Lack of breastfeeding, low birth weight and smoking during pregnancy can predict behavioral difficulties for child

A study, published in the Journal of Affective Disorders, used artificial intelligence to analyze data from nearly 6,000 children in a longitudinal study of UK households. The aim was to use pregnancy and newborn risk factors to predict emotional and behavioral difficulties in children at the age of five.

The study found differences between the sexes: boys were more affected by maternal smoking, while girls were more vulnerable to fussiness in infancy. According to the study’s lead author, doctoral researcher Xu Zong from the University of Helsinki, these findings highlight the need for gender-responsive approaches in early screening and policy design.

“Lack of breastfeeding, low birth weight and maternal smoking during pregnancy are factors that can be used to react very early, even before the child shows symptoms,” says Zong. “Our findings are particularly relevant in a time when both concern for children’s mental health and demand for early intervention are increasing.”

The researchers’ observations are based on machine learning and big data. Using an advanced AI approach, the study revealed complex relationships and classified the predictive significance of 14 prenatal and postnatal factors.

“Based on our results, greater investments should be made in preventive care right at the beginning of pregnancy,” Zong points out.

Children’s emotional and behavioral difficulties are increasing worldwide. The study provides up-to-date information on how early life risk factors, which begin as early as pregnancy and the newborn, can shape children’s mental well-being years later.

The new international study, led by researchers from the University of Helsinki, was conducted in collaboration with Stockholm University, Karolinska Institutet Sweden and the University of Essex (UK).