Analysis of the body mass index of Ethiopian children: General linear mixed-effect model

Abstract

Abstract Background The BMI, formerly known as the Quetelet Index, is a measure of the nutritional status of adults. It is defined as a person’s weight in kilograms divided by the square of their height in meters (kg/m2). According to research in Ethiopia, 44% of children under the age of 5 are underdeveloped (growth retarded), and 21% of them are severely underdeveloped. The incidence of stunting increases as the age of the child increases, with the highest incidence of chronic malnutrition found in children aged 24–35 months (57%) and the lowest in children under six months (ten percent). Male children are slightly more likely to be malnourished than female children (46 and 43 respectively). Methods The study design was surveyed by the central statistical agency to conduct the Ethiopian Rural Socioeconomic Survey, and the study design for this study is clarified. CSA followed the scientific sampling method, a total of 12 households were randomly nominated from the sample of 30 Health Sample Survey households. A generalized linear mixed model was proper to conduct data analysis, and statistical software to run the analysis should be incorporated. The average value, variances, and covariances of body mass index are estimated over the study periods using a generalized linear mixed-effect model. Unstructured is selected for covariance structure which approves the dynamic and monotonic nature of variance displayed by variance function. Result The study presents data on the mean and standard deviation of children's body mass index categorized by demographic factors. It reveals that 52% of the children are male while 48% are female, with the average family size being 5.61 children and the mean age at the first wave being 27.32 months. The research also observes that 34.5% of the children reside in small towns and 65.5% in rural areas, with the overall average body mass index showing a decline from the first wave to the third wave. Regional variations are evident, with SNNP having the highest average body mass index and Somalia the lowest. Additionally, the study examines average trends based on location, region, income, gender, and maternal educational level, and concludes that there is a decreasing trend in the average body mass index, with decreasing variability over the study period. The findings provide valuable insights into the factors that impact children's body mass index. Conclusion In general, current study focused on assessing children's body mass index (BMI) in the Ethiopian Rural Socio-economic Survey data. A random intercept and trend model with an unstructured covariance structure for random effects was employed. The results indicated a decreasing average BMI over the study period, with variations across regions and lower average BMI for females. Both genders showed a decline in BMI over time. Modeling the covariance structure and fixed effects were integral to the analysis. The final model highlighted significant effects of region and age on BMI, implying decreasing average BMI over time with regional variations, influenced mainly by region and age. This study yields valuable insights into children's BMI, presenting compelling evidence for further research.