Individual and household correlates of Helicobacter pylori infection among Young Ethiopian children in Ziway, Central Ethiopia.

Background Investigating distinct individual- and household-level risk factors for acquiring Helicobacter pylori (H. pylori) infection can inform disease prevention efforts and implicate possible routes of transmission. This study determined the magnitude of H. pylori infection among schoolchildren in Ziway, central Ethiopia and identified personal and household correlates of H. pylori infection in young Ethiopian children. Methods A total of 434 schoolchildren participated in this cross-sectional study. Infection status was assessed using antigen and antibody rapid tests. Demographic and lifestyle information was obtained from parents via an interviewer-led questionnaire. Univariate and multivariate logistic regressions were performed to assess the relationships between potential individual- and household-level risk factors and H. pylori infection. Results The prevalence of H. pylori infection was 65.7% (285/434). Of the personal variables assessed, the age group 10–14 years was found to be significantly associated with higher odds of H. pylori infection in univariate analysis (COR = 2.22, 95% CI: 1.06–4.66, p = 0.03) and remained positively correlated after adjusting for confounding factors. Of the household-level factors explored, having a traditional pit or no toilet was found to be significantly associated with 3.93-fold higher odds of H. pylori infection (AOR = 3.93, 95% CI: 1.51–10.3, p = 0.01), while the presence of smokers in the household was associated with 68% lower odds of infection (AOR = 0.32, 95% CI: 0.11–0.89, p = 0.03). Conclusion This study from a developing country provides additional evidence for older age as a personal risk factor for H. pylori infection and identifies correlations between socioeconomic and sanitation household factors and positive childhood infection status. The associations reported here support the hypothesized fecal-oralroute of transmission for H. pylori.


Background
Helicobacter pylori (H. pylori) is a pathogenic bacterium that infects human gastric mucosa [1]. It is known to be one of the leading causes of dyspepsia, active gastritis, peptic ulcers, and gastric adenocarcinomas [1]. Because of its role in gastric cancers, it is classified as a World Health Organization Group 1 carcinogen [2]. Nearly one half of the global population is infected with H. pylori, with some countries reporting over 80% infection rates [3,4]. Although prevalence rates have been generally declining in recent years, infection remains more prevalent in developing countries in comparison to developed countries [3,4]. The reasons for this are not yet fully established, though are speculated to be related to poorer sanitation and living conditions in resourcelimited settings [3].
Close person-to-person contact, especially between family members living together, has been shown to be important for H. pylori transmission [5,6]. The exact mode of person-to-person transmission remains unclear though, with literature proposing both oral-oral and fecal-oral routes. Although H. pylori has been detected in dental plaque, saliva, gastric juice, and fecal matter, samples outside of the stomach have proven difficult to culture [7,8]. Identifying associated risk factors for infection can create a more comprehensive understanding of probable H. pylori transmission mechanisms.
Most H. pylori infections are acquired during childhood and persist into adulthood [9]. Infection often remains latent for years, with symptomatic disease developing up to several decades later [1]. Accordingly, H. pylori transmission is more highly associated with childhood living conditions than adult living conditions [9,10]. Previous studies have thus sought to identify select risk factors for H. pylori infection by examining childhood environments [10][11][12][13][14][15]. Common risk factors examined in these studies include age, indicators of socioeconomic status, and poor hand and cooking sanitation [10][11][12][13][14][15][16]. Both individual and familial characteristics have been shown to influence infection status. Investigating separate personal-and household-level risks for acquiring H. pylori infection can elucidate appropriate levels of disease prevention efforts. While personal attributes are crucial to investigate, factors affecting entire households have a greater impact on the health of families and groups of people living in close proximity. Focusing on distinct household risk factors can help inform public health interventions that are able to prevent a comparatively larger number of cases.
Despite the unequal worldwide distribution of H. pylori infection, previous studies on risk factors for childhood infection have almost mostly been carried out in industrialized countries [10][11][12][13][14][15]. More research needs to be done to identify and compare risk factors in children in more resource-limited settings, as these are the demographics with the highest rates of infection. H. pylori prevalence is comparatively understudied in many African countries, though the studies that have been completed show higher than average prevalence rates [3,4,17,18]. This makes studying childhood H. pylori infection in these countries important for disease surveillance and intervention. Understanding which contributing factors are inherent or mutable on the personal and household levels will enable public health workers to effectively maximize prevention by adding important clarifying details to the preexisting body of knowledge regarding H. pylori infections. This cross-sectional study therefore aimed to determine the magnitude of H. pylori infection among schoolchildren in Ziway, central Ethiopia and provide evidence for distinct personal and household risk factors associated with positive H. pylori childhood infection status.

Study area and design
This study was conducted in Ziway, Ethiopia along Lake Ziway located 160 km south of the country capital, Addis Ababa. Ziway has a population of 43,660 per the 2007 census and has experienced population and economic growth since its establishment in 1961. The population of Ziway is majority male (52.6%) and Orthodox Christian (51.0%). Participant study eligibility included children less than 15 years old who lived in Ziway and were enrolled in primary school. Data was collected from June 6 to July 30, 2016 at four different locations: Sher Elementary School, Sher Ethiopia Hospital, Batu Hospital, and Batu Health Clinic. A cross-sectional study was designed to examine the prevalence of H. pylori in Ziway schoolchildren as well as assess risk factors for acquiring H. pylori infection. A convenient sampling technique was used in order to enroll participants in the study; a child's school attendance on a particular day determined participation. Enrollment in mass deworming programs is based on World Health Organization (WHO) recommendation, in which anthelmintic drugs are administrated to all schoolchildren regardless of their demographic difference.

Data collection and laboratory analyses
Informed consent forms were signed by the parents of children involved in the study. Demographic and lifestyle information was collected through a questionnaire developed previously [19], and administered to the parents by a trained interviewer. Demographic questions included information on mother's education, occupation, and place of residence. Lifestyle information included household size, sanitary conditions, child health practices, and latrine usage. Deworming status was ascertained from school records. Active H. pylori infection was determined using an antigen rapid test enzyme immunoassay (Human Helicobacter Pylori Rapid Test Cassette Antigen RapiCard InstaTest, Diagnostic Automation Inc., USA). Each participant was given a leak-proof plastic container and clear instructions on how to collect a fecal sample. Once collected, samples were placed in a container with 10% formalin and transported to Sher Ethiopia Hospital for detection of H. pylori antigens. For this, a small amount of stool was homogenized with buffer solution and added to test wells. A monoclonal mix of anti-H. pylori antibodies was used for the capture antibody and a mix of monoclonal peroxidaseconjugated anti-H. pylori antibodies was used for the detection antibody. The test was read after 15 min of contact with antibodies. A positive result was defined as the presence of two lines, even if the test line was noticeably fainter than the control line. A negative result was defined as the presence of only the control line. The sensitivity and specificity of the test are 94 and 96.7%, respectively.
An antibody rapid test (Diagnostic Automation Inc., USA) was used to detect any past or present H. pylori infections through a double antigen chromatographic lateral flow immunoassay. This test is unable to distinguish between current and past infection. A 5 mL whole blood sample was collected from each participant using sterile disposable syringes. Within two hours of blood collection, serum samples were separated at Sher Hospital laboratory to detect presence of H. pylori antibodies. For this assay, each test well received 1-2 drops of serum and was read 15 min after the serum was added. A positive result was defined as the presence of two lines, even if the test line was noticeably fainter than the control line. A negative result was defined as the presence of only the control line.

Exposure definitions
Both personal and household-level risk factors were examined in order to understand the different levels of risk for H. pylori infection. Personal factors are individual attributes or behaviors that can affect the likelihood of developing a disease. Personal variables examined in this study include, but are not limited to, age, sex, medication history, and laboratory test results. All individual variables were able to be measured directly from participants.
Household factors are familial living conditions that affect health opportunities and outcomes for all people living in the same household. Some of the household factors assessed include, but are not limited to, indicators of socioeconomic status, housing conditions, cooking practices, water source, sanitation, and animal ownership. Several household factors are difficult to measure directly, so in these cases proxy measurements were used as estimates. Socioeconomic status was determined using maternal education and household electricity use. Quality of housing materials was defined by wall type and floor type. Housing conditions were evaluated by the presence of a bed for the child and household cooking location. Sanitation was assessed through type of latrine in household and materials used for cooking.

Outcome definitions
A positive stool antigen rapid test indicates bacterial antigens in the body and is used in this study to define current H. pylori infection. A positive antibody rapid test from a blood serum sample indicates antibodies for the bacteria are present in the bloodstream following infection and is used in this study to define past infection. Positive H. pylori infection status is classified as a positive outcome for either test signifying H. pylori infection at some point during the child's life.

Statistical analysis
Data were entered, cleaned, and coded for analysis using SPSS Statistics (IBM Corp., Armonk, N.Y., USA). Descriptive statistics were used to characterize data and assess distribution of study variables. Univariate logistic regressions were completed in order to identify factors to include in further analysis. Multivariate logistic regressions were completed to assess the relationships between potential risk factors and H. pylori infection status after adjusting for confounding variables. Both direct multivariate analysis using all factors and stepwise regression including only variables associated with positive H. pylori infection status at p < 0.3 during univariate analysis were performed. Odds ratios were calculated to determine the magnitude of associations for each risk factor. Statistical significance was defined as p < 0.05. Similar patterns of demographic and lifestyle distributions were observed among study subject who had complete outcome data and all respondents using sensitivity analysis (data not shown).

Data quality assurance
Questionnaires were inspected by two separate reviewers for completeness and consistency of biological data. Incomplete surveys were returned to participants to complete. Physical samples were analyzed immediately after collection. Two senior lab technicians read the data. Quality control samples were used for both the rapid antigen test and rapid antibody test.

Prevalence of H. pylori infection
Positive H. pylori infection status was defined as either current or previous infection. As seen in Table 2, the prevalence of H. pylori infection among tested schoolchildren was 65.7% (285/434).

Association of Personal Factors and H. pylori infection
The crude associations between personal factors and H. pylori infection were assessed using univariate logistic regressions (  (Table 3). Vaccination status was the only variable with an OR that greatly changed from crude to adjusted, with unvaccinated children showing a moderately strong non-significant association with higher odds of H. pylori infection (AOR = 2.13, 95% CI: 0.38-12.1, p = 0.39). Stepwise multivariate analysis including only individual factors with a COR p < 0.3 also yielded no statistically significant results, but again demonstrated that older age, female sex, and current antibiotic use were correlated with higher odds of H. pylori infection. Most ORs remained relatively stable throughout univariate and multivariate analysis.

Association of Household Factors and H. pylori infection
The crude associations between household-level factors and H. pylori infection were assessed using univariate logistic regressions. Out of all household variables assessed, univariate analysis found that having a traditional pit or no toilet was the only factor associated with statistically significant increased odds of H. Direct and stepwise multivariate logistic regressions were performed in order to determine associations with H. pylori infection after adjusting for confounding factors (Table 4). Direct multivariate analysis yielded similar trends as in univariate analysis, with traditional pit or no toilet being significantly associated with positive H.

Discussion
This study identified the prevalence of H. pylori infection in central Ethiopian schoolchildren and provides additional evidence for associations between distinct personal and household correlates and infection status. Positive infection status was significantly associated with older age at the individual level and having a traditional pit or no toilet at the household level. Presence of smokers in the household showed statistically significant decreased odds of infection. There was a borderline significant negative correlation between infection and pit or field waste disposal as well.
The H. pylori prevalence rate in schoolchildren in Ziway, Ethiopia was 65.7%, signifying that over half of children less than 15 years old have been infected at some point in their life. This rate is higher than reports of the average global prevalence [3,4], though comparable to recent studies in rural Ethiopia [17,18]. However, previously reported rates are inclusive of all age ranges, not just school-aged children. Assuming the trend of increased infection with older age continues, this suggests an overall higher disease prevalence in central Ethiopia than calculated in this study.
Of all the personal variables assessed, older age was associated with higher odds of positive H. pylori infection status, with ages 10-14 having a stronger association than ages 5-9. Older age is a common risk factor for H.    [9, 13-15, 18, 20-24]. There is no consensus yet whether this trend is due to cumulative increased risk over a lifetime [11,22] or decreased risk of infection in each subsequent birth cohort causing older cohorts to have higher infection rates comparatively [20,24]. Due to the nature of this study, the data presented here cannot support one hypothesis over the other. Of all the household-level factors investigated, having a traditional pit or no toilet at all was significantly associated with 3.93-fold higher odds of H. pylori infection after adjusting for confounding factors. To date, contradictory literature exists regarding associations between latrine use and H. pylori infection status [14,18]. Having no toilet or a traditional pit could increase the possibility of contact with fecal material as compared to flush toilets or ventilated pits. In a study on traditional pit sanitation in Malawi, it was found that traditional pits can promote transmission of fecal-oral diseases due to lack of maintenance, poor disposal of feces, fouled washwater, and potential for pit surcharging [25]. Flush toilets, on the other hand, have been proposed as a mechanism for disease vector removal, which can help prevent further infection [14]. Since H. pylori can be detected in fecal material [7,8], using traditional pits or no toilet may increase the potential for contact with the bacteria and facilitate infection. The significant association between toilet type and infection status reported in this study thus supports the proposed fecal-oral route of transmission for H. pylori. Accordingly, hygiene education and availability of alternative toilet types are potential areas of focus for disease control and prevention. Strikingly, at the household level the presence of cigarette smokers in the household was significantly correlated with 68% decreased odds of H. pylori infection after adjusting for confounding factors. Previous studies have published conflicting results on the relationship between smoking as a personal attribute and H. pylori infection status [18,26], though there is little data addressing infection rates in those living with smokers. In other developing African countries, smoking is more common among wealthy people [27]. Considering the small number of children in this study with family members that smoke, the presence of smokers in the household could thus be indicative of good socioeconomic status. Poorer socioeconomic status has been previously good socioeconomic status defined as formal maternal education and any household electricity use C ack of paracetamol defined as lack of affordability or availability d poor housing materials defined as mud and wood walls and uncovered mud floor e poor housing conditions defined as no child bed and cooking inside main living area in house f natural materials defined as dung, leaves, and wood g farm animals defined as hen, cow/ox, sheep, horse, goat, pig, and mule/donkey h poor sanitation defined as tradition pit or no toilet and using dung for cooking reported to be associated with increased H. pylori infection [11,13,21,23], perhaps due to limited access to food, water, healthcare, and soap, among other things. This suggests that poverty and access to education may be possible targets for H. pylori prevention. Yet, as another explanation, Ogihara et al. proposed that chemical interactions between nicotine smoke and gastric mucosa can increase gastric acidity and mucosal atrophy, assisting in the auto-eradication of H. pylori infections [26]. It is unknown whether secondhand smoke exposure can lead to similar results. It is also possible that if adults who smoke have lower prevalence rates, there is less potential for household transmission. However, since this study does not address cigarette smoking as a personal variable, it remains unclear how the presence of smokers in the household influences H. pylori infection status. The effects of smoking on bacterial transmission in parents and children warrants further investigation.
Additionally, pit or field waste disposal was borderline significantly associated with decreased odds of positive H. pylori infection status as compared to burning or using garbage bins. A previous study in Asella, central Ethiopia describes limited access to door-to-door waste collection services in Ethiopian towns outside of the capital city [28]. Waste disposed in garbage bins in Ziway may thus not be frequently collected and instead stored in the home for extended periods of time. A study in Ghana found that storing solid waste in open containers in the home increased the presence of houseflies, which was in turn associated with increased gastrointestinal disease occurrence in children [29]. Similarly, Parente et al. found that the absence of garbage collection services is positively associated with H. pylori infection [30]. Houseflies are known vectors for disease transmission from fecal matter to food. It is thus possible that garbage bin waste disposal increases vector-borne transmission of H. pylori, again aligning with the proposed fecal-oral route of transmission. However, further investigation on waste management and disposal practices is needed to explore this possibility.
These findings should be interpreted in light of a few study limitations. First, a cross-sectional study design was used as there were no previously available baseline data on H. pylori infection status. Because of the observational nature of this study, it is difficult to conclude temporality and causality between personal and household variables and infection. A longitudinal study is needed to confirm our findings and assess the possibility of reverse causality. Additionally, only a small region of central Ethiopia was sampled, limiting the generalizability of the reported prevalence rates to the entire country. It is also probable that some residual confounding personal or household factors were excluded from analysis despite the large number of variables examined. Surveying the parents instead of the participants themselves limited the collection of information on child-specific variables such as frequency of handwashing after using the latrine or whether a child helps with food preparation. Our study participants were apparently healthy children enrolled in primary school, so the observed individual and household risk factors for H pylori infection might be different if we included children who dropped out or never attended the school. A further limitation of the current study is that some subgroup categories had small sample sizes which led to wide and imprecise confidence intervals for those observed measures of association.
Despite these limitations, this study is the first to differentiate between personal-and household-level risk factors for H. pylori infection in a developing country. This distinction allows for a more thorough assessment of disease transmission and determination of which prevention efforts are most useful for disease control. The high rate of participation and large sample size used serve to strengthen the study findings. Additionally, because separate antigen and antibody rapid tests were performed, analyses were able to assess risk factors for both current and past infection.

Conclusion
Our data from a developing country provides additional support for older age as a personal risk factor for H. pylori infection and identifies correlations between socioeconomic and sanitation household factors and positive childhood infection status. The associations reported here support the hypothesized fecal-oral route of transmission for H. pylori. Further longitudinal research is needed to establish causality and exact mechanisms.
Abbreviations H. pylori: Helicobacter pylori; WHO: World Health Organization Funding Colgate University Research Council funded this study. The funding body had no role in study design; collection, analysis, and interpretation of data; writing of the report; or the decision to submit the paper for publication.

Availability of data and materials
The datasets analyzed in the current study will be available from the corresponding author upon reasonable request.
Ethics approval and consent to participate Ethical clearance was given by Addis Ababa University Department of Medical Laboratory Science departmental research and ethics review committee. Prior to study enrollment, parents of participants gave written informed consent via signature or witnessed thumbprint. Written assent was received when appropriate. Confidentiality of participant identities was ensured during data analysis through the use of numerical identifiers. Electronic data were protected with passwords and hard copies were locked away during and after completion of the study.

Consent for publication
Not Applicable.

Competing interests
We declare that we do not have any conflicts of interest.
Author details