I enjoyed her article a lot. It was a good reminder that correlation does not mean causation.

The bottom line for me is that there's no reason to assume animal products are anything but healthy. There's also no reason to assume we have to eat tons of animal products every day. As always, a wide variety of whole, natural foods seems the best choice.

Copy and paste from comments section. Note: I did retract "I'm not going to read this crap" and I did read it, but the rest of it stands. Also a disclaimer that I do know that her conclusion isn't that he is right, just clearing up a misunderstanding.

Denise, you are funny and smart and I like you, but I’m not going to read this crap because it is completely meaningless. Saturated fat DOES contribute to cardiovascular disease, but only in certain contexts. Unless we have data that controls for context we are seriously at risk of false positives. Even if we have extremely good data showing that there is basically no way that an association is explained by anything other than causation, we still don’t have -necessary- causation. There’s the obvious point, that saturated fat is associated with affluence which is associated with processed crap and stress, possibly pollution at the time. But then there is also the very important biochemistry. Saturated fat does impair endothelial function, but only in the context of an omega-3 deficiency http://www.nrcresearchpress.com/doi/abs/10.1139/H10-020?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%3dpubmed

The informed and intelligent conclusion from clinical nutritionist Byron Richards is that DHA helps saturated fat function properly in the body http://www.wellnessresources.com/health/articles/dha_helps_saturated_fat_function_properly_in_your_body/

If you want to talk red meat we can do that, if you want to talk alcohol or sugar, we can do that too, if you want to talk any number of things that people think is unhealthy, we can do that too, and we’re always going to find mitigating factors. Epidemiology is the conversation starter.

The concept of the mitigating factor has set epidemiology back another 50 years in my mind. Until they identify, quantify, and incorporate mitigating factors into epidemiology, it isn’t worth much to me other than as a place to start looking.

And I'll just add to that now that if we don't want to waste our time we shouldn't be asking whether or not Keys' study supported his conclusion or not, but whether or not his methods should be given the credence that they have. Looking back on that era it is hard to say that he was justified in his conclusion, even in that context where all of the real science was still young, but now it just seems absurd. It's not Denise's fault that she gets sucked into this stuff, it's the entire zeitgeist's fault for wanting her to.

Here's what I got out of the article

-Paleos accuse Key's of cherry-picking

-Even without the cherry-picking, his data stil supports his conclusion

-Here are a list of problems with that conclusion

-So there are problems, but there is also a point to the evidence. Life's not so simple.

Which is how I feel. Life's not so simple, saturated fat can probably range from harmless and even healthy, to unhealthy. We simply don't know yet, and it would be hubris to recommend slathering everything with butter just because you can. Read that very carefully: no reason to be afraid of butter/saturated fats, but that doesn't mean you should go too far in the other direction.

If I had to guess how this will all turn out, I would probably guess that saturated fats will be shown to have deleterious health effects, sometimes. I think that this certainly won't hold true in every situations. For example, saturated fats don't change HDL/LDL in trained cyclists (there's a study) as long as they're in energy balance. So I expect, leanness, activity, energy balance (losing weight is heart-protective), amount consumed (poison's in the dose), and genetics to all play a role in determining the health of saturated fats.

To whom it may concern: As I write this, I'm drinking whole milk.

My takeaway:

It frightens me that such pervasive government policy & medical mindset is based on such flawed studies/research.

No, she's not saying Keys is right. The first part of the article may seem confusing because she brings up anti-Paleo arguments of why Keys was NOT wrong with his cherry-picking because there's still a correlation when using all 22 countries. The rest of the article proceeds to provide counter-arguments for this. So, it's like you posited: she's only saying his data is unreliable but not for the reasons we have previously believed. She made a big point about using food/fat that was available rather than what was actually consumed (which tends to be overestimated, as she puts it, in developed countries). She also points out faulty medical filing; if someone died of heart disease but it was filed under non-cardiac related deaths, that would appear to make it look like that country has an overall low rate of deaths from heart disease.
But then again, she is a raw-foodist, so maybe I don't really know what point she's trying to make....
Still a nicely written article.

From what I understand, it's impossible to draw any conclusions from the data because there are too many cofounding factors, such as level of advancement/technology, diagnostic procedures etc etc (ie: countries that have a lot of meat available tend also to be more advanced societies, more economically stable, have better methods of diagnosis and so on and so forth). Also, the data showed how much of each type of food was available for consumption - not how much was actually consumed by actual people.

Also, correlation was there for heart related deaths, but not for non-heart related deaths... basically, you can't draw any conclusions from the data at all.

Gotta love Denise! She is so GOOD at this stuff.