Back

Modelling the risk factors of malnutrition amongst children below five years of age in Uganda. A GSEM based analysis

Maniragaba, V. N.; Leonard, A.; Rutayisire, P. C.

2022-04-19 epidemiology
10.1101/2022.04.18.22273977 medRxiv
Show abstract

This study aimed at establishing the causal mechanism through which the risk factors of malnutrition act to influence the undernutrition of under-fives in Uganda. A generalized structural equation model (GSEM) was estimated based on the multifaceted nature of the risk factors according to the UNICEF 1991 conceptual framework for child malnutrition. Uganda Demographic and Health Survey (UDHS-KR) dataset was used for the analysis. An index of malnutrition which was the overall response variable in the GSEM model was derived from the three anthropometric variables of stunting, underweight, and body wasting. The findings indicate that sex of the child, age of the child, the perceived size of the child at birth, the weight of the child at birth, duration of breastfeeding, mothers education level, working status of the mother, wealth quintile of the household and mothers weight (BMI) were the factors that had a direct and significant influence on undernutrition. Further analysis of the causal mechanism reveals about five main paths of causal mechanism and these were; (i) the mothers level of education influences the mothers weight which influences the birth weight of the child and then this combination influences under-nutrition, (ii) mothers education levels also influence the household wealth status and as observed above, household wealth significantly influences under-nutrition, (iii) residence whether rural or urban setting influences under-nutrition through its influence on the household wealth, (iv) being in urban areas is also associated with employment opportunities which in turn influences the duration of breast feeding and then influences under-nutrition, (v) size at birth which is key in influencing under-nutrition is influenced by the working status of the mother which is dependent on the place of residence. Based on the findings, the recommendation is that a multi-faceted approach that addresses all the factors at different levels be implemented in order to get a lasting solution to the silent killer of under-fives.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.