Smoking and Parkinson's disease--a new perspective.

This paper examines fresh the evidence for an inverse relationship between smoking behaviour and Parkinson's disease. The frequent epidemiological observation, that the relative risk of a smoker having Parkinson's disease is approximately 0.5 that of a non-smoker, is fairly well established. Possible explanations are examined in the light of available epidemiological and biochemical evidence. Despite much biochemical work suggesting a protective role for cigarette smoke against the development of Parkinson's disease, the epidemiological data points away from such a mechanism. Instead, the epidemiological work favours two alternative hypotheses: either that there is selective mortality of smokers with Parkinson's disease such that their conditions deteriorate and death occurs at an accelerated rate (resulting in fewer smokers with Parkinson's disease-especially in the older age groups when Parkinson's disease is most common), or that the latent phase of Parkinson's disease (before symptoms become apparent) involves subtle personality changes producing the so-called Parkinsonian personality and an associated aversion to cigarette smoking. The latter changes may be a consequence of progressive nigrostriatal dopamine depletion resulting eventually in Parkinson's disease. Future approaches that would allow a more definitive understanding of the processes underlying the intriguing link between smoking and Parkinson's disease are considered.