Study on the relationship between air quality and mortality due to respiratory diseases at global level

The aim of this study is to bring new knowledge and verify some assumptions about how different environmental and socio-economic conditions affect mortality worldwide. From the three main causes of mortality, only one mortality due to join effects of household and ambient air pollution is analysed in this work. The study is carried out for a selection of 165 countries. Open data from several sources (the United Nations, The World Bank, Science Direct and Expansión datos macro) is combined bringing together information about geography, demography, socio-economy, environment (including conditions, emissions and natural resources), energy, education, standard of living, health and mortality. Unsupervised and supervised learning models were used: clustering and PCA first, and then ANN and regression models. With the cross-sectional data studied, a main conclusion is that a clear relationship between the air pollution (presence of PM25) and mortality by respiratory disease is found, as well as GDP_PPP of the countries. Whereas air pollution is associated with increasing mortality, wealth is associated with a decrease in such mortality. Poorest countries show highest mortality rates; thus income levels have the greatest impact. Nevertheless, the analysis of covariance has indicated that between countries with the same income group, climate has an impact and warmer countries have higher mortality rates. All in all, there is still remaining variability that might be explained by other factors, not included in the current database.