[ExI] new nutrition thread

William Flynn Wallace foozler83 at gmail.com
Fri Aug 28 22:56:56 UTC 2015


Someone criticized correlations in an earlier mailing.  Yes, they don't
show cause and effect, or at least not directly.  But they are not just the
first step in determining cause.  Cause is somewhere in there in
significant correlations, and dozens of studies are not likely to have the
same defects, so some might cancel out.

But to do experiments with people and nutrition?  Hah.

A few things;  it is nearly impossible to do true experiments with people,
especially longitudinal ones,  for a lot of reasons:  great variability, so
many variables, people dropping out, and much more.

One of the most important reasons is that people will lie.  Humans are so
good at it, and highly like to do so in certain areas.  Ask them about what
they eat, when they eat etc. and you are likely to get a lot of false
data.  I doubt that one single person accurately logs everything he eats in
years of the study.  Many people just cannot remember all they ate during
the day, esp. if they 'cheated'. Almost worse than trying to do studies
(never mind experiments) of sexuality.

Given these conditions, doing long term experiments is just impossible.

So we have to live with correlations.  As we all know there are no
experiments in humans showing tobacco causing cancers, emphysema, or
anything else.  It is very likely that that statement is true of every
carcinogen studied (all experiments were done with lab animals, right?).
But the correlations are overwhelming.  Good enough for me.

As The China Study and many others have pointed out, there are severe
problems of bias in food studies because  of the montrous amount of money
involved, and thus bias in studies.

It reminds me of 30 years ago when I did some research on cannabis in
connection with a course I was teaching.  Every single study funded by the
feds showed ill effects of smoking.  Every single independent study found
no significant effects.  Night and day.  The longest study, done in Jamaica
over 20 something years, showed no differences at all even though the
cannabis used there was much stronger than what we had here.  The only
difference, nonsignificant, was that the smokers were 5 pounds lighter.
(Who would have predicted that?).  So it's hard for me to trust government
or industry funded studies.

If we are going to do experiments with the closest animal to people re
digestive tract, we would use pigs.  True.

If we could not base our knowledge of people on anything but experiments,
there would be very little known about us, and that probably not very
generalizable.  You people in the physical sciences don't know just how
good you've got it.

bill w
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20150828/2dc2a49f/attachment.html>


More information about the extropy-chat mailing list