The Weight Loss Hype: Why Counting Calories Never Works

We've known for more than a decade that the key to weight-loss is to consume fewer calories than you're burning--in other words, eat less, exercise more, or both. That dietary adage was confirmed last week by researchers from the Harvard School of Public Health, with a widely reported study in the New England Journal of Medicine.

In the study, researchers put 811 overweight adults on one of four weight loss plans, which were supposed to vary widely in fat, protein, and carbohydrate content. Most of the reporting discussed Atkins-style high protein diets as similar to the diet's high protein plan and Ornish-style low fat vegetarian diets as similar to the study's low fat plan. Since everyone who cut out 750 calories per day from their diets lost basically the same amount of weight, the take home message seemed to be that none of the popular diets are any better than any of the others.

But upon closer analysis, a very different conclusion emerges.

First, all the tested diets strived to be "heart healthy," which means that they limited saturated fat, limited cholesterol, and contained at least 20 daily grams of fiber (in the form of whole grains, fruits, and vegetables). Anyone with even a passing familiarity with Atkins-style diets knows that all three of these requirements are virtually impossible on such diets--so this study should not be read, in any way, as endorsing an Atkins (or similar high-meat) diet for weight loss.

Second, although the caloric restriction worked for everyone who stuck with it--so it certainly is confirmation that caloric restriction is the way to lose weight--participants at two years were already consuming more than the allowed calories and gaining back weight. In fact, all four groups were on track to be right where they started by year three. In other words, for long term weight loss, all of the diets failed.

The reason for the high failure rate seems obvious to me: All four diets used similar foods and required precise caloric accounting, so all four diets were confusing and very hard to follow. Basically, adherents were asked to be absurdly careful with caloric counts (dropping precisely 750 calories per day) and proportions, but were told to eat identical foods--just in different amounts. So far, diets that require rigorous participant logs and calorie counting have always failed in peer reviewed studies, so this shouldn't have come as a big surprise.

In fact, there is a diet that works--consistently--at helping adherents to lose weight and keep it off, and which has a very high compliance rate: a very low fat, vegetarian diet, as recommended by Dr. Dean Ornish, Dr. John McDougall, Dr. Neal Barnard, and many others.

The very-low-fat vegetarian diets work long-term because they focus on the consumption of fiber and complex carbohydrates, which make you feel full without a lot of empty fat calories, so adherents needn't keep food logs, restrict food intake, or count calories--in other words, they take advantage of the nature of food.

The Harvard study got off to a good start by requiring (in all four groups) 20 grams of fiber per day and by limiting fat and cholesterol, but the reason all four groups failed in the end is that all four diets included meat, which has no fiber at all, and which is packed with fat, relative to whole grains, beans, fruits and vegetables.

As explained by Dr. Ornish in the 2001 foreword to Eat More, Weigh Less, which is the Bible of this way of eating: "When you go from a high fat to a low fat diet, even if you eat the same amount of food, you consume fewer calories without feeling hungry and deprived. Also, because the food is high in fiber, you get full before you consume too many calories. You can eat whenever you're hungry and still lose weight."