Smoking Study Adds Fuel to Parkinson's Debate

Male British physicians who smoked tobacco in 1951 had a 30% lower risk of death from Parkinson's disease, an analysis of data from the British Doctors Study showed.

Moreover, doctors who continued to smoke over the years had a 40% lower risk of Parkinson's mortality, reported Robert Clarke, MD, of the University of Oxford in England, and co-authors.

As shown in their study online in Neurology, Parkinson's mortality risk was inversely associated with the amount of tobacco smoked, and for those who quit smoking, the effect was attenuated with increasing time since quitting.

The findings emerged from 65 years of follow-up with nearly 30,000 male physicians in Great Britain and should not be interpreted to promote smoking, but rather to foster a closer look at why smoking continues to be linked to lower Parkinson's risk, Clarke said.

"Current smoking is the leading cause of premature death and disability worldwide, and any such hazards would greatly exceed any beneficial effects of current smoking on risk of Parkinson's disease," Clarke wrote in an email to MedPage Today.

The report "demonstrates a causally protective effect of current smoking on risk of Parkinson's disease," he continued. While the underlying mechanism is unclear, "the most likely explanation is that the nicotine content in tobacco smoke may have protective properties, possibly by stimulating the release of dopamine, but the effects of other components of tobacco cannot be excluded," he added.

Researchers first proposed that smoking may reduce Parkinson's risk 60 years ago, noted Yuan Cheng, MD, and Yan-Jiang Wang, MD, PhD, both of Daping Hospital of Third Military Medical University in Chongqing, China, in an accompanying editorial.

"Since then, mounting evidence has suggested that tobacco smoking, which is a major cause of premature deaths and a risk factor for almost all other noncommunicable diseases, is associated with a reduced risk of sporadic Parkinson's disease," Cheng and Wang wrote. "However, the causal relationship between tobacco smoking and the risk of Parkinson's disease remains uncertain due to the inherent limitations of retrospective studies and insufficient case numbers or inadequate follow-up duration in the previous prospective studies."

Clarke and colleagues evaluated data on 29,737 male British physicians from the British Doctors Study, excluding those with a pre-existing diagnosis of Parkinson's disease. An initial questionnaire was mailed to registered doctors in 1951, and six resurveys were sent between 1958 and 1998, with response rates varying between 94% and 98%. Cause-specific mortality was followed until December 2016.

The researchers excluded the first 10 years of follow-up from analyses to minimize reverse causality bias. Doctors who withdrew before the end of the study or were lost to follow-up were included until absent, and then censored.

With each resurvey, smoking prevalence declined among doctors of all ages. Current smoking fell from 67% in 1951 to 8% in 1998 among physicians 65 to 69 years old, for example.

Excluding the first 10 years of follow-up, 25,379 deaths were reported in the study. Of these, 283 (1.1%) had Parkinson's disease listed as their underlying cause of death. Doctors who died of Parkinson's were followed for 42 years on average and died at a mean age of 82. Doctors who died of other causes were followed for 35 years on average and died at a mean age of 77.

Physicians who smoked at baseline showed a 30% lower risk of death from Parkinson's disease (RR 0.71, 95% CI 0.60-0.84), compared with physicians who never smoked. Continuing smokers had a 40% lower risk (RR 0.60, 95% CI 0.46-0.77) than never-smokers.

Long follow-up, large numbers, and high response rates were strengths in this study, the editorialists observed, noting that the analysis was based only on 283 people with Parkinson's disease as the underlying cause of death and the incidence of Parkinson's overall was likely underestimated. Criteria and accuracy of diagnosing Parkinson's disease also changed over time, which may have confounded the results, Cheng and Wang added.

"Despite these limitations, the findings of this study support the protective effects of tobacco smoking on Parkinson's disease risk," the editorialists wrote. "The results do not, however, provide a mechanism by which this effect occurs, so it is premature to assume that there is a direct causative relationship between tobacco use and a reduced risk of Parkinson's disease."

Comment