels in sera of 509 Israeli Arab children aged 1 to 19 years. Serum ferritin levels were measured using the immunoturbidity assay. ELISA was used to detect serum IgG antibodies to H pylori and IgG antibodies to H pylori CagA⁺ strains. Low ferritin level was found between 14.5% and 8.6% of the H pylori seropositive and seronegative participants, respectively (P = 0.035). A trend for a higher frequency of low ferritin was detected among H pylori-CagA seropositive children. Multivariate analysis showed an increased frequency of low ferritin only among children younger than or equal to 5 years: adjusted OR 4.62 (95% CI 1.53–13.99) P = 0.007.

Epidemiologic observational studies and intervention trials ^(1–5) reported supportive findings linking between Helicobacter pylori infection and iron stores, whereas other studies could not demonstrate such an association ⁽⁶⁾. These controversial results prompted the present study to be carried out in a pediatric population with enhanced growth and iron demands. We examined the association between low ferritin levels and H pylori seropositivity among Israeli Arab children, a defined population, with a high prevalence of 50% of H pylori infection ⁽⁷⁾ and high anemia rates of approximately 20% ⁽⁸⁾.

MATERIALS AND METHODS

Residual sera of Israeli Arab children from the serum bank of the Israel Center for Disease Control, collected from 2000 to 2001, were examined. Sera were kept frozen (−20°C) until tested. The samples were anonymous and had details only on sex, age, and address (at the level of town). The socioeconomic index of the place of residence defined by the Israel Central Bureau of Statistics ⁽⁹⁾ was used as an indicator for socioeconomic status. In calculating the sample size we estimated 10% and 20% frequencies of low ferritin level in H pylori seronegative and seropositive children, respectively, and have assumed a type 1 error of 5% and 80% power. Given these assumptions, the minimal sample size of each H pylori–infected and –uninfected group was 199 (ie, a total of 398). To allow power for performing a multivariate analysis we studied all of the 509 sera collected in 2000 to 2001. The institutional review board of the Tel Aviv University approved the study protocol.

Serum H pylori–specific IgG antibodies were measured using the Dade Behring's (Marburg, Germany) Enzygnost Anti-Helicobacter pylori II/IgG commercial kit. The kit's sensitivity and specificity in children are 92.7% and 95.7%, respectively. In children younger than 6 years the respective values are 91.6% and 97.1% ⁽¹⁰⁾. Results of ≥0.250 IU were considered H pylori IgG seropositive. In a subanalysis, results of <0.250 IU, between 0.250 and 0.581 IU and ≥0.582 IU, were classified as “negative,” “medium seropositive,” and “high seropositive,” respectively. A random subsample of seropositive sera to H pylori (N = 71) was tested for presence of IgG antibodies to Cag-A H pylori strains, using CagA IgG ELISA kit (Genesis Diagnostics Ltd, UK) following the manufacturer's instructions.

Serum ferritin was tested on the Olympus 5420 and 5430 analyzers according to the manufacturer's (Olympus-Life and Material Science Europa GmbH, Hamburg, Germany) procedures. Ferritin was quantified by the immunoturbidimetric test. Low ferritin level was defined as a serum ferritin level lower than12 μg/L ⁽¹¹⁾.

The data were analyzed with SPSS version 13 (SPSS Inc, Chicago, IL). The differences between H pylori seropositive and seronegative subjects in percentage of low ferritin levels were examined using the χ² test and Fisher exact test when appropriate. Logistic regression models were used to examine the independent association between H pylori infection and low ferritin levels.

RESULTS

The serum samples included in the study (N = 509) belonged to children with a mean age of 9.1 years. Of these children, 257 (50.5%) were females. The H pylori seropositivity was 47.3% (241 subjects). A low ferritin level (ferritin <12 μg/L) was found among 58 (11.4%) of the subjects. In H pylori–seropositive children, 35 (14.5%) had low ferritin levels (ferritin <12 μg/L) compared with 23 (8.6%) among H pylori–seronegative subjects (P = 0.035). When applying the cutoff of ferritin <10 μg/L, 11.2% and 5.6% of H pylori–seropositive and seronegative children had low ferritin levels, respectively (P = 0.022). A dose-response association was observed: 16.1% of the H pylori high-positive subjects had low ferritin levels (<12 μg/L) compared with 12.8% and 8.6% of the H pylori medium-positive and seronegative subjects, respectively (P = 0.024). A similar relation was found when using the cutoff of ferritin <10 μg/L: of the high-positive, medium-positive, and H pylori–seronegative subjects, 11.3%, 11.1%, and 5.6% had low ferritin levels, respectively (P = 0.035). We found a trend of higher frequency of low ferritin levels (<12 μg/L) in H pylori-CagA seropositive subjects compared with H pylori seropositive subjects lacking CagA antibodies, and H pylori–seronegative children: 17.2%, 9.5%, and 8.6%, respectively (P = 0.16). When applying the cutoff of ferritin <10 μg/L, the corresponding values were 13.8%, 9.5%, and 5.6% (P = 0.068).

Stratified Analysis

Sex

In males, 10.7% and 7.1% of the H pylori–seropositive and seronegative subjects had low ferritin levels (<12 μg/L), respectively (P = 0.31). The corresponding values in females were 17.8% and 10.2% (P = 0.07) (Table 1).

TABLE 1: Stratified analysis of the association between H pylori serostatus and low ferritin level by sex, age group, and socioeconomic status

Age

In children aged younger than or equal to 5 years, 20.8% and 5.9% of seropositive and seronegative subjects had low ferritin levels, respectively (P = 0.008). Such an association was not found in children aged 6 to 10 and 11 to 15 years. In the age group of 16+ years, 22.9% and 5.3%, respectively, of the H pylori–seropositive and seronegative subjects had low ferritin levels (P = 0.15) (Table 1).

Socioeconomic Status

H pylori–infected children, living in localities and towns of high or low socioeconomic status, had higher frequency of low ferritin levels compared with noninfected patients; however, these differences were not statistically significant (Table 1).

Multivariate Analysis

We ran separate multivariate analyses in each age group, controlling for sex and socioeconomic status. In children aged 5 years or younger, H pylori infection was associated with an increased frequency for low ferritin levels: adjusted OR 4.62 (95% CI 1.53–13.99) P = 0.007 (Table 2). In children aged 6 to 10 years and 11 to 15 years, there was no significant association between H pylori infection and ferritin levels in the multivariate analysis. In adolescents aged 16+ years, H pylori seropositivity was associated with 4-fold higher odds for low ferritin level, although not statistically significant, adjusted OR 4.19 (95% CI 0.48–36.44) P = 0.19 (Table 2).

TABLE 2: Adjusted odds ratios and 95% CI of the association between H pylori seropositivity and low ferritin level in age groups ≤5 years and 16+ years

DISCUSSION

We found an increased frequency of low ferritin levels among H pylori–seropositive children and a dose-response relation between low ferritin and anti-H pylori serum IgG levels. We assume that H pylori acquired in early childhood could compete with the host for iron and thus may play an important role in the iron stores depletion. It has been already proposed that H pylori infection may be a stressor of decreased iron biomarkers in children and adolescents ^(2,3). The exact mode by which H pylori alter the iron stores is not fully recognized. Higher serum IgG levels may reflect a more severe gastric infection ⁽¹²⁾ and may be linked to a stronger reduction in the serum ferritin levels. A study from Israel showed that Arab children formed a great portion of the subjects with peptic ulcer disease presenting to gastroenterology units ⁽¹³⁾. It is possible that breaks in gastric mucosa may induce gastrointestinal blood loss. Other mechanisms such as decreased absorption of dietary iron and enhanced uptake of the iron by the bacterium were also proposed ⁽¹⁴⁾.

We found higher frequency of low ferritin levels among H pylori-CagA seropositive subjects. It has been shown that CagA-positive strains were associated with duodenal ulcers, intensity, and activity of gastritis in children ⁽¹⁵⁾, suggesting more severe illness. Previous studies showed no association between CagA strains and reduced iron biomarkers ^(1,16). However, the presence of H pylori-CagA antibodies was related to reduced levels of gastric ascorbic acid ⁽¹⁶⁾, a potent enhancer of dietary iron absorption ⁽¹⁷⁾.

Few clinical trials assessed the efficacy of anti-H pylori therapy on resolution of iron deficiency anemia or iron deficiency (ID). The large randomized controlled trial ⁽¹⁸⁾ in children showed no superiority for anti-H pylori treatment on iron supplementation in resolving mild ID in highly endemic populations for both H pylori and ID. In other studies, the sample size was small ⁽⁴⁾ or a concurrent control group was lacking ⁽⁵⁾.

A potential limitation of our study could be the use of sera of children referred for laboratory investigation with illnesses that could have been associated with serum ferritin. However, there is no reason to assume that these illnesses could have been associated with the independent variable, namely H pylori serostatus and in this way alter the observed association.

In summary, our findings indicate that in a pediatric population with high prevalence of both H pylori infection and anemia, H pylori seropositivity is associated with low serum ferritin levels, with young children being at highest risk. To determine whether this is a causal association, large controlled trials examining the efficacy of anti-H pylori therapy on the reduction of ID and iron deficiency anemia are needed.