Spatial Distribution and Modeling of malnutrition among under-five Children in Ethiopia

Background: Majority of this world is concerned by malnutrition. Ethiopia is one of the Sub Saharan Africancountries known by poverty, childhood diseases, high mortality and poor infrastructures and technology. The study aimed to examine differences within individuals and between clusters in nutritional status of under-five children and to identify socioeconomic factors using adequate nutrition of children in Ethiopia. Method: Data was obtained from Ethiopian 2019 Mini Demographic and Health Survey surveyed by Ethiopian Public Health Institute. A weighted sub- sample of 8768 under-five children was drawn from the dataset. Spatial statistics was used to analysis spatial variations of malnutrition of children in clusters of regional areas of Ethiopia. Multilevel modeling was used to look at demographic, socioeconomic factors at individuals and clusters levels. Result: At national level the proportion of stunting, underweight and wasting among under-five children were 39.5 percent, 29.8 percent and 15.4 percent respectively. The Global Moran Index’s value for children malnutrition result in Ethiopia was (for stunting I = 0.204, P-value =<0.0001, for underweight I = 0.195, P-value = < 0.0001 and for wasting I = 0.152, P-value=<0.0001). Spatial variability of malnutrition of under-five children across the clusters of Ethiopia observed. Result of heterogeneity between clusters obtained was 𝑋 2 = 147.25, 𝑋 2 = 211.43 𝑎𝑛𝑑 𝑋 2 = 201.43 respectively for stunting, underweight and wasting with P=< 0.0001 providing evidences of variation among regional clusters with respect to the status of nutrition of under-five children.Multilevel model result revealed that high differences of malnutritionin individual households and regional clusters in the under-five children in Ethiopia. Conclusion: The model showed that there were spatial variations in malnutrition among clusters in Ethiopia. Child age in month, breast feeding, family educational level, wealth index, place of residence, media access and region were highly significantly associated with childhood malnutrition. Inclusion of explanatory variables in multilevel model has shown that a significant impact on variation in malnutrition among individual households and regional clusters. Accessible resources, promoting education,use media to expand activities regarding nutritional and health services and through health workers and health institutions in Ethiopia is significant.


Background
Malnutrition is allied with the images of starving children who are suffering from severe acute malnutrition with their bellies bloated and their arms and legs pain fully boney.It can indeed refer to short-term acute malnutrition. But further malnutrition is chronic and lifelong, a highly preventable condition that begins in early childhood and continues into old age, divesting one generation and passing the miserable legacy on to the next. Malnutrition is a significant influence for maternal and child health and is the outcome of deficient of diet, poor care and infection diseases [1,2].
According to World Health Organization (WHO) malnutrition indicated in three categories named as stunting, wasting and underweight considered by height for age, weight for age and weight for age indexed respectively [3] Most under nutrition begins during pregnancy referred to as intrauterine growth retardation and the first two years of life, leading to higher infant mortality, stunting, low birth weight, and premature delivery. Each of indicatorsmeasures somewhat different aspects of nutritional status.
Note that higher values of a z score indicates better nutritional status and vice versa.
Therefore, a decrease of z-scores indicates an increase in malnutrition [4].
Ethiopia has known shows potential progress in dropping levels of malnutrition over two past decades. However, the baseline levels of malnutrition remain so high that the country still needs to continue substantial investment in nutrition. According to Ethiopian Demographic and Health Survey (DHS), there is a substantial variation of under-five children nutrition in Ethiopia. For instance, there is regional variation in the prevalence of stunting in children; the estimated prevalence of chronic malnutrition is above the national average in Affar regional administration (49 percent), Tigray regional administration (44 percent), SNNP of Ethiopia (44 percent) and Amhara regional administration years [5,6] Variations in under-five malnutrition proportion is associated with individuals socio-economic factors because it determines the amount of resources such as food, good sanitation, and health care that are available to infants and neglected temporal and geographic gradients and other variations in risk, in order to generate hypothesis towards the cause of malnutrition.
For example, the exposure to stunting is higher in rural areas, among children because of mothers in urban have more awareness than in rural [7,8]. Educated mothers and fathers have better health-seeking behavior for childhood illness as compared to uneducated which can help prevent malnutrition [9,10]. indicates that statistical significance of spatial autocorrelation in model [11].

Multilevel Modeling
A multilevel logistic regression model also referred to as a hierarchal model, can account for lack of independence across levels of nested data (individuals within groups).
Multilevel modeling relaxes this assumption and allows the effects of these variables to vary across groups. It tolerates also not only independent variables at any level of a hierarchical structure, but also at least one random effect above level one [12].

Multivariate Response Multilevel modeling
Multivariate response multilevel models may possibly be necessary when one is interested in two or more outcomes measured on individuals within group. The researcher is interested in drawing conclusions about the degree to which the residual correlations depend on the individual and the group level; investigate specific effect of a covariate across two or more outcomes; and interest in conducting a single test of a joint effect of a covariate on two or more outcomes.
Then the two-level model can be written as; Where is the random effect at level-two.
Equation (1) can be a standard logistic model without . Therefore, and 0 = 0 + Comparing to equations (2) and (3), equation (1) is said to be combined model [12]. respectively.It is thought that exposure of media like radio, television and newspapers helps to improve health care and practiced of feeding.

Result
Accordingly, households who were not exposed to any kind of media (40.0%) were found with stunted children than those who were exposed (35.5%).Households who were not exposed to any kind of media (16.8% and 29.5%) were found with more wasted and underweight children than those exposed to media (13.1%

Multivariate Multilevel Regression Model
To identify determinant factors of child  Table   2).

Discussion
In model two individual factors included and in model three cluster level factors are included (Table 3).  Table 3). Therefore, childhood malnutrition was found to be high in mothers and fathers with not educated formally.   Ethical approval and consent to participate:

Indicators
Ethical approval and permission was obtained from ethical committee research of Ambo University.