Four months ago the so-called “Eco-Atkins” diet made a splash with the publication of a study in the Archives of Internal Medicine. The “Atkins” diet is a low-carbohydrate, high-protein diet that typically includes unhealthy amounts of saturated fat from meat, which the “Eco-Atkins” diet avoids by relying on vegetable sources of protein.
The study compared two vegetarian diets, one low-carbohydrate, high-protein, the other high-carbohydrate, low-fat. The reported result was that both diets could reduce body weight, but only the low-carb diet lowered cholesterol and LDL and it was more effective in lowering blood pressure. Although the study was small - only 22 subjects in each diet group completed the four-week program - the design was properly randomized and the results were said to be statistically significant.
But a look at the details suggests that other factors may be at work. The two diets were:
Low-carb: 27 percent of calories from carbohydrate, 31 percent from protein (mainly wheat gluten and soy), 42 percent from fat; vegan (no animal products), supplemented with vitamins and minerals, and more fiber than in the other diet.
Low-fat: 58 percent of calories from carbohydrate, 16 percent from protein, 25 percent from fat; vegetarian but not vegan because protein sources included egg white and low-fat or fat-free dairy products, less fiber, and no vitamin or mineral supplements.
Apparently the study authors not only turned the tables, but overturned the tables. Where the Atkins diet has more animal content than its rivals, their high-protein diet has less. Not only that, it has more fiber and includes vitamin and mineral supplements.
We know fiber is good, we know vitamin supplements can be helpful, so it’s not clear what difference the proportions of the nutrients might have made. To me, your editor, the big news is simply that the better diet had no animal products at all.
Something funny happened on the way to the study design: Solae, LLC, a company in the soy-based food industry, supported the study and employs two of the study authors. Can any study be trusted?