For two decades, Americans have been preached at from almost every corner to reduce their intake of fatty foods in order to lower the risk of heart disease and keep their weight at appropriate levels. Over those two decades, the rate of heart disease has failed to fall, the rate of Type 2 diabetes has skyrocketed and begun to appear more frequently in children, and the nation has become the land of the large and home of the obese. What gives?

Writer Gary Taubes, correspondent for the journal Science, and the experts he consulted think they know the answer, and it is a shocker: Americans would have been better off staying with their higher-fat habits.

Writing in the New York Times Magazine, Taubes puts forward the view, gaining ground in the scientific and medical communities, that Americans have substituted loads of refined carbohydrates and starches for the fat they have given up. The effects have been disastrous: greater obesity, higher levels of dangerous triglycerides, an epidemic of what used to be called "adult-onset diabetes."

Put simply, Taubes' tentative conclusion is that if you have a choice of sausage and eggs or a muffin for breakfast, you'd be better off choosing the sausage and eggs. And don't even look at one of those Big Gulp soft drinks.

It appears to work this way: People like food that tastes good, and fat tastes good. So as Americans worked to reduce the fat in their diets, they looked for substitutes that appealed to their taste buds. Starch and carbohydrates, especially refined carbohydrates, were the substitute of choice: pasta, bread, sodas, potatoes. The shift was reinforced by government urgings to eat less fat and meat, and by a low-fat food industry that necessarily turned to refined carbs as a fat replacement.

Refined carbohydrates move quickly into your bloodstream, sending your blood sugar level into a spike. That triggers the release of lots of insulin to bring the sugar level down; some of the sugar gets burned, some gets stored as fat. You get energy that is quick but short-lived. Soon you are hungry again, but your insulin levels are still high, and that prevents your body from reaching into its stores of fat to burn them. So you eat more carbohydrates and go through the cycle again.

But when you eat, say, a steak that has a high fat level and lots of protein, the food is broken down more slowly. You feel full longer, and your blood sugar avoids the roller coaster, so insulin levels remain lower. As the effects of your meal wear off, your body begins to burn fat.

What about cholesterol? Taubes and the experts he talked to say that the link between cholesterol and heart disease was never actually proven -- a controversial assertion -- and that at any rate the health threat from a high-carbohydrate diet is much greater than the threat from a high-fat diet. In other words, Robert Atkins and his carb-averse, fat-embracing diet may have been right all along -- or at least closer to right than just about every other dietary authority.

"More than two-thirds of the fat in a porterhouse steak, for instance," Taubes writes, "will definitively improve your cholesterol profile (at least in comparison with the baked potato next to it); it's true that the remainder will boost your [bad cholesterol], but it will also boost your [good cholesterol]. The same is true for lard. If you work out the numbers, you come to the surreal conclusion that you can eat lard straight from the can and conceivably reduce your risk of heart attack."

But it's not necessary to go that far. The key to the problem Taubes lays out is the substitution of carbs and starches for fat. The diet Americans ate before the low-fat boom may be better than a high-carb diet, but the best diet may be something altogether different: lean meat, fish, poultry, legumes, vegetables, whole grains, select fruits and unsaturated oils like olive oil. One Wall Street Journal writer calls this the "caveman diet" because it more closely resembles what human beings ate before the development of agriculture.

The central premise of Taubes' assertion is breathtaking: that the dietary change pushed on Americans to get control of weight and heart disease has made things much worse. If that's so, why are Americans learning about it chiefly from diet books and the popular press? Taubes' Times article is an updated and popularized version of a report that appeared in Science magazine more than a year ago, but to most Americans this will come as a complete and infuriating surprise.

Why is that? Taubes suggests two reasons: First, because the American medical community, having spent three decades vilifying Atkins, now finds itself paralyzed by the idea it must embrace his heresies and admit that the low-fat orthodoxy may have caused great harm. And second, only recently have the Atkins-type diets actually been studied against traditional low-fat diets.

Initial results support Atkins, but studies supported by the National Institutes of Health and others are incomplete. If the low-fat dietary recommendations resulted from bad science, the error can't be rectified with more bad science and a leap to conclusions.

But as the headline on Taubes' Times piece asks, "What if it's all been a big fat lie?" Then health-conscious Americans are in for a big adjustment in their eating habits. So are many American dietary experts: They're going to be eating an awful lot of crow.