Children's Health Express Report
Primary Care Conference
Boston, MA

Dietary Treatment of Obesity in Children

This report was reviewed for medical and scientific accuracy by William C. Waters, III, MD , Internist-Nephrologist, Piedmont Medical Center, Atlanta, Georgia.

According to Dr. David Ludwig of the Division of Endocrinology, Children's Hospital, Boston, adult and child obesity has reached epidemic proportions in the United States. Currently, one in four children and half of all adults are overweight-rates that have increased by 50 percent since the 1960s.

Along with obesity comes the dramatically increased risk for serious disease. "The relative risk of cardiac disease and death is a function of the body mass index (BMI*)" Dr. Ludwig told delegates. "A BMI of 20 to 25 is normal, 25 to 30 is overweight, over 30 is obese. Even modest increases in BMI are associated with significant increases in the relative risk for either cardiac illness or death." Other obesity-related conditions are also on the rise. Type II diabetes mellitus (DM), once rare in children, is now commonly seen in this group. In some populations, approximately one-half of new diabetes cases diagnosed in children are Type II DM, or adult onset, rather than Type I DM, or juvenile onset.

Also, research suggests that children and adolescents who are obese are less likely to achieve success at school, more likely to live in poverty as adults, and less likely to marry.

*[BMI = weight (kg) / height2 (m2)]

Current Treatments Bring Limited Success

Current medical treatment for obesity is largely unsuccessful. The findings of a recent National Institutes of Health (NIH) technology consensus conference found that people remaining in weight loss programs lose about ten percent of their weight. "And while ten percent is certainly better than nothing," Dr. Ludwig remarked, "it is a relatively small percent of the [weight that needs to be lost] in most of these patients. Of the ten percent of weight that is lost, about half is typically regained within one year, and there is virtually complete relapse by about five years."

Not long ago obesity was viewed as weakness of character, however, research in the past two decades has suggested a genetic basis for the disorder. But genes are certainly not the overriding cause of the high rates of obesity seen today, Dr. Ludwig said. "In the last 50 years, our genes have not mutated that much; environmental factors must underlie this current observation of obesity," he told delegates.

Although low levels of physical activity certainly contribute to the obesity epidemic, the most commonly cited environmental cause is excessive dietary fat. But cross sectional and longitudinal studies do not support the hypothesis that an individual's weight is dependent on how much fat he or she eats. Weight loss on low fat diets is characteristically transient and modest, Dr. Ludwig noted. And although the consumption of dietary fat has decreased from an average of about 42 percent of calories in the 1960s to about 30 to 34 percent of calories in the 1990s, obesity has continued to increase. "These data should suggest that a primary focus on decreasing dietary fat is not going to give us an answer to the obesity epidemic," Dr. Ludwig remarked.

The Glycemic Index: A New Way of Looking at Carbohydrates

Current United States Department of Agriculture (USDA) dietary recommendations recognize two basic types of carbohydrates (CHO), simple sugars and complex CHO (e.g. grains, starchy vegetables, legumes). Because most foods high in complex CHO contain more nutrients than foods containing high quantities of sugar, the USDA dietary pyramid places complex CHO at the bottom of the pyramid, as a major component of the diet, and simple sugars at the top, to be eaten sparingly.

From a chemical standpoint, physiological response to CHO may not be so well-categorized. A 1983 landmark study in the New England Journal of Medicine (NEJM) showed no difference in glucose response to a meal of sucrose and a meal of wheat flour among diabetic subjects. In fact, a 1998 study of children with Type I DM, published in Pediatrics , showed an increase in glycemic control in these subjects after substituting starch with sucrose.

A more physiologic way of classifying CHO is the glycemic index (GI), Dr. Ludwig announced. "The concept of GI was formed in 1981 and is based on the rate of CHO absorption after a meal. The GI is affected by a variety of factors, including macronutrient composition (proportion of CHO, fat, and protein), fiber content, and extent of food refinement," he said.

The GI of a particular food is calculated by comparing the two-to-three hour blood glucose response to that food and a reference food, usually white bread. To calculate the GI, the area under the response curve for the test food is expressed as a percent of the mean value for the reference food. Foods with a high degree of refined or rapidly absorbed CHO, such as starch, potatos, or rice cakes, have a high GI, whereas unrefined CHO such as whole grain flour, beans, fruits, and vegetables, tend to have a low GI. Dr. Ludwig pointed out that starch actually has a lower GI than table sugar because the latter is comprised solely of sucrose, a dimer of glucose plus fructose. Fructose is a low GI sugar because it has to be metabolized in the liver, whereas starch is digested almost immediately into glucose and rapidly absorbed into the bloodstream, causing an insulin response.

Evidence for the Beneficial Effects of Low GI Foods

Dr. Ludwig noted that 15 out of 16 studies done on the physiological effects of the GI show beneficial effects on appetite, hunger, and food intake with low GI foods. For example, one study performed by Dr. Ludwig and colleagues was designed to compare three different meals that had the same amount of calories but different GIs. The hypothesis of the study was that high GI foods would elicit a sequence of hormonal changes that would limit access to the body's two main metabolic fuels-glucose and fatty acids-and cause obesity. The study enlisted 12 males between the ages of 15 and 18 years. These participants were given either a low GI meal (egg, low-fat cheese, spinach, tomato, grapefruit, and apple slices), a medium GI meal (steel-cut oats, two percent milk, half-and-half cream, fructose, saccharine, and water), or a high GI meal (instant oatmeal, two percent milk, half-and-half cream, dextrose, saccharine, and water). Dr. Ludwig described each meal as "a low-fat meal consistent with current nutritional recommendations." All three meals contained the same proportion of macronutrients.

Participant blood glucose levels after eating the high GI index meal remained higher than those after the medium or low GI meals for about five to six hours. Dr. Ludwig said, "By hours four and five, there was a relative hypoglycemia after the high GI meal that was statistically significant compared to the other meals." Fatty acids were also suppressed to a greater degree after the high GI meal. Dr. Ludwig commented, "We had three meals with identical nutrient compositions, which left the subjects in dramatically different metabolic and hormonal states. When we provided free access to food for these subjects for five to six hours, they ate significantly more after the high GI meal than after the medium or low GI meals. In fact, there was a 53 percent difference in calorie consumption between the low GI and medium GI meals and an 81 percent difference between the high GI and the low GI meals."

In a second study, Dr. Ludwig and colleagues examined the physiological effects of the GI over a seven-day period. The study was motivated by the very poor long-term outcomes of conventional weight-loss diets, which have given rise to the concept of a body weight "set point." According to this theory, Dr. Ludwig explained, changes in body weight elicit physiological adaptations that make it harder to lose weight.

"To examine this issue, we examined ten overweight subjects and gave them their usual diets for a couple of days and then assigned them to either a high or low GI diet that was calorie-restricted by 50 percent-so an individual who would normally get 2,500 calories would get 1,250 calories," Dr. Ludwig explained. "This is a significant, but not severe, energy restriction. Then they continued for a few days on these diets. Body weight decreased during this period of calorie restriction in both groups by a similar amount, as expected. However, the resting metabolic rate decreased by the expected 10 percent on the high GI diet but by less than half as much on the low GI diet. Nitrogen balance was negative on the high GI diet, suggesting breakdown of lean body tissue, whereas it was actually positive on the high GI diet, despite the fact that both treatments provided the recommended daily allowances of protein."

Dr. Ludwig concluded that, compared to conventional low-fat diets, the low GI diet in this study showed beneficial effects on energy expenditure, preservation of lean body tissue (e.g. muscle), and hunger. The body weight "set point" may be adjusted by the composition of the diet, and the low GI diet may be beneficial for the long-term treatment of obesity.

In view of these findings, Dr. Ludwig and his colleagues set out to compare the effectiveness of a low GI diet compared to standard therapy in the treatment of pediatric obesity. In this study, patients in the Optimal Weight for Life (OWL) program at Children's Hospital, Boston, were assigned to one of two treatment teams (not formally randomized). One team prescribed a low GI diet to their patients, and the other prescribed the USDA-recommended reduced-fat diet. Patients on both teams received similar behavioral therapy interventions, physical activity recommendations, and treatment intensity.

The reduced-fat diet was a balanced hypocaloric diet designed to achieve 0.5 lb weight loss per week, Dr. Ludwig explained. Modeled from the USDA food guide pyramid, it was also designed to increase grain products, fruits, and vegetables, and to decrease fat, sugar, and energy-dense foods.

Patients on the low GI diet were instructed to eat until satiety and to snack when hungry. This diet, however, was based on a "low GI pyramid" devised by Dr. Ludwig, with an increase in most fruits, vegetables, legumes, and nuts, and a decrease in refined grain products, starchy vegetables, and concentrated sugar. The proportion of macronutrients in this diet was similar to the reduced fat diet consumed by the comparator group.

The children had a mean age of about 10.5 years and were medically obese (BMI 32.5 to 34.6). Children in the reduced fat group showed essentially no change in BMI over the course of their treatment, which represents a modest improvement over the expected increase in BMI that occurs in growing children over time (one to two units per year). However, the children in the low GI group showed a significant decrease in their BMI, which remained significant after adjustment for age, sex, ethnicity, duration of follow-up, etc.

Other Benefits of Low GI Diets

Low GI diets seem to decrease the risk for cardiovascular disease. They are associated with a lowering of total cholesterol, an increase in high-density lipoprotein (HDL), and a lowering of triglycerides. In addition, a recent study published in the American Journal of Clinical Nutrition suggests a reduction of myocardial infarction rates after adopting a GI diet. "This was after controlling for changes in body weight," Dr. Ludwig explained. "We know that weight loss also improves these outcomes, so a low GI diet may have a synergistic effect if it independently improves these outcomes and promotes weight loss at the same time."

Pointedly, Type II DM rates are much higher in the United States, as is the high proportion of carbohydrates in the American diet. Two major epidemiology studies (Nurses' Health Study [women only] and the Health Professionals Follow-up Study [men only]), suggest that low GI diets, independent of affecting changes in body weight, may decrease the risk of developing Type II DM. There was a 37 percent reduction in both studies.

Dr. Ludwig concluded by saying, "Clearly, more research needs to be done, but the currently recommended grain-based, low-fat diet may be potentially high in GI. When this diet is translated into typical food patterns, the results can be extremely high in GI. Thus, our current obsession with fat-free foods may paradoxically be contributing to the obesity epidemic. An alternative low GI diet, with abundant quantities of fruits, vegetables, nuts and legumes, moderate amounts of protein, moderate amounts of whole grain products, and reductions in the intake of highly refined, starchy food, and high GI sugars, might convey significant benefits by reducing obesity and its associated morbidity."