another problem with it
is that the absolute non-drinkers tend to have a higher population of people who are not drinking *for a reason* (not just taking statins, but there's a lot of people who used to be alcoholics and are teetotal now).
So the non-drinker group would be slightly more likely to have historic health issues than the people who drink small amounts, for example.
So it might look like having 1 or 2 drinks is healthier (in fact it might be the case), but the data just doesn't reflect that.
(
admiral crunch holds the newsreader's nose squarely, Wed 18 May 2016, 14:41,
Reply)
Oddly enough, the statisticians have thought about that...
There's a concise overview, with links to references, here:
velvetgloveironfist.blogspot.co.uk/2016/01/doubt-is-their-product.html
(, Wed 18 May 2016, 15:38, Reply)