Smoking and dementia in male British doctors: prospective study

Objective To assess the possible association between smoking and dementia. Design Prospective study Setting Cohort of British male doctors followed up since 1951. Subjects 34 439 male British doctors, with 24 133 deaths recorded. Results For all types of dementia combined the relative risk was 0.96 (95% confidence interval 0.78 to 1.18), based on 473 deaths at a mean age of 81 years. For probable or definite Alzheimer's disease, the relative risk in c ontinuing smokers was 0.99 (0.78 to 1.25), based on 370 deaths at a mean ag e of 82 years. In aggregate, however, the other prospective studies indicat e a direct, although not clearly significant, association between smoking a nd the onset of dementia in general and of Alzheimer's disease in particula r. Conclusions Contrary to previous suggestions persistent smoking does not su bstantially reduce the age specific onset rate of Alzheimer's disease or of dementia in general. If anything, it might increase rather than decrease t he rate, but any net effect on severe dementia cannot be large in either di rection.