> "Zheng says that it's not necessary to stop eating meat. Nor it is necessary
> to gobble up huge quantities of vegetables. A balanced diet, he says, is all
[quoted text clipped - 15 lines]
>
> http://my.webmd.com/content/Article/83/97796.htm
> Yes, and people who eat all those veggies also tend to be wealthier, live in
> neighborhoods without a lot of toxic waste and be more educated about health
> and have better access to medical care.
My grandfather died from Non-Hodgekins Lymphoma. I was given to
understand that the cause was likey radiation exposure as he was one of
the lead engineers building the Richland nuclear power plant for WUPPS.
In the course of working for Bonneville, he was also exposed to
whatever is causing the cancer cluster in and around The Dalles OR.
Annecdotes are *definitely* as reliable as spurrious correlations from
dietary studies, but he was middle class and of the first group to be
placed on Low Fat diet to control cholesterol. I remember the shift in
their eating when I was a kid in the 70's. My grandparents had used to
eat eggs for breakfast every morning and suddenly they were eating weird
chemo-eggs (egg beaters), lots of cereal, lean meat and white bread.
> But nutritional studies never factor that stuff out.
Thank you for saying that. One of the biggest difficulties in
determining causation is factoring out every other environmental factor
that may contribute to phenomenon you are trying to study. This is why
you almost never see a reputable study that states that X is *caused* by
Y. What you are more likely to see is a numerical value for the
correlation between X and Y - and as we have already established,
correlation *does not equal* causation.
The best information is usually going to come from long term
longitudinal reasearch studies, nurses studies are among the best of
those as nurses are relatively easy to keep track of over time.
Well that was more detail than I meant to get into,
brigid
Ignoramus20562 - 25 Mar 2004 19:09 GMT
>> But nutritional studies never factor that stuff out.
>
[quoted text clipped - 5 lines]
> correlation between X and Y - and as we have already established,
> correlation *does not equal* causation.
Good studies are based on randomization and control, not on anecdotes
or epidemiology (correlation).
i
> The best information is usually going to come from long term
> longitudinal reasearch studies, nurses studies are among the best of
> those as nurses are relatively easy to keep track of over time.
>
> Well that was more detail than I meant to get into,
> brigid
Saffire - 25 Mar 2004 22:27 GMT
> Thank you for saying that. One of the biggest difficulties in
> determining causation is factoring out every other environmental factor
[quoted text clipped - 3 lines]
> correlation between X and Y - and as we have already established,
> correlation *does not equal* causation.
There was a professor who was doing experiments with a frog. He was
teaching a frog to jump. The training went on for a while and finally
when he said "Jump!" the frog would jump high in the air. He thought
it was time to take some measurements and publish the results.
He started his measurements with a twine, a ruler and a knife. He
placed the frog on a wooden cutting plate and said "jump". It jumped
and he measured the height it jumped. He wrote in his observation
note book: "Height jumped (with 4 legs): 14 inches. Inference: None".
Then he cut one of the legs of the frog and said jump. It jumped to
a height of 10 inches. Inference: None. Then he cut the next leg, and
measured the height jumped. Because it had only 2 legs the height
jumped was only 5 inches. The he cut one more leg and the frog now
had only one leg. The height jumped was just 1 inch with one leg.
Again the inference was none. Then he cut the last leg of the frog
and said "jump!". It didn't move at all. He wrote his inference in
the note book: "When all four legs are removed, frogs go deaf."

Signature
Saffire
205/163/125 - 5'2.5"
Atkins since 6/14/03
Progress photo: http://photos.yahoo.com/saffire333