Seroprevalence and determinants of Helicobacter pylori infection among asymptomatic under-five children at a Tertiary Hospital in the South-Western region of Nigeria

Background The epidemiology of Helicobacter pylori (H. pylori) infection among under-five children in the South West Nigeria remains largely understudied. There is no data on the subject from the study area. Objectives This study was conducted to assess the seroprevalence of H. pylori infection among under-five children at a tertiary hospital in the South Western, Nigeria and to determine its associated socio-demographic factors. Methods Sera of 360 children were analyzed for anti H. pylori Ig G using enzyme linked immunosorbent assay test kit (BQ ELISA Ig G KIT) and H. pylori infection risk factors were determined. Determinants of H. pylori infection was determined using binary logistic regression analysis and p-values < 0.05 were taken as statistically significant. Results H. pylori infection seroprevalence rate was 32.8% and increased with age. Living in one room accommodation, large families, playing with soil, family history of dyspepsia, practice of premastication, sharing of plates and cutlery, and water closet toilet were associated with H. pylori Ig G seropositivity (p<0.05) on binary regression analysis. Conclusion The seroprevalence of H. pylori infection in under -five children is high, increasing as the age of the children increased. This may suggest that instituting preventive measures at young age, targeting identified factors may be effective in reducing the burden of H. pylori infection.


Introduction
Helicobacter pylori (H. pylori) infection is one of the most common human bacterial infection worldwide with about 50% of the global population infected 1 . The prevalence of infection varies both between and within countries in relation to race, ethnicity and geographical area 2 such that developed countries have significantly lower rates of H. pylori infection than developing countries. This difference has been attributed to differing rates of acquisition of the organism in childhood as well as sanitation, socio economic and hygiene status of the population3. Most of the infected people develop no clinical symptoms and live their lives with superficial chronic gastritis. However, about 17% of infected subjects will develop peptic ulcers and about 1% will progress to gastric cancer which is the third leading cause of cancer-related deaths in the world, especially in the adult population. 4 The childhood period may be critical for the acquisition of H. pylori infection 2 . It is believed that once the organism is acquired, it persists for life unless there is an intervention 1,5 . However, some authors have reported loss of infection either spontaneously in some rare cases, or as a result of inadvertent exposure to antibiotics 6,7 . Documented risk factors for H. pylori infection include poverty, 8 poor sanitation, overcrowding 9,10 and unsafe water source 11 , which are prevalent in most communities in Nigeria. Since the epidemiology of this bacterial infection differs from one geographical area to another and even within countries, it is therefore pertinent that more studies are conducted on people from different backgrounds in order to have better understanding of the burden of H. pylori infection. Hence, this study assessed the burden of H. pylori infection as well as the socio-economic and demographic factors that may increase the risk of childhood acquisition of infection in a semi-urban community in South Western Nigeria.

Study location
The study was conducted in two units of the Obafemi Awolowo Teaching Hospitals Complex (OAUTHC), Ile-Ife, Nigeria, These were the Wesley Guild Hospital unit (WGH) Ilesa and the Urban Comprehensive Health Centre unit (UCHC) Ile-ife. Ile -Ife town lies on longitude 40 69'E and latitude 700 50'N. 12 According to the 2006 population census, Ile-Ife has a population of 355,281 and an area of 283 sq.km. Ilesa town is located between latitude 70 37' N and longitude 40 40' E. 13 with a population of about 277,904. 14 The services rendered at these units include Infant Welfare, Nutrition Rehabilitation, School Health, Immunization, Dental, and Physical Rehabilitation. On average, about 1,000 children aged 0-59 months are seen at the various clinics at the WGH and 2,800 children at the UCHC monthly.

Study design
We conducted a cross-sectional study between July 2014 and March 2015.

Study population
The study population comprised of children aged 6-59 months attending the child welfare clinics.

Sampling methods
On average, about 1,000 children aged 0-59 months are seen at the various clinics at the WGH and 2,800 children at the UCHC monthly. Based on this, proportionate sampling was employed in recruiting study participants. A total of 360 children between the ages of 6 months and 59 months were recruited, 265(73.6%) from UCHC and 95(26.4%) from WGH. Recruitment was done based on the average population of children that attend each facility. All consecutive children who met eligibility criteria were recruited and questionnaires were administered to their parents or caregivers.

Data collection
Parental social class was determined using Oyedeji 15 classification method. Children who had taken antibiotics one month before presentation, those on proton pump inhibitors and those on H2 receptor blockers were excluded from the study.

Anthropometric measurements
Weight, length and height were measured by standard techniques as described in the supplementary materials.
H .pylori detection Two milliliters of whole blood was collected aseptically through venepuncture from each participant into plain bottles. The sample was then left to clot and serum separated after spinning for 10 minutes at 4,000 revolutions per minute. The separated serum was then frozen at -20 0 C till the time of analysis. The sample analysis was done in the laboratory of the Department of Biochem-istry, Obafemi Awolowo University, Ile-Ife, Nigeria. The frozen samples were thawed at room temperature. The "BQ Helicobacter pylori Ig G ELISA" kit (BQ Kits Inc. San Diego USA).was used. Each of the samples was run twice and the mean H. pylori Ig G recorded. According to the manufacturer's manual, the antibody index was interpreted as: < 0.9 -no detectable antibody to H. pylori Ig G by ELISA, 0.9-1.1-Borderline positive > 1.1-detectable antibody to H. pylori Ig G by ELISA

Data analysis
Data analysis was performed using the Statistical Package for Social Sciences (SPSS) version 20.0. Frequencies and proportions were used to summarize socio-demographic variables and this was presented as tables and charts. The sero-prevalence of H. pylori infection was determined using percentages. Chi square test of statistic was then used to determine any association between sero-prevalence and the socio demographic factors. Binary Logistic re-gression with odds ratio at 95% confidence intervals was used to assess the determinants of H. pylori seropositivity. Significance level was set at p-value < 0.05 for all tests.

Ethical approval
Ethical approval was obtained from the OAUTHC Ethics and Research Committee. Also, written informed consent was obtained from the parent or guardian of each study participant before the questionnaire was administered.

Sero-prevalence of H. pylori infection:
Of the 360 study participants, 118 (32.8%) were serologically positive for H. pylori Ig G antibodies while 242 (67.2%) were negative. Socio-demographic characteristics of the study participants: As shown in Table 1, there was slight male preponderance. Most of the mothers were aged between 20-39 years and virtually all the mothers (98.1%) had formal education. Majority of parents (52%) were in social class 3. No family was in social class 5 and most households had between one or two children.

Socio-demographic factors associated with H. pylori infection:
As shown in Table 2, the sero-prevalence of H. pylori infection increases with advancing age of the children. Also, its prevalence increases with increasing number of children in the households (p<0.001). However, there was no difference in the prevalence of H. pylori infection with respect to mothers` age, level of formal education and social class of the parents (p>0.05).
Relationships between child care practices, environmental characteristics and H. pylori sero-positivity: The signifi-cant factors associated with H. pylori sero-positivity were: practice of premastication, sharing of plates and cutlery, and accommodation type (p<0.05) Table III.   23 reported an overall sero-prevalence of 45% among children up to 15 years old in India. Also, Hestvik et al. 22 reported a prevalence rate of 46.0% among less than 3 years children in Kampala, Uganda.
However, the sero-prevalence rate of H. pylori infection in this study is higher than those reported in developed countries [24][25][26] . Naito and colleagues reported 4.0-6.7% among Japanese children using urine H. pylori antibody based test. 24 Dore et al. 25 obtained 13.3% amongst children aged 6-15 years in Italy. Similarly, Tam and coworkers 26 using 13C-urea breath test reported 13.1% among Chinese children in Hong Kong. These lower prevalence rates in developed countries could be due to the better living conditions and improved hygiene status over there.
In this study it was found that the sero-prevalence of H. pylori infection increases with age. This has been corroborated by many authors 19-21, 27,28 . The reason for this, is probably the increased opportunities for acquisition of the microorganism as a result of increased exposure to the various sources of the infection especially in children as they grow. Children are usually more mobile with increasing age and particularly in under-fives, they have less regard for hygienic practices which may facilitate acquisition of infection especially through faeco-oral mode of transmission. Some authors, however, did not observe any relationship with age 9,16,26 . They opined that this may be due to spontaneous elimination of infection and loss of infection as indirect benefits from use of antibiotics for some other purposes 29 .
In this study, the factors found to be associated with the high prevalence of H. pylori infection among children included premastication of food by parents, sharing of plates and cutlery with family members, communal residing in one room apartment by families, family history of dyspepsia and having many children in the household. These findings are mixed regarding literature. Sheikhian et al. 27 in Iran reported association between premastication of food by mothers and H. pylori infection but found no association with sharing of plates. Nabwera et al. 31 reported sharing of plates as an independent risk factor for H. pylori infection. In contrast, Langat et al. 9 reported no association between premastication of food by mother and H. pylori infection.
Ikpeme et al. 32 in Uyo, Nigeria reported an association between H. pylori seropositivity and a family history of dyspepsia. This association probably corroborates the causative role of H. pylori in dyspepsia and possible intra-family spread of the organism which was also observed by Osaki et al. 33 and Nahar et al. 34 in Japan and in Bangladesh respectively.
This study revealed that children who reside with parents in one room apartments were found to be seven times more likely to be infected with H. pylori than those who reside in houses with two or more rooms thus highlighting impacts of overcrowding in H pylori infection as earlier observed 9,20,30,35 .
Surprisingly, the use of water closet as means of sewage disposal was found to be a risk factor for H. pylori infection in this study. This might be related to improper and unhygienic use of the water closet toilet facilities especially in developing countries such as Nigeria with poor access to water supply. Similar to this study, Senbanjo et al. 18 in Lagos also found that people who used water closet toilet facilities had higher H. pylori seropositivity rate than those who did not. This is in contrast with the findings of Etukudo and coworkers 16 in Uyo, who found that use of pit latrine was more likely to be associated with H. pylori infection in children than water closet. There was no association between H. pylori seropositivity and mother's level of education. This is in agreement with findings by Oleastro et al in Lisbon 36 .
Our finding of no association may be because most of the mothers of the study participants were literate. However, previous authors have established link between low mother education and increased H. pylori infection 9,26 . This study also revealed no association between types of feeding practice (bottle feeding, breastfeeding and use of cup and spoon) and H. pylori seropositivity rates. Sial et al. 30 in Tunisia reported no association between H. pylori and breastfeeding but found prolonged bottle feeding as a risk factor, thus suggesting that the feeding bottle may be a vehicle of infection. Langat et al. 9 also found no association between H. pylori infection and breast feeding.

Conclusion
Sero-prevalence of H. pylori infection was found to be high in this study and logistic regression model indicated that water closet toilet facilities, sharing of plates and cutlery, one room accommodation, playing with soil outside the house, number of children in the households, practice of premastication of food and family history of dyspepsia were associated with H. pylori infection. Therefore, health education targeted at parents and guardians with regards to preventing these factors will be helpful in reducing the prevalence of H. pylori infection in children in the study area.
In addition, since vaccination against the infection has been proposed by some researchers in developed countries as a way to combat the scourge of H. pylori infection, 37 the findings from this study, raises the need for more longitudinal studies focusing on the evolution of dyspeptic symptoms during childhood and possibly identification of serotypes commonly responsible for H. pylori infection in Nigeria for targets for vaccination. Also, such studies may help to identify other determinants of H. pylori infection. Hence, this may help government in formulating policies such as inclusion of vaccination against H. pylori infection in the routine childhood immunization schedule and improving the general housing conditions of the populace. Finally, there is a need for guidelines that could assist the clinicians in developing countries in identifying which category of children to screen and treat for the H. pylori infection.

Limitation of the study
This study was hospital based and has the possibility of being skewed towards severe cases of the health targeted conditions. However, the primary health care units where the study was conducted mainly care for children who are not sick but merely came for well child visitation, for growth and nutritional monitoring and immunization services. Hence, are presumed to closely reflect the community prevalence of H. pylori infections.