Assessing long-term effects of gaseous air pollution exposure on mortality in the United States using a variant of...

Long-term mortality effects of particulate air pollution have been investigated in a causal analytic frame, while causal evidence for associations with gaseous air pollutants remains extensively lacking, especially for carbon monoxide (CO) and sulfur dioxide (SO₂). In this study, we estimated the causal relationship of long-term exposure to nitrogen dioxide (NO₂), CO, SO₂, and ozone (O₃) with mortality. Utilizing the data from National Morbidity, Mortality, and Air Pollution Study, we applied a variant of difference-in-differences (DID) method with conditional Poisson regression and generalized weighted quantile sum regression (gWQS) to investigate the independent and joint effects. Independent exposures to NO₂, CO, and SO₂ were causally associated with increased risks of total, nonaccidental, and cardiovascular mortality, while no evident associations with O₃ were identified in the entire population. In gWQS analyses, an interquartile range-equivalent increase in mixture exposure was associated with a relative risk of 1.067 (95% confidence interval: 1.010-1.126) for total mortality, 1.067 (1.009-1.128) for nonaccidental mortality, and 1.125 (1.060-1.193) for cardiovascular mortality, where NO₂ was identified as the most significant contributor to the overall effect. This nationwide DID analysis provided causal evidence for independent and combined effects of NO₂, CO, SO₂, and O₃ on increased mortality risks among the US general population.