Exhaled carbon monoxide tracks with lower Parkinson's risk in people who never smoked

Design

Prospective population cohort with Cox regression, about 12 years of follow-up

Population

512,701 adults aged 30 to 79 across ten areas of China; 1,131 incident Parkinson's disease cases

Primary outcome

Incident Parkinson's disease by smoking status, and by measured exhaled carbon monoxide among never-smokers

Effect

Regular smoking HR 0.70 (95% CI 0.62 to 0.79); never-smokers with exhaled CO 3.0 to under 5.0 ppm HR 0.62 (0.53 to 0.73), p for trend under 0.001

The China Kadoorie Biobank followed 512,701 adults aged 30 to 79 for about 12 years, recording 1,131 incident cases of Parkinson's disease and 2,949 of other neurodegenerative disease through death, disease registry and insurance linkage. Regular smoking was associated with raised risks of lung cancer, ischaemic heart disease, stroke and death — and with lower Parkinson's disease risk, adjusted HR 0.70 (95% CI 0.62 to 0.79). That inverse association has been reported for decades without a mechanism.

The new part is what happened among people who never smoked. Measured exhaled carbon monoxide, which in never-smokers reflects endogenous production and household exposure rather than tobacco, was associated with lower Parkinson's risk in a broadly graded way: 2.0 to under 3.0 ppm, HR 0.85 (0.73 to 0.99); 3.0 to under 5.0 ppm, 0.62 (0.53 to 0.73); 5.0 to under 11.5 ppm, 0.68 (0.56 to 0.83); 11.5 ppm or more, 0.65 (0.45 to 0.92); p for trend under 0.001. Exhaled carbon monoxide showed no association with the smoking-related diseases or with other neurodegenerative conditions, which is the observation that argues against it being a simple proxy for tobacco.

This is an association in an observational cohort, adjusted for solid fuel use and passive smoke exposure but not immune to reverse causation — prodromal Parkinson's disease alters autonomic and respiratory physiology years before diagnosis. Nothing here justifies a word to a patient about smoking, and nothing here is a test to order. It is a mechanistic lead that supports the carbon monoxide trials already running.

• Do not translate this into any advice about smoking; the same cohort links regular smoking to raised lung cancer, stroke and mortality.
• Exhaled carbon monoxide is not a Parkinson's risk test and has no validated threshold for that purpose.
• Solid fuel use and passive smoke exposure were adjusted for — relevant where indoor biomass cooking is common.
• Expect patients who have read the smoking headline to ask; the honest answer is that the protective factor may be a gas, not the habit.
• Watch the carbon monoxide interventional trials rather than the epidemiology for anything actionable.

The statistics, in plain English

A hazard ratio of 0.62 with an interval of 0.53 to 0.73 excludes 1.0 comfortably, so chance is an unlikely explanation of that band — but an observational design cannot separate the exposure from whatever else travels with it. The strongest internal check here is the specificity of the finding: if raised exhaled carbon monoxide in never-smokers were simply marking hidden tobacco exposure, it should also have tracked lung cancer and ischaemic heart disease, and it did not. The graded dose-response and the null result for other neurodegenerative diseases point the same way. That still falls well short of showing that carbon monoxide protects anyone.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.