Dietary Guidelines for Americans: We don’t need no stinkin’ science

I know, I know. I never post. I never call. I don’t bring you flowers. It’s a wonder we’re still together. I have the usual list of excuses:


But before I disappear off the face of the interwebz once again, I thought I share with you a quickie post on the science behind our current Dietary Guidelines. Even as we speak, the USDA and DHHS are busy working on the creation of the new 2015 Dietary Guidelines for Americans, which are shaping up to be the radically conservative documents we count on them to be.

For just this purpose, the USDA has set up a very large and impressive database called the Nutrition Evidence Libbary (NEL), where it conducts “systematic reviews to inform Federal nutrition policy and programs.” NEL staff collaborate with stakeholders and leading scientists using state-of-the-art methodology to objectively review, evaluate, and synthesize research to answer important diet-related questions in a manner that allows them to reach a conclusion that they’ve previous determined is the one they want.

It’s a handy skill to master. Here’s how it’s done.

The NEL question:

What is the effect of saturated fat intake on increased risk of cardiovascular disease or type 2 diabetes?

In the NEL, they break the evidence up into “cardiovascular” and “diabetes” so I’ll do the same, which means we are really asking: What is the effect of saturated fat (SFA) intake on increased risk of cardiovascular disease?

Spoiler alert–here’s the answer: “Strong evidence” indicates that we should reduce our intake of saturated fat (from whole foods like eggs, meat, whole milk, and butter) in order to reduce risk of heart disease. As Gomer Pyle would say, “SUR-PRIZE, SUR-PRIZE.”

Aaaaaaaand . . . here’s the evidence:

The 8 studies rated “positive quality” are in blue; the 4 “neutral quality” studies are in gray. The NEL ranks the studies as positive and neutral (less than positive?), but treats them all the same in the review. Fine. Whateverz.

According the exclusion criteria for this question, any study with a dropout rate of more than 20% should be eliminated from the review. These 4 studies have dropout rates of more than 20%. They should have been excluded. They weren’t, so we’ll exclude them now.

Also, according to NEL exclusion criteria for this question, any studies that substituted fat with carbohydrate or protein, instead of comparing types of fat, should be excluded. Furtado et al 2008 does not address the question of varying levels of saturated fat in the diet. In fact, saturated fat levels were held constant–at 6% of calories–for each experimental diet group. So, let’s just exclude this study too.

One study–Azadbakht et al 2007–was conducted on teenage subjects with hypercholesterolemia, a hereditary condition that affects about 1% of the population. Since the U.S. Dietary Guidelines are not meant to treat medical conditions and are meant for the entire population, this study should not have been included in the analysis. So let’s take care of that for those NEL folks.


In one study–Buonacorso et al 2007–total cholesterol levels did not change when dietary saturated fat was increased: “Plasma TC [total cholesterol] and triacylglycerol levels were NS [not significantly] changed by the diets, by time (basal vs. final test), or period (fasting vs. post-prandial) according to repeated-measures analysis.” This directly contradicts the conclusion of the NEL. Hmmmm. So let’s toss this study and see what’s left.

In these four studies, higher levels of saturated fat in the diet made some heart disease risk factors get worse, but other risk factors got better. So the overall effect on heart disease risk was mixed or neutral. As a result, these studies do not support the NEL conclusion that saturated fat should be reduced in order to reduce risk of heart disease.


That leaves one lone study. A meta-analysis of eleven observational studies. Seeing as the whole point of a meta-analysis is to combine studies with weak effects to see if you end up with a strong one, if saturated fat was really strongly associated with heart disease, we should see that, right? Right. What this meta-analysis found was that among women over 60, there is no association between saturated fat and coronary events or deaths. Among adult men of any age, there is no association between saturated fat and coronary events or deaths. Only in women under the age of 60 is there is a small inverse association between risk of coronary events or deaths and the reduction of saturated fat in the diet. That sounds like it might be bad news—at least for women under 60—but this study also found a positive association between monounsaturated fats—you know, the “good fat,” like you would find in olive oil—and risk of heart disease. If you take the results of this study at face value–which I wouldn’t recommend–then olive oil is as bad for you as butter.

So there’s your “strong” evidence for the conclusion that saturated fat increases risk of heart disease.


Just recently, Frank Hu of the 2015 Dietary Guidelines Advisory Committee was asked what we should make of the recent media attention to the idea that saturated fat is not bad for you after all (see this video at 1:06:00). Dr. Hu reassured us that, no, saturated fat still kills. He went on to say that the evidence to prove this, provided primarily by a meta-analysis created by USDA staffers (and we all know how science-y they can be), is MUCH stronger than that used by the 2010 Committee.

Well, all I can say is:  it must be.  Because it certainly couldn’t be any weaker.