Childhood stunting and associated factors among irrigation and non- irrigation user kebeles in Mecha district, northwest, Ethiopia: a comparative cross-sectional study

is a critical of developing like Different interventions like there is scanty of its associated A community-based comparative cross-sectional study design was employed from October to November 2019. A systematic sampling was used to draw mothers with children age 6–59 months (582 irrigated and 582 non- irrigated household users). Data were collected using questionnaire and anthropometric measurement tools. Multivariable logistic regression was used to identify the predictors of stunting. Adjusted odds ratios with 95% CI were used to determine the degree of association between independent and outcome variable. A-p-value < 0.05 was used as cutoff point to declare statistically signicant variables with the outcome variable.


Introduction
Globally, stunting remains one of the major public health problems among children under ve years of age (1). It is an indicator of malnutrition in early childhood and is strongly associated with numerous short term and long-term conditions (2,3). It is the best of overall indicator of children's well-being and an accurate re ection of social inequalities. Use of stunting as a direct indicator of nutritional status have shifted focus away from considering a broad set of environmental and social determinants of child growth to a disproportionate emphasis on dietary determinants of linear growth (4,5). It is the result of the complex interaction between the food we eat, our overall state of health, and the environment in which we live in (6,7).
The burden of stunting is multidimensional which are related with retarded growth, behavior development, poor school performance, lower working capacity, losses in cognitive function, losses in income, decreases productivity, and increased likelihood of chronic disease, if not corrected, it can slow down economic growth and increase poverty levels (5,8,9). In the past twenty years, 55.5 billion Ethiopian birr (ETB) are estimated with the costs associated with child under nutrition, equivalent to 16.5% of Gross Domestic Product (GDP) (4,10,11).
In 2019, the prevalence of stunting in Ethiopia was estimated to be 37%. In the same, the percentage of stunting is higher in Amhara region compared to national gure (41%), indicating inappropriate child feeding practices in the region. Stunting for children under age 5 sharply increases between age 6 and 23 months, and peaks at age 24-35 months; this represents the impact of undernutrition in the rst 1,000 days of life (12).
Different intervention with national nutrition strategies, nutrition speci c and nutrition sensitive intervention activities with focus on improving Infant and Young Child Feeding have been implemented to reduce chronic malnutrition. Increasing agricultural production through irrigation system to meet food security at household level is also one of the strategies to address the basic needs of the community.
Studies on prevalence of childhood stunting so far have been conducted in general community (5,13,14), which are either irrigated and/or non-irrigated users. To the best of our knowledge, there is not speci c study of prevalence of childhood stunting and associated factors among irrigation and nonirrigation users. This study aimed to compare the prevalence of childhood stunting and its associated factors among irrigation and non-irrigation users in North Mecha district, Amhara Region, Ethiopia. The ndings from this study will help to end users in identify the gap and maximize the potential of irrigation for their health bene ts.

Study setting and period
The study area is Mecha district which is located at 530 kms in Northwest direction of Addis Ababa, the capital city of Ethiopia, and 35 kms in the southwest direction of Bahir Dar, capital city of Amhrara National Region State. It is one of the thirteen districts found in West Gojjam Administrative Zone. The district comprises three climatic zone; high land "Dega", mid-altitude "Wena Dega" and lowlands "Kola". The mean annual rainfall ranges from 1,000 mm to 2,000 mm . The district has 156,027 hectares area, of   which 72,178 hectares are used for cultivation and about 1,386 hectares covered by water bodies. The  district has 40 Kebeles (10 irrigation users and others are non-irrigated). Based on the 2007 national  census, the estimated population of the district in 2019, is about 303,208, of which 150,088 are males and 153,120 are females. The district has 10 health center and 38 health posts and 1 government hospital during the time of the study. The district is one of the food surplus areas and the commonest staple foods are maize, millet, teff and barley (15). The study was conducted from October to December in 2019,

Study design and population
A community-based comparative cross-sectional study design was employed to assess the prevalence and its associated factors of childhood stunting. All mothers/care-takers with children age 6-59 months in both irrigation users and non-irrigation were source of population in the district. All mothers/care takers with children age 6-59 months in the randomly selected Kebele were study populations.

Sample size determination
The adequate sample size for this study was determined using two-population proportion formula [n (in each group) = f (α, β) (p1q1 + p2q2) / (p1 -p2)²] with the following assumptions: f (α, β) = 7.84, when the power = 80%, the level of signi cance = 5% and the prevalence of stunting among irrigating (P1) was 10% and non-irrigated (P2) was17% from previous study in Kenya (16), design effect of 1.5 and 5% of nonresponses rate. Based on the above assumptions, a sample size of 1164 HHs (582 irrigation users and 582 non-irrigation users) was calculated for the study.
Sampling technique and sampling procedure First, Kebeles (the smallest administrative unit in the country) in the district were strati ed based on irrigation status (irrigation users and non-irrigation users). Six Kebeles from each stratum were selected randomly and proportion to size allocation was made to determine the required households from each randomly selected kebeles. Households from randomly selected kebeles were selected using systematic sampling technique. Sampling interval was determined by dividing the total number of households in each kebele with the allocated sample size.

Data collection procedure
Socio-demographic and other data were collecting by using structured questionnaire adapted from different standard questionnaires and literatures. The anthropometric data was collected by using WHO Recommendations for data collection, analysis and reporting on anthropometric indicators in children under 5 years old (17). A vertical or horizontal measuring board reading a maximum of 175 cm and capable of measuring to 0.1 cm was used to take the height of a child. The child stands on the measuring board barefoot; have hands hanging loosely with feet parallel to the body, and heels, buttocks, shoulders and back of the head touching the board. The head was held comfortably erect with the lower border of the orbit of the eye being in the same horizontal plane as the external canal of the ear. The headpiece of the measuring board is then pushed gently, crushing the hair and making contact with the top of the head. Height is then read to the nearest 0.1 cm. For those with written evidence, date of birth was obtained from clinic cards, child health cards and immunization status certi cates of the children, unless by asking mother/care takers age was determined.

Data quality assurance
Pretest of the questionnaire, training and calipering of the weigh scale were done to assure data quality. The questionnaire was developed from previous literature and eld pre-tested on ve percent of sample size was done. Before the actual data collection, training was given for data collectors and supervisors. During the training, the trainers give instructions on the questions to be asked, their meaning, how to ask them, and how to record the answers. Data collectors and supervisors were also trained using role play practices. The equipment that was used to measure the anthropometric variables was calibrated between ve respondents for actual data collection by using a known weight material. At the end of every data collection day, each questionnaire was examined for its completeness and consistency by the principal investigator and pertinent feedback was given to the data collectors and supervisors. All completed Epi (survey solution software) collected data entered questionnaire was conducted daily and Examined by principal investigator.

Data processing and analysis
Data were collected using Epi collect survey solution software and directly downloaded to Excel, exported to SPSS version 23. Data was cleaned before data analysis. The anthropometric data was exported to WHO Anthro plus software to computed individual level of nutritional status. Frequencies, proportions and chi-square were computed for description of the study population in relation to socio-demographic and other variables. Bivariate and multivariable analysis logistic regressions were done to determine the association between stunting and explanatory variables. Those independent variables whose p-value less than 0.2 at the bivariable regression were included in a multivariable logistic regression model. Adjusted odds ratio with 95% CI was calculated to measure the degree of association between stunting and explanatory variables. A-p-value < 0.05 was used as cutoff point to declared statistically signi cant variables with the outcome variable.

Ethical consideration
The study was approved by the Ethics Review Board of College of Medicine and Health Sciences, Bahir Dar University. A support letter from the University was written to the Amhara Regional Health Bureau.
Permission and a letter of support for the study were then obtained from the Amhara Region Health Bureau and the Regional Health Bureau letter was given to the district health o ce to commence the study. Information related ethical issue was given for mothers/caretakers (since most of them are farmers and could not able to read and write) by reading its content attached together in the rst part of the questionnaire. Each study participant was briefed on the objective of the study, and informed about the right to refuse to participate in the study or to discontinue at any time. Participants were also informed that all data was con dential and personal identi ers were not registered. Interview was conducted after the parents and/or legal guardians of the child participants on their behalf provided verbal consent to participate in the study.

Socio-demographic and economic status of participants
Of 1164 Households (HHs) randomly selected to participate in the study, 582 HHs from each irrigation and non-irrigation users were included in the analysis. Sixty eight percent of the respondent were found at greater and equal to 30 years, 71.0% from irrigated and 65% from non-irrigated users. Seventy-six from (76.3%) irrigation users and 72.2% from non-irrigation users had greater than four family members. Above eighty percent of mothers from irrigation users (83.0%) and non-irrigation (80.9%) were unable to read and write (Table 1).
Forty-one percent of mothers from irrigation users and 36.4% of mothers from non-irrigation users had greater than three ANC visit during their pregnancy. Eighty-three percent of mothers from irrigation users and 88.0% of mothers from non-irrigation users had no PNC visit after birth. Regarding health institution location, 24.4% of mothers from irrigation users and 64.4% of mothers from non-irrigation users took one hour and above (Table 1).
Two-third of HHs (75.3%) from irrigation users and 73.9% of HHs from non-irrigation users had improved water source. Sixty-eight of HHs (67.7%) from irrigation users and 38.1% of HHs from non-irrigation users had improved latrine. Regarding hand washing habits, thirteen percent of mothers from irrigation users and 5.2% of mothers from non-irrigation users use always soap and water (Table 1). Nearly one fourth of children (24.6%) from irrigation users and 30.1% of mothers from non-irrigation users were between 24 and 35 months. Seventeen percent of children from irrigation users and 20.4% of mothers from non-irrigation users had diarrhea in the past two weeks before preceding the survey. Eighteen percent of children (24.6%) from irrigation users and 22.7% of mothers from non-irrigation users had fever in the past two weeks before preceding the survey (Table 2).   (19). However, this result is lower than the prevalence reported from the previous study conducted in West Gojjam (43.5%) (20) and in Awi Zone (50%), Amhara region (21). According to the WHO's classi cation, the prevalence of stunting in the study area is high (> 40% is very high, 30 to 39% is high, 20 to 29% is medium and < 20% is classi ed as a low prevalence (22).
The variation in the prevalence of stunting in different studies may be attributed to a broad range of interrelated factors such as environmental or geographical factors (e.g. population density, rainfall and temperature variation and disease environment), socio-cultural differences in relation to child feeding and care practices, maternal under nutrition, HH food insecurity, economic growth and maternal education (19,(23)(24)(25).
This study found that signi cant variation of child stunting did not observe between irrigated and nonirrigated user. The nding of the study is inconsistent with previous studies, which showed that irrigation is associated with a reduced risk of stunting (26,27). Studies revealed that irrigation schemes can ensure household food security (26,27). Although food security may be a necessary prerequisite for good nutrition outcomes, it is insu cient by its own to improve nutritional outcomes (28). The nutritional status of children can be addressed by adopting a multi-sectorial approach (29,30) and determined by mother knowledge on child nutrition practice, maternal nutritional status, intra household food allocation and utilization practices, and access to health services and healthy environmental conditions. For example, the head of the family gets rst priority in eating while mothers and children get a smaller share of the family's food relative to their nutritional need due to hierarchical position with intra-family food distribution (31). A study also revealed that some farmers using irrigation schemes engaged in marketoriented crops (32) which might not be necessary improve the nutritional status of children. Thus, further study is recommended to clearly understand the root cause of child malnutrition in the study area community.
Children aged > 12 months were found a signi cant association with child stunting compared to infants of less than one year in this study which is consistent with other studies (33,34). This is obvious that symptom of stunting occurs after prolonged time if a child is being weaned off of exclusive breastfeeding, does not get adequate diet and proper care, and expose to inadequate environment health facilities such as house with poor hygiene, unsafe water supply and sanitation. Literature indicates improved behavioral practices such as hand washing before eating early on are important to consider along with other factors closely associated with the age of the child (33).
In this study, children whose mother had no ANC visit increase the odds of child stunting compared to multiple contacts during ANC visits which is consistent with previous studies (35,36). Mothers who made frequent and regular contacts with health care provider get a chance for interactive health education sessions. Thus, multiple contacts during ANC enhance mother knowledge of appropriate feeding for their infants after delivery, including breastfeeding and feeding a diverse diet, and practice preventive and curative child care activities such as timely seeking of health care services, which in turn can have a positive effect on children's nutritional status.
The result of this study indicated that male children were slightly more stunted than female children (53% and 47%, respectively) but not at a level of statistical signi cance, which is consistent with other ndings (37,38). There are studies which show males are at higher risk for stunting (39)(40)(41)(42) whereas another study found that females are at higher risk for stunting (43). These sex-related differences may require further study, but seem to be less important than other factors that have been identi ed as affecting children's nutritional status.
In this study, improved water supply and sanitation did not show a signi cant reduction in child stunting, however, hand washing practice with soap was one of the signi cant factors associated with a reduction in child stunting. Inconsistent ndings on bene ts of improved water supply and sanitation and hygiene in child stunting reduction were reported (33,44,45). Some studies showed a strong association between the level of environmental WASH (Water, Sanitation and Hygiene) and child linear growth (46,47). The concept of a clean play and infant feeding area as an additional component of early child development programs along with WASH is one of the recommended strategies to reduce child stunting (47). The literature indicates that sanitation and water supply improvements, hand washing with soap, ensuring a clean play and infant feeding environment and food hygiene are all important to interrupt the multiple transmission pathways of fecal-oral pathogens (48). Better hygiene practices are essential to prevent diarrhea and other infections among children, which in turn contribute to reducing child stunting.
Previous studies on the association between improved drinking water stunting found inconsistent ndings, while other studies have shown that the de nition of "improved" does not reliably predict the microbiological safety of the water (49,50). A study done by Ran et al. in rural India found that household access to an improved source of drinking water or piped water was not associated with child stunting (51). On the other hand, a study done by Fink et al. using the merged data set of 171 Demographic and Health Surveys found access to improved drinking water was associated with a lower risk of mild or severe stunting (52). A longitudinal study conducted in Peru by Checkleyet al. found that there was a positive association between improved water sources and reduction of child stunting only when it was accompanied by improved sanitation and water storage practices (53), so that multiple routes of transmission are addressed. A lack of synergistic effects of improved drinking water and sanitation was also found in a large prospective cohort study in Sudan (54). The possible reasons for this variation may be related with initial water quality at the source, and with water handling practices during collection, transport and storage at the household level. There is evidence that improved water sources do not correlate well with water that is safe for human consumption (55,56) and that water quality deteriorates from source to point of consumption due to unsafe collection, transport, storage and handling practices (57,58). Thus, further research is required to determine if an improved water source that ful lls the recommended quality standard for consumption, along with safe drinking water handling and storage practices, and the combined effect with improved sanitation, have independent or synergistic effects on child stunting.
The study has strengths and limitations. One strength is that this study employed multivariable logistic regression models to control confounding effect and determine the independent effect of irrigation users on child stunting. There might be a social desirability bias and a recall bias during answering of questions related to hygienic practices and events happening in the past, such as the child's history of illness and child care practices. Moreover, information on some important confounding variables such as mothers' nutritional status during pregnancy, parasitic infection and the child's birth weight were not collected, and these could also have in uenced children's nutritional status. Interpretations of factors associated with child stunting should be with caution since a cause and effect relationship cannot be established using a cross-sectional study design.

Conclusions
In conclusion, there is a high prevalence of stunting in children from 6-59 months in the study area, with no variation between irrigation and non-irrigation users. Mothers who had no ANC visit, children whose age was between 12 and 47 months, mothers who did not practice hand washing with soap always and children who had fever were the factors associated with higher child stunting. As the government of Ethiopia confront to reduce food insecurity and child stunting through irrigation practices. This study provides insight that irrigation only may not su cient to improve the nutritional status of children. Therefore, decision makers and program implementers may bene t from the ndings of this study to revise strategies and integrate agricultural practice with nutrition education. Strengthening implementation of the government's agricultural practice, combined with interventions to empower women during ANC visit, could accelerate further reductions in child stunting.
Abbreviations AOR: Adjusted Odds Ratio; CI:Con dence Interval; COR:Crude Odds Ratio; SD:Standard Deviations; and WHO:World Health Organization.