High-dose carbon monoxide exposure, as in acute poisoning, is neurotoxic and can cause delayed parkinsonism. It damages the basal ganglia (especially the globus pallidus) and white-matter tracts through hypoxia, mitochondrial inhibition, and oxidative stress. Cohort studies, such as Lai et al. (2015), report substantially elevated later risk of Parkinson’s disease after CO intoxication (roughly 9- to 15-fold in some analyses).
In contrast, low, chronic CO levels – comparable to those in cigarette smokers – are associated with lower Parkinson’s risk and appear neuroprotective in experimental models. Low-dose CO reduces dopaminergic neuron loss, preserves striatal dopamine, and decreases α-synuclein pathology, in part by activating haeme oxygenase-1 (HO-1) and related stress-response pathways that limit oxidative damage and promote protein clearance (Rose et al., 2024).
The most recent large human evidence comes from Bueno Lopez et al. (JAMA Neurology, 2026). In the China Kadoorie Biobank of more than 512 000 adults followed for about 12 years, regular smoking was linked to lower Parkinson’s risk (adjusted HR 0.70). Among never-smokers, higher exhaled CO showed a broadly dose-dependent inverse association with Parkinson’s risk (strongest effects around 3–11.5 ppm), without elevating risks of smoking-related diseases or other neurodegenerative conditions. These findings support CO as a plausible contributor to the long-observed smoking–Parkinson’s paradox and bolster interest in controlled low-dose CO as a research target, while underscoring that smoking itself remains harmful.
