فصل 7

7

THE HUNGER NEURON

“I started my fellowship in 1987,” reminisces obesity researcher Mike Schwartz about his training at the University of Washington, “and was immediately indoctrinated into the idea that there is an adiposity control system.” Schwartz’s fellowship was part of a very unusual research program led by obesity and diabetes researchers Dan Porte and Steve Woods. “What I was working on was considered way out of the mainstream at the time,” explains Schwartz. “Everyone just assumed that obesity is a problem where people eat too much, and if they could control their eating and be normal, they wouldn’t have this problem.” Although some scientists at the time believed that adiposity is regulated, most didn’t. When Schwartz began his fellowship, Leibel and Friedman hadn’t yet identified leptin, and researchers knew very little about the lipostat. Schwartz’s goals were to raise awareness about the fact that adiposity is biologically regulated and eventually target it therapeutically. The only way to accomplish these two goals was to understand the underlying brain systems. Schwartz’s research, and that of many others, would eventually explain much about how the brain regulates body fatness, how the lipostat changes in the brain of a person with obesity, and why some people are particularly prone to overeating and fat gain.

Three years before the start of Schwartz’s fellowship, a researcher named Satya Kalra discovered that a small protein called neuropeptide Y (NPY) causes massive overeating when it’s injected into the brain of a rat. Adding to the excitement, researchers discovered that NPY is naturally produced by neurons in the arcuate nucleus, a tiny region of the hypothalamus near the VMN satiety center, and it became more abundant after fasting, suggesting that it could be involved in hunger. Together, this led Schwartz, Porte, and Wood to hypothesize that NPY might be part of the brain apparatus that regulates eating and adiposity.

At the time, Schwartz’s research was focused on the hormone insulin, which plays an important role in regulating the levels of sugar and fat in the blood. Because of the fact that circulating insulin levels increase when a person overeats and gains fat, and decrease when a person undereats and loses fat, Schwartz viewed it as a possible signal to the brain that participates in regulating adiposity. His team was able to show that injecting insulin into the brains of rats reduces the production of NPY in the hypothalamus and also reduces food intake. This was the first time anyone had been able to draw a biological road map from food intake, to a circulating hormone, to a brain circuit, and back to food intake.

Yet, explains Schwartz, “we knew insulin wasn’t enough.” They were well aware of Coleman’s parabiosis work in obese mice, and also aware that insulin couldn’t explain his findings. Something else had to be out there—something much bigger.

When Friedman’s team published the identity of the ob gene, Schwartz, Porte, and Woods immediately realized leptin might be the missing link they were looking for. “It seemed logical to speculate that leptin might do something similar to insulin,” recalls Schwartz, “which was to inhibit neurons that stimulate feeding.” After four months of grueling work, and thousands of microscope slides, the answer was no longer speculation: Leptin reduced hunger-promoting NPY levels exactly as predicted, supporting the idea that leptin controls food intake (in part) by reducing NPY levels in the brain. His finding was one of the first steps toward understanding how the lipostat works.

Schwartz submitted his team’s findings to the journal Science—the highest-impact scientific journal in the world. “I was just beginning my assistant professorship,” explains Schwartz. “No one knew who I was, so this was an important opportunity.” A month after submission, he received a mailed letter from an editor at Science, the contents of which are permanently seared into his memory: “It has come to our attention that a paper has been accepted for publication elsewhere that significantly compromises the novelty of your findings.” A competing group, led by Mark Heiman at the pharmaceutical company Eli Lilly, had scooped him. Schwartz ended up publishing his team’s version of the finding in the journal Diabetes, which is a good journal, but not as visible as Science. This setback galvanized Schwartz and his team. “We had been burned, and there was a whole series of things we knew were going to be a race. We ended up winning most of those races because we knew what was at stake.” What followed was a virtual avalanche of data from Schwartz, Porte, Woods, Woods’s new postdoc Randy Seeley, and several other competing groups. Shortly after the NPY study, Schwartz published a paper showing that the hypothalamus, and particularly the arcuate nucleus, contains high levels of the receptor for leptin. Even more tantalizing was the accumulating evidence that another group of proteins, called melanocortins, play the opposite role of NPY in the brain: When injected into the brains of rodents, melanocortins powerfully suppressed food intake. Like NPY, melanocortins are located in distinct neurons of the arcuate nucleus, which were named POMC neurons, after the POMC protein that is the precursor of melanocortins. Schwartz’s group showed that melanocortin levels are also regulated by leptin—yet in the opposite direction of NPY. NPY and melanocortins are key cellular pathways by which leptin regulates food intake and adiposity via the brain.

What emerged from these studies was a remarkably logical explanation for how leptin regulates the lipostat: It turns off neurons that drive eating, and it turns on neurons that inhibit eating. And, by implication, when leptin levels decline, neurons that drive eating turn on and neurons that inhibit eating turn off, increasing the drive to eat. This “push-pull” system is redundant and extremely robust, and only disrupting major nodes in the signaling pathway can derail it.

Disrupting a major node is precisely what Albert Hetherington and Stephen Ranson did by damaging the VMN “satiety center” in the rat studies we encountered early in the last chapter. The VMN contains neurons that stimulate POMC neurons, and when it’s destroyed, POMC neurons become less active and cease to restrain appetite. As Hetherington and Ranson showed, this makes rats overeat tremendously and grow to impressive proportions. But it doesn’t require brain damage to disrupt a major node in the signaling pathway. Getting rid of leptin is another way to do so, and as we’ve seen, the result is the same.

Schwartz’s basic explanation for how leptin works has withstood the test of time, and NPY and POMC neurons have remained at the center of the story, although we now know a lot more about how the system operates. For example, NPY neurons in the arcuate nucleus don’t just secrete NPY; they secrete at least two other substances that stimulate feeding. Because of this synergy of appetite-stimulating substances, released in just the right downstream brain regions, NPY neurons are the most powerful driver of eating known to science. If there is such thing as a “hunger neuron” that drives pure, visceral hunger, the NPY neuron is it. Scott Sternson, a neuroscience researcher at the National Institutes of Health’s Janelia research campus, knows this well. His group was the first to specifically stimulate NPY neurons in awake, normally behaving mice. When they turn on NPY neurons, mice eat. A lot. I’ve replicated this experiment myself, and it’s quite impressive. With the flick of a switch, a mouse will stuff its face with whatever food is around—eating up to ten times what it would normally eat in the same period of time.

Furthermore, Sternson’s work shows that the way NPY neurons compel a mouse to seek food is by making the mouse feel bad until it eats. Like humans, mice don’t like to feel hungry, and relieving that hunger state by eating—or, in Sternson’s case, by turning off NPY neurons directly—is itself a reward. If we tie this together with what we covered in chapter 3, it becomes clear that eating motivates us in two distinct ways that reinforce one another: Unpleasant hunger neurons get turned off, and food reward neurons get turned on.

Richard Palmiter, the University of Washington researcher who keeps popping up in this book, has a neat trick up his sleeve: He can destroy almost any neuron population in the brain with pinpoint accuracy, without harming nearby neurons. When he uses this trick to destroy NPY neurons in obese mice, their appetites normalize, they lose weight, and ultimately, they become difficult to distinguish from normal mice. “The major symptoms are all corrected,” explains Palmiter. What this suggests is that the primary reason obese animals overeat and become tremendously fat is that their NPY neurons are in constant overdrive because there isn’t any leptin around to keep them in check. Get rid of the NPY neurons, and the animals slim down, even without a trace of leptin. This has a second, even more remarkable implication: The hunger, the obsession with food—many of the physiological and psychological effects that we see in people who are dieting, starving, or born without leptin—could be largely due to a population of hyperactive NPY neurons that is small enough to fit on the head of a pin.

At this point, there has been an enormous amount of research on the brain systems that regulate appetite and adiposity—far too much for me to summarize in this book. There are many other hormones and neurons that play roles in the system. Yet you don’t have to know all those details to have a basic understanding of how the lipostat works. Think of it as an hourglass, with NPY and POMC neurons at its narrow center. On the top of the hourglass, we have signals entering the brain that communicate the body’s current energy status. They include things like leptin and insulin. These signals converge, mostly through indirect routes, on NPY and POMC neurons, and collectively, they determine the activity of these neurons.

The bottom of the hourglass represents the outputs of NPY and POMC neurons. These are the responses that the brain uses to regulate the body’s energy status, such as hunger, food reward, metabolic rate, and physical activity.

As far as we currently know, NPY and POMC neurons are the most important convergence points between the adiposity-regulating inputs and outputs of the brain, and as such, they have attracted a disproportionate amount of attention from the scientific community. Many researchers, including Sternson, Palmiter, and Brad Lowell, a neuroscience researcher at Harvard Medical School, are working on deciphering the inputs and outputs of these neurons—and they are making remarkable progress. “The ability to identify individual neurons and then use this technology to do circuit mapping,” says Schwartz, “is ultimately going to take the field to the next level.”

In many ways, we’ve already reached the next level Schwartz is referring to. In fact, we’ve cured obesity countless times—in rodents. We now have the ability to take genes from almost any species, manipulate them to make them do what we want, insert them into the mouse genome so that they are expressed in specific cell populations of the brain, and use them to influence food intake, adiposity, and many other things. We can precisely activate, silence, or even kill specific populations of neurons in the mouse brain, controlling appetite and adiposity as if the mouse were a marionette. Modern neuroscience has achieved feats that would have seemed like science fiction to researchers only a few decades ago.

We know from the work of Mohr, Leibel, Friedman, O’Rahilly, Farooqi, and many others that the brain circuits that regulate eating and adiposity in humans have much in common with those of rodents. With time, we could undoubtedly adapt the techniques I described for use in ourselves. So what’s holding us back from curing human obesity? In a word, ethics. While we already have the technical ability to genetically engineer humans, and probably directly manipulate the brain circuits that control eating, we don’t currently consider it ethical. And there are many good reasons for this, one of which is that we haven’t yet demonstrated the safety of this technology.

We do consider it ethical to use drugs to nudge the brain circuits that control eating, and researchers have already developed many weight-loss drugs that do precisely that. Unfortunately, drugs are an extremely blunt tool, poorly suited for targeting an organ as complex as the brain. This is because when we take a pill or an injection, it bathes the entire brain in the drug, including all eighty-six billion neurons, and trillions of connections between them, which together perform countless unique tasks. Adding to the challenge, most of the chemical signals that regulate eating and adiposity are used for other purposes elsewhere in the brain and body. This makes it very difficult to target your circuit of interest without collateral damage. Imagine trying to drive nails into your wall with a sledgehammer: you may get a nail in sometimes, but you’re also going to put holes in your drywall. Because of this, most drugs that impact food intake have unacceptable side effects, such as the dangerous psychological effects of the drug rimonabant (“reverse marijuana”) we encountered earlier. Very few of the weight-loss drugs identified so far have acceptable side effects, and the few that do aren’t the silver bullets we wish they were. But we keep looking, hoping for a lucky break, and it’s possible that our expanding knowledge will someday make that possible.

At this point, Schwartz thinks the field has accomplished one of his goals: raising awareness about energy homeostasis. There is little remaining doubt among researchers, and even many doctors, that appetite and body fatness are biologically regulated by nonconscious regions of the brain. As far as his second goal, Schwartz hopes that our greater understanding of the lipostat will soon fulfill its promise to prevent and reverse obesity in humans. For me, this goal doesn’t seem very far off, but, as Schwartz cautions, “that’s what I thought when they discovered NPY.” Although this work explains a lot about how the lipostat works under normal conditions, it doesn’t tell us what changes cause the brain of a person with obesity to defend a higher level of adiposity, or how we might reverse those changes. For that, the field would need a different approach.

SCARY IMPLICATIONS

Licio Velloso, an obesity researcher at the University of Campinas in Brazil, is determined to understand the changes in the brain that underlie obesity. In the early 2000s, he decided to take a new approach to tackling the problem, one that wasn’t constrained by existing ideas about what might be happening in the brain. To do this, Velloso turned to a technique called an RNA microarray that tells the researcher which genes are turned on, which are turned off, and to what degree. By looking at the pattern of gene expression, we can peer into the inner workings of the cell and understand, to some extent, what it’s up to at the moment.

Velloso wanted an answer to a simple question: What are the cells of the hypothalamus doing when an animal becomes obese? To do so, he used an RNA microarray to compare gene expression in the hypothalami of lean rats versus rats made obese by diet. When Velloso’s team analyzed the data, a striking trend emerged: Many of the genes that were more active in obese mice were related to the immune system, and particularly a type of immune system activation called inflammation. As Velloso noted in his 2005 paper, this makes perfect sense. Previous research had already implicated chronic inflammation in insulin resistance—a condition in which tissues like liver and muscle have a harder time responding to the glucose-controlling hormone insulin—and this process had already been linked to increased diabetes risk. It wasn’t a major leap to suppose that inflammation in the hypothalamus might cause resistance to leptin and insulin, increasing the adiposity set point and contributing to obesity risk.

To further test this idea, Velloso’s team blocked a major inflammatory pathway in the brains of obese rats. They reasoned that if inflammation in the hypothalamus is really causing obesity, then blocking this inflammation should reduce food intake and body weight. And that’s exactly what they observed. Since Velloso’s discovery, other researchers have followed up on the finding, confirming that inflammation in the hypothalamus blocks leptin signaling, leading to leptin resistance and weight gain. Yet inflammation isn’t the only thing going wrong in the hypothalami of obese rodents. In 2012, my colleagues Josh Thaler and Mike Schwartz and I published a study in which we looked more closely at the cellular changes that occur in the hypothalamus during the development of obesity. Much of the study focused on two types of cells in the brain called astrocytes and microglia. While neurons are the cells that do most of the information processing of the brain, astrocytes and microglia play a supporting role in keeping delicate neurons happy—protecting them from threats, helping them heal, giving them energy, and cleaning up after them. When the brain is injured, these cells go into overdrive, increasing in size and number to counter the threat and accelerate healing. “All conditions that damage the brain,” explains Thaler, “such as trauma, stroke, neurodegenerative disease, even infections to some extent, cause this effect.” In a healthy brain, astrocytes are small cells that send out a web of thin filaments to monitor surrounding cells, and these filaments don’t overlap with the filaments of neighboring astrocytes. In an injured brain, astrocytes multiply and grow in size, and their filaments enlarge and overlap those of neighboring astrocytes. Microglia undergo similar changes. The activation of microglia and astrocytes is a universal marker of brain injury, and it’s visible under a microscope, so we decided to look for it in the hypothalami of our obese rats and mice.

And we found it: Astrocytes in the hypothalami of obese rats and mice were enlarged, and their filaments were tangled together in a thick mat. Microglia had also enlarged and multiplied. Both changes were specifically located in the same area as NPY and POMC neurons (the arcuate nucleus), but not elsewhere. Our results suggest that obese rodents suffer from a mild form of brain injury in an area of the brain that’s critical for regulating food intake and adiposity. Not only that, but the injury response and inflammation that developed when animals were placed on a fattening diet preceded the development of obesity, suggesting that this brain injury could have played a role in the fattening process.

This is pretty interesting stuff if you’re a rat, but what does it have to do with humans? To see if humans with obesity show evidence of injury in the hypothalamus, we called on our colleague Ellen Schur, an obesity researcher at the University of Washington. She specializes in a technique called magnetic resonance imaging (MRI), which allows researchers and doctors to observe the structure of live tissues without harming them, sort of like an X-ray that can see soft tissues in great detail.

One of the conditions doctors use MRI to diagnose is brain damage resulting from a past injury, such as stroke or physical trauma. This is because when the brain is injured, astrocytes go into overdrive to support the healing process, and they eventually form a scar that is visible on an MRI scan for a long time after the injury. This is analogous to how your skin grows scar tissue as it heals a cut.

Although we weren’t expecting to see stroke-like changes in the hypothalami of people with obesity, we thought it was worth looking for a subtler version of the same scarring—similar to what we found in rats and mice. And that’s exactly what we saw. Schur’s analysis showed that the more signs of damage we found in a person’s hypothalamus, the more likely he was to have obesity. What’s more, this effect was once again located in the part of the hypothalamus that harbors NPY and POMC neurons. “The scariest implication,” explains Schur, “is that the food we eat may cause damage in areas of the brain that we need to regulate our body weight and our appetite, as well as our blood sugar and, to some degree, our reproductive health.” Just as a tumor in Elisa Moser’s hypothalamus caused her to develop obesity, a milder form of brain damage could be contributing to our own expanding waistlines.

Now it’s time for a reality check: We still don’t know to what extent these changes actually contribute to obesity, rather than being a result of obesity, or simply passively associating with the fattening process. We’ll have to do more research to figure that out. However, what we can conclude is that the hypothalamus is under duress in obesity and that this is likely caused (at least in part) by the unhealthy food we eat. In response to this challenge, the hypothalamus activates a broad swath of cellular stress response pathways, and some of these have the potential to dampen leptin signaling and contribute to the fattening process. This likely operates in parallel with the set-point-altering effects of food reward and protein intake we encountered in the last chapter.

Brain damage is a daunting term, and it may make the situation seem hopeless for people who would like to lose weight. Yet our research also suggests that the process is reversible—at least in mice. When we switch mice off a fattening diet and back on to a strict healthy diet, even without restricting their calorie intake, they lose their excess fat, and their astrocytes and microglia go back to normal. This is true even if they’ve been obese for a long time. We still don’t know if the same is true for humans, but there is reason to hope.

What’s so fattening about the diets we use to make rodents obese in a research setting, and how do they injure the hypothalamus? In many ways, they are similar to the diets of affluent humans. They’re made of refined ingredients; they have a high calorie density; they’re highly rewarding (to rodents); they’re high in fat and often sugar. The diet Schwartz and I used in our research is a light blue pellet that has the texture of greasy cookie dough. Rodents greatly prefer this stuff to normal, unrefined, low-fat food, and when we give them unrestricted access to it, they gorge on it for the first week or so.

Many researchers have tried to narrow down the mechanisms by which this food causes changes in the hypothalamus and obesity, and they have come up with a number of hypotheses with varying amounts of evidence to support them. Some researchers believe the low fiber content of the diet precipitates inflammation and obesity by its adverse effects on bacterial populations in the gut (the gut microbiota). Others propose that saturated fat is behind the effect, and unsaturated fats like olive oil are less fattening. Still others believe the harmful effects of overeating itself, including the inflammation caused by excess fat and sugar in the bloodstream and in cells, may affect the hypothalamus and gradually increase the set point. In the end, these mechanisms could all be working together to promote obesity. We don’t know all the details yet, but we do know that easy access to refined, calorie-dense, highly rewarding food leads to fat gain and insidious changes in the lipostat in a variety of species, including humans. This is particularly true when the diet offers a wide variety of sensory experiences, such as the hyperfattening “cafeteria diet” we encountered in chapter 1.

Personally, I believe overeating itself probably plays an important role in the process that increases the adiposity set point. In other words, repeated bouts of overeating don’t just make us fat; they make our bodies want to stay fat. This is consistent with the simple observation that in the United States, most of our annual weight gain occurs during the six-week holiday feasting period between Thanksgiving and the new year, and that this extra weight tends to stick with us after the holidays are over. Thanksgiving dinner is the definition of overeating, and Christmas Eve, Christmas Day, and New Year’s Eve aren’t far behind. Throughout that entire period, well-meaning family and friends inundate us with cookies, pies, and other tempting calorie-rich treats that tend to hang around the kitchen until we eat them.

Because of some combination of food quantity and quality, holiday feasting ratchets up the adiposity set point of susceptible people a little bit each year, leading us to gradually accumulate and defend a substantial amount of fat. Since we also tend to gain weight at a slower rate during the rest of the year, intermittent periods of overeating outside of the holidays probably contribute as well.

How might this happen? We aren’t entirely sure, but researchers, including Jeff Friedman, have a possible explanation: Excess leptin itself may contribute to leptin resistance. To understand how this works, I need to give you an additional piece of information: Leptin doesn’t just correlate with body fat levels; it also responds to short-term changes in calorie intake. So if you overeat for a few days, your leptin level can increase substantially, even if your adiposity has scarcely changed (and after your calorie intake goes back to normal, so does your leptin). As an analogy for how this can cause leptin resistance, imagine listening to music that’s too loud. At first, it’s thunderous, but eventually, you damage your hearing, and the volume drops. Likewise, when we eat too much food over the course of a few days, leptin levels increase sharply, and this may begin to desensitize the brain circuits that respond to leptin. Yet Rudy Leibel’s group has also shown that high leptin levels alone aren’t enough—the hypothalamus seems to require a second “hit” for high leptin to increase the set point of the lipostat. This second hit could be the brain injury process we, and others, have identified in obese rodents and humans.

Let’s recap what I’ve proposed thus far. We overeat because we’re surrounded by seductive, calorie-dense food that’s a great deal. The food’s high reward value increases the set point of the lipostat, though not necessarily permanently, and this further facilitates overeating. At the same time, overeating itself spikes leptin levels and injures the hypothalamus by a mechanism we have yet to nail down (likely involving diet quality in addition to quantity). These two simultaneous hits cause the hypothalamus to lose sensitivity to the leptin hormone, meaning that it requires more leptin, and therefore more body fat, to hold off the starvation response that drives us to overeat. This time, the increase in your set point is permanent, or at least difficult to reverse. The lower limit of your comfortable weight creeps up.

EFFICIENTLY MANAGING OVEREATING

One practical implication of this research is that if you want to control your weight over the long term, focusing on the six-week holiday period will give you the greatest return on your effort. Developing strategies to avoid holiday overeating, such as getting rid of holiday snacks in the kitchen and cooking lighter versions of traditional recipes, might go a long way toward curbing the inexorable upward arc of adiposity that most of us experience over our lifetimes.

To be clear, this is a working hypothesis that needs to be tested further before we can hang our hats on it. We don’t yet have a complete understanding of how obesity develops and is maintained, but each year brings us closer to the answer.

THE HUNGRY BRAIN

It’s four in the afternoon, and I haven’t eaten since breakfast. I also just rode my bicycle for over an hour to get to the University of Washington in Seattle. As I look at images of junk food inside an MRI machine in the mazelike basement of the Health Sciences Building, I can’t help but think how surreal the experience is. I’m lying inside a Philips Achieva 3.0T, which looks like a giant white doughnut standing on its outer edge. My head, which is in the center of the doughnut hole, is secured by an inflatable cap, and I’m wearing earplugs to protect my ears from the deafening sound of the machine. I’m trying hard not to move.

Researchers and doctors use MRI to look for structural features of the brain, but they can also use it to measure brain activity. This technique is called functional MRI, or fMRI for short. My colleague Ellen Schur and her team are using fMRI to understand the brain regions that determine hunger and satiety, ultimately influencing how much we eat.

In this particular experiment, I’m looking at blocks of images that fall into three categories: 1) rewarding, high-calorie foods, such as pastries, pizza, and potato chips; 2) low-calorie healthy foods, such as strawberries, celery, and apples; and 3) nonfood items like shoes and cars. While I look at the images, the MRI machine measures my brain activity using a magnetic field six hundred times stronger than a typical refrigerator magnet. By comparing my brain’s response to the three categories of images, we can see which areas are activated by each.

The next week, I stop by Schur’s office to go over the images. Schur and research scientist Susan Melhorn pull them up on a computer monitor. First, we look at an image in which they’ve compared my brain’s response to high-calorie foods and my response to nonfood images. This allows us to subtract out the brain activity that happens when we look at pictures in general, and focus on changes that relate specifically to looking at high-calorie food.

“You have a classic response,” notes Schur. First, she points out my VTA. If you recall from chapters 2 and 3, this is the brain region that sends dopamine into the ventral striatum. Both the VTA and the ventral striatum are central to motivation and reinforcement—for example, the desire we feel when we smell freshly baked brownies. “My VTA is lit up!” I exclaim, unable to contain myself. There’s a bright, colorful blob where my VTA is located, indicating that my brain’s dopamine system was very excited by the high-calorie food images. You can see a black-and-white version of the image in the upper left-hand panel of figure 37.

Next, we move on to my ventral striatum, which should also be activated since it receives input from the VTA. This time, the blob is even larger, as you can see in the upper right-hand panel of figure 37. “That’s monstrous!” proclaims Schur.

“That’s the biggest ventral striatum response I’ve ever seen,” adds Melhorn.

They explain that skipping lunch and riding my bike to the medical center had created a much larger energy deficit than they typically see in their subjects, therefore provoking a heightened state of food motivation in my brain.

The next region we examine is my OFC, the place where economic value is computed during the decision-making process. Again, it’s partially obscured by a colorful blob. “You need to make a decision,” Schur offers by way of explanation, “and from here you’re going to activate a plan to obtain the food.” The fourth region Schur examines is my insular cortex. This is a part of the brain that processes taste information. Again, it’s covered by a colorful blob. I find this puzzling since I hadn’t tasted the food, but Schur points out that the insular cortex is often activated by looking at food, similar to how our motor cortex fires when we think about movements. “We use some of the same neurons to think about movements that we use to make movements,” explains Schur. Evidently, my brain was going through the motions of eating—rehearsing that glorious moment when I’d be stuffing pizza into my mouth (sadly for my long-suffering insular cortex, that never happened).

The diagnosis was clear: My hungry brain wanted food. A lot. And it wouldn’t be satisfied by low-calorie foods, because the pattern of activation was almost nonexistent when I gazed at fruits and vegetables. “When we’re hungry,” explains Schur, “our bodies don’t want healthy food.” Instead, powerful, instinctive brain regions draw us toward concentrated, quick, easy calories. “That’s what we’re all up against.”

My fMRI results are consistent with what Schur’s studies have shown. When people are hungry, their brains react strongly to calorie-dense foods. However, after they eat, this reaction to food cues subsides. Schur explains: “At the end of a meal, what you’re really experiencing is the fact that food doesn’t look good to you anymore. The taste of it isn’t as good as it used to be; you look at your plate and say, ‘Ugh, I don’t want any more of that.’” As a meal progresses, something in the brain receives information about what we’ve eaten and shuts down the circuits that make us want more food. How does this work, and can we exploit it to help curb our tendency to overeat?

WILL THE REAL SATIETY CENTER PLEASE STAND UP?

Harvey Grill, a neuroscientist at the University of Pennsylvania, studies the brain stem—a complex region of the brain where it joins the spinal cord. The brain stem is the most ancient part of the brain, evolutionarily speaking, and it tends to govern deeply instinctual, nonconscious functions, such as digestion, breathing, and basic movement patterns. It is also, according to Grill’s research over the last forty years, the most important brain region for satiety.

“At the time I started my postdoc in 1974 at the Rockefeller Institute with Ralph Norgren,” explains Grill, “there really wasn’t a basis in data yet—there was only an idea.” At the time, the prevailing view was that the hypothalamus is the only brain region that regulates food intake. Yet Grill and Norgren knew that the brain stem receives numerous inputs from the gut and mouth that are relevant to eating—and also contains outputs that regulate eating-related movements like chewing. “The question was,” recalls Grill, “to what extent are these linked?” Grill spent a year perfecting a technique that allowed his team to surgically inactivate everything in the rat brain except the brain stem and nearby structures. These “decerebrate” rats were unable to use most of the circuits in their brains, including their hypothalami. Surprisingly, when Grill’s team placed food into the front of their mouths, they would chew and swallow normally. Even more impressive, when offered food continuously, decerebrate rats would eat the same amount as intact rats normally eat during a meal and then abruptly refuse additional food. “They would take meals!” exclaims Grill, still excited by his seminal finding four decades later.

The similarity with normal rats didn’t end there. Decerebrate rats reacted to a variety of satiety-related signals in the same way as normal rats: They ate less at a meal when Grill’s team gave them a “snack” first, and they ate less in response to satiety hormones that the gut normally produces when we eat. This demonstrated, without a shadow of a doubt, that the brain stem is single-handedly capable of monitoring what’s happening in the gut and generating the satiety response that ends a meal. Thanks to the research of Grill and many others, we now have a reasonably clear picture of how this happens. When you eat food, it enters your stomach and stretches it. After partially digesting the food, your stomach gradually releases it into the small intestine. Here, specialized cells in the intestinal lining detect the nutrient content of what you ate, for example, the amount of carbohydrate, fat, and protein. These stretch and nutrient signals are relayed to the brain, primarily via the vagus nerve, which plays a major role in bidirectional gut-brain communication. At the same time, incoming nutrients cause the gut and pancreas to release a number of hormones that either activate the vagus nerve or act on the brain directly.

These signals that encode the quantity and quality of food you just ate converge on a brain region called the nucleus tractus solitarius (NTS), which is the junction point of the vagus nerve with the brain stem. The NTS integrates the various signals ascending from the digestive tract and produces a level of satiety that’s appropriate for what you ate. These complex computations happen beyond your conscious awareness, and the only information your conscious brain receives is whether or not you feel full.

Despite the fact that part of the hypothalamus was once called the satiety center, and leptin was dubbed the satiety factor, we now believe the brain stem is the primary brain region that directly regulates meal-to-meal satiety, while leptin and the hypothalamus primarily regulate long-term energy balance and adiposity. Grill’s research shows that although decerebrate rats take normal-sized meals, if they are underfed, they are unable to compensate normally by increasing the size of subsequent meals. In other words, their satiety system works great, but their lipostat is out of the picture, once again suggesting that the hypothalamus may be required for that function. A more accurate name for the hypothalamus would be the adiposity center, and leptin, the adiposity factor. Yet this distinction isn’t black and white: Grill’s research shows that the brain stem has a hand in regulating adiposity, and the hypothalamus may also play a role in regulating meal-to-meal food intake.

At some point, this information has to reach the action selection circuits that tell the brain whether or not to eat food. Although there are multiple connections from the hypothalamus and brain stem to the basal ganglia and associated structures, we still don’t have a very clear picture of how they influence food-related decisions. However, Palmiter’s and Lowell’s research is increasingly converging on a small area of the brain stem called the parabrachial nucleus that receives input (direct and indirect) from NPY neurons, POMC neurons, and NTS neurons. The parabrachial nucleus could end up being a master regulator of hunger and satiety that plugs into action selection circuits—but this remains to be seen. Ultimately, the size of your meals has a major impact on your total calorie intake, and whether or not you gain weight over time. This is partially determined by how your brain stem generates satiety as you eat, eventually causing you to lose interest in food. Yet we know the hypothalamus impacts adiposity in large part by modifying food intake. How does this work? The hypothalamus influences brain stem satiety circuits in response to long-term changes in adiposity. In other words, if you’re dieting and you’ve lost fat, the hypothalamus ensures that it takes more food to feel full at a meal than it did before you lost fat. Your brain dampens the feeling of satiety so you won’t feel satisfied until you’ve eaten enough calories to start regaining fat. This is why people who are dieting often seem to have a bottomless appetite and never feel full. Conversely, if you’ve overeaten and gained fat, your brain enhances the feeling of satiety so your meals will be smaller for a while. This is how we think the hypothalamus and brain stem work together to regulate appetite and adiposity.

Since the hypothalamus influences brain stem satiety circuits, any impairment of the leptin system should increase the amount of food it takes a person to reach satiety. This is exactly what researchers have found in people with obesity, consistent with the idea that their brains are leptin resistant. When a person with obesity eats food, it doesn’t suppress her brain’s response to food cues as much as it does in a lean person. The brain regions that govern hunger and motivation just keep firing, driving her to overeat. This might seem grim, but there are ways to mitigate it.

The system of gut-brain communication that governs satiety doesn’t do a perfect job of transmitting the calorie value of a meal to the brain. In other words, some foods make us feel more full than others, even if they contain the same number of calories. Because of this, we can exploit the quirks of the satiety system to help naturally reduce (or increase) our calorie intake, without discomfort.

In 1995, Susanna Holt and colleagues published a groundbreaking paper that gives us powerful insight into how we can use food to “trick” the brain into feeling full with fewer calories. The idea behind it is quite simple. Holt and her team recruited volunteers and fed them 240-Calorie portions of thirty-eight common foods, such as bread, oatmeal, beef, peanuts, candy, and grapes. Over the next two hours, the volunteers recorded how full they felt every fifteen minutes. Holt’s team used the resulting data to calculate a “satiety index” for each food—representing how filling it is per calorie. Then, they analyzed the data set as a whole to determine which food properties are the most strongly related to satiety.

White bread, as expected, had a low satiety index relative to other foods, meaning it delivers little satiety per unit calorie. Whole-grain bread, in contrast, had a significantly higher satiety index. Calorie-dense bakery products, like cake, croissants, and doughnuts, had the lowest satiety index of all foods tested. Fruit, meat, and beans tended to have a high satiety index. Plain potatoes were off the charts—far more filling than any other food. Holt and colleagues noted that “simple ‘whole’ foods such as the fruits, potatoes, steak and fish were the most satiating of all foods tested.” Holt’s team found that the sating ability of each item was largely explained by a few simple food properties. The first is calorie density; in other words, the volume of food per calorie For example, oatmeal is mostly water, so it has a much lower calorie density than crackers, which are nutritionally similar but contain little water. The lower the calorie density of the item, the more satiety it produced per unit calorie—and the effect was extremely robust. This makes sense, because stomach distension is one of the main signals that the NTS monitors to regulate satiety. If your stomach contains more food volume, you’ll feel more full, even if that food doesn’t contain more calories. This only works up to a point, however; you can only trick the brain so much. A belly full of lettuce isn’t going to cut it.

Neck and neck with calorie density was another factor we’ve encountered before: palatability. The more palatable a food, the less filling it was. Again, this makes sense. Palatable foods are those that the brain intuitively views as highly valuable, and the brain is quite good at removing barriers to their consumption. We even have ideas about how this might work. Within the hypothalamus lies a region called the lateral hypothalamus (LH), which is a nexus between energy balance and food reward functions (among other things). Researchers have known for a long time that stimulating the LH causes animals to eat voraciously, and disrupting it makes them lean. As it turns out, palatable food activates neurons in the LH. Furthermore, the LH sends fibers directly to the NTS of the brain stem, where it inhibits neurons that play a role in satiety—so it’s not much of a leap to suppose that eating palatable foods might inhibit the very NTS neurons that make us feel full. This may be one of the reasons why we tend to overeat highly palatable foods and magically grow a “second stomach” for dessert. Sticking with simple foods can help us restrain our calorie intake without feeling hungry.

The third most influential factor Holt and colleagues identified is a food’s fat content. The more fat it contained, the less filling it was per calorie. People often find this counterintuitive, because when they eat high-fat foods, they feel extremely full. The key to understanding this is to remember that we’re talking about fullness per unit calorie. If you eat a stick of butter, you may feel full, but you will also have eaten over 800 Calories—the equivalent of two and a half large baked potatoes. Isolated fats like butter and oil are the most calorie-dense substances in the human diet by far, mostly because fat delivers nine Calories per gram versus only four for carbohydrate and protein. Isolated fats also increase the palatability of food. For these reasons, adding fat to food is a highly effective way to increase your calorie intake without increasing your satiety much, and limiting added fat helps reduce calorie intake without sacrificing satiety.

That being said, fat in general isn’t necessarily something you need to avoid if you want to control your calorie intake. Research has shown that the reason fat makes us eat more is precisely because of its high calorie density and palatability. When high-fat foods aren’t calorie dense or highly palatable, they provide the same level of satiety per calorie as high-carbohydrate foods. What this means is that if we eat fat in the context of unrefined, filling foods like meat, fish, eggs, dairy, nuts, and avocados, a higher fat intake can be compatible with a naturally slimming diet pattern. While these foods are high in fat, they don’t have the deadly combination of calorie density and extreme palatability that characterize other high-fat foods like potato chips and cookies.

A fourth critical factor that Holt’s team identified is fiber. The more fiber a food contained, the more filling it was. This explains why whole-grain bread is more filling than white bread, despite their similar calorie density.

Finally, the protein content of a food was a major contributor to satiety. This is consistent with a large body of research showing that protein is more filling than carbohydrate or fat, per unit calorie. Both the lining of the small intestine and the pancreas have the ability to detect dietary protein, and they relay this signal to the NTS. For reasons that aren’t entirely clear, this protein signal seems to play a disproportionate role in satiety. Coupled with the effects of protein on the lipostat we discussed in the last chapter, this may explain why high-protein diets help people eat less and lose fat without feeling hungry.

Holt’s results go a long way toward explaining why both rats and humans overeat spectacularly on a cafeteria diet, and why we overeat without intending to in our daily lives. The nonconscious parts of the brain that regulate satiety, including the NTS, respond to specific food properties, such as food volume, protein, fiber, and palatability. Many of our modern processed foods have properties that don’t stimulate satiety circuits to the same degree as traditional whole foods. These foods, such as pizza, ice cream, cake, soda, and potato chips, invariably boast a combination of properties that make them less filling per calorie. Since most people use the sensation of satiety as a signal to stop eating, these foods allow us to blow past the point where we’ve had enough to satisfy our calorie needs—yet we don’t even realize we’re overeating because we don’t feel any fuller at the end of the meal.

Traditional diets from many different cultures tend to have the opposite qualities: lower calorie density and palatability, and higher fiber (although not necessarily higher protein). The most striking antithesis of modern junk food, however, is the popular Paleolithic diet, inspired by the diets of our hunter-gatherer ancestors. This whole-food-based diet combines multiple sating food properties, including high protein, high fiber, low calorie density, and moderate palatability. In contrast to the caricature that most people are familiar with, the Paleolithic diet isn’t an all-meat diet (hunter-gatherers didn’t eat bacon, sorry), isn’t necessarily low in carbohydrate, and contains large amounts of whole plant foods. Research confirms that it’s highly sating and naturally promotes a lower calorie intake. In clinical trials, the Paleolithic diet has often outperformed conventional diets both for weight loss and metabolic improvements, explaining its popularity.

At the end of the paper, Holt and her colleagues put the pieces together for us: “The results therefore suggest that ‘modern’ Western diets which are based on highly palatable, low-fibre convenience foods are likely to be much less satiating than the diets of the past or those of less developed countries.” Fortunately, Holt’s findings also empower us to do something about it.

EATING FOR SATIETY

If your goal is to eat fewer calories without going hungry, high-satiety foods will help. These are items that have some combination of low calorie density, moderate palatability, high protein, and/or high fiber, such as beans, lentils, fresh fruit, vegetables, potatoes and sweet potatoes, fresh meat and seafood, oatmeal, avocados, yogurt, and eggs. The exceptionally high satiety value of potatoes is undoubtedly one of the reasons why the “potato diet” we encountered in chapter 3 is effective. But if you cover your potato with calorie-dense flavorings such as butter and cheese, or deep-fry it into french fries, all bets are off.

Yet not everyone reacts the same way to the modern food environment. Some people don’t overeat in the first place, most of us tend to overeat and gain weight, and a few lucky folks don’t gain weight despite eating prodigious amounts of food. What explains these differences?

BORN TO BE FAT

In 1976, Mats Börjeson, a researcher at Lund University, Sweden, published a paper that rocked the research community—and whose implications continue to challenge many of our dearly held notions about obesity. Börjeson wanted to understand the importance of genes in determining who develops obesity and who doesn’t—a question that had not yet been seriously investigated. To answer this question, he exploited a neat trick of biology: the fact that identical twins are genetically identical, while nonidentical (fraternal) twins share only half their genetic material. In both cases, the siblings develop in the same womb, are born at the same time, and are raised in the same family, so the only significant difference between identical and fraternal twins is their degree of genetic relatedness.

Researchers can exploit this principle to understand what proportion of a trait is genetically determined. For example, if identical twins have skin colors that are more similar than fraternal twins, it implies that skin color is genetically influenced. This makes sense: The less two people are related, the less similar they tend to be. As common sense suggests, skin color is heavily influenced by genetics.

Börjeson recruited forty identical and sixty-one fraternal twins and measured their body weights. He found that identical twins tended to have very similar body weights, while fraternal twins were more divergent. “Genetic factors,” he concluded, “apparently play a decisive role in the origin of obesity.” Since Börjeson’s study, many others have confirmed that genes have an outsized influence on adiposity. In fact, in modern affluent nations like the United States, genetic differences account for about 70 percent of the difference in body weight between individuals. They also play a prominent role in many of the details of our eating behavior, such as how much food we eat at a sitting, how responsive we are to the sensation of fullness, and how much impact food reward has on our food intake. In other words, whether a person is lean or fat in today’s world has less to do with willpower and gluttony and more to do with genetic roulette. If you want to be lean, the most effective strategy is to choose your parents wisely.

Genes also explain that friend of yours who seems to eat a lot of food, never exercises, and yet remains lean. Claude Bouchard, a genetics researcher at the Pennington Biomedical Research Center in Baton Rouge, Louisiana, has shown that some people are intrinsically resistant to gaining weight even when they overeat, and that this trait is genetically influenced. Bouchard’s team recruited twelve pairs of identical twins and overfed each person by 1,000 Calories per day above his calorie needs, for one hundred days. In other words, each person overate the same food by the same amount, under controlled conditions, for the duration of the study.

If overeating affects everyone the same, then they should all have gained the same amount of weight. Yet Bouchard observed that weight gain ranged from nine to twenty-nine pounds! Identical twins tended to gain the same amount of weight and fat as each other, while unrelated subjects had more divergent responses. Furthermore, not only did twins gain a similar amount of fat, they even gained it in the same places. If one person stored the excess fat inside his abdominal cavity—the most dangerous place to gain fat—his twin usually did the same. Not only do some people have more of a tendency to overeat than others, but some people are intrinsically more resistant to gaining fat even if they do overeat.

The research of James Levine, an endocrinologist who works with the Mayo Clinic and Arizona State University in Scottsdale, Arizona, explains this puzzling phenomenon. In a carefully controlled overfeeding study, his team showed that the primary reason some people readily burn off excess calories is that they ramp up a form of calorie burning called “non-exercise activity thermogenesis” (NEAT). NEAT is basically a fancy term for fidgeting. When certain people overeat, their brains boost calorie expenditure by making them fidget, change posture frequently, and make other small movements throughout the day. It’s an involuntary process, and Levine’s data show that it can incinerate nearly 700 Calories per day! The “most gifted” of Levine’s subjects gained less than a pound of body fat from eating 1,000 extra Calories per day for eight weeks. Yet the strength of the response was highly variable, and the “least gifted” of Levine’s subjects didn’t increase NEAT at all, shunting all the excess calories into fat tissue and gaining over nine pounds of body fat. For Levine, the study highlighted the importance of light physical activity throughout the day, which inspired him to invent the treadmill desk. Together, these studies offer indisputable evidence that genetics plays a central role in obesity and dispatch the idea that obesity is primarily due to acquired psychological traits. Yet they don’t tell us which genes are responsible for individual differences in eating behavior and obesity. For that, researchers would need different approaches.

Stephen O’Rahilly and Sadaf Farooqi haven’t been twiddling their thumbs since they identified humans who lack leptin; in the meantime, they’ve located a number of other single-gene mutations that cause severe obesity. As it turns out, nearly all of them are in the leptin signaling pathway. These genes disrupt either leptin itself, the leptin receptor, or downstream signals that mediate leptin’s actions in the brain. What’s more, now that researchers have identified the genes responsible for fattening our various rodent models of single-gene obesity, they’ve found that these genes are in the leptin signaling pathway as well. It’s not every day in science that results from several related fields of research converge in such a powerful way. In this case, they all say in chorus: Leptin signaling in the brain is a key component of the biological control of adiposity.

O’Rahilly and Farooqi focus on children with extreme obesity. These are the cases where something really seems to be broken, and it isn’t so farfetched to suppose that a severe genetic disruption could be responsible. Among these patients, about 7 percent show evidence of such genetic disruptions. The most common mutation disrupts the melanocortin-4 receptor, which is primarily responsible for the appetite-suppressing effects of melanocortins in the brain (the substances released by POMC neurons). In these children, melanocortin release triggered by their own circulating leptin fails to restrain their appetites, so they eat substantially more than typical children. However, according to Farooqi, known mutations only account for about 1 percent of adults admitted to an obesity clinic. They don’t explain the obesity epidemic, yet, Farooqi adds, “we think there’s an awful lot more to find.” Most researchers think destructive single-gene mutations only account for a small fraction of overeating and obesity. Yet at the same time, we know that genetics explains much of the differences in eating behavior, and most of the differences in adiposity, between individuals. Where are the missing genes? To answer this question, we have to turn to another type of genetics research that studies common gene variation rather than rare, catastrophic mutations that disrupt entire signaling pathways.

Many genes come in several different flavors, called alleles, each of which is encoded by a slightly different DNA sequence. Sometimes, all alleles of a particular gene are functionally identical, but in other cases, they have subtly different functions. For example, different alleles of the same set of genes determine our eye colors and blood types. This common form of genetic variation explains much of why individual humans look and act differently from one another.

Researchers have figured out ways to determine which genes have different alleles that lead to different body weights. Unlike O’Rahilly and Farooqi’s studies, which focus on selected cases of extreme obesity, these methods are aimed at identifying the genetic factors that determine adiposity in the general population.

Remarkably, it turns out that a boatload of genes influence adiposity in humans, but each individual gene only has a small impact. So far, researchers have identified nearly one hundred genes that influence adiposity—yet together, these genes explain less than 3 percent of the differences in adiposity between people. Clearly, there’s still work to do. However, by looking at the genes that have been identified so far (which were identified precisely because they’re the most influential), we can gain unbiased insights into the biological processes that underlie garden-variety obesity. By now, you should have a pretty good guess what organ these genes tend to influence: the brain. Although some genes relate to other processes such as fat metabolism, the lion’s share exert their influence in the brain, and of those, many act via brain circuits that are already known to regulate food intake and adiposity (such as POMC neurons and their downstream targets). This suggests that genetic differences in brain function are the primary reason why some people are fatter than others.

Well, I suppose that’s that: If you have fat genes, you’re destined to be fat. Or are you? A century ago in the United States, people carried the same genes we do today, yet few people had obesity. What has changed isn’t our genes, it’s our environment—our food, our cars, our jobs. This leads us to a critical conclusion about obesity genes: In most cases, they don’t actually make us fat, they simply make us susceptible to a fattening environment. In the absence of a fattening environment, they rarely cause obesity. As Francis Collins, geneticist and director of the National Institutes of Health, is fond of saying, “Genetics loads the gun, and environment pulls the trigger.” Unless you have a faulty gun, which is rare, if you don’t pull the trigger, it doesn’t discharge.

A few lucky people are so genetically resistant to obesity that they’re unlikely to develop it under any circumstances. A few others are so genetically susceptible that they may carry excess fat even in a very healthy environment. For the rest us, the environment in which we find ourselves has a major impact on our weight.