Statistical and Machine Learning Analysis of Maternal Smoking's Impact on Neonatal Birth Weight

Maternal smoking during pregnancy is a significant public health concern, known to adversely affect neonatal birth outcomes, particularly birth weight. This study investigates the impact of maternal smoking on neonatal birth weight using data from the CDC National Vital Statistics System, which includes 3,737 birth records and 15 maternal and infant variables. The primary objective is to evaluate the extent to which maternal smoking influences birth weight and predict neonatal outcomes. Traditional statistical methods, including ANOVA and multiple comparison techniques like Bonferroni correction and Tukey's HSD, were employed alongside machine learning models such as Random Forest and Neural Networks. The results indicate a significant reduction in birth weight for infants born to mothers who smoked, confirming smoking's detrimental impact on neonatal health. The Random Forest model achieved high predictive accuracy, explaining over 99% of the variance in birth weight. These findings underscore the critical need for public health interventions that target smoking cessation in maternal smoking. Future research should address confounding factors and expand the study to more diverse populations to improve generalizability.