First things first: science does not deal in absolutes (I suppose “absolute zero” may be an exception, but frankly, probably not). Famously, everything is relative. Even relativity. Einstein’s groundbreaking theory has been subject to on-going revisions in the minds and machinations of his successors ever since he passed the baton.
Second, saturated fat is neither satan, nor saint. It is not the scapegoat some were seeking or perhaps believed it to be, nor any kind of silver bullet- in coffee, or elsewhere. It is not the one thing wrong with our diets, any more than it is the one thing to fix them. In fact, it isn’t one thing at all- it is a diverse class of compoundswith very different metabolic effects. The best of its number appear to be innocuous. That’s important- there really is no evidence of meaningful benefit in any instance, just lack of harm. The less savory of its number remain convincingly associated with harmful effects. This is really what the current state of the science tells us, despite the pop culture contest that is all about who can shout the loudest.
Which saturated fatty acids are most decisively innocuous? Stearic acid, predominant in dark chocolate; and lauric acid, predominant in coconut. This begs the question: do Americans (or most people eating in industrialized countries) get their saturated fat principally from chunks of raw coconut coated in pure, unDutched dark chocolate? Of course not.
In the real world, when saturated fat intake is high, it is from the customary sources: pepperoni pizza; pastrami sandwiches; and ice cream sandwiches for that matter. It is from meat, much of it processed; dairy, much of it processed as well; and food both fast, and fried. Since that is where saturated fat is coming from for the most part, the health effects of saturated fat at varying levels pertain mostly to variations in the intake of, as a shorthand, “meat, butter, and cheese,” noting that much of the intake is actually processed and fast food incorporating these. Therefore, population-level studies looking at variation in saturated fat intake in the U.S. are not studies about eating more or less dark-chocolate covered coconut; they are studies looking at variations in meat, butter, and cheese, and the foods with which folks replace these.
That’s exactly what a new study in the Journal of the American College of Cardiology did. Harvard researchers looked at change in saturated fat intake over time in a large cohort of Americans. They then addressed the key issue missing in prior, and much over-hyped studies on this general topic: with what did they replace saturated fat when they ate less of it?
The results turned out exactly as I would have predicted. In fact, they turnedexactly as I did predict, when I was pretending months ago that such a study had already been done. The study showed that if saturated fat calories were replaced with sugar and/or refined starches (which tend to go together, as in a donut), the risk for cardiovascular disease was equally high both times. Let’s pause there a moment: equally high both times.
This means that those inclined to argue that saturated fat is the only concern when it comes to coronary disease are just as wrong as those inclined to suggest that sugar is the root of all dietary evil, and vice versa. Neither camp of absolute advocacy- it’s all about saturated fat; it’s all about sugar– finds any comfort, or cover, in this study.
Moving on, the study also showed, predictably to some of us, that when saturated fat was replaced with whole grains, polyunsaturated fats, and/or monounsaturated fats, cardiovascular risk went down significantly- with the largest benefit associated with polyunsaturated fats. What are the food sources of such calories apt to be? In the case of whole grains, whole grains, obviously. In the case of PUFAs and MUFAs, they would be the usual suspects: nuts, seeds, olive oil and other cooking oils, avocado, seafood and fish.
So, this study is potentially a nice antidote to the overcooked rhetoric about meat, butter and cheese; and related excesses regarding saturated fat and sugar.
But, of course, the study is flawed. This is not much of a concession, and certainly not an epiphany. Every study is flawed.
How do I know? It’s my job to know. I am formally trained in epidemiology, have written textbooks on research methods and the interface of evidence and clinical decisions, and have run a clinical research lab for nearly 20 years. It’s my job to know; all studies are flawed.
They can be flawed in various ways, and we haven’t the space here to explore them all. We may note generally that the more internally valid a study tends to be (i.e., the more reliably and precisely it generates a true answer to a narrow question), the less generalizable it tends to be to real people in the real world. Randomized, controlled trials are considered a very high standard of evidence, but in fact they often suffer from quite poor external validity, or real-world relevance. In contrast, studies that are less strictly controlled may offer up more realistic findings, but with less precision and more uncertainty. There are trade-offs, these and many others.
Does this mean that we are consigned to be forever clueless and confused about the basic care and feeding of Homo sapiens? Of course not. It means that no one study, nor one kind of study, can be the basis for what we know. We must, rather, acknowledge both the strengths and weaknesses of every study, aggregate the evidence, and lean with the weight of it. And if it begins to tip in a new direction, so must we, if we are honest.
Is that what people do? Heavens no! This is the age of echo chamber dogma, and the loudest wins. We seem to have forgotten entirely the precautionary wisdom ofBertrand Russell. Everyone is competing to be the most foolishly sure, and fanatically certain.
So, what happens is this: when people like the results of a study, they highlight the results with recourse to every New Age megaphone at their disposal, and mums the word about methods. When they don’t like the results, they highlight the methods, with emphasis on the (inevitable) flaws, and mums the word about any strengths. Often, when methods are being criticized this way, the concerns are hypothetical- maybe there was a bias, maybe there wasn’t- but they are propounded as established fact.
So it is that the camp currently advocating for the canonization of saturated fat and a shift to more meat, butter, and cheese will talk only about the flaws of this new study. That would be fine if they did the same with studies that supported their platform, but we don’t roll that way. As noted, when one likes the results, mums the word about methodologic limitations. For instance? The much over-hyped meta-analysis that allegedly (but not really) told us that saturated fat is “good” now suffers at least all of the same limitations as the current study. What’s good for the goose is good for the gander, or really ought to be.
Regarding that overhyped meta-analysis, the senior author was Dr. Dariush Mozaffarian, who has made quite clear that he does not take the results to mean we should be adding saturated fat to our diets. Commenting on the new study, Dr. Mozaffarian had a quibble: the study didn’t differentiate among the various saturated fatty acids. That’s true, but the study didn’t compare different varieties of sugar or fiber, or polyunsaturated fats, either. It was asking a question from altitude, and generated a clear answer. While it’s true that health effects of saturated fatty acids vary, that is not germane to the question: what happens when NET saturated fat calories are replaced by this, that, or the other thing? That’s the question this study asked, and answered with considerable clarity.
Among the clear answers was this, as noted: when net saturated fat calories were replaced with sugar, things got no better, but they got no worse either. In other words, the net effect of saturated fat calories on health was just about equivalent to the net effect of commensurate sugar calories. This doesn’t say that all saturated fatty acids are the same. But it does say what happens when diets are higher overall in saturated fats from the usual sources, namely processed meat and dairy.
Unfortunately, in this age of cyberspatial echo chambers and accelerating news cycles, the true consensus of genuine scientists is overwhelmed by a clamor in the service of notoriety and profiteering. Every study is a hyperbolic headline, and an invitation for some faction or other to hoist their victory flag. And then along comes the next hyperbolic headline, refuting the first, and another faction claims the high ground.
Science doesn’t work this way; it is about the incremental additions of each new study to what we knew before. It requires uncertainty, and allows for mistakes- even as it allows for us to know enough. We can act based on what we know about nutrition, and despite our doubts, to do enormous good; just as we can act, based on what we know of physics and despite our many doubts, to bounce radio waves off of orbiting satellites, and beam electrons through cyberspace.
But between us and that opportunity are the parasites of science who have no apparent interest in the weight of evidence, and instead feed selectively on studies depending on whether they endorse the opinion they’ve already decided is true. You will find them out here every day, in close quarters with the fools and fanatics, busily reTweeting one another’s absolutes. They are a greater peril to us all than either saturated fat or sugar.
Director, Yale University Prevention Research Center; Griffin Hospital
President, American College of Lifestyle Medicine
Founder, The True Health Coalition