Aaron:

The thought "I may have to talk to some of these people at AHS 2012" crossed my mind.  I've endeavored to keep my discussion civil and science-based.

Yes, I'm absolutely willing to participate in such a panel discussion -- and I'd love to take it to the next level, to wit: "Apparently we've known for over twenty years (since Salvini et.al.) that the food consumption data from the Nurses' Health Study ranges from "wildly inaccurate" to "complete bunk".  Why are we still basing research on it -- and, even worse, why are we basing public policy recommendations on it?"

Zak:

I could make a similar argument about my grandfather: smoked two packs a day for over 50 years, died of a non-smoking-related illness in his 80s.  Smoking is good for you!

Ed:

Roger Williams was an eminent scientist (he discovered pantothenic acid -- vitamin B5 -- among other things), but I'm unable to find any of his nutrition writing online.

I don't think variety for its own sake is a good idea: "a little bit of everything" is only a good strategy when you don't know what's good and what's bad.  I think it's best to eat more of the good stuff and less of the bad stuff!  But since I have no way to know what Williams' actual words or intent are, I won't press the issue.

And yes, cholesterol is a vitamin, not a poison: it's absolutely necessary to life.

Uncephalized:

Good catch.  I added it back in.

peter aris:

Which comment do you mean?

Tom:

Most "health food" is indeed bad for us, but I don't believe the converse is true: everyone agrees that Twinkies and Mountain Dew are unhealthy, but the Twinkie/Dew Diet won't leave you in good health.

Also note that "health food" is indeed better than the SAD (Standard American Diet).  My favorite analogy is low-tar cigarettes: they're indeed less harmful than regular cigarettes...but you're still smoking cigarettes.

BT:

Most people aren't interested in hearing anything that contradicts their existing viewpoint...paleo eaters included.

However, I'd be cautious about taking your secret to the grave...we could easily end up like Denmark, with a tax on saturated fat, if we don't continue to represent ourselves proudly.  People are less likely to push for something like that if they believe that there's a reasonable alternative viewpoint, even if they don't subscribe to it themselves.

Oliver:

Thank you.  It means a lot to have my work approved by a professional.

eddie:

I've been hoping for an excuse to post an xkcd cartoon for some time.

Thanks for sharing it on Facebook!  ("Likes" are nice, but "shares" are what get the word out to people who aren't already fans of mine.)  

Beowulf:

Exactly!  "Vegetarian" has become synonymous with "healthy eating", due to tireless vegetarian propaganda...we need to change that.

Wayne:

I'm not confident enough in my knowledge of statistics to teach the multivariate analysis class: perhaps Oliver can step in?

George:

I appreciate the support!  Don't lose hope: judging from the number of page views this article is getting, I'd say that there are quite a few people questioning the science.  

No, I'm not up there with the Daily Telegraph or the LA Times -- but there are a lot more of us than you might think.

JS

Roger:

Industrial agriculture is destructive and unsustainable whether the end product is fed to people or to cattle.  I've written about this topic at length.

In contrast, grass-fed ruminants have negative carbon impact if properly managed -- something that cannot be said for any form of agriculture.  (It must be so: otherwise the Serengeti and the Great Plains would have blown away to dust millions of years ago from the impact of all those wildebeest and bison, and the Pleistocene wouldn't have been a continual global cooling trend, complete with Ice Ages.)

Also, consider that most of the figures for carbon impact of even the worst industrial feedlot beef are sheer baloney -- they assume that cattle are grain-fed their entire life, which is impossible.  (They die if you feed them on grain for more than a few months.)

JS

Ric:

It might be impossible to "prove" cause and effect, but RCTs can get us to a reasonable functional facsimile...

...and those who decide to test whether the train will "really" run them over generally get run over.

JS

Susan:

Personally, I blame M. Night Shyamalan.

JS

SP:

These articles seem to come out every year or two: yet another mining of the same known-bad data sets to produce another spurious association between reported meat consumption and something negative.

I think the latest one purports to show that meat-eaters beat their wives more often (not kidding).  Of course, the Third Factor in play is that vegetarians are many times more likely to be college-educated and of higher socioeconomic strata, and therefore much less likely to engage in domestic violence.

JS