A biobank study asks whether carbon monoxide explains smoking's odd link to Parkinson's
In half a million Chinese adults, higher exhaled carbon monoxide tracked with lower Parkinson's risk even among people who never smoked. It is an association, not a therapy.
| Group | Value (value) |
|---|---|
| <2.0 ppm | 1 |
| 2.0–<3.0 ppm | 0.85 |
| 3.0–<5.0 ppm | 0.62 |
| 5.0–<11.5 ppm | 0.68 |
| ≥11.5 ppm | 0.65 |
For decades, epidemiology has carried an uncomfortable finding: people who smoke are less likely to develop Parkinson's disease. A large study in JAMA Neurology proposes a candidate for what, inside tobacco smoke, might account for it — carbon monoxide — and reports that higher exhaled carbon monoxide tracked with lower Parkinson's risk even in people who never smoked [s1]. This is an association drawn from observational data, not evidence that carbon monoxide prevents or treats anything, and the paper does not claim otherwise [s1].
The lower Parkinson's rate among smokers is one of the most replicated associations in the field, and also one of the most misused — it is not a reason to smoke, because smoking's harms dwarf any hypothetical benefit. The open question has been mechanism: is the association real biology, or an artefact of who smokes and who is later diagnosed? Isolating a single component of smoke and testing it in non-smokers is one way to probe that.
The cohort
The study used the China Kadoorie Biobank, which recruited 512,724 adults aged 30 to 79 across 10 areas of China between June 2004 and July 2008 [s1]. In the main analyses of 512,701 participants, the mean age was 52 and 302,041 (58.9%) were women [s1]. Smoking was heavily sex-patterned: 156,650 men (74.5%) but only 9,944 women (3.3%) had ever smoked regularly [s1]. Exhaled carbon monoxide, measured at recruitment, was higher in regular smokers (mean 11.1 ppm) than in those who never smoked (3.5 ppm), smoked occasionally (3.8 ppm) or formerly smoked regularly (3.7 ppm) [s1]. Over roughly 12 years of follow-up, 1,131 people were recorded with Parkinson's disease and 2,949 with other neurodegenerative diseases [s1].
What it found
Regular smoking carried the expected penalties — higher risks of lung cancer, ischaemic heart disease, stroke and death — but a lower risk of Parkinson's (adjusted hazard ratio 0.70; 95% CI, 0.62 to 0.79) [s1]. The novel part is what happened among people who never smoked. There, higher exhaled carbon monoxide was associated with lower Parkinson's risk (675 cases) in a broadly dose-dependent way: against the lowest band, hazard ratios fell to 0.85, 0.62, 0.68 and 0.65 across rising carbon-monoxide categories (P for trend <0.001) [s1]. The same carbon-monoxide gradient was not associated with smoking-related diseases or with other neurodegenerative diseases, and it appeared again within regular smokers [s1]. The association was more pronounced in women who never smoked [s1].
How much weight it bears
The internal logic is the study's strength: if exhaled carbon monoxide were merely a proxy for smoking, it should also predict lung cancer and heart disease in this data, and it did not [s1]. That specificity to Parkinson's is what makes the carbon-monoxide hypothesis worth stating.
But specificity is not proof. In never-smokers, exhaled carbon monoxide comes from ordinary sources — the body's own metabolism, household air, background exposure — and those sources may travel with other things that affect Parkinson's risk. Reverse causation is a live concern too: Parkinson's has a long silent phase, and its earliest changes can alter behaviour and physiology years before diagnosis, which could nudge a measured biomarker. An association in a single cohort, however large, does not establish that carbon monoxide protects dopamine neurons. Other environmental lines of Parkinson's research, such as green space and where people live, sit in the same evidential category — suggestive, not settled.
What the study does earn is a mechanism worth chasing in the laboratory, where the causal question can actually be asked. Whether carbon monoxide, at the low levels the body handles safely, does anything to the biology of Parkinson's is precisely what an observational study cannot tell you [s1].
Sources
- [s1] Smoking, Exhaled Carbon Monoxide, and Risk of Parkinson Disease, JAMA Neurology, 8 September 2026.
Sources
- Smoking, Exhaled Carbon Monoxide, and Risk of Parkinson Disease — JAMA Neurology , September 8, 2026
More on
Greenness near home tracked with less Parkinson's — and more Alzheimer's
A 220,000-person cohort found opposite associations for two neurodegenerative diseases depending on the type and distance of greenery. Two other February papers ask who can actually use it.
An AI read one night of sleep and forecast 130 diseases. Read the fine print.
Stanford's SleepFM, published in Nature Medicine in January 2026, is the most consequential sleep paper of the year. Its own limitations section explains why it is not yet a clinical test.
A hand squeeze and five chair stands predicted nine long-term conditions
Pooling 94 cohort studies, grip strength tracked with later dementia, depression and disability. The authors rate their own confidence in the numbers as limited to moderate.
9/11 and cancer: what 25 years of surveillance actually shows
Registries show a real but narrow signal: a few cancers, led by prostate and thyroid, run high in 9/11 responders, while overall cancer incidence sits at or below the rate expected.