Using the Binary Logistic Regression Model to Predict Underweight Among Children Under Five "An Analytical Study of Social and Behavioral Determinants in Zawiya City"

Abstract

Background: Childhood malnutrition among children under five years remains one of the most serious global health challenges, contributing to approximately 45% of child deaths. This study aims to identify the social and behavioral determinants associated with malnutrition using binary logistic regression analysis.

Methods: A cross-sectional analytical study was conducted involving 150 children under five years (52% male, 48% female) randomly selected from three primary healthcare centers. Data were collected using a structured questionnaire validated for reliability (Cronbach's α = 0.82). Binary logistic regression was employed to calculate odds ratios (OR) with 95% confidence intervals.

Results: Statistical analysis revealed that the prevalence of underweight reached 33.3%, compared to 13.3% for children with overweight. Three main factors were identified as significantly increasing the likelihood of malnutrition, namely low family income, lack of exclusive breastfeeding, and spending more than two hours daily on electronic screens. It was found that children from low-income families are the most vulnerable, followed by those deprived of exclusive breastfeeding, and then those who spend long hours in front of screens. Moreover, the statistical model demonstrates good explanatory power, accounting for more than 42% of the factors influencing underweight, with an overall classification accuracy of approximately 79%, alongside high sensitivity in detecting affected cases and specificity in excluding healthy ones.