Fed versus fasted: how hormones switch your cells between burning and building
A biochemist, an obesity researcher and a longevity geneticist describe the same hormonal switch, with insulin, leptin and related signals telling cells whether to store, build or burn fuel. Understanding it clarifies why fasting, weight regain and metabolic disease are so tightly linked.
The Short Version
Your body runs on a hormonal switch that tells every cell whether food has just arrived or is running short, and each cell responds by storing, building or burning fuel. Biochemist Jared Rutter describes insulin and GLP-1 as the fed signal and glucagon as the fasting signal, with fat cells storing after a meal and releasing fuel during a fast [1]. Obesity researcher Zachary Knight describes the long-term version of the same switch: leptin, made by fat in proportion to how much you carry, tells the brain how full the reserves are, and when it falls after weight loss the brain launches a starvation response that drives weight back up [2]. Longevity geneticist David Sinclair describes a third face of it: low insulin during fasting switches on the sirtuin longevity genes and turns down the growth pathway mTOR, while constant feeding keeps those defenses off [3]. Seen together, these accounts explain why fasting, weight regain and diseases like diabetes, heart failure and cancer are so tightly linked: they are all consequences of how the body reads, and sometimes misreads, the fed-versus-fasted signal [1][2][3].
Table of contents
What They Said
Jared Rutter: every cell decides whether to burn or build
Jared Rutter, a University of Utah biochemist and Howard Hughes Medical Institute investigator, starts from a reframing: you don't have one metabolism [1]. As the episode introduction puts it, "Your metabolism, as it's called, is actually a reflection of the constellation of all the metabolisms of all the cells in your body" [1]. Whole-body metabolism, Rutter says, is basically "the sum total of what we ingest, what we eat, what we drink, what we breathe," which gets broken down into molecules like "amino acids and sugars and so forth that then distribute throughout the body, go into individual cells" [1]. Each of roughly 30 trillion cells then decides what to do with them [1].
Hormones coordinate the fed state. After a meal, signals like GLP-1 and insulin tell cells throughout the body that food has arrived [1]. Each cell responds differently: "The response of different cells to the fed state is different depending on the needs and functions of that cell. Again, some cells don't care at all, they're gonna just go about and do their business, and some cells completely rewire their function depending on the metabolic state, the fed fasted state of the organism" [1]. Fat cells respond strongly, taking up glucose, converting it to fat and storing it safely [1]. Rutter sees an evolutionary logic: "those adipocytes, full of fat, from when we ate, probably kept our ancestors alive when they went through the periods of prolonged fasting" [1].
Huberman found this reframed insulin for him. Most people, he said, think of insulin as a shuttle for sugar, but "the size of that signal is saying what's likely to be there," which can prepare each cell for its job, whether that means a fat cell squirreling energy away or a brain cell getting ready to fire [1].
Glucagon flips the switch. "There are hormones to do the opposite," Rutter says. "Glucagon is called a fasting hormone," and "glucagon in many ways does the opposite of insulin" [1]. It tells fat cells to release their stored fat so other tissues can use it [1]. He notes glucagon has become more popular recently because it's now being combined into some of the newer GLP-1 drugs [1].
Different tissues burn different fuels. The heart is a major destination for released fat: an estimated 70 to 80 percent of its energy extraction comes from fat, and Rutter calls it an omnivore that will burn fats, glucose, lactate, ketones and amino acids [1]. Neurons, by contrast, rely mostly on glucose [1]. That explains a grim asymmetry in diabetes: high blood sugar does damage over years, while blood sugar that drops too low can kill within minutes [1]. Rutter adds that fatty acids tend to be burned first because excess free fatty acids are acutely toxic, while high glucose causes harm more slowly [1].
Pyruvate is the fork in the road. Glucose is broken down through glycolysis into pyruvate, which Rutter calls a pivot point [1]. Pyruvate can enter the mitochondria and be burned for as much ATP as possible, as heart muscle cells do, or it can feed into building materials for new cells [1]. "Food can either be converted to energy or it can be converted to biomass," he says, calling this a simplified but useful baseline [1]. Huberman offers an analogy they both use: lumber that can either build more house or be burned for heat [1]. Cells don't follow a fixed split; they constantly monitor their resources, especially ATP, and when ATP runs low they shut down ATP-consuming processes and pull in more glucose [1].
Cells are specialized for this choice. A heart muscle cell, which barely turns over and must contract for a lifetime, has mitochondria wired to burn whatever is available [1]. Intestinal stem cells, which renew the gut lining every five to seven days, have mitochondria wired to produce biomass [1]. Work from Craig Thompson at Sloan Kettering suggests a single cell can even contain two kinds of mitochondria, one more biosynthetic and one more focused on ATP [1].
When the switch goes wrong, disease follows. Rutter co-discovered the mitochondrial pyruvate carrier (MPC), the gate pyruvate uses to enter mitochondria [1]. Mice lacking it in the heart live for weeks, then die of heart failure with massively enlarged hearts [1]. They don't appear to die from lack of ATP, since they can still burn fat [1]. Instead, Rutter suggests somewhat speculatively that the heart cells make a pathological allocation decision, turning glucose into biomass and growing [1]. Nearly every human who dies of heart failure has a big, dilated heart [1]. Cancer shows the same tilt: Otto Warburg saw in the 1920s that cancer cells consume less oxygen than expected and blamed broken mitochondria, but Rutter says those mitochondria aren't broken; they are very good at making stuff, which is what a dividing cancer cell needs [1]. He lists cancer, enlarged hearts and hyperactivated immune cells that drive inflammatory disease as settings where shifting toward building rather than burning is pathological [1].
Too much fuel carries its own cost. Running metabolism "is expensive not only in terms of what we need to eat to fuel it, but it's expensive in terms of the damage that can come as a side effect of that," Rutter says, and "the accumulation of that damage over time is certainly correlated strongly with aging" [1]. Responding to Layne Norton's idea of energy toxicity, he describes a widely accepted, though not universal, hypothesis: overpowered mitochondria are prone to generating reactive oxygen species, which damage DNA and proteins and may contribute to pathologies including aging [1].
Zachary Knight: leptin, the long-term fat gauge
Zachary Knight, a UCSF physiologist and HHMI investigator, describes the brain side of the fed-fasted switch across two timescales [2]. A fast brainstem system handles a single meal, deciding when to stop eating, while a slow hypothalamic system tracks body fat over weeks, months and years [2]. The key evidence comes from Harvey Grill's decerebrate rat experiment about 50 years ago: rats with the brainstem cut off from the rest of the brain still regulate meal size, but after a couple of days of fasting they don't eat a bigger meal [2]. The forebrain is what reads the body's longer-term energy need [2].
Why fat? Fat is the body's energy reserve, so the brain needs to know how much remains to judge how urgent the next meal is [2]. The signal is leptin, cloned in 1994 by Jeff Friedman [2]. Its story began at Jackson Laboratory, where Doug Coleman connected the circulations of two strains of obese mice [2]. "The obese mouse, that strain dramatically loses weight. In fact, within a couple of weeks, it looks like a normal mouse. It just stops eating," Knight says, while "the DB mouse, nothing really happens" [2]. Coleman guessed one strain lacked a hormone and the other its receptor, and he was right [2].
Leptin's design makes it an ideal gauge. "How much it's expressed is directly proportional to how much body fat you have. So as you gain weight, the expression of this hormone increases in a linear manner," Knight explains; the level in your blood "is a direct read out of your body fat reserves" [2]. Its functional receptor is expressed almost exclusively in the brain, in the regions known to control appetite, so "the expression of this receptor gives you a map in the brain of the neurons that control hunger" [2]. When you lose fat, leptin falls, and the brain launches a full starvation response: more hunger, lower energy expenditure, lower body temperature, less spontaneous movement and even reduced fertility [2].
Leptin resistance mirrors insulin resistance. Amgen paid about $20 million up front for the leptin patent, but obese people given leptin did not lose much weight [2]. Most obese people already "have high levels of leptin. And so what they have is the state of leptin resistance," Knight says. "So it's analogous to someone who has type two diabetes. It's not because they lack insulin. It's because they actually have over time a high level of insulin. And so target tissue stop responding to insulin. The thought is that it's the same way in obesity and leptin" [2]. A reanalysis confirmed it: "the people with the lowest levels of leptin, they lost the most weight when you gave them this drug. And the people with the highest levels of leptin lost the least weight" [2].
AgRP neurons carry the fasting signal. In the hypothalamus, a few thousand AgRP neurons drive hunger: stimulate them in a well-fed mouse and it eats voraciously; silence them and it will starve even with food available [2]. Leptin inhibits them, so they are more active when fat is low [2]. They also respond to calories, not their source: equal calories of sugar, fat or protein in the stomach inhibit them to the same extent [2].
Why lost weight comes back. Here the switch works against the dieter. After weight loss, AgRP neurons become more active, Knight says, partly because leptin levels in the blood are lower, creating "that counter regulatory drive that drives you to then consume more food" [2]. Efforts to design diets that consistently keep weight off have, in his words, "been very unsuccessful" [2]. He cites two lines of evidence:
- Energy expenditure falls. For every kilogram lost, daily energy expenditure falls by about 30 kilocalories [2]. In Rudy Leibel's study about 25 years ago, people who had lost around 100 pounds burned about 25% less energy than never-obese people of the same size and body composition, so "they have to eat 25% less than someone who looks the same as them" to hold their weight, and the gap persisted for at least a year [2]. It is unclear whether those people always had a slower metabolism or whether "being in a higher weight for a longer period of time changes the brain so that then once you lose the weight it's irreversible" [2].
- Hunger rises more. Kevin Hall used SGLT2 inhibitors, diabetes pills that make people lose about 90 grams of glucose a day in urine without noticing, to produce covert weight loss, then back-calculated food intake [2]. The result: "for every two pounds or so of weight you lose your hunger goes up by 100 calories per day" [2]. Knight concludes that "the increased hunger seems to be the main reason people find it so difficult to keep weight off" [2].
The fed-state hormones as drugs. GLP-1 began as part of the incretin effect, the gut's extra signal that amplifies insulin release after glucose is eaten rather than injected [2]. Natural GLP-1 lasts about two minutes, and DPP-4 inhibitors that raise it about threefold treat diabetes but don't cause weight loss, which tells Knight that GLP-1's natural job isn't really body weight control [2]. Engineered drugs like semaglutide, with a seven-day half-life, produce about 16% loss of body weight over roughly a year, acting mainly on brainstem appetite regions at concentrations a thousand to ten thousand times higher than natural [2]. Knight notes the old dogma held that only the long-term hypothalamic system could change body weight; GLP-1 drugs proved that hitting the short-term system 24 hours a day works [2]. He suggests a future approach might use a GLP-1 drug to lose weight and then switch to a leptin or hypothalamus-based drug to defend the new weight [2].
The GIST of every new summary in your inbox, daily or weekly. Free, and you can unsubscribe any time.
David Sinclair: fasting flips on the longevity genes
David Sinclair, a Harvard geneticist, describes what the fasted side of the switch does for aging [3]. He calls the idea "that people should never be hungry" a big mistake [3]. "We live in a world now with at least three meals a day and then we've got companies selling bars and snacks in between. So the feeling of hunger, some people never experience hunger in their whole lives. It's really, really bad for them" [3]. He traces the habit to "the 20th century view that you don't want to stress out the pancreas and you try to keep insulin levels pretty steady" [3].
Animals that don't eat constantly live longer. "When you look at first all animals, whether it's a dog or mouse or a monkey, the ones that live the longest by far, 30% longer and stay healthy are the ones that don't eat all the time," Sinclair says [3]. He recalls Clive McCay's 1930s experiment, which put cellulose in rats' food so they got fewer calories, and "those rats lived 30% longer" [3].
Low insulin switches on the sirtuins. Sinclair's lab "showed in 2005 in a science paper that if you have low levels of insulin and another molecule called insulin-like growth factor, those low levels turn on the longevity genes" [3]. These are the sirtuins, a family of seven genes [3]. The flip side: "by having high levels of insulin all day, being fed means your longevity genes are not switched on" [3]. He describes a study by Rafael de Cabo at the NIH in which mice fed within one hour a day lived dramatically longer than mice eating freely, regardless of diet, and concludes: "it's not as important what you eat. It's when you eat during the day" [3].
mTOR is the building sensor. Sirtuins respond mainly to sugar and insulin, while mTOR senses amino acids, especially leucine, isoleucine and valine [3]. "And they talk to each other," Sinclair says; when you fast, you get sirtuin activation, and "through lack of amino acids" the body turns down mTOR, which "turns on all of the body's defenses," including clearing old proteins, improving insulin sensitivity, giving more energy and repairing cells [3]. He calls these two pathways the most important for longevity [3].
Growth now, cost later. Asked whether leucine, because it triggers cellular growth, is pro-aging, Sinclair answers: "Well, it could be. That's what the evidence suggests" [3]. He puts it in the same category as growth hormone and testosterone: "You'll bulk up more. You'll feel better immediately. But based on the research, it's at the expense of long-term health. So my view of longevity, the way I treat my body is, I don't burn both candles" [3]. On growth hormone specifically: "it's like burning your candle at both ends. Ultimately, if you want to live longer, you want less of that" [3]. Mice with low growth hormone live longest; one, given calorie restriction plus a dwarf mutation, lived five years instead of the usual "two and a bit" [3].
Pulsing rather than constant restriction. Sinclair's key idea is alternating periods of fasting, eating, supplements and exercise so the body perceives adversity [3]. Huberman relays Ori Hofmekler's view that "fasting is wonderful, but these pulses where you nourish the body or even slightly overnourish the body provided they aren't too frequent. They have a tremendous effect on vitality," describing it as "famine and feast in lowercase letters" [3]. Sinclair skips breakfast, drinks water, tea and coffee through the day, eats one mostly vegetable-based dinner, often within about a two-hour window, and fasts for two days about once a month, since longer fasts trigger a deeper cleanup process called chaperone-mediated autophagy [3]. He warns the first two to three weeks are hard and advises building up gradually rather than going cold turkey [3]. He also takes metformin, a prescription diabetes drug that mimics low energy, and calls berberine the poor man's metformin, citing human trials showing it improves insulin sensitivity [3].
Obesity and the fuel for longevity genes. Sirtuins need NAD as fuel, and NAD declines with age, obesity and never being hungry [3]. Huberman asks why obesity would lower NAD, noting fat "is stored fuel that's acting as an endocrine organ, essentially" [3]. Sinclair is candid: "that we don't know, but I'll give you my best answer, which is that obesity comes along with a lot of problems that include a lot of senescent cells" [3].
Where They Agree, and Where They Don't
They describe the same switch from three vantage points. Rutter's cells hear insulin and GLP-1 as a sign that food has arrived and glucagon as a cue to release the stores [1]. Knight's brain hears leptin as a readout of fat reserves and responds to its fall with hunger and energy saving [2]. Sinclair's genes hear low insulin as a cue to switch on defenses and turn down growth [3]. In each case a hormone reports the body's fuel state and tissues decide whether to store, build, burn or protect.
They agree that insulin is a signal, not just a sugar shuttle. Huberman makes this point explicitly with Rutter [1], Knight uses insulin resistance to explain leptin resistance [2], and Sinclair treats insulin levels as the switch for longevity genes [3].
They agree that resistance is central to metabolic disease. Knight draws the direct parallel: chronically high insulin leads tissues to stop responding in type 2 diabetes, and chronically high leptin does the same in obesity [2]. Sinclair's case for fasting is partly that constant feeding keeps insulin high all day [3], and he lists improved insulin sensitivity among the defenses fasting turns on [3].
They agree that building has a price. Rutter describes pathological shifts toward biomass in cancer, enlarged hearts and inflammatory immune cells [1]. Sinclair argues that growth signals like leucine, growth hormone and testosterone trade long-term health for short-term gains [3]. Huberman draws the connecting thread with Rutter, suggesting "there does seem to be this balance between size and the use of fuel to to build more of oneself and the use of fuel to just be oneself" [1]. Rutter frames it more cautiously, as a question of cell identity [1].
Where they pull in different directions: hunger. Sinclair treats hunger as healthy and says never feeling it is "really, really bad" for people [3]. Knight treats hunger after weight loss as the main force that defeats dieters, rising about 100 calories a day for every two pounds lost [2]. These aren't strictly contradictory. Sinclair is describing short, deliberate fasts in someone at a stable weight, which switch on repair pathways [3]; Knight is describing the chronic, leptin-driven hunger of someone whose fat stores have shrunk, which the brain reads as starvation [2]. But the tension matters for anyone using fasting to lose weight: the same fasted signal that may help cells can also push a lighter body to eat its way back.
Where they differ on protein and growth. Sinclair presents lower mTOR activity, partly from fewer amino acids like leucine, as beneficial for longevity [3]. Knight's everyday advice is to get adequate protein, because of its greater satiety and higher thermic effect, and because people tend to keep eating until they reach a protein minimum [2]. He also notes that 25 to 33% of weight lost through diet or drugs is typically muscle unless people eat enough protein and do resistance training [2]. Sinclair himself lifts weights to maintain muscle and hormones [3] and uses pulsing, not permanent restriction, which partly bridges the gap.
Where the evidence is firmer or thinner. Knight's weight-regain numbers come from human studies [2]. Rutter's heart-failure interpretation is labeled somewhat speculative, and his energy-toxicity idea is widely but not universally accepted [1]. Much of Sinclair's longevity evidence comes from animals, and he openly says the link between obesity and low NAD is not understood [3].
Why It Works
Fat as the reserve, leptin as the gauge. Because fat is the body's main energy reserve, a hormone made in proportion to fat is a natural way to tell the brain how much fuel remains [2]. Low leptin disinhibits AgRP neurons, and the brain responds as if facing starvation: more hunger, less energy spent, lower body temperature, less movement and lower fertility [2].
Insulin as the fed-state broadcast. After a meal, insulin and GLP-1 tell cells nutrients are available; fat cells store them, other cells rewire or ignore the signal based on their jobs [1]. Glucagon reverses the message so fat is released for tissues like the heart [1].
Pyruvate as the burn-or-build valve. Inside each cell, pyruvate either enters mitochondria through the MPC and is burned for ATP, or is diverted into building material [1]. Making lactate instead of burning pyruvate keeps carbon available for building, since burning releases it as carbon dioxide we breathe out [1]. Blocking the MPC in heart cells tilts them toward growth and leads to enlarged, failing hearts [1].
Sirtuins and mTOR as the repair-or-grow switch. Low insulin and IGF turn on sirtuins, and scarce amino acids turn mTOR down; together they activate cleanup, insulin sensitivity and cell repair [3]. High insulin all day keeps sirtuins off [3], and growth signals push toward short-term building at a long-term cost [3].
Resistance from chronic signals. When insulin or leptin stays high for a long time, target tissues stop responding [2]. That is why obese people with high leptin barely respond to more leptin, while people with low leptin do [2].
Defended set points explain regain. Lower leptin, more active AgRP neurons, about 30 fewer kilocalories burned per kilogram lost and about 100 more calories of appetite per two pounds lost combine to push weight back up [2]. Knight and Huberman sum it up: "It's hard to beat homeostasis" [2].
Excess fuel generates damage. Overpowered mitochondria may generate reactive oxygen species that damage DNA and proteins [1], and running metabolism carries a cost in accumulated damage correlated with aging [1]. This is one proposed link between constant surplus and aging [1][3].
Where It Applies, and Where It Doesn't
It applies to understanding weight regain. The fed-fasted switch explains why diets so often fail in the long run: after fat loss, leptin drops and the brain defends the old weight through hunger and reduced energy use [2]. This is biology, not a failure of willpower.
It applies to metabolic disease. Insulin resistance in type 2 diabetes and leptin resistance in obesity reflect chronic signals that tissues stop hearing [2]. Rutter adds that neurons' dependence on glucose explains why both high and low blood sugar are dangerous, on very different timescales [1], and that misallocation toward building shows up in cancer and heart failure [1].
It applies to thinking about fasting. Sinclair's case for skipping meals rests on low insulin activating longevity genes, supported mostly by animal studies of eating less often [3].
It doesn't mean fasting is a weight-loss shortcut. Knight's evidence suggests the body resists fat loss regardless of method, with rising hunger as the main obstacle [2]. And some people do keep weight off for reasons Knight says aren't understood [2].
It doesn't mean building is always bad. Intestinal stem cells and activated immune cells must build to do their jobs [1], and muscle preservation through protein and resistance training matters during weight loss [2]. The problem is building in the wrong place or at the wrong time [1].
It doesn't reduce to calories in, calories out. Rutter reframes metabolism as trillions of cellular decisions rather than a simple equation [1], and Knight's findings show that the brain actively changes both sides of that equation after weight loss [2].
Many details remain open. Rutter is honest about how much remains unknown in cell metabolism [1], Knight doesn't know whether reduced energy expenditure after weight loss is lifelong or reversible [2], and Sinclair doesn't know why obesity lowers NAD [3].
Anyone with diabetes, a history of eating disorders, or on medication that affects blood sugar should talk to a doctor before trying fasting or a new diet.
What To Do With It
Leave room to feel hungry. Sinclair argues constant eating keeps insulin high and longevity genes off, and suggests skipping one meal a day, ideally breakfast or dinner so the fast joins sleep [3]. Build up gradually, dropping breakfast first and then shrinking lunch, rather than going cold turkey, and expect the first two to three weeks to be hard [3].
Stop snacking between meals. Sinclair singles out bars and snacks between three daily meals as part of the problem [3], and Rutter's account shows every meal sends a fresh fed signal to your cells [1].
Plan for hunger after weight loss. Knight's numbers suggest appetite rises about 100 calories a day for every two pounds lost, and energy expenditure falls about 30 kilocalories per kilogram [2]. Set up your food environment in advance so that rising hunger meets whole foods rather than easy snacks.
Eat enough protein and lift weights while losing weight. Of weight lost through dieting or drugs, 25 to 33% is typically muscle, and adequate protein plus resistance training can almost completely prevent that [2]. Sinclair also lifts to maintain muscle [3].
Favor whole foods over ultra-processed ones. In Kevin Hall's inpatient studies, people ate much more of ultra-processed food and gained weight, then lost it when switched [2]. Knight's own everyday advice is fewer ultra-processed foods and more whole foods [2].
Pulse rather than restrict forever. Sinclair's approach alternates fasting with periods of eating well, and Huberman relays the idea of occasional, not too frequent, periods of generous nourishment [3].
Be cautious with constant growth signals. Sinclair suggests that supplementing leucine, growth hormone or testosterone may trade long-term health for short-term gains [3]. Discuss any hormone therapy with a physician.
Track the signals, with your doctor. Sinclair recommends tracking HbA1c for average blood sugar and high-sensitivity CRP for inflammation [3]; metformin, which he takes, is a prescription drug that should only be used under medical supervision [3].
See weight regain as biology, not failure. Knight's evidence shows the brain actively defends lost fat through leptin and AgRP neurons [2]. If regain keeps happening, talk with a doctor about options, including the GLP-1 drugs, which work where diets alone often don't [2].