Bob in CT wrote in message ...
On 23 Sep 2004 06:51:40 -0700, Chet Hayes wrote:

Kevin Stevens wrote in message
...
In article ,
(Roman Bystrianyk) wrote:

In a study in today's Journal of the American Medical Association,
mortality rates were 65% lower among elderly people who combined a
so-called Mediterranean diet with 30 minutes of daily exercise,
moderate drinking and no tobacco use.

What a useless freaking study! How much lower was the mortality rate
among elderly people who combined ANY diet plan or WOE with 30 minutes
of daily exercise, moderate drinking, and no tobacco use?!

Don't like your initial results? Keep adding factor elements until you
see a number you like. Ridiculous.

KeS


I wouldn't blast the study based on short excerpts from news
organizations. The news usually goes for the simple, easy, overall
message. If you look at the actual study, it was done to determine the
effects of the diet, excercise, moderate drinking, no smoking, both
together and seperately. It appears to be well designed and covered a
10 year period. There were benefits to all components, the combined
effect was just the best result.

But without data like true mortality (not the BS "relational" mortality),
the study is useless.

And did you read the actual study to see what data was recorded and
reported before coming to the conclusion that the study was useless?

It's like the study that gave two drugs to two
different groups of people. The average LDL level dropped farther with
one drug, and the relative number of deaths due to heart disease also
dropped farther with that drug. The authors said that this "proved" that
lowering LDL was beneficial, when that's not what the study proved at
all.

And what does drawing incorrect inferences have to do with this? In
this example, you're right, the data doesn't show that the lowering of
LDL was the mechanism, only that the drug reduced the incidence of
heart disease. That would certainly be jumping to conclusions. What
evidence do you have that this was done with the current study?





(What it indicated was that if you took one drug and not another
your relative risk of heart disease was lower.) Without access to the
real data, none of us know what the results of this study are.


OK, but without access how do you know it's useless based on a news
report excerpt?