Assessing the impact of meteorological factors and air pollution on respiratory disease mortality rates: a random forest model analysis (2017-2021)
Yousef Dowlatabadi,Shaghayegh Abadi,Maryam Sarkhosh,Mitra Mohammadi,Seyed Mohammad Mahdi Moezzi
DOI: https://doi.org/10.1038/s41598-024-74440-2
2024-10-19
Abstract:Air pollution poses a significant threat to the health of all living beings on our planet. It has been scientifically established as a crucial factor affecting mortality rates, respiratory illnesses, mental well-being, and overall health. This study aimed to investigate the impact of air pollution and meteorological factors on respiratory disease mortality rates in Mashhad in 2017-2021 using a Random Forest (RF) model. At first, the daily statistics of meteorological parameters (pressure, humidity, temperature, solar radiation) during 2017-2021 were collected. The information related to pollutants pollutants such as PM2.5 (which is defined as particulate matter with less than 2.5 micrometer diameter and potentially harmful to humans), PM10 (Particles with a diameter of 10 micrometers or less that can negatively impact both human health and environmental conditions.), sulfur dioxide (SO2), nitrogen dioxide (NO2), and carbon monoxide (CO) was collected. the mortality statistics from respiratory diseases were collected from the Health system registaration (Sina). Then we used the RF regression model in Excel and Python software to investigate the interaction between the mentioned variables. The escalating trend of air pollution in Mashhad has led to an expected increase in respiratory-related hospitalizations. This necessitates urgent air pollution control measures. Concurrently, the study of pollutants and climatic elements, as substantiated by global epidemiological studies, is crucial. In Mashhad, the second most polluted city in Iran, climatic factors like humidity, sunshine duration, temperature, pressure, and sunlight intensity exacerbate atmospheric pollution levels, impacting human health and ecosystems. The R2, RSME, and MAE of RF model are 0.73, 2.52, and 2 which indicate that the model has successfully identified the relationship between input variables and the target variable, and it will exhibit high accuracy in predictions. In this study, the most influential factor was identified when the Variance Inflation (VI) factor reached a value of 0.548. This indicates a strong correlation between this factor and the death rate of patients during the specified period. Furthermore, we analyzed by excluding the day and month plans from our model. The results showed that the factor with the highest coefficient in the executive model was related to pressure, with a VI value of 0.049. This suggests that pressure plays a significant role in our model and has a substantial effect on the death rate of patients. In the study of various pollutants, it was found that PM10 had the most significant impact on the mortality rate of patients with respiratory conditions, with a VI of 0.039. Following PM10, the pollutants with the next highest coefficients of importance were NO2 (VI = 0.034), SO2 (VI = 0.033), PM2.5 (VI = 0.029), and CO (VI = 0.025), respectively. Furthermore, the study observed a notable increase in the mortality rate of respiratory patients over time. Specifically, for every five days, the death rate increased by 35.5%. Indeed, climate change and air pollution significantly contribute to the mortality rate from respiratory diseases. Therefore, it is crucial for individuals, particularly those with respiratory conditions, to heed meteorological warnings.