The risk of COVID-19 infection and death may be associated with a heterogeneous distribution of environmental, socioeconomic, and demographic factors at a small-area level. Our aim in this study was to investigate, at a small-area level controlled by socioeconomic and demographic factors, whether long-term exposure to atmospheric pollutants, such as particles and nitrogen dioxide, would increase the risk of both of COVID-19 infection and death in Catalonia, Spain. To this end, we used a mixed longitudinal ecological design applied between 25 February 2020 and 16 May 2020, in which the study population came from small areas of Catalonia. We estimated generalized, mixed, linear models in which we controlled a large number of confounders, both observed and unobserved, as well as the temporal and spatial dependence.
The results of this study suggest that long-term exposure to nitrogen dioxide (NO2) and, to a lesser extent, particles, has been an independent predictor of the spatial spread of COVID-19. In our study, an increase of 1 μm/m3 in the concentration of NO2 resulted in an increase of 2.7% in the mean risk of testing positive for COVID-19. In the case of particles, this risk increased by 3.0%. The regions falling within the third and fourth quartile of NO2 exposure had a 28.8% and 35.7% higher risk of death, respectively, compared with the regions falling within the first two quartiles.
Although there may be biological mechanisms that explain, at least partially, the association between long-term exposure to atmospheric pollutants and COVID-19, our hypothesis was that the spatial spread of COVID-19 in Catalonia was attributed to the different ease with which some human hosts of the virus infected others. This ease of transmission is dependent on the heterogeneous distribution, at the small-area level, of variables such as population density, precarious housing, and mobility of its residents, for whom exposure to pollutants was an intermediate variable.
The study is relevant owing to the fact that it was the first study in Spain, and one of the first studies in the world, that evaluated the effects of long-term exposure to air pollution on the incidence of COVID-19 and the risk of death at the level of small areas in Catalonia, controlling for socioeconomic and demographic factors and using spatiotemporal models. In these models, we controlled a large number of confounders, both observed and unobserved, as well as the temporal and spatial dependence. Furthermore, we exclusively used Open Data for this study.