Abstract
We examined the effect of annual exposure to fine particulate matter (PM(2.5)), nitrogen dioxide (NO(2)), and ozone (O(3)), on the rate of first hospitalization with a PD-related diagnosis (hospitalization with PD) among Medicare Fee-for-Service beneficiaries (2001-2016). Machine learning-derived annual air pollution concentrations were linked to residential ZIP codes. For each exposure, we fitted four models: 1) traditional outcome stratification, 2) marginal structural, 3) doubly robust, and 4) generalized propensity score matching Poisson regression models, adjusted for sociodemographic and meteorological confounders and long-term trends. Among 49,121,026 beneficiaries, incidence rate ratios of 1.08 (95% CI: 1.07, 1.10), 1.07 (95% CI: 1.05, 1.08), and 1.03 (95% CI: 1.02, 1.05) for an interquartile range increase in PM(2.5) (3.72 µg/m(3)), NO(2) (13.84 ppb), and O(3) (10.09 ppb), respectively, were estimated from doubly robust models. Results were similar across modeling approaches. In this nationwide study, higher air pollution exposure increased the rate of hospitalizations with PD.