Prevalence and risk factors of Helicobacter pylori infection among children in Kuichong Subdistrict of Shenzhen City, China

Ethics approval

The study was approved by the research ethics committee of the First Affiliated Hospital of Shenzhen University, Shenzhen Second People’s Hospital. Written and informed consents were obtained from all participants and their legal guardians.

Sample size determination

According to a previous H. pylori prevalence study conducted in Guangzhou, we assumed that 22.9% of asymptomatic children would be infected with H. pylori (Chen et al., 2007). The samples size required for this study was estimated based on a 95% confidence interval and a 2.5% admissible error rate using the following formula:

Sample size, $n=Z_{\alpha /2}^2\times \pi \times (1-\pi )/{\delta }^2$, where Z_α/2 is the 2-tailed normal deviate for an α level of 0.05; π is the prevalence ratio; δ is the admission error rate.

A sample size of 1,152 was therefore required for this study. We added at least 25% to the estimate for contingency purpose, raising the final sample size to 1,355.

Study population

To avoid selection bias, from September 2018 to October 2018, all school children aged 6–12 years from four primary schools located in Kuichong Subdistrict of Shenzhen City, China, who did not meet any of the exclusion criteria and had provided parental or guardian consent, were recruited. Exclusion criteria included intake of antibiotics within the last 4 weeks, intake of proton pump inhibitors within the last 2 weeks and previous history of H. pylori eradication therapy.

The ¹³C-UBT (Beijing Boran Pharmaceutical, Changping, China) was performed according to the manufacturer’s instructions to determine H. pylori infection status. All breath samples were taken and analyzed immediately in the school by the same technician. Briefly, an initial baseline breath was collected from each study participant following an overnight fast of at least 4 h. Each participant was then requested to ingest a tablet containing 75 mg ¹³C labeled urea with 80–100 mL of water. After 30 min of sitting, exhalation was again collected. The ¹³CO₂ content within the initial and 30-min expiratory air bags were analyzed using an HG-IRIS13C infrared spectrometer (Beijing Richen-Force Science & Technology Co., Beijing, China). The concentration of ¹³CO₂ following 30 min of administration that exceeded the baseline by 4% was regarded as a positive indicator of H. pylori infection. The weight, height, and both head and chest circumferences of each participant were also recorded. Body mass index (BMI) was calculated as body weight divided by height squared (kg/m²). The BMI category was classified according to the corresponding percentiles as recommended by The Centers for Disease Control and Prevention.

Questionnaires

On behalf of each study participant, parents or legal guardians were requested to complete a questionnaire for data collection on sociodemographic and living conditions, birth method and childhood feeding practices, lifestyle habits and health parameters. A blank copy of the questionnaire is available as File S2. The outcomes are summarized in Table S1. Sociodemographic and living conditions consisted of age and gender of all children, number of siblings (three categories: 1, 2–3 and >3), total annual income (two categories: ≤200,000 or >200,000 RMB), household population (two categories: ≤3 or >3 persons), household space (two categories: ≤40 or >40 m²/person), parents’ education level (two categories: both were high school graduates or either one had completed tertiary education) and pet ownership (two categories: yes or no).

The birth method of each child was assessed in two categories (via natural delivery or a cesarean section). Childhood feeding practices comprised the primary feeder of young children (four categories: parents, grandparents, both parents and grandparents, or other caregivers), feeding methods (three categories: feeding pre-chewed food, feeding by using shared cutleries or feeding by using a separate set of cutleries), breast feeding duration (three categories: never, up to 6 months or more than 6 months) and how the milk bottle temperature was tested (three categories: by dripping a few onto the wrist, taking a swig or using a thermometer).

Lifestyle investigations included the teeth brushing frequency (three categories: ≤3, 4–6 or ≥7 times per week), hand washing before meal and after toilet (three categories: rarely, sometimes or always), dining out frequency (three categories: ≤1, 2–4 or ≥5 times per week), how fruit was eaten (two categories: peeled or unpeeled), drinking water for consumption (two categories: raw or boiled), toothbrush sharing (two categories: yes or no), snacking regularly (two categories: yes or no) and thumb sucking (two categories: yes or no). The health conditions of each child were evaluated in three categories (≤1, 2–4 or ≥5 times per week) for bloating, abdominal discomfort, burping and diarrhea, respectively, and in two categories (present or absent) for halitosis, anemia, asthma and skin allergy, respectively.

Statistical analysis

Unpaired, two-tailed Student’s t-test was used for statistical comparison of continuous variables between H. pylori positive and negative groups. For comparison of categorical variables, the chi-square test was employed, followed by univariate logistic regression analysis using JASP software v0.11.1 (https://jasp-stats.org/). Variables significantly associated with H. pylori infection in the univariate test (P < 0.05) were selected as candidates for entry into the multivariable logistic regression models using the forward selection process. During this iterative selection process, any variable with P > 0.1 was eliminated. Subsequently, variables previously not included in the original model were introduced, one at a time, to identify any variable that could have an important role in H. pylori infection in the presence of significant variables retained earlier. The selection process was repeated but only for the additionally loaded variables until a final model was obtained. Both univariate and multivariate analyses were adjusted for age. Results were presented as odds ratio (OR) with 95% confidence intervals (CI). A P-value of less than 0.05 was regarded as statistically significant.

Prevalence of H. pylori infection

In this study, 1,355 children aged from 6 to 12 years were recruited, among which 747 (55.1%) were males and 608 (44.9%) were females (Table 1). The overall H. pylori prevalence rate was 16.7% (226/1355; 95% CI [14.7–18.7]). There was no distinguishable difference in H. pylori infection rate in gender and between children who were locally and elsewhere born. Interestingly, while 8-year-old children had the highest prevalence of H. pylori infection at 23.2%, lower than the overall prevalence positivity rates were seen in children aged 6 and 10 years old, at 10.6% and 12.6%, respectively. No significant differences in anthropometric measurements were observed among the infected and non-infected children.

We performed chi-square test to examine for any significant difference in different sociodemographic, lifestyle and clinical characteristics between H. pylori-positive and H. pylori-negative children. Variables tested with no significant difference between both groups are available in Table S2. As demonstrated in Table 2, the following variables contributed to greater H. pylori prevalence among children: having more than one child, both parents attended no education after high school, smaller living space, testing milk bottle temperature by taking a swig, feeding infant with pre-chewed food or by using feeder’s cutlery, sharing toothbrush, eating out at least 5 times a week, snacking and eating fruits after peeling.

Protective and risk factors for H. pylori infection

Risk factors for H. pylori infection were further investigated by univariate and multivariable logistic regression analyses adjusting for age. The final model of multivariable analysis revealed six variables significantly associated with H. pylori in children (Table 3). Children whose parent(s) with tertiary education level (OR: 0.64; 95% CI [0.46–0.89]; P = 0.008) and larger living space at home (OR: 0.54; 95% CI [0.28–1.07]; P = 0.078) were protective factors against H. pylori infection. Conversely, testing milk bottle temperature by taking a swig as opposed to dripping a few onto the wrist (OR: 1.79; 95% CI [1.19–2.68]; P = 0.005), sharing of cutlery between the feeding person and young children rather than using separate cutlery each during meals (OR: 1.84; 95% CI [1.22–2.78]; P = 0.004), and snacking (OR: 1.43; 95% CI [1.01–2.01]; P = 0.043) resulted in lower risks for H. pylori infection. Only eating fruits peeled than unpeeled (OR: 2.56; 95% CI [1.4–4.71]; P = 0.002) and eating out frequently for at least 5 times a week (OR: 3.13; 95% CI [1.46–6.68]; P = 0.003) would result in substantially higher risks for H. pylori infection.