Jim Swayze wrote:

> I'll ask you again for the one thousandth time.

I've already answered your question. 

The Bristol researchers have an untested and unproven theory -- a conjecture
-- which they advanced in an effort to explain why their results did not
agree with the results of the majority of studies into the same subject. You
may choose to accept that conjecture as truth, but don't pretend that you
are accepting it based on the results of any scientific test of the theory. 

Firstly, the Bristol researchers presented no evidence to support their
conjecture that non-drinkers are more likely than moderate drinkers to be
sickly people with shorter life-expectancies. In fact the opposite seems
just as likely to be true: it's certainly possible if not likely that sickly
people tend to drink to compensate for the pain and misery of their
illnesses. 

Secondly, even if the researchers could present such evidence, they would
still need to present evidence that the effect they are speculating about
was significant enough to sway the statistics.

In sum: they have no evidence whatsoever to support their conjecture about
why their results differed from those of many other researchers who examined
the same subject. They did not, according to the article, control for the
confounding variable that they now suppose might exist. 

But I understand that a lot of people are determined to believe whatever
they want to believe, evidence be damned.

-gts