Diarrhea Prevalence and Associated Factors among Under-Five Children in the Periphery Area of Azezo Sub-city, Gondar, Northwest Ethiopia: A community based cross-sectional study


 Background

Although there has been a global decrease in childhood diarrheal disease in parallel with improvements in the standard of living, it still remains a significant public health problem that occurs due to poor WASH status and other related factors that cause massive childhood morbidity and mortality particularly in sub-Saharan Africa countries including Ethiopia.
Objective

To assess diarrheal disease prevalence and associated factors among under-five children in periphery area of Azezo sub-city, Gondar, northwest Ethiopia.
Methods

A community-based cross-sectional study was conducted in April 2019. A simple random sampling technique was applied to recruit the eligible clusters/villages. Structured interview questionnaires were used to collect data. The completed data were entered into Epi Info version 7 and exported to SPSS version 20 for further analysis. A binary logistic regression model was used to process bivariate and multivariable analysis of the data to establish the association between dependent and independent variables. The adjusted odds ratio (AOR) with 95% CI was used for the interpretation of data after controlling the confounders.
Results

The two-week prevalence of U5C diarrheal disease was 24.9% with 95% CI: (20.4 – 29.7). Age group of 1-12months [AOR: 9.22, 95%CI: (2.93-29.04)], age group of 13-24months [AOR: 4.44, 95%CI: (1.87-10.56)], low monthly income (AOR: 3.68, 95% CI: (1.81-7.51)], small family size [AOR: 0.32, 95% CI: (0.16-0.65)], poor hand washing practice [AOR: 8.37, 95% CI: (3.12-22.52), and immediate feed for cooked foods [AOR: 0.39, 95%CI: (0.19-0.81)] were significantly associated factors with childhood diarrhea.
Conclusion

Diarrhea was a common health problem among under-five children in the periphery area of Azezo sub-city. Age of children, family size, monthly income, hand-washing practice, and feeding time for cooked food were identified statistically significant risk factors. Therefore, an appropriate intervention program through health education would be recommended with a focus on identified risk factors to reduce the burden of diarrheal diseases.

Diarrheal illness is a main public health problem around the globe. Internationally, 525, 000 under-5-yearold children die from diarrhea annually, approximately 2195 every single day (5). It accounts for 9% of child deaths worldwide and making childhood diarrhea the second leading cause of mortality next to pneumonia(6). Although there is a global reduction in the death rates of under-ve children, the risk of a child for death earlier the age of under ve years remains the highest in sub-Saharan African countries (90 per 1000 live births), which is approximately 7 times higher than that in European Regions (12 per 1000 live births) (7). That implies the burden of diarrheal diseases in developing countries is higher than in more developed countries (8)(9)(10)(11)(12).
In Ethiopia, the toll of diarrhea is another killer of under-ve children following pneumonia (11). A previous study showed that in Ethiopia the incidence of diarrhea declined from 2000 to 2016, the problem is extremely high for children under ve years old by causing loss of uids with substantial mortality in infancy stages (13). Various studies revealed that diarrheal illness in unindustrialized countries like Ethiopia is due to many reasons such as unsafe water supply, lack of water linked to inadequate hygiene, poor personal and domestic hygiene, poor sanitation, failure to wash hands at the critical time, accessibility of toilet facilities, housing environments, educational status of mothers, and poor feeding practices (8- 12,14,15). Still, there is limited evidence in the study area towards diarrheal disease and predictors among under-ve children. Therefore, this investigation aimed to evaluate the prevalence of diarrheal disease and associated factors of under-ve children in the periphery area of Azezo sub-city, Gondar, northwest Ethiopia.

Methods And Materials
Study design and period A community-based cross-sectional study design was employed to collect data from households having U5C in April 2019.

Study area and population
The study was conducted at the periphery area located at Azezo sub-city, Gondar, about 727 km far from Addis Ababa, northwest Ethiopia. According to the 2019 Ethiopian budget year, the total residents of this study area were estimated to be about 68,538 populations. Total population of 68,538 of which urban residents account for about 40,430 while the remaining 28,108 are rural inhabitants. In Azezo periphery area there are more than 513 U5C exists from a total population of 4198 and a total of 901 households with 14 villages.

Sample size determination and sampling technique
The optimal sample size was determined using a single population proportion formula, [n=Z 2 α/2 * P(1-P)*D/d 2 ] based on the following assumptions as 95% con dence interval (Z=1.96), the proportion of childhood diarrhea 22.1% a study nding from central Gondar zone(16), and 5% margin of error (d), and design effect (D =1.5). Then after adding the non -response rate (5%), the total sample size was 417 households paired with at least one U5C. Cluster sampling technique was implemented to select ten legible villages with a simple random sampling technique from the total number of fourteen villages present in the study area and then all households with U5C were recruited in randomly selected villages.
The questionnaires were divided into three sub-sections to address a different part of the questions.
Section-I: Socio-demographic and economic factors: age, sex, number of family members, level of education, occupation, religion, housing condition, and household income. Section-II: Water-related factors: distance of water source to the household and waste disposal sites, type of water storage containers, type of water source, eating food prepared, and utensils washed with contaminated water, bathing, and washing in unhygienic water. Section-III: Sanitation factors: availability of human and animal waste and household garbage disposal sites, latrine availability, type of latrine, utilization of latrine, cleanliness of latrine, availability of sewage system, presence of domestic animals and ies inside the house. Section-IV: Behavioral factors: hand-washing practice at critical time, raw and prepared food and drink storage places and handling habits, unclean feeding bottles, washing utensils, cleanliness of domestic places, water drawing practice, cleaning practice of the oor, waste disposal practice, vaccination status and breastfeeding.

Data quality control
The questionnaire was rst prepared in English and then translated to the local language Amharic and back to English for data entry purposes to evaluate the consistency of the questionnaire. Three B.Sc. in Environmental Health professionals were involved in data collection with close supervision after delivering two-day training. Pretest was done from ve percent of samples outside of the study area to check the response, language clarity, and appropriateness of the questionnaires before starting the actual data collection. Then the identi ed ambiguous questions had been corrected before the interview. Every day after data collection each questionnaire was reviewed and checked for its completeness by the supervisors and principal investigators. Double-entry of data and frequency checks were also done during the data entry, processing, and analysis. The test result of Cronbach-alpha was 0.8, which indicates the acceptable internal consistency of the measurement reliabilities.
Data processing and analysis Data were checked for completeness and relevance, coded and entered into Epi Info version 7 computer software, and analyzed using SPSS for Windows version 26. The descriptive statistical (frequency, percentages, mean and standard division ) analysis was used to compute the ndings of this study. The results were then presented using tables, graphs, and texts. The strength of association of predictors with the outcome variable had been computed using the odds ratio with a 95% con dence interval. Bivariate analysis was performed to assess the binary relation between the outcome and independent variables. All variables with P-value less than 0.2 were included in the multivariable analysis model and then a P-value of less than 0.05 was considered as statistically signi cant predictors.

Results
Socio-demographic and economic characteristics of respondents Four hundred and seven participants were enrolled in this study with a 100% response rate. All the respondents were women with a mean (±SD) age of 32.5 (±5.5) years. Most (96.4%) of participants were biological mothers. Ninety-two percent (384/417) of the mothers were married and had a median family size was six persons per household. The average family income of participants was 4458 ETB. The majority (62.6%) of participants could not read and write (Table 1).

Diarrheal prevalence among under-ve Children
The two-week prevalence of U5C diarrheal disease was 24.9% with 95% CI: (20.4, 29.7) in which watery diarrhea takes a higher proportion, which was 78.8% of the overall prevalence. The prevalence of underve diarrhea is higher in children in the age group 13-14 months of age 42(40.4%) followed by age group 1-12months 40(38.5%). The rest 22 (21.1%) was aged 25 and higher.    Open eld 332 79.6 @ One household disposed of the waste by using more than one method.
Behavioral/hygienic practices of the participants The majority of 293(70.3%) household members were poorly practiced hand-washing at the critical time. However, only water is the most accessible hand-wash material for the majority of households accounting for 329 (78.9%) of the total participants. The larger proportion of the study participants did not have separate food preparing materials for their children. More than three-quarters (76.7%) of the study participants fed cooked foods immediately after cooking to their children. Of the one-third (33.6%) of the participants were had leftover food storing habits. From which about 87 (62.1%) of participants reheat leftover foods before consumption (Table 5).   (20) and Afar region, (26.1%) (21). However, This nd is higher than the ndings in Serobo town Jimma (12%) (22), Wolaita Sodo (11%)(23) and the 2016 EDHS national report (12%)(16), Tanzania (6.1%)(24), Northern Nigeria, (12.7%) (9). These differences might be due to the variation in the sociodemographic characteristics, study setting, environmental factors, study period, behavioral characteristics of the study households. This showed that the occurrence of diarrheal disease was varied among children in different study areas.
In factor analysis outcome, the present study revealed that young children were more exposed to the burden of diarrheal disease. The odds of childhood diarrhea were higher than in children whose age is less than two years when compared with two years and above age groups. This could be due to multiple reasons such as the young children started crawling and walking which increases environmental exposure for infectious agents, children's transition period from exclusive breastfeeding to complementary food supplementation that may be contaminated, and less development of immune systems that may easily be affected by diarrheal diseases. This nding is supported by the ndings from the study was done in different parts of Ethiopia such as Kersa district, Afar, Benishangul Gumuz, and Sidama zone (15,21,(25)(26)(27).
As shown in this study, the burden of diarrheal disease was about 3.5 times higher among children with lower family incomes than their high-income counterparts. This might be related to the inability to access water, sanitation, and hygiene technologies and to cover direct medical costs as well as it might be a bottleneck to ful ll the necessary dietary components to their children as a result they might become malnourished and increase the risk of acquiring the diarrheal disease. Our study nding is consistent with the ndings in India, Bangladesh, Senegal, Nigeria, Rwanda, and Eretria (21,(28)(29)(30)(31)(32).
The nding of the current study revealed that lower family size was a protective factor for the burden of childhood diarrhea. This is the fact that high family size might be positively associated with under-ve diarrheal disease in which as the number of siblings increases mothers/caretakers are exposed to high burden responsibility in different household activities. As a result, this might be diverted the mother's attention of care to their children and this may create an abandoned opportunity for children to expose unhygienic environments. This study nding is consistent with ndings of other similar studies in Kersa district, Sengale, Wolitta Soddo Town, and Eritrea (15,20,23,28).
The current nding was also obtained a statistically signi cant association of diarrheal disease with poor hand hygiene practices of child guardians at a critical time. The occurrence of childhood diarrhea was increased by eightfold due to poor hand-washing practices after latrine visits, during child meal preparation, and after refuse handling. The possible explanation might be the high microbial load may be exposed to various health threats as contamination of hands by various types of pathogens. This nding was in line with studies in India, Somalia, Ethiopia in which children's whose mothers/caretakers are failed to practice good hand-washing at critical times were at higher risk of diarrhea as compared with children's whose mothers/caretakers who are practicing good hand washing abilities (3,18,33).
The outcome of this study was also indicated that odds of diarrheal disease were strongly associated with the time elapsed for the prepared food given to children. The odds of diarrhea among children who feed cooked foods immediately were lower when compared with children who feed cooked foods that are waiting for an unde ned period. This might be due to some strains of pathogenic agents having the capacity to survive and reproduce in prepared foods through time which leads to diarrheal diseases. In addition, the prepared child food may be exposed to recontamination when stored for a long time. Studies were done in Afar and Dessie Ethiopia showed that long storage of cooked food-related problems was strongly associated with under-ve diarrheal disease (34,35).

Limitations
It is very di cult to establish a causal relationship due to the inherent nature of the cross-sectional study design. Recall and/or social desirability biases may be introduced in interview-based data collection. The self-report of mothers/caretakers may cause under or over-estimation of diarrheal disease outcomes.

Conclusions
This study demonstrated that childhood diarrheal disease was a common health problem among underve children in the periphery area of the Azezo sub-city like another part of Ethiopia. Compared with the national and other similar study ndings, a higher prevalence of diarrhea was reported in this study area. This study was also identi ed that children in the young age group, a high family size, low household income, poor hand-wash practice, and longtime storage for cooked food were statistically signi cant potential risk factors of childhood diarrheal diseases. Therefore, an appropriate intervention program through health education would be recommended with a focus on identi ed risk factors to reduce the burden of diarrheal diseases. The participants that are mothers/caretakers with child paired had been informed about the aim of the study and the right to withdraw at any time if they were felt discomfort. The participant's privacy and con dentiality were strongly maintained using codes rather than personal identi ers. A statement to con rm that all methods were carried out following relevant guidelines and regulations of the crosssectional study.

Consent for publication
Not applicable in consent for publication due to participants' images or identity revealing data is not used.

Availability of data
All data generated or analyzed during this study are included in this to be published article [and its supplementary information les].

Con icts of Interest
The authors declared that they have no con icts of interest in nancial and non-nancial perspectives. Funding