Chapter Seven.

FIBER.

The thing is, it's very dangerous to have a fixed idea. A person with a fixed idea wil always find some way of convincing himself in the end that he is right.

ATLE SELBERG, winner of the 1950 Fields Medal in Mathematics THE HYPOTHESIS THAT SUGAR AND refined carbohydrates cause chronic disease peaked as a subject of serious consideration in late April 1973, when George McGovern's Senate Select Committee on Nutrition and Human Needs held its first hearing on diet and what the committee took to cal ing kil er diseases. The testimony would have little impact on the content of McGovern's Dietary Goals for Americans, in part because none of the staff members who organized the hearings would stil be working for the committee three and a half years later, when the Dietary Goals would be drafted. Equal y important, neither McGovern nor his congressional col eagues could reconcile what they were hearing from the a.s.sembled experts with what they had now come to believe about the nutritional evils of modern diets.

The committee had initial y planned a series of hearings in 1972 on dietary fat, cholesterol, and heart disease, but the plans changed because McGovern ran for president. When the committee returned to the diet-and-chronic-disease issue after McGovern's defeat, the subject that seemed most urgent-thanks in part to the publication of John Yudkin's Sweet and Dangerous-was sugar in the diet, diabetes, and heart disease.

The hearings were a surprisingly international affair. Aharon Cohen from Jerusalem testified on diabetes and heart disease among the Yemenite Jews.

George Campbel testified on his studies of diabetes in Zulus and Natal Indians in South Africa. Peter Bennett, an NIH epidemiologist, testified on the Pima Indians of Arizona, who had the highest incidence of diabetes ever recorded at the time: half of the Pima over thirty-five years old were diabetic.

"The only question that I would have," Bennett said, "is whether we can implicate sugar specifical y or whether the important factor is not calories in general, which in fact turns out to be real y excessive amounts of carbohydrates." Walter Mertz, chairman of the USDA Human Nutrition Inst.i.tute, testified, as did his col eague Carol Berdanier, explaining that refined sugar seemed to play particular havoc with health, at least in laboratory rats. It elevated blood sugar and triglycerides, and caused subjects to become diabetic, Berdanier said, "and they die at a very early age."

When the testimony focused on sugar and diabetes, the committee members found it compel ing. They occasional y solicited suggestions as to how Americans might reduce the 120-odd pounds of sugar they were eating on average in 1973, to the less than seventy pounds that Campbel said could be safely consumed without triggering an epidemic of diabetes and obesity.

With the subject of heart disease, however, controversy arrived. Cohen testified that there was no "direct relationship" linking heart disease to dietary fats, and that he had been able to induce the same blood-vessel complications seen in heart disease merely by feeding sugar to his laboratory rats. Peter Cleave testified to his belief that the problem extended to al refined carbohydrates. "I don't hold the cholesterol view for a moment," Cleave said, noting that mankind had been eating saturated fats for hundreds of thousands of years. "For a modern disease to be related to an old-fashioned food is one of the most ludicrous things I have ever heard in my life," said Cleave. "If anybody tel s me that eating fat was the cause of coronary disease, I should look at them in amazement. But, when it comes to the dreadful sweet things that are served up...that is a very different proposition." Yudkin blamed heart disease exclusively on sugar, and he was equal y adamant that neither saturated fat nor cholesterol played a role. He explained how carbohydrates and specifical y sugar in the diet could induce both diabetes and heart disease, through their effect on insulin secretion and the blood fats known as triglycerides. McGovern now struggled with the difficulty of getting some consensus on these matters.

"Are you saying that you don't think a high fat intake produces the high cholesterol count?" McGovern asked Yudkin. "Or are you even saying that a person with high cholesterol count is not in great danger?"

"Wel , I would like to exclude those rare people who have probably a genetic condition in which there is an extremely high cholesterol," Yudkin responded. "If we are talking about the general population, I believe both those things that you say. I believe that decreasing the fat in the diet is not the best way of combating a high blood cholesterol.... I believe that the high blood cholesterol in itself has nothing whatever to do with heart disease."

"That is exactly opposite what my doctor told me," said McGovern.

"If men define situations as real," the sociologist Wil iam Isaac Thomas observed in the 1920s, "they are real in their consequences." Embracing a hypothesis based on incomplete evidence or ideological beliefs is risk enough. But this also makes it extremely difficult to entertain alternative possibilities, unless we can reconcile them with what we have now convinced ourselves is indisputable.

By the early 1970s, al potential causes of heart disease, or potential y any chronic disease, had to be capable of coexisting with the belief that dietary fat was the primary cause of coronary heart disease. The notion that refined or easily digestible carbohydrates caused chronic disease could not be so reconciled.

The evidence that led Peter Cleave to propose this alternative theory-the disparity in disease rates among populations, the intimate relationship of atherosclerosis, hypertension, obesity, and diabetes, and the apparent absence of chronic disease in populations relatively free of Western influences -had to be explained in other ways if they were to be consistent with Keys's hypothesis. Fiber, the indigestible carbohydrates in vegetables, starches, and grains, now replaced refined carbohydrates and sugar in the debate about the nutritional causes of chronic diseases. The fiber hypothesis captured the public's nutritional consciousness by virtue of the messianic efforts of a single investigator, a former missionary surgeon named Denis Burkitt, who proposed that this indigestible roughage was a requisite component of a healthy diet. The notion was consistent with Keys's hypothesis, which was not the case with Cleave's or Yudkin's hypothesis, and it resonated also with the era's countercultural leanings toward diets heavy in vegetables, legumes, and cereal grains.

Burkitt's fiber hypothesis was based original y and in its entirety on Cleave's saccharine-disease hypothesis, but simply inverted the causal agent.

Rather than proclaim, as Cleave did, that chronic disease was caused by the addition of sugar and refined carbohydrates to diets that we had evolved natural y to eat, Burkitt laid the blame on the subtraction of the fiber from those evolutionarily ideal diets, which in turn led to constipation and then, through a variety of mechanisms, al the chronic diseases of civilization. The fiber deficiency itself was caused either by the removal of fiber during the refining of carbohydrates or by the consumption of refined carbohydrates in lieu of the fibrous, bulky roughage we should be eating. The fiber hypothesis and the refined-carbohydrate hypothesis of chronic disease were photographic negatives of each other, and yet the fiber hypothesis caught on immediately upon appearing in the journals. The refined-carbohydrate hypothesis, which was the only one of the two that was capable of explaining the actual evidence, remained a fringe concept.

Denis Burkitt began his career as a missionary surgeon in Uganda in 1947. In the early 1960s, he earned his renown-"one of the world's best-known medical detectives," as the Washington Post would cal him-for his studies of a fatal childhood cancer that came to be known as Burkitt's lymphoma and would be the first human cancer ever linked to a viral cause. That discovery alerted Burkitt to the lessons to be learned by tracking the geographical distribution of disease. Burkitt spent five years gathering information about the lymphoma from hundreds of African hospitals, and made a legendary ten-thousand-mile, sixty-hospital trek from Kampala to Johannesburg and back as part of his research.

In 1966, Burkitt returned to England, where he worked as a cancer epidemiologist for the Medical Research Council. There Richard Dol told him about Cleave and his saccharine-disease hypothesis. Burkitt met with Cleave and read Diabetes, Coronary Thrombosis and the Saccharine Disease, which he found revelatory. Cleave possessed "perceptive genius, persuasive argument and irrefutable logic," Burkitt wrote.

What he was saying was that many of the common diseases in post-industrialized western countries are rare throughout the third world, were rare even in England or New York until about the First World War, are equal y common in black and white Americans, and therefore must be due not to our skin color or our genes, but to the way we live. Now, this made an enormous amount of sense to me because I knew from my experience in Africa that he was perfectly right saying this.

On a tour of the United States, Burkitt visited hospitals and observed, as George Campbel had a decade earlier, that African-American patients in these hospitals were often obese, diabetic, or atherosclerotic, conditions virtual y nonexistent among the black Ugandans Burkitt had treated.

Burkitt considered himself in the ideal position to test Cleave's hypothesis on a wider scale. He had already established a network of 150 African hospitals, mostly missionary hospitals in rural areas, that mailed him monthly reports on their cancer cases: "I was able to ask them al : 'Do you see gal stones, appendicitis, diverticular disease, coronary heart disease....'" Burkitt also sent his questionnaire to mission hospitals through out the world, and over eight hundred faithful y returned them. The results confirmed the basics of Cleave's hypothesis. Whereas Cleave had anecdotal evidence, Burkitt recal ed, he now had "anecdotal multiplied by a thousand," and it was al consistent. Moreover, he had the necessary reputation to be taken seriously, whereas Cleave did not. Cleave, Campbel , and others had been "written off as cranks," Burkitt said. "Now, just because there happened to be a Burkitt's lymphoma, when Burkitt said, 'What about looking at this,' people listened to me when they hadn't listened to far better guys."

Through the early 1970s, Burkitt published a series of articles expanding on Cleave's hypothesis. "These 'western' diseases are certainly a.s.sociated geographical y and in many instances tend to be related to one another in individual patients," Burkitt wrote in the Journal of the National Cancer Inst.i.tute in 1971. "My epidemiological studies in Africa and elsewhere substantiate Cleave's basic hypothesis. Changes made in carbohydrate food may of course be only one of many etiological factors, but in some instances they would appear to be the major one."

But Burkitt was beginning to revise Cleave's hypothesis. Now Burkitt's working a.s.sumption, as he explained in the JNCI, was that any dietary factors responsible for benign conditions such as appendicitis or diverticulitis were likely to be responsible as wel for related malignant conditions-in particular, colon and rectal cancer. Burkitt's research had led him to Thomas Al inson, who in the 1880s argued that white flour caused constipation, hemorrhoids, and other il s of modern societies. It also led him to a 1920 article by the Bristol University surgeon Arthur Rendle Short, doc.u.menting a dramatic increase in the incidence of appendicitis that Rendle Short also blamed on white flour and the lack of fiber in modern diets. Burkitt believed he could draw a direct line of causation from the absence of fiber in refined carbohydrates to constipation, hemorrhoids, appendicitis, diverticulitis, polyps, and final y malignant colon and rectal cancer.

Burkitt's African correspondents had reported that appendicitis increased dramatical y in urban populations-at Burkitt's Mulago Hospital in Kampala, the number of yearly appendectomies had increased twenty-fold from 1952 to 1969-whereas polyps, diverticular disease, and colorectal cancer, al common in the United States and Europe, wrote Burkitt, were stil "very rare in Africa and almost unknown in rural communities." Burkitt concluded that appendicitis, just as it appeared in Western nations typical y in children, appeared in Africans, both adults and children, within a few years of the adoption of Western diets.

Burkitt focused now on constipation. He theorized that removing the fiber from cereal grains would slow the "transit time" of the stool through the colon.

Not only would any carcinogens in the stool therefore have more time to inflict damage on the surrounding cel s, but it was conceivable that the overconsumption of refined carbohydrates would increase the bacterial flora of the stool, and that in turn could lead to carcinogens being metabolized by the bacteria out of "normal bowel const.i.tuents." Burkitt could offer no explanation for why this might cause appendicitis, but he was confident that some combination of al these factors played a role.

In the summer of 1969, Burkitt began studying stool characteristics in available subjects. "Finished bowel transit tests on family," he recorded in his diary on July 4. The fol owing month, he visited Alec Walker, who ran the human biology department at the South African Inst.i.tute of Medical Research.

Walker had been studying the rising tide of chronic diseases in urban Bantus in South Africa since the late 1940s, and he was the rare investigator who shared with Burkitt an interest in human feces and constipation. Walker had done extensive studies linking the relative lack of constipation among black convicts in the local prison, as wel as the lack of appendicitis in the Bantus at large, to their traditional high-fiber diets. (Walker publicly dismissed the hypothesis that sugar or refined carbohydrates caused heart disease, but he also reported that the Bantus developed chronic disease only after they moved into the city and began consuming "more white bread, sugar, soft drinks and European liquor.") Walker had also just submitted an article to the British Medical Journal linking the very low mortality rates from colon cancer among the Bantus to their bowel motility, a characteristic, he wrote, that was "largely lost" among Western societies. Walker's research gave Burkitt the confidence to devote his efforts to the study of stool characteristics and bowel behavior, hoping to a.s.sociate in a scientific manner fiber deficiency, constipation, and the presence of chronic diseases.

It was precisely this work that led to the fiber hypothesis and its present place in our nutritional consciousness. In 1972, Burkitt and Walker published an article in The Lancet supporting their theory and discussing their measurements of transit time and stool characteristics in twelve hundred human subjects. In rural areas, unaffected by industrialization, they reported, "diets containing the natural amount of fiber are eaten and result in large, soft stools that traverse the intestine rapidly. By contrast, the refined low-fiber foods of the economical y-developed countries produce smal firm stools which pa.s.s through the gut very slowly." Thus, the relative constipation endemic in the developed world, they suggested, appeared to play a causative role in bowel-related disorders: appendicitis, diverticulitis, and both benign and malignant tumors of the colon and r.e.c.t.u.m, al of which showed the cla.s.sic distribution of diseases of civilizations. "Al these diseases are very closely a.s.sociated epidemiological y," Burkitt and Walker explained. "These diseases are stil rare in developing countries and in rural j.a.pan, where eating habits have changed but little, but they are al seen increasingly in j.a.panese who live in Hawaii and California and are increasing in j.a.pan in those who have changed to a Western diet. In no country or region is one of these diseases common and the others rare save that appendicitis, which afflicts the young, appears about a generation before the other conditions."

Within two years, Burkitt had extended his hypothesis from appendicitis, diverticulitis, and colon cancer to al chronic diseases of civilization. In the process, Cleave's refined-carbohydrate hypothesis of saccharine diseases was transformed into Burkitt's fiber hypothesis of Western diseases. This transformation of the causal agent of disease from the presence of carbohydrates to the absence of fiber may have been influenced by factors other than science-Burkitt's close a.s.sociation with Harold Himsworth in particular. Himsworth had been secretary of the Medical Research Council when Burkitt was hired, and he had been publicly effusive about Burkitt's contributions to modern medicine. It was Himsworth's research that had been responsible for convincing diabetologists that sugar and other carbohydrates were not the cause of diabetes. Indeed, Cleave and Campbel had presented their saccharine-disease theory in the context of diabetes as a refutation of Himsworth's scholarship as much as Joslin's. That Burkitt would find Cleave's general thesis compel ing but the details unacceptable in light of Himsworth's own work and beliefs is quite possible. Burkitt would often tel the story of how Himsworth had convinced him of the importance of paying attention to those factors that were absent in searching for the causative agents of disease. "Denis," Burkitt recal ed Himsworth tel ing him, "do you remember the story in Sherlock Holmes when Holmes said to Watson: 'The whole clue, as I see it, to this case lies in the behavior of the dog.' And Watson said: 'But, sir, the dog did nothing at al .' 'That,' said Holmes, 'is the whole point.' And it often is in medicine.... The clue can lie in what is not there rather than what is there." In this case, fiber was not there. Burkitt also seemed motivated by the simple expediency of emphasizing the positive benefits of fiber rather than the negative effects of sugar and flour, which seemed like a hopeless cause. "[Sugar] is simply an integral part of the daily diet and emphatical y is here to stay," Burkitt's col aborator Alec Walker said. Better to say Don't Forget Fibre in Your Diet, which was the t.i.tle of Burkitt's 1979 diet book, than to say, Don't eat sugar, flour, and white rice, and drink less beer.

The final transformation of Cleave's refined-carbohydrate hypothesis into Burkitt's fiber hypothesis came primarily through the efforts of Burkitt's col eague Hugh Trowel , who had spent thirty years as a missionary physician in Kenya and Uganda, beginning in 1929. This had been a time, as Trowel later explained, when scores of British doctors working for the colonial service and missionary hospitals in the Kenyan highlands had the unprecedented experience of watching the native population of "three mil ion men, women and children...emerge from pre-industrial life and undergo rapid westernization." After Trowel retired to England in 1959, he published Non-infective Disease in Africa, which was the first rigorous attempt to draw together the entire body of medical literature on the spectrum of diseases afflicting the native population of Africa.*36 The Western diseases-a list almost identical to Cleave's-were conspicuous by their absence.

Trowel 's experiences in East Africa had left him with the characteristic awareness of the diseases-of-civilization phenomenon. When he arrived in Kenya in 1929, he said, he had noticed that the Kenyans were al as thin as "ancient Egyptians," yet when he dined with the native tribes, they always left food at the end of the meal and fed it to their domestic animals, which suggested that their relative emaciation was not caused by food shortages or insufficient calories. During World War I , according to Trowel , a team of British nutritionists was dispatched to East Africa to figure out how to induce the Africans in the British Army to put on weight, since they would not or could not do it. "Hundreds of x-rays," Trowel recal ed, "were taken of African intestines in an effort to solve the mystery that lay in the fact that everyone knew how to fatten a chicken for the pot, but no one knew how to make Africans...put on flesh and fat for battle. It remained a mystery." Nonetheless, by the 1950s, fat Africans were a common sight, and in 1956 Trowel himself reported the first clinical diagnosis of coronary heart disease in a native East African-a Banting-esque high-court judge (five foot two and 208 pounds) who had lived in England and had been eating a Western diet for twenty years. In 1970, Trowel returned to East Africa and described what he saw as "an amazing spectacle: the towns were ful of obese Africans and there was a large diabetic clinic in every city. The twin diseases had been born about the same time and are now growing together."

Burkitt and Trowel had been friends since the late 1940s, when Burkitt first arrived in Uganda. In 1970 the two began working together on Burkitt's fiber hypothesis and a textbook on diseases of civilization, which Burkitt and Trowel now cal ed "Western diseases."*37 To explain how obesity could be induced by the fiber deficiency of modern refined-carbohydrate foods, Trowel reasoned that the causal factor was an increased ratio of energy to nondigestible fiber in the Western diet. Ninety-three percent of the nutrients in a typical Western diet were available for use as energy, Trowel calculated, compared with only 88 or 89 percent of those in a typical primitive diet containing copious vegetables, fruits, and wholemeal bread. The lower figure, Trowel wrote, is "the figure that is the natural, inherited evolutionary figure." Over the course of a few decades, he said, we would unknowingly eat 4 percent more calories than would be evolutionary appropriate and therefore gain weight. (Later investigators would build on this idea by adding that fibrous foods were bulky, and thus more fil ing, and they also took longer to chew and digest, which supposedly led to an inevitable decrease in calories consumed, at least per unit of time.) As for heart disease, Trowel accommodated Keys's logic: if the relevant epidemiology suggested that a low-fat, high-carbohydrate diet protected against heart disease, then carbohydrates obviously protect against heart disease, with the critical caveat that those carbohydrates must contain "their ful complement of dietary fiber." Those "partial y depleted" of fiber provide only "partial protection," Trowel said; those ful y depleted, sugar and white flour, offer no protection.

More attention would have been paid to Cleave's hypothesis, Trowel explained, had Cleave accepted the validity of Keys's research and "not dismissed completely the role of saturated animal fats" in heart disease. (Burkitt later said as much, too.) Trowel didn't make the same mistake. He accepted that diets rich in fat, especial y saturated fat, raise cholesterol levels in the blood and so raise heart-disease risk, but then noted that the epidemiological evidence also implicated a low consumption of starchy high-fiber foods. So both fat and the absence of fiber could be blamed. (As Cleave and Yudkin had pointed out, exactly the same evidence can be used to implicate sugar and refined carbohydrates.) Burkitt and Trowel cal ed their fiber hypothesis a "major modification" of Cleave's ideas, but they never actual y addressed the reasons why Cleave had identified refined carbohydrates as the problem to begin with: How to explain the absence of these chronic diseases in cultures whose traditional diets contained predominantly fat and protein and little or no plant foods and thus little or no fiber-the Masai and the Samburu, the Native Americans of the Great Plains, the Inuit? And why did chronic diseases begin appearing in these populations only with the availability of Western diets, if they weren't eating copious fiber prior to this nutrition transition? Trowel did suggest, as Keys had, that the experience of these populations might be irrelevant to the rest of the world. "Special ethnic groups like the Eskimos," he wrote, "adapted many mil ennia ago to special diets, which in other groups, not adapted to these diets, might induce disease." Trowel spent three decades in Kenya and Uganda administering to the Masai and other nomadic tribes, Burkitt had spent two decades there, and yet that was the extent of the discussion.

Unlike the reaction to Cleave's hypothesis, which garnered little attention even after Cleave testified to McGovern's Select Committee, the media pounced on the fiber hypothesis almost immediately. After Trowel published a pair of articles on fiber and heart disease in The American Journal of Clinical Nutrition in 1972, Robert Rodale, a national y syndicated columnist, wrote a series of articles on the research, touting fiber as the answer to heart disease and obesity. Rodale was president of Rodale Press and the Rodale Inst.i.tute, both dedicated to furthering the cause of organic foods and chemical-and pesticide-free "regenerative" agriculture. Rodale saw Burkitt and Trowel 's fiber hypothesis as validation of the wisdom of organic foods and the agrarian lifestyle. "The natural fiber in whole processed foods may be instrumental in keeping cholesterol levels low and preventing the onset of heart disease," he wrote.

Burkitt and Alec Walker fol owed up Trowel 's articles with an August 1974 review in The Journal of the American Medical a.s.sociation discussing the causal chain from fiber to constipation and "changes in gastrointestinal behavior" to the entire spectrum of Western diseases. The Washington Post wrote up the JAMA article on the day of its release, cal ing fiber "the tonic for our time." That December, Reader's Digest published an article on Burkitt and the fiber hypothesis; a year later, the magazine claimed that sales of fiber-rich products had more than doubled since the article. The breakfast-cereal industry, led by Kel ogg and General Foods, immediately started pushing bran and fiber as inherent heart-healthy aspects of their products. In 1975, Burkitt and Trowel published a book, Refined Carbohydrate Foods and Disease.

Burkitt then spent the next decade lecturing on the dangers of fiber-poor diets. He would condemn modern diets equal y for their "catastrophic drop in starch," for their high fat content-"We eat three times more fat than communities with a minimum prevalence of [Western] diseases," he would say; "We must reduce our fat!"-and for their lack of fiber, which he considered "the biggest nutritional catastrophe in [the United Kingdom] in the past 100 years."

Not everyone bought into it. For public-health authorities and health reporters, dietary fat and/or cholesterol continued to be the prime suspects in chronic disease, and dietary fat had already been linked through international comparisons to colon cancer, as wel as breast cancer. Burkitt recal ed memorable disputes with researchers in the United States who blamed colon cancer on dietary fat, but he insisted that the absence of fiber was responsible. Eventual y, they compromised. His opponents, said Burkitt, conceded "that the fact that fat happened to be causative...did not preclude the possibility that fiber might be protective." Harvard nutritionist Jean Mayer also discounted the significance of fiber, after Burkitt, Walker, and Trowel 's early papers sparked the "furor over fiber" in the United States. But then Mayer, too, saw the wisdom of compromise. The ideal diet, he noted, would minimize the risk of both heart disease and cancer. It would be low in fat, or at least low in saturated fat, and so would be low in meat and dairy products. And it would be high in fiber. "A good diet," Mayer wrote, "high in fruits and vegetables and with a reasonable amount of undermil ed cereals-wil give al you need of useful fiber." The a.s.sumption that it would lead to long life and good health, however, was based more on faith and intuition than on science.

Over the last quarter-century, Burkitt's fiber hypothesis has become yet another example of Francis Bacon's dictum of "wishful science"-there has been a steady acc.u.mulation of evidence refuting the notion that a fiber-deficient diet causes colon cancer, polyps, or diverticulitis, let alone any other disease of civilization. The pattern is precisely what would be expected of a hypothesis that simply isn't true: the larger and more rigorous the trials set up to test it, the more consistently negative the evidence. Between 1994 and 2000, two observational studies-of forty-seven thousand male health professionals and the eighty-nine thousand women of the Nurses Health Study, both run out of the Harvard School of Public Health-and a half-dozen randomized control trials concluded that fiber consumption is unrelated to the risk of colon cancer, as is, apparently, the consumption of fruits and vegetables. The results of the forty-nine-thousand-women Dietary Modification Trial of the Women's Health Initiative, published in 2006, confirmed that increasing the fiber in the diet (by eating more whole grains, fruits, and vegetables) had no beneficial effect on colon cancer, nor did it prevent heart disease or breast cancer or induce weight loss.

"Burkitt's hypothesis got accepted pretty wel worldwide, quite quickly, but it has gradual y been disproved," said Richard Dol , who had endorsed the hypothesis enthusiastical y in the mid-1970s. "It stil holds up in relation to constipation, but as far as a major factor in the common diseases of the developed world, no, fiber is not the answer. That's pretty clear."

As we have seen with other hypotheses, the belief that dietary fiber is an intrinsic part of any healthy diet has been kept alive by factors that have little to do with science: in particular, by Geoffrey Rose's philosophy of preventive medicine-that if a medical hypothesis has a chance of being true and thus saving lives, it should be treated as if it is-and by the need to give the public some positive advice about how they might prevent or reduce the risk of cancer. This was immediately evident in a New England Journal of Medicine editorial that accompanied back-to-back April 2000 reports on two major trials-one on fourteen hundred subjects of the Phoenix [Arizona] Colon Cancer Prevention Physicians' Network, and one $30 mil ion trial from the National Cancer Inst.i.tute-both of which confirmed that fiber had no effect on colon cancer. The editorial was written by Tim Byers, a professor of preventive medicine at the University of Colorado, who said that the two trials had been short-term and focused only on the early stages of cancer. For this reason, they should not be interpreted as "evidence that a high-fiber cereal supplement or a low-fat high-fiber diet is not effective in protecting against the later stages of development of colorectal cancer." Byers was wrong, in that the results certainly were evidence that a high-fiber diet would not protect against the later stages of colorectal cancer; they simply weren't sufficient evidence for us to accept the conclusion wholeheartedly as true.

Burkitt's hypothesis lived on, and it would continue to live, as the fat/ breast-cancer hypothesis continued to live on, in part because the original data that led to it remained unexplained: "Observational studies around the world," wrote Byers, "continue to find that the risk of colorectal cancer is lower among populations with high intakes of fruits and vegetables and that the risk changes on adoption of a different diet, but we stil do not understand why." It would always be possible to suggest, as Byers had, that the trials could have been done differently-for longer or shorter duration, on younger subjects or older subjects, with more, less, or maybe a different kind of dietary fiber-and that the results would have been more promising. The American Cancer Society and the National Cancer Inst.i.tute continued to suggest that high-fiber diets, high in fruits and vegetables, might reduce the risk of colon cancer, on the basis that some evidence existed to support the hypothesis and so a prudent diet would stil include these ingredients.

The media would also contribute to keeping the fiber hypothesis alive, having first played a significant role in transforming Burkitt's hypothesis into dogma without benefit of any meaningful long-term clinical trials. "Scientists have known for years that a diet rich in vegetables, fruits and fiber, and low in fat, can greatly reduce-or eliminate-the chances of developing colon cancer," as a 1998 Washington Post article put it-four years after the Harvard a.n.a.lysis of forty-seven thousand male health professionals suggested it was not true.

Although the New York Times ran articles on the negative results from the Nurses Health Study (by Sheryl Gay s...o...b..rg) and the Phoenix and NCI fiber trials (by Gina Kolata), neither was written by the two reporters who had fol owed the subject for decades and traditional y wrote about diet and health for the paper: Jane Brody, who wrote the Times personal-health column, and Marian Burros, who had begun endorsing the benefits of fiber as a Washington Post reporter in the 1970s and had joined the Times in 1981. Rather, Burros and Brody chose to respond to the negative news about Burkitt's hypothesis by continuing to defend it with the fal back position that it stil might be true in other ways. "If preventing colon cancer was the only reason to eat fiber,"

wrote Jane Brody after the publication of the Phoenix and NCI studies, "I would say you could safely abandon bran m.u.f.fins, whole-grain cereals, beans and peas and fiber-rich fruits and vegetables and return to a pristine diet of pasty white bread. But dietary fiber...has myriads of health benefits." After s...o...b..rg's 1999 report on the Nurses Health Study, the Times published an article by Brody ent.i.tled "Keep the Fiber Bandwagon Rol ing, for Heart and Health," pointing out that fiber was certainly good for constipation and that earlier results from the Nurses Health Study had suggested that women who ate "a starchy diet that was low in fiber and drank a lot of soft drinks developed diabetes at a rate two and a half times greater than women who ate less of these foods." This, according to Brody, const.i.tuted the motivation to keep fiber in a healthy diet.

Five days after Kolata's article on the negative results from the Phoenix and NCI trials, the Times published an explanatory article by Kolata-"Health Advice: A Matter of Cause, Effect and Confusion"-in which she discussed why the public had come to be misled on the benefits of fiber. She suggested that one reason was the loose use of language: "Scientists and the public alike use words like 'prevents' and 'protects against' and 'lowers the risk of'

when they are discussing evidence that is suggestive, and hypothesis-generating, as wel as when they are discussing evidence that is as firm as science can make it." Burkitt's fiber hypothesis, she said, had been based on hypothesis-generating data-international comparisons, in particular-and had then been refuted by the best studies science could do. "Yet even in the aftermath of the high-fiber diet studies," Kolata noted, "researchers were speaking confidently about other measures people could take to 'prevent' colon cancer, like exercising and staying thin. And they were saying that there were reasons to keep eating fiber because it could 'reduce the risk' of heart disease. When asked about the evidence for these statements, the researchers confessed that it was, of course, the lower level hypothesis-generating kind."

The very next day, the Times ran an article by Burros ent.i.tled "Plenty of Reasons to Say, 'Please Pa.s.s the Fiber,'" in which she suggested, based on what Kolata would have cal ed "hypothesis-generating data," that eating fiber "significantly" lowers the risk of heart attack in women, and that "fiber is also useful in preventing the development of diabetes," "helps control obesity," and "may also be useful in reducing hypertension." Less than a month later, Brody fol owed with an article ent.i.tled "Vindication for the Maligned Fiber Diet," noting that, although fiber had "been knocked around a bit lately, after three disappointing studies failed to find that a high-fiber diet helped to prevent colon cancer," a recent study published in the New England Journal of Medicine of thirteen subjects fol owed for six weeks suggested it helped them to better control their diabetes and so should be eaten on that basis.

"Since diabetes greatly increases a person's risk of developing heart disease and other disorders caused by fat-clogged arteries," Brody wrote, "the results of this study are highly significant to the 14 mil ion Americans with Type 2 diabetes." By 2004, Brody was advocating high-fiber diets solely for their al eged ability, untested, to induce long-term weight loss and weight maintenance. In effect, fiber had now detached itself from its original hypothesis and existed in a realm always a step beyond what had been tested. Cleave's hypothesis that refined carbohydrates and sugars were the problem, the single best explanation for the original data, had been forgotten entirely.

Chapter Eight.

THE SCIENCE OF THE CARBOHYDRATE HYPOTHESIS.

Forming hypotheses is one of the most precious faculties of the human mind and is necessary for the development of science. Sometimes, however, hypotheses grow like weeds and lead to confusion instead of clarification. Then one has to clear the field, so that the operational concepts can grow and function. Concepts should relate as directly as possible to observation and measurements, and be distorted as little as possible by explanatory elements.

MAX KLEIBER, The Fire of Life: An Introduction to Animal Energetics, 1961 AFTER THE UNITED STATES EXPLORATION EXPEDITION under Captain Charles Wilkes visited the Polynesian atols of Tokelau in January 1841, the expedition's scientists reported finding no evidence of cultivation on the atols, and confessed their surprise that the islanders could thrive on a diet composed primarily of coconuts and fish. Tokelau came under the administration of New Zealand in the mid-1920s, but the atol s remained isolated, visited only by occasional trading ships from Samoa, three hundred miles to the north. As a result, Tokelau lingered on the fringes of Western influence. The staples of the diet remained coconuts, fish, and a starchy melon known as breadfruit (introduced in the late nineteenth century) wel into the 1970s. More than 70 percent of the calories in the Tokelau diet came from coconut; more than 50 percent came from fat, and 90 percent of that was saturated.

By the mid-1960s, the population of Tokelau had grown to almost two thousand and the New Zealand government, concerned about the threat of overpopulation, initiated a voluntary migration program during which more than half the Tokelauans moved to the mainland. From 1968 to 1982, a team of New Zealand anthropologists, physicians, and epidemiologists led by Ian Prior took the opportunity to study the health and diet of the emigrants as they resettled, as wel as those who remained behind on the islands as their diets were progressively Westernized. This Tokelau Island Migration Study (TIMS) was a remarkably complete survey of the health and diet of al men, women, and children of Tokelauan ancestry. It was also quite likely the most comprehensive migration study ever carried out in the history of nutrition-and-chronic-disease research.

On Tokelau, the primary changes during the course of the study came in the mid-1970s, with the establishment of a cash economy and trading posts on the atol s. The year-round availability of imported foods led to a decrease in coconut consumption to roughly half of al calories. This was offset by a sevenfold increase in sugar consumption*38 and a nearly sixfold increase in flour consumed-from twelve pounds per person annual y to seventy pounds. The islanders also began eating canned meats and frozen foods, which they stored in freezers donated by the United Nations; by 1980, six pounds of mutton per capita, three pounds of chicken backs, and five pounds of tinned corned beef had been consumed. (In comparison, 270 pounds of fish were caught per islander in 1981.) By then, the trading ships were also delivering annual y some eighteen pounds per person of crackers, biscuits, and Twisties, a cheese-flavored corn snack. Smoking increased dramatical y, as did alcohol consumption.

Through the 1960s, the only noteworthy health problems on the islands had been skin diseases, asthma, and infectious diseases such as chicken pox, measles, and leprosy. (Modern medical services and a trained physician had been available in Tokelau since 1917.) In the decades that fol owed, diabetes, hypertension, heart disease, gout, and cancer appeared. This coincided with a decrease in cholesterol levels, consistent with the decrease in saturated-fat consumption. Average weights increased by twenty to thirty pounds in men and women. A similar, albeit smal er, trend was seen in Tokelauan children. The only conspicuous departure from these trends was in 1979, when the chartered pa.s.senger-and-cargo ship Cenpac Rounder ran aground and the islanders went five months without a food or fuel delivery. "There was no sugar, flour, tobacco and starch foods," reported the New Zealand Herald, "and the atol hospitals reported a shortage of business during the enforced isolation. It was reported that the Tokelauans had been very healthy during that time and had returned to the pre-European diet of coconut and fish. Many people lost weight and felt very much better including some of the diabetics."

As for the migrants to New Zealand, the move brought "immediate and extensive changes" in diet: bread and potatoes replaced breadfruit, meat replaced fish, and coconuts virtual y vanished from the diet. Fat and saturated-fat consumption dropped, to be replaced once again by carbohydrates, "the difference being due to the big increase in sucrose consumption." This coincided with an almost immediate increase in weight and blood pressure, and a decrease in cholesterol levels -al more p.r.o.nounced than the increases witnessed on Tokelau. Hypertension was twice as common among the migrants as among the Tokelauans who remained on the islands. The migrants also had an "exceptional y high incidence" of "diabetes, gout, and osteoarthritis, as wel as hypertension." Electrocardio-graphic evidence suggested that the "migrants were at higher risk for coronary heart disease than were non-migrants."

A number of factors combined to make this higher disease incidence among the migrants difficult to explain. For one thing, the Tokelauans who emigrated smoked fewer cigarettes than those who remained on the atol s, so tobacco was unlikely to explain this pattern of disease. The migrants tended to be younger, too, which should have led to the appearance of less chronic disease on the mainland. And though the weights of the Tokelauan migrants were "substantial y higher" than those of the atol -dwel ers and, "in fact, obesity became a problem for some," the migrant lifestyle was definitively the more rigorous of the two. The men worked in the forest service and casting shops of the railway; the women worked in electrical-a.s.sembly plants or clothing factories, or they cleaned offices during the evening hours, and they walked "some distance to and from the shops with their purchases." Final y, the original Tokelauan diet had been remarkably high in fat and saturated fat, but the migrants consumed considerably less of both. If Keys's hypothesis was correct, the migrants should have manifested less evidence of heart disease, not more.

In fact, the migrant experience had led to an increased incidence over the entire spectrum of chronic diseases. Prior and his col eagues acknowledged that their data made this difficult to explain in any simple manner. They suggested "that a different set of relevant variables might account for observed differences in incidence." Excess weight, whatever the cause, could explain at least part of the increased incidence of hypertension, diabetes, coronary heart disease, and gout among the migrants. They appeared to get more salt in their diets than the islanders did, so that might also explain the increased incidence of hypertension, as might the stress of a.s.similating to a new culture. The red meat consumed on the mainland might have contributed to the increased incidence of gout as wel . The greater incidence of asthma could be explained by the presence of al ergens in New Zealand that were absent in Tokelau.

As in the Tokelau study, the dominant approach over the past fifty years toward understanding the chronic diseases of civilization has been to a.s.sume that they are only coincidental y related, that each disease has its unique causal factors a.s.sociated with the Western diet and lifestyle, although dietary fat, saturated fat, serum cholesterol, and excess weight invariably remain prime suspects.

The less common approach to this synchronicity of diseases has been to a.s.sume, as Peter Cleave did, that related diseases have related or common causes; that they are manifestations of a single underlying disorder. Cleave cal ed it the saccharine disease because he believed sugar and other refined carbohydrates were responsible. By this philosophy, if diabetes, coronary heart disease, obesity, gout, and hypertension appear simultaneously in populations, as they did in the Tokelauan experience, and are frequently found together in the same patients, then they are very likely to be manifestations of a single underlying pathology. If nothing else, Cleave argued, this common-cause hypothesis was the simplest possible explanation for the evidence, and thus the one that should be presumed true until compel ing evidence refuted it. This was Occam's razor, and it should be the guiding principle of al scientific endeavors.

In the early 1950s, clinical investigators began to characterize the physiological mechanisms that would underlie Cleave's saccharine-disease hypothesis of chronic disease, and that could explain the appearance of diseases of civilization going back over a century-the basis, in effect, of this carbohydrate hypothesis. The research evolved in multiple threads that resulted in some of the most fundamental discoveries in heart-disease and diabetes research. Only in the late 1980s did they begin to come together, when the Stanford diabetologist Gerald Reaven proposed the name Syndrome X to describe the metabolic abnormalities common to obesity, diabetes, and heart disease, al , at the very least, exacerbated by the consumption of sugar, flour, and other easily digestible carbohydrates. Syndrome X included elevated levels of the blood fats known as triglycerides; low levels of HDL cholesterol, now known as the good cholesterol; it included hypertension, and three phenomena that are considered precursors of adult-onset diabetes-chronical y high levels of insulin (hyperinsulinemia), a condition known as insulin resistance (a relative insensitivity of cel s to insulin), and the related condition of glucose intolerance (an inability to metabolize glucose properly). Over the years, other abnormalities have been added to this list: the presence of predominantly smal , dense LDL particles, and high levels of a protein cal ed fibrinogen that increases the likelihood of blood-clot formation. Elevated uric-acid concentrations in the blood, a precursor of gout, have been linked to Syndrome X, as has a state of chronic inflammation, marked by a high concentration in the blood of a protein known as C-reactive protein.

In the last decade, Syndrome X has taken on a variety of names as authorities, inst.i.tutions, and a.s.sociations have slowly come to accept its validity. It is often referred to as insulin resistance syndrome. The National Heart, Lung, and Blood Inst.i.tute belatedly recognized the existence of Syndrome X in 2001, cal ing it metabolic syndrome. It has even been referred to as insulin resistance/metabolic syndrome X, or MSX, by those investigators attempting to cover al bases.*39 By any name, this metabolic syndrome is as much a disorder of carbohydrate metabolism as is adult-onset diabetes, and is certainly a consequence of the carbohydrate content of the diet, particularly, as Cleave would have predicted, such refined, easily digestible carbohydrates as sugar and white flour.

It wasn't until the late 1990s that the evolving science of metabolic syndrome began to have any significant influence outside the field of diabetes, at which point the media final y began to take notice.40 The potential implications of metabolic syndrome for heart disease and other chronic diseases have only just begun to be appreciated by the research community. As a result, a hypothesis that emerged from research in the 1950s as an alternative explanation for the high rates of heart disease in Western nations has been accepted by medical researchers and public-health authorities a half-century later as a minor modification to Keys's dietary-fat/cholesterol hypothesis, even though this alternative hypothesis implies that Keys's hypothesis is wrong. The bulk of the science is no longer controversial, but its potential significance has been minimized by the a.s.sumption that saturated fat is stil the primary evil in modern diets.

The Tokelau experience stands as an example. The current accepted explanation for the pattern of disease among the Tokelauans is that the increased sugar and flour in their diets caused metabolic syndrome, and in turn heart disease and diabetes, at least according to Scott Grundy, who is a nutritionist and specialist in the metabolism of blood lipids at the University of Texas Southwestern Medical Center and the primary author of the 2003 cholesterol guidelines published by the National Cholesterol Education Program (NCEP). This does not mean, however, that Grundy believes that Cleave's saccharine-disease hypothesis of chronic disease is correct, or that Keys was incorrect. Rather, as he explained it, in the United States the situation was less straightforward than in Tokelau. "What you're faced with," Grundy said, "is a historical change in people's habits. Going back to the 1940s, '50s, and '60s, people ate huge amounts of b.u.t.ter and cheese and eggs, and they had very high LDL levels [the "bad cholesterol"] and they had severe heart disease early in life, because of such high cholesterol levels. What's happened since then is, there has been a change in population behavior, and they don't consume such high quant.i.ties of saturated fat and cholesterol anymore, and so LDL has come down a great deal as our diets have changed. But now...we have got obesity, and most of the problem is due to higher carbohydrate consumption or higher total calories. And so we're switching more to metabolic syndrome."

Grundy's explanation is a modern version of the changing-American-diet story, in this case invoked as a rationale to explain how metabolic syndrome could be the primary cause of heart disease today, while Keys's hypothesis could stil be correct, but no longer particularly relevant to our twenty-first-century health problems. Grundy's explanation al ows both Keys and Cleave to be right-by suggesting that their hypotheses addressed two different but relevant nutrition transitions-and therefore does not require that we question the credibility of our public-health authorities. His explanation might be valid, but it relies on a number of disputable a.s.sumptions and a selective interpretation of the evidence. It could also be true that we faced very much the same problem fifty years ago that we do today, and that a continuing acc.u.mulation of evidence exonerates the fats in the diet and incriminates refined, easily digestible carbohydrates and starches instead. The implications are profound.

The appropriate response to any remarkable proposition in science is extreme skepticism, and the carbohydrate hypothesis of chronic disease offers no exception. But looking at the hypothesis in the context of a concept cal ed homeostasis, which is of fundamental importance for understanding the nature of living organisms, gives us great insight. Much of the progress in physiology in the mid-twentieth century could be described as the transferral of this "concept of the nature of the wholeness," as the n.o.bel Laureate chemist Hans Krebs suggested in 1971, "from the realm of philosophy and theory of knowledge to that of biochemical and physiological experimentation." Though physiologists were aware of this paradigmatic shift, clinical investigators studying chronic disease have paid little attention, which means that the greater implications of the fundamental idea of homeostasis have been slighted.

In the mid-nineteenth century, the legendary French physiologist Claude Bernard observed that the fundamental feature of al living organisms is the interdependence of the parts of the body to the whole. Living beings are a "harmonious ensemble," he said, and so al physiological systems have to work together to a.s.sure survival. The prerequisite for this survival is that we maintain the stability of our internal environment, the milieu interieur, as Bernard famously phrased it -including a body temperature between 97.3F and 99.1F and a blood-sugar level between 70 mg/dl and 170 to 180 mg/dl-regardless of external influences. "Al the vital mechanisms, however varied they may be," Bernard wrote, "have only one object, that of preserving constant the conditions of life in the internal environment." (As the British biologist J.B.S. Haldane noted a half-century later, "No more pregnant sentence was ever framed by a physiologist.") And this stability of the milieu interieur is accomplished, Bernard said, by a continual adjustment of al the components of this living ensemble "with such a degree of perfection that external variations are instantly compensated and equilibrated."

In 1926, Bernard's concept was reinvented as homeostasis by the Harvard physiologist Walter Cannon, who coined the term to describe what he cal ed more col oquial y "the wisdom of the body." "Somehow the unstable stuff of which we are composed," Cannon wrote, "had learned the trick of maintaining stability." Although "homeostasis" technical y means "standing the same," both Cannon and Bernard envisioned a concept more akin to what systems engineers cal a dynamic equilibrium: biological systems change with time, and change in response to the forces acting on them, but always work to return to the same equilibrium point-the roughly 98.6F of body temperature, for instance. The human body is perceived as a fantastical y complex web of these interdependent homeostatic systems, maintaining such things as body temperature, blood pressure, mineral and electric-charge concentration (pH) in the blood, heartbeat, and respiration, al sufficiently stable so that we can sail through the moment-to-moment vicissitudes of the outside world. Anything that serves to disturb this harmonic ensemble wil evoke instantaneous compensatory responses throughout that work to return us to dynamic equilibrium.

Al homeostatic systems, as Bernard observed, must be amazingly interdependent to keep the body functioning properly. Maintaining a constant body temperature, for example, is critical because biochemical reactions are temperature-sensitive -they wil proceed faster in hotter temperatures and slower in colder ones. But not al biochemical reactions are equal y sensitive, so their rates of reaction wil not change equal y with changes in temperature. A biological system like ours that runs ideal y at 98.6F can spin out of control when this temperature changes and al the myriad biochemical reactions on which it depends now proceed at different rates. Our body temperature is the product of the heat released from the chemical reactions that const.i.tute our metabolism. It is balanced in turn by the cooling of our skin in contact with the outside air. On cold days, we wil metabolical y compensate to generate more heat, and so more of the calories we consume go to warming our bodies than they would on hot days. Thus, the ambient temperature immediately affects, among other things, the regulation of blood-sugar and of carbohydrate and fat metabolism. Anything that increases body heat (like exercise or a hot summer day) wil be balanced by a reduction of heat generated by the cel s, and so there is a decrease in fuel use by the cel s. It wil also be balanced by dehydration, increased sweating, and the dilation of blood vessels near the surface of the skin. These, in turn, wil affect blood pressure, so another set of homeostatic mechanisms must work, among other things, to maintain a stable concentration of salts, electric charge, and water volume. As the volume of water in and around the cel s decreases in response to the water lost from sweating or dehydration, our bodies respond by limiting the amount of water the kidneys excrete as urine and inducing thirst, so we drink water and replenish what we've lost. And so it goes. Any change in any one homeostatic variable results in compensatory changes in al of them.

This whole-body homeostasis is orchestrated by a single, evolutionarily ancient region of the brain known as the hypothalamus, which sits at the base of the brain. It accomplishes this orchestral task through modulation of the nervous system -specifical y, the autonomic nervous system, which controls involuntary functions-and the endocrine system, which is the system of hormones. The hormones control reproduction, regulate growth and development, maintain the internal environment-i.e., homeostasis-and regulate energy production, utilization, and storage. Al four functions are interdependent, and the last one is fundamental to the success of the other three. For this reason, al hormones have some effect, directly or indirectly, on fuel utilization and what's known technical y as fuel part.i.tioning, how fuel is used by the body in the short term and stored for the long term. Growth hormone, for example, wil stimulate the mobilization of fat from fat cel s to use as energy for cel repair and tissue growth.

Al other hormones, however, are secondary to the role of insulin in energy production, utilization, and storage. Historical y, physicians have viewed insulin as though it has a single primary function: to remove and store away sugar from the blood after a meal. This is the most conspicuous function impaired in diabetes. But the roles of insulin are many and diverse. It is the primary regulator of fat, carbohydrate, and protein metabolism; it regulates the synthesis of a molecule cal ed glycogen, the form in which glucose is stored in muscle tissue and the liver; it stimulates the synthesis and storage of fats in fat depots and in the liver, and it inhibits the release of that fat. Insulin also stimulates the synthesis of proteins and of molecules involved in the function, repair, and growth of cel s, and even of RNA and DNA molecules, as wel .

Insulin, in short, is the one hormone that serves to coordinate and regulate everything having to do with the storage and use of nutrients and thus the maintenance of homeostasis and, in a word, life. It's al these aspects of homeostatic regulatory systems-in particular, carbohydrate and fat metabolism, and kidney and liver functions-that are malfunctioning in the cl.u.s.ter of metabolic abnormalities a.s.sociated with metabolic syndrome and with the chronic diseases of civilization. As metabolic syndrome implies, and as John Yudkin observed in 1986, both heart disease and diabetes are a.s.sociated with a host of metabolic and hormonal abnormalities that go far beyond elevations in cholesterol levels and so, presumably, any possible effect of saturated fat in the diet.

This suggests another way to look at Peter Cleave's saccharine-disease hypothesis, or what I'l cal , for simplicity, the carbohydrate hypothesis of chronic disease. As Cleave pointed out, species need time to adapt ful y to changes in their environment-whether shifts in climate, the appearance of new predators, or changes in food supply. The same is true of the internal environment of the human body-Bernard's milieu interieur. By far the most dramatic change to this internal environment over the past two mil ion years is due to the introduction of diets high in sugar and refined and other easily digestible carbohydrates. Blood-sugar levels rise dramatical y after these meals; insulin levels rise in response and become chronical y elevated-hyperinsulinemia-and tissues become resistant to insulin. And because half of every molecule of table sugar (technical y known as sucrose) is a molecule of the sugar known as fructose, which is found natural y only in smal concentrations in fruits and some root vegetables, the human body has also been confronted with having to adjust to radical y large amounts of fructose. In this sense, al of the abnormalities of metabolic syndrome and the accompanying chronic diseases of civilization can be viewed as the dysregulation of homeostasis caused by the repercussions throughout the body of the blood-sugar, insulin, and fructose-induced changes in regulatory systems. (As the geneticist James Neel wrote in 1998 about adult-onset diabetes, "The changing dietary patterns of Western civilization had compromised a complex homeostatic mechanism.") It's possible that obesity, diabetes, heart disease, hypertension, and the other a.s.sociated diseases of civilization al have independent causes, as the conventional wisdom suggests, but that they serve as risk factors for each other, because once we get one of these diseases we become more susceptible to the others. It's also possible that refined carbohydrates and sugar, in particular, create such profound disturbances in blood sugar and insulin that they lead to disturbances in mechanisms of homeostatic regulation and growth throughout the entire body.

Any a.s.sumptions about regulatory mechanisms and disease, as Claude Bernard explained, have to be understood in the context of the entire harmonic ensemble. "We real y must learn, then, that if we break up a living organism by isolating its different parts, it is only for the sake of ease in experimental a.n.a.lysis, and by no means in order to conceive them separately," Bernard wrote. "Indeed when we wish to ascribe to a physiological quality its value and true significance, we must always refer it to this whole, and draw our final conclusion only in relation to its effects in the whole." When Hans Krebs paraphrased this lesson a century later, he said that if we neglect "the wholeness of the organism-we may be led, even if we experimented skil ful y, to very false ideas and very erroneous deductions."

Perhaps the simplest example of this kind of erroneous deduction is the common a.s.sumption that the cause of high blood pressure and hypertension is excess salt consumption.

Hypertension is defined technical y as a systolic blood pressure higher than 140 and a diastolic blood pressure higher than 90. It has been known since the 1920s, when physicians first started measuring blood pressure regularly in their patients, that hypertension is a major risk factor for both heart disease and stroke. It's also a risk factor for obesity and diabetes, and the other way around-if we're diabetic and/or obese, we're more likely to have hypertension. If we're hypertensive, we're more likely to become diabetic and/or obese. For those who become diabetic, hypertension is said to account for up to 85 percent of the considerably increased risk of heart disease. Studies have also demonstrated that insulin levels are abnormal y elevated in hypertensives, and so hypertension, with or without obesity and/or diabetes, is now commonly referred to as an "insulin-resistant state." (This is the implication of including hypertension among the cl.u.s.ter of abnormalities that const.i.tute metabolic syndrome.) Hypertension is so common in the obese, and obesity so common among hypertensives, that textbooks wil often speculate that it's overweight that causes hypertension to begin with. So, the higher the blood pressure, the higher the cholesterol and triglyceride levels, the greater the body weight, and the greater the risk of diabetes and heart disease.

Despite the intimate a.s.sociation of these diseases, pub