It has been shown that people who smoke are less likely to develop Parkinson's disease; the evidence is fairly convincing, although this "protective" influence of smoking is not strong. Several small studies have also been done that suggest a similar effect of smoking on Alzheimer's disease. These were retrospective studies, i.e. information was gathered by looking back into the past - in this case people with Alzheimer's disease had their previous smoking habits compared with those of people without Alzheimer's.

In order to get a better estimate of any protection offered by smoking, researchers in the UK studied the smoking habits of male physicians for over 40 years have looked at the occurrence of Alzheimer's disease and other forms of dementia. This is what is called a prospective study, and is far more reliable than retrospective studies.

In 1951 over 34,000 male physicians on the British medical register who were living in the UK gave information on their smoking habits. Changes in these habits were recorded at 6 - 11 year intervals up to 1998. By that time, 24,000 of the physicians had died. Dementia (Alzheimer's or other forms) was mentioned in the death certificates of 473 of those who died in the last 35 years. (Deaths earlier than this time were not analyzed, as they were too close to the first record of smoking habits, i.e. smoking might not have had time to exert an effect.)

The occurrence of dementias in those physicians who were smokers was compared with that in physicians who had never smoked or were ex-smokers. (The ex-smokers who died with dementia had stopped smoking on average 34 years earlier, so that it is most unlikely that earlier smoking played any role in their deaths.)

The two main types of dementia recorded were Alzheimer's disease and vascular dementia. This was based on the diagnoses on the death certificates, in some cases backed up by autopsy reports.

Each subject who died due to dementia was matched with four non-dementia physicians born in the same year, and each subject who died where dementia was mentioned as an associated condition, and not the cause, was matched with four physicians having the same year of birth who died in the same year as the subject. (This was done to make allowance for the well-known fatal effects of smoking unconnected with dementia).
Analyses of the information obtained showed that the mean age at death was 81 years in the 473 physicians who died with dementia; only 2% of them died before age 65. The likelihood of any type of dementia occurring in a smoker was the same as in non-smokers and ex-smokers (in fact, the likelihood was 0.96 times that in non-smokers, a trivial, non-significant difference).
The presence of Alzheimer's disease was also unrelated to smoking, the likelihood being 0.99 (or virtually the same) in smokers, compared with non- and ex-smokers; the average age at death of the Alzheimer subjects was 82 years. For the other types of dementia (vascular dementia, Lewy-body dementia etc) there was also no evidence of a different risk between smokers and non- and ex-smokers.

This prospective study provides different results from those of the smaller earlier retrospective studies, showing conclusively that prolonged smoking does not increase or decrease the likelihood that men will develop dementia of any type, including Alzheimer's disease. The study was large, prolonged, and follow-up was just about complete, with extremely careful checks and balances to exclude any factors that might inflence the results unfairly.

We can draw two important conclusions. First, we should not attach too much importance to retrospective analyses, provided we can afford the time to wait for a well-designed prospective study to be completed. Second, these results should remove one more excuse that smokers have used to continue their habit. The slight protective effect that smoking appears to have on Parkinson's disease (and possibly also on ulcerative colitis) is clearly outweighed by the serious adverse effects it has on other organs.