Abstract: Numerous studies have been conducted to examine the association between exposure to various air pollutants, including carbon dioxide (CO2), carbon monoxide (CO), and ozone (O3), and their impact
on the overall rate of completed suicides. However, the specific mechanisms through which air pollution influences the increased suicide rate remain unclear.Hence, the objective of our study was to utilize publicly accessible open air pollution to forecast the escalation in suicide rates.We extracted relevant air pollutant data using the public data provided by the Organization for Economic Co-operation and Development
(OECD). Our study included countries of Turkey, Greece, Slovak Republic, United Kingdom, Canada, Luxembourg, Poland, Japan, Hungary, Lithuania, Italy, Spain, Portugal, Ireland, Denmark, Germany, Chile,
Netherlands, Canada, Austria, Czech Republic, Switzerland, Australia, Sweden, Iceland, United States, Finland, Latvia, Belgium, Estonia, Slovenia, and South Korea.Random Forest (RF) Machine Learning (ML)
model showed that we predicted a low or high suicide rate status with an accuracy of 90% (area under curve (AUC): 94%). The level of CO2 was the most predictive factor according to the RF model.In conclusion, specific levels of air pollutants can potentially be utilized to predict whether a region has a high or low rate of completed suicides.
Keywords: air pollution, suicide, prevalence, air pollutants, carbon monoxide, greenhouse gas, nitrogen
oxides, sulfur dioxide.
link: https://www.airitilibrary.com/Article/Detail/P20220411001-N202410090003-00005
cite: Uludag, K. (2024). Leveraging Machine Learning to Investigate the Link between Exposure to Major Air Pollutants and the Escalation of Suicide Rates in OECD Countries. Journal of Suicidology, (in press), 835-841. https://doi.org/10.30126/JoS.202406_19(2).0004