Let us begin with a short quiz: a few questions to ponder during the 30 (or 60 or 90) minutes a day you spend burning off excess calories at the gym, or perhaps while feeling guilty because you’re not so engaged. If lean people are more physically active than fat people—one fact in the often-murky science of weight control that’s been established beyond reasonable doubt— does that mean that working out will make a fat person lean? Does it mean that sitting around will make a lean person fat? How about a mathematical variation on these questions: Let’s say we go to the gym and burn off 3,500 calories every week—that’s 700 calories a session, five times a week. Since a pound of fat is equivalent to 3,500 calories, does that mean we’ll be a pound slimmer for every week we exercise? And will we continue to slim down at this pace for as long as we continue to exercise?

For most of us, fear of flab is the reason we exercise, the motivation that drives us to the gym. It’s also why public-health authorities have taken to encouraging ever more exercise as part of a healthy lifestyle. If we’re fat or fatter than ideal, we work out. Burn calories. Expend energy. Still fat? Burn more. The dietary guidelines of the U.S. Department of Agriculture, for instance, now recommend that we engage in up to 60 minutes daily of “moderate to vigorous intensity” physical activity just to maintain weight—that is, keep us from fattening further. Considering the ubiquity of the message, the hold it has on our lives, and the elegant simplicity of the notion—burn calories, lose weight—wouldn’t it be nice to believe it were true? The catch is that science suggests it’s not, and so the answer to all of the above quiz questions is “no.”

Just last month, the American Heart Association and the American College of Sports Medicine published joint guidelines for physical activity and health. They suggested that 30 minutes of moderate physical activity five days a week is necessary to “promote and maintain health.” What they didn’t say, though, was that more physical activity will lead us to lose weight. Indeed, the best they could say about the relationship between fat and exercise was this: “It is reasonable to assume that persons with relatively high daily energy expenditures would be less likely to gain weight over time, compared with those who have low energy expenditures. So far, data to support this hypothesis are not particularly compelling.” In other words, despite half a century of efforts to prove otherwise, scientists still can’t say that exercise will help keep off the pounds.

The 30 minutes recommended by the AHA-ACSM report is a departure from the recent guidelines of other authoritative organizations—the Institute of Medicine of the National Academies and the International Association for the Study of Obesity—both of which, like the USDA, have recommended that we exercise for up to 60 minutes a day to avoid what the USDA calls “unhealthy weight gain.” But the reason for this 60-minute recommendation is precisely that so little evidence exists to support the notion that exercising less has any effect.

The report that these experts cite most often as grounds for their assessments was published in 2000 by two Finnish researchers who surveyed all the relevant research on exercise and weight of the previous twenty years. Yet the Finnish report, the most scientifically rigorous review of the evidence to date, can hardly be said to have cleared up the matter. When the Finnish investigators looked at the results of the dozen best-constructed experimental trials that addressed weight maintenance—that is, successful dieters who were trying to keep off the pounds they had shed—they found that everyone regains weight. And depending on the type of trial, exercise would either decrease the rate of that gain (by 3.2 ounces per month) or increase its rate (by 1.8 ounces). As the Finns themselves concluded, with characteristic understatement, the relationship between exercise and weight is “more complex” than they might otherwise have imagined.

This is not to say that there aren’t excellent reasons to be physically active, as these reports invariably point out. We might just enjoy exercise. We may increase our overall fitness; we may live longer, perhaps by reducing our risk of heart disease or diabetes; we’ll probably feel better about ourselves. (Of course, this may be purely a cultural phenomenon. It’s hard to imagine that the French, for instance, would improve their self-esteem by spending more time at the gym.) But there’s no reason to think that we will lose any significant amount of weight, and little reason to think we will prevent ourselves from gaining it.

The one thing that might be said about exercise with certainty is that it tends to makes us hungry. Maybe not immediately, but eventually. Burn more calories and the odds are very good that we’ll consume more as well. And this simple fact alone might explain both the scientific evidence and a nation’s worth of sorely disappointing anecdotal experience.

It’s difficult to get health authorities to talk about the disconnect between their official recommendations and the scientific evidence that underlies it because they want to encourage us to exercise, even if their primary reason for doing so is highly debatable. Steve Blair, for instance, a University of South Carolina exercise scientist and a co-author of the AHA-ACSM guidelines, says he was “short, fat, and bald” when he started running in his thirties and he is short, fatter, and balder now, at age 68. In the intervening years, he estimates, he has run close to 80,000 miles and gained about 30 pounds.

When I asked Blair whether he thought he might be leaner had he run even more, he had to think about it. “I don’t see how I could have been more active,” he said. “Thirty years ago, I was running 50 miles a week. I had no time to do more. But if I could have gone out over the last couple of decades for two to three hours a day, maybe I would not have gained this weight.” And maybe he would have anyway. If we trust the AHA-ACSM report he co-authored, there is little reason to believe that the amount he runs makes any difference. Nonetheless, Blair personally believes he would be fatter still if he hadn’t been running. Why?

There was a time when virtually no one believed exercise would help a person lose weight. Until the sixties, clinicians who treated obese and overweight patients dismissed the notion as naïve. When Russell Wilder, an obesity and diabetes specialist at the Mayo Clinic, lectured on obesity in 1932, he said his fat patients tended to lose more weight with bed rest, “while unusually strenuous physical exercise slows the rate of loss.”

The problem, as he and his contemporaries saw it, is that light exercise burns an insignificant number of calories, amounts that are undone by comparatively effortless changes in diet. In 1942, Louis Newburgh of the University of Michigan calculated that a 250-pound man expends only three calories climbing a flight of stairs—the equivalent of depriving himself of a quarter-teaspoon of sugar or a hundredth of an ounce of butter. “He will have to climb twenty flights of stairs to rid himself of the energy contained in one slice of bread!” Newburgh observed. So why not skip the stairs, skip the bread, and call it a day?

More-strenuous exercise, these physicians further argued, doesn’t help matters—because it works up an appetite. “Vigorous muscle exercise usually results in immediate demand for a large meal,” noted Hugo Rony of Northwestern University in his 1940 textbook, Obesity and Leanness. “Consistently high or low energy expenditures result in consistently high or low levels of appetite. 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 appetite.” If a tailor becomes a lumberjack and, by doing so, takes to eating like one, why assume that the same won’t happen, albeit on a lesser scale, to an overweight tailor who decides to work out like a lumberjack for an hour a day?

Credit for why we came to believe otherwise goes to one man, Jean Mayer, who began his career at Harvard in the early fifties, went on to become the most influential nutritionist in the country, and then, for sixteen years, served as president of Tufts University (where there is now a Jean Mayer USDA Human Nutrition Research Center on Aging). As an authority on human weight regulation, Mayer was among the very first of a new breed, a type that has since come to dominate the field. His predecessors—Wilder, Rony, Newburgh, and others—had all been physicians who worked closely with obese and overweight patients. Mayer was not. His training was in physiological chemistry; he had obtained a doctorate at Yale with a dissertation 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 why we get fat, but he never had to reduce obese patients as part of his clinical obligation, and so his hypotheses were less fettered by anecdotal or real-life experience.

As early as 1953, after just a few years of research on laboratory mice, Mayer began extolling the virtues of exercise for weight control. By 1959, the New York Times was crediting him with having “debunked [the] popular theories” that exercise played little role in weight control. Mayer knew that the obese often eat no more than the lean and occasionally even less. This seemed to exclude gluttony as a cause of their weight gain, which meant that these fat people had to be less physically active. Otherwise, how could they take in more calories than they expend and so become fat?

Through the sixties, Mayer documented the relationship between inactivity and the overweight. He noted that fat high-school girls ate “several hundred calories less” than lean classmates. “The laws of thermodynamics were, however, not flouted by this finding,” he wrote, because the obese girls expended less energy than the lean: They were much less active; they spent four times as many hours watching television. Mayer also studied infants. “The striking phenomenon is that the fatter babies were quiet, placid babies that had moderate intake,” Mayer reported, “whereas the babies who had the highest intake tended to be very thin babies, cried a lot, moved a lot, and became very tense.” Thus, Mayer concluded, “some individuals are born very quiet, inactive, and placid and with moderate intake get fat, and some individuals from the very beginning are very active and do not get particularly fat even with high intakes.”

It was Mayer who pioneered the now-ubiquitous practice of implicating sedentary living as the “most important factor” leading to obesity and the chronic diseases that accompany it. Modern Americans, said Mayer, were inert compared with their “pioneer forebears” who were “constantly engaged in hard physical labor.” Every modern convenience, by this logic, from power windows to the electric toothbrush, only serves to minimize the calories we expend. “The development of obesity,” Mayer wrote in 1968, “is to a large extent the result of the lack of foresight of a civilization which spends tens of billions annually on cars, but is unwilling to include a swimming pool and tennis courts in the plans of every high school.”

Mayer’s hypothesis always had shortcomings, but they were ignored for the same reasons they still are—who wants to openly question the idea that physical activity is a panacea? The first issue is a logical one: That conclusion that the fatter we are, the more sedentary we’re likely to be is actually a correlation; it tells us nothing about what is cause and what is effect. “It is a common observation,” noted Rony in 1941, “that many obese persons are lazy, i.e., show decreased impulse to muscle activity. This may be, in part, an effect that excess weight would have on the activity impulse of any normal person.” Equally possible is that obesity and physical inactivity are both symptoms of the same underlying cause.

This logical problem was then obscured by Mayer’s all-out attack on the role of hunger. Mayer acknowledged that exercise could make us hungrier, but he said it wasn’t necessarily the case. This was the heart of Mayer’s message—a purported loophole in the relationship between appetite and physical activity. “If exercise is decreased below a certain point, food intake no longer decreases,” said Mayer. “In other words, walking one half-hour a day may be equivalent to only four slices of bread, but if you don’t walk the half-hour, you still want to eat the four slices.”

Mayer based this conclusion on two (and only two) of his own studies from the mid-fifties. The first purported to demonstrate that laboratory rats exercised for a few hours every day will eat less than rats that don’t exercise at all. But this would never be replicated. In more recent experiments, the more rats run the more rats eat; weights remain unchanged. And when rats are retired from these exercise programs, they eat more than ever and gain weight with age more rapidly than rats that were allowed to remain sedentary. With hamsters and gerbils, exercise increases body weight and body-fat percentage. So exercising makes these particular rodents fatter, not leaner.

Mayer’s second study was an assessment of the diet, physical activity, and weights of workers and merchants at a mill in West Bengal, India. This article is still cited—by the Institute of Medicine, for instance—as perhaps the only existing evidence that physical activity and appetite do not necessarily go hand in hand. But it, too, has never been replicated, despite (or perhaps because of) a half-century of improvements in methods of assessing diet and energy expenditure in humans.

It helped that Mayer promoted his pro-exercise message with a fervor akin to a moral crusade. In 1966, Mayer was the primary author of a U.S. Public Health Service report advocating increased physical activity along with diet as the best way to lose weight. In 1969, Mayer chaired Richard Nixon’s White House Conference on Food, Nutrition, and Health. “The successful treatment of obesity must involve far-reaching changes in lifestyle,” the conference report concludes. “These changes include alterations of dietary patterns and patterns of physical activity.” In 1972, Mayer began writing a syndicated newspaper column on nutrition: Exercise, Mayer now wrote, sounding suspiciously like a diet doctor selling a patent claim, will make “weight melt away faster,” and “contrary to popular belief, exercise won’t stimulate your appetite.”

Our culture of physical exercise began only in the late sixties, coincident with Mayer’s crusade, which explains why our parents might not have been quite so devoted to the idea of spending their leisure time perspiring profusely. In 1977, the New York Times was covering the “exercise explosion” that had come about because the conventional wisdom of the sixties that exercise was “bad for you” had been transformed into the “new conventional wisdom—that strenuous exercise is good for you.” When the Washington Post estimated in 1980 that 100 million Americans were partaking in the “new fitness revolution”—coincident with the start of the current obesity epidemic—it also noted that most of them “would have been derided as ‘health nuts’" only a decade earlier.

Meanwhile, the evidence simply never came around to support Mayer’s hypothesis, even though our beliefs did. My favorite study of the effect of physical activity on weight loss was published in 1989 by a team of Danish researchers. Over the course of eighteen months the Danes trained nonathletes to run a marathon. At the end of this training period, the eighteen men in the study had lost an average of five pounds of body fat. As for the nine women subjects, the Danes reported, “no change in body composition was observed.” That same year, F. Xavier Pi-Sunyer, then director of the St. Luke’s–Roosevelt Hospital Obesity Research Center in New York, reviewed the studies on exercise and weight, and his conclusion was identical to that of the Finnish review’s eleven years later: “Decreases, increases, and no changes in body weight and body composition have been observed,” Pi-Sunyer reported.

Granted, all this still doesn’t explain why we bought into Mayer’s idea that we could exercise more and not compensate by eating more. One simple reason is that the health reporters bought it, and we were reading their articles, not the research literature itself. In 1977, for instance, the National Institutes of Health hosted its second conference on obesity and weight control. “The importance of exercise in weight control is less than might be believed,” the assembled experts concluded, “because increases in energy expenditure due to exercise also tend to increase food consumption, and it is not possible to predict whether the increased caloric output will be outweighed by the greater food intake.” That same year, The New York Times Magazine reported that there was “now strong evidence that regular exercise can and does result in substantial and—so long as the exercise is continued—permanent weight loss.” By 1990, a year after Pi-Sunyer’s pessimistic assessment of the evidence, Newsweek was declaring exercise an “essential” element of any weight-loss program, and the Times had stated that on those infrequent occasions “when exercise isn’t enough” to lose weight, “you must also make sure you don’t overeat.”

As for the authorities themselves, the primary factor fueling their belief in the weight-maintaining benefits of exercise was their natural reluctance to acknowledge otherwise. Although one couldn’t help but be “underwhelmed by” the evidence, as Mayer’s student Judith Stern, a UC Davis nutritionist, wrote in 1986, it would be “shortsighted” to say that exercise was ineffective because it meant ignoring the possible contributions of exercise to the prevention of obesity and to the maintenance of weight loss that might be induced by diet. These, of course, had never been demonstrated either, but they hadn’t been ruled out. This faith-based philosophy came to dominate scientific discussions on exercise and weight, but it couldn’t be reconciled with the simple notion that appetite and calories consumed will increase with an increase in physical activity. Hence, the idea of working up an appetite was jettisoned. Clinicians, researchers, exercise physiologists, even personal trainers at the local gym took to thinking and talking about hunger as though it were a phenomenon exclusive to the brain, a question of willpower (whatever that is), not the natural consequence of a body trying to replenish itself with energy.

Ultimately, the relationship between physical activity and fatness comes down to the question of cause and effect. Is Lance Armstrong excessively lean because he burns off a few thousand calories a day cycling, or is he driven to expend that energy because his body is constitutionally set against storing calories as fat? If his fat tissue is resistant to accumulating calories, his body has little choice but to burn them as quickly as possible: what Rony and his contemporaries called the “activity impulse”—a physiological drive, not a conscious one. His body is telling him to get on his bike and ride, not his mind. Those of us who run to fat would have the opposite problem. Our fat tissue wants to store calories, leaving our muscles with a relative dearth of energy to burn. It’s not willpower we lack, but fuel.

For the last 60 years, researchers studying obesity and weight regulation have insisted on treating the human body as a thermodynamic black box: Calories go in one side, they come out the other, and the difference (calories in minus calories out) ends up as either more or less fat. The fat tissue, in this thermodynamic model, has nothing to say in the matter. Thus the official recommendations to eat less and exercise more and assuredly you’ll get thinner. (Or at least not fatter.) And in the strict sense this is true—you can starve a human, or a rat, and he will indeed lose weight—but that misses the point. Humans, rats, and all living organisms are ruled by biology, not thermodynamics. When we deprive ourselves of food, we get hungry. When we push ourselves physically, we get tired.

Our bodies, like all living organisms, have evolved a fantastically complex web of feedback loops. These physiological mechanisms serve fundamentally to work against the inevitable pull of thermodynamics (which is entropy, a.k.a. death) and so make life possible. The necessary condition of life, as the great French physiologist Claude Bernard noted 140 years ago, is to keep the internal environment of an organism stable and conducive to life, regardless of what’s happening on the outside. This is what the Harvard physiologist Walter Cannon, in the thirties, called homeostasis—or the “wisdom of the body,” as he put it. “Somehow the unstable stuff of which we are composed,” Cannon wrote, “had learned the trick of maintaining stability.”

The key is that among the many things regulated in this homeostatic system—along with blood pressure and blood sugar, body temperature, respiration, etc.—is the amount of fat we carry. From this biological or homeostatic perspective, lean people are not those who have the willpower to exercise more and eat less. They are people whose bodies are programmed to send the calories they consume to the muscles to be burned rather than to the fat tissue to be stored—the Lance Armstrongs of the world. The rest of us tend to go the other way, shunting off calories to fat tissue, where they accumulate to excess. This shunting of calories toward fat cells to be stored or toward the muscles to be burned is a phenomenon known as fuel partitioning.

The job of determining how fuels (glucose and fatty acids) will be used, whether we will store them as fat or burn them for energy, is carried out primarily by the hormone insulin in concert with an enzyme known technically as lipoprotein lipase—LPL, for short. (Sex hormones also interact with LPL, which is why men and women fatten differently.)

In the eighties, biochemists and physiologists worked out how LPL responds to exercise. They found that during a workout, LPL activity increases in muscle tissue, and so our muscle cells suck up fatty acids to use for fuel. Then, when we’re done exercising, LPL activity in the muscle tissue tapers off and LPL activity in our fat tissue spikes, pulling calories into fat cells. This works to return to the fat cells any fat they might have had to surrender—homeostasis, in other words. The more rigorous the exercise, and the more fat lost from our fat tissue, the greater the subsequent increase in LPL activity in the fat cells. Thus, post-workout, we get hungry: Our fat tissue is devoting itself to restoring calories as fat, depriving other tissues and organs of the fuel they need and triggering a compensatory impulse to eat. The feeling of hunger is the brain’s way of trying to satisfy the demands of the body. Just as sweating makes us thirsty, burning off calories makes us hungry.

This research has never been controversial. It’s simply been considered irrelevant by authorities, all too often lean, who have been dead set on blaming fatness on some combination of gluttony, sloth, and perhaps a little genetic predisposition thrown in on the side. But contemplating the means by which we might lose weight without considering the hormonal regulation of fat tissue is like wondering why children grow taller without considering the role of growth hormones. Or, for that matter, like trying to explain the record-breaking triumphs of modern athletes—Barry Bonds, say—and never considering the possibility that steroid hormones (or human growth hormone or insulin) might be involved.

If it’s biology, and not a lack of willpower, that explains why exercise fails so many of us as a weight-loss tool, then we can still find reason for optimism. Since insulin is the primary hormone affecting the activity of LPL on our cells, it’s not surprising that insulin is the primary regulator of how fat we get. “Fat is mobilized [from fat tissue] when insulin secretion diminishes,” the American Medical Association Council on Foods and Nutrition explained back in 1974, before this fact, too, was deemed irrelevant to the question of why we gain weight or the means to lose it. Because insulin determines fat accumulation, it’s quite possible that we get fat not because we eat too much or exercise too little but because we secrete too much insulin or because our insulin levels remain elevated far longer than might be ideal.

To be sure, this is the same logic that leads to other unconventional ideas. As it turns out, it’s carbohydrates—particularly easily digestible carbohydrates and sugars—that primarily stimulate insulin secretion. “Carbohydrates is driving insulin is driving fat,” as George Cahill Jr., a retired Harvard professor of medicine and expert on insulin, recently phrased it for me. So maybe if we eat fewer carbohydrates—in particular the easily digestible simple carbohydrates and sugars—we might lose considerable fat or at least not gain any more, whether we exercise or not. This would explain the slew of recent clinical trials demonstrating that dieters who restrict carbohydrates but not calories invariably lose more weight than dieters who restrict calories but not necessarily carbohydrates. Put simply, it’s quite possible that the foods—potatoes, pasta, rice, bread, pastries, sweets, soda, and beer—that our parents always thought were fattening (back when the medical specialists treating obesity believed that exercise made us hungry) really are fattening. And so if we avoid these foods specifically, we may find our weights more in line with our desires.

As for those people who insist that exercise has been the key to their weight-loss programs, the one thing we’d have to wonder is whether they changed their diets as well. Rare is the person who decides the time has come to lose weight and doesn’t also decide perhaps it’s time to eat fewer sweets, drink less beer, switch to diet soda, and maybe curtail the kind of carb-rich snacks—the potato chips and the candy bars—that might be singularly responsible for driving up their insulin and so their fat.

For the rest of us, it may be time to take a scientific or biological view of our excesses rather than a biblical one. The benefits of exercise include the joys of virtuousness. I worked out today, therefore I can eat fattening foods to my heart’s content. But maybe the causality is reversed here too. Maybe it’s because we eat foods that fatten us that the workout becomes a necessity, the best we can do in the battle against our own fat tissue.

Gary Taubes is the author of Good Calories, Bad Calories: Challenging the Conventional Wisdom on Diet, Weight Control, and Disease (Knopf, October 2), from which this article is adapted.