John Tierney has the latest:
What we have to keep in mind here is that nutrition is a science (or at least should be) and science is about generating hypotheses, making predictions from our hypotheses, and then seeing if they hold true. The relevant hypothesis here — i.e., what we’ve believed for the past 30-odd years — is that saturated fat causes heart disease by elevating either total cholesterol or LDL cholesterol, specifically. So our prediction is that the diet with the higher saturated fat content will have a relatively deleterious effect on cholesterol. We do the test; we repeat it a half dozen times in different populations. Each time it fails to confirm our prediction. So maybe the hypothesis is wrong. That seems like a reasonable conclusion. No one is proving anything here — as some of your respondents like to decry — we’re just looking at the evidence and trying to decide which hypotheses it supports and which it tends to refute.
…These latest trials just happen to be the best data we have on the long-term effects of saturated fat in the diet, and the best data we have says that more saturated fat is better than less. It may be true that if we lowered saturated fat further — say to 7 % of all calories as the American Heart Association is now recommending — or total fat down to 10 percent, as Dean Ornish argues, or raised saturated fat to 20 percent of calories, as Keys did, that we’d see a different result, but that’s just another hypothesis. The trials haven’t been done to test it. It’s also hard to imagine why a small decrease in saturated fat would be deleterious, but a larger decrease would be beneficial.
I think that what the nutrition industry and the FDA have done over the past decades with their pseudoscience war on dietary fat borders on the criminal. I’m pretty much convinced at this point that the biggest culprit in both our health and weight is starch and refined sugars, and that the FDA “food pyramid” has been, and remains (despite recent improvements) quackery, not science.