Analyzing Compression of Comorbidity in Italy: A Mixture Model with Regional Random Effects

This study pursues three objectives. Firstly, we propose a comorbidity index defined as a latent variable incorporating the disability weights from the Global Burden of Disease (GBD) projects into the estimation process. Secondly, we model the nonlinear relationship between this novel comorbidity index and socio-demographic features using a mixed-mixture model, which accommodates zero inflation and variability between the Italian regions. Lastly, an initial exploration of the concept of comorbidity compression in socio-demographic subpopulations is provided by analyzing the PASSI data across twelve years.