Wouldn't taking these kind of factors into account be quite basic part of the analysis?

They acknowledge that they are just looking at the number of prescriptions and that the research doesn't show that the drugs themselves were reducing people's risk.

It could equally be the amount of sex that is reducing the risk.