Multidimensionality of Health Inequalities: A Cross-Country Identification of Health Clusters through Multivariate Classification Techniques.

Despite major efforts in scientific literature to explain and understand the social determinants of health inequalities, the complex association between social causes and health outcomes remains empirically questionable and theoretically puzzling. To date, the studies on social determinants of health has mainly been generated by research techniques and methods that were developed to answer specific questions about the causes and effects of particular indicators on specific health outcomes. The present exploratory study follows a complex system approach to capture the interdependence between socioeconomic status, lifestyles, and health in a single measure that enables international comparisons of population health. Specifically, this study is aimed to: (a) classify individuals' state of health according the usage of multidimensional data on physical and mental health, SES, lifestyles and risk behaviors, in order to (b) compare the relative strength of the different predictors of health groups (or clusters) at the individual-level and, finally, (c) to measure the level of health inequalities between different countries. From a complex system approach, this study uses multivariate classification methods to compare health groups in a sample of 29 countries and shows that interdependence models may be useful to describe and compare between-country health inequalities that are not visible through techniques for the analysis of dependence. The present work offers two fundamental contributions. On the one hand, this study compares the relative relevance of different indicators that are susceptible to affect individual health outcomes; on the other hand, the resulting multidimensional classification of countries according health clusters provides an alternative for inter-country health comparisons.