You cannot get a good answer to a lousy question.
The current diet study making headlines purportedly asked, and answered this question: Which is better for weight loss and improving cardiac risk, a low-fat or a low-carb diet? For starters, that is a truly lousy question, resurrected from something like the Stone Age. I doubt even the Paleo clan find the question attractive, since they like prehistoric food, not prehistoric research questions about food.
Why prehistoric? Because it is long known and well established that dietary fats run the gamut from good to bad to ugly. No good diet should willfully exclude the monounsaturated fats and omega-3s in nuts and seeds and avocados. I'm pretty sure everybody not stuck under a boulder knows that.
There is ongoing debate today about specific effects of specific fats, but the wholesale cutting of dietary fat intake was pretty much yesterday's news yesterday. The relevant concept today would be plant-based eating, which at the extreme of veganism, tends to be low in fat -- but as an effect rather than an objective. This was not a study of a vegan diet.
The concept of low-carb is also terribly outdated, and was silly when it was first spawned. Everything from lentils to lollipops is carbohydrate. Why on earth would anyone want to treat such a vast expanse of the food supply as if it were just one thing? Sillier still, all plant food is a carbohydrate source. A truly "low-carb" diet is, of necessity, low in all plant foods -- including vegetables, fruits, nuts, seeds, beans, and lentils along with whole grains. This is directly at odds with everything we know about diet and health across the lifespan.
So the research question in this case was, in a word, dumb. But now let me tell you what I really think.
The study didn't even really ask this question. If you only read the headlines, you will believe it did -- and frankly, most people will only read the headlines. I read the study -- and before you start arguing with me, I invite you to do the same.
It was published in the Annals of Internal Medicine, and frankly redounds much to the shame of this generally prestigious journal. Allegedly, the researchers compared a low-fat to a low-carb diet. But in fact, they compared a diet that allowed up to 30 percent of calories from fat to a diet that allowed up to 40 grams of daily carbohydrate.
The baseline diets were, reportedly, roughly 2,000 calories per day on average among the nearly 140 obese (i.e., BMI > 30) study participants. (This is a bit suspect, since calorie intake would be predicted to be higher in obese adults.) That means the allegedly low-fat diet assignment allowed up to 600 calories per day from fat, while the low-carb assignment allowed only about one-quarter that much carbohydrate, 160 calories. The baseline fat intake of the participants in the low-fat assignment was just over 35 percent of calories, so this was, essentially, a diet intervention that didn't intervene much with their diets.
In contrast, baseline carbohydrate intake was 240 grams per day, so while fat intake was "trimmed" 5 percent, carbohydrate intake in that assignment was slashed 75 percent. This might have been billed "a study to compare a really big change from baseline diet to a really small change from baseline diet." I suppose we can all guess why it wasn't called that.
That would be bad and biased enough if the researchers had made any attempt to compare comparably good, or comparably bad versions of the two diets; but they did not. The "low-fat" diet was, for starters, not much lower in fat than the typical American diet, which as we all know -- is basically crap. Shockingly, the fiber intake was virtually identical, at about 15 to 16 grams per day, in both groups throughout the study. You cannot possibly eat any variant on the theme of "good" low-fat, mostly plant-based eating and fix the fiber intake at that pitiful level. The only way to do that is to combine modestly low fat with preferentially crummy foods made mostly from refined starches and added sugars. The study did not provide this level of detail about the diets, but it's clear that the low fat diet was (A) not low fat; and (B) rather crummy. So another title option was: "a comparison of the best low-carb diet to the worst low-fat diet we could come up with." Again, I think it's clear why they didn't go with that one.
And finally, the low-carb diet, since it was actually low-carb, obviously was much more restrictive than the low-fat diet, which wasn't actually low-fat. That had the predictable result: those on the low-carb assignment took in many fewer calories (this information in summarized in Table 2 in the article). Over the first several months of the study, when everyone was probably on their best behavior, the low-carb group took in about 200 fewer calories per day. All the way out at the 12-month mark, when folks were falling off the wagon, the low-carb assignees were still taking in nearly 100 fewer calories per day.
And so, the results were a foregone conclusion. Over the span of a year, obese people who ate less, lost more weight. And those who lost more weight had more improvement in their cardiac risk measures -- which were mostly a mess in the first place due to obesity. Ta-da!
Folks, this is not a defense of low-fat diets. I am not an advocate of low-fat diets. I think the concept is obsolete. I am an advocate, based on the evidence, of wholesome foods in sensible combinations. That dietary pattern can be low or high in fat, relatively lower or higher in carbohydrate. The theme is pretty universal, but the variant on the theme that best suits you and your family is really up to you, and frankly -- that's the beauty of it. All of us can hope to love the food that loves us back.
But I am an advocate of research that is fair, unbiased, and relevant in the real world. As the head of a clinical research lab, and the author of textbooks on research methods -- I am an advocate of sound answers to sensible questions. This is a defense of sense, and good science. Neither of which was on display in this study. As best I can tell, neither was fairness -- nor even honesty. The researchers themselves called their interventions "low-fat" and "low-carbohydrate," and must have known from the start they were cooking those definitions over very different fires.
As a medical journalist, I am a proponent of reading past headlines. Headlines want to titillate you, not educate you. If you are comfortable, they are designed to afflict you. If you are afflicted, they are designed to comfort you. The truth, the whole truth, and nothing but the truth are for the most part -- entirely expendable. Proceed accordingly.
If this had actually been a study comparing comparably good (or comparably bad), genuinely low-fat and low-carb diets, it would still be research stuck in the Stone Age. Such questions have been asked and answered before -- and we can do much better. Had it been such a study, it would still belong on the walls of a cave, rather than the pages of a generally high-quality, peer-reviewed journal. But it might at least have been antique honesty.
But it was not such a study at all. It is both prehistoric and propaganda. It was a comparison of a quite restricted, lower-calorie, low-fiber diet; to a less restricted, higher calorie, equally low-fiber diet. The first worked better for weight loss. Ignored in the mix? Was the diet sustainable when the intervention ended? Could families join in? Would the diet reliably improve health and prevent disease across a lifespan?
Not relevant, because this was a study designed to generate a predictably useless, misleading, and potentially harmful answer to an egregiously silly and perhaps even willfully disingenuous question. If science were generally this bad, we would never have exited the Stone Age.
David L. Katz, MD, MPH, FACPM, FACP is the founding director of Yale University's Prevention Research Center, President of the American College of Lifestyle Medicine, Editor-in-Chief of the journal Childhood Obesity, and the Childhood Obesity expert for About.com. He is the author of 15 books, including three editions of a nutrition textbook, four editions of an epidemiology textbook, and a text on clinical research methods. His most recent book is Disease Proof.