Modeling Determinants of Anemic Status for Women among Reproductive Age in Ethiopia

Abstract

Anemia is a problem characterized by insufficient red blood cell volume and a low concentration of hemoglobin in the blood (WHO, 2014). Anemia is a global public health problem that affects low, middle and high-income countries and has significant adverse health consequences, as well as adverse impacts on social and economic development (WHO, 2015). According to the 2011 EDHS, 17 percent of Ethiopianwomen age 15-49 are anemic. A higher proportion of anemic women were pregnant (22%) compared to others. Objective: The objective of this study is to model anemic status with associated factors in women among reproductive age in Ethiopia. In addition, we assess the significant variations of anemic patients within and between regions and estimate the prevalence of anemia in women among reproductive age in Ethiopia. Method: A cross-sectional but cluster study carried out based on the secondary data of the Ethiopia Demographic Health Survey 2011. For the categorized response variable, generalized linear model, generalized estimating equations, and generalized linear mixed models are compared to model anemic status of women among reproductive age in Ethiopia to identify the most candidate predictors. That shows us each of them estimates parameters from among different statistical models and comment on the interpretation of parameters and the statistical properties of the methods involved. Data is mainly analyzed using SAS 9.3 and R 3.41 version software offers for the analysis of binary responses for correlated data and both marginal and cluster-specific effects take into account. Results: The study shows that 19.9% of women in the reproductive age group are anemic. The generalized estimating equation is best fits the data for population-averaged effects for given factors of anemic status in women among reproductive age than that of the two models. Generalized linear mixed model with two random intercepts revealed that there is variation between and within regions of anemic status. The result of best model revealed that the variables: Occupation women employed (odds ratio OR=0.6808, 95% CI: 0.6266, 0.7396), higher education level women attain: primary education level (OR=0.7014, 95% CI: 0.6336, 0.7764), secondary education level (OR=0.567, CI: 0.4642, 0.6962) and Higher education level (OR=0.7167, 95% CI: 0.5689, 0.9028), marital status: married (OR=1.5498, 95%CI: 1.6398, 1.7535), living with partner (OR=1.4521, 95%CI: 1.1513, 1.8316), widowed (OR=1.7521, 95%CI: 1.3982, 2.1957), divorced (OR=1.3524, 95%CI: 1.1025, 1.6589) and separated (OR=1.5484, 95%CI: 1.1599, 2.067), wealth index: rich (OR=0.8642, 95%CI: 0.7825, 0.9544), contraceptive use method: modern (OR=0.4966, 95%CI: 4389, 0.5697), pregnant women (OR=1.3338, 95CI: 1.1799, 1.5334) and BMI of women: with 18.5≤BMI<25 (OR=0.7773, 95%CI: 0.7101, 0.8508) and BMI≥25 (OR=0.6849, 95%CI: 0.5686, 0.8249) are significant determinant factors for anemic status at 5% level of significance. Conclusion: Anemia in women among reproductive age is the finaloutcome of the collective effects of health, socio-demographic and economic factors. GEE model with measure of association exhibited the best fit for this data than GLM and GLMM models. The GLMM provided interesting relationships that would not be evident from a standard logistic model. This concluded that there is heterogeneity of anemic status between and within regions.