Risk Factors for Anaemia, Iron Deficiency, and Iron Deficiency Anaemia in Women of Reproductive Age Using Logistic Regression

Abstract

Women of reproductive age (WRA) are vulnerable to anaemia, iron deficiency (ID), or iron deficiency anaemia (IDA). To identify the factors influencing anaemia, ID, and IDA to WRA in Indonesia, logistic regression analysis was employed. This study aims to determine the prevalence of anaemia, ID, and AID among WRA, as well as to identify influencing factors and evaluate the classification results produced by Logistic Regression methods. The data used were obtained from the Research and Development Agency, Ministry of Health of Indonesia. Haemoglobin data, demographic, and socioeconomic data were derived from the Basic Health Research 2013, and ferritin (Fe) and CRP data used stored serum samples collected in 2013 and analyzed in 2016. The results of this study found that the prevalence of anaemia among WRA in Indonesia is 11%, ID 14%, and AID 9%. Significant factors influencing health conditions include BMI, marital status, family size, malaria, and ARI. Individuals with overweight or obesity have a lower chance of experiencing anaemia, ID, and IDA compared to those who are thin, while individuals who are divorced have a higher risk than those who are unmarried. Additionally, individuals affected by malaria or ARI also have a higher risk of experiencing anaemia. Consumption of animal protein and education also emerges as significant factors affecting ID conditions. Although the model using Multinomial Logistic Regression shows higher accuracy than the binary model, both still have weaknesses in identifying cases of anaemia, ID, and IDA with low sensitivity. Model evaluation indicates that despite proficiency in recognizing normal cases, they still struggle to detect cases of anaemia, ID, and IDA.