Exposure to atmospheric particulate matter and nitrogen dioxide has been linked to SARS-CoV-2 infection and death. We hypothesized that long-term exposure to farming-related air pollutants might predispose to an increased risk of COVID-19-related death. To test this hypothesis, we performed an ecological study of five Italian Regions (Piedmont, Lombardy, Veneto, Emilia-Romagna and Sicily), linking all-cause mortality by province (administrative entities within regions) to data on atmospheric concentrations of particulate matter (PM_{2.5} and PM_{10}) and ammonia (NH_{3}), which are mainly produced by agricultural activities. The study outcome was change in all-cause mortality during March-April 2020 compared with March-April 2015-2019 (period). We estimated all-cause mortality rate ratios (MRRs) by multivariate negative binomial regression models adjusting for air temperature, humidity, international import-export, gross domestic product and population density. We documented a 6.9% excess in MRR (proxy for COVID-19 mortality) for each tonne/km^{2} increase in NH_{3} emissions, explained by the interaction of the period variable with NH_{3} exposure, considering all pollutants together. Despite the limitations of the ecological design of the study, following the precautionary principle, we recommend the implementation of public health measures to limit environmental NH_{3} exposure, particularly while the COVID-19 pandemic continues. Future studies are needed to investigate any causal link between COVID-19 and farming-related pollution.
【저자키워드】 COVID-19, SARS-CoV-2, Mortality, Particulate matter, livestock, Pollution, Import and export, ammonia, gross domestic product, Farming,