LEDA Home Page Harvard Law School LEDA at Harvard Law School



Carbophobia: The Dieting Public’s Obsession with Carbohydrates and the U.S. Government’s Response

By Jeremy Berry
J.D., Harvard Law School, 2003
May 16, 2003

Carbophobia: The Dieting Public’s Obsession with Carbohydrates and the U.S. Government’s Response [1]

Abstract:

This paper examines the recent low-carbohydrate diet craze, and analyzes the reaction of the United States government to new scientific evidence potentially indicating that the government’s current food recommendations may actually be contributing to the obesity epidemic in America. After a brief introduction, the paper examines the history of food science, focusing on the government’s efforts to develop the food recommendations now in the Food Guide Pyramid, as well as recent studies indicating that other diets may be more effective for weight loss and cardiovascular health. The paper then discusses the generally-accepted science of what constitutes the various macronutrients, as well as how carbohydrates are processed by the human body. Next, four of the most popular diets in America today are analyzed, including their recommendations and the science, sometimes controversial, behind these diet recommendations. The four diets analyzed are a low-fat diet, a low-carbohydrate diet such as the Atkins Diet, a mixed diet such as the Zone Diet, and the SugarBusters Diet and other low-glycemic index diets. The science background concludes by noting where the science seems to be converging, despite vast differences of opinion amongst the various diet advocates. Next, the U.S. government’s recommendations and response to these studies are analyzed, particularly in three contexts: the recommendation that Americans consumer more “whole grains;” the problem the government has had in defining what constitutes a “complex carbohydrate;” and the choices the government faces in defining a “carbohydrate,” including why the government may have made the wrong choice. The paper concludes with several proposals the government should take to update its food recommendations and increase the health of the American public. Generally speaking, the recommendations are for the USDA to revise its Food Guide Pyramid to reflect current knowledge about carbohydrates, and for the FDA to allow food manufacturers to list the glycemic load on food labels. If the government makes these changes, the health of the American public should improve, or, at a minimum, American consumers will find it easier to improve their health.

I. Introduction

In the United States alone, over 97.1 million adults (54.9% of the adult population) are overweight, defined as having a body mass index greater than 25. 22.3% of the adult population, or 39.8 million, are obese, defined as having a body mass index greater than 30. From 1960 to 1994, obesity increased in the U.S. by over 50%, from from 13.4 to 22.3 percent of the adult population, with most of this increase coming in the last decade. This fattening of America costs $99.2 billion each year, with $51.6 billion as direct costs of obesity and overweight. This is equivalent to 5.7% of the total US health expenditure each year. Americans may be fatter than ever, but they’re spending more and more on weight loss products and services- $33 billion annually throughout the 1990’s.[2]

One avenue Americans are increasingly turning towards is limiting their intake of carbohydrates, as advocated by the popular and controversial Dr. Atkins. The health agencies of the United States government have been some of Dr. Atkins’ staunchest opponents, continuing to advocate a high-carbohydrate low-fat diet despite recent evidence indicating the pitfalls of such an approach. The obsession with carbohydrate intake, and the government agencies’ response to it, is the focus of this paper.

II. History of Food Science

A. Governmental Developments, Including the Development of the Dietary Guidelines and Food Guide Pyramid

Mothers and grandmothers have undoubtedly been telling their offspring for centuries what to eat, but using science to determine nutritional needs is a relatively new phenomenon. In 1894 W.O. Atwater, the first director of the Office of Experiment Stations in the U.S. Department of Agriculture, published tables of food composition and dietary standards for Americans. The tables provided information on protein, fat, carbohydrate, ash, and “fuel” (calorie) content for some of the most common foods. Atwater was the first scientist to link dietary intake to health, and though he did not publish a food guide as such, he did recognize the importance of scientifically determining what foods would lead to a proper intake of essential nutrients, stating that “... for the great majority of people in good health, the ordinary food materials- meats, fish, eggs, milk, butter, cheese, sugar, flour, meal and potatoes and other vegetables- make a fitting diet, and the main question is to use them in the kinds and proportions fitted to the actual needs of the body.”[3]

Building upon Atwater’s work, Caroline L. Hunt, a nutrition specialist in USDA’s Bureau of Home Economics, developed the first food guide in 1916. The guide categorized foods into five food groups: milk and meat; cereals; vegetables and fruits; fats and fat foods; and sugars and sugary foods. Although they did have recommended daily servings for each of the five food groups, Hunt’s publications were generally more similar to buying guides, listing foods to buy each week, as well as recipes and menus.[4]

The weekly buying guides would continue to dominate until the 1940’s, when daily dietary guides first began to appear. In 1941 the Food and Nutrition Board of the National Academy of Sciences released the first Recommended Dietary Allowances (RDA’s) at the National Nutrition Conference for Defense. RDA’s were given for calories, protein, and eight vitamins and minerals. The Bureau of Home Economics of the USDA issued a daily food guide to assist with obtaining these essential nutrients, entitled “Eat the Right Food to Help Keep You Fit.”

The daily guides were updated until 1958, when the USDA issued its Food for Fitness—A Daily Food Guide, introducing the idea of the “Basic Four,” or the four food groups- milk and milk products; meat, fish, poultry, eggs dry beans and nuts; fruits and vegetables; and grain products. The “Basic Four” set minimum consumption for these four food groups, but gave little guidance on how many calories to consume, what else to consume, or on fat and sugars. The Basic Four would remain the focal point of nutrition education until 1979, when the USDA would add a fifth food group- fats, sweets, and alcohol.[5]

Federal involvement in American nutrition reached a turning point in February 1977 when the U.S. Senate Select Committee on Nutrition and Human Needs issued its Dietary Goals for the United States . The Committee set goals for the intake of protein, carbohydrate, fat, fatty acids, cholesterol, sugars, and sodium. Because USDA studies showed that the diets necessary to reach those goals were too different from usual food patterns of Americans to be useful, the Goals were not adopted by the USDA as the basis for food plans and guides.

Although the Goals were not embraced by the USDA for its food guides, the adoption of the Goals by Congress sent a clear signal that the American public needed a better, more authoritative food guide than what was being used for educational purposes until then. Therefore, in 1980, the Department of Health and Human Services, in conjunction with the United States Department of Agriculture, issued Nutrition and Your Health: Dietary Guidelines for Americans , first edition. The Guidelines were based on the most current science at the time, focusing on the nutritional needs of healthy Americans. For the second, 1985 edition, USDA and HHS appointed a Dietary Guidelines Advisory Committee of prominent experts in nutrition and health to review the scientific and medical knowledge current at the time and recommend to the Secretaries revisions to the Guidelines. The Committee has been created each of the times since then that revisions have been made, every five years.

In 1990 Congress passed legislation officially requiring the USDA and HHS to publish jointly every 5 years a report entitled The Dietary Guidelines for Americans, thereby requiring what the agencies had previously done voluntarily.[6] The report must contain nutritional and dietary information and guidelines for the general public, and be based on the preponderance of scientific and medical knowledge current at the time of publication. The report will be promoted by each Federal agency in carrying out any Federal food, nutrition, or health program.

The original 1980 Dietary Guidelines for Americans contained 7 principles: 1) eat a variety of foods; 2) maintain ideal weight; 3) avoid too much fat, saturated fat, and cholesterol; 4) eat food with adequate starch and fiber; 5) avoid too much sugar; 6) avoid too much sodium; and 7) if you drink alcohol, do so in moderation.[7]

In the early 1980’s nutritionists began expressing displeasure with the “Basic Four” food guide, and desired something more specific as a food guide. In response, the USDA created a new daily food guide. This new food guide was to be consistent with the new Dietary Guidelines for Americans , and also help the public to achieve the RDA’s as published by the National Academy of Sciences. The food guide, created to be a reference for total daily nutrition, was to be easy to use and simple to understand, so that everybody would use it. The result, first published in 1992, was the USDA Food Guide Pyramid.

The Food Guide Pyramid, last updated in 1996, emphasized grains such as bread, cereal, rice and pasta, with these making up the bottom layer of the pyramid at 6 to 11 servings per day.[8] The next layer up is divided between the vegetable group, at 3 to 5 servings per day, and the fruit group, at 2 to 4 servings per day. The layer after that is the milk, yogurt, and cheese group (2 to 3 servings per day), as well as the meat, poultry, fish, dry beans, eggs, and nuts group (2 to 3 servings per day). The top of the pyramid consists of fats, oils, and sweets, all to be used sparingly.[9]

B. Private Developments in the Modern History of Food Science

In 1972, Dr. Robert C. Atkins published his own dietary guidelines with Dr. Atkins’ Diet Revolution . Although the diet guide and updated versions have sold over 16 million copies, it did not garner positive scientific treatment until recently, nor did it originally attract the phenomenal publicity it now does.[10] From the start Atkins was denounced by the mainstream scientists, especially by the Journal of the American Medical Association, possibly due to his flamboyant claims about his diet. Atkins did not help his acceptance by the mainstream nutritionists and dieticians by boasting that his diet was “the high-calorie was to stay thin forever” and by advocating bacon, eggs, cheese, and other sources of saturated fats.[11] In the 1970’s, Dr. Atkins was arguing that people could lose weight while eating foods laden with saturated fat, such as lobster with butter sauce or bacon cheeseburgers. Unfortunately for Dr. Atkins, the late 1970’s also marked the beginning of the low-fat trend by nutritionists, and studies linking saturated fat to heart disease were being released, making the Atkins diet’s claims sound preposterous and dangerous.

The 1980’s marked the height of the low-fat movement, and once the USDA put fat at the top of the Food Guide Pyramid and even noted on the cover of the brochure to limit fat intake to no more than 30% of calories, commercial forces took over, ensuring that the reduced-fat recommendations of the government would not go unnoticed. The food industry began producing thousands of products with less fat than their leading competitors, and soon “low-fat” versions of America’s favorite foods were available everywhere. Usually the fats were replaced with sugars, of equally dubious health claims, but so long as the food industry could label the products “low-fat,” the American public would buy them, usually with the belief that it was better for them than the non-reduced fat versions. Since these products had to compete with regular, non-reduced fat versions of the same foods, advertising and labeling was greatly increased, trumpeting the low-fat message of greater health benefits. Soon the majority of the American public was consuming less fat, in an effort to be healthier, as recommended to them by the USDA and HHS.[12]

By the 1990’s, Americans were eating less fat but still not getting any thinner. Dr. Atkins published his updated book in 1992, and this time the American public, if not the nutritionists and scientists, began embracing the low-carbohydrate diet. The mass paperback of Dr. Atkins’ book sold over 15 million copies in the 1990’s, making it one of the top sellers of the decade.[13] The Atkins Diet had arrived.

The Atkins Diet supposedly works by restricting carbohydrates so low that people actually go into a state of ketosis, in which insulin levels fall so low that the muscles and tissues of the body burn body fat for energy, as during starvation. Mainstream food scientists said that ketosis was dangerous, and that it can be excessively hard on the kidneys.[14]

During the 1990’s, other diets also became popular with the American public. In 1995 Dr. Barry Sears published his diet book, A Week in the Zone, which, combined with other titles by Dr. Sears, has now sold over 4 million copies.[15] The Zone Diet, as Dr. Sears has called his recommendations, is based on the body’s hormonal responses to the foods it is fed. The diet strives to have insulin levels within a certain zone (hence the name), neither too high nor too low. To accomplish this, Dr. Sears recommends eating a low-fat protein source with each meal, as well as lots of fruits and vegetables. The meal is to be finished off by adding a touch of fat, usually monounsaturated fats in olive oil, nuts, or some salad dressings. Most importantly, the Zone Diet does not recommend consuming large amounts of grains and other complex carbohydrates, as they tend to release large amounts of insulin, causing insulin spikes which are followed by crashes. Because Dr. Sears advocates replacing some of the lost complex carbohydrates with protein and fat, his diet is neither a low-fat diet such as that prescribed by the USDA, nor a low-carbohydrate diet such as that prescribed by Dr. Atkins.[16] Dr. Sears states that the most important principles of his diet are balance and moderation- balancing the foods you eat with each meal, and limiting portion sizes of any one food group.[17]

In 1998 H. Leighton Steward, Morrison Bethea, M.D., Sam Andrews, M.D., and Luis Balart, M.D., introduced a new diet book into the mainstream, Sugar Busters, which, combined with other titles, has sold 2.5 million copies.[18] Like Dr. Sears and the Zone Diet, Sugar Busters attempts to restrict the release of insulin into the body by avoiding refined carbohydrates and sugars. Sugar Busters brought the notion of a glycemic index into the mainstream; the glycemic index is a measure of how much insulin a certain food product will cause the body to release. According to Sugar Busters, foods with higher glycemic indices cause more insulin, which in turn forces those calories into fat rather than being used for energy, and which also makes the person hungry again sooner.

The Sugar Busters diet recommends more protein than the USDA but less than Atkins or Sears, and on the Sugar Busters diet one still consumes bread, pasta, and cereal, as long as they’re whole wheat and sugarless.[19] Sugar Busters is more cautious of fat than the Zone Diet, although it does allow for unsaturated fats used in moderation.[20] Sugar Buster is easier to follow than the Zone Diet, also, in that it categorizes foods into “acceptable” and “to avoid.”[21]

It is important to note that all of the diets described above (Atkins, Zone, and Sugar Busters) disagree with the USDA Food Guide Pyramid and blame it for the obesity epidemic facing America. All of them disagree with the base of the Food Guide Pyramid, that of bread, cereal, rice and pasta. Atkins and the Zone argue that this is too many carbohydrates, and Sugar Busters insists on whole wheat carbohydrate sources, as well as more fruits and vegetables and less refined carbohydrates.

July 7, 2002 was one of the most gratifying days of Dr. Atkins’ life.[22] On that day, the New York Times Magazine ran a cover story suggesting that Dr. Atkins’ diet might be both effective at weight loss and good for you. The story, “What If It’s All Been A Big Fat Lie,” is inconclusive but suggestive that the mainstream food scientists have been wrong all these years, and that they may wake up to someday find that Dr. Atkins has been right all along. The story gave instant credibility to Dr. Atkins, who rode the wave of publicity to the bank with his company, Atkins Nutritionals.[23]

The story also brought to the forefront of the public mind the scientific studies on the various diets that have been done and are being done. Because scientific studies have been inconclusive at best, with some proving a low-fat diet is best, others proving a low-carbohydrate diet is best, and still others proving that a balance is necessary, the public seems to be unsure of where to turn for nutrition advice. On the one hand, they saw the government sticking with its Dietary Guidelines for Americans, and the corresponding Food Guide Pyramid with bread, cereal, rice and pasta still at the base. On the other hand, each of these other popular diets had millions of adherents, as well as scientific evidence to back up their claims. The American public has been told, at various times, that fat is bad for your heart and will make you gain weight, that protein is bad for your kidneys and may increase the risk of heat disease, and that carbohydrates make you fat and increase the risk of heart disease.

III. Scientific Background Information on Nutrition and Bodily Processes

A. The Science of Macronutrients

The science behind these diets, and the American government’s response to the various scientific studies, will be discussed shortly. But first, it is necessary to acquire an understanding of the basic nutritional concepts on which each of the studies builds. Thus, before we begin examining the controversial studies proposed by the proponents of each of the diets, let us begin by examining the basic, non-controversial science behind nutrition and food science.

First, the body needs energy, in the form of calories, to survive. Energy can come from fat stores or even the body’s digestion of its own muscle tissue, but the most common source of energy is food. All calories humans ingest can be traced to one of four sources, or macronutrients: protein, carbohydrates, fats, and alcohol. Protein and carbohydrates have roughly 4 calories per gram, fats have roughly 9 calories per gram, and alcohol has a little more than 7 calories per gram. For the most part, we will ignore alcohol as a source of calories, as it is devoid of most of the vitamins and minerals the body needs to survive and has its own deleterious health consequences.

Proteins are composed of amino acids linked end to end, with the number and sequence of the amino acids giving each protein its unique characteristics. There are 20 different amino acids, and of these, the body can manufacture all but nine. These nine are therefore called the essential amino acids, as the body must ingest them and cannot make them from other amino acids. The remaining eleven amino acids are nonessential amino acids, meaning that the body could use a surplus of one to alleviate a deficiency of another nonessential amino acid.[24] Animal products such as meat, eggs, and milk provide the richest sources of both essential and nonessential amino acids, meaning that they contain complete proteins. Some plant products contain significant amounts of amino acids, but no one plant product contains all of the essential amino acids, making plant proteins incomplete proteins.

Carbohydrates are long chains of sugar molecules (also known as saccharides) bound together. Carbohydrates can be classified based on how many sugars are linked together. Monosaccharides are single sugars, such as fructose, galactose and glucose. Disaccharides are two sugars linked together, including sucrose (table sugar) and lactose (milk sugar). Polysaccharides, frequently described as being complex carbohydrates, are three or more sugars linked together, and include starch, glycogen, and fiber.[25]

The human digestive system breaks starches down into disaccharides, and disaccharides down into monosaccharides. These monosaccharides are released through the small intestine into the blood stream, where the liver breaks down fructose and galactose into glucose, the body’s primary energy source. If more glucose is available than the body currently needs for energy purposes, it is turned into glycogen, a stored form of carbohydrate. Glycogen is stored in the muscles and liver, or can be converted into fat and stored in the body’s fat deposits. Unlike all other carbohydrates, the body is incapable of breaking down fiber into monosaccharides, and thus fiber passes intact through the digestive system. Therefore, fiber is not a significant source of energy for the human body.

Fat is a group of chemical compounds that contain fatty acids. There are four types of fatty acids: saturated, monounsaturated, polyunsaturated, and trans-fatty acids. All fatty acids are composed mostly of carbon and hydrogen atoms. A saturated fatty acid has the maximum possible number of hydrogen atoms attached to each carbon atom. A monounsaturated fatty acid is missing one pair of hydrogen atoms in the fatty acid molecule. A polyunsaturated fatty acid is missing more than one pair of hydrogen atoms. Trans fatty acids have had some of the missing hydrogen repairs replaced through partial hydrogenation.[26]

Saturated fatty acids are most commonly found in meat products, especially red meat, butter, and cheese. Unsaturated fatty acids tend to be found from plant sources, especially olive oil, avocados, and nuts. However, salmon and other fatty fish often contain significant amounts of omega-3 polyunsaturated fatty acids, which have been shown in some studies to have beneficial effects on cholesterol and triglyceride levels, possibly leading to decreased chances of heart attacks.[27]

B. Scientific Examination of Carbohydrate Digestion and Bodily Processes

Next, let’s examine how the body processes carbohydrates a little more in-depth, so that we can better understand the debates between the various parties and the scientific studies behind them. As previously stated, all carbohydrates, from table sugar to starches, can be digested and converted into glucose, which is then sent throughout the body via the bloodstream. Until relatively recently, digestion rate and therefore blood glucose levels was thought to depend upon the length of the saccharide chains of the carbohydrate being digested. Thus, it was thought that sugars and simple carbohydrates would be digested faster, and starches and other complex carbohydrates would be digested slower. Although the government was never able to propose an adequate definition of “complex carbohydrates,” nor of “simple carbohydrates,” it did seemingly acknowledge this assumption about the correlation between chain length and digestion rate in the Food Guide Pyramid, which has starchy carbohydrates at the base and sugars at the top.[28]

Recently, scientists have questioned the relevance of chain length to carbohydrate digestion rates. A study published in November of 1978 found that blood glucose and insulin levels were not different by any statistically significant amount between ingestion of carbohydrates as monosaccharides, disaccharides, and polysaccharides. The study concluded that dietary form in which starch is administered, rather than chain length, was probably most important in determining blood glucose and insulin levels.[29] Another study, this one in 1983, also found similar evidence that so-called simple carbohydrates were digested no faster than so-called complex carbohydrates. In this study, subjects were given five meals, each containing similar amounts of carbohydrates, but in different forms. The five test carbohydrates were glucose, sucrose, fructose, wheat starch, and potato starch. Glucose levels for sucrose (table sugar) were significantly different from that of wheat starch, potato starch, and glucose. Fructose had slightly lower glucose levels, sometimes achieving significance, and sometime not. These researchers concluded that table sugar did not raise blood glucose levels any more than most other forms of carbohydrates.[30]

Despite this evidence, scientists have long known that some carbohydrates cause greater changes in blood glucose levels than others. This led researchers to the glycemic index, a system of classifying carbohydrate-containing foods according to their glycemic responses. Although glycemic indices vary, and cannot be predicated based on saccharide chain length, there are certain principles one can follow. In general, most highly refined, starchy carbohydrates have a high glycemic index, and nonstarchy vegetables, fruit, and legumes tend to have a lower glycemic index. Eating fat or protein with your carbohydrates appears to lower the glycemic index of the overall meal, but the glycemic response to mixed meals can still be prediced from the glycemic index of the constituent foods. The term “glycemic load” is also frequently used to describe the effects of certain foods on blood glucose levels. The glycemic load, defined as the weighted average glycemic index of individual foods multiplied by the percentage of dietary energy as carbohydrate, is used to describe the impact of foods with different macronutrient compositions on glycemic response.[31]

As discussed, meals high on the glycemic index or glycemic load will cause the body to release more glucose into the bloodstream than low-glycemic index foods. In fact, blood glucose levels two hours after ingestion may be at least twice as high after a high-glycemic index meal than a low-glycemic index meal containing identical nutrients and energy (calories).[32] These high levels of blood glucose concentration stimulate the pancreas to release insulin, which transports nutrients into the cells of the body. Insulin also dictates energy storage, basically where the calories consumed go and to what degree. High levels of insulin can lead the body into storing energy as body fat, with negative consequences.[33] Because of the rapid uptake of nutrients due to high insulin levels, soon (after several hours) the energy is absorbed and blood glucose levels fall rapidly, much more rapidly than after low-glycemic index meals. Blood glucose levels fall so far that the body goes into a hypoglycemic state where it needs to elevate the levels. In an effort to achieve homeostasis and return the blood glucose levels to normal, the body gets hungry and signals for the person to eat again. Had the person eaten a meal with a lower glycemic index, the hypoglycemic state would never have occurred, and blood glucose levels would have remained more stable over a longer period of time.[34] Put simply, the higher the glycemic index (more accurately, load) of your meal, the sooner you will get hungry again, even with the exact same amount of macronutrients and calories.

All of this said about increased fat storage from consumption of carbohydrates is mainstream, accepted food science. That said, most food scientists believe that people gain weight from eating more calories than they burn, and that it doesn’t matter which macronutrient they get these calories from. That is to say, if you eat more calories from any macronutrient (protein, carbohydrates, or fat) than you burn, you will gain weight. However, some recent studies have begun to challenge this notion that a calorie is a calorie.[35]

We should pause here to examine why humans store fat, and why some humans seem predisposed to store more fat than others. The current mainstream hypothesis as to why some humans are getting fatter in today’s society but some aren’t is called the Thrifty Gene Hypothesis. This hypothesis supposes that genes that lead to more fat storage provided an appreciable advantage to our evolutionary ancestors. Those families with the thrifty gene would store excess food as fat during times when food was plentiful, and then rely on this fat storage for energy during times of famine or when food is less plentiful. It’s easy to see that those families without the gene might be more likely to die off during periods of famine.[36]

Equally important, however, is that our ancestors did not dine in restaurants or live sedentary lifestyles. They went on hunts for meat, with hunts lasting for days and covering vast territory. They traveled for miles to get to gather wild sources of fruits, grains, and nuts. Because their chief sources of protein were wild animals, they ingested less fat than we do, eating animals raised in feedlots with much higher fat contents. Neither did they eat highly refined carbohydrates, instead primarily eating whole grains that had more fiber to slow digestion.[37] Thus, we see that evolution could be used to support either the low-fat or the low-carbohydrate diets.

IV. The Scientific Basis for Four Modern Recommended Diets

Now that we understand how the body processes carbohydrates, let us move to the more controversial science that each of these diets claims as evidence of their effectiveness at weight loss and health improvements.

First, a quick note about agreements between the scientists. All of the diets described above recommend moving away from simple carbohydrates, which is difficult to define precisely but usually means avoiding sugary foods and highly refined carbohydrates such as white bread and those made with white flour. These sugars and refined carbohydrates are usually devoid of nutrients and, as discussed previously, cause blood glucose and insulin levels to rise, usually causing one to be hungry again before he/she normally would need more food. Instead of refined carbohydrates, nearly all scientists agree that fiber-rich whole foods are much healthier. Even better, according to all nutritionists, are fruits and vegetables, although several of the diets prefer that one choose vegetables and fruits with a low glycemic index, to avoid rising blood glucose and insulin levels. Everybody also agrees that omega-3 fatty acids are a healthy way to lower the risks of coronary heart disease; these fats are frequently found in cold-water fish such as salmon, and even those advocating low-fat diets recommend consuming this type of fat. One type of fat most experts think should be avoided is trans-fatty acids; these fats are widely believed to cause heart disease.[38]

A. The Low-Fat Diet

The chief advocate of a low-fat diet similar to that proposed by the Dietary Guidelines for Americans and the USDA Food Guide Pyramid for the last twenty years is Dr. Dean Ornish. Dr. Ornish has sold over 25 million copies of his books on reversing and preventing heart disease, and advocates a low-fat high-carbohydrate diet rich in fruits and vegetables and whole grains. Recognizing that fat has more than twice as many calories per gram as carbohydrates, Dr. Ornish believes that people should fill up on complex carbohydrates since gram for gram they won’t consume as many calories. For reversing heart disease, Dr. Ornish recommends an extremely low-fat diet with fat less than 10% of total daily caloric intake. For simple weight loss and heart disease prevention, Dr. Ornish still advocates cutting back on fat, but allows up to 15% of calories to come from fat.[39]

Dr. Ornish bases his diet on the simple premise that a calorie is a calorie, whether it comes from protein, carbohydrates, or fat. If we assume that people are going to eat the same number of grams of food no matter what the macronutrient composition, people will consume less calories if they eat low-fat less-dense meals rather than high-fat calorie-dense meals. Since fat has 9 calories per gram and carbohydrates and protein only have 4 calories per gram, people can eat twice as many grams of carbohydrates than they can fat. Thus, a meal with 15 grams of protein, 65 grams of carbohydrates, and 20 grams of fat would have 500 calories, whereas a meal with an equal number of grams of food but a macronutrient mix of 30 grams of protein, 30 grams of carbohydrates, and 40 grams of fat would have 600 calories.[40] According to Dr. Ornish, since you’ve consumed 100 less calories, you’ll lose weight. Dr. Ornish’s prescription for weight loss, in fact, couldn’t be simpler- consume fewer calories than you burn each day. “Here’s how you lose weight: 1. Burn more calories. 2. Eat fewer calories. That’s it.”[41]

For evidence of effectiveness, Dr. Ornish cites to a study he performed that had an experimental group of patients with heart disease adopt a 10% fat, vegetarian high-carbohydrate diet, exercise, and stress reduction techniques, comparing the results of that group to a control group (also with heart disease) that had more minor changes in their diet and no other lifestyle changes. The low-fat group lost 23.9 pounds after one year, and kept an average of 12.8 pounds off after 5 years. In contrast, the control group’s weight changed little from the baseline. The experimental group lowered their LDL (bad) cholesterol levels 40% after one year, and LDL levels were still 20% lower than the baseline after 5 years; the control group had their LDL cholesterol levels decrease by 1.2% after one year, and 19.3% after 5 years. The difference between the two groups after 5 years was not statistically significant, possibly because 60% of the control group, and none of the experimental group, took lipid-lowering drugs between years 1 and 5. The drop in LDL cholesterol levels is of note, because high LDL cholesterol levels are thought to be one predictor of heart disease. There was no significant changes in triglycerides, which are also widely considered to be another predictor of heart disease. The low-fat group also experienced a 91% decrease in frequency of angina pectoris (chest pain) after one year, and maintained a 72% reduction after 5 years. Control group patients had a 186% increase in angina after one year, but then three of the control group patients underwent coronary angioplasties, with the result being a 36% decrease in control group chest pain after five years, not statistically different from the experimental group. Interestingly, HDL (good) cholesterol levels also decreased for the experimental group, but the ratio of LDL to HDL levels was still improved. While low levels of LDL cholesterol help prevent heart disease, as do high levels of HDL cholesterol, the most important factor, according to Dr. Ornish and most doctors, is the ratio of good to bad cholesterol, which improved for the low-fat vegetarian diet group. Lastly, 20 of 28 (71%) of the experimental patients maintained the lifestyle changes for 5 years, and 15 of 20 (75%) of patients in the control group made more minor changes and completed the 5-year follow-up. This is important in that a frequent criticism of the low-fat diet Dr. Ornish prescribes is that it is too hard to follow and people give up easily.[42]

Speaking just in terms of weight loss, Dr. Ornish believes that any diet that restricts calories can cause weight loss, since a calorie is a calorie. He advocates a low-fat diet because of the health benefits associated with such a diet, including a decreased risk of heart disease and improved cholesterol levels. On his point that weight loss can occur regardless of macronutrient composition of the diet, he does have some studies to back up his claim. Noakes and Clifton divided subjects into three groups. The first group ate a low-fat diet with fat less than 10% of daily calories, and saturated fat only accounting for less than 3% of daily calories. The second group had a high saturated fat diet, with total fat intake at 32% of calories, and saturated fat intake at 17% of calories. The third group had a high unsaturated fat diet, with fat intake at 32% of calories and 6% of calories coming from saturated fat. All three groups consumed the same number of calories, and all three lost an average of 9.7% of their weight over 12 weeks, with no significant differences between groups. Interestingly, LDL (bad) cholesterol decreased roughly 20% with the low-fat and high-unsaturated fat groups, but only decreased 7% for the saturated fat group, which is significant at the .05 level. This would seem to support Dr. Ornish’s claim that any calorie-restricted diet can cause weight loss, but that a low-fat diet has other health benefits over a diet high in saturated fat, such as the Atkins diet.[43]

Similarly, Alford et al performed an experiment with three groups of women on 1200 calorie diets, with the groups eating diets that contained either 25%, 45%, or 75% carbohydrate. After 10 weeks, no significant differences were found between groups for weight loss, body composition, cholesterol, triglycerides and percent body fat. Alford et al concluded: “Selection of a weight reduction diet can be designed around preferred food patterns and nutrition needs of individuals rather than a proportion of the energy provided by carbohydrate in healthy adult women.”[44] This would seem to prove Dr. Ornish’s point that a calorie is a calorie, and he would argue the 75% carbohydrate group had less hunger cravings than the 25% carbohydrate group.

As even more evidence of there being no significant difference in weight loss between dietary compositions, Golay et al. put subjects on a 1200 calories diet, some with 25% of calories from carbohydrates, some with 45% of calories from carbohydrates. Weight loss and fat loss occurred similarly for both groups, and there were also similar decreases between groups for waist/hip ratio, blood glucose levels, cholesterol levels, and triglycerides levels. The authors concluded: “Neither diet offered a significant advantage when comparing weight loss or other, metabolic parameters over a 12 [week] period.”[45]

B. The Low-Carbohydrate (Atkins) Diet

Largely due to the fact that it has been so long ignored by the mainstream food scientists, there are not as many studies on the Atkins diet as there are on the low-fat diets. Nonetheless, low-carbohydrate diets are being increasingly studied, as doctors find themselves with little information to give patients asking about the Atkins diet.[46]

One of the most basic arguments made by Atkins proponents doesn’t involve science at all, but rather an argument based on correlation. In 1980, the U.S. Department of Agriculture made the official recommendation in the U.S. Dietary Guidelines For Americans that fat should account for no more than 30% of daily calories, and the Food Guide Pyramid still prominently displays this limit on the front cover of the Food Guide Pyramid brochure. The theory was that fat was making Americans fat, and that carbohydrates should replace the fat that we stop eating. Since then, Americans have reduced their average fat intake from 40% of daily calories to 34%, but heart disease incidence hasn’t declined.[47] When the Food Guide Pyramid came out in 1992 with its recommendation of 6 to 11 servings of grains, breads, and cereals, 56% of American adults were overweight; today, 64% of Americans are overweight, despite the increase in percentage of calories from fat. The rate of obesity has fared even worse- since 1991 the rate of obesity has increased by 74%, with 21% of the adult population obese today.[48] Atkins supporters ask why Americans started getting fatter right after the USDA Food Guide Pyramid came out recommending increased carbohydrate consumption and less fat consumption. Dr. Ornish points out that actual fat grams consumed has increased over the last 10 years, but fat as a percentage of calories has decreased because we have so drastically increased our daily caloric intake. Also, Dr. Ornish think 30% of calories from fat is still too high, and recommends it being lowered to 10 to 15%.[49]

Despite the relative ignoring of the Atkins diet by most health professionals, some scientific studies have been performed, and more studies are currently under way. First, a note about the Atkins diet is in order. Dr. Atkins recommends consuming very low levels of carbohydrates (less than 50 grams) the first week of his diet, which puts the body in a state of ketosis. Ketosis, which can also be caused by prolonged fasting, is when whole-body metabolism shifts towards obtaining a greater percentage of energy from fat stores, which can result in the production of ketone bodies in the liver. Ketones, which are large protein molecules, can serve as an alternative fuel for tissues to spare carbohydrate and protein. Ketone body production indicates that fat metabolism has been accelerated and that all the enzymes involved in the metabolism of fat are operational.[50] There have been some concerns that ketones are hard for the kidneys to process, and that ketosis may increase the risk of kidney stones.

One of the more recent studies involving the Atkins diet, led by University of Pennsylvania’s Gary Foster, divided overweight subjects into two classes, putting one on a high-carbohydrate diet and another on a low-carbohydrate diet. After three months, the low-carbohydrate group had lost an average of 19 pounds, 10 more than people on the standard high-carbohydrate diet. Even better was the differences in cholesterol; the Atkins dieters’ bad cholesterol did go up seven points, but their good cholesterol rose almost 12 points. Changes in the high-carbohydrate group weren’t as dramatic; their bad cholesterol went down slightly while good cholesterol levels were unchanged. Most significantly, the Atkins dieters’ triglycerides dropped 22 points, but the high-carbohydrate group’s didn’t change.[51]

One study found that increased carbohydrate intake as a percentage of daily calories lowered HDL (good) cholesterol levels. More specifically, the study found HDL levels were positively correlated with fat intake and saturated fatty acid intake, but negatively correlated with carbohydrate intake and simple carbohydrate intake, all significant at the .001 level.[52] Thus, this study would seem to suggest that consumption of carbohydrates would lower good cholesterol and possibly lead to an increased risk of heart disease. Similarly, studies have found that a diet low in fat and high in carbohydrates may raise triglyceride levels,[53] and may actually increase LDL (bad) cholesterol levels.[54]

Another study looking favorable for the Atkins diet tried to discover the effect of a low-carbohydrate ketogenic (meaning that it induces ketosis) diet on healthy men who are not overweight and don’t have high cholesterol. The study forced the men to keep their weight steady while on the low-carbohydrate diet (30% of calories from protein, 8% from carbohydrate and 61% from fat), and the control group maintained their previous diet (17% of calories from protein, 47% from carbohydrate and 32% from fat). After six weeks, there were no changes in blood profiles for the control group, but HDL (good) cholesterol had increased 11.5% for the low-carbohydrate dieters (p=.066). This increase in HDL cholesterol is important in that it cannot be attributed solely to the weight loss, as the researchers forced the men to maintain their previous weight.[55]

Another study involving healthy, normal-weight men had an astonishing finding. Volek et al. had 12 healthy men switch from their normal high-carbohydrate diet (48% of calories from carbohydrates, 32% from fat) to a very-low-carbohydrate diet (8% of calories from carbohydrate, 61% from fat), and 8 men remained on their normal diet as a control group. After six weeks, there were no significant changes in the control group. The low-carbohydrate group, however, lost an average of 3.3 kg of fat (p<.05), despite consuming more calories than the high-carbohydrate group (2335 to 2190 calories per day, respectively). The low-carbohydrate group also had an increase in lean body mass of 1.1 kg, suggesting that very-low-carbohydrate diets favor loss of fat, and help to preserve lean muscle tissue over longer periods of time. The fact that 12 normal, healthy men lost fat and increased lean body mass while consuming more calories than the control group, without any change in physical activity, suggests that the very-low-carbohydrate ketogenic Atkins diet deserves further study, as the sample size of 12 and 8 men was rather small.[56] Although the precise mechanism for how the body lost fat and gained lean body mass is not yet understood, the test subjects also exhibited a decrease in insulin levels of 34%, which explained 70% of the variability of fat loss for the very-low-carbohydrate group. Studies like this one may also cast doubt on the theory that all calories are equivalent, since the low-carbohydrate group consumed more calories yet lost more fat.

C. The Zone Diet

Dr. Sears and his Zone diet also have scientific studies proving effectiveness. One study divided women into two groups- one consuming 40% of daily calories as carbohydrate, 30% as protein, and 30% as fat, and the other consuming 55% of calories as carbohydrate, 15% as protein, and 30% as fat.[57] This ratio of 40% carbohydrate, 30% protein, and 30% fat is what the Zone diet recommends for optimum blood glucose and insulin levels.[58] The Zone dieters ate 125 grams of protein per day, 171 grams of carbohydrates, and 50 grams of fat; the high carbohydrate group ate 68 grams of protein per day, 240 grams of carbohydrate, and 50 grams of fat, giving both groups roughly 1650 calories per day. After 10 weeks, weight loss was similar between the two groups, but insulin levels two hours after ingestion of a meal were significantly (p<.05) higher for the high-carbohydrate group than the Zone group.[59] We have already discussed how greater insulin levels after eating causes increased absorption of nutrients into the body, causing the subject to feel hungry again sooner, and how insulin encourages the storage of extra energy as fat. Thus, the Zone diet effectively stabilized blood glucose levels and reduced the insulin response to meals.[60]

Another study involving the Zone diet sought to explore changes in body composition during weight loss and changes in blood profiles between diets. Once again, women were separated into two groups, the first consuming a diet of 41% carbohydrate, 30% protein, and 29% fat, the other (high-carbohydrate) group consuming 58% carbohydrate, 16% protein, and 26% fat. Daily calories consumption was equal for the two groups at roughly 1650 calories. Both groups lost about 7 pounds of weight, but the ratio of loss of fat to loss of lean body mass was significantly improved for the Zone dieters, at 6.36 to the high-carbohydrate group’s 3.92 (p<.05). “The ratio of fat/lean loss demonstrated that the higher protein [Zone] diet partitioned a significantly greater percentage of the weight loss to body fat while sparing lean tissue [such as muscle].”[61] The two groups lost a similar amount of total cholesterol (about 10%) and of LDL (bad) cholesterol, with neither group showing a significant change in HDL (good) cholesterol. Women in the high-carbohydrate group had higher insulin responses to meals, and the Zone women reported greater levels of satiety. The Zone diet thus appears to help preserve lean body mass while losing fat, and, as before, aids in the stabilization of blood glucose and insulin levels.

Another study limited subjects to either a high-carbohydrate diet (58% of calories from carbohydrate, 12% from protein, 30% from fat) or a diet similar to the Zone diet (45% carbohydrate, 25% protein, 30% fat). Weight loss after six months was 5.1 kg in the high-carbohydrate group and 8.9 kg in the Zone group, with the Zone group’s weight loss significantly greater (p<.001). Fat loss amounted to 4.3 kg for the high-carbohydrate group and 7.6 kg for the Zone group, again a significant difference (p<.0001). This study is especially important since it wasn’t conducted in a clinically controlled setting but rather allowed the subjects to go home and eat as many calories as they wanted. Thus, the Zone diet appears to cause greater weight loss and fat loss when calories are not limited to a specific number as with previous studies.[62]

D. The SugarBusters Diet

The SugarBusters diet, with its emphasis on the glycemic index of foods, has also received attention from the food scientists. One study first put subjects on either the American Heart Association (AHA) Step 1 diet (similar to that recommended by USDA Food Guide Pyramid, restricting fat intake to 30% of calories and saturated fat to 10%) or a low-glycemic index-low-fat-high-protein diet. In the first stage of the experiment, subjects were allowed to eat as many calories as they wanted, until they felt full. Subjects on the AHA Step 1 diet experienced an increase of triglyceride levels of 28% (p<.05), and a 10% reduction in HDL (good) cholesterol levels (p<.01), resulting in a significant increase in cholesterol:HDL-cholesterol (p<.05), this ratio being commonly used to assess the risk of coronary heart disease. In contrast, subjects on the low-glycemic index-low-fat-high-protein diet spontaneously decreased daily caloric consumption 25% (p<.001), and decreased triglyceride levels 35% (p<.0005). In the second phase of the experiment, some subjects were fed the AHA Step 1 diet, but calorie consumption was restricted to the same as that consumed by the low-glycemic index-low-fat-high-protein diet group. These subjects experienced a marked increase in hunger (p<.0002) and a significant decrease in satiety (p<0.007) as opposed to the low-glycemic group.[63]

In another study, 107 obese but otherwise healthy children were put on either a low-fat or low-glycemic index diet. The body mass index decreased 1.53 points in the low-glycemic index group, and only an insignificant .06 points, with the difference significant at the .001 level. Body weight decreased 2.03 kg for the low-glycemic index group and increased 1.31 kg for the low-fat group, with the difference again significant at the .001 level. The differences between the two diets remained significant during multivariate analysis as well, controlling for BMI baseline, baseline weight, age, sex, and ethnicity (p<.01).[64] Again, we see that the low-glycemic index diet resulted in a significant decrease in body mass index and body weight over the standard low-fat diet.

In a multiple linear regression reanalysis of a cross-sectional study, a relationship was found between HDL (good) cholesterol levels and dietary glycemic index. Total cholesterol and LDL (good) cholesterol levels were not influenced by glycemic index or any other component of dietary carbohydrate or fat. “In this study, only body-mass index, smoking, and the glycemic index of the diet were identified as potentially modifiable risk-factor variables for HDL-cholesterol concentration.” Thus, it would seem that modifying one’s diet based on the glycemic index is the best, and perhaps only dietary, way to increase HDL cholesterol levels, thereby improving the LDL:HDL cholesterol ratio.[65]

A recent study, studying the dietary habits of over 75,000 U.S. women with a 10 year follow-up, found that glycemic load was directly associated with risks of coronary heart disease after adjusting for age, smoking status, total energy intake, and other coronary disease risk factors. Interestingly, “[c]arbohydrate classified by glycemic index, as opposed to its traditional classification as either simple or complex, was a better predictor of CHD [coronary heart disease] risk.” The authors concluded: “These epidemiologic data suggest that a high dietary glycemic load from refined carbohydrates increases the risk of CHD [coronary heart disease], independent of known coronary disease risk factors.”[66]

E. A Convergence of Food Science?

Given all these seemingly conflicting studies, one may wonder if there is any agreement amongst all the studies and individual diets. The great majority of the studies supporting all of these diets seems to indicate that highly refined carbohydrates (such as white bread, white rice, and soft drinks with high-fructose corn syrup) generally increase fat storage due to their high glycemic indices (or glycemic loads). Since we know that being overweight or obese greatly increases the risks of having heart disease, we can safely conclude that the science agrees that highly refined carbohydrates are bad for us.

V. Analysis of the U.S. Government’s Efforts and Response to Glycemic Index Studies

A. The Food Guide Pyramid, Dietary Guidelines, and “Whole Grains”

Now knowing that refined carbohydrates are not healthy, especially in large quantities due to an even greater glycemic load and consequently greater insulin response, let us examine the USDA Food Pyramid and the Dietary Guidelines for Americans, 5th Edition. As previously noted, the USDA Food Guide Pyramid made its debut in 1992 as an easy and quick way for Americans to visualize and absorb the dietary advice of the Dietary Guidelines for Americans. The Food Guide Pyramid was last updated in 1996, and despite the plethora of scientific evidence available by then showing the dangers and negative health consequences of highly refined carbohydrates the pyramid still fails to differentiate between highly-refined and whole-grain carbohydrates.[67]

One may wonder why the government would ignore the concerns of food scientists over foods high in glycemic index. Indeed, considering that the glycemic index was developed in 1981 as a better way to classify carbohydrates rather than classifying them by the saccharide chain length, one would assume the concept of glycemic index was rather prominent by 1992, but definitely by the time of the Food Guide Pyramid update of 1996.[68] To understand why the USDA would ignore concerns over the glycemic index of highly refined carbohydrates, one needs to remember that the government at that time was focusing its efforts on reducing the amount of fat that Americans consumed, as evidenced in both the Food Guide Pyramid and the Dietary Guidelines for Americans. Also, the developers of the Food Guide Pyramid wanted it to be understandable and memorable, especially for uneducated Americans who probably need the nutrition advice the most. Given these priorities, it is little wonder that the Food Guide Pyramid fails to distinguish between highly refined carbohydrates and whole grain carbohydrates- the USDA had too many other concerns to bother with also making this distinction in an easy-to-use memorable graphic.[69]

What seems harder to understand is why the brochure accompanying the USDA Food Guide Pyramid fails to distinguish these sources of carbohydrates. Although this brochure still needed to be easy to understand, it was 17 pages long and this distinction easily could have been made without complicating things too much.[70] Indeed, the only time the brochure appears to make a distinction is when it states: “To get the fiber you need, choose several servings a day of foods made from whole grains, such as whole-wheat bread and whole-grain cereals.”[71] While it is true that whole grain foods do contain more fiber than refined carbohydrates, no mention is made of the lower glycemic index of these foods and the accompanying benefits of choosing foods low in glycemic index. Considering that 60% of Americans are familiar with the Food Guide Pyramid,[72] a simple paragraph in the brochure discussing choosing whole wheat bread over white bread, brown rice over white rice, whole-grain pasta over refined pasta, etc., would have been tremendously helpful in assisting Americans to lose weight and stay healthy.

The Dietary Guidelines for Americans, 5th Edition, does do a much better job of encouraging the consumption of whole grain foods. Page 20 of the brochure is devoted almost entirely to encouraging an increased intake of whole grain foods. The first paragraph describes why, according to the government, people should choose whole grains:

Choose a variety of grains daily, especially whole grains

F oods made from grains (wheat, rice, and oats) help form the foundation of a nutritious diet. They provide vitamins, minerals, carbohydrates (starch and dietary fiber), and other substances that are important for good health. Grain products are low in fat, unless fat is added in processing, in preparation, or at the table. Whole grains differ from refined grains in the amount of fiber and nutrients they provide, and different whole grain foods differ in nutrient content, so choose a variety of whole and enriched grains. Eating plenty of whole grains, such as whole wheat bread or oatmeal (see box 11 ), as part of the healthful eating patterns described by these guidelines, may help protect you against many chronic diseases. Aim for at least 6 servings of grain products per day—more if you are an older child or teenager, an adult man, or an active woman (see box 7 )—and include several servings of whole grain foods. See box 8 for serving sizes. [73]

Several aspects of this paragraph are interesting. First, like the Food Guide Pyramid brochure, the Guidelines recommend consuming whole grain foods, but only for one reason. They both fail to note that whole grain carbohydrates have a lower glycemic index and thus help to prevent an insulin rush which usually leads to increased body fat. Second, the paragraph is of note because it recommends choosing a variety of “whole and enriched” grains (emphasis added). Had the committee carefully read the reports on the glycemic index of refined carbohydrates, it would have recommended whole grains over enriched grains. Lastly, the paragraph concludes by recommending to consume at least 6 servings of grain products per day, and to include several servings of whole grain foods. When the recommendation for grains is 6 to 11 servings per day, only recommending several servings of whole grain foods means that most of the carbohydrates people get will be from enriched grains, most likely in the form of high-glycemic index refined carbohydrates. According to those food scientists associated with the glycemic index, this is not a message that should be conveyed to the American public.

Americans seem to know the Food Guide Pyramid, and they have taken its advice to heart. From 1970 to 1994, the intake of wheat flour increased by 35%, corn flour by 79%, grain mixtures by 115%, high carbohydrate snacks by 200%, and ready-to-eat cereals by 41%. Today, 78% of total dietary carbohydrate in the USA is in the form of grains, grain products and sweeteners.[74] Given this massive consumption of carbohydrates, one would expect Americans to be consuming whole grain foods as well as refined carbohydrates, but that is evidently not the case. A 1995 study found that dietary intake of whole grain foods averaged only one-half serving per person per day.[75] Interestingly, the Dietary Guidelines Advisory Committee noted the increased health benefits of substituting whole grain foods for refined grain foods, but did not understand itself why these benefits occurred. “Specific mechanism that explain the associations of higher intakes of whole grains with reductions of chronic disease risk are not defined well.” The Committee also noted that the “risk reduction in the cited studies [on coronary heart disease and hypertension] was associated with higher levels of whole grain intake and could not be explained by adjustments for fiber intake. This suggests that components of whole grain nutrients other than fiber help reduce risk for coronary heart disease.”[76]

The Committee fails to even mention that one likely reason for the beneficial effects of substituting whole grain foods for refined grain foods is a reduced glycemic index. This would, of course, help explain the decrease in risk of coronary heart disease and hypertension. That the Committee would fail to note this aspect of whole grain foods is peculiar, especially since the glycemic index has been around since 1981. Nonetheless, since the previous edition of the Dietary Guidelines did not even include advice about choosing whole grain foods over refined grain foods, the 5th Edition of the Dietary Guidelines should be seen as a significant step towards the recognition of the beneficial health effects of eating a low-glycemic index diet.

As for the Dietary Guidelines and Food Guide Pyramid recommending that Americans consume both whole grain foods and refined grain foods, that can be explained by the USDA’s process to determine the amounts of each food group to include in the food guide:

Step 5. Determination of the Numbers of Servings.

Determination of the amounts of each food group to include in the food guide was a two-phase process. The first focused on concerns about nutrient adequacy and the second focused on concerns about moderation. The first phase involved determining the number of servings from each nutrient-bearing food group and subgroup needed to meet the nutritional goals for protein, vitamins, and minerals.[77]

The second phase was supposed to focus on food components for which moderation was a concern. However, the USDA calculated the number of calories necessary to achieve the desired levels of the vitamins, minerals, and protein, subtracted that from the daily caloric targets, and allowed the remainder to be from fats and sugars. Thus, when the USDA was supposed to be concerned with moderation, it really was only concerned with moderation of fats and sugars, not moderation of refined carbohydrates.

The USDA’s process of first ensuring adequate protein, vitamin, and mineral intake and then worrying about moderation practically guarantees that enriched, highly refined grain foods will be over-emphasized in the Food Guide Pyramid. Interestingly, the problem of malnutrition from a lack of vitamins or minerals is no longer a pressing problem in America, nor was it when the Food Guide Pyramid and Dietary Guidelines were created. [78] Instead, the most pressing nutrition and health problem in America today would have to be either coronary heart disease or obesity, both of which are caused (at least in part) by nutritional problems that have nothing to do with vitamin and mineral intake. Thus, we see that the Food Guide Pyramid reflects, at least in part, outdated priorities that address problems long since overshadowed by larger, more serious problems. Should the USDA change its process of determining servings to reflect the fact that obesity and coronary heart disease are much worse problems in America today than vitamin and mineral deficiencies, the amount of whole grain foods recommended would probably be increased while decreasing the amount of vitamin-enriched highly-refined grain foods recommended.

B. The “Complex Carbohydrate” Controversy

As evidenced by the Advisory Committee notes, the federal government is aware that some forms of carbohydrates are healthier than others. In the past, many food scientists used to divide carbohydrates into two groups, simple carbohydrates and complex carbohydrates. The term “simple carbohydrate” was usually used to refer to foods that had a high sugar content and were thought to be unhealthy. The term “complex carbohydrate” was used to refer to those carbohydrates that didn’t contain very many sugars and were thought to be healthy. The USDA Food Guide Pyramid recommends breads, cereals, rice and pasta precisely because they “provide complex carbohydrates (starches), which are an important source of energy, especially in lowfat diets.”[79]

In an effort to aid consumers in choosing foods with healthy carbohydrates rather than simple carbohydrates, Congress passed the Nutrition Labeling and Education Act of 1990, which gave FDA the authority to require the now-ubiquitous food labels.[80] Congress originally required the following nutrients to be included in the food label: total fat saturated fat, cholesterol, sodium, total carbohydrates, complex carbohydrates, sugars, dietary fiber, and total protein.[81] The Act gave the Secretary of the DHHS the authority to require other nutritional values on the label, and also allowed the Secretary to remove the information required by Congress to be on the label if the Secretary determined that the information relating to that nutrient is not necessary to assist consumers in maintaining healthy dietary practices.[82] It was expected that the Secretary would provide definitions of the nutrients at its discretion, consistent with scientific principles.[83]

As alluded to earlier, the definition of complex carbohydrate heretofore had been rather vague and inconsistent. Therefore, the FDA had considerable trouble creating a suitable definition for “complex carbohydrate.” The FDA wanted to provide a definition that was consistent with the physiological effects attributed to complex carbohydrates in the various scientific reports, but none of the reports had provided a chemical definition of the term, and there was no commonly accepted chemical definition. Therefore, FDA tried to provide a chemical definition that would enable consumers to receive the health benefits commonly ascribed to “complex carbohydrates.”

FDA proposes that "complex carbohydrate" be defined as the sum of dextrins and starches. Thus, complex carbohydrate, as defined, includes those carbohydrate components that contain 10 or more saccharide units (exclusive of dietary fiber). FDA is aware that including dextrins within the definition of "complex carbohydrate" may result in the classification of certain components of a few nutritive sweeteners as complex carbohydrates. The agency, therefore, requests comments on this proposed definition and solicits suggestions for alternative definitions of "complex carbohydrate."[84]

The FDA received numerous complaints about this proposed definition. Many comments pointed to the lack of existing analytical methodology to support the proposed definition, raising concerns about the feasibility of compliance and the economic burden of developing methods and databases. Others criticized the FDA on the basis that the cutoff at 10 saccharide units was arbitrary, as there are no known nutritional or physiological differences, nor a methodological justification, to make a distinction between polysaccharides smaller than 10 saccharide units and those with 10 or more saccharide units. Listening to these complaints, FDA decided to abandon trying to define “complex carbohydrate,” and instead created a new category of “other carbohydrate.”[85] Replacing “complex carbohydrate,” “other carbohydrate” is defined as the difference between total carbohydrate and the sum of dietary fiber and sugars or, if sugar alcohol is declared, the difference between total carbohydrate and the sum of dietary fiber, sugars, and sugar alcohol.[86]

The FDA’s inability to define complex carbohydrate is significant, since numerous studies tout the myriad health benefits of complex carbohydrates and Americans, hearing about these studies, wish to obtain these health benefits. They read in the USDA Food Guide Pyramid brochure that they should replace fats and sugars with complex carbohydrates, yet when they go to the grocery store to purchase foods high in complex carbohydrates they run into a problem. Without having complex carbohydrate grams listed in the food label, consumers are unaware whether any given food will provide the health benefits of complex carbohydrates to them or not. Thus, even consumers wishing to increase their complex carbohydrate intake cannot, under current regulations, just by looking at the food label; instead, they must search the ingredients list for whole grain foods such as whole wheat, brown rice, and pearl barley. Given the emphasis on complex carbohydrates throughout the Food Guide Pyramid, forcing consumers to memorize key ingredients for what may amount to “complex carbohydrates” seems unnecessarily difficult.

Nonetheless, because the Secretary is given the discretion to determine whether inclusion of “complex carbohydrate” in the food label would assist consumer in maintaining healthy dietary practices, any legal challenge would almost certainly fail. Courts will defer to the Secretary’s expertise, and there is no litigation private parties can bring that would likely result in forcing the Secretary to create a definition of “complex carbohydrates.”

“Deference to an agency's technical expertise and experience is particularly warranted with respect to questions involving ... scientific matters." Friends of the Clearwater v. Dombeck, 222 F.3d 552, 556 (9th Cir.2000) (quoting United States v. Alpine Land and Reservoir Co., 887 F.2d 207, 213 (9th Cir.1989) ). "When specialists express conflicting views, an agency must have discretion to rely on the reasonable opinions of its own qualified experts even if, as an original matter, a court might find contrary views more persuasive." Greenpeace Action v. Franklin, 14 F.3d 1324, 1332 (9th Cir.1992) (quoting Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 378, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989) ). Once the court is satisfied that an agency's exercise of discretion is informed, the court must defer to that informed discretion. Id. [87]

Given the courts’ usual deference to agency decisions involving their particular expertise, and given that the FDA did take comments on the proposal and explained their rationale behind choosing to change “complex carbohydrate” to “other carbohydrate” in their final rule, any legal challenge to that action faces a stiff uphill battle.

C. The Definition of “Carbohydrate”

Not only did FDA have to try to define complex carbohydrate, it also had, as a preliminary matter, to define the term “carbohydrate” itself. To do this, FDA relied on the pioneering work of W.O. Atwater, who first described the process of determining carbohydrate content of foods by difference. The FDA adopted Atwater’s method, and currently defines total carbohydrate by difference:

Total carbohydrate content shall be calculated by subtraction of the sum of the crude protein, total fat, moisture, and ash from the total weight of the food. This calculation method is described in A. L. Merrill and B. K. Watt, "Energy Value of Foods--Basis and Derivation," USDA Handbook 74 (slightly revised 1973) pp. 2 and 3, which is incorporated by reference in accordance with 5 U.S.C. 552(a) and 1 CFR part 51 (the availability of this incorporation by reference is given in paragraph (c)(1)(i)(A) of this section).[88]

Thus, under current FDA definitions, anything that isn’t protein, fat, moisture or ash in a food, must be by definition carbohydrate. Again, the agency based this definition on accepted scientific principles, so its choice of definition is to be accorded great deference by the courts. This definition of carbohydrate worked well for quite a while, but because of the recent increased focus on reducing carbohydrate consumption (or high glycemic index carbohydrate consumption) the definition has been at the heart of some controversy.

The huge surge in popularity of low-carbohydrate diets such as the Atkins diet inevitably led to the development of new products intended to capitalize on the craze. Numerous companies began selling low-carbohydrate meal replacement bars, with the first big seller being Carbolite’s sugar-free chocolate bar. The chocolate bar, which was first produced in 1999, was so popular that orders went from 10,000 to 1 million in just two months. In 2000, Carbolite persuaded 7-Eleven to place the bar next to candy bars such as Snickers and Milky Way, and the bar became the chain’s number 3 seller. Dr. Atkins has his own nutritional line, Atkins Nutritionals, which sold about $35 million in 1999.[89] Indeed, seeing the potential for huge sales, numerous companies have gotten in the business of selling low-carbohydrate bars: Atkins Nutritionals, Carbolite, EAS, Country Life/Biochem, American Body Building, Next Proteins, Solid Protein, Optimum Nutrition, and various smaller companies. In short, by 2000, if you wanted to sell meal replacement bars, you needed a low-carbohydrate version, and every major company was either already selling them or looking to start selling them.

The problem was that these low-carbohydrate meal replacement bars weren’t actually low in carbohydrates. The companies all produced bars claiming to have a few grams of carbohydrate or less, and on the food label they would only list several grams of carbohydrate under the “Total Carbohydrate” heading. However, most of the products would have a small footnote under or near the food label explaining that the product contained a carbohydrate-substitute, usually glycerin, oligofructose, sugar alcohols, and/or malitol. For example, the wrapper for the Designer Whey Protein Bar (Chocolate Triplemint flavor) manufactured by Next Proteins lists only 7 grams of carbohydrates on the food label; under the food label is the statement: “This product contains glycerine. Glycerine is not a carbohydrate but has a caloric value of 4.32 calories per gram.”

Most labels did not include the grams of the carbohydrate-substitute on the wrapper, but very astute consumers could calculate the grams of the carbohydrate-substitute; to do so, one needed to add up the calories from the grams of fat, protein, and listed carbohydrate (of course, one would need to know that protein and carbohydrates have 4 calories per gram, and fat 9 calories per gram), and then subtract that number from the total calories listed a the top of the food label. That would give the calories from the carbohydrate substitute, and one could then find the number of grams by dividing the calories by the energy density of the substitute, usually around 4 calories per gram. In the Designer Whey example, the bar lists 7 grams of fat, 7 grams of carbohydrates, and 30 grams of protein. This would give us (7*9)+(7*4)+(30*4) or 211 calories accounted for in the food label. Since the top of the label gives total calories of 250, we know there are 250-211, or 39 calories from glycerine. Dividing the 39 calories by 4.32 calories per gram, we find that there are roughly 9 grams of glycerine in our protein bar.

Given the FDA’s definition of carbohydrate being everything that’s not protein, fat, ash, or moisture, one would have a hard time excluding glycerine from the carbohydrate content. The makers of the low-carbohydrate bars claimed, sometimes on the wrapper itself, that these carbohydrate substitutes did not count as carbohydrates because they didn’t raise blood glucose levels, which is what makes normal carbohydrates so unhealthy. Although there is little evidence in the public domain of the impact of these carbohydrate substitutes on blood glucose levels, given that the bars implementing these carbohydrate substitutes do have lower glycemic loads than other bars not using them, we must assume that these carbohydrate substitutes do, indeed, have a minimal impact on blood glucose levels. Nonetheless, the FDA definition of carbohydrates does not mention blood glucose levels, and so the stage was set for a showdown between the bar manufacturers and the FDA.

In mid-2000 the FDA issued warning letters to all the manufacturers of these low-carbohydrate bars, warning them that the products were misbranded under the Federal Food, Drug, and Cosmetic Act because the “low carbohydrate” claim was an unauthorized nutrient content claim, and because the glycerine was not included as a carbohydrate. As one FDA warning letter put it, “Glycerine is a carbohydrate. ... We advise that glycerine must be included in the value declared for “Total Carbohydrate.”[90]

The various low-carbohydrate bar manufacturers were clearly not happy with the FDA’s decision, but most of the large manufacturers have started complying with the decision. As with dietary supplements, smaller companies that have little to lose are refusing to update their labels, remaining in violation of the FDA’s decision. One such bar clearly in violation of the Federal Food, Drug, and Cosmetic Act as interpreted by the FDA is Biochem’s Ultimate Lo Carbohydrate Bar. As if the name wasn’t in violation of the law’s prohibition against unauthorized implied nutrient claims, the bar’s front says prominently “Only 2g Carbohydrates!” The food label on the back also lists only 2 grams of carbohydrates under “Total Carbohydrates,” and there is a small footnote noting that glycerine has a caloric value of 4.32 calories per gram. Because companies such as Biochem are small and only manufacture this particular bar, they will probably not comply with the FDA mandate until absolutely forced, creating unfair competition for the larger manufacturers that have already started to comply.

The Atkins Nutritionals website discusses their approach to complying with the FDA decision, which is pretty typical for most of the larger companies.[91] Atkins Nutritionals changed the front label to read that there are only 2 grams of Net Carbohydrates in this product. Let us use the Atkins Chocolate Mocha Crunch Advantage Bar as an example. On the back, the carbohydrate substitutes glycerine and malitol are included in the “Total Carbohydrates” portion, as well as fiber, for a total of 19 grams of carbohydrate in the food label. The Net Carbohydrates sidebar has the 11 grams of polydextrose/fiber subtracted, as well as the 4.5 grams of glycerine, leaving only 3.5 grams of Net Carbohydrates. Atkins says that these are the grams of carbohydrate which should be closely monitored, as they are the ones with a significant impact on blood glucose levels, spikes of which can lead to overeating and obesity.

Interestingly, all three of the diets profiled in this paper (the Atkins diet, the Zone diet, and the SugarBusters diet) agree that keeping blood glucose levels low and stable is an important component of weight loss. However, the FDA’s decision to include these carbohydrate substitutes as carbohydrates has made it more difficult for consumers to watch their blood glucose levels and to follow these diets. Nonetheless, the FDA’s definition of carbohydrate is based on scientific research performed by Atwater years ago, and was based on accepted science, so it seems virtually impossible for the companies to challenge the decision on administrative law grounds. The FDA could, of course, try to define carbohydrates differently in the future, so that those people wishing to watch their blood glucose levels (which includes diabetics) would have an easier time of it. One possible definition would be to provide a glycemic index cutoff, so that those ingredients which don’t increase blood glucose levels would not count towards the carbohydrate content. As FDA notes, its original proposal for defining complex carbohydrates, using saccharide chain length, wouldn’t work for defining carbohydrates that have an impact on blood glucose levels, as there seems to be no relation between saccharide chain length and blood glucose response levels.

Most European countries define carbohydrates differently than we do in North America. As already discussed, FDA mandates that carbohydrate be calculated as the difference between total grams of the food, and the sum of protein, fat, ash, and moisture. Canada utilizes a similar methodology. By contrast, European countries usually determine the available carbohydrate, by doing a direct analysis to find the amount of mono-, di-, and polysaccharides in a given food. The North American method results in an overstatement of available carbohydrate content because it includes non-carbohydrate compounds and non-available carbohydrates or fiber. As an example, the carbohydrate and energy content of a portion of spaghetti calculated directly is 50.6 grams of carbohydrate and 237 calories, but calculation by difference (the North American method) yields 64.6 grams of carbohydrate and 321 calories. This overstates available carbohydrates by almost 28%, and overstates caloric content by over 35%. Thus, descriptions of carbohydrate consumption in North America are probably overstating the available carbohydrate content, due to the FDA requiring carbohydrate calculation by difference.[92] Of course, determining available carbohydrate content by direct analysis would cost more money and require all new databases, so it seems unlikely that there will be large-scale support amongst food manufacturers for a change in definitions.

VI. Conclusions and Proposals for Change

A. Conclusion

In an era when every single major diet book advocates restricting some or most carbohydrates (even the Atkins diet allows vegetables once desired weight is attained), the response of the FDA, USDA and DHHS has been disappointing. Despite mainstream acceptance of the wisdom of choosing foods with low a low glycemic index, the Dietary Guidelines for Americans and Food Guide Pyramid continue to advocate 6 to 11 servings of grains each day, without distinguishing between healthy, low-glycemic index carbohydrates and unhealthy high-glycemic index refined carbohydrates. The FDA has not helped matters by clinging to an outdated definition of carbohydrates that overstates the available carbohydrate of any given food, nor by failing to provide an adequate definition of complex carbohydrates to enable Americans to pursue the benefits of increasing their intake of complex carbohydrates. As time goes by, and more and more consumers hear of successful weight loss on the Atkins diet, or the Zone or SugarBusters diet, the advice of the government agencies in charge of the public health will begin to seem antiquated and irrelevant. If the agencies continue to ignore the differences amongst carbohydrates, and continue to demonize fat without even criticizing over-consumption of carbohydrates, the agencies may lose legitimacy in the eyes of the public.

B. Proposals for Change

Going forward, the agencies can do a lot to improve their image in the public eye and simultaneously improve the health of the American public. First, the Food Guide Pyramid and Dietary Guidelines for Americans must be updated to reflect the current knowledge about refined carbohydrates such as white rice, potatoes, and white bread. The Harvard School of Public Health has proposed a new, healthier food pyramid that reflects the latest research on optimal diet.[93] The new pyramid emphasizes whole grain foods and plant oils, also recommending increased consumption of vegetables. Alcohol is recommended in moderation, and a calcium supplement can replace dairy foods on the pyramid. The very top of the pyramid is split between red meat and butter, and highly refined carbohydrates white rice, white bread, potatoes, pasta and sweets. No less of an institution than Harvard has accepted the scientific evidence that Americans are consuming too many refined carbohydrates. When will the USDA catch on and revise its Food Guide Pyramid?

Second, the Secretary of the Department of Health and Human Services should reconsider its decision to define carbohydrates by difference, possibly adopting the European approach to defining carbohydrates. Given what we know about blood glucose levels and fat storage, lumping together those nutrients that cause no rise in blood glucose levels (e.g., fiber), with those that cause rapid increases in blood glucose levels, simply makes no sense and only serves to annoy dieters and diabetics. Admittedly, this would change the food labels of thousands of foods, and cause companies to perform more testing of their products to determine the new carbohydrate content. While this may not be inexpensive, the benefits to diabetics and low-carbohydrate dieters would be immense, and American carbohydrate consumption would cease to be overstated. As to claims that this would confuse the American public, if the Atkins web page can explain why fiber doesn’t count as a carbohydrate, surely the USDA could do the same.

Finally, the FDA can also help Americans to get healthier. Although there are some general principles for determining whether a food has a high glycemic index or a low one, sometimes it is not easy to tell which is the healthy option. Potatoes, long thought to be a great source of starchy complex carbohydrates, are now on the list of bad, high-glycemic index foods, while sweet potatoes have a low glycemic index. White rice has a high glycemic index, while brown rice has a fairly low glycemic index. Bananas have a high glycemic index, whereas most legumes have a low one. It is difficult for consumers wishing to monitor their blood glucose levels to remember the data for all of the possible foods they may wish to eat.

FDA can assist these consumers by allowing voluntary declaration of the glycemic load of a food. The glycemic load is preferable to the glycemic index for these purposes because the glycemic load takes into account the carbohydrate, protein and fat content of a food. For example, carrots have a high glycemic index, but a rather low glycemic load, because while the saccharides digest rather quickly, the fiber in them slows down the digestion process such that blood glucose levels don’t rise very much. If the FDA would allow food manufacturers to include glycemic load voluntarily, consumers could choose those foods that have the least impact on blood glucose levels, thus helping them to eat healthier and possibly lose weight. This would allow the previously-called low-carbohydrate bars to differentiate themselves from sugar-laden high-protein bars, and allow consumers to comparison shop for bars based on glycemic load. The voluntary declaration of glycemic load would also greatly assist diabetics in choosing food choices that help them to treat diabetes through diet, keeping their blood glucose levels stable by carefully choosing foods with a low glycemic load.

Although the benefits of declaring the glycemic load of foods would be even greater if all foods were required to declare the glycemic load, this would require food manufacturers to create a new database of the blood glucose responses to each food they produce, which could prove costly. Nonetheless, the procedure for measuring glycemic load is not very complex nor costly, and books are available cheaply that give the glycemic index of common foods. Also, many common foods have their glycemic index and glycemic load in a database maintained by the University of Sydney’s GI website, available at www.glycemicindex.com. The database is searchable by food and/or glycemic index or glycemic load. If databases containing hundreds of foods are already in existence, it cannot cost that much to calculate the glycemic index or glycemic load of a given food. Food manufacturers that produce healthy, low-glycemic load foods would eagerly pay the costs of testing to declare their healthy glycemic load on their label, and other manufacturers may be shamed into trying to find ways to lower the glycemic load of their foods, which would obviously improve the overall health of Americans.

The inclusion (whether voluntary or mandatory) of the glycemic load on food labels would obviate the need for a definition of complex carbohydrates, since most of the benefits reported for complex carbohydrates are actually those benefits for all low-glycemic index foods. In theory, the government could define complex carbohydrates as foods with a glycemic load less than a certain cutoff, but if the glycemic load is on the food label then this would be superfluous. For those food manufacturers that choose not to put the glycemic load on their foods, consumers may simply assume that the glycemic load is rather high, or may choose foods that have the glycemic load listed. Therefore, the market would create pressure on manufacturers to list the glycemic load, and a definition for complex carbohydrates would be unnecessary.

C. The Future

The health of the country is at stake. Americans are getting fatter than ever, and yet the government seems to be ignoring information that virtually every nutritionist and American dieter “knows” to be true. If the government’s recommendations aren’t changed soon, people may wonder why they should listen to the government when it comes to nutrition at all. Fortunately, it’s not too late. If the government adopts the proposals outlined here, it would increase its perceived legitimacy as guardian of the public’s nutrition, and would be seen as finally responding to criticisms of the current system. It’s time for the government to make some changes regarding nutritional recommendations.

Appendix A: Current USDA Food Guide Pyramid

Appendix B: Harvard School of Public Health Recommended Food Pyramid [94]


[1] By Jeremy Berry, J.D., Harvard Law School, 2003

[2] See generally “Statistics Related to Overweight and Obesity,” National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (available at http://www.niddk.nih.gov/health/nutrit/pubs/statobes.htm#other) (2003).

[3] USDA’s Food Guide: Background and Development, p. 1 (1993), quoting W.O. Atwater, Principles of Nutrition and Nutritive Value of Food, U.S. Department of Agriculture, Farmers’ Bull. No. 142, p. 38 (1902).

[4] See generally, USDA’s Food Guide: Background and Development, p. 1 (1993) (discussing the publications of Caroline Hunt and other early nutritionists).

[5] See Hassle-Free Guide to a Better Diet, in Food, USDA/SEA (1979).

[6] See National Nutrition Montioring and Related Research Act of 1990, 7 U.S.C. 5341 (2003).

[7] See Dietary Guidelines for Americans, 1980 to 2000, Center for Nutrition Policy and Promotion, USDA, May 30, 2000 (available at http://www.usda.gov/cnpp/Pubs/DG2000/Dgover.PDF). The 2000 Dietary Guidelines for Americans now has 10 principles: 1) aim for a healthy weight; 2) be physically active each day; 3) let the Pyramid guide your food choices; 4) choose a variety of grains daily, especially whole grains; 5) choose a variety of fruits and vegetables daily; 6) keep food safe to eat; 7) choose a diet that is low in saturated fat and cholesterol, and moderate in total fat; 8) choose beverages and foods to moderate your intake of sugar; 9) choose and prepare foods with less salt; and 10) if you drink alcoholic beverages, do so in moderation.

[8] For pictorial depiction of USDA’s Food Guide Pyramid, see Appendix A.

[9] See USDA Food Guide Pyramid, 1992 (revised 1996), p. 3 (available at http://www.usda.gov/cnpp/pyrabklt.pdf).

[10] See “Dr. Atkins Is Getting Fat”, Susan Orenstein, at http://www.business20.com/articles/mag/print/0,1643,47902,00.html.

[11] See generally id.

[12] See generally, “What If It’s All Been a Big Fat Lie?” Gary Taubes, New York Times, Section 6, p. 22, Column 1 (July 7, 2002).

[13] See Orenstein, supra note 7.

[14] See Taubes, supra note 9.

[15] See Orenstein, supra note 7.

[16] See generally “Zone Summary” on Dr. Sears’ web site, at http://www.drsears.com/site/Tools/Summary/Summary.nsf.

[17] See “How Much Fat Can I Eat?” Natural Health, No. 3, Vol. 32, p. 64 (April 1, 2002) (comments of Dr. Sears in expert roundtable discussion of nutrition).

[18] See Orentstein, supra note 7.

[19] See “Losing Weight the Sugar Busters Way: It’s Not the Fat, It’s the Sugar,” Judith Weinraub, Washington Post, p. E01, November 11, 1998.

[20] See “Sugar Busters! Concept” at http://www.sugarbusters.com/filessb/concept.html.

[21] See Weinraub, supra note 16.

[22] See “It Really Has Been a Big Fat Lie,” Dr. Robert Atkins, at http://atkinscenter.com/Archive/2002/7/17-676178.html (stating “Sunday, July 7, 2002, was one of the most gratifying days of my life—and one that validated the controlled carbohydrate nutritional approach to weight management and good health.”).

[23] See Orenstein, supra note 7.

[24] See generally The Columbia University College of Physicians and Surgeons Complete Home Medical Guide, Chapter 5 “Essential Nutrients and Their Functions”, Tapley et al. (available online at http://cpmcnet.columbia.edu/texts/guide/hmg05_0002.html#5.2) (links to this chapter are given on the Federal government’s Food and Nutrition Information Center’s web site).

[25] For discussion of carbohydrates generally, see “Carbohydrates in the Diet,” Janice Hermann, Oklahoma Cooperative Extension Service, available at http://pearl.agcomm.okstate.edu/fci/health/t-3117.html, also linked to from the Food and Nutrition Information Center web site.

[26] Information on fats was taken from “A Consumer’s Guide to Fats,” by Eleanor Mayfield, and first appeared in the May 1994 FDA Consumer; it has since been updated in January of 1999 and the updated version is available online at http://www.pueblo.gsa.gov/cic_text/food/fatguide/fatguide.html.

[27] See id.

[28] For information on blood glucose and insulin levels after consumption of carbohydrates, see generally “The Glycemic Index: Physiological Mechanisms Relating to Obesity, Diabetes, and Cardiovascular Disease,” Journal of American Medical Association, Vol. 287(18), May 8, 2002 pp. 2414-2423.

[29] See “The effect of chain length on glucose absorption and the related metabolic response,” Wahlqvist et al, Am. J. Clin. Nutr. Vol. 31, p. 1998-2001, 1978.

[30] See “Postprandial glucose and insulin responses to meals containing different carbohydrates in normal and diabetic subjects,” Bantle et al, N. Engl. J. Med. Vol. 309, p. 7-12 (1983).

[31] See Ludwig, supra note 25.

[32] See id.

[33] See Sports Supplement Review, 4th Issue, Vince Andrich, p. 29.

[34] See Ludwig, supra note 25.

[35] See scientific studies supporting the Atkins diet, infra.

[36] See “Cracking the Fat Riddle: SHOULD YOU COUNT CALORIES OR CARBS? IS DIETARY FAT YOUR BIGGEST ENEMY? THE LATEST RESEARCH MAY SURPRISE YOU,” Time, J. Madeleine Nash, Sept. 2, 2002, p. 46.

[37] See generally id.

[38] See “How Much Fat Can I Eat,” Natural Health, Section 3, Vol. 32, p. 64 (April 1, 2002) (expert panel of four dietary experts, one of which advocating the low-fat Ornish diet, and one advocating the low-carbohydrate Atkins diet, coming to these points of agreement).

[39] See generally “The Heart Healthy Lifestyle Program,” Dr. Dean Ornish, transcript of comments at Healthy Heart Cyber Conference, 2/19/03 available at http://my.webmd.com/content/article/61/67423.htm.

[40] See generally “How Much Fat Can I Eat,” supra note 35. These are the percentage macronutrient breakdowns recommended by the experts advocating the Ornish and Atkins diets, respectively.

[41] “Fed Up With Diets,” Dr. Dean Ornish, available at http://my.webmd.com/content/article/62/71792.htm.

[42] See generally “Intensive Lifestyle Changes for Reversal of Coronary Heart Disease,” Ornish et al., Jour. Am. Med. Ass. Vol. 280(23):2001-2007 (1998).

[43] See “Changes in plasma lipids and other cardiovascular risk factors during 3 energy-restricted diets differing in total fat and fatty acid composition,” Noakes and Clifton, American Journal of Clinical Nutrition, 71(3):706-12, 2000.

[44] “The effects of variations in carbohydrate, protein, and fat content of the diet upon weight loss, blood values, and nutrient intake of adult obese women,” Alford et al., Journal of the American Dietetic Association. 90(4):534-40, 1990 Apr.

[45] “Weight-loss with low or high carbohydrate diet?” Golay et al, International Journal of Obesity & Related Metabolic Disorders: Journal of the International Association for the Study of Obesity. 20(12):1067-72, 1996 Dec.

[46] See “New Research on Atkins Diet Challenges 30 years of Nutritional Dogma,” Daniel Haney, Associated Press, available at http://atkinscenter.com/Archive/2003/3/12-103820.html.

[47] See “How Much Fat Can I Eat?”, supra note 37.

[48] See “It’s Time to Build An Improved Food Pyramid,” Atkins, available at http://atkinscenter.com/Archive/2003/2/10-425583.html. For purposes of this information, the usual convention of recording individuals with a body mass index (BMI) between 25 and 29 as overweight, and individuals with a BMI 30 or over as being obese.

[49] See “How Much Fat Can I Eat,” supra note 37.

[50] See “Very-low-carbohydrate weight-loss diets revisitied,” Volek and Westman, Cleveland Clinic Journal of Medicine, Vol. 69(11), p. 849, at 858 (Nov. 2002).

[51] See Haney, supra note 43.

[52] See “Greater dietary intake of simple carbohydrate is associated with lower concentrations of high-density-lipoprotein cholesterol in hypercholesterolemic children,” TJ Starc et al., American Journal of Clinical Nutrition, Vol 67, 1147-1154 (1998).

[53] See Parks, E. J. & Hellerstein, M. K. “Carbohydrate-induced hypertriacylglycerolemia: historical perspective and review of biological mechanisms,” Am. J. Clin. Nutr. 71: 412-433 (2000).

[54] See Dreon, D. M., Fernstrom, H. A., Miller, B. & Krauss, R. M. (1994) Low-density lipoprotein subclass patterns and lipoprotein response to a reducedfat diet in men. FASEB J. 8: 121-126; see also Dreon, D. M., Fernstrom, H. A., Williams, P. T. & Krauss, R. M. (1999) A very low-fat diet is not associated with improved lipoprotein profiles in men with a predominance of large, low-density lipoproteins. Am. J. Clin. Nutr. 69: 411-418.

[55] See “A ketogenic diet favorably affects serum biomarkers for cardiovascular disease in normal-weight men,” Mathew Sharman et al., J. Nutr. 132: 1879-1885, July 2002.

[56] See “Body composition and hormonal responses to a carbohydrate-restricted diet,” JS Volek et al., Metabolism: Clinical & Experimental. 51(7):864-70, 2002 Jul.

[57] See “Increased Dietary Protein Modifies Glucose and Insulin Homeostasis in Adult Women during Weight Loss,” Donald K. Layman et al., J Nutr 2003 Feb;133(2):405-10.

[58] See “How Much Fat Can I Eat,” supra note 37 (comments and diet prescription of Dr. Sears, creator of the Zone diet).

[59] See Layman, supra note 54 (finding insulin levels of 207 ± 21 pmol/L for the hig-carbohydrate group and 75 ± 18 pmol/L for the Zone group).

[60] See id.

[61] “A Reduced Ratio of Dietary Carbohydrate to Protein Improves Body Composition and Blood Lipid Profiles during Weight Loss in Adult Women,” Layman et al., J. Nutr. 133:411-417, February 2003.

[62] See “Randomized trial on protein vs carbohydrate in ad libitum fat reduced diet for the treatment of obesity,” A.R. Skov et al., International Journal of Obesity & Related Metabolic Disorders: Journal of the International Association for the Study of Obesity. 23(5):528-36, 1999 May.

[63] See “Effect of a low-glycaemic index--low-fat--high protein diet on the atherogenic metabolic risk profile of abdominally obese men,” JG Dumesnil et al., British Journal of Nutrition. 86(5):557-68, 2001 Nov.

[64] See “A Low–Glycemic Index Diet in the Treatment of Pediatric Obesity,” L.E. Spieth et al., Archives of Pediatrics & Adolescent Medicine. 154(9):947-51, 2000 Sep.

[65] “Glycaemic index as a determinant of serum HDL-cholesterol concentration,” G. Frost et al., Lancet. 353(9158):1045-8, 1999 Mar 27.

[66] “A prospective study of dietary glycemic load, carbohydrate intake, and risk of coronary heart disease in US women,” S Liu et al., American Journal of Clinical Nutrition. 71(6):1455-61, 2000 Jun.

[67] For scientific studies relating benefits of low-glycemic index foods, see, e.g., Jenkins, DJ, Wolever, TM, Taylor, RH, Barker, H, Fielden, H, Baldwin, JM, Bowling, AC, Newman, HC, Jenkins, AL, Goff, DV. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981; 34 : 362-366; Jenkins, DJ. Wholemeal versus wholegrain breads: proportion of whole or cracked grain and the glycaemic response. BMJ 1988; 297 : 958-960; Jenkins, DJ, Wolever, TM, Jenkins, AL. Starchy foods and glycemic index. Diabetes Care 1988; 11 : 149-159; Wolever, TM. The glycemic index. World Rev Nutr Diet 1990; 62 : 120-185; Slabber, M, Barnard, HC, Kuyl, JM, Dannhauser, A, Schall, R. Effects of a low-insulin-response, energy-restricted diet on weight loss and plasma insulin concentrations in hyperinsulinemic obese females. Am J Clin Nutr 1994; 60 : 48-53; Leathwood, P, Pollet, P. Effects of slow release carbohydrates in the form of bean flakes on the evolution of hunger and satiety in man. Appetite 1988; 10 : 1-11; Spitzer, L, Rodin, J. Effects of fructose and glucose preloads on subsequent food intake. Appetite 1987; 8 : 135-145; Holt, SH, Miller, JB. Increased insulin responses to ingested foods are associated with lessened satiety. Appetite 1995; 24 : 43-54; Holt, SH, Miller, JB. Particle size, satiety and the glycaemic response. Eur J Clin Nutr 1994; 48 : 496-502; Holt, S, Brand, J, Soveny, C, Hansky, J. Relationship of satiety to postprandial glycaemic, insulin and cholecystokinin responses. Appetite 1992; 18 : 129-141.

[68] See Jenkins, DJ, Wolever, TM, Taylor, RH, Barker, H, Fielden, H, Baldwin, JM, Bowling, AC, Newman, HC, Jenkins, AL, Goff, DV. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981; 34 : 362-366.

[69] See USDA’s Food Guide: Background and Development (1993) (available at http://www.usda.gov/cnpp/Pubs/Pyramid/FoodGuideDevt.pdf).

[70] See USDA Food Guide Pyramid, 1992 (revised 1996) (available at http://www.usda.gov/cnpp/pyrabklt.pdf).

[71] See. Id., p. 19.

[72] See “Are All Food Pyramid Created Equal,” Alyson Escobar, Center for Nutrition Policy and Promotion, 1999 Nutrition Insights, Vol. 12, No. 3, p. 75 (1999).

[73] Dietary Guidelines for Americans, 5th Edition, US Department of Agriculture and US Department of Health and Human Services, Home and Garden Bulletin No. 232, p. 20 (2000).

[74] See “Should obese patients be counselled to follow a low-glycaemic index diet? Yes,” D.B. Pawlak et al., Obesity Reviews, Vol. 3(4), p. 235 (Nov. 2002).

[75] See Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2000 Dietary Advisory Committee, Feb. 2, 2000 p. 37 (available at http://www.usda.gov/cnpp/Pubs/DG2000/Full%20Report.pdf) (citing Albertson and Tobelmann, 1995).

[76] See id., p. 38 (citations omitted).

[77] See USDA’s Food Guide: Background and Development, supra note 1, at 12.

[78] See Emily J. Schaffer, “Is the Fox Guarding the Henhouse? Who Makes the Rules in American Nutrition Policy?” 57 Food & Drug L.J. 371 at 387.

[79] See USDA Food Guide Pyramid, supra note 67, at p. 19.

[80] See Nutrition Labeling and Education Act of 1990, PLAINTIFF 101-535, Nov. 8, 1990, 104 Stat 2353.

[81] See 21 USCA 343(q)(1)(D) (2003).

[82] See 21 USCA 343(q)(2)) (2003).

[83] See H.R. Rep. No. 101-538, 1990 U.S.C.C.A.N. 3336, 3343.

[84] Food Labeling; Mandatory Status of Nutrition Labeling and Nutrient Content Revision, 55 Fed. Reg. 29487-01, 29497 (July 19, 1990).

[85] See Food Labeling: Mandatory Status of Nutrition Labeling and Nutrient Content Revision, Format for Nutrition Label, 58 Fed. Reg. 2079, at 2100 (Jan. 6, 1993).

[86] See 21 C.F.R. 101.9(c)(6)(iv) (2003).

[87] Harlan Land Co. v. U.S. Dept. of Agr. 186 F.Supp.2d 1076, 1084 E.D.Cal.,2001.

[88] 21 C.F.R. 101.9(c)(6) (2003).

[89] See Orenstein, supra note 7.

[90] See John B. Foret of FDA’s Center for Food Safety and Applied Nutrition, Warning Letter to David McCabe of Worldwide Sports Nutritional Supplements, Inc. (July 18, 2000) (available at http://www.fda.gov/foi/warning_letters/m4025n.pdf).

[91] See “A Message from Atkins: All Carbohydrates Are Not Equal,” October 2001 (available at http://atkinscenter.com/Archive/2001/12/26-133646.html).

[92] See Alison M. Stephen, “Carbohydrate Intake in Canada: What We Know and What We Don’t,” Carbohydrate News 2002, p. 2.

[93] For pictorial depiction of Harvard School of Public Health’s Food Pyramid, see Appendix B (also available online at http://www.hsph.harvard.edu/nutritionsource/pyramids.html).

[94] Available online at http://www.hsph.harvard.edu/nutritionsource/pyramids.html.