After half a century of research on smoking and PD, where do we go now?

Half a century has passed since the initial report of an inverse association between Parkinson disease (PD) and smoking.¹ While counter to the expected health effects of tobacco smoke, this finding has since been confirmed in meta-analyses and pooled analyses comprising data from more than 50 high-quality observational studies^2–4 suggesting that smokers are only half as likely as those who have never smoked (never smokers) to develop PD, with an evident dose response as a function of pack-years.

In this issue of Neurology®, Chen et al.⁵ provide new insight into this association by examining the effects of duration and intensity of smoking independently, rather than combined, as with the pack-years estimation. Their prospective cohort design alleviates most concerns about selection and survival biases that might produce spurious associations but it does not guarantee the absence of residual bias from unmeasured or unknown confounding factors. They observed no association with intensity (daily number of cigarettes) at any duration, suggesting that only the duration component of the pack-years measure previously reported may be responsible for the observed lower PD risk in smokers. They further confirmed prior observations that fewer patients …