Statins could substantially reduce the risk of developing dementia, either by delaying its onset or by decelerating age-related changes that result in the disease, say American researchers. The researchers investigated three groups of patients, all aged between 50 and 89 years, from the UK-based General Practice Research Database - those taking any lipid-lowering drug (n= 24,480), those with hyperlipidaemia who were taking no drugs (n=11,421) and a control group (25,000 people without hyperlipidaemia and not taking lipid lowering drugs). They subdivided the first group into those who were taking statins and those prescribed non-statin lipid-lowering drugs. The subsequent incidence of dementia was then assessed over a six-year period. Dr Hershel Jick (Boston Collaborative Surveillance Programme, Lexington) and colleagues estimate that the incidence of dementia in patients who were taking statins was between 37 and 70 per cent lower than in the other groups. They identified 284 patients who developed dementia and matched them with 1,080 controls. The relative risk estimates for developing dementia were as follows: current statin users 0.29; individuals with untreated hyperlipidaemia 0.72; current users of non-statin lipid-lowering agents 0.96. The authors comment that the mechanism by which statins might reduce the risk of developing dementia is not known but it might not be related to reductions in low-density lipoprotein levels. They suggest that the mechanism might involve beneficial effects on the cerebral capillary endothelium "or other properties of the drugs". Dr Jick and colleagues say that the reduction in risk seemed to be similar for all statins (Lancet 2000;356:1627).