Good Calories, Bad Calories Part 3

One striking fact about this evolution is that the low-fat diets now being recommended for the entire nation had only been tested twice, as I've said, once in Hungary and once in Britain, and in only a few hundred middle-aged men who had already suffered heart attacks. The results of those trials had been contradictory. The diets tested since then had been exclusively cholesterol-lowering diets that replaced saturated fats with unsaturated fats.

The rationale for lowering the total fat content of the diet to 30 percent was the tangential expectation that such a diet would help us control our weight. In 1984, the year of the NIH Consensus Conference, Robert Levy and Nancy Ernst of the NHLBI had described the state of the science this way: "There has been some indication that a low-fat diet decreases blood cholesterol levels," they wrote. "There is no conclusive proof that this lowering is independent of other concomitant changes in the diet (for example, increased dietary fiber or complex carbohydrate...or decreased cholesterol or saturated fatty acid level).... It may be said with certainty, however, that because 1 g fat provides about 9 calories-compared to about 4 calories for 1 g of protein or carbohydrate-fat is a major source of calories in the American diet. Attempts to lose weight or maintain weight must obviously focus on the content of fat in the diet." Though this was an untested conjecture (however obvious it might seem), the official healthy diet of the nation was now a low-fat diet. A new generation of diet doctors, the most influential of whom was Dean Ornish, were even prescribing 10-percent-fat diets, if not lower.

Another striking aspect of the low-fat diet recommendations is how little any individual might benefit from lowering his cholesterol.*19 Keys and others had argued that heart disease had to be prevented because its first symptom was often a fatal heart attack. But in twenty-four years of observation, the Framingham Heart Study had detected no relations.h.i.+p between cholesterol and sudden cardiac death. The likelihood of suffering a fatal first heart attack was no less for those with a cholesterol level of 180 mg/dl than for those with 250. "The lack of a.s.sociation between serum cholesterol level and the incidence of sudden death suggests that factors other than the atherosclerotic process may be of major importance in this manifestation of coronary artery disease," explained Thomas Dawber.

There is also little to gain from lowering cholesterol even in less catastrophic manifestations of the disease. This was made clear in 1986, when Stamler published a rea.n.a.lysis of his MRFIT data in JAMA. As Stamler reported it, the MRFIT investigators had continued to track the health of the 362,000 middle-aged men who had original y been screened as potential candidates for MRFIT, including death certificates. Stamler reported that the cholesterol/heart-disease a.s.sociation applied at any level of cholesterol, and so anyone would benefit from lowering cholesterol.

Using the MRFIT data, however, it is possible to see how large or smal that benefit might be (see chart, below). For every one thousand middle-aged men who had high cholesterol-between, say, 240 and 250 mg/dl-eight could expect to die of heart disease over any six-year period. For every thousand men with cholesterol between 210 and 220, roughly six could expect to die of heart disease. These numbers suggest that reducing cholesterol from, say, 250 to 220 would reduce the risk of dying from a heart attack in any six-year period from .8 percent (eight in a thousand) to .6 percent (six in a thousand). If we were to stick rigorously to a cholesterol-lowering diet for thirty years-say, from age forty to seventy, at which point high cholesterol is no longer a.s.sociated with an increased risk of heart disease-we would reduce our risk of dying of a heart attack by 1 percent.

The data from the MRFIT trial showing the relations.h.i.+p between heart-disease mortality and cholesterol levels in the blood.

The data from the MRFIT trial showing the relations.h.i.+p between total mortality-i.e., death by all causes-and cholesterol levels in the blood.

Whether we would actual y live longer by lowering our cholesterol is, of course, a different question. People die from myriad causes. Though Stamler neglected to include total mortality data in his JAMA article, a second group of MRFIT investigators did include it in an article published in The Lancet just a month earlier.

Their data revealed that for every thousand men with cholesterol around 240 to 250 mg/dl, twenty to twenty-three would likely die of any cause within six years. For those whose cholesterol was approximately 220, between nineteen and twenty-one were likely to die. In other words, for every thousand middle-aged men who successful y lower their cholesterol by diet from, say, 250 to 220, at most four (although perhaps none) can expect to avoid death during any six-year period. Nineteen or twenty of these men can expect to die whether they diet or not. For the remaining 98 percent, they wil live regardless of their choice. Moreover, lowering cholesterol further would not help. The death rate for men whose cholesterol is below 200 appears little different from that of men whose cholesterol fal s between 200 and 250. Only for those men whose cholesterol is above 250 mg/dl does it appear that lowering cholesterol might improve the chances of living longer.

There is another way to interpret this statistical a.s.sociation between cholesterol, heart disease, and death. The a.s.sociation, as doc.u.mented by Framingham, MRFIT, and other studies, only says that, the higher our cholesterol, the greater our risk of heart disease. It does not tel us whether the benefit from lowering cholesterol is shared by the entire population or only by a smal percentage. The latter is the implicit a.s.sumption of the above a.n.a.lysis. But what if the benefit of lowering cholesterol is indeed shared democratical y among al who do it? Perhaps we may al live longer by lowering our cholesterol. But how much longer?

Between 1987 and 1994, independent research groups from Harvard Medical School, the University of California, San Francisco, and McGil University in Montreal addressed the question of how much longer we might expect to live if no more than 30 percent of our calories came from fat, and no more than 10 percent from saturated fat, as recommended by the various government agencies. Al three a.s.sumed that cholesterol levels would drop accordingly, and that this low-fat diet would have no adverse effects, which was stil speculation rather than fact.

The Harvard study, led by Wil iam Taylor, concluded that men with a high risk of heart disease-such as smokers with high blood pressure-might gain one extra year of life by shunning saturated fat. Healthy nonsmokers, however, might expect to gain only three days to three months. "Although there are undoubtedly persons who would choose to partic.i.p.ate in a lifelong regimen of dietary change to achieve results of this magnitude, we suspect that some might not," the Harvard investigators noted.

The UCSF study, led by Warren Browner, was initiated and funded by the Surgeon General's Office. This study concluded that cutting fat consumption in America would delay forty-two thousand deaths each year, but the average life expectancy would increase by only three to four months. To be precise, a man who might otherwise die at sixty-five could expect to live an extra month if he avoided saturated fat for his entire adult life. If he lived to be ninety, he could expect an extra four months.*20 The McGil study, published in 1994, concluded that reducing saturated fat in the diet to 8 percent of al calories would result in an average increase in life expectancy of four days to two months.

Browner reported his results to the Surgeon General's Office, and only then submitted his article to JAMA. J. Michael McGinnis, the deputy a.s.sistant secretary for health, then wrote to JAMA trying to prevent publication of Browner's article, or at least to convince the editors to run an accompanying editorial that would explain why Browner's a.n.a.lysis should not be considered relevant to the benefits of eating less fat. "They would have liked it to come out the other way," explained Marion Nestle, who had edited the Surgeon General's Report on Diet and Health and had recruited Browner to do the a.n.a.lysis. This put Browner in the awkward position of protecting his work from his own funding agents. As he wrote McGinnis at the time, "I am sensitive to the needs of your office to put forward a consistent statement about what Americans should do, and to your dismay when a project that you have sponsored raises some questions about current policy. I am also concerned that the impacts of recommendations that apply to 240 mil ion Americans are clearly understood. This ma.n.u.script estimates the effects of one such recommendation-altering dietary fat intake to 30 percent of calories-based on the a.s.sumptions that underlie that recommendation. Shooting the messenger-or creating a smoke screen-does not change those estimates." JAMA published Browner's article-"What If Americans Ate Less Fat?"-without an accompanying editorial.

That cholesterol-lowering provides little benefit to the individual was not unknown to the authors of these expert reports. This rationale was elucidated in Diet and Health, which explained that the purpose of preventive medicine in public health was to achieve the greatest good by treating entire populations rather than individuals. In this case, that meant addressing the situation of the 85 or 90 percent of the population with normal or low cholesterol. Though the actual benefit to these individuals "might be smal or negligible," as Diet and Health explained, "because these people represent the great majority of the population, the benefit for the total population is likely to be paradoxical y large."

This strategy is credited to the British epidemiologist Geoffrey Rose, a longtime veteran of the dietary-fat controversy. "The ma.s.s approach is inherently the only ultimate answer to the problem of a ma.s.s disease," Rose explained in 1981.

But, however much it may offer to the community as a whole, it offers little to each partic.i.p.ating individual. When ma.s.s diphtheria immunization was introduced in Britain 40 years ago, even then roughly 600 children had to be immunized in order that one life would be saved-599 "wasted"

immunizations for the one that was effective.... This is the kind of ratio that one has to accept in ma.s.s preventive medicine. A measure applied to many wil actual y benefit few.

When it came to dietary fat and heart disease, according to Rose's calculation, only one man in every fifty might expect to avoid a heart attack by virtue of avoiding saturated fat for his entire adult life: "Forty-nine out of fifty would eat differently every day for forty years and perhaps get nothing from it."

And thus the dilemma: "People wil not be motivated to any great extent to take our advice because there is little in it for each of them, particularly in the short and medium term."*21 The best way around this problem, Rose explained, is to create social pressure to change. Consider young women who diet, he suggested, "not for medical reasons but because thinness is social y acceptable and obesity is not." So the task confronting public-health authorities is to create similar social pressure to induce "healthy behavior." And to do that, the benefits-or the risks of "unhealthy" behavior-have to be made to seem dramatic. "The modern British diet is kil ing people in their thousands from heart attacks," Rose told the BBC in 1984.

The a.s.sumption underpinning this public-health philosophy, as Rose explained in an influential 1985 International Journal of Epidemiology article ent.i.tled "Sick Individuals and Sick Populations," is that the entire population chronical y overconsumes fat, and al of us have cholesterol levels that are unnatural y high. This is why attempts to uncover an a.s.sociation between fat consumption and cholesterol within a population like Framingham, Ma.s.sachusetts, inevitably failed. Imagine, Rose suggested, if everyone smoked a pack of cigarettes every day. Any study trying to link cigarette smoking to lung cancer "would lead us to conclude that lung cancer was a genetic disease...since if everyone is exposed to the necessary agent, then the distribution of cases is whol y determined by individual susceptibility." The only way to escape this misconception, as with dietary fat, cholesterol, and heart disease, is to study the "differences between populations or from changes within populations over time." This "sick population" logic also explained why lowering cholesterol by 10 or 20 percent wil have little effect on a single individual-just as smoking sixteen or eighteen cigarettes a day instead of twenty wil do little to reduce individual lung-cancer risk-but would significantly affect the burden of heart disease across the entire population, and so should be widely recommended.

The arguments on sick populations and preventive public health are compel ing, but they come with four critical y important caveats.

First, Rose's logic does not differentiate between hypotheses. It would invariably be invoked to explain why studies failed to confirm Keys's fat hypothesis, and would be considered extraneous when similar studies failed to generate evidence supporting competing hypotheses. It is precisely to avoid such subjective biases that randomized control ed trials are necessary to determine which hypotheses are most likely true.

Second, as Rose observed, al public-health interventions come with potential risks, as wel as benefits-unintended or unimagined side effects. Smal or negligible risks to an individual wil also add up and can lead to unacceptable harm to the population at large. As a result, the only acceptable measures of prevention are those that remove what Rose cal ed "unnatural factors" and restore "'biological normality'-that is...the conditions to which presumably we are genetical y adapted." "Such normalizing measures," Rose explained, "may be presumed to be safe, and therefore we should be prepared to advocate them on the basis of a reasonable presumption of benefit."

This facet of Rose's argument effectively underpins al public-health recommendations that we eat low-fat or low-saturated-fat diets, despite the negligible benefits. It requires that we make a.s.sumptions about what is safe and what might cause harm, and what const.i.tutes "biological normality" and "unnatural factors." The evidence for those a.s.sumptions wil always depend as much on the observers' preconceptions and belief system as on any objective reality.

By defining "biological normality" as "the conditions to which presumably we are genetical y adapted," Rose was saying that the healthiest diet is (presumably) the diet we evolved to eat. That is the diet we consumed prior to the invention of agriculture, during the two mil ion years of the Paleolithic era-99 percent of evolutionary history-when our ancestors were hunters and gatherers. "There has been no time for significant further genetic adaptation," as the nutritionists Nevin Scrimshaw of MIT and Wil iam Dietz of the Centers for Disease Control noted in 1995. Any changes to this Paleolithic diet can be considered "unnatural factors," and so cannot be prescribed as a public-health recommendation.

The Paleolithic era, however, is ancient history, which means our conception of the typical Paleolithic diet is wide open to interpretation and bias. In the 1960s, when Keys was struggling to have his fat hypothesis accepted, Stamler's conception of the Paleolithic hunter-gatherer diet was mainly "nuts, fruits and vegetables, and smal game." We only began consuming "substantial amounts of meat," he explained, and thus substantial amounts of animal fat, twenty-five thousand years ago, when we developed the skil s to hunt big game. If this was the case, then we could safely recommend, as Stamler did, that we eat a low-fat diet, and particularly low in saturated fats, because animal fats in any quant.i.ty were a relatively new addition to the diet and therefore unnatural.

This interpretation, shared by Rose, was established authoritatively in 1985, the year after the NIH Consensus Conference, when The New England Journal of Medicine published a quant.i.tative a.n.a.lysis of hunter-gatherer diets by two investigators-Boyd Eaton, a physician with an amateur interest in anthropology, and Melvin Konner, an anthropologist who had recently earned his medical degree. Eaton and Konner a.n.a.lyzed the diets of hunter-gatherer populations that had survived into the twentieth century and concluded that we are, indeed, genetical y adapted to eat diets of 2025 percent fat, most of which would in the past have been unsaturated. Eaton and Konner's article has since been invoked to support low-fat recommendations-in Diet and Health, for instance-as Rose's argument suggests it should.

But Eaton and Konner "made a mistake," as Eaton himself later said. This was only corrected in 2000, when Eaton, working now with John Speth and Loren Cordain, published a revised a.n.a.lysis of hunter-gatherer diets. This new a.n.a.lysis took into account, as Eaton and Konner's hadn't, the observation that hunter-gatherers consumed the entire carca.s.s of an animal, not just the muscle meat, and preferential y consumed the fattest parts of the carca.s.s -including organs, tongue, and marrow-and the fattest animals. Reversing the earlier conclusion, Eaton, Speth, and Cordain now suggested that Paleolithic diets were extremely high in protein (1935 percent of calories), low in carbohydrates "by normal Western standards" (2240 percent of energy), and comparable or higher in fat (2858 percent of energy). Eaton and his new col aborators stated with certainty that those relatively modern foods that today const.i.tute more than 60 percent of al calories in the typical American diet-cereal grains, dairy products, beverages, vegetable oils and dressings, and sugar and candy-"would have contributed virtual y none of the energy in the typical hunter-gatherer diet." This latest a.n.a.lysis makes it seem that what Rose and the public-health authorities considered biological normality in 1985-a relatively low-fat diet-would now have to be be considered abnormal.*22 The third critical caveat of Rose's logic is that it makes it effectively impossible to chal enge the underlying science once it is invoked to defend a particular hypothesis, one that is said to benefit the public health. Policy and the public belief are often set early in a scientific controversy, when the subject is most newsworthy. But that's when the evidence is by definition premature and the demand for clarification most urgent. As the evidence acc.u.mulates, it may cease to support the hypothesis, but altering the conventional wisdom by then can be exceedingly difficult. (The artificial sweetener saccharine is stil widely considered unhealthy, despite being absolved of any carcinogenic activity in humans over twenty years ago.) Rose's logic demonstrates why good science and public policy are often incompatible.

The fourth caveat is closely related. The philosophy of population-wide preventive medicine implies that the public health is not served by skepticism of the science or the reporting of contradictory evidence, both of which are essential to the process of science. A campaign to convince the public to embrace a public-health recommendation requires unconditional belief in the promised benefits. This was the motivation for creating the appearance of a consensus in the dietary-fat controversy and, as Arno Motulsky had told the Was.h.i.+ngton Post, for publis.h.i.+ng the National Academy of Sciences Diet and Health report as wel .

But if the underlying science is wrong-and that possibility is implied by the lack of a true consensus-then this tendency of public-health authorities to rationalize away al contradictory evidence wil make it that much harder to get the science right. Once these authorities insist that a consensus exists, they no longer have motivation to pursue further research. Indeed, to fund further studies is to imply that there is stil uncertainty. But the public's best interest wil be served only by the kind of skeptical inquiry and attention to negative evidence that are necessary to learn the truth. "If the public's diet is going to be decided by popularity pol s and with diminis.h.i.+ng regard for the scientific evidence," remarked Pete Ahrens in 1979, "I fear that future generations wil be left in ignorance of the real merits, as wel as the possible faults, in any given dietary regimen aimed at prevention of [coronary heart disease]."

Among the more conspicuous examples of the kind of scientific and social quagmire to which the logic of sick populations and preventive public health can lead is the proposition that dietary fat causes breast cancer. This possibility was suggested in 1976 in George McGovern's "Diet and Kil er Disease"

hearings, and then was cited in Dietary Goals for the United States as one reason why Americans should eat a low-fat diet (30-percent fat calories) as opposed to a cholesterol-lowering diet, in which the total fat content itself doesn't change. By 1982, the proposition that dietary fat causes cancer was considered so likely true that a National Academy of Sciences report ent.i.tled Diet, Nutrition, and Cancer not only recommended that Americans cut fat consumption to 30 percent, but noted that the evidence was sufficiently compel ing that it "could be used to justify an even greater reduction." In 1984, the American Cancer Society released its first cancer-fighting, low-fat-diet prescription, and then both The Surgeon General's Report on Nutrition and Health and Diet and Health embraced the hypothesis.

The proposition had emerged original y from the same international comparisons that led to Keys's fat/heart-disease hypothesis-in particular, low rates of breast cancer and low fat consumption in j.a.pan compared with high breast-cancer rates and high fat consumption in the United States.

Moreover, when j.a.panese women immigrate to the United States, their breast-cancer rates quickly rise and, by the second generation, are equal to those of other American ethnic groups. As fat consumption increased in j.a.pan from the 1950s to the early 1970s, breast-cancer rates there increased.

These a.s.sociations were given substance by the observation, original y made in the 1940s, that adding fat to the diet of laboratory rats promotes the growth of tumors, a phenomenon known technical y as fat-induced tumorigenesis.

Considerable evidence also argued against the hypothesis. As John Higginson, founding director of the International Agency for Research on Cancer, noted in 1979, the international comparisons were as contradictory as they were confirmatory. In urban Copenhagen, breast-cancer rates were four times higher than in rural Denmark, but fat consumption was 50 percent lower. Large-population studies in Framingham; Honolulu; Evans County, Georgia; Puerto Rico; and Malmo, Sweden, had al reported low cholesterol levels a.s.sociated with higher cancer rates. Since low cholesterol is al egedly the product of low-fat diets, it was "difficult to reconcile" this evidence, as the Framingham investigators noted in 1981, with the hypothesis that high-fat diets cause cancer.

The publication of the National Academy of Sciences report Diet, Nutrition, and Cancer in 1982 prompted the National Cancer Inst.i.tute and the NAS to make funding available to test the hypothesis. A critical test would come from the Nurses Health Study, led by the Harvard epidemiologist Walter Wil ett, which began tracking diet, lifestyle, and disease in nearly eighty-nine thousand nurses around the country in 1982. Such a prospective study is no subst.i.tute for a randomized clinical control trial, but it const.i.tutes the best that observational epidemiology can do. Wil ett and his col eagues published his first report on fat and breast cancer in January 1987 in The New England Journal of Medicine. Over six hundred cases of breast cancer had appeared among the eighty-nine thousand nurses over the first four years of the study. If anything, the less fat the women confessed to eating, the more likely they were to get breast cancer. In a New York Times article reporting the results of the study, Peter Greenwald, director of the National Cancer Inst.i.tute Division of Cancer Prevention, said that the Nurses Health Study was "a good study, but not the only one," and so NCI would continue to recommend -despite what was then, by far, the best evidence available-that Americans eat less fat to prevent breast cancer. Eight months later, NCI researchers themselves published the results of a study similar to the Nurses Health Study but smal er, also suggesting that eating more fat and more saturated fat correlated with less breast cancer. The NCI study went virtual y unnoticed, as Science later noted, "perhaps because no one wanted to hear the message that a promising avenue of research was turning into a blind al ey, and perhaps because it swam against the 'medical y political y correct' idea that fat is bad."

In 1992, Wil ett published the results from eight years of observation of the Nurses cohort. Fifteen hundred nurses had developed breast cancer, and, once again, those who ate less fat seemed to have more breast cancer. In 1999, the Harvard researchers published fourteen years of observations. By then almost three thousand nurses had contracted breast cancer, and the data stil suggested that eating fatty foods (even those with copious saturated fat) might protect against cancer. For every 5 percent of saturated-fat calories that replaced carbohydrates in the diet, the risk of breast cancer decreased by 9 percent. This certainly argued against the hypothesis that excessive fat consumption caused breast cancer.

Despite this acc.u.mulation of contradictory evidence, Peter Greenwald and the administrators at NCI refused to let their hypothesis die. This was Rose's philosophy at work. After Wil ett's publication of the first Nurses Health Study results, Greenwald and his NCI col eagues had responded with an article in JAMA ent.i.tled "The Dietary FatBreast Cancer Hypothesis Is Alive." The NCI administrators argued that any study that generated evidence refuting the hypothesis could be flawed. The existence of any positive evidence, they argued, even if it came from admittedly rudimentary studies-in other words, studies that almost a.s.suredly were flawed-was sufficient to keep such a critical hypothesis alive.

The only evidence that Greenwald and his col aborators considered "indisputable" was that laboratory rats fed "a high-fat, high-calorie diet have a substantial y higher incidence of mammary tumors than animals fed a low-fat, calorie-restricted diet." In this they were right, but they did not rule out the possibility that it was the calories or whatever caused weight gain (what they implied by the adjective "high-calorie") and not the dietary fat itself that was to blame, which was very likely the case. Even in 1982, when the authors of Diet, Nutrition, and Cancer had reviewed the animal evidence for fat-induced tumor growth, it had been less than indisputable. Adding fat to the diets of lab rats certainly induced tumors or enhanced their growth, but the most effective fats by far at this carcinogenesis process were polyunsaturated fats-saturated fats had little effect unless "supplemented with" polyunsaturated fats. This raised questions about the applicability of these observations to Western diets, which were traditional y low in polyunsaturated fats, at least until the 1960s, when the AHA started advocating polyunsaturated fats as a tool to lower cholesterol. Adding fat to rat chow also caused the rodents to gain weight, which was among the foremost reasons why obesity researchers came to believe that dietary fat caused human obesity. But it was hard to determine in these experiments whether it was the fat or the weight gain itself that led to increased tumor growth.

This laboratory evidence that dietary fat caused breast cancer began to evaporate as soon as Diet, Nutrition, and Cancer was published, and researchers could get funding to study it. By 1984, David Kritchevsky, one of the authors of Diet, Nutrition, and Cancer, had published an article in Cancer Research reporting on experiments that had been explicitly designed to separate out the effects of fat and calories on cancer, at least in rats. As Kritchevsky reported, low-fat, high-calorie diets led to more tumors than high-fat, low-calorie diets, and tumor production was shut down entirely in underfed rats, regardless of how fatty their diet was. Kritchevsky later reported that if rats were given only 75 percent of their typical daily calorie requirements, they could eat five times as much fat as usual and stil develop fewer tumors. Mike Pariza of the University of Wisconsin had published similar results in 1986 in the Journal of the National Cancer Inst.i.tute. "If you restrict calories just a little bit," Pariza later said, "you completely wipe out this so-cal ed fat enhancement of cancer." This observation has been confirmed repeatedly. Demetrius Albanes of the National Cancer Inst.i.tute later described the data as "overwhelmingly striking." And he added: "Those data have very largely been ignored and strongly downplayed."

By 1997 when the World Cancer Research Fund and American Inst.i.tute for Cancer Research released a seven-hundred-page report t.i.tled Food, Nutrition and the Prevention of Cancer, the a.s.sembled experts could find neither "convincing" nor even "probable" reason to believe that fat-rich diets increased the risk of cancer. A decade later stil , Arthur Schatzkin, chief of the nutritional epidemiology branch at the National Cancer Inst.i.tute, described the acc.u.mulated results from those trials designed to test the hypothesis as "largely nul ."

Nonetheless, the pervasive belief that eating fat causes breast cancer has persisted, partly because it once seemed undeniable. Purveyors of health advice just can't seem to let go of the notion. When the American Cancer Society released its nutrition guidelines for cancer prevention in 2002, the doc.u.ment stil recommended that we "limit consumption of red meats, especial y those high in fat," because of the same epidemiologic a.s.sociations that had generated the fat-cancer hypothesis thirty years earlier. By 2006, with the next release of cancer-prevention guidelines by the American Cancer Society, the ACS was acknowledging that "there is little evidence that the total amount of fat consumed increases cancer risk." But we were stil advised to eat less fat and particularly meats ("major contributors of total fat, saturated fat and cholesterol in the American diet"), because "diets high in fat tend to be high in calories and may contribute to obesity, which in turn is a.s.sociated with increased risks of cancers." (Saturated fats, in particular, the ACS added, "may have an effect on increasing cancer risk," a statement that seemed to be based solely on the belief that if saturated fat causes heart disease it probably causes cancer as wel .) Belief in the hypothesis persists also because of the time lag involved in research of this nature. In 1991, the National Inst.i.tutes of Health launched the $700 mil ion Women's Health Initiative to test the hypothesis (and also the hypothesis that hormone-replacement therapy protects against heart disease and cancer). The WHI investigators enrol ed forty-nine thousand women, aged fifty to seventy-nine. They randomly a.s.signed twenty-nine thousand to eat their usual diets, and twenty thousand were prescribed a low-fat diet. The goal was to induce these women to consume only 20 percent of their calories from fat; to do this, they were told to eat more vegetables and fresh fruits, as wel as whole grains, in case fiber was beneficial as wel . If the diet succeeded in preventing breast cancer, or any chronic disease, the WHI investigators wouldn't know if it was because these women ate less fat or because they ate more fruits, vegetables, and grains. It's conceivable that a diet of fruits, vegetables, grains, and more fat, or of vegetables and fruits but less grains, could be even more protective. The women on the diet also consumed fewer calories-averaging 120 calories a day less than the controls over the eight years of the study.*23 So, similarly, if this diet appeared to prevent cancer, the WHI investigators wouldn't know whether it did so because it contained less fat (or more fruits and vegetables) or fewer calories. To induce those on the diet to stick to it for the better part of a decade, the WHI investigators provided them with an intensive nutritional and behavioral-education program. The women a.s.signed to eat their usual diets received no such attention, which means they would be considerably less likely to change their lives in other ways that might also have an effect on breast cancer-to exercise or maintain their weight, stay away from sweets, refined flour, fast-food joints, and smoky bars. This disparity in counseling is known as an intervention effect, and it is precisely to avoid such an effect that drug trials must be done with placebos and double-blind.

Al of these effects would be expected to bias the trial in favor of observing a beneficial effect where none exists, but the WHI trial stil came up negative.

In the winter of 2006, the WHI investigators reported that those women who were eating what we today consider the essence of a healthy diet-little fat, lots of fiber, considerable fruits, vegetables, and whole grains, fewer calories-had no less breast cancer than those who ate their typical American fare.

(The women on the diet had no less heart disease, colon cancer, or stroke, either.) The results confirmed those of every study that had been done on diet and breast cancer since 1982. This, however, was stil not general y perceived as a definitive refutation of the hypothesis. Rose's logic of preventive medicine held fast (it stil does). In a press release on the findings, NHLBI Director Elizabeth Nabel stated, "The results of this study do not change established recommendations on disease prevention." In editorials that accompanied the WHI articles in JAMA, in virtual y every press report, and even in the World Health Organization's official statement on the trial, it was said that this particular study may have failed to show a beneficial effect of a low-fat, high-fiber diet on breast cancer (and heart disease, stroke, colon cancer, and weight), but that was not a reason to disbelieve the hypotheses. (The WHO press release was ent.i.tled "The World Health Organization Notes the Women's Health Initiative Diet Modification Trial, but Reaffirms That the Fat Content of Your Diet Does Matter.") Rather than enumerate the ways the WHI trial was biased to find a positive relations.h.i.+p, which was one facet of the controversy in the early 1990s over whether the trial should be funded to begin with, the WHI investigators and those like-minded now enumerated al the reasons why the study might have failed to find an effect.

At the core of al such ongoing scientific controversies is the inability to measure accurately the phenomenon at issue-the effect of dietary fat, for instance, on heart disease or cancer-either because it is negligible or nonexistent, or because the epidemiological tools available lack sufficient resolution for the task. Even clinical trials, unless done with meticulous attention to detail, double-blind, and placebo-control ed, cannot do the job. And if fat consumption has no effect whatsoever on heart disease or breast cancer, the available clinical and epidemiological tools wil always be incapable of demonstrating such a fact, because it is impossible in science to prove the nonexistence of a phenomenon. So the effect of saturated fat on heart disease-or the benefit of replacing saturated fat in the diet with carbohydrates or unsaturated fats-wil remain beyond the realm of science to demonstrate unambiguously. Investigators and public-health authorities wil continue to base their conclusions on their personal a.s.sessment of the totality of the data or the consensus of opinion among their col eagues.

One chal enge in this kind of controversy is to determine whether those skeptical of the established wisdom are incapable of accepting reality, closed-minded, or self-serving, or whether their skepticism is wel founded. In other words, is the evidence invoked to support the established wisdom the product of sound scientific thinking and reasonably unambiguous, in which case the skeptics are wrong, or is it what Francis Bacon would have cal ed "wishful science," based on fancies, opinions, and the exclusion of contrary evidence, in which case the skeptics are right to be so skeptical? Bacon offered one viable suggestion for differentiating. Good science, he observed, is rooted in reality, and so it grows and develops, and the evidence grows increasingly more compel ing, whereas wishful sciences remain "stuck fast in their tracks," or "rather the reverse, flouris.h.i.+ng most under their first authors before going downhil ever since."

Wishful science eventual y devolves to the point where it is kept alive simply by the natural reluctance of its advocates to recognize or acknowledge error, rather than compel ing evidence that it is right. "These are cases where there is no dishonesty involved," explained the n.o.bel Prizewinning chemist Irving Langmuir in a celebrated 1953 lecture, "but where people are tricked into false results by a lack of understanding about what human beings can do to themselves in the way of being led astray by subjective effects, wishful thinking or threshold interactions." Whereas good science would blossom over time, Langmuir noted, this "pathological science" would not. The most concise statement of this philosophy may be an unwritten rule of experimental physics credited original y to Wolfgang Panofsky, a former Manhattan Project physicist and presidential science adviser. "If you throw money at an effect and it doesn't get larger," Panofsky said, "that means it is not real."

That has certainly been the case with the dietary-fat/breast-cancer hypothesis. The relations.h.i.+p between dietary fat, cholesterol, and heart disease is more complicated, because the hypothesis const.i.tutes three independent propositions: first, that lowering cholesterol prevents heart disease; second, that eating less fat or less saturated fat not only lowers cholesterol and prevents heart disease but, third, that it prolongs life.

Since 1984, the evidence that cholesterol-lowering drugs, particularly those known as statins, are beneficial-proposition number one-has certainly blossomed, particularly regarding people at high risk of heart attack. These drugs reduce serum-cholesterol levels dramatical y, and they seem to prevent heart attacks, although whether they actual y do so by lowering cholesterol levels or by other means as wel is stil an open question. ("Most drugs have multiple actions," notes the University of Was.h.i.+ngton biostatistician Richard Kronmal. Saying that statins reduce heart-disease risk by lowering cholesterol, he adds, is like "saying that aspirin reduces heart-disease risk by reducing headaches.") There is also a legitimate question as to whether they wil prolong the life of anyone who is not in imminent danger of having a heart attack, but new trials consistently seem to confirm their benefits. Al this may be irrelevant to the question of a healthy diet, however, because there is no compel ing reason we should believe that a drug and a diet wil have equivalent effects on our health, even if they both happen to lower cholesterol.

The evidence supporting the second and third propositions-that eating less fat, or less saturated fat, makes for a healthier and longer life-has remained stubbornly ambiguous. The message of the 1984 consensus conference and the ensuing expert reports was that the benefits of low-fat diets were effectively indisputable, and so pursuing further research on these questions was unnecessary. This in turn led to the ubiquitous belief in the validity of Keys's hypothesis and the unwholesome nature of saturated fat, but the reality is that since the early 1980s the evidence has become progressively less compel ing.

Keys's own experience stands as an example. In the early 1950s, Keys had based his dietary-fat hypothesis of heart disease to a great extent on the congruence between the changing-American-diet story and the appearance of a heart-disease epidemic. By the early 1970s, however, he had publicly acknowledged that the heart-disease epidemic may indeed have been a mirage. There was "no basis" to make the claim, he admitted, that trends in heart-disease mortality in the United States reflect changes in the consumption of any item in the diet.

In the late 1950s, Keys supported his fat hypothesis with the disparity in fat consumption, cholesterol levels, and heart-disease mortality he found among j.a.panese men living in j.a.pan, Hawaii, and Los Angeles. This a.s.sociation was then confirmed, more or less, in his Seven Countries Study, in which the j.a.panese vil agers stil had remarkably little fat in their diets, low cholesterol levels, and fewer heart-disease deaths over ten years than any other population with the exceptions of those of the islands of Crete and Corfu and the vil age of Velika Krsna in what is now Serbia. By the mid-1990s, however, the j.a.panese contingent of the Seven Countries Study, led by Yos.h.i.+nori Koga, reported that fat intake in j.a.pan had increased from the 6 percent of calories they had measured in the farming vil age of Ta.n.u.s.h.i.+maru thirty-five years earlier, to 22 percent of calories. "There have been progressive increases in consumption of meats, fish and shel fish and milk," they reported. Mean cholesterol levels rose in the community from 150 mg/dl to nearly 190 mg/dl, which is only 6 percent lower than the average American values (202 mg/dl as of 2004). Yet this change went along with a "remarkable reduction" in the incidence of strokes and no change in the incidence of heart disease. In fact, the chance that a j.a.panese man of any particular age would die of heart disease had steadily diminished since 1970. "It is suggested that dietary changes in Ta.n.u.s.h.i.+maru in the last thirty years have contributed to the prevention of cardiovascular disease," Koga and his col eagues concluded.

In the late 1950s, Keys had dismissed the possibility that misdiagnosis might have contributed to the extremely low heart-disease death rates in j.a.pan they had observed initial y. In 1984, Keys reversed himself, saying that the j.a.panese cardiologists who had worked with his Seven Countries Study "might have been misled by the local physicians who signed the death certificates and provided details."

Three years later, Keys acknowledged to the New York Times that he had re-evaluated his hypothesis. "I've come to think that cholesterol is not as important as we used to think it was," he said, "Let's reduce cholesterol by reasonable means, but let's not get too excited about it."

As in j.a.pan, increases in fat consumption with coincident decreases in heart disease have occurred recently in Spain and Italy, which has prompted the observation that the French paradox-a nation that eats a high-fat diet and has little heart disease-has evolved into the French-Italian-Spanish paradox.*24 Through the mid-1990s, according to John Powles, an epidemiologist with the British Inst.i.tute of Public Health, France and Italy both showed declines in death rates from stroke and heart disease that were greater than those in most European countries, while the decline in mortality in Spain lagged only slightly behind. And studies of Mediterranean immigrants to Australia suggest that the low heart-disease rates of these immigrants fal even lower in Australia, despite a considerable increase in their meat consumption.

In the late 1970s, the World Health Organization launched a research project known as MONICA, for "MONItoring CArdiovascular disease," that was similar in concept to Keys's Seven Countries Study but considerably larger. The study tracked heart disease and risk factors in thirty-eight populations in twenty-one countries-a total population of roughly six mil ion people, which unlike previous studies included both men and women. Hugh Tunstal -Pedoe, the MONICA spokesman, has described the project as "far and away the biggest international col aborative study of cardiovascular disease ever carried out" and noted that, "whatever the results, n.o.body else has better data." By the late 1990s, MONICA had recorded 150,000 heart attacks and a.n.a.lyzed 180,000 risk-factor records. Its conclusion: heart-disease mortality was declining worldwide, but that decline was independent of cholesterol levels, blood pressure, or even smoking habits.

The MONICA investigators suggested reasons why their study might not have confirmed Keys's hypothesis, among them the possibility, as Tunstal - Pedoe noted, that with populations "the contribution of cla.s.sical risk factors is swamped by that of other dietary, behavioral, environmental, or developmental factors." He also discussed something that may have contributed initial y to the widespread belief in Keys's hypothesis: the tendency to publish or pay attention to only that evidence that confirms the existing beliefs about heart disease and risk factors. "If you do a study in your population and you show a perfect correlation between risk factors and heart disease, you rush off and publish it. If you don't, unless you have great confidence in yourself, you worry that perhaps you didn't measure something properly, or perhaps you'd better keep quiet, or perhaps there's something you haven't thought about. And by doing this, there is a risk of myths' becoming self-perpetuating." "There are people," Tunstal -Pedoe said, "who want to believe that if we find anything less than 100-percent correlation between traditional risk factors and trends in heart disease, we are somehow traitors to the cause of public health, and what we say should be suppressed, and we should be ashamed of ourselves. Whereas we are asking a perfectly reasonable question, and we came up with results. That is what science is about."

In the two decades since the NIH, the surgeon general, and the National Academy of Sciences first declared that al Americans should consume low-fat diets, the research has also failed to support the most critical aspect of this recommendation: that such diets wil lead to a longer and healthier life. On the contrary, it has consistently indicated that these diets may cause more harm than good. In 1986, the year before the National Cholesterol Education Program recommended cholesterol-lowering for every American with cholesterol over 200 mg/dl, the University of Minnesota epidemiologist David Jacobs visited j.a.pan, where he learned that j.a.panese physicians were advising patients to raise their cholesterol, because low cholesterol levels were linked to hemorrhagic stroke. At the time, j.a.panese men were dying from stroke almost as frequently as American men were succ.u.mbing to heart disease. Jacobs looked for this inverse relations.h.i.+p between stroke and cholesterol in the MRFIT data and found it there, too. And the relations.h.i.+p transcended stroke: men with very low cholesterol seemed p.r.o.ne to premature death; below 160 mg/dl, the lower the cholesterol, the shorter the life.

In April 1987, the Framingham investigators provided more reason to worry when they final y published an a.n.a.lysis of the relations.h.i.+p between cholesterol and al mortality. After thirty years of observation, there was a significant a.s.sociation between high cholesterol and premature death for men under fifty. But for those over fifty, both men and women, life expectancy showed no a.s.sociation with cholesterol. This suggested, in turn, that if low cholesterol did prevent heart disease, then it must raise the risk of dying from other causes.

This was compounded by what may have been the single most striking result in the history of the cholesterol controversy, although it pa.s.sed without comment by the authorities: those Framingham residents whose cholesterol declined over the first fourteen years of observation were more likely to die prematurely than those whose cholesterol remained the same or increased. They died of cardiovascular disease more frequently as wel . The Framingham investigators rejected the possibility that the drop in cholesterol itself was diet-related-the result of individuals' fol owing AHA recommendations and eating low-fat diets. Instead, they described it as a "spontaneous fal ," and insisted that it must be caused by other diseases that eventual y led to death, but they offered no evidence to support that claim.

The a.s.sociation between low cholesterol and higher mortality prompted administrators at the National Heart, Lung, and Blood Inst.i.tute once again to host a workshop and discuss it. Researchers from nineteen studies around the world met in Bethesda, Maryland, in 1990 to report their results. The data were completely consistent (see charts on fol owing page): when investigators tracked al deaths, not just heart-disease deaths, it was clear that men with cholesterol levels above 240 mg/dl tended to die prematurely because of their increased risk of heart disease. Those whose cholesterol was below 160 mg/dl tended to die prematurely with an increased risk of cancer, respiratory and digestive diseases, and trauma. As for women, if anything, the higher their cholesterol, the longer they lived.*25 The proponents of Keys's hypothesis said the results could not be meaningful. The excess deaths at low cholesterol levels had to be due to pre-existing conditions; chronic il ness leads to low cholesterol, they concluded, not vice versa, and then the individuals die from the il nesses, which confuses the mortality issue. This was the a.s.sumption the Framingham researchers had made. At the one end of the population distribution of cholesterol, low cholesterol is the effect and disease is the cause. At the other end of the distribution, high cholesterol is the cause and disease is the effect. This, of course, is a distinction based purely on a.s.sumptions rather than actual evidence, and one consistent with the universal recommendations to lower cholesterol by diet. When NIH Administrator Basil Rifkind offered this interpretation during my interview with him in 1999, he pointed to the report of the 1990 conference as the definitive doc.u.ment in support of it. But the report, which Rifkind co-auth.o.r.ed, states unequivocal y that this interpretation was not supported by the available evidence.

The relations.h.i.+p between blood cholesterol (horizontal axes) and all deaths (total mortality) or just heart disease deaths, as reported in a 1990 NIH conference.

In an alternate interpretation, both ends of the cholesterol distribution are treated identical y. Whether high or low, either our cholesterol levels directly increase mortality or they're a symptom of an underlying disorder that itself increases our risk of disease and death. In both cases, diet leads to disease, although whether it does so directly, via its effect on cholesterol, or through other mechanisms would stil be an open question. In this interpretation, what a cholesterol-lowering diet does to cholesterol levels, and what that in turn does to arteries, may be only one component of the diet's effect on health. So lowering cholesterol by diet might help prevent heart disease for some individuals, but it might also raise susceptibility to other conditions-such as stroke and cancer-or even cause them. This is what had always worried those investigators who were skeptical of Keys's hypothesis. "Questions should be pursued about biological mechanisms that might help explain low [total cholesterol]: disease a.s.sociations," noted the report from the 1990 NHLBI workshop. Nonetheless, public-health recommendations to eat low-fat diets and lower cholesterol would remain inviolate and unconditional.

In 1964, when the physicist Richard Feynman presented what would become a renowned series of lectures at Cornel University, he observed that it was a natural condition of scientists to be biased or prejudiced toward their beliefs. That bias, Feynman said, would ultimately make no difference, "because if your bias is wrong a perpetual acc.u.mulation of experiments wil perpetual y annoy you until they cannot be disregarded any longer." They could be disregarded, he said, only if "you are absolutely sure ahead of time" what the answer must be.

In the case of Keys's hypothesis, the annoying evidence was consistently disregarded from the beginning. Because the totality of evidence was defined as only those data that confirmed the hypothesis, Keys's hypothesis would always appear monolithic. Annoying observations could not force a rea.n.a.lysis of the underlying a.s.sumptions, because each of those observations would be discarded immediately as being inconsistent with the totality of the evidence. This was a self-fulfil ing phenomenon. It was unlikely, however, to lead to reliable knowledge about either the cause of heart disease or the routes to prevention. It did not mean the hypothesis was false, but its truth could never be established, either.

One other method can be employed to judge the validity of the hypotheses that dietary fat or saturated fat causes heart disease, and that cholesterol-lowering diets prevent it. This is a technique known as meta-a.n.a.lysis, viewed as a kind of last epidemiological resort in these kinds of medical and public-health controversies: if the existing studies give ambiguous results, the true size of a benefit or harm may be a.s.sessed by pooling the data from al the studies in such a way as to gain what's known as statistical power. Meta-a.n.a.lysis is controversial in its own right. Investigators can choose, for instance, which studies to include in their meta-a.n.a.lysis, either consciously or subconsciously, based on which ones are most likely to give them the desired result.

For this reason, a col aboration of seventy-seven scientists from eleven countries founded the Cochrane Col aboration in 1993. The founders, led by Iain Chalmers of Oxford University, believed that meta-a.n.a.lyses could be so easily biased by researchers' prejudices that they needed a standardized methodology to minimize the influence of such prejudice, and they needed a venue that would al ow for the publication of impartial reviews. The Cochrane Col aboration methodology makes it effectively impossible for researchers to influence a meta-a.n.a.lysis by the criteria they use to include or exclude studies. Cochrane Col aboration reviews must include al studies that fit a prespecified set of criteria, and they must exclude al that don't.

In 2001, the Cochrane Col aboration published a review of "reduced or modified dietary fat for preventing cardiovascular disease." The authors combed the literature for al possibly relevant studies and identified twenty-seven that were performed with sufficient controls and rigor to be considered meaningful.*26 These trials encompa.s.sed some ten thousand subjects fol owed for an average of three years each. The review concluded that the diets, whether low-fat or cholesterol-lowering, had no effect on longevity and not even a "significant effect on cardiovascular events." There was only a "suggestion" of benefit from the trials lasting more than two years. In 2006, the Cochrane Col aboration published a review of multiple-risk-factor interventions-including lowering blood pressure and cholesterol-for the prevention of coronary heart disease. In this case, thirty-nine trials were identified of which ten (comprising over nine-hundred thousand patient years of observation) included sufficient data and were carried out with sufficient rigor to draw meaningful inferences. "The pooled effects suggest multiple risk factor intervention has no effect on mortality," the authors concluded.

Although, once again, a "smal " benefit of treatment, perhaps "a 10 percent reduction in CHD mortality," may have been missed, they added.

If we believe in Rose's philosophy of preventive medicine, this suggestion of benefit or the possibility that even a "smal " benefit was missed stil const.i.tutes sufficient motivation to advocate cholesterol-lowering diets to the entire population, as indeed the authors of the first Cochrane review suggested. We could also a.s.sume that if a suggestion of a benefit can be induced after two years on such a diet, we might do considerably better after ten or twenty years, although we would stil need trials to test that a.s.sumption.

We might also compare this conclusion to the original predictions of Keys's hypothesis in the mid-1950s. When Keys first suggested that eating fat caused heart disease, as we discussed, he did so partly on the basis of the experience in wartime Europe, where food shortages of a few years' duration coincided with dramatic decreases in the incidence of heart disease. Keys had attributed those decreases to the reduced availability of meat, eggs, and dairy products. Other investigators pointed out that the war changed many other aspects of diet and lifestyle. Mortality from infectious diseases, diabetes, tuberculosis, and cancer al dropped during the war. Stil , to Keys it was the fat, particularly saturated fat, that was crucial. "A major lesson gained from World War I ," he wrote in 1975, "is the proof that in a very few years the incidence of CHD [coronary heart disease] could drop to a level of the order of one-fourth the preceding rate." If this were indeed possible, or even vaguely possible based on the col ective European experience during World War I , as the evidence indeed suggested, then something considerably more profound had been happening than was reflected in only the "suggestion" of a reduction in mortality seen in the clinical trials of cholesterol-lowering diets. Other factors of diet or lifestyle that had changed during wartime must have played far more significant roles in improving the health of the populations.

Part Two

THE CARBOHYDRATE HYPOTHESIS.

The world is gradual y going carbohydrate. That is because there are more people than there have ever been before (one would like to add "or ever wil be again") so there must be more food. You can get about eight times as many calories from an acre of corn as you can from the flesh of pigs fed on this same corn. Because of population pressure, certain sections of the world are progressively using more of the vegetable and less of animal materials. This means that the carbohydrates, from sugar and cereals particularly, are increasing steadily in quant.i.ty. One does not need to view this with alarm, but it is not amiss to point out that this tendency is not the best road to health. Not that starches and sugars are harmful, but they are low in the essentials we must have for good health. If the proportion of carbohydrates is high then the amount of something else of greater importance is low. Nutrition is a six-way teeter-totter. Have you ever tried to balance such a device?

C. C. FURNAS AND S. M. FURNAS, Man, Bread & Destiny: The Story of Man and his Food, 1937

Chapter Five.

DISEASES OF CIVILIZATION.

The potato took 200250 years, in spite of organized encouragement, to become accepted in England. It took only fifty years in Ireland. Maize and ca.s.sava have come to be accepted in parts of Africa in considerably less time.... Tea, white bread, rice and soft drinks have entered many African dietaries in even shorter time and the extent to which they have spread and their consequences to nutrition have been rather severe.

F. T. SAI, Food and Agricultural Organization regional nutrition officer for Africa, 1967 ON APRIL 16, 1913, ALBERT SCHWEITZER arrived at Lambarene, a smal vilage in the interior lowlands of West Africa, to establish a missionary hospital on the banks of the Ogowe River. Attended by his wife, Helene, who had trained as a nurse, he began treating patients the very next morning. Schweitzer estimated that he saw almost two thousand patients in the first nine months, and then averaged thirty to forty a day and three operations a week for the better part of four decades. The chief complaints, at least in the beginning, were endemic diseases and infections: malaria, sleeping sickness, leprosy, elephantiasis, tropical dysentery, and scabies.

Forty-one years after Schweitzer's arrival, and a year and a half after he received the n.o.bel Peace Prize for his missionary work, Schweitzer encountered his first case of appendicitis among the African natives. Appendicitis was not the only Western disease to which the natives seemed to be resistant. "On my arrival in Gabon," he wrote, "I was astonished to encounter no cases of cancer.... I can not, of course, say positively that there was no cancer at al , but, like other frontier doctors, I can only say that if any cases existed they must have been quite rare." In the decades that fol owed, he witnessed a steady increase in cancer victims. "My observations inclined me to attribute this to the fact that the natives were living more and more after the manner of the whites."

As Schweitzer had suggested, his experience was not uncommon for the era. In 1902, Samuel Hutton, a University of Manchestertrained physician, began treating patients at a Moravian mission in the town of Nain, on the northern coast of Labrador, or about as far from the jungles of West Africa as can be imagined, in both climate and the nature of the indigenous population. As Hutton told it, his Eskimo patients fel into two categories: There were those who lived isolated from European settlements and ate a traditional Eskimo diet. "The Eskimo is a meat eater," he wrote, "the vegetable part of his diet is a meager one." Then there were those Eskimos living in Nain or near other European settlers who had taken to consuming a "settler's dietary,"

consisting primarily of "tea, bread, s.h.i.+p's biscuits, mola.s.ses, and salt fish or pork." Among the former, European diseases were uncommon or remarkably rare. "The most striking is cancer," noted Hutton on the basis of his eleven years in Labrador. "I have not seen or heard of a case of malignant growth in an Eskimo." He also observed no asthma and, like Schweitzer, no appendicitis, with the sole exception of a young Eskimo who had been "living on a 'settler'

dietary." Hutton observed that the Eskimos who had adopted the settlers' diet tended to suffer more from scurvy, were "less robust," and endured "fatigue less easily, and their children are puny and feeble."

What both Schweitzer and Hutton had witnessed during their missionary years was a "nutrition transition," a term now commonly used to describe a population's Westernization in diet, lifestyle, and health status. The World Health Organization recently described the current version of the nutrition transition this way: Changes in the world food economy have contributed to s.h.i.+fting dietary patterns, for example, increased consumption of an energy-dense diet high in fat, particularly saturated fat, and low in carbohydrates. This combines with a decline in energy expenditure that is a.s.sociated with a sedentary lifestyle.... Because of these changes in dietary and lifestyle patterns, diet-related diseases-including obesity, diabetes mel itus, cardiovascular disease, hypertension and stroke, and various forms of cancer-are increasingly significant causes of disability and premature death in both developing and newly developed countries.

This is little more than an updated version of the changing-American-diet story Ancel Keys and others had invoked to advocate low-fat diets: we eat fewer carbohydrates and ever more fat then we did in some idealized past, and we pay the price in chronic disease. Keys's reference point was the American diet circa 1909 (as portrayed by USDA estimates), or the j.a.panese or Mediterranean diets of the 1950s. When it was suggested to Keys that other nutrition transitions, including those witnessed by Schweitzer and Hutton, could be edifying, he argued that not enough was known about the diets or about the health of those isolated populations for us to draw reliable conclusions. He also insisted that in many of these populations-particularly the Inuit -relatively few individuals were likely to live long enough to develop chronic disease, so little could be learned.

This argument, too, has taken on the aura of