Summary

A poor start in the first 1,000 days of life can have lifelong consequences for physical and mental health, as well as for economic and social participation. Although early preventive interventions and well-designed policies can yield substantial long-term benefits, vulnerable pregnant women, their partners, and children are often identified too late or receive support that is insufficiently tailored to their needs. This project investigates how the analysis of linked, nationally representative administrative data using machine learning techniques can improve the early identification of vulnerable circumstances in the first 1000 days of life and support more targeted and effective policies. Specifically, we assess the added value of machine learning compared with traditional statistical approaches, with the aim of strengthening prevention and personalized care in the first 1,000 days of life.