فصل 6

6

THE SATIETY FACTOR

At age fifty-seven, Elisa Moser was admitted to a hospital in Würzburg, Germany, with a collection of troubling symptoms. Her family members explained to the doctors that over the last three years, she had increasingly suffered from headaches, memory loss, poor vision, and childish behavior. Most strangely, over this same period Moser had developed “uncommonly extreme obesity.” Moser’s condition continued to deteriorate, and four weeks after admission to the hospital, she died. Presumably due to the unusual nature of her case, a professor named Bernard Mohr decided to perform an autopsy on her body. In his autopsy notes, he noted that Moser’s “abdomen of extreme dimensions” contained “uncommonly large fat deposits.” Mohr then examined her brain. As he lifted it out of her skull and turned it over to inspect its bottom surface, he noticed a tumor that had damaged her pituitary gland as well as the brain region immediately above it, the hypothalamus. The year was 1839.

Although Mohr couldn’t have known it at the time, his finding may have been the first in a long line of research that helps explain why we overeat, why some people weigh more than others, and why weight loss is challenging and often temporary.

THE SEARCH FOR THE SATIETY CENTER

By the turn of the twentieth century, other researchers had begun to replicate Mohr’s findings. In 1902, the Austrian American pharmacologist Alfred Fröhlich defined a cluster of symptoms, including obesity and sex hormone dysfunction, which were associated with brain tumors in the same location Mohr had described. This condition came to be known as Fröhlich’s syndrome.

Initially, researchers attributed the obesity of Fröhlich’s syndrome to a disruption of the pituitary gland, which at the time was already known to play an important role in growth and development. This view predominated for three decades after Fröhlich’s study was published, yet cracks in the hypothesis had already appeared shortly after Fröhlich’s discovery. In 1904, the Austrian pathologist Jakob Erdheim reported that some patients with obesity had tumors located in the hypothalamus, above the pituitary, but with no apparent damage to the pituitary itself. When several research groups confirmed that damaging the hypothalamus but not the pituitary gland produces obesity in dogs and rats, the case was closed. The obesity of Fröhlich’s syndrome was due to damage of the hypothalamus, not the pituitary.

Yet the scientific odyssey into the regulation of adiposity by the brain had only just begun. In the 1940s, Albert Hetherington and Stephen Ranson performed a series of studies that ushered in the modern era of obesity neuroscience research. They accomplished this by using a remarkable device called the stereotaxic apparatus, invented by British neurosurgeons at the beginning of the twentieth century. A stereotaxic apparatus—pictured in figure 27—fixes the skull in place and allows researchers (and neurosurgeons) to perform brain surgery in an incredibly precise and reproducible manner, which is why versions of it continue to be used today. Hetherington and Ranson adapted it for use in rats and quickly discovered that the critical location for obesity was not the hypothalamus as a whole but rather a subregion of the hypothalamus called the ventromedial hypothalamic nucleus (VMN).

As shown in figure 29, rats with VMN lesions became incredibly obese, with some exceeding two pounds. Physiologist John Brobeck, who was the first to perform careful studies on the feeding behavior of these rats during his medical studies at Yale University in the early 1940s, described them as “ravenously hungry.” They were so eager to eat that they began bingeing even before the anesthesia of their surgery had worn off. This initial binge would typically go on for several hours without interruption, and they would continue eating two to three times their normal food intake for the next month. Brobeck found that the degree of overeating was well correlated with the degree of weight gain and that restricting the rats to a normal food intake largely prevented weight gain. This led to the conclusion that VMN-lesioned rats were obese primarily as a result of their excessive food intake.

Researchers dubbed the VMN the satiety center, because disrupting it seemed to cause animals to lose the ability to feel full—and rapidly eat themselves to obesity. It was damage to this satiety center that caused obesity in Elisa Moser, in Fröhlich’s patients, and in Hetherington’s and Brobeck’s rats. Yet although they had identified its location, researchers still didn’t know how the satiety center worked.

THE SEARCH FOR THE SATIETY FACTOR

In 1949—only a few years after Hetherington published his seminal VMN lesion studies—researchers at the Jackson Laboratory in Bar Harbor, Maine, puzzled over a mouse that appeared to be pregnant. The puzzling part was that the mouse never delivered a litter—and upon closer inspection, also happened to be a male. As it turned out, they had stumbled upon an obese strain of mice that had arisen due to a spontaneous genetic mutation. Dubbed the obese mouse, it was extremely fat and had an appetite to match. It also had a low energy expenditure for its size and metabolic disturbances reminiscent of human obesity. The inheritance pattern of its obesity suggested that the effect was caused by a single gene, which they called the ob gene.

This was only the beginning of the era of obesity genetics research. In 1961, Lois and Theodore Zucker identified a strain of obese rats very similar to the obese mouse. The obesity-causing gene had the same inheritance pattern as the ob gene, and the rats became enormously fat, mostly (but not entirely) due to their prodigious appetites. In fact, they closely resembled rats with VMN lesions, with some animals exceeding two pounds. This strain was named the Zucker fatty rat. Over the ensuing decades, a number of other genetically obese rodent models would be identified, including diabetes and agouti mice. At the time these mutations were discovered, no one had any idea what the functions of the mutated genes were, or whether they were related to one another.

In 1959, just ten years after the obese mouse was identified, University of Leeds physiologist Romaine Hervey initiated a series of studies that would begin to uncover the cause of obesity in VMN-lesioned rats, obese mice, and Zucker fatty rats. To do this, he used a gruesome surgical technique called parabiosis. In parabiosis, two animals are surgically fused together at their flanks, effectively creating conjoined twins. The key property of parabiosis is that it causes the circulatory systems of the two animals to exchange blood at a slow rate so that certain hormones released by one animal will affect the other. This allows researchers to determine if a phenomenon of interest—in this case, obesity—involves a circulating hormone.

To find out if the obesity produced by VMN lesions involved a hormone, Hervey lesioned the VMN of one parabiosed rat while leaving its twin intact. The results were striking: As expected, the VMN-lesioned rats became voracious and rapidly gained fat—yet unexpectedly, the nonlesioned twins lost interest in food, became emaciated, and often died of starvation. Upon autopsy, the bodies of VMN-lesioned animals were overflowing with fat, while remarkably, their intact twins contained no visible fat at all.

For Hervey, the result suggested that a circulating factor was passing from the VMN-lesioned rat to its intact twin, suppressing the twin’s appetite and adiposity. Building on a hypothesis recently developed by Gordon Kennedy, Hervey proposed that fat tissue secretes a hormonal satiety factor whose blood levels reflect adiposity, such that the more fat tissue a person carries, the higher his level of the hormone. This hormone then travels through the bloodstream and acts in the satiety center of the brain to constrain appetite and adiposity. The idea is that when body fat levels increase, the satiety factor increases, and this suppresses appetite and reduces adiposity back to its initial level. Conversely, when body fat levels decrease, the satiety factor decreases, and this stimulates appetite and fat gain back up to its initial level. Together, Hervey hypothesized, the satiety factor and the VMN form a feedback system that regulates adiposity, working to keep it at a stable level. They named this fat-regulating system the lipostat, after the Greek words for “fat” and “stationary.”

Lesioning the satiety center of one parabiosed rat had caused its brain to become unresponsive to the satiety factor, making it think the rat was starving, and driving it to eat voraciously and become obese. In turn, Hervey reasoned, the lesioned rat’s massive adiposity greatly increased the concentration of its now-ineffective satiety factor. Yet the intact twin was still responsive to the satiety factor, and as the hormone came pouring into its circulation from the obese twin, its satiety center received the signal and caused it to stop eating, become emaciated, and gradually starve.

In the early 1970s, at the Jackson Laboratory, a researcher named Doug Coleman set out to learn more about the mysterious obese mouse that had been identified there in 1949. He hypothesized that, like VMN-lesioned rats, obese mice have a defect in the lipostat system. Using parabiosis, Coleman joined obese mice with normal mice. In contrast to what Hervey had reported in his experiments with VMN-lesioned rats, the normal mice continued eating, and their weight remained stable. However, the obese twins underwent a remarkable transformation: Their appetite declined, they didn’t grow as fat, and their obesity-related metabolic disturbances improved. This led Coleman to conclude that obese mice lacked the elusive satiety factor and, therefore, that the satiety factor is encoded by the ob gene that is damaged in these mice. Joining them with normal mice had restored their damaged hormone, normalizing their appetite, body weight, and metabolism. Coleman published his findings in 1973, yet the identity of the ob gene remained shrouded in mystery.

Coleman’s findings were a critical stepping-stone in the history of obesity research, because they gave us a defined entry point to unravel how the brain regulates appetite and body fatness. The ob mouse carries a mutation that inactivates the satiety factor hormone. If researchers could locate the gene, identify the hormone it produced, and discover how it works, they might be able to unlock the secrets of adiposity that had thus far eluded them.

Yet while Hervey and Coleman had shown that extreme disruptions of the lipostat, such as brain lesions and genetic mutations, can influence food intake and adiposity, no one had convincingly demonstrated that the satiety factor plays an important role in the normal, everyday regulation of food intake and adiposity. Ruth B. Harris, one of Hervey’s former graduate students, set out to fill this knowledge gap in follow-up studies at the University of Georgia in the early 1980s. Harris and her team parabiosed two normal rats and overfed one of them using a feeding tube—similar to how farmers overfeed geese to make foie gras. The thinking was that if the satiety factor operates in normal animals, then as the overfed animal gains fat, it should produce more factor, and that should cause the twin to eat less and lose fat. As expected, as Harris’s team overfed one twin and caused it to gain fat, the other twin lost fat. Further experiments showed that a small decline in food intake was required for fat loss to occur in the non-overfed twin. Again, this suggested that fat tissue (or something associated with it) was secreting a powerful hormone that constrains food intake and adiposity. Not only was this hormone relevant to animals with lesions and mutations but it was probably important for the everyday regulation of appetite and body fatness in normal animals.

Gradually, researchers converged on a remarkable conclusion: Even though each obesity model had been developed independently, VMN lesions, the obese mutation, the Zucker fatty mutation, and overfeeding all appeared to impact the same fat-regulating system. Obese mice are unable to produce the satiety factor; VMN-lesioned animals and Zucker fatty rats are unable to respond to it; and overfed animals overproduce it. These independent models all supported the fundamental importance of the same fat-regulating system Hervey had hypothesized in 1959: the lipostat.

Despite this profound convergence, Harris and other researchers still didn’t know what hormone was responsible, as they had already ruled out every suspect that was known at the time. This satiety factor was clearly central to the regulation of appetite and adiposity—yet no one had any idea what it was.

“I’m a baby doctor,” explains Rudy Leibel in a slightly gravelly New York accent, “an endocrinologist baby doctor who got interested in how babies and children get obese, and decided to do something about it. I like that story. It happens to be true.” Leibel is currently an obesity researcher and professor of pediatrics at Columbia University. At the time of his medical training in the late 1960s and early 1970s, obesity research was still in its infancy, and this knowledge void allowed all sorts of harebrained theories to flourish. Obesity was often attributed to a slow metabolic rate or mysterious hormonal imbalances. Worse, it was frequently viewed through a psychoanalytical lens as a “neurosis, the physical expression of which is the accumulation of fat.” At best, it was a moral failure resulting from gluttony or insufficient willpower.

Leibel was part of a growing contingent of researchers who were dissatisfied with these views. During his training at Massachusetts General Hospital, he became acquainted with obese mice, which didn’t appear to require neuroses or moral failure to overeat and grow fat. In addition, he was familiar with a large body of evidence, gradually accumulated over the last century, suggesting that human body weight may be regulated—though it wasn’t entirely clear by what. One of the earliest pieces of evidence came from Rudolf Neumann, a physiology researcher in Hamburg, Germany, who from 1895 to 1897 obsessively measured his own calorie intake and body weight. Neumann found that over a three-year period, his body weight remained surprisingly stable without any conscious effort to control it, despite the fact that his calorie intake fluctuated up and down in the short term. “That was probably the first exposure I had to the idea that there might be some very sophisticated regulatory mechanism for the maintenance or control of body weight in humans,” recalls Leibel.

He had also been impressed by a number of studies suggesting that the human body vigorously resists large, short-term changes in weight brought about by underfeeding or overfeeding. One of the earliest and most influential studies was the Minnesota Starvation Experiment conducted by the prolific nutrition researcher Ancel Keys in the latter years of World War II. The goal was to understand the effects of starvation on the human body and mind. Over the course of six months of semistarvation, thirty-six young male conscientious objectors lost approximately one-quarter of their initial body weight. While this weight loss was no surprise, what happened after their food restriction was lifted is more interesting: Due primarily to a prodigious appetite, their body weights and adiposity rebounded rapidly. As they regained weight, their appetites normalized, and they eventually settled close to their original weights. It seemed as if a powerful internal control system was regulating their appetite and adiposity.

The system seemed to work in the opposite direction as well, pushing back against short-term fat gain. Obesity researcher Ethan Sims tremendously overfed a group of lean prison inmates in the 1960s, causing their weight to increase by up to 25 percent over a four- to six-month period. Despite the fact that these men were not particularly overweight even after overfeeding, their bodies vigorously resisted the weight gain. Sims found that he had to feed his subjects up to 10,000 Calories per day to get them to sustain their weight gain—nearly four times what most adult men require. After the experiment was over, most of them hardly had any appetite for weeks afterward, and the majority slimmed back down to their former weights. This suggested, again, that an internal control system was regulating their appetite and body weight.

Leibel also knew about the research of Fröhlich, Hetherington, Ranson, Hervey, and Coleman, which suggested that the brain plays an important role in appetite and adiposity. And so in 1978, his curiosity got the better of him, and he took a position at Rockefeller University in New York City to look for the elusive satiety factor, under the tutelage of established obesity researcher Jules Hirsch. That decision cost Leibel an assistant professor appointment at Harvard University, and half his salary.

After Leibel spent several years investigating candidate factors without much success, technical advances in genetics research promised to make it possible to identify disease-causing mutations in unknown genes. Today, this is a routine feat, but with the technology of the time, it was herculean, and success was far from certain. With Coleman’s findings on the obese mouse in mind, Leibel explains, “I got to thinking that maybe we should try to identify one of these genes from a mouse.” Researchers had already determined, due to its inheritance pattern, that the satiety factor hormone that was missing in the obese mouse was encoded by a single gene, the ob gene. The discrete genetic defect of the obese mouse offered a molecular entry point from which researchers could begin systematically unraveling the mysterious biology of appetite and body weight regulation.

Leibel had the concept and the drive, but he didn’t have the technical skill. What he needed was a technical wizard who had mastered the rapidly evolving techniques of molecular biology. This man was Jeff Friedman, a bright and driven assistant professor at Rockefeller. In 1986, Leibel and Friedman initiated a collaboration to “clone” the ob gene—in other words, to identify its location in the genome and its DNA sequence.

What followed was a grueling eight-year investigation, requiring more than four thousand mice and one hundred person-years of work. The result would shake obesity science to its foundation and eventually rebuild it into a more mature and sophisticated discipline.

As Leibel and Friedman’s team narrowed in on the location of the ob gene, Friedman grew increasingly concerned that the more senior Leibel would receive the lion’s share of the credit for the gene’s discovery. Friedman insisted that Leibel stay out of the lab where the work was being conducted, although Leibel continued to play a guiding role in the project.

On December 1, 1994, without Leibel’s knowledge, Friedman published the identity of the ob gene in the journal Nature. He reported that the gene codes for a small protein hormone that’s secreted by fat tissue and circulates in the blood. Friedman named this hormone leptin, after the Greek word leptos, meaning “thin.” Furthermore, he showed that humans carry an almost identical gene. The paper concludes, presciently: Identification of ob now offers an entry point into the pathways that regulate adiposity and body weight and should provide a fuller understanding of the [development] of obesity.

The satiety factor had been identified. And obesity was on its way to becoming a biological problem.

Leibel, and most of the other researchers who played key roles in the project, were excluded as authors on the paper. The day before the paper was published, Friedman had filed a patent for leptin. After a ferocious bidding war between pharmaceutical companies, that patent was sold to Amgen for a $20 million initial payment so it could develop leptin as the ultimate weight-loss drug.

AN INCREDIBLE DRIVE TO EAT

The identification of leptin triggered a feeding frenzy among researchers and the pharmaceutical industry. The race was on to understand how leptin works and to explore its potential as a weight-loss drug.

The first step was to purify leptin and see how administering it to rodents affects their adiposity. Echoing Coleman’s parabiosis findings, Friedman and his collaborators showed that leptin injections curbed the prodigious appetite of obese mice and caused them to slim down, exactly as predicted. Leptin precisely plugged the physiological hole that drove them to gain fat. Most intriguingly—especially for the pharmaceutical industry and the public—injecting large doses of leptin into normal mice caused their body fat to melt away almost completely, without affecting their muscle mass. Just as Kennedy, Hervey, and Coleman had predicted over the last half century, leptin was a hormone produced by fat tissue, which acts in the brain to regulate appetite and adiposity.

That is, in rodents. The significance of leptin in humans remained unclear until a remarkably lucky finding by Stephen O’Rahilly, professor of clinical biochemistry and medicine at the University of Cambridge, in 1996. O’Rahilly studies the genetics of diabetes and obesity, sifting through countless cases of these disorders to look for individuals who truly stand out—“clinical exceptions” who just might carry genetic mutations that shed light on the mechanisms of disease. Shortly after Friedman’s team published their paper, O’Rahilly was searching for humans with leptin mutations, and he had a pair of candidates: two cousins of Indian descent, both of whose parents were first cousins in the same family (inbreeding increases the likelihood that a rare mutation will occupy both copies of a gene, resulting in a genetic disorder). One of the cousins weighed 189 pounds by the age of eight and had to use a wheelchair to get around. Liposuction had failed to attenuate her extreme horizontal growth. The other cousin weighed 64 pounds by the age of two. Both children had an extraordinary obsession with food from a very young age, and seemingly insatiable appetites. To O’Rahilly, this wasn’t garden-variety obesity, and it had nothing to do with psychology. Something was seriously wrong with the biology of these children.

It was Sadaf Farooqi’s first month as a clinical fellow in O’Rahilly’s lab, and her first assignment was to look for leptin in the two cousins. She was unable to detect it. Suspecting that she might have run the experiment incorrectly, she repeated it. Again, the children showed undetectable levels of leptin. O’Rahilly and Farooqi had found human counterparts of the obese mouse on their first try.

Eventually, Farooqi and O’Rahilly were able to show that the children’s insatiable appetites and extraordinary obesity were caused by the omission of a single guanine nucleotide—one letter in the 3.2 billion-letter genetic code—that happened to inactivate the leptin gene. “That was really the first evidence that a defect in a single human gene could cause obesity,” explains Farooqi, “but also that a complete lack of leptin could cause obesity in humans.” At this point, Farooqi and O’Rahilly have worked extensively with leptin-deficient humans, who are extremely rare. “Usually, they are of normal birth weight,” says Farooqi, “and then they’re very, very hungry from the first weeks and months of life.” By age one, they have obesity. By age two, they weigh fifty-five to sixty-five pounds, and their obesity only accelerates from there. While a normal child may be about 25 percent fat, and a typical child with obesity may be 40 percent fat, leptin-deficient children are up to 60 percent fat. Farooqi explains that the primary reason leptin-deficient children develop obesity is that they have “an incredible drive to eat,” resulting in an exceptionally high intake of calories. In addition, the reward regions of their brains show an exaggerated response to images of calorie-dense, high-reward foods. Leptin-deficient children are nearly always hungry, and they almost always want to eat, even shortly after meals. Their appetite is so exaggerated that it’s almost impossible to put them on a diet: If their food is restricted, they find some way to eat, including retrieving stale morsels from the trash can and gnawing on fish sticks directly from the freezer. This is the desperation of starvation.

Furthermore, leptin-deficient children have a powerful emotional and cognitive connection to food. “They really enjoy food,” explains Farooqi, “and if you provide them with food, they’re incredibly happy. It doesn’t matter what the food is.” Even the dreaded hospital cafeteria doesn’t faze them. Conversely, they become distressed if they’re out of sight of food, even briefly. If they don’t get food, they become combative, crying and demanding something to eat.

Unlike normal teenagers, those with leptin deficiency don’t have much interest in films, dating, or other teenage pursuits. They want to talk about food, about recipes. “Everything they do, think about, talk about, has to do with food,” says Farooqi. This shows that the lipostat does much more than simply regulate appetite—it’s so deeply rooted in the brain that it has the ability to hijack a broad swath of brain functions, including emotions and cognition, and put them to work seeking food.

There’s another condition that causes similar behavior: starvation. Let’s return to the Minnesota Starvation Experiment for a moment, but this time, let’s focus on the psychological responses that occurred. Over the course of their weight loss, Keys’s subjects developed a remarkable obsession with food. In addition to their inescapable, gnawing hunger, their conversations, thoughts, fantasies, and dreams revolved around food and eating—part of a phenomenon Keys called “semistarvation neurosis.” They became fascinated by recipes and cookbooks, and some even began collecting cooking utensils. Like leptin-deficient adolescents, their mental lives gradually began to revolve around food. Also like leptin-deficient adolescents, they had very low leptin levels due to their semistarved state. The striking similarity between leptin deficiency and starvation suggested that low leptin levels might be responsible for the human brain’s response to starvation, including the hunger, the obsession with food, the activation of brain reward regions, and, as we will soon see, the reduced metabolic rate. It seemed as if the absence of leptin in leptin-deficient children prevented their brains from “seeing” their fat, triggering a powerful starvation response despite the fact that they had extreme obesity.

This phenomenon of starvation in the face of abundant body fat is a paradox with which Rudy Leibel and Jules Hirsch are very well acquainted. In 1984, they published a seminal paper showing that people with garden-variety obesity who then lose weight show evidence of a starvation response. Beginning with a group of twenty-six people with an average body weight of 336 pounds, Leibel and Hirsch slimmed them down to 220 pounds using a rigorous low-calorie diet. Although 116 pounds is an impressive amount of weight to lose, at the end of the weight-loss period, the volunteers were still considered to have obesity. Remarkably, after weight loss, the number of calories their bodies consumed was only about three-quarters of what it should have been based on their new, slimmer body size. What’s more, they were ravenously hungry. Something was shutting down their metabolic rate and ramping up their appetites—fighting their weight loss—even though they remained quite fat.

Leibel and Hirsch have spent much of their careers following up on this puzzling finding. In further studies, they found that weight loss, both in people who are lean and obese, triggers a powerful suite of biological and psychological responses that work together to restore the lost fat. To do this, the brain curtails the activity of the sympathetic nervous system and reduces thyroid hormone levels, both of which slow the metabolic rate, accounting for the cold and sluggish feeling some people experience after weight loss. The brain cuts back the number of calories a muscle burns during a given contraction, reducing the calories expended in physical activity. And most important, the brain ramps up hunger and increases the response to food cues that signal high-calorie, high-reward foods. Before weight loss, you might have been able to stroll by the ice cream aisle without any problem, but after weight loss, the temptation to buy and eat ice cream can be overwhelming. In effect, substantial weight loss triggers a starvation response, whether a person is lean, overweight, or obese—and this response continues until the fat comes back.

If you’ve never had the experience of fighting your own body’s starvation response, Jeff Friedman provides a helpful analogy: Those who doubt the power of basic drives, however, might note that although one can hold one’s breath, this conscious act is soon overcome by the compulsion to breathe. The feeling of hunger is intense and, if not as potent as the drive to breathe, is probably no less powerful than the drive to drink when one is thirsty. This is the feeling the obese must resist after they have lost a significant amount of weight.

Friedman’s analogy is an important lesson for people who think weight loss is as easy as deciding to eat less and exercise more. The brain was forged in the flames of highly competitive natural selection, and it doesn’t take weight loss lightly. “[The starvation response] is there for preservation,” says Leibel. “Evolutionarily, it has played a very important role in our survival.” Remarkably, Leibel and Hirsch found that this starvation response could be almost completely eliminated by injecting their volunteers with small doses of leptin, just enough to keep leptin levels where they were before weight loss. This shows that the drop in leptin that occurs with weight loss is the key signal that triggers the starvation response. Although this powerful self-preservation mechanism evolved to keep us alive and fertile, it often seems to backfire in the modern affluent world, where excess adiposity is a greater threat than starvation.

The findings of Leibel, Hirsch, Friedman, O’Rahilly, Farooqi, and many others lead us inescapably to the conclusion that appetite and adiposity are, to a large extent, biological phenomena that are regulated by nonconscious parts of the brain. Their research has brushed away a variety of archaic hypotheses that were blissfully unconstrained by the need for supporting evidence, and left us with a clear understanding that food intake and adiposity are not simply the result of conscious, voluntary decisions. Yet the leptin story also delivered a major disappointment.

DETECTING DEFICIENCY

Because of a tiny genetic error that disrupted the leptin gene, the brains of O’Rahilly and Farooqi’s patients were unable to perceive their vast fat stores, and as a result, triggered the ultimate starvation response. No amount of food was able to satisfy these children whose brains thought they were in the final, deadliest stage of starvation.

Fortunately for the leptin-deficient cousins, O’Rahilly and Farooqi obtained approval to treat them with injections of isolated leptin. The effect was immediate and dramatic. While before leptin treatment no amount of food was enough, within four days of treatment, they began turning down food. Their obsession with food subsided too, and their brain responses to tempting foods normalized. They lost much of their excess fat, and after a few years, they looked and acted similar to normal kids.

This raises a $20 million question: Why aren’t we all taking leptin to lose weight?

It turns out that people with garden-variety obesity—as opposed to obesity caused by a rare genetic mutation—already have high levels of leptin. And researchers have found that leptin isn’t the miraculous obesity cure the pharmaceutical industry hoped it would be. While leptin therapy does cause some amount of fat loss, it requires enormous doses to be effective (up to forty times the normal circulating amount). Also troubling is the extremely variable response, with some people losing over thirty pounds and others losing little or no weight. This is a far cry from the powerful fat-busting effect of leptin in rodents. The new miracle weight-loss drug never made it to market. This disappointment forced the academic and pharmaceutical communities to confront a distressing possibility: The leptin system defends vigorously against weight loss, but not so vigorously against weight gain. “I have always thought, and continue to believe,” explains Leibel, “that the leptin hormone is really a mechanism for detecting deficiency, not excess.” It’s not designed to constrain body fatness, perhaps because being too fat is rarely a problem in the wild. Many researchers now believe that while low leptin levels in humans engage a powerful starvation response that promotes fat gain, high leptin levels don’t engage an equally powerful response that promotes fat loss.

Yet something seems to oppose rapid fat gain, as Ethan Sims’s overfeeding studies (and others) have shown. Although leptin clearly defends the lower limit of adiposity, the upper limit may be defended by an additional, unidentified factor—in some people more than others. We’ll return to this in the next chapter. For now, let’s explore how this fat-regulating system works and what the implications are for those of us who would like to lose weight or stay lean.

THE FAT THERMOSTAT

The leptin system functions by the same principle as the thermostat in your home, which measures the ambient temperature and compares it to the temperate set point you’ve programmed. If the temperature dips too low, your thermostat engages the heating system; if it rises too high, it engages the air conditioner. This feedback system serves to maintain the stability, or homeostasis, of your home’s interior temperature. The body maintains homeostasis for a number of variables, including temperature, blood pressure, blood pH, respiration, and pulse rate. These variables are regulated because they’re so important for survival.

And so, like your home thermostat, the brain maintains temperature homeostasis by measuring skin and core body temperature and taking action to heat or cool the body as necessary. This includes a variety of physiological and behavioral strategies like constricting blood vessels in your skin to reduce heat loss or dilating them to increase it, revving up a special heat-producing tissue called brown fat, making you shiver to generate heat, or prompting you to put on a sweater or gravitate toward shade and ice water. This coordinated strategy is so effective that it manages to keep your core temperature within about one degree Fahrenheit, regardless of the weather.

As it turns out, the body’s thermostat resides in the hypothalamus. It receives temperature information from sensors in the body and coordinates the physiological and behavioral responses necessary to maintain ideal body temperature. Similarly, the hypothalamus (and other brain regions, to a lesser extent) is the body’s lipostat—the brain region that regulates appetite and body fatness. It receives information about the size of fat stores from signals, including leptin, and coordinates the physiological and behavioral responses necessary to maintain adiposity. As Leibel and Hirsch observed in their weight-loss studies, if a person loses fat, the lipostat engages a coordinated suite of responses that work to increase energy intake, reduce energy expenditure, and thereby regain the lost fat. This is a milder form of the same starvation response Farooqi and O’Rahilly described in leptin-deficient children. Yet the analogy with a thermostat isn’t perfect: In humans, the lipostat isn’t as good at preventing fat gain, as if your home thermostat has very good heat to prevent the temperature from dropping, but weak air conditioning to prevent the temperature from rising. As you may have noticed, our modern understanding of the lipostat closely resembles the model that Romaine Hervey proposed in 1959.

Leibel and Hirsch’s findings suggest that the lipostat isn’t broken in garden-variety obesity—it simply regulates adiposity around a higher set point, analogous to turning up the thermostat in your home. When the adiposity set point is high, the brain requires a higher level of leptin to restrain the starvation response, and in the long run, the only way to get more leptin is to have more fat. In other words, for the brain of a person with obesity, obese is the new lean. Researchers call this phenomenon leptin resistance, because the brain seems to have a hard time “hearing” normal levels of leptin.

What are the implications for us? The first is that once a person develops obesity, it becomes a self-sustaining state, and the person has to overeat to feel the same satisfaction that a lean person feels after eating a smaller meal. In essence, once we gain weight, the lipostat becomes one of the primary reasons why we continue to overeat, undermining our conscious desire to be lean and healthy.

A second key implication is that weight loss is hard because it requires us to fight deeply wired impulses. Long-term diet trials suggest that the hypothalamus is remarkably good at undermining fat-loss efforts. Weight regain afflicts all the most popular diet styles, including portion control, low-fat diets, and low-carbohydrate diets. One of the best examples of this comes from the hit TV series The Biggest Loser. On this reality show, contestants with obesity adopt an extreme diet and exercise regimen, and the person who loses the highest percentage of her starting weight at the end of the season wins a cash prize of $250,000. Many contestants lose over 100 pounds of body weight, including Ali Vincent. In 2008, Vincent dropped 112 pounds, securing a win in the show’s fifth season. Yet since her nadir of 122 pounds, she has regained most of the weight. “I feel like a failure,” says Vincent, understandably frustrated by gradually losing the war of attrition against her hypothalamus. Her experience is by no means unique. Suzanne Mendonca, who lost 90 pounds in the show’s second season in 2005, quipped, “NBC never does a reunion. Why? Because we’re all fat again.” The hypothalamus doesn’t care what you look like in a bathing suit next summer, and it doesn’t care about your risk of developing diabetes in ten years. Its job is to keep your energy balance sheet in the black, and it takes that task very seriously because it was essential for survival and reproduction in the time of our distant ancestors. The tools at its disposal, including hunger, increased food reward, and slowing your metabolic rate, are extremely persuasive. When the hypothalamus squares off with the conscious, rational mind, it usually wins in the end. This doesn’t mean dieting is hopeless, but to be successful, it’s important to understand, respect, and work with what you’re up against. The good news is that the lipostat responds to the cues we give it through our diet and lifestyle, and we can use this to our advantage.

APPEASING THE LIPOSTAT

If the brain regulates adiposity, then how does a person go from being lean to overweight or obese? And can this process be reversed? We know that in each instance of homeostasis, such as heart rate, body temperature, and adiposity, the body defends a certain set point value against changes—but that doesn’t mean set points are inflexible. For example, your body temperature set point can increase in response to infections, a phenomenon we call fever. When you have a fever, your brain doesn’t lose control of temperature regulation—it deliberately regulates it around a higher set point to fight the infection. Your body turns up its thermostat. Similarly, the set point of the lipostat can be turned up and, some evidence suggests, turned down.

Leibel’s and Hirsch’s weight-loss studies lend scientific rigor to a common-sense conclusion: Different people defend different adiposity set points. In lean people, the lipostat defends a low-adiposity set point against fat loss, which makes great sense from an evolutionary standpoint—a lean person can’t afford to lose much fat. Yet what makes less sense is that among people with obesity, the lipostat defends a high-adiposity set point. Somehow, the hypothalamus has “decided” to defend an obese body type rather than a lean one, even though a person with obesity carries much more fat than is necessary to avoid starvation and infertility. In fact, it makes even less sense than that, because excess adiposity is one of the leading causes of infertility and premature death in the affluent world.

Many researchers, including myself, speculate that the lipostat behaves abnormally because it’s been placed in an unfamiliar situation: The hypothalamus is wired to keep us healthy and fertile in an environment that disappeared long ago. In today’s environment of plentiful, tasty, refined, calorie-dense food, low physical activity requirements, and other unnatural characteristics, the lipostat essentially misfires, driving many of us inexorably to overeat and gain weight. Yet others seem to remain lean no matter what they do—a topic we’ll return to in the next chapter.

The set point doesn’t just differ by individual; it can change over time in the same person. Most people in affluent nations gain fat over the course of a lifetime, showing that the set point can move upward, gradually increasing the lower limit of our comfortable weight. This malleability explains how a lean nation like the nineteenth-century United States can become dangerously overweight over the course of a few generations, without a significant change in genetic makeup. Our body weight isn’t completely determined by our genes. Just like the temperature set point, the adiposity set point responds to the conditions of our lives.

In 2000, Barry Levin, an obesity and diabetes researcher at Rutgers University, published a paper clearly demonstrating this effect in rats. Starting with a genetically diverse strain of rats, he fed them either ordinary rat pellets or a high-calorie palatable diet. On the palatable diet, some of the rats gained weight and fat, while others didn’t. Levin’s team took the rats that had gained weight and restricted their food intake while keeping them on the palatable diet, which caused them to lose weight and fat. So far this is what you might expect, but what they found next is more interesting: When they lifted the calorie restriction, allowing the animals to eat as much as they wanted again, the rats didn’t just resume their gradual upward trajectory of weight gain—they quickly bounced back up to the weight of rats that had been eating an unrestricted palatable diet and gaining weight the entire time. Levin’s findings suggest that there is a certain weight that the lipostat “wants” an animal to maintain, defending it against changes, and that this weight depends both on genetics and on the specific diet the animal is eating.

Levin’s team went further, testing the effects of rotating an obesity-susceptible strain of rats between three different diets. The first diet was ordinary rat pellets, the second diet was the same palatable diet as the previous experiment, and the third diet was a highly palatable milkshake-like meal-replacement beverage called Ensure (specifically, the chocolate flavor). As expected, the rats began overeating and gaining weight and fat on the palatable diets. In fact, the rats eating Ensure almost doubled in weight over ten weeks—an impressive feat.

Yet when Levin’s team made the rats switch diets, they found that the rats seemed to defend dramatically different body weights depending on which diet they were currently eating. For example, when they were switched from Ensure to ordinary pellets, their food intake dropped impressively and they rapidly lost weight, until they approximated the weight of animals that had been eating pellets the whole time. When these same animals were switched back to Ensure, they gorged and quickly bounced back up to the weight of animals that had been eating Ensure the whole time. Again, it seemed as if the diets were not just passively causing weight gain but actually changing the set point of the lipostat. Levin ascribed much of this effect to the diet’s palatability, in part because the rats would only overeat and gain weight on chocolate-flavored Ensure—not vanilla or strawberry! This effect hasn’t been studied as thoroughly in humans, but there are tantalizing clues that it might apply to us as well. Let’s return to the studies we encountered in chapter 3, in which volunteers with obesity lost weight rapidly on an unrestricted bland liquid diet dispensed by a machine through a straw. The volunteers were specifically instructed to drink as much of the liquid diet as they needed to feel satisfied, yet once they began the diet, their calorie intake spontaneously dropped because they simply weren’t very hungry (even though lean people on the same regimen drank a normal amount of calories, showing that it wasn’t physically difficult to do so). They lost weight rapidly, yet their starvation response never seemed to kick in. Something about the bland diet was allowing their bodies to feel comfortable at a lower weight, suggesting that just as with Levin’s rats eating plain old rat pellets, the low reward value of the diet may have lowered their adiposity set point.

Five years after the machine-feeding studies, Michel Cabanac, a physiology researcher at Laval University in Canada, published another study that supported and expanded upon these findings. Cabanac’s team had a group of volunteers eat an unrestricted bland liquid diet for three weeks, which caused them to spontaneously eat fewer calories and lose just under seven pounds. Then the researchers had a second group of volunteers lose the same amount of weight over the same period of time by deliberately restricting the portion size of their normal diet. Cabanac found that the portion control group developed the expected hunger response to weight loss—but the bland diet group didn’t. He reported that the bland diet volunteers “reduced their intake voluntarily and were always in good spirits,” while the portion control group “had to continually fight off their hunger and would spend the night dreaming of food.” On the bland diet, the starvation response never kicked in. Cabanac concluded that diet palatability influences the set point of the lipostat in humans.

Returning to the brain, we know that there are important connections between the hypothalamus and reward regions such as the ventral striatum, because hunger magnifies food reward. This is encapsulated in the old adage “Hunger is the best sauce.” Yet we don’t know as much about the mechanisms underlying the reverse connection: how food reward might affect the brain regions that determine appetite and adiposity.

RESTRICTING REWARD

High-reward foods tend to increase food intake and adiposity, while lower-reward foods tend to have the opposite effect. This suggests a weight management “secret” you’ll rarely find in a diet book: eat simple food. The reason you’ll rarely find it in a diet book is that, by definition, lower-reward food is not very motivating. It doesn’t get us excited about a diet, and it doesn’t make books fly off the shelves. We want to hear that we can lose weight while eating the most delicious food of our lives, and the weight-loss industry is happy to indulge us. The truth is that there are many ways to lose weight, but all else being equal, a diet that’s lower in reward value will control appetite and reduce adiposity more effectively than one that’s high in reward value. The trick, as with all diets, is sticking with it, because just as the set point can go down, it can go right back up if you return to your former eating habits. And that means designing an eating plan you can live with for the long haul. For most people, a “bland liquid diet” like those I described isn’t a viable long-term solution, but keeping added fats, sugars, salt, and calorie-dense highly rewarding foods to a modest level may be.

We do, however, have enough information to arrive at some practical conclusions. First, calorie-dense, highly rewarding food may favor overeating and weight gain not just because we passively overeat it but also because it turns up the set point of the lipostat. This may be one reason why regularly eating junk food seems to be a fast track to obesity in both animals and humans. Second, focusing the diet on less rewarding foods may make it easier to lose weight and maintain weight loss because the lipostat doesn’t fight it as vigorously. This may be part of the explanation for why all weight-loss diets seem to work to some extent—even those that are based on diametrically opposed principles, such as low-fat, low-carbohydrate, Paleo, and vegan diets. Because each diet excludes major reward factors, they may all lower the adiposity set point somewhat.

Are there other ways to appease the lipostat? Researchers have often noted that people who exercise more frequently gain less weight over time. There’s a seemingly straightforward explanation for this: They’re burning more calories, so they remain in energy balance. This is probably at least partially correct, but there may be more to the story, as Barry Levin’s research suggests. His findings show, not surprisingly, that exercise attenuates weight gain in rats when they’re offered a fattening diet. Yet Levin’s data also reveal that fit rats aren’t just leaner—they actively defend a lower adiposity set point than sedentary rats on the same diet. This is actually quite consistent with human studies, in which physically fit people are better able to resist fat gain in the face of overeating. It appears that exercise helps keep the lipostat happy at a lower set point.

Yet many people have pointed out that exercise doesn’t work very well as a weight-loss tool in humans, and they have plenty of data to back them up. If you send volunteers home with advice to exercise regularly, most of them hardly lose any weight. Superficially, this presents a striking contrast to the rodent studies. However, I’ve gradually come to believe that there’s more to this story than meets the eye. The problem with many of the human studies is that they simply offer people exercise advice, without having any way to enforce the advice, and often without even accurately measuring how much exercise was actually performed.

In contrast, when we only consider studies in which volunteers had to regularly report to a research gym and exercise under supervision—ensuring compliance—a different picture emerges. In these studies, fat loss is often substantial, and it increases with the intensity and duration of the exercise regimen. So it appears that many of us in the research world, including myself at one time, may have misjudged exercise: It really does cause fat loss.

But the evidence is more nuanced than it may initially appear. The research of John Blundell, professor of psychobiology at the University of Leeds, shows that not everyone loses the same amount of fat as a result of exercise. Intrigued by earlier observations that people respond differently to exercise, Blundell and colleagues recruited thirty-five overweight and obese sedentary men and women and assigned them to exercise five times per week for twelve weeks. Each session was designed to burn 500 Calories and was supervised by the researchers to ensure compliance. At the end of the twelve-week period, the average participant had lost more than eight pounds of fat. However, this average conceals some very interesting information: Changes in adiposity ranged from a loss of twenty-one pounds to a gain of six pounds! To be fair, only one person out of thirty-five gained fat, and we don’t know what else was happening in that person’s life at the time, but it does show that fat gain is possible in the face of a vigorous exercise program. Two others lost less than a pound of fat—a paltry reward for so much effort.

How is it possible to expend 2,500 extra Calories a week yet gain fat? Again, our answer must lie in the energy balance equation we encountered in chapter 1. The only way for adiposity to increase despite increasing calorie expenditure is if calorie intake increases by an even greater amount. And that’s exactly what Blundell’s team observed. When they measured calorie intake in the volunteers, they found that those who lost less weight than expected were inadvertently increasing their calorie intake in response to exercise. This isn’t particularly surprising, since most of us have probably had the experience of “working up an appetite” after playing sports or doing yard work. What’s remarkable is what happened in people who lost as much, or more, weight than expected: They actually decreased their calorie intake in response to the exercise regimen. In the end, about half of the volunteers ate more as a result of the exercise, and half didn’t.

Presumably, this reflects the effects of exercise on the lipostat, as suggested by Levin’s studies in rats. On one hand, exercise depletes the body’s fat reserves and therefore triggers the lipostat to increase appetite. On the other hand, exercise may lower the adiposity set point among people who carry excess fat, reducing appetite and facilitating fat loss. The strength of each of these opposing forces, which varies from person to person, determines the net change in appetite in response to exercise. So while exercise can cause substantial fat loss, it works better for some people than others.

Another fact that’s often overlooked is the difference between weight loss and fat loss. When someone tries to slim down, the goal is not usually to lose weight—it’s specifically to lose fat. It turns out that exercise helps preserve muscle mass during weight loss. Although slow progress on the scale may be frustrating, changes in the mirror and in health as a result of exercise can be better than what the scale suggests. In the end, the evidence suggests that if you can maintain a high level of physical activity, it will probably help you prevent fat gain, accelerate fat loss, and maintain that loss. But it only works if you actually do it—and even then, the degree of fat loss depends on how effectively your brain compensates for the lost calories by increasing your appetite.

Low-carbohydrate dieting is one of the most popular ways to lose weight, and numerous studies suggest that although it’s no miracle obesity cure, it is more effective than a traditional low-fat portion-controlled diet over periods of about a year. This is actually a big deal, because it’s a major reversal of the view that’s prevailed for most of the last half century, which is that fat is fattening, and the best way to lose weight is to cut back on fat. In fact, many people report that low-carbohydrate diets help them curb their appetite and cravings, and the research backs this up. When people go on a low-carbohydrate diet, their spontaneous calorie intake drops substantially—even though they usually aren’t making any deliberate effort to eat fewer calories.

Why? As you may have noticed, this effect looks a whole lot like what happens when the adiposity set point goes down. If we take a closer look at the diets of people who eat a low-carbohydrate diet, what we see is that when they reduce their carbohydrate intake, the proportion of protein in the diet tends to go up. As it turns out, amino acids, the building blocks of protein, act directly in the hypothalamus, influencing the lipostat system. Although most of the direct evidence comes from rodent studies, a substantial amount of indirect evidence suggests that a high intake of protein may be able to lower the adiposity set point in humans too.

A 2005 study by University of Washington researcher Scott Weigle and colleagues illustrates just how striking this effect can be. Weigle’s team started by determining the habitual calorie intake of a group of nineteen volunteers, after which the researchers fed them a high-protein diet (30 percent of calories) for twelve weeks under tightly controlled conditions. On the high-protein diet, the volunteers’ calorie intake spontaneously declined by an average of 441 Calories per day, and their weight dropped by nearly eleven pounds—despite the fact that it wasn’t a weight-loss study and no one had asked them to eat less. As expected, their leptin levels declined as they lost weight—yet the starvation response never seemed to kick in. The effect can’t be attributed to reducing carbohydrate intake, because Weigle’s team increased protein at the expense of fat, not carbohydrate.

Adding to the case is the work of Maastricht University researchers Klaas Westerterp and Margriet Westerterp-Plantenga, which supports the idea that high-protein diets attenuate the starvation response that typically undermines weight loss. Their studies show that people who lose weight by eating a high-protein diet experience less hunger than those who lose weight by other means, and extra protein also largely prevents the reduction of energy expenditure that often accompanies dieting. Consistent with this, their findings demonstrate that restricting carbohydrate without increasing protein doesn’t cause the same weight-loss effect as a typical higher-protein, low-carbohydrate diet, suggesting that carbohydrate restriction per se isn’t actually the key ingredient in low-carbohydrate diets. Rather, advice to eat a low-carbohydrate diet may be effective simply because it’s an easy way to get people to eat high-protein foods and reduce major food reward culprits.

The lipostat system Bernard Mohr first stumbled upon in 1839 is a key nonconscious regulator of food intake and body fatness that often drives us to overeat. It helps explain why weight loss is so hard and why our appetites and waistlines react in certain ways to the cues we give them through our food and lifestyle. But we have much left to uncover as we delve deeper into the machinery of the lipostat. In the next chapter, we’ll explore how genetics affects the lipostat, why some people can eat as much as they want and not gain fat, and how a related system in the brain stem affects how many calories you eat at a sitting.