In a small study that Ludwig and colleagues conducted several years ago, he fed teens a meal that dramatically raised blood sugar levels. (Instant oatmeal, bagels, and a number of breakfast cereals—many of the foods we believe to be good diet choices—provoke sharp spikes followed by sudden crashes in blood sugar.) Overall, the volunteers in the high-glycemic (glycemia refers to the presence of glucose—sugar—in the bloodstream) group described themselves as "very hungry" throughout the day and ended up eating 600 to 700 more calories than those who ate foods that didn't cause wild blood sugar swings, like vegetable omelets, low-fat cheese, apples, and grapefruit. In subsequent studies, Ludwig found that keeping rodents on a high-glycemic diet prompted most—but not all—of them to overeat and gain weight. What set the weight gainers apart was a high insulin response. One of insulin's jobs is to direct excess blood sugar to the liver and muscles, where it can be stored for later use. Because this hormone also helps signal when we've had enough to eat, a sudden surge of insulin followed by a sharp dip could trigger the sensation of hunger and the need to eat.
"We decided to return to humans to test our hypothesis that people vary in their reaction to a blood sugar spike," Ludwig says, and that variation could explain their weight gain and loss. In a study published earlier this year—the one that LeeAnn Henn joined—Ludwig and his colleagues recruited 73 obese adults between the ages of 18 and 35 to take part in an 18-month trial. First, the researchers measured the participants' insulin response. Then volunteers were asked to follow one of two eating approaches: a low-glycemic diet that included certain fruits and vegetables, whole grains such as barley, and other carbs that help keep blood sugar levels even (see The 5 Commandments of Low-Glycemic Eating), or a low-fat diet in which carbohydrate foods were not as restricted but fat was limited to 20 percent of total intake. There were no other restrictions—the volunteers could eat until they were satisfied.
Ludwig's theory held up: By the end of the 18 months, people whose bodies reacted most strongly to a blood sugar spike—they secreted the most insulin—lost an average of 13 pounds on the low-glycemic diet but only about 3 pounds on the low-fat plan. People who secreted less lost an average of three pounds no matter which diet they followed. "These findings provide an explanation for why some people do so poorly on standard low-fat and low-calorie diets," Ludwig explains. "Variability in weight loss isn't just because some people are more motivated."
LeeAnn Henn knows this is true. As she found out later, her body overreacts to blood sugar spikes. She was lucky enough to be assigned to the low-glycemic group, and she lost 75 pounds over the 18 months of the study. Better yet, she has maintained the loss since she finished the study over a year ago. "It's surprising how easy it is to eat this way," says Henn. "I'm evangelical about the plan to my friends." Henn's markers for heart disease and diabetes also improved, as did, on average, those of the entire low-glycemic group, regardless of their insulin patterns: Good (HDL) cholesterol rose, and levels of the potentially harmful blood fats called triglycerides decreased. However, bad (LDL) cholesterol dropped lower among the low-fat dieters.