Smoking, Exhaled Carbon Monoxide, and Risk of Parkinson Disease

Abstract

Importance Smoking has been associated with lower risks of Parkinson disease (PD) in epidemiological studies, but the underlying mechanisms remain poorly elucidated. Objective To investigate the association of exhaled carbon monoxide (CO) with risk of PD. Design, Setting, and Participants This nationwide prospective cohort study leveraged data from the China Kadoorie Biobank across 10 geographically diverse areas in China. A total of 512 724 adults aged 30 to 79 years were recruited between June 2004 and July 2008. Prospective analyses were restricted to first events occurring between ages 40 and 94 years and before January 1, 2019. Data were analyzed from May 2025 to March 2026. Exposure Self-reported smoking status and measured exhaled CO. Main Outcomes and Measures During approximately 12 years’ follow-up, 1131 individuals with PD and 2949 with other neurodegenerative diseases were recorded through linkage to national death and disease registries and the health insurance system. Cox regression was used to estimate adjusted hazard ratios (HRs) for associations of smoking and, among individuals who never smoked, exhaled CO levels with PD, other neurodegenerative diseases, and several known smoking-related diseases. Covariates included sociodemographic characteristics, lifestyle factors, and other confounders (ie, solid fuel use and passive exposure to smoking). Results Among the 512 701 participants included in the main analyses, the mean (SD) age was 52 (11) years and 302 041 (58.9%) were female. Overall, 156 650 male individuals (74.5%) and 9944 female individuals (3.3%) ever smoked regularly, with mean exhaled CO levels higher among those who regularly smoked (11.1 ppm) than those who never smoked (3.5 ppm), those who occasionally smoked (3.8 ppm), and those who formerly smoked regularly (3.7 ppm). Regular smoking was significantly associated with elevated risks of lung cancer, ischemic heart disease, stroke, and all-cause mortality but with lower risk of PD (adjusted HR, 0.70; 95% CI, 0.62-0.79). In individuals who never smoked, higher exhaled CO levels were associated with lower HRs of PD (n = 675) in a broadly dose-dependent manner (exhaled CO <2.0 ppm: 1.00 [95% CI, 0.85-1.18]; 2.0 to <3.0: 0.85 [95% CI, 0.73-0.99]; 3.0 to <5.0: 0.62 [95% CI, 0.53-0.73]; 5.0 to <11.5: 0.68 [95% CI, 0.56-0.83]; ≥11.5: 0.65 [95% CI, 0.45-0.92]; P for trend < .001) but were not associated with risks of smoking-related diseases or other neurodegenerative diseases. The associations with PD were more pronounced in female individuals who never smoked. Similar exhaled CO-PD associations were found in those who regularly smoked. Conclusion and Relevance Among individuals who never smoked, higher exhaled CO levels were associated with lower PD risk. These findings demonstrate a potentially protective role of CO in PD etiology, which help explain the association between smoking and PD observed in this and other epidemiological studies and support ongoing clinical trials of CO in the treatment of PD.