Abstract

Low birth weight (LBW) is a major determinant of infant mortality, morbidity, and future child development. This study aims to develop a predictive score to identify pregnant women at risk of giving birth to a low birth weight child. A case-control study was conducted among 126 mother-newborn pairs at four health centers in Cotonou. The cases were newborns with a weight of less than 2500 g, while the controls had a normal weight. Logistic regression was used to identify the predictive factors of LBW. Score performance was evaluated using the ROC curve, Hosmer-Lemeshow test, and Youden index. The analysis revealed six predictive factors for LBW: BMI less than 18.5, socioeconomic position index less than 35, energy intake less than 2200 Kcal, unskilled occupation (housewife), low level of education, and age below 25 years. The score demonstrated good performance with an area under the ROC curve of 0.930, a sensitivity of 0.857, a specificity of 0.921, a positive predictive value of 0.915, and a Youden index of 0.778. The Hosmer-Lemeshow test indicated good agreement between the observed and predicted values for each subject (p=0.701). This simple and accessible predictive score could be a valuable tool for healthcare professionals to identify pregnant women at risk of LBW and target preventive nutritional interventions to improve birth outcomes.

Key words: Cotonou; low birth weight; risk score; prediction; pregnant.