according to an NIH report ent.i.tled The Pima Indians: Pathfinders for Health, "people became more overweight." "If the Pima Indians could return to some of their traditions," explained one NIH authority, "including a high degree of physical activity and a diet with less fat and more starch, we might be able to reduce the rate, and surely the severity, of unhealthy weight in most of the population."

The problem with this version of the Pima history is that obesity and overweight had been evident a century ago, when the relevant nutrition transition was from relative abundance to extreme poverty. From November 1901 to June 1902, the Harvard anthropologist Frank Russel lived on the Pima reservation south of Phoenix studying the tribe and its culture. Many of the older Pima, Russel noted in a report of the Bureau of American Ethnology, "exhibit a degree of obesity that is in striking contrast with the 'tal and sinewy' Indian conventionalized in popular thought."

Obesity among the Pima is not a new phenomenon, as demonstrated by this photo of "Fat Louisa" taken in 1901 or 1902 by the Harvard anthropologist Frank Russell.

Russel 's a.s.sessment of the Pima's relative corpulence was then confirmed by the anthropologist and physician Ale Hrdli ka, who visited the Pima reservation in 1902 and 1905. "Especial y wel -nourished individuals, females and also males, occur in every tribe and at al ages," Hrdli ka reported, "but real obesity is found almost exclusively among the Indians on reservations."

For perhaps two mil ennia, the Pima had lived as both hunter-gatherers and agriculturalists. Game was abundant in the region, as were fish and clams in the Gila River. When the Jesuit missionary Eusebio Kino arrived among the Pima in 1787, the tribe was already raising corn and beans on fields irrigated with Gila River water. In the decades that fol owed, they took to raising cattle, poultry, wheat, melons, and figs. They also ate mesquite beans, the fruit of the saguaro cactus, and a mush of what Russel later cal ed "unidentified worms." In 1846, when a U.S. Army battalion pa.s.sed through Pima lands, the battalion's surgeon John Griffin described the Pima as "sprightly" and in "fine health." He also noted that the Pima had "the greatest abundance of food, and take care of it wel , as we saw many of their storehouses ful of pumpkins, melons, corn &c."

Life began to change dramatical y the fol owing year, when a wagon route was opened to California "by way of Tucson and the Pima vil ages." This became the southernmost overland route for the California gold rush that began in 1849; tens of thousands of travelers pa.s.sed through the Pima vil ages on the way west over the next decade. They relied on the Pima for food and supplies.

With the arrival of Anglo-American and Mexican settlers in the late 1860s, the prosperity of the Pima came to an end, replaced by what the tribe referred to as "the years of famine." Over the next quarter-century, these newcomers hunted the local game almost to extinction, and the Gila River water, on which the Pima depended for fishing and irrigating their own fields, was "entirely absorbed by the Anglo settlements upstream." By the mid-1890s, the Pima were relying on government rations to avoid starvation, and this was stil the situation when Hrdli ka and Russel arrived in the early 1900s.

Both Hrdli ka and Russel struggled with the dilemma of poverty coincident with obesity. Russel knew that the life of these Indians was arduous; sedentary behavior could not be a cause of obesity in the Pima. Instead, he proposed that a dietary factor was responsible. "Certain articles of their diet appear to be markedly flesh producing," Russel wrote. Hrdli ka suggested that "the role played by food in the production of obesity among the Indians is apparently indirect." He suggested that life on the reservation might be relatively sedentary and this could play a role-"the change from their past active life to the present state of not a little indolence"-but he did not appear particularly confident about it. After al , he wrote, obesity was quite rare among the Pueblo, "who have been of sedentary habits since ancient times." And obesity among the Pima was found "largely but not exclusively" in the women, and the women of the tribe worked considerably harder than the men, spending their days harvesting the crops, grinding corn, wheat, and mesquite beans and carrying whatever burdens were not carried by pack animals.

Hrdli ka also noted that by 1905 the Pima diet already included "everything obtainable that enters into the dietary of the white man," which raises the possibility that this might have been responsible for the obesity. At the half-dozen trading posts that opened on the Pima reservation after 1850, the Indians took to buying "sugar, coffee and canned goods to replace traditional foodstuffs lost ever since whites had settled in their territories."

Neither Hrdli ka nor Russel suggested that the U.S. government rations might be the cause of obesity. But if the Pima diet on government rations was anything like that of tribes reduced to similar situations at the time on which data exist-including the Sioux on the Standing Rock Reservation in the Dakotas-then almost 50 percent of their calories came from sugar and flour.

Obesity in a.s.sociation with "widespread poverty" was doc.u.mented again on the Pima reservation in the early 1950s by Bertram Kraus, a University of Arizona anthropologist working with the Bureau of Indian Affairs. According to Kraus, more than 50 percent of the children on the Pima reservation could legitimately be described as obese by their eleventh birthday. The local Anglos, Kraus wrote, got leaner as they got older (at the time, at least); this was not the case with the Pima. Kraus lamented the absence of dietary data to a.s.sess the nutritional state of the tribe, but this situation was remedied a few years later by Frank Hesse, a physician at the Public Health Service Indian Hospital on the Gila River Reservation. Hesse noted that the Pima diet of the mid-1950s was remarkably consistent from family to family and consisted of "mainly beans, tortil as, chili peppers and coffee, while oatmeal and eggs are occasional y eaten for breakfast. Meat and vegetables are eaten only once or twice a week." Hesse neglected to a.s.sess sugar consumption, but he did note that "a large amount of soft drinks of al types is consumed between meals." Hesse then concluded that 24 percent of the calories consumed by the Pima (the soft drinks not included) were from fat, which is certainly low by modern standards.*69 Over the next twenty years, the prevalence of obesity and diabetes among the Pima continued to rise, now coincident with a change in the foods distributed by government agencies and sold in the reservation trading posts. By the late 1950s, according to the Indian Health Service in Tucson, "large quant.i.ties of refined flour, sugar, and canned fruits high in sugar" were being distributed widely on the reservations, courtesy of a surplus commodity food program run by the U.S. Department of Agriculture. When mechanization of the local agriculture industry brought a cash economy to the Pima, the local stores and trading posts "started to carry high caloric pre-packed sweets, such as carbonated beverages (i.e., 'soda pop'), candy, potato chips, and cakes." "Soda pop is used in immense amounts," as one 1962 study described it.

In April 1973, when the evils of dietary fat were stil widely considered hypothetical, the NIH epidemiologist Peter Bennett appeared before George McGovern's Senate Select Committee on Nutrition and Human Needs to discuss diabetes and obesity on the Pima reservation. The simplest explanation for why half of al adult Pima were diabetic, said Bennett, was the amount of sugar consumed, which represented 20 percent of the calories in the Pima diet. "The only question that I would have," Bennett had 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." Bennett's opinion was consistent with that of Henry Dobyns of the D'Arcy McNickle Center for the History of the American Indian, who is considered the foremost authority on Pima history. In 1989, Dobyns described obesity and diabetes in the tribe as being "to some extent a result of inadequate nutrition" and added that this inadequate nutrition had come about because "many of the poorer individuals subsist on a diet of potatoes, bread, and other starchy foods. Their traditional diet is beyond their reach, for they cannot catch fish in a dry riverbed and they cannot afford to buy much meat or many fresh fruits and vegetables."

Studies of the Sioux of the South Dakota Crow Creek Reservation in the 1920s, Arizona Apaches in the late 1950s, North Carolina Cherokees in the early 1960s, and Oklahoma tribes in the 1970s al reported levels of obesity comparable to that in the United States today, but in populations living in extreme poverty. "Men are very fat, women are even fatter," as the University of Oklahoma epidemiologist Kel y West said of the local tribes of the 1970s.

"Typical y, their lifetime maximum weight has been 185 percent of standard."

The early study of the Sioux, by two investigators from the University of Chicago, is particularly interesting, because it was one of the few published studies of diet, health, and living conditions in such a population, and it appeared the same year that the U.S. Department of the Interior released the results of a lengthy investigation of Native American living conditions. "An overwhelming majority of the Indians are poor, even extremely poor," the Interior Department reported, "living on lands from which a trained and experienced white man could scarcely wrest a reasonable living." The University of Chicago report said most of the Sioux lived in one-or two-room shacks; 40 percent of the children lived in homes without toilet facilities; water had to be hauled from the river. Little milk was consumed, although canned milk was included in the government rations. b.u.t.ter, green vegetables, and eggs were almost never eaten. No fruit was consumed.*70 Twenty-five to forty pounds of beef were issued per person as government rations each month, but this was "not an indication of the amount consumed by each person," the report noted, "for the families who receive rations are not left alone to eat them. Issue day is visiting day for the families not on the ration rol , and often the visit lasts until the friends' or relatives' rations of meat are gone. The ration family, therefore, may be compel ed to live on bread and coffee for the remainder of the month."

The staple of the Sioux diet on the reservation was "grease bread," fried in fat and made from white flour, supplemented by oatmeal, potatoes, and beans, some squash and canned tomatoes, black coffee, canned milk, and sugar. "Almost two-thirds of the families, including 138 children, were receiving distinctly inadequate diets," the report concluded. Fifteen families, with thirty-two children among them, "were living chiefly on bread and coffee."

Nonetheless, 40 percent of the adult women, over 25 percent of the men, and 10 percent of the children "would be termed distinctly fat," the University of Chicago investigators reported, whereas 20 percent of the women, 25 percent of the men, and a slightly greater percentage of the children were "extremely thin."

By the 1970s, when studies of obesity in populations began in earnest, the general att.i.tude was that obesity was simply a fact of life in developed nations. "Even a brief visit to Czechoslovakia," reported a Prague epidemiologist at the first International Conference on Obesity, in 1974, "would reveal that obesity is extremely common and that, as in other industrial countries, it is probably the most widespread form of malnutrition."

The observation that this was also true in poor populations in nonindustrialized countries, that obesity frequently coexists side-by-side with malnutrition and undernutrition, shows up with surprising consistency. In a 1959 study of African Americans living in Charleston, South Carolina, nearly 30 percent of the adult women and 20 percent of the adult men were obese although living on family incomes of from $9 to $53 a week. In Chile in the early 1960s, a study of factory workers, most of whom were engaged in "heavy labor," revealed that 30 percent were obese and 10 percent suffered from "undernourishment." Nearly half the women over forty-five were obese. In Trinidad, a team of nutritionists from the United States reported in 1966 that one-third of the women older than twenty-five were obese, and they achieved this condition eating fewer than two thousand calories a day-an amount lower than the United Nations' Food and Agriculture Organization recommendation to avoid malnutrition. Only 21 percent of the calories in the diet came from fat, compared with 65 percent from carbohydrates.

In Jamaica, high rates of obesity, again among adult women in particular, were first reported in the early 1960s by a British Medical Research Council diabetes survey. By 1973, according to Rolf Richards of the University of the West Indies, Kingston, 10 percent of al Jamaican men and nearly two-thirds of the women were obese in a society in which "malnutrition in infancy and early childhood remains one of the most important disorders contributing to infant and childhood mortality."

Similar observations were made in the South Pacific and throughout Africa. In Rarotonga in the South Pacific, for instance, in the mid-1960s, on a diet of only 25 percent fat, over 40 percent of the women were obese and 25 percent were "grossly obese." Among Zulus living in Durban, South Africa, according to a 1960 report, 40 percent of adult females were obese. Women in their forties averaged 175 pounds. In a population of urban Bantu "pensioners," the mean weight of women over the age of sixty was reported in the mid-1960s to be 165 pounds. "Although dietary habits vary widely amongst the African countries, tribes and vil ages," wrote B. K. Adadevoh from Nigeria's University of Ibadan in 1974, "it is general y established that the African diet is rich in carbohydrates. Caloric intake for most is low and protein fal s short of the recommended al owance."

It seems fair to a.s.sume that the lives of market women in West Africa in the 1960s or poor Jamaicans of the same era were nontoxic by any of the definitions that are commonly a.s.sociated with the current obesity epidemic. The Sioux of the mid-1920s, or the Pima of the 1900s or 1950s, living on reservations and relying on government rations to survive, clearly lived in a state of poverty that most of us today would find almost unimaginable.

Obesity in Africa is not a.s.sociated with prosperity. These photos from Nigeria, of market women and an obese eleven-year-old, date to the early 1970s.

So why were they fat? "It is difficult to explain the high frequency of obesity seen in a relatively impecunious society such as exists in the West Indies, when compared to the standard of living enjoyed in the more developed countries," Rolf Richards wrote about Jamaica in the 1970s. "Malnutrition and subnutrition are common disorders in the first two years of life in these areas, and account for almost 25 per cent of al admissions to pediatric wards in Jamaica. Subnutrition continues in early childhood to the early teens. Obesity begins to manifest itself in the female population from the 25th year of life and reaches enormous proportions from 30 onwards."

The question of what causes obesity in these impoverished populations has typical y been ignored by obesity researchers, other than to suggest that there is something unique about given groups of people that exacerbates the problem of obesity. The a.s.sumption, as The New Yorker writer Malcolm Gladwel wrote about the Pima in 1998, is that they are "different only in degree, not in kind."

The idea of specific populations predisposed to obesity is encapsulated in a notion now known as the thrifty gene-technical y, the thrifty-genotype hypothesis-that is now commonly invoked to explain the existence of the obesity epidemic and why we might all gain weight easily during periods of prosperity but have such difficulty losing it. The idea, initial y proposed in 1962 by the University of Michigan geneticist James Neel, is that we are programmed by our genes to survive in the paleolithic hunter-gatherer era that encompa.s.sed the two mil ion years of human evolution before the adoption of agriculture-a mode of life stil lived by many isolated populations before extensive contact with Western societies. "Such genes would be advantageous under the conditions of unpredictably alternating feast and famine that characterized the traditional human lifestyle," explained the UCLA anthropologist Jared Diamond in 2003, "but they would lead to obesity and diabetes in the modern world when the same individuals stop exercising, begin foraging for food only in supermarkets and consume three high-calorie meals day in, and day out." In other words, the human body evolved to be what Kel y Brownel has cal ed an "exquisitely efficient calorie conservation machine." And so, by this hypothesis, we suck up calories when they are abundant and store them as fat until they are cal ed upon in a time of need. "Your genes match nicely with a scarce food supply," Brownel explains, "but not with modern living." Such populations as the Pima and the descendants of African tribes, according to this logic, were until very recently stil trapped in this cycle of feast and famine and scarce food in general, and so their thrifty genes have yet to evolve to deal with times of continual plenty. The NIH researchers who study the Pima, as Gladwel reported, "are trying to find these genes, on the theory that they may be the same genes that contribute to obesity in the rest of us."

For the first few decades of its existence, this notion that we have evolved "thrifty mechanisms to defend energy stores during times of privation" was invariably referred to as a hypothesis. That qualification is now often dropped, but the thrifty gene remains only a hypothesis, and one that rests on many a.s.sumptions that seem unjustifiable.

James Neel initial y proposed the idea of a "thrifty genotype rendered detrimental by progress" to explain why diabetes was so prevalent in Western societies and yet apparently absent in primitive tribes, including the Yanomamo of the Brazilian rain forest, who were then the subject of Neel's research.

Neel was addressing the diseases of civilization and the kind of observations that led Peter Cleave to propose his saccharine-disease hypothesis. (Neel was unaware of Cleave's work at the time.) The enigma of Type 2 diabetes, Neel observed, is that it bestows significant evolutionary disadvantages upon anyone who has it. Diabetic women are more likely to die in childbirth and more likely to have stil births than healthy women; their children are more likely to be diabetic than those of healthy women. This implies that any genes that might predispose someone to become diabetic would evolve out of the population quickly, but this did not seem to have happened. One way to reconcile these observations is to imagine a scenario in which having a genetic predisposition to become diabetic is advantageous in some circ.u.mstances. (In a similar way, having the gene for sickle-cel anemia, normal y a disadvantage, provides protection against malaria, a major advantage in malarial areas, as Neel himself reported.) Since diabetic mothers are known to give birth to heavier children, Neel speculated that these diabetic genes bestowed an exceptional ability to use food efficiently, and thus an exceptional ability to convert calories into fat. Those with such thrifty genes, Neel explained, "might have, during a period of starvation, an extra pound of adipose reserve" that would keep them alive when those who failed to fatten easily would die of starvation. So it would be beneficial to have such genes in the event of famine or prolonged food deprivation, which Neel now a.s.sumed must have been the case throughout our evolutionary history. Those same genes would lead to obesity and diabetes in an environment in which food was plentiful.

"If the considerable frequency of the disease is of relatively long duration in the history of our species," Neel had asked to begin his discussion, "how can this be accounted for in the face of the obvious and strong genetic selection against the condition? If, on the other hand, this frequency is a relatively recent phenomenon, what changes in the environment are responsible for the increase?"

The thrifty gene could be the answer only if diabetes was of long duration in the species-and there is no evidence of that. The disease seems to appear only after populations have access to sugar and other refined carbohydrates. In the Pima, diabetes appeared to be "a relatively recent phenomenon," as Neel himself later noted. When Russel and Hrdli ka discussed the health of the Pima in the early 1900s, they made no mention of diabetes, even while noting the presence of such "rare" diseases as lupus, epilepsy, and elephantiasis.*71 As late as 1940, when El iott Joslin reviewed the medical records of the hospitals and physicians in Arizona, he concluded that the prevalence of diabetes was no higher among the Pima and other local tribes than anywhere else in the United States. Only in the 1950s, in studies from the Bureau of Indian Affairs, was there compel ing reason to believe that diabetes had become common. When Neel tested adolescent Yanomamo for the condition known as glucose intolerance, which might indicate a predisposition to diabetes, he found none, so had no reason to believe that diabetes existed before such isolated populations began eating Western foods. The same was true of an isolated tribe of Pima, discovered living in the Sierra Madre Mountains of northern Mexico. "The high frequency of [Type 2 diabetes] in reservation Amerindians," Neel later explained, "must predominantly reflect lifestyle changes."

By 1982, Neel had come to side with Peter Cleave in believing that the most likely explanation for the high rates of obesity and diabetes in populations like the Pima that had only recently become Westernized was their opportunity to "overindulge in high sugar content foods."

This left open the question of what biological factors or genes might determine who got obese and diabetic and who didn't in the presence of such foods, but it eliminated any reason to suggest that thrifty genes had ever bestowed some evolutionary advantage. "The data on which that (rather soft) hypothesis was based has now largely col apsed," Neel observed. He now suggested that either a tendency for the pancreas to oversecrete insulin and so cause hyperinsulinemia, or a tendency toward insulin resistance, which in turn would result in hyperinsulinemia, was the problem, which is consistent with the carbohydrate hypothesis of chronic disease. Both of these, Neel suggested, would be triggered by the "composition of the diet, and more specifical y the use of highly refined carbohydrates."

It wasn't until the late 1970s, just a few years before Neel himself publicly rejected his hypothesis, that obesity researchers began invoking thrifty genes as the reason why putting on weight seems so much easier than losing it. Jules Hirsch of Rockefel er University was among the first to do so, and his logic is noteworthy, because his primary goal was to establish that humans, like every other species of animal, had apparently evolved a homeostatic system to regulate weight, and one that would do so successful y against fluctuations in food availability. We eat during the day, and yet have to supply nutrients to our cel s al night long, while we sleep, for example, so we must have evolved a fuel storage system that takes this into account. "To me, it would be most unthinkable if we did not have a complex, integrated system to a.s.sure that a fraction of what we eat is put aside and stored," Hirsch wrote in 1977. To explain why these components might cause obesity so often in modern societies, he a.s.sumed as fact something that Neel had never considered more than speculation. "The biggest segment of man's history is covered by times when food was scarce and was acquired in unpredictable amounts and by dint of tremendous caloric expenditure," Hirsch suggested. "The long history of food scarcity and its persistence in much of the world could not have gone unnoticed by such an adaptive organism as man. h.o.a.rding and caloric miserliness are built into our fabric."

This was one of the first public statements of the notion that would evolve into the kind of unconditional proclamation made by Kel y Brownel a quarter century later, that the human body is an "exquisitely efficient calorie conservation machine." But it depended now on an a.s.sumption about human evolution that was contradicted by the anthropologic evidence itself-that human history was dominated by what Jared Diamond had cal ed the "conditions of unpredictably alternating feast and famine that characterized the traditional human lifestyle." Reasonable as this may seem, we have no evidence that food was ever any harder to come by for humans than for any other organisms on the planet, at least not until our ancestors began radical y reshaping their environment ten thousand years ago, with the invention of agriculture.

Both the anthropological remains and the eyewitness testimony of early European explorers suggest that much of the planet, prior to the last century or two, was a "paradise for hunting," in the words of the Emory University anthropologist Melvin Konner and his col aborators, with a diversity of game, both large and smal , "present in almost unimaginable numbers."*72 Though famines have certainly been doc.u.mented among hunter-gatherer populations more recently, there's little reason to believe that this happened prior to the industrial revolution. Those isolated populations that managed to survive as hunter-gatherers wel into the twentieth century, as the anthropologist Mark Nathan Cohen has written, were "conspicuously wel -nourished in qualitative terms and at least adequately nourished in quant.i.tative terms."

Hunter-gatherers lived in equilibrium with their environment just as every other species does. The oft-cited example is the !Kung Bushmen of the semi-arid Kalahari desert, who were studied by Richard Lee of the University of Toronto and a team of anthropologists in the mid-1960s. Their observations, Lee noted, were made during "the third year of one of the most severe droughts in South Africa's history." The United Nations had inst.i.tuted a famine-relief program for the local agriculturalists and pastoralists, and yet the Bushmen stil survived easily on "some relatively abundant high-quality foods," and they did not "have to walk very far or work very hard to get them." The !Kung women would gather enough food in one day to feed their families for the next three, Lee and his col eagues reported; they would spend the remaining time resting, visiting, or entertaining visitors from other camps.

The prevailing opinion among anthropologists, not to be confused with that of nutritionists and public-health authorities, is that hunting and gathering al ow for such a varied and extensive diet, including not just roots and berries but large and smal game, insects, scavenged meat (often eaten at "levels of decay that would horrify a European"), and even occasional y other humans, that the likelihood of the simultaneous failure of al nutritional resources is vanishingly smal . When hunting failed, these populations could stil rely on foraging of plant food and insects, and when gathering failed "during long-continued drought," as the missionary explorer David Livingstone noted of a South African tribe in the mid-nineteenth century, they could relocate to the local water holes, where "very great numbers of the large game" also congregated by necessity. This resiliency of hunting and gathering is now thought to explain why it survived for two mil ion years before giving way to agriculture. In those areas where human remains span the transition from hunter-gatherer societies to farmers, anthropologists have reported that both nutrition and health declined, rather than improved, with the adoption of agriculture. (It was this observation that led Jared Diamond to describe agriculture as "the worst mistake in the history of the human race.") Although famines were both common and severe in Europe until the nineteenth century, this would suggest that those with European ancestry should be the most likely to have thrifty genes, and the most susceptible to obesity and diabetes in our modern toxic environments. Rather, among Europeans there is "a uniquely low occurrence of Type 2 diabetes," as Diamond puts it, more evidence that the thrifty-gene hypothesis is incorrect.

Species adapt to their environment over successive generations. Those that don't, die off. When food is abundant, species multiply; they don't get obese and diabetic.

When earlier generations of obesity researchers discussed the storage of fat in humans and animals, they a.s.sumed that avoiding excessive fat is as important to the survival of any species as avoiding starvation. Since the average 150-pound man with a body fat percentage of only 10 percent is stil carrying enough fat calories to survive one month or more of total starvation, it seems superfluous to carry around more if it might have negative consequences. "Survival of the species must have depended many times both on the ability to store adequate yet not excessive amounts of energy in the form of fat [my italics], and on the ability of being able to mobilize these stores always at a sufficient rate to meet the body's needs," observed George Cahil and Albert Renold, considered two of the leading authorities on the regulation of fat metabolism, in 1965. The total amount of fat stored, they suggested, "should be kept sufficiently large to al ow for periods of fasting to which a given species in a given environment is customarily exposed, yet sufficiently smal to preserve maximum mobility."

The thrifty-gene hypothesis, on the other hand, implies that we (at least some of us) are evolutionarily adapted to survive extreme periods of famine, but a.s.signs to humans the unique concession of having evolved in an environment in which excess fat acc.u.mulation would not be a burden or lead to untimely death-by inhibiting our ability to escape from predators or enemies, for instance, or our ability to hunt or perhaps even gather. It presupposes that we remain lean, or at least some of us do, only as long as we remain hungry or simply lack sufficient food to indulge our evolutionary drive to get fat-an explanation for leanness that the British metabolism researchers Nancy Rothwel and Michael Stock described in 1981 as "facile and unlikely," a kind way of putting it. The "major objection" to the thrifty-genotype hypothesis, noted Rothwel and Stock, "must be based on the observation that most wild animals are in fact very lean" and that this leanness persists "even when adequate food is supplied," just as we've seen in hunter-gatherers. If the thrifty-gene hypothesis were true of any species, it would suggest that al we had to do was put them in a cage with plentiful food available and they would fatten up and become diabetic, and this is simply not the case.

Proponents of the thrifty-gene hypothesis, however, wil invoke a single laboratory model-the Israeli sand rat-to support the notion that at least some wild animals wil get fat and diabetic if caged with sufficient food. "When this animal is removed from the spa.r.s.e diet of its natural environment and given an abundant, high-calorie diet," wrote Australian diabetologist Paul Zimmet in a 2001 article in Nature, "it develops al of the components of the metabolic syndrome, including diabetes and obesity."

But the sand-rat experiments themselves, carried out in the early 1960s at Duke University by the comparative physiologist Knut Schmidt-Nielsen, suggested that the abundance of food was not the relevant factor. Schmidt-Nielsen was trying to establish what aspect of the laboratory diet might be responsible for the obesity and diabetes that appeared in his sand rats. He had taken two groups of rats freshly trapped in Egypt and raised one on Purina Laboratory Chow-"49.4% digestible carbohydrates, 23.4% protein and 3.8% fat"-supplemented with "fresh mixed vegetables," and the other on the fresh vegetables alone. Both had access to as much food as they desired, but only the chow-eating rats got diabetic and obese. This suggested that something about Purina Chow was the determining factor. Perhaps the rats liked it better than vegetables, and so they ate more, although that, too, could be a physiological effect related to the nutrient composition of the chow. It might have been the density of calories in the rat chow, which have less water content than vegetables and so more calories per gram. It was also possible that the cause of the diabetes and obesity in these rats, as Schmidt-Nielsen suggested, was "a carbohydrate intake that is greater than that occuring in the natural diet."*73 Depending on the researchers' preconceptions, the Israeli sand rats could have been considered an animal model of the carbohydrate hypothesis, rather than the thrifty-gene hypothesis. Monkeys in captivity, by the way, wil also get obese and diabetic on high-carbohydrate chow diets. One of the first reports of this phenomenon was in 1965, by John Brobeck of Yale, whose rhesus monkeys got fat and mildly diabetic on Purina Monkey Chow-15 percent protein, 6 percent fat, and 59 percent digestible carbohydrates. According to Barbara Hansen, who studies diabetes and obesity and runs a primate-research laboratory at the University of Maryland, perhaps 60 percent of middle-aged monkeys in captivity are obese by monkey standards. "This is on the kind of diet recommended by the American Heart a.s.sociation," she says, "high-fiber, low-fat, no-cholesterol chow."

The world is ful of species that do fatten regularly, always to serve a purpose-long-distance migrations, reproduction, or survival during periods when food is either unavailable or too risky to procure. Hibernators seem to be an obvious choice to shed light on the a.s.sumptions underlying the thrifty-gene hypothesis. These animals acc.u.mulate enormous fat deposits in response to an environment that offers up periods of feast-spring, summer, and fal -and famine in the winter. Yet this acc.u.mulation goes unaccompanied by the chronic il s, such as diabetes, that appear in obese humans. Hibernating ground squirrels, for instance, wil double their weight and body fat in a few weeks of late summer. Dissecting such squirrels at their peak weight is akin to "opening a can of Crisco oil," as the University of California biologist Irving Zucker, a pioneer of this research, has described it, "enormous gobs of fat, al over the place."

Investigators who study hibernators, like Nicholas Mrosovsky, a University of Toronto zoologist, point out that weight gain, maintenance, and loss in these animals, and so perhaps in al species, is genetical y pre-programmed and particularly resilient to variations in food availability. This program is characterized by its ability to adjust readily to changing circ.u.mstances and the unpredictability of the environment. Ground squirrels wil gain weight through the summer at the same rate whether they're in the wild or in the laboratory. They wil lose it at the same rate during the winter whether they are kept awake in a warm laboratory or are in ful hibernation, eating not a bite, and surviving solely off their fat supplies. "It is very hard to prevent them gaining and losing weight" on schedule, explains Mrosovsky, who did much of the original research in this area. When researchers surgical y remove a sizable portion of fat from experimental animals-a procedure known as a lipectomy-the animals wil restore the lost fat so that within months of the surgery they wil be just as fat as they would have been without the surgery.*74 Even the type of fat found in animals and humans is regulated in a way that accommodates differing internal and external environments. The fat in our limbs, for instance, is less saturated than the fat around our organs, and so is less likely to stiffen in cold weather. We wil also change the fatty-acid composition of our subcutaneous fat with temperature-the colder it gets, the more unsaturated the fats. This same phenomenon, independent of the type of fat consumed, has been observed in pigs, rats, and hibernators. Another example of the evolutionary specificity of fat deposits can be seen in those desert animals that do not store fat subcutaneously, as humans and most animals do, apparently because it would inhibit heat loss and cooling. So there are fat-rumped and fat-tailed sheep, and fat-tailed marsupial mice, al desert-dwel ers that carry their fat almost exclusively in the so-named locations.

The storage of fat, it seems clear, like al evolutionary adaptations, tends to be exquisitely wel suited to the environment-both internal and external-in a way that maximizes benefits while minimizing risks. This is why most investigators who considered these issues in the 1970s and 1980s a.s.sumed that a tendency to gain any excessive weight during periods of abundance would be the kind of obvious liability that evolution would work to select out of the species rather than select in. The thrifty-gene hypothesis does not hold up. But without a thrifty gene, rendered detrimental by the abundance of food in modern societies and the absence of physical labor needed to procure it, how do we explain why gaining weight in modern societies stil seems so much easier than losing it?

Chapter Fifteen.

HUNGER.

Khrushchev, too, looks like the kind of man his physicians must continual y try to diet, and historians wil some day correlate these sporadic deprivations, to which he submits "for his own good," with his public tantrums. If there is to be a world cataclysm, it wil probably be set off by skim milk, Melba toast, and mineral oil on the salad.

A.J. LIEBLING, The Earl of Louisiana, 1961 IN OCTOBER 1917, FRANCIS BENEDICT, director of the Carnegie Inst.i.tution of Washington's Nutrition Laboratory (located, as it happens, in Boston), put twelve young men on diets of roughly fourteen hundred to twenty-one hundred calories a day with the intention of lowering their body weights by 10 percent in a month. Their diets would then be adjusted as necessary to maintain their reduced weights for another two months, while Benedict and his col eagues meticulously recorded their psychological and physiological responses. A second squad of twelve men was studied as a comparison and then they were put on similar calorie-restricted diets. The results were published a year later in a seven-hundred-page report ent.i.tled Human Vitality and Efficiency Under Prolonged Restricted Diet.

Benedict hoped to establish whether humans could adjust to this lower nutritional level and thrive. His subjects lost the expected weight, but they complained constantly of hunger-"a continuous gnawing sensation in the stomach," as described by the Carnegie report-and of being cold to the extent that several found it "almost impossible to keep warm, even with an excessive amount of clothing." They also experienced a 30-percent decrease in metabolism. Indeed, Benedict's subjects reduced their energy expenditure so dramatical y that if they consumed more than twenty-one hundred calories a day-a third to a half less than they had been eating prior to the experiment-they would begin to regain the weight they had lost. The men also experienced significant decreases in blood pressure and pulse rate; they suffered from anemia, the inability to concentrate, and marked weakness during physical activity. They also experienced "a decrease in s.e.xual interest and expression, which, according to some of the men, reached the point of obliteration." That these phenomena were caused by the diet itself rather than the subsequent weight loss was demonstrated by the experience of the second squad of men, who manifested, according to the Carnegie report, "the whole picture...with striking clearness" after only a few days of dieting.

"One general feature of the post-experimental history," the Carnegie researchers reported, "is the excess eating immediately indulged in by the men."

Despite repeated cautions about the dangers of overindulgence after such a strict diet, the men "almost invariably over-ate." As the Carnegie report put it, "the circ.u.mstances militated against" any acquisition of "new dietetic habits." In particular, the cravings for "sweets and accessory foods of al kinds,"

-i.e., snacks-were now free to be indulged, and so they were. Perhaps for this reason, Benedict's young subjects managed to regain al the lost weight and body fat in less than two weeks. Within another three weeks, they had gained, on average, eight pounds more, and came out of this exercise in calorie restriction considerably heavier than they went in. "In practical y every instance the weight prior to the beginning of the experiment was reached almost immediately and was usual y material y exceeded," Benedict and his col agues wrote.

In 1944, Ancel Keys and his col eagues at the University of Minnesota set out to replicate Benedict's experiment, although with more restrictive diets and for a greater duration. Their goal was to reproduce and then study the physiological and psychological effects of starvation of the kind that Al ied troops would likely confront throughout Europe as the continent was liberated. Thirty-two young male conscientious objectors would serve as "guinea pigs," the phrase Keys used in this context. These volunteers would eventual y spend twenty-four weeks on a "semi-starvation diet," fol owed by another twelve to twenty weeks of rehabilitation.

The subjects consumed an average of 1,570 calories each day, split between two meals designed to represent the daily fare of European famine areas. "The major food items served," the researchers noted, "were whole-wheat bread, potatoes, cereals, and considerable amounts of turnips and cabbage. Only token amounts of meats and dairy products were provided."*75 This diet provided roughly half the calories that the subjects had been consuming to maintain their weight. It was expected to induce an average weight loss of 20 percent-or forty pounds in a two-hundred-pounder-aided by a routine that required the subjects to walk five to six miles each day, which would burn off another two to three hundred calories.

Keys's conscientious objectors lost, on average, a dozen pounds of fat in the first twelve weeks of semi-starvation, which const.i.tuted more than half of their original fat tissue, and they lost three more pounds of body fat by the end of twenty-four weeks. But weight loss, once again, was not the only physiological response to the diet. Nails grew slowly, and hair fel out. If the men cut themselves shaving, they would bleed less than expected, and take longer to heal. Pulse rates were markedly reduced, as was the resting or basal metabolism, which is the energy expended by the body at rest, twelve to eighteen hours after the last meal. Reflexes slowed, as did most voluntary movements: "As starvation progressed, fewer and fewer things could stimulate the men to overt action. They described their increasing weakness, loss of ambition, narrowing of interests, depression, irritability, and loss of libido as a pattern characteristic of 'growing old.'" And, like Benedict's subjects, the young men of the Minnesota experiment complained persistently of being cold.

Keys's conscientious objectors reduced their total energy expenditure by over half in response to a diet that gave them only half as many calories as they would have preferred. This was a reasonable response to calorie deprivation, as Keys and his col eagues explained, "in the sense that a wise man reduces his expenditure when his income is cut."

More than fifty pages of the two-volume final report by Keys and his col eagues, The Biology of Human Starvation, doc.u.ment the "behavior and complaints" induced by the constant and ravenous hunger that obsessed the subjects. Food quickly became the subject of conversations and daydreams.

The men compulsively col ected recipes and studied cookbooks. They chewed gum and drank coffee and water to excess; they watered down their soups to make them last. The antic.i.p.ation of being fed made the hunger worse. The subjects came to dread waiting in line for their meals and threw tantrums when the cafeteria staff seemed slow. Two months into the semi-starvation period, a buddy system was initiated, because the subjects could no longer be trusted to leave the laboratory without breaking their diets.

Eventual y, five of the subjects succ.u.mbed to what Keys and his col eagues cal ed "character neurosis," to be distinguished from the "semi-starvation neurosis" that al the subjects experienced; in two cases, it "bordered on a psychosis." One subject failed to lose weight at the expected rate, and by week neurosis" that al the subjects experienced; in two cases, it "bordered on a psychosis." One subject failed to lose weight at the expected rate, and by week three was suspected of cheating on the diet. In week eight, he binged on sundaes, milk shakes, and penny candies, broke down "weeping, [with] talk of suicide and threats of violence," and was committed to the psychiatric ward at the University Hospital. Another subject lasted until week seven, when "he suffered a sudden 'complete loss of wil power' and ate several cookies, a bag of popcorn, and two overripe bananas before he could 'regain control' of himself." A third subject took to chewing forty packs of gum a day. Since his weight failed to drop significantly "in spite of drastic cuts in his diet," he was dropped from the study. For months afterward, "his neurotic manifestations continued in ful force." A fifth subject also failed to lose weight, was suspected of cheating, and was dropped from the study.

With the relaxation of dietary restriction, Keys avoided the dietary overindulgence problem that had beset Benedict's subjects by restricting the rehabilitation diets to less than three thousand calories. Hunger remained unappeased, however. For many of the subjects, the depression deepened during this rehabilitation period. It was in the very first week of rehabilitation, for instance, that yet another subject cracked-his "personality deterioration culminated in two attempts at self-mutilation."

Even during the last weeks of the Minnesota experiment, when the subjects were final y al owed to eat to their hearts' content, they remained perversely unsatisfied. Their food intake rose to "the prodigious level of 8,000 calories a day." But many subjects insisted that they were stil hungry, "though incapable of ingesting more food." And, once again, the men regained weight and body fat with remarkable rapidity. By the end of the rehabilitation period, the subjects had added an average of ten pounds of fat to their pre-experiment levels. They weighed 5 percent more than they had when they arrived in Minneapolis the year before; they had 50 percent more body fat.

These two experiments were the most meticulous ever performed on the effects on body and mind of long-term low-calorie diets and weight reduction.

The subjects were selected to represent a range of physiological types from lean to overweight (albeit al young, male, and Caucasian). They were also chosen for a certain strength of character, suggesting they could be trusted to fol ow the diets and remain dedicated to the scientific goals at hand.

The diets may seem severe in the retel ing, but, in fact, fourteen to sixteen hundred calories a day for weight loss could be considered generous compared with the eight-to-twelve-hundred-calorie diets that are now commonly prescribed, what the 1998 Handbook of Obesity refers to as "conventional reducing diets." Nonetheless, such diets were traditional y known as semi-starvation diets, a term that has fal en out of use, perhaps because it implies an unnatural and uncomfortable condition that few individuals could be expected to endure for long.

In both experiments, even after the subjects lost weight and were merely trying to maintain that loss, they were stil required to eat considerably fewer calories than they would have preferred, and were stil beset by what Keys and his col eagues had cal ed the "persistent clamor of hunger." Of equal importance, simply restraining their appet.i.tes, independent of weight loss, resulted in a dramatic reduction in energy expenditure. This could be reversed by adding calories back into the diet, but then any weight or fat lost returned as wel . One lesson learned was that, for the weight reduction to be permanent, some degree of semi-starvation has to be permanent. These experiments indicated that would never be easy.

Obese patients also get hungry on semi-starvation diets. If they have to restrict their calories to lose weight, then by definition they are forcing themselves to eat less than they would otherwise prefer. Their hunger is not being satisfied. As with lean subjects, their energy expenditure on a semistarvation diet also "diminishes proportionately much more than the weight," as the Pittsburgh clinician Frank Evans reported in 1929 of his obese subjects. This same observation was reported in 1969 by George Bray, who was then at the Tufts University School of Medicine in Boston, and who ent.i.tled his article, for just this reason, "The Myth of Diet in the Management of Obesity." "There is no investigator who has looked for this effect and failed to find it," the British obesity researcher John Garrow wrote in 1978.

The latest reiteration of these experiments, using obese subjects, was conducted by Jules Hirsch at Rockefel er University, and the results were published in The New England Journal of Medicine in 1995. Calorie restriction in Hirsch's experiment resulted in disproportionate reductions in energy expenditure and metabolic activity. Increasing calorie consumption resulted in disproportionate increases in metabolic activity.

Hirsch and his col eagues interpreted their observations to mean that the human body seems surprisingly intent on maintaining its weight-resisting both weight gain and weight loss-so that the obese remain obese and the lean remain lean. As Hirsch explained it, the obese individual appears to be somehow metabolical y normal in the obese state, just as Keys's and Benedict's young men were metabolical y normal in their lean or overweight states before their semi-starvation diets. Once Hirsch's obese subjects took to restricting their calories, however, they experienced what he cal ed "al the physiological and psychological concomitants of starvation." A semi-starvation diet induces precisely that-semi-starvation-whether the subject is obese or lean. "Of al the d.a.m.n unsuccessful treatments," Hirsch later said, "the treatment of weight reduction by diet for obese people just doesn't seem to work."

Over the course of a century, a paradox has emerged. Obesity, it has been said, is caused, with rare exceptions, by an inability to eat in moderation combined with a sedentary lifestyle. Those of us who gain excessive weight consume more calories than we expend, creating a positive caloric balance or a positive energy balance, and the difference acc.u.mulates as excessive pounds of flesh. But if this reconciles with the equal y "indisputable" notion that "eating fewer calories while increasing physical activity are the keys to control ing body weight," as the 2005 USDA Dietary Guidelines for Americans suggest, then the problems of obesity and the obesity epidemic should be easy to solve. Those few individuals for whom obesity is a preferred condition, such as sumo wrestlers, would remain obese through their voluntary program of overeating, and the rest would create a negative energy balance, lose the excess weight, and return to leanness. The catch, as Hirsch pointed out, is that this doesn't happen.

The doc.u.mented failure of semi-starvation diets for the obese dates back at least half a century. It begins with Albert Stunkard's a.n.a.lysis of the relevant research in the mid-1950s, motivated by his desire to resolve what he cal ed the "paradox" between his own failure to reduce obese patients successful y by diet at New York Hospital and "the widespread a.s.sumption that such treatment was easy and effective." Stunkard managed to locate eight reports in the literature that al owed for an accurate a.s.sessment of whether semi-starvation diets worked. In 1959, he reported that the existing evidence confirmed his own failures: semi-starvation diets were "remarkably ineffective" as a treatment for obesity. Only 25 percent of the subjects discussed in these articles had lost as much as twenty pounds on their semi-starvation diets, "a smal weight loss for the grossly overweight persons who are the subjects of these reports." Only 5 percent successful y lost forty pounds. As for Stunkard's own experience with a hundred obese patients, al prescribed "balanced" diets of eight to fifteen hundred calories a day, "only 12% were able to lose 20 lb., and only 1 patient lost 40 lb....Two years after the end of treatment only two patients had maintained their weight loss."*76 A decade later, when Stunkard was invited to discuss obesity at Richard Nixon's White House Conference on Food, Nutrition, and Health, he had come to believe that the adverse effects caused by semi-starvation diets as a treatment for obesity often outweighed any benefits. "Attempts at weight reduction are often accompanied by anxiety and depression, at times severe enough to warrant discontinuation," he said. "Many obese persons today might wel be better off if they learned to live with their condition and stopped subjecting themselves over and over to painful and frustrating attempts to lose weight."

More recent a.s.sessments of the efficacy of semi-starvation diets tend to be studies that set out to evaluate the efficacy of low-fat, calorie-restricted diets, but because they do so by comparing these diets with more balanced calorie-restricted diets, they provide evidence for the efficacy of the latter as wel . In 2002, a Cochrane Col aboration review of the evidence concluded that low-fat diets induced no more weight loss than calorie-restricted diets, and in both cases the weight loss achieved "was so smal as to be clinical y insignificant." A similar a.n.a.lysis was published in 2001 by the U.S. Department of Agriculture. In this case, the authors identified twenty-eight relevant trials of low-fat diets, of which at least twenty were also calorie-restricted. The overweight subjects consumed, on average, less than seventeen hundred calories a day for an average weight loss of not quite nine pounds over six months. Only one of these studies tracked its partic.i.p.ants for more than a year, and in that case the subjects reportedly reduced their caloric intake to thirteen hundred calories for eighteen months. In other words, these subjects reportedly consumed fewer calories per day than had Keys's conscientious objectors, they maintained this semi-starvation regimen for three times as long-and they emerged from the trial having gained, on average, a pound. In the Women's Health Initiative, discussed earlier (see Chapter 4) twenty thousand women were prescribed a low-fat diet and reportedly reduced their calorie consumption by an average of 360 calories a day. After almost eight years of this regimen, they weighed only two pounds less than when they started, and their average waist circ.u.mference, which is a measure of abdominal fat, had increased.

The evidence for the failure of semi-starvation as a treatment of obesity hasn't stopped obesity researchers from recommending the approach. The Handbook of Obesity, published in 1998 and edited by three of the most prominent authorities in the field-George Bray, Claude Bouchard, and W.P.T.

James-says that "dietary therapy remains the cornerstone of treatment and the reduction of energy intake continues to be the basis of successful weight reduction programs." It also notes, in contradiction, that the results of such calorie-restricted diets "are known to be poor and not long-lasting." The chapter on obesity in the latest edition of Joslin's Diabetes Mellitus, written by two clinical investigators from Harvard Medical School, also describes "reduction of caloric intake" as "the cornerstone of any therapy for obesity." It then notes that reducing energy intake to a level substantial y below that of energy expenditure "is difficult to accomplish despite a wide variety of specific dietary approaches." A deficit of seventy-five hundred calories, the authors explain, "is predicted to produce a weight loss of [2.2 pounds]," and so a reduction in food intake of a hundred calories a day "should bring about [an eleven-pound] weight loss over 1 year." But this doesn't seem to happen. "It is clear from common experience, however, that attempts at dieting that rely on such smal reduction in food intake are rarely successful. Thus, more severe reductions in energy intake are typical y prescribed," the Harvard physicians write. These more severe regimens include total starvation, but "the extreme nature of the therapy," the loss of muscle rather than fat tissue, and the many complications "have led to the virtual disappearance of this approach." They also include very low-calorie diets of two to six hundred calories a day, which wil inevitably lead to weight loss, but the weight loss diminishes as the diet progresses, once again because metabolism and energy expenditure both decrease, and when the patients go off the diet, they regain the weight lost. Final y, there are the "many different diets" that provide eight hundred to a thousand calories and are in common use, al of which "should result in weight loss." "None of these approaches," the authors say, "has any proven merit."

In response to these pessimistic a.s.sessments, it is commonly suggested that the obese would ameliorate their problem, or prevent it, if they merely exercised-perhaps sixty or ninety minutes a day, as now prescribed by the USDA Dietary Guidelines. A negative energy balance can be created, according to this logic, by increasing energy expenditure as wel as by eating less. Advice to engage in daily physical activity is now ever-present in public-health messages and popular writing on the problems of obesity and overweight. It's reinforced by the existence of the ubiquitous electronic displays on stair-climbers, treadmil s, and other exercise apparatus that tel us how many calories we al egedly expended in our latest workout.

The belief in physical activity as a method of weight control is relatively new, however, and it has long been contradicted by the evidence. When Russel Wilder of the Mayo Clinic lectured on obesity in 1932, he noted that his patients tended to lose more weight with bed rest, "while unusual y strenuous physical exercise slows the rate of loss." "The patient reasons quite correctly," Wilder said, "that the more exercise he takes the more fat should be burned and that loss of weight should be in proportion, and he is discouraged to find that the scales reveal no progress."

Until the 1960s, clinical investigators routinely pointed out that moderate exercise would lead only to insignificant increases in energy expenditure, and these could be easily matched by slight and comparatively effortless changes in diet. A 250-pound man wil expend three extra calories climbing a flight of stairs, as Louis Newburgh of the University of Michigan calculated in 1942, and this in turn is the equivalent of depriving himself of one-fourth of a teaspoon of sugar or a tenth of an ounce of b.u.t.ter. "He wil have to climb twenty flights of stairs to rid himself of the energy contained in one slice of bread!"

Newburgh observed.

Though more strenuous exercise would burn more calories, it would also lead to a significant increase in appet.i.te. This is the implication of the phrase "working up an appet.i.te." "Vigorous muscle exercise usual y results in immediate demand for a large meal," noted the Northwestern University endocrinologist Hugo Rony in 1940. "Consistently high or low energy expenditures result in consistently high or low levels of appet.i.te. Thus men doing heavy physical work spontaneously eat more than men engaged in sedentary occupations. Statistics show that the average daily caloric intake of lumberjacks is more than 5,000 calories while that of tailors is only about 2,500 calories. Persons who change their occupation from light to heavy work or vice versa soon develop corresponding changes in their appet.i.te."*77 If a tailor became a lumberjack and, by doing so, took to eating like one, there was little reason to think that the same wouldn't happen, albeit on a lesser scale, to an obese tailor who chose to work out like a lumberjack for an hour a day.

In 1960, when the epidemiologist Alvan Feinstein examined the efficacy of various obesity treatments in a lengthy review in the Journal of Chronic Diseases, he dismissed exercise in a single paragraph. "There has been ample demonstration that exercise is an ineffective method of increasing energy output," Feinstein noted, "since it takes far too much activity to burn up enough calories for a significant weight loss. In addition, physical exertion may evoke a desire for food so that the subsequent intake of calories may exceed what was lost during the exercise."

By this time, though, exercise had a profoundly influential proponent: the Harvard nutritionist Jean Mayer, who would almost single-handedly overturn a century of clinical evidence and anecdotal experience. In the 1950s, when Mayer established himself as the leading authority on obesity in the United States, he did so based more on the romance of his background than his expertise as a clinical scientist: he was the son of the famous French physiologist Andre Mayer, and he had fought in the French resistance during World War I .

Mayer represented a new breed of obesity authority, of a kind that would now come to dominate the field. His predecessors-among them Louis Newburgh, Hugo Rony, Hilde Bruch, Frank Evans, Julius Bauer, and Russel Wilder-had al been physicians who worked closely with obese patients.

Col ectively, they had treated thousands of them. Their views on the cause of obesity differed, often radical y, but their firsthand experience was unquestionable. Mayer was not a clinician. His training was in physiological chemistry; he had obtained a doctorate at Yale on the interrelationship of vitamins A and C in rats. In the ensuing decades, he would publish hundreds of papers on different aspects of nutrition, including obesity, but he never treated obese patients, so his hypotheses were less fettered by any anecdotal or real-life experience.

As early as 1953, after just two years of research on genetical y obese mice, Mayer was extol ing the virtues of exercise for weight control. By the end of the decade, he was getting credit from the New York Times for having "debunked" the "popular theories," argued by clinicians and their obese patients, that exercise had little influence on weight. Mayer knew that the obese often eat no more than the lean, and often even less. This seemed to exclude overeating, which meant the obese had to be less physical y active. Otherwise, how could they have achieved positive energy balance and become obese? Mayer himself first reported this phenomenon in a strain of laboratory mice that were p.r.o.ne to both obesity and diabetes. They ate little more than their lean littermates, he noted, but their activity was "almost nil" this sedentary behavior could explain their propensity to grow fat.

Through the 1960s, Mayer doc.u.mented this relationship between energy intake, inactivity, and obesity in a series of human studies. He noted that high-school girls who were overweight ate "several hundred calories less" than those who weren't. "The laws of thermodynamics, however, were not flouted by this finding," he said, because the obese girls spent only a third as much time in physical activity as the lean girls; they spent four times as many hours watching television. Mayer studied adolescent girls at summer camp and reported that