A toolkit for dynaMic health Impact analysiS to predicT disability-Related costs in the Aging population based on three case studies of steeL-industry exposed areas in europe (R-13545)

The environment is one of the most crucial determinants of health. The Global Burden of Disease report estimates an emerging impact in terms of disability and reducing thequality of life worldwide, particularly for the aging populations. One of the root causes of this decline is likely to derive from the interaction of socio-environmental risk factorsand sub-clinical conditions and the consequent increase of the primary non-communicable disease burden (e.g., dementia, COPD, cerebrovascular and chronic ischemic heartdiseases). The multi-dimensional causal pathways of these interactions are still mostly unknown. In this complex scenario, where the relationship between exposure andoutcomes is so different and multifaceted, the Health Impact Assessment (HIA) process is the standard tool that provides an overview of the matter, from the screening of healthrisk factors to the introduction of new health policies and the monitoring of effects. A complete digital approach for HIA that could dynamically adapt to the variability of thedeterminants and their interaction is still poorly investigated. Artificial Intelligence algorithms offer innovative and high-performance possibilities for HIA implementations,improving the elaboration and resizing of complex information and data. This proposal aims to develop a technological, dynamic and intelligent HIA toolkit to predict the healthimpact of health-related features, forecasting the trajectories of disability and quality of life reduction. This method will use environmental, socio-economic, geographical, andclinal characteristics, managed and elaborated with a federated learning architecture. The generated models will be adjusted for lifestyle and individual conditions data sourcedfrom large population-based digital surveys. The models will be trained and validated on three different exposures to the steel plants' pollution: Taranta in southern Italy, Rybnikin Poland, and Flanders in Belgium

Keywords:Federated Data, Non-Communicable Diseases, Pollution, Virtual Infrastructure

Disciplines:Epidemiology

Project type:Collaboration project