Epidemiological studies can often throw out intriguing and sometimes erroneous correlations between a variable and a disease, but UK scientists say a finding that Alzheimer’s disease seems to be less common in Viagra users warrants further investigation.