Good Calories, Bad Calories Part 12

Another issue that complicates this issue of calorie versus carbohydrate restriction is that the effect of weight-loss diets changes over time. The modest benefits of semi-starvation slowly diminish with time, as the calorie restriction induces a compensatory inhibition of energy expenditure. Moreover, much of the initial weight loss comes from losing water, not fat (see Chapter 8). Because of this "tendency to retain water on a carbohydrate diet and to give it out on a rich fat diet," as Dunlop described it, restricting carbohydrate calories specifical y wil induce a more dramatic and immediate loss of water.

Testing diets for only a few weeks wil demonstrate that carbohydrate-restricted diets induce weight loss at a greater rate than calorie-or fat-restricted diets, but whether they induce fat loss at a greater rate is a different question. "Changes in body-weight are to be taken, therefore, as of significance only when the experiment continues for a period of several weeks," as Francis Benedict cautioned in 1910. "Certainly, for short experiments, body-weight is for the most part whol y without significance."

For this reason, the first meaningful report on the efficacy of carbohydrate restriction for weight loss was one published in 1936, by Per Hanssen of the Steno Memorial Hospital in Copenhagen. Hanssen reported treating twenty-one obese patients over two years with an 1,850-calorie diet that contained only 450 calories of carbohydrates, or a little less than 25 percent. Nearly 60 percent of the calories came from fat: 65 grams of cream, 65 grams of b.u.t.ter, and 25 grams of olive oil every day, along with two eggs, cheese, and a liberal portion of meat or fish. Some of his patients, Hanssen reported, were so fat initial y that they "could scarcely move when they arrived at the hospital, and were unable to work." On the diet from one to four months, the patients lost an average of two pounds a week. "During the stay in the hospital the patients never felt hungry," he reported. "The fatigue, a prominent and disturbing symptom, improved often very rapidly, and before the occurrence of any considerable reduction in weight." Hanssen compared his results with those reported five years earlier by physicians at the nearby University Clinic using a diet consisting of half the calories but twice the proportion of carbohydrates (over 50 percent). "At Steno Memorial Hospital," Hanssen noted, "a diet of 1,850 calories wil reduce weight as quickly as a diet of 950 calories at the University Clinic of Copenhagen."

If obese individuals can lose weight and keep it off, without hunger, on a diet of 1,850 calories, it's a reasonable a.s.sumption that they wil find it easier to sustain such a diet than one that al ows only 950 calories, or even less, and a.s.sumes, as Evans put it, that the obese "should be hungry most of the time as this is normal." A diet "relatively poor in carbohydrates," Hanssen suggested, might "not be so difficult to adhere to as the diets commonly used."

What additional y complicates any a.s.sessment of the role of carbohydrate restriction in reducing diets is that the composition of a diet is never quite so simple as merely being high or low in carbohydrates or refined carbohydrates. Proteins, fats, and calories a.s.sume different roles depending on the diet.

Also, carbohydrates in these diets can be restricted, but the standard thinking is that they have to remain sufficiently high so that the brain and central nervous system derive al their necessary fuel from this dietary source of glucose. Nutritionists wil often insist that 130 grams a day of carbohydrates are the minimal safe amount in a human diet.

Though glucose is a primary fuel for the brain, it is not, however, the only fuel, and dietary carbohydrates are not the only source of that glucose. If the diet includes less than 130 grams of carbohydrates, the liver increases its synthesis of molecules cal ed ketone bodies, and these supply the necessary fuel for the brain and central nervous system. If the diet includes no carbohydrates at al , ketone bodies supply three-quarters of the energy to the brain.

The rest comes from glucose synthesized from the amino acids in protein, either from the diet or from the breakdown of muscle, and from a compound cal ed glycerol that is released when triglycerides in the fat tissue are broken down into their component fatty acids. In these cases, the body is technical y in a state cal ed ketosis, and the diet is often referred to as a ketogenic diet. Whether the diet is ketogenic or anti-ketogenic-representing a difference of a few tens of grams of carbohydrates each day-might influence the response to the diet, complicating the question of whether carbohydrates are responsible for some effect or whether there is another explanation. (Ketosis is often incorrectly described by nutritionists as "pathological." This confuses ketosis with the ketoacidosis of uncontrol ed diabetes. The former is a normal condition; the latter is not. The ketone-body level in diabetic ketoacidosis typical y exceeds 200 mg/dl, compared with the 5 mg/dl ketone levels that are typical y experienced after an overnight fast-twelve hours after dinner and before eating breakfast-and the 520 mg/dl ketone levels of a severely carbohydrate-restricted diet with only 510 percent carbohydrates.) For fifty years after Wil iam Banting publicized Wil iam Harvey's prescription for a carbohydrate-restricted diet in 1863, the primary clinical disagreements were on the role of fat in the diet. Banting's original prescription was a high-fat diet, but then it was modified by Harvey himself and by the German clinicians Felix von Niemeyer and Max Oertel into lower-fat, higher-protein versions, and by Wilhelm Ebstein into a version featuring stil more fat. "The fat of ham, pork or lamb is not only harmless but useful," Ebstein wrote.

The notion of a carbohydrate-restricted diet based exclusively on fatty meat was publicized after World War I by the Harvard anthropologist-turned-Arctic-explorer Vilhjalmur Stefansson, who was concerned with the overal healthfulness of the diet, rather than its potential for weight loss. Stefansson had spent a decade eating nothing but meat among the Inuit of northern Canada and Alaska. The Inuit, he insisted, as wel as the visiting explorers and traders who lived on this diet, were among the healthiest if not the most vigorous populations imaginable.

Among the tribes with whom Stefansson lived and traveled, the diet was primarily caribou meat, "with perhaps 30 percent fish, 10 percent seal meat, and 5 or 10 percent made up of polar bear, rabbits, birds and eggs." The Inuit considered vegetables and fruit "not proper human food," Stefansson wrote, but they occasional y ate the roots of the knotweed plant in times of dire necessity.

The Inuit paid little attention to the plants in their environment "because they added nothing to their food supply," noted the Canadian anthropologist Diamond Jenness, who spent the years 191416 living in the Coronation Gulf region of Canada's Arctic coast. Jenness described their typical diet during one three-month stretch as "no fruit, no vegetables; morning and night nothing but seal meat washed down with ice-cold water or hot broth." (The ability to thrive on such a vegetable-and fruit-free diet was also noted by the lawyer and abolitionist Richard Henry Dana, Jr., in his 1840 memoirs of life on a sailing s.h.i.+p, Two Years Before the Mast. For sixteen months, Dana wrote, "we lived upon almost nothing but fresh beef; fried beefsteaks, three times a day...[in] perfect health, and without ailings and failings."

None of Stefansson's observations would have been controversial had not the conventional wisdom at the time been-as it is stil -that a varied diet is essential for good health. A healthy diet, it is said, must contain protein, fats, and carbohydrates, the latter because of the misconception that the brain and central nervous system require dietary glucose to function, and the debatable a.s.sumption that fresh vegetables and fruit, which contain carbohydrates, are essential to prevent deficiency diseases.

Because it is stil common to a.s.sume that a meat-rich, plant-poor diet wil result in nutritional deficiencies, it's worth pausing to investigate this issue.

The a.s.sumption dates to the early decades of the twentieth century, the golden era of research on vitamins and vitamin-deficiency diseases, as one disease after another-scurvy, pel agra, beriberi, rickets, anemia-was found to be caused by a lack of essential vitamins and minerals. This was The Newer Knowledge of Nutrition, as it was cal ed by the Johns Hopkins nutritionist Elmer McCol um; it dictated that the only way to ensure al the essential elements for health was to eat as many types of foods as possible, and nutritionists stil hold by this logic today. "A safe rule of thumb," as it was recently described, "is that the more components there are in a dietary, the greater the probability of balanced intake."

This philosophy, however, was based almost exclusively on studies of deficiency diseases, al of which were induced by diets high in refined carbohydrates and low in meat, fish, eggs, and dairy products. When the Scottish naval surgeon James Lind demonstrated in 1753 that scurvy could be prevented and cured by the consumption of citrus juice, for example, he did so with British sailors who had been eating the typical naval fare "of water gruel sweetened with sugar in the morning, fresh mutton broth, light puddings, boiled biscuit with sugar, barley and raisins, rice and currants." Pel agra was a.s.sociated almost exclusively with corn-rich diets, and beriberi with the eating of white rice rather than brown. When beriberi broke out in the j.a.panese navy in the late 1870s, it was only after the naval fare had been switched from vegetables and fish to vegetables, fish, and white polished rice.

The outbreak was brought under control by replacing the white rice with barley and adding meat and evaporated milk. Pel agra, too, could be cured or ameliorated, as Carl Voegtlin demonstrated in 1914, by adding fresh meat, milk, and eggs to a pel agra-causing diet, which in Voegtlin's experiments const.i.tuted primarily wheat bread, cabbage, cornmeal and corn syrup, turnips, potatoes, and sugar. Nutritionists working with lab animals also found that they could induce deficiency diseases by feeding diets rich in refined grains and sugar. Guinea pigs were given scurvy in a series of laboratory experiments in the 1940s when they were fed diets of mostly crushed barley and chickpeas.

This research informed the conventional wisdom of the era that fresh meat, milk, and eggs were what the Scottish nutritionist Robert McCarrison cal ed "protective foods" (which is how they were known before Ancel Keys and his contemporaries established them as the fat-rich agents of coronary disease), but it also bolstered the logic that a "balanced" diet, with copious vegetables, fruits, and grains, was necessary for health. Because diets of mostly grains and starches, or diets of refined grains, fish, and vegetables, such as the j.a.panese sailors consumed, might be deficient in a vitamin or vitamins essential for health, nutritionists considered it a reasonable a.s.sumption that this might be true of any such "unbalanced" diets, including those that were made up exclusively of animal products.

What the nutritionists of the 1920s and 1930s didn't then know is that animal foods contain al of the essential amino acids (the basic structural building blocks of proteins), and they do so in the ratios that maximize their utility to humans.*94 They also contain twelve of the thirteen essential vitamins in large quant.i.ties. Meat is a particularly concentrated source of vitamins A, E, and the entire complex of B vitamins. Vitamins D and B12 are found only in animal products (although we can usual y get sufficient vitamin D from the effect of sunlight on our skin).

The thirteenth vitamin, vitamin C, as...o...b..c acid, has long been the point of contention. It is contained in animal foods in such smal quant.i.ties that nutritionists have considered it insufficient and the question is whether this quant.i.ty is indeed sufficient for good health. Once James Lind demonstrated that scurvy could be prevented and cured by eating fresh fruits and vegetables, nutritionists a.s.sumed that these foods are an absolutely essential dietary source of vitamin C. What had been demonstrated, they wil say, is that scurvy is "a dietary deficiency resulting from lack of fresh fruit and vegetables." To be technical y accurate, however, Lind and the nutritionists who fol owed him in the study of scurvy demonstrated only that the disease is a dietary deficiency that can be cured by the addition of fresh fruits and vegetables. As a matter of logic, though, this doesn't necessarily imply that the lack of vitamin C is caused by the lack of fresh fruits and vegetables. Scurvy can be ameliorated by adding these to the diet, but the original lack of vitamin C might be caused by other factors. In fact, given that the Inuit and those Westerners living on the Inuit's vegetable-and fruit-free diet never suffered from scurvy, as Stefansson observed, then other factors must be involved. This suggested another way of defining a balanced diet. It's possible that eating easily digestible carbohydrates and sugars increases our need for vitamins that we would otherwise derive from animal products in sufficient quant.i.ties.

This was the issue that Stefansson was raising in the early 1920s. If the Inuit thrived in the harshest of environments without eating carbohydrates and whatever nutrients exist in fruits and vegetables, they, by definition, were consuming a balanced, healthy diet. If they did so solely because they had become evolutionarily adapted to such a diet, which was a typical rejoinder to Stefansson's argument, then how can one explain those traders and explorers, like Stefansson himself and the members of his expeditions, who also lived happily and healthful y for years at a time on this diet?

Nutritionists of the era a.s.sumed that al -meat diets were unhealthy because (1) excessive meat consumption was al eged to raise blood pressure and cause gout; (2) the monotony of eating only meat-or any other single food-was said to induce a physical sense of revulsion; (3) the absence of fresh fruit and vegetables in these diets would cause scurvy and other deficiency diseases, and (4) protein-rich diets were thought to induce chronic kidney damage, a belief based largely on early research by Louis Newburgh.

None of these claims were based on compel ing evidence. Newburgh, for instance, had based his conclusions largely on experiments in which he fed excessive quant.i.ties of soybean, egg whites, and beef protein to rabbits, which, as critics would later observe, happen to be herbivores. Their natural diet is buds and bark, not their fel ow animals, and so there was little scientific value in force-feeding them meat or animal protein. Nonetheless, the dangers of an al -meat diet were considered sufficiently likely that even Francis Benedict, as Stefansson told it, claimed that it was "easier to believe" that Stefansson and al the various members of his expeditions "were lying, than to concede that [they] had remained in good health for several years on an exclusive meat regimen."

In the winter of 1928, Stefansson and Karsten Anderson, a thirty-eight-year-old Danish explorer, became the subjects in a yearlong experiment that was intended to settle the meat-diet controversy. The experiment was planned and supervised by a committee of a dozen respected nutritionists, anthropologists, and physicians.*95 Eugene Du Bois and ten of his col eagues from Cornel and the Russel Sage Inst.i.tute of Pathology would oversee the day-to-day details of the experiment.

For three weeks, Stefansson and Anderson were fed a typical mixed diet of fruits, cereals, vegetables, and meat while being subjected to a battery of tests and examinations. Then they began living exclusively on meat, at which point they moved into Bel evue Hospital in New York and were put under twenty-four-hour observation. Stefansson remained at Bel evue for three weeks, Anderson for thirteen weeks. After they were released, they continued to eat only meat for the remainder of one year. If they cheated on the diet, according to Du Bois, the experimenters would know it from regular examinations of Stefansson's and Anderson's urine. "In every individual specimen of urine which was tested during the intervals when they were living at home," Du Bois wrote, "acetone [ketone] bodies were present in amounts so constant that fluctuations in the carbohydrate intake were practical y ruled out."

The experimental diet included many types of meat. To test the argument that the vitamins necessary in such a diet to avoid scurvy and remain healthy could be obtained only by eating raw meat, as was incorrectly a.s.sumed to be the practice of the Inuit, al of the meat was cooked. (In fact, the Inuit only occasional y ate raw meat.) Stefansson and Anderson each consumed an average of almost two pounds of meat per day, or twenty-six hundred calories: 79 percent from fat, 19 percent protein, and roughly 2 percent from carbohydrates (a maximum of fifty calories a day), which came from glycogen contained in the muscle meat. (Glycogen is the compound that stores glucose, a carbohydrate, in the liver and the muscle.) "The only dramatic part of the study was the surprisingly undramatic nature of the findings," wrote Du Bois, when he later summarized the results. "Both men were in good physical condition at the end of the observation," he reported in 1930, in one of the nine articles he and his col eagues published on the study. "There was no subjective or objective evidence of any loss of physical or mental vigor." Stefansson lost six pounds over the course of the year, and Anderson three, even though "the men led somewhat sedentary lives." Anderson's blood pressure dropped from 140/80 to 120/80; Stefansson's remained low (105/70) throughout. The researchers detected no evidence of kidney damage or diminished function, and "vitamin deficiencies did not appear." Nor did mineral deficiencies, although the diet contained only a quarter of the calcium usual y found in mixed diets, and the acidic nature of a meat-rich diet was supposed to increase calcium excretion and so deplete the body of calcium. Among the minor health issues reported by Du Bois and his col eagues was the observation that Stefansson began the experiment with mild gingivitis (inflammation of the gums), but this "cleared up entirely, after the meat diet was taken."

When Stefansson published Not by Bread Alone, a popular treatise on fat-and-protein diets, in 1946, a New York Times reviewer wrote, "Mr.

Stefansson makes the mixed-diet technicians and the nuts-and-fruits addicts look terribly sil y." Du Bois, who supervised the experiments, wrote an introduction to Stefansson's book. After Stefansson and Anderson were living exclusively on meat, he said, "a great many dire predictions and bril iant theories faded into nothingness." A diet that should have left Stefansson and Anderson deathly il from scurvy had left them as healthy as or healthier than the balanced diet they had been eating in the years immediately preceding the study. "Quite evidently we must revise some of our text book statements,"

Du Bois concluded.

The textbook statements on vitamins would go unrevised, however, despite laboratory research that has confirmed Stefansson's speculations.

Nutritionists would establish by the late 1930s that B vitamins are depleted from the body by the consumption of carbohydrates. "There is an increased need for these vitamins when more carbohydrate in the diet is consumed," as Theodore Van Ital ie of Columbia University testified to McGovern's Select Committee in 1973. A similar argument can now be made for vitamin C. Type 2 diabetics have roughly 30 percent lower levels of vitamin C in their circulation than do nondiabetics. Metabolic syndrome is also a.s.sociated with "significantly" reduced levels of circulating vitamin C, which suggests that vitamin-C deficiency might be another disorder of civilization. One explanation for these observations-described in 1997 by the nutritionists Julie Wil and Tim Byers, of the Centers for Disease Control and the University of Colorado respectively, as both "biological y plausible and empirical y evident"-is that high blood sugar and/ or high levels of insulin work to increase the body's requirements for vitamin C.

The vitamin-C molecule is similar in configuration to glucose and other sugars in the body. It is shuttled from the bloodstream into the cel s by the same insulin-dependent transport system used by glucose. Glucose and vitamin C compete in this cellular-uptake process, like strangers trying to flag down the same taxicab simultaneously. Because glucose is greatly favored in the contest, the uptake of vitamin C by cel s is "global y inhibited" when blood-sugar levels are elevated. In effect, glucose regulates how much vitamin C is taken up by the cel s, according to the University of Ma.s.sachusetts nutritionist John Cunningham. If we increase blood-sugar levels, the cel ular uptake of vitamin C wil drop accordingly. Glucose also impairs the reabsorption of vitamin C by the kidney, and so, the higher the blood sugar, the more vitamin C wil be lost in the urine. Infusing insulin into experimental subjects has been shown to cause a "marked fal " in vitamin-C levels in the circulation.

In other words, there is significant reason to believe that the key factor determining the level of vitamin C in our cel s and tissues is not how much or little we happen to be consuming in our diet, but whether the starches and refined carbohydrates in our diet serve to flush vitamin C out of our system, while simultaneously inhibiting the use of what vitamin C we do have. We might get scurvy because we don't faithful y eat our fruits and vegetables, but it's not the absence of fruits and vegetables that causes the scurvy; it's the presence of the refined carbohydrates.*96 This hypothesis has not been proven, but, as Wil and Byers suggested, it is both biological y plausible and empirical y evident.

When we discuss the long-term effects of diets that might reverse or prevent obesity, we must not let our preconceptions about the nature of a healthy diet bias the science and the interpretation of the evidence itself.

Chapter Twenty.

UNCONVENTIONAL DIETS.

Here was a treatment, that, in its encouragement to eat plentiful y, to the ful satisfaction of the appet.i.te, seemed to oppose not only the prevailing theory of obesity but, in addition, principles basic to the biological sciences and other sciences as wel . It produced a sense of puzzlement that was a mighty stimulant to thought on the matter.

ALFRED PENNINGTON, talking about a high-fat, high-protein diet, unrestricted in calories, in the American Journal of Digestive Diseases, 1954 Does it help people lose weight? Of course it does. If you cannot eat bread, bagels, cake, cookies, ice cream, candy, crackers, m.u.f.fins, sugary soft drinks, pasta, rice, most fruits and many vegetables, you wil almost certainly consume fewer calories. Any diet wil result in weight loss if it eliminates calories that previously were overconsumed.

JANE BRODY, talking about a high-fat, high-protein diet, unrestricted in calories, in the New York Times, 2002 A. J. LIEBLING, THE CELEBRATED AUTHOR of The New Yorker's "On Press" column, once wrote that he had enunciated a journalistic truth with such clarity that it was suitable for framing. "There are three kinds of writers of news in our generation," Liebling wrote. "In inverse order of worldly consideration, they are: 1. The reporter, who writes what he sees.

2. The interpretive reporter, who writes what he sees and what he construes to be its meaning.

3. The expert, who writes what he construes to be the meaning of what he hasn't seen.

"To combat an old human prejudice in favor of eyewitness testimony," Liebling wrote, "the expert must intimate that he has access to some occult source or science not available to either reporter or reader. He is the Priest of Eleusis, the man with the big picture.... Al is manifest to him, sincehis conclusions are not limited by his powers of observation."

Leibling was talking about journalism, but a similar ranking holds true in medicine. In fact, the medical experts have the further advantage that they can disseminate their opinions with considerably greater influence. They can make their case with the imprimatur of the inst.i.tutions that employ them-the American Medical a.s.sociation, for instance, or Harvard University. They can easily attract the media's attention. Physicians' case reports and the patients' anecdotal experience have a fundamental role in medicine, but if these conflict with what the experts believe to be true, the experts' opinions win out.

This conflict between expertise and observational evidence has had a significant influence in the science of obesity. Reliable eyewitness testimony has come only from those who have weight problems themselves, or the clinicians who regularly treat obese patients, and neither group has ever garnered much credibility in the field. (The very a.s.sumption that obesity is a psychological disorder implies that the obese cannot be trusted as reliable witnesses to their own condition.) But it is these individuals who have the firsthand experience. When Hilde Bruch reported in 1957 that a fine-boned girl in her teens, "literal y disappearing in mountains of fat," lost nearly fifty pounds over a single summer eating "three large portions of meat" a day, it was easier for the experts to ignore the testimony as a freakish phenomenon than to contemplate how such a thing was possible. But the process of discovery in science, as the philosopher of science Thomas Kuhn has put it, only begins with the awareness that nature has violated our expectations. Often it is the unconventional events-the anomalous data, as these are cal ed in science-that reveal the true nature of the universe.

In 1920, while Vilhjalmur Stefansson was just beginning his campaign to convince nutritionists that an al -meat diet was a uniquely healthy diet, it was already making the transition into a reducing diet courtesy of a New York internist named Blake Donaldson. Donaldson, as he wrote in his 1962 memoirs, began treating obese patients in 1919, when he worked with the cardiologist Robert Halsey, one of four founding officers of the American Heart a.s.sociation. After a year of futility in trying to reduce these patients ("fat cardiacs," he cal ed them) with semi-starvation diets, he spoke with the resident anthropologists at the American Museum of Natural History, who told him that prehistoric humans lived almost exclusively on "the fattest meat they could kil ," perhaps supplemented by roots and berries. This led Donaldson to conclude that fatty meat should be "the essential part of any reducing routine,"

and this is what he began prescribing to his obese patients. Through the 1920s, Donaldson honed his diet by trial and error, eventual y settling on a half-pound of fatty meat-three parts fat to one part lean by calories, the same proportion used in Stefansson's Bel evue experiment-for each of three meals a day. After cooking, this works out to six ounces of lean meat with two ounces of attached fat at each meal. Donaldson's diet prohibited al sugar, flour, alcohol, and starches, with the exception of a "hotel portion" once a day of raw fruit or a potato, which subst.i.tuted for the roots and berries that primitive man might have been eating as wel . Donaldson also prescribed a half-hour walk before breakfast.

Over the course of four decades, as Donaldson told it, he treated seventeen thousand patients for their weight problems. Most of them lost two to three pounds a week on his diet, without experiencing hunger. Donaldson claimed that the only patients who didn't lose weight on the diet were those who cheated, a common a.s.sumption that physicians also make about calorie-restricted diets. These patients had a "bread addiction," Donaldson wrote, in that they could no more tolerate living without their starches, flour, and sugar than could a smoker without cigarettes. As a result, he spent considerable effort trying to persuade his patients to break their habit. "Remember that grapefruit and al other raw fruit is starch. You can't have any," he would tel them.

"No breadstuff means any kind of bread.... They must go out of your life, now and forever." (His advice to diabetics was equal y frank: "You are out of your mind when you take insulin in order to eat Danish pastry.") Had Donaldson published details of his diet and its efficacy through the 1920s and 1930s, as Frank Evans did about his very low-calorie diet, he might have convinced mainstream investigators at least to consider the possibility that it is the quality of the nutrients in a diet and not the quant.i.ty of calories that causes obesity. As it is, he discussed his approach only at in-house conferences at New York Hospital. Among those who heard of his treatment, however, was Alfred Pennington, a local internist who tried the diet himself in 1944-and then began prescribing it to his patients.

After the war, Pennington worked for the industrial-medicine division of E. I. du Pont de Nemours & Company, and specifical y for George Gehrmann, the company's medical director and a pioneer in the field of occupational health.*97 Gehrmann founded and was the first president, from 1946 to 1949, of the American Academy of Occupational Medicine, an organization that has since merged and evolved into the American Col ege of Occupational and Environmental Medicine. By 1948, according to Gehrmann, DuPont as a corporation had become anxious about the apparent epidemic of heart disease in America. Just as Ancel Keys said he was prompted to pursue dietary means to prevent heart disease after perusing the obituaries, Gehrmann said he was prompted by the heart attack of a DuPont executive. Gehrmann decided to attack overweight and obesity, hoping heart-disease risk would diminish as a result.

"We had urged our overweight employees to cut down on the size of the portions they ate," Gehrmann said, "to count their calories, to limit the amounts of fats and carbohydrates in their meals, to get more exercise. None of those things had worked." These failures led Gehrmann and Pennington to test Donaldson's meat diet on overweight DuPont executives.

In June 1949, Pennington published an account of the DuPont experience in the journal Industrial Medicine. He had prescribed Donaldson's regimen to twenty executives, and they lost between nine and fifty-four pounds, averaging nearly two pounds a week. "Notable was a lack of hunger between meals," Pennington wrote, "increased physical energy and sense of wel being." Al of this seemed paradoxical: the DuPont executives lost weight on a diet that did not restrict calories. The subjects ate a minimum of twenty-four hundred calories every day, according to Pennington: eighteen ounces of lean meat and six ounces of fat divided over three meals. They averaged over three thousand calories. Carbohydrates were restricted in their diet-no more than eighty calories at each meal. "In a few cases," Pennington reported, "even this much carbohydrate prevented weight loss, though an ad-libitum [unrestricted] intake of protein and fat, more exclusively, was successful."*98 In June 1950, Holiday magazine cal ed Pennington's diet a "believe it or not diet development" and "an eat-al -you-want reducing diet." Two years later, Pennington discussed his diet at a smal obesity symposium hosted by the Harvard department of nutrition and chaired by Mark Hegsted. "Many of us feel that Dr. Pennington may be on the right track in the practical treatment of obesity," Hegsted said afterward. "His high percentage of favorable results is impressive and cal s for more extensive and for impartial comparative trials by others"-although, Hegsted concluded, "any method of [obesity] treatment other than caloric restriction stil requires study by al methods that can be brought to bear on the problem."

The Harvard symposium led to the publication of Pennington's presentation in The New England Journal of Medicine, and this, along with the Vogue article, prompted the competing medical journals to address it. In a scathing editorial cal ed "Freak Diets!" The Journal of the American Medical a.s.sociation (JAMA) took the position that calorie restriction was the only legitimate way to induce weight loss, and that what Hegsted had cal ed "impartial comparative trials by others" were not necessary. "The proposed high-fat diet wil probably add unduly to the patient's weight and thus, in addition to the other harmful effects of obesity, increase the hazard of atherosclerosis," wrote JAMA. In Britain, The Lancet wrote, "A low calorie intake is the best way to restore the composition of the body to normal, and this is most easily arranged by eliminating fat from the diet." If Pennington's diet worked, according to The Lancet, it did so only because "any monotonous diet leads to a loss of weight."

Clinicians-doctors who actual y treated obese patients-pushed back against the experts. After The Lancet's editorial, local clinicians wrote that the diet was successful in "a surprisingly large proportion of cases," as one Devonport physician put it. "Results so far certainly seem to support the work of Pennington which you rather lightly dismiss." "Pennington's idea of cutting out the carbohydrate but al owing plenty of protein and fat works excel ently...,"

wrote the prominent British endocrinologist Raymond Greene, "and al ows of a higher caloric intake than a proportionate reduction of protein, fat and carbohydrate.... The diet need not be monotonous. Many patients come to prefer it." By early 1954, The Lancet's editors were backpedaling, just as they had with Banting a century earlier. "Pennington has hardly proved his case," the journal argued, but it accepted the possibility that his diet worked, and perhaps not through the usual method of restricting calories.

The chal enge to JAMA came from a physician within the American Medical a.s.sociation itself-from George Thorpe, a Kansas doctor who both treated obese patients and chaired the AMA's Section on General Practice. At the AMA annual meeting in 1957, Thorpe charged that semi-starvation diets would inevitably fail, because they work "not by selective reduction of adipose deposits, but by wasting of al body tissues," and "therefore any success obtained must be maintained by chronic undernourishment." Thorpe had tried Pennington's diet, he said, after "considering a personal problem of excess weight." He then began prescribing the diet to his patients, who experienced "rapid loss of weight, without hunger, weakness, lethargy or constipation." Even with smal portions of salad and vegetables included, Thorpe said, weight losses of six to eight pounds a month could be obtained.

"Evidence from widely different sources," he concluded, "seems to justify the use of high-protein, high-fat, low-carbohydrate diets for successful loss of excess weight."

In response to Thorpe's testimonial, JAMA could no longer claim outright that a high-fat, carbohydrate-restricted diet would actual y increase weight, as it had a.s.serted five years earlier, but it stil insisted in a 1958 editorial that the diet would endangered health, whatever else it might accomplish.*99 Pennington's diet failed to fulfil the criterion of being "adequate in al essential nutrients," JAMA wrote. Thus, "the most reasonable diet to employ for weight reduction is one that maintains normal proportions of fat, proteins, and carbohydrates and simply limits the total quant.i.ty of the mixture." As it would do for the next fifty years, JAMA disregarded firsthand testimony from clinicians and trivialized the scientific issues; it promoted diets not because they were effective, but because they were supposedly "least harmful"-invariably basing its notion of harm on ideas that had been and would be strongly chal enged for decades.

Al the while, the DuPont experience would be confirmed in the literature repeatedly. The first confirmation came from two diet.i.tians, Margaret Ohlson and Charlotte Young, who published their observations in the Journal of the American Dietetic a.s.sociation in 1952. Ohlson was chair of the food and nutrition department at Michigan State University. Young had studied with Ohlson in the 1940s and then moved to Ithaca, New York, to become a nutritionist at Cornel . Young also worked with Cornel 's Student Medical Clinic, and it was in this capacity, along with struggles to control her own weight (she was five ten and weighed 260 pounds), that she had become dissatisfied with calorie-restricted diets.

Ohlson began her research by testing Pennington's diet on members of her own laboratory. "The edibility of the food mixture, the feeling of wel -being of the subjects and the ease with which meal pattern could be fitted into a daily schedule involving business and social engagements, suggested a further trial with patients," Ohlson reported. She then prepared a version of Pennington's diet that restricted both carbohydrates and calories, on the mistaken a.s.sumption that the diet must work by restricting calories. This was the diet that Young would also use at Cornel . It al owed only fourteen to fifteen hundred calories a day, out of which 24 percent was protein, 54 percent was fat, and 22 percent was carbohydrates.*100 Because the diet was also calorie-restricted, it did not actual y test Pennington's observation that weight would be lost even without such a calorie limitation. Nor did Ohlson or Young address the question of why their subjects never reported feeling hungry even though it provided no more calories than a typical semi-starvation diet. Stil , their observations are relevant, particularly because they came in an era when high-fat diets were not yet widely considered deadly, so that researchers were not biased by this perception.

Ohlson initial y tested a twelve-hundred-calorie low-fat diet on four overweight young women. This was eight hundred to a thousand calories less than these women normal y ate to maintain their weight, Ohlson reported, so they should have lost at least twenty-two pounds each over the fifteen weeks of the trial. Rather, the four women lost zero, six, seven, and seventeen pounds. The "subjects reported lack of 'pep' throughout...[and] they were discouraged because they were always conscious of being hungry."

Ohlson then tested her calorie-restricted version of Pennington's diet on seven women who ranged from mildly overweight to obese. These women fol owed the diet for sixteen weeks and lost between nineteen and thirty-seven pounds. In a comparison of the low-fat diet of twelve hundred calories with the carbohydrate-restricted diet of fourteen to fifteen hundred calories, the former resulted in an average weight loss of a half-pound a week, whereas the latter diet, higher in calories, induced an average weight loss of almost three pounds weekly. "Without exception, the low-carbohydrate reducing diet resulted in satisfactory weight losses," Ohlson wrote. "The subjects reported a feeling of wel -being and satisfaction. Hunger between meals was not a problem."

Over a ten-year period, Ohlson's laboratory tested a range of dietary compositions on nearly 150 women, including between 50 and 60 women on her version of Pennington's diet. She also tested low-protein diets and diets low in fat (only 180 calories, or less than 15 percent fat) but high in carbohydrates. Her subjects considered these low-fat diets to be "dry, uninteresting, [and] hard to eat," no more satisfying than those regimens of turnips, bread, and cabbage that Ancel Keys had fed his conscientious objectors. Diets with 360 calories of fat proved "sufficient to provide acceptability," she added, but her subjects "uniformly" preferred the high-fat diets, with seven to eight hundred calories of fat. At that level, the women "did not appear to give as much thought to forbidden foods," and "they also appeared to be more successful in control ing appet.i.te during col ege vacations." Simply put, Ohlson's subjects were not as hungry on the high-fat, low-carbohydrate diet as they were on the low-fat, high-carbohydrate regimens.

On these high-fat, high-protein diets, according to Ohlson, her subjects appeared to add muscle or lean-tissue ma.s.s, rather than losing it, which she believed to happen inevitably with both balanced semi-starvation diets and low-protein diets. On Ohlson's version of Pennington's diet, her subjects stored nitrogen while losing one to three pounds of weight a week. This "can only mean that replenishment of the lean muscle ma.s.s is taking place,"

Ohlson said, an observation reinforced in some of her subjects by "a reduction in dress size [that] appeared to be greater than seemed reasonable on the basis of pounds lost."

Meanwhile, Charlotte Young at Cornel first tested Ohlson's version of Pennington's diet on sixteen overweight women, who lost between nine and twenty-six pounds in ten weeks, averaging nearly two pounds per week. They were "unanimous in saying that they had not been hungry," Young wrote. She reported that her subjects seemed unexpectedly healthy while on the diets, "despite an unusual y heavy siege of colds and 'flu' on the campus," and that several "reported that their skins had never looked better than during the reducing regimen." "No excessive fatigue was evident; there was a sense of wel -being unusual during weight reduction." In 1957, Young published the results of a second trial with eight overweight male students, and the results were comparable. Young fed these men an eighteen-hundred-calorie version of Ohlson's diet. After nine weeks, the men had lost between thirteen and twenty-eight pounds, averaging almost three pounds each week. Their weight loss, Young said, "in every case" actual y exceeded that expected purely from the reduction in calories. Ohlson's and Young's journal articles were ignored.

As with virtual y al weight-loss diet studies until the last decade, these were not the kind of randomized, wel -control ed trials necessary to establish whether a particular diet actual y extends life or prevents chronic disease. Subjects were not randomly selected to fol ow a low-carbohydrate diet, or a low-calorie diet, or no diet at al , and then fol owed for months or years to compare the treatments and their respective risks and benefits. Rather, the logic behind them was that obese patients were themselves the controls because they had tried calorie-restricted diets and they hadn't been successful.

For an obese person, it's a reasonable a.s.sumption that they have tried to weigh less by eating less-i.e., calorie restriction. If that approach had worked, as Hilde Bruch noted, that person would not be obese. When Bruch described a fifty-pound weight loss in a young patient eating Pennington's diet, she also reported that the woman had described her life, as Bruch's obese patients often did, as a constant, ongoing failure to control her appet.i.te and restrict her calories to a level that would maintain or reduce her weight.

In 1961, Wil iam Leith of McGil University reported his clinical experience with forty-eight patients on Pennington's diet al of whom had previously tried low-calorie diets "without measurable success." Half had used appet.i.te-suppressant drugs ("anorectic agents," as Leith cal ed them), seven had taken "bulk subst.i.tutes," and "eight had partic.i.p.ated in group psychotherapy for a period of eight months," and yet "none of them showed a sustained loss of weight." Twenty-eight, by contrast, lost a significant portion of their excess weight on Pennington's diet-between ten and forty pounds, averaging one and a half pounds each week. "Our results do show that satisfactory weight loss may be accomplished by a ful caloric, low carbohydrate diet," Leith concluded. "The patients ingested protein and fat as desired." For the successful dieters, a significant success had fol owed a lifetime of failure.

Neither the individuals who wish to lose weight nor the clinicians who prescribe the diet need a randomized trial to tel them if it works. Such a trial is necessary only to establish that the diet works better than some other diet, and whether it leads to sustained benefits in health and longevity.

Until recently, few nutritionists or clinicians considered it worth their time and effort to test weight-reducing diets. Instead, they spent their careers studying the physiological and psychological abnormalities a.s.sociated with the condition of obesity, comparing food consumption and physical activity in obese and lean individuals, and studying obesity in animals. They tried to induce fat people to endure semi-starvation by behavioral modification; they studied pharmacological methods of suppressing hunger, or surgical methods of reducing the amount of food that could be consumed or digested.*101 Testing diets or even treating obese patients was regarded as lesser work. "To be honest, obesity treatment is extremely boring," said Per Bjorntorp, who was among the most prominent European authorities on obesity in the 1970s and 1980s. "It's very difficult and unrewarding." When obese individuals came to his biochemistry laboratory at the University of Goteborg, they were referred to the local nutritionists to be taught how to count and restrict calories. Since everyone knew that obesity was caused by overeating, why bother with diet trials? "There's no point wasting your time on them," George Bray, considered one of the world's leading authorities, said in a recent interview. "If you get restriction of energy you wil lose weight, unequivocal y. It's not an issue."

When clinical investigators did test the efficacy of high-fat, carbohydrate-restricted diets, however, the results were remarkably consistent. Every investigator reported weight losses of between one and five pounds a week even when the investigators running the trial seemed more concerned with establis.h.i.+ng that the diets caused deleterious side effects. Every investigator who discussed the subjective experiences of the test subjects reported that they suffered none of the symptoms of semi-starvation or food deprivation-"excessive fatigue, irritability, mental depression and extreme hunger," as Margaret Ohlson described them.

The last of these symptoms may be the most tel ing. The diets induced significant weight loss without hunger even when the patients ate only a few hundred calories a day, as Russel Wilder's did at the Mayo Clinic in the early 1930s, or 650800 calories per day, as was the case with the patients treated by George Blackburn and Bruce Bistrian of MIT's department of nutrition and food science and the Harvard Medical School in the 1970s. Wilder was treating his obese patients with the very low-calorie diet developed by Frank Evans, princ.i.p.al y meat, fish, and egg white, with 80100 calories' worth of green vegetables. "The absence of complaints of hunger has been remarkable," Wilder wrote. Bistrian and Blackburn reported in 1985 that they had prescribed their diet of lean meat, fish, and fowl-almost 50 percent protein calories and 50 percent fat-to seven hundred patients. On average, the patients lost forty-seven pounds over a period of four months; nearly three pounds a week. "People loved it," said Blackburn.102 Significant weight loss without hunger was also reported when the diet was prescribed at 1,000 calories, as the University of Wurzburg clinicians Heinrich Kasper and Udo Rabast did in a series of trials through the 1970s; at 1,200 calories, as the University of Iowa nutritionist Wil ard Krehl reported in 1967; at 1,320 calories, as Edgar Gordon of the University of Wisconsin reported in JAMA in 1963; at 1,400 or 1,800 calories, as Young and Ohlson did; at 2,200 calories, as the Swedish clinician Bertil Sjoval reported in 1957, and even when the diet provided more than 2,700 calories a day, as reported also in 1957 by Weldon Walker, who would later become chief of cardiology at the Walter Reed Army Medical Center in Was.h.i.+ngton. The same has invariably been the case even when patients are simply "encouraged to eat as much as [is] necessary to avoid feeling hungry," but to avoid carbohydrates in doing so, as John LaRosa, now president of the State University of New York Downstate Medical Center, reported in 1980.

Every investigator who compared these carbohydrate-restricted diets with more balanced low-calorie diets also reported that the carbohydrate-restricted diet performed at least as wel , and usual y better, even when the caloric content of the carbohydrate-restricted diet was significantly greater -say, 1,850 calories versus 950 calories, as Per Hanssen reported in 1936; or 2,200 calories versus 1,200 calories, as Bertil Sjoval reported in 1957; or even an "eat as much as you like" diet compared with a 1,000-calorie diet, as Trevor Silverstone of St. Bartholomew's Hospital in London reported in 1963 in a study of obese diabetics. The same held true for children, too. In 1979, L. Pena and his col eagues from the Higher Inst.i.tute of Medical Sciences in Havana reported that they had randomized 104 obese children to either an "eat as much as you like" high-fat, high-protein diet with only 80 calories of carbohydrates, or an 1,100-calorie diet of which half the calories came from carbohydrates. The children on the carbohydrate-restricted diet lost almost twice as much weight as those who were semi-starved on the balanced diet.

Between 1963 and 1973, Robert Kemp, a physician at Walton Hospital in Liverpool, published three articles reporting his clinical experience with a low-carbohydrate, unrestricted-calorie diet. Kemp reported that his obese patients craved carbohydrates and were invariably puzzled and frustrated by two aspects of their condition: "that other people can eat just the same diet and remain thin," and "that they themselves in earlier life may wel have been thin on the same amount and type of food on which they subsequently became fat." These observations led Kemp to formulate "a working hypothesis that the degree of tolerance for carbohydrate varies from patient to patient and indeed in the same patient at different periods of life." He then translated this hypothesis into a carbohydrate-restricted, calorie-unrestricted diet. Doing so, he said, made it "possible for the first time in [his] experience to produce worth-while results in obesity treatment."

Beginning in 1956, Kemp prescribed this diet to 1,450 overweight and obese patients. More than seven hundred (49 percent) were "successful y reduced" in his practice, which Kemp defined as having lost more than 60 percent of their excess weight. These patients averaged twenty-five pounds of weight loss after a year on the diet. Another 550 patients (38 percent) defaulted, which means they stopped appearing at Kemp's monthly counseling sessions. Nearly two hundred patients (13 percent) failed to lose significant weight while apparently fol owing through with the treatment. This failure suggested to Kemp that the diet may not work on everyone, despite some claims by popular diet books that it can.

Stil , even if we a.s.sume that al of Kemp's patients who defaulted on the diet also failed to lose significant weight, Kemp's track record would stil suggest that his carbohydrate-restricted diet was at least four times more effective than the balanced semi-starvation diet that Albert Stunkard used when reporting on his clinical experiences in 1959.

The last decade has witnessed a renewed interest in testing carbohydrate-restricted diets as obesity levels have risen and a new generation of clinicians have come to question the prevailing wisdom on weight loss. Six independent teams of investigators set out to test low-fat semi-starvation diets of the kind recommended by the American Heart a.s.sociation in randomized control trials against "eat as much as you like" Pennington-type diets, now known commonly as the Atkins diet, after Robert Atkins and Dr. Atkins' Diet Revolution. Five of these trials tested the diet on obese adults, one on adolescents. Together they included considerably more than six hundred obese subjects. In every case, the weight loss after three to six months was two to three times greater on the low-carbohydrate diet-unrestricted in calories-than on the calorie-restricted, low-fat diet.

In 2003, seven physicians from the Yale and Stanford medical schools published an article in JAMA that claimed to be the "first published synthesis of the evidence" in the English-language medical literature on the efficacy and safety of carbohydrate-restricted diets. These doctors concluded that the evidence was "insufficient to recommend or condemn the use of these diets," partly because there had been no long-term randomized control ed trials that established the safety of the diets. Nonetheless, they did report the average weight loss from the trials that the authors had cul ed from the last forty years of medical research. "Of the 34 of 38 lower-carbohydrate diets for which weight change after diet was calculated," they noted, "these lower-carbohydrate diets were found to produce greater weight loss than higher-carbohydrate diets"-an average of thirty-seven pounds when carbohydrates were restricted to less than sixty grams a day, as Pennington had prescribed, compared with four pounds when they were not.*103 Accepting that high-calorie diets can lead to greater weight loss than semi-starvation diets requires overturning certain common a.s.sumptions. One is that a calorie is a calorie, which is typical y said to be al we need to know about the relations.h.i.+p between eating and weight. "Calories are al alike," said the Harvard nutritionist Fred Stare, "whether they come from beef or bourbon, from sugar or starch, or from cheese and crackers. Too many calories are too many calories." But if a calorie is a calorie, why is it that a diet restricted in carbohydrates-eat cheese, but not crackers-leads to weight loss, largely if not completely independent of calories? If significant weight can be lost on al these carbohydrate-restricted diets, even when subjects eat twenty-seven hundred or more calories a day, how important can calories be to weight regulation? Wouldn't this imply that the quant.i.ty of carbohydrates is at least a critical factor, in which case there must be something unique about these nutrients that affects weight but fal s outside the context of energy content? Isn't it possible, as Max Rubner suggested a century ago, that "the effect of specific nutritional substances upon the glands" might be a factor when it comes to weight regulation, and perhaps the more relevant one?

Look at this another way. When Bruce Bistrian and George Blackburn instructed their patients to eat nothing but lean meat, fish, and fowl-650 to 800 calories a day of fat and protein-half of them lost at least forty pounds each. That success rate held true for "thousands of patients" from the 1970s on, Bistrian said. "It's an extraordinarily effective and safe way to get large amounts of weight loss." But had they chosen to balance these very low-calorie diets of fat and protein with carbohydrates-say, by adding another 400 calories of "wonderful fruits and vegetables," as Bistrian phrased it-they would then be consuming the kind of semi-starvation diet that inevitably fails: 1,200 calories evenly balanced between protein, fat, and carbohydrates. "The likelihood of losing forty pounds on that diet is one percent," Bistrian said.

The bottom line: If we add 400 calories of fat and protein to 800 calories of fat and protein, we have a 1,200-calorie high-fat, carbohydrate-restricted diet that wil stil result in considerable weight loss. If we add 400 calories of carbohydrates to the 800 calories of fat and protein, we have a balanced semi-starvation diet of the kind commonly recommended to treat obesity-and we reduce the efficacy by a factor of fifty. We now have a diet that wil induce forty pounds of weight loss in perhaps one in a hundred patients rather than one in two.

This striking contrast also relates to hunger. One obvious explanation for the failure of balanced semi-starvation diets is hunger. (Another, as I noted earlier, is that our bodies adjust to caloric deprivation by reducing energy expenditure.) We're semi-starved, and so we eventual y break the diet. We cannot withstand the "nagging discomfort," as Wil iam Leith put it. This is why clinicians like Pena and Leith believed that the carbohydrate-restricted diets were more successful: their obese patients could eat whenever they got hungry and would sustain the diet longer. It's why Per Hanssen in 1936 suggested that the 1,800-calorie carbohydrate-restricted diet was likely to make weight maintenance easier than a 900-calorie balanced diet. But, as Wil ard Krehl noted, the diet at 1,200 calories also abated hunger: the desire for food, he wrote, was "more than amply satisfied." Bistrian and Blackburn were able to reduce or eliminate hunger even at 650 to 800 calories. Had hunger remained acute, as Bistrian said, it's likely that the patients would have eventual y cheated, and this would have thwarted the weight loss if they cheated with carbohydrates. If the cheaters reached daily for a few hundred calories of carbohydrates-say, a bagel or a couple of a sodas-they would be eating a balanced semi-starvation diet with its 1-percent success rate. The 50-percent success rate on the half-protein, half-fat diet suggests that these dieters do not feel hunger, or certainly do not feel it as acutely as they would had they been eating a diet that came with carbohydrates as wel . "Isn't the proof of the pudding in the eating?" asked Bistrian.

These observations would suggest that we can add 400 calories to a diet of 800 calories-400 calories of fruits and vegetables on top of our 800 calories of meat, fish, and fowl-and be less satisfied. But, again, this wil happen only if the initial diet is protein and fat and the added calories come from carbohydrates. If we add more fat and protein, we have a 1,200-calorie carbohydrate-restricted diet that wil satisfy our hunger. So is the amount of calories consumed the critical variable, or is there something vital y important about the presence or absence of carbohydrates? The implication is that there is a direct connection between carbohydrates and our experience of hunger, or between fat and protein and our experience of satiety, which is precisely what Ethan Sims's overfeeding experiments had suggested-that it's possible to eat up to 10,000 calories of mostly carbohydrates and be hungry at the end of the day, whereas eating a third as many calories of mostly fat and protein wil more than satiate us.

Now take into account the experience of prolonged starvation. In 1963, Walter Bloom, then director of research at Atlanta's Piedmont Hospital, published a series of articles on starvation therapy for obesity, noting that total starvation-i.e., fasting, or eating nothing at al -and carbohydrate restriction had much in common. In both cases, our carbohydrate reserves are used up quickly, and we have to rely on protein and fat for fuel. When we fast, the protein and fat come from our muscle and fat tissue; when we restrict carbohydrates, they're provided by the diet as wel . "At a cel ular level, the major characteristic of fasting is limitation of available carbohydrate as an energy source," Bloom wrote. "Since fat and protein are the energy sources in fasting, there should be little difference in cel ular metabolism whether the fat and protein come from endogenous [internal] or exogenous [external]

sources." And this turns out to be the case. The metabolic responses of the body are virtual y identical.

And, once again, there is "little hunger" during prolonged starvation. "In total starvation," Keys wrote in The Biology of Human Starvation, "the sensation of hunger disappears in a matter of days." This a.s.sessment was confirmed in the early 1960s by Ernst Drenick at UCLA, when he starved eleven obese patients for periods of twelve to 117 days. "The most astonis.h.i.+ng aspect of this study," wrote Drenick and his col eagues in JAMA, "was the ease with which prolonged starvation was tolerated. This experience contrasted most dramatical y with the hunger and suffering described by individuals who, over a prolonged period, consume a calorical y inadequate diet." As the editors of JAMA suggested in an accompanying editorial, this absence of hunger made starvation seem to be a viable weight-loss therapy for severely obese patients: "The gratifying weight loss without hunger may bring about the desired immediate results and help establish a normal eating pattern where other dietary restrictions may fail."

The implication is that we wil experience no hunger if we eat nothing at al -zero calories-and our cel s are fueled by the protein and fat from our muscle and fat tissue. If we break our fast with any amount of dietary protein and fat, we'l stil feel no hunger. But if we add carbohydrates, as Drenick noted, we'l be overwhelmed with hunger and wil now suffer al the symptoms of food deprivation. So why is it when we add carbohydrates to the diet we get hungry, if not irritable, lethargic, and depressed, but this wil not happen when we add only protein and fat? How can the amount of calories possibly be the critical factor?

In the early 1950s, Alfred Pennington noted the paradoxes engendered by a diet restricted in carbohydrates and relatively rich in fat and protein, and described them as a "mighty stimulant to thought on the matter." But this is not how the medical-research establishment has perceived them. Rather, the accepted explanation for the success of carbohydrate-restricted diets is that they work via the same mechanism as calorie-restricted diets-they restrict calories, creating a negative energy balance. Either they so limit the choices of food that dieters simply find it too difficult to consume as many calories as they might otherwise prefer, or they bore the dieters into eating less, or both. "Many individuals spontaneously and unconsciously reduce their energy intakes by as much as 30% when placed on low carb diets," Johanna Dwyer, a Tufts University nutritionist, explained in 1985. They do this "because there is insufficient carbohydrate permitted for them to eat many common and highly palatable foods in which they might otherwise indulge." So where's the paradox?

"The fact remains that some patients have lost weight on the low-carbohydrate diet 'unrestricted in calories,'" the AMA Council on Food and Nutrition conceded in 1973 in a critique of such diets. "When obese patients reduce their carbohydrate intake drastical y, they are apparently unable to make up the ensuing deficit by means of an appreciable increase in protein and fat." By this logic, weight loss on a diet "unrestricted in calories" does not represent a refutation of the hypothesis that calorie restriction itself-creating a negative energy balance-is the only way to lose weight, because it suggests that a carbohydrate-restricted diet is a calorie-restricted diet in disguise. And the sensation of hunger isn't an issue, because it can apparently be ignored.*104 This rationale, which has been invoked frequently over the past four decades, is curious on many levels. First of al , it seems to contradict the underlying principle of low-fat diets for weight control and the notion that we get obese because we overeat on the dense calories of fat in our diets. One reason that bread has always been considered the ideal staple of a low-fat reducing diet, as Jean Mayer noted, is that it only has about sixty calories a slice. "If you put a restaurant-size pat of b.u.t.ter on your toast, for example, you triple the calories," Mayer said. If we avoid the dense calories of fat in the b.u.t.ter, the argument goes, we wil natural y eat fewer total calories and lose weight accordingly. (This was the fal back position in 1984 for the official NIH recommendation of a low-fat diet for heart disease: if nothing else, we'd lose weight on such a diet, and so that would reduce heart-disease risk.) To explain the peculiar efficacy of carbohydrate-restricted reducing diets, the circuitous reasoning is that if we avoid the not-nearly-so-dense calories of bread and potatoes, we wil also not consume the dense calories in the b.u.t.ter. We could stil eat the dense cal