The Modern Guide to Metabolism

For most of the twentieth century, metabolism was discussed in the vocabulary of a furnace. Fuel went in, heat came out, and the rate at which the furnace ran determined the leftover energy that became fat. The framing was tidy and largely wrong — not in its arithmetic, but in its physiology. A real human metabolism is a network of regulated, adaptive systems with multiple feedback loops, responding to caloric availability with the kind of dynamic adjustment that frustrates anyone trying to model it with a single number. The shift from the furnace model to the regulated-network model has been driven by half a century of careful measurement, and what follows is the working map that emerges from that literature.

What Is Metabolism?

Metabolism, in the strict biochemical definition, is the sum of all chemical reactions that occur in a living organism to maintain life. Those reactions split into two broad categories. Catabolism breaks larger molecules into smaller ones, releasing energy in the process — the breakdown of glucose into pyruvate, the oxidation of fatty acids, the digestion of dietary protein into amino acids. Anabolism does the opposite, using energy to assemble smaller molecules into larger ones — building muscle protein from amino acids, synthesising glycogen from glucose, constructing the lipid membranes of new cells. Catabolic and anabolic reactions run simultaneously in every tissue, and the balance between them determines whether the body is net storing or net breaking down at any given moment.

In the popular sense, however, "metabolism" usually means something narrower: total daily energy expenditure (TDEE), the calories a person burns over twenty-four hours. TDEE has three main components, and understanding their relative sizes is the first useful corrective to most metabolism folklore.

Basal metabolic rate (BMR) is the largest component, accounting for roughly 60 to 70% of TDEE in most adults. BMR represents the energy required to keep the body running at complete rest — to maintain body temperature, pump blood, repair tissue, run the kidneys and liver, fire the nervous system, and sustain the thousands of background biochemical processes that continue whether or not a person moves. Even someone lying perfectly still in bed for twenty-four hours would expend approximately their BMR.

The thermic effect of food (TEF) — sometimes called diet-induced thermogenesis — accounts for roughly 10% of TDEE. TEF is the energy cost of digesting, absorbing, and processing the food a person eats. It is not constant across foods: protein has the highest thermic effect (about 20 to 30% of the calories in a protein-rich meal are spent processing that meal), carbohydrates a moderate one (5 to 10%), and fat the lowest (about 0 to 3%). This is why high-protein meals modestly raise total energy expenditure compared with isocaloric high-fat ones, and why some popular claims about food being "metabolism-boosting" usually trace back, at best, to a small TEF effect.

The remaining 15 to 30% of TDEE comes from physical activity, which splits into two pieces. Deliberate exercise is one. The other, larger for most non-athletes, is non-exercise activity thermogenesis (NEAT): the energy spent in spontaneous movement — fidgeting, standing, gesturing, walking to the kitchen. NEAT varies dramatically between individuals and within the same individual across different periods, and is one of the components the body quietly adjusts during caloric restriction, which is part of why simple calorie arithmetic diverges from real-world weight outcomes.

The popular framing of "fast" versus "slow" metabolism almost always conflates these components. When someone says a friend "has a fast metabolism," they usually mean one of several things: the friend is taller or heavier, has more lean muscle, moves more in unmeasured ways, eats less than the observer estimates, or some combination. Outright differences in BMR between two adults of identical body composition, age, and sex are real but modest — usually within plus or minus 200 to 300 calories per day. Meaningful at the margin, but not the dramatic gap people often invoke to explain weight differences.

At the cellular level, the engine driving all of this is the mitochondrion. Each one converts the chemical energy in glucose, fatty acids, and (sometimes) amino acids into ATP — the universal energy currency the cell uses to power every active process. The efficiency of conversion is not fixed. Mitochondria can decouple electron transport from ATP synthesis through uncoupling proteins, dissipating the energy as heat. Brown adipose tissue is densely packed with UCP1, the most studied uncoupling protein, which is why brown fat has become a popular target for metabolism speculation. The proportion of metabolism running through brown fat in adult humans turns out to be much smaller than the early enthusiasm suggested.

Metabolism is, in short, not a single number a person has. It is a system with multiple compartments, each regulated separately. The interesting question is rarely "is my metabolism fast or slow" — it is usually "which compartment am I asking about, and what has been happening to it recently."

Basal Metabolic Rate Explained

Basal metabolic rate is the most carefully defined of metabolism's components. It is measured under strictly standardised conditions: the subject must be awake but at complete rest, in a thermoneutral environment, having fasted for at least twelve hours, and in a stable physiological state. Anything less than this, and what is measured is technically resting metabolic rate (RMR), which runs about 10% higher than true BMR because the body is not in its fully basal state. In clinical and research practice the two terms are often used interchangeably, though purists distinguish them.

Two measurement techniques dominate. Indirect calorimetry measures oxygen consumption and carbon dioxide production; because cellular metabolism has predictable stoichiometry, gas exchange yields a calculable energy expenditure. A modern metabolic cart produces a BMR estimate in under thirty minutes. The gold standard for total daily expenditure is doubly-labeled water, introduced by Dale Schoeller at Wisconsin in the early 1980s — subjects drink water labeled with two stable isotopes, and the differential rate at which the isotopes leave the body yields an integrated CO₂ production measure over one to two weeks. It is the only technique that captures real-world TDEE without confining the subject to a calorimeter, and most modern cross-cultural energy expenditure data (including the Pontzer work discussed below) rests on it.

What determines a person's BMR? The dominant variable, by a wide margin, is fat-free mass — sometimes called lean body mass. Skeletal muscle, organs, bone, and the brain are all metabolically active tissues. The brain alone, despite weighing about 2% of body mass, accounts for roughly 20% of resting energy expenditure in adults. The liver and kidneys together account for another 30 to 40% per unit weight, which is why these organs punch above their mass in metabolic terms. Skeletal muscle, by contrast, is metabolically modest at rest — about 13 kcal per kilogram per day — though its sheer mass means it still contributes meaningfully to BMR in muscular individuals.

Steven Heymsfield, at Pennington Biomedical Research Center, has spent much of his career quantifying how organ and tissue masses contribute to BMR. His work shows that variation in BMR between individuals is largely accounted for by variation in the relative proportions of these tissues. A person with more muscle, larger organs, and a larger brain has a higher BMR; a person who is smaller across the board has a lower BMR. This relationship is so consistent that BMR-prediction equations based on body weight, height, age, and sex achieve reasonable accuracy in most populations.

Beyond lean mass, several other variables matter. Age exerts a modest effect, though as we will see in the section on Pontzer's 2021 work, the magnitude of this effect has been substantially overestimated. Sex matters because men, on average, have higher lean mass and lower fat mass than women of the same body weight, producing a higher BMR per kilogram. Genetics contribute meaningfully — twin studies attribute roughly 30 to 40% of the variance in BMR to inherited factors after controlling for body composition. Thyroid hormone exerts the largest single hormonal effect on BMR; both clinical hyperthyroidism and hypothyroidism produce measurable BMR shifts, though within the normal range of thyroid function the contribution is smaller than popular discussion implies.

This is why two people of the same body weight can have different BMRs, sometimes substantially. Body weight is the rough sum of fat mass, lean mass, and water. Two people who weigh 80 kg can have very different ratios — one with 50 kg of lean mass and 25 kg of fat, another with 40 kg of lean and 35 kg of fat. The first person will have a higher BMR despite the identical scale weight. Add in differences in organ size, thyroid status, and genetic background, and the spread becomes wider still.

The clinical prediction equations work imperfectly within this spread. The Harris-Benedict equation, derived in 1919 by James Arthur Harris and Francis Benedict from a relatively small sample of mostly young, lean Americans, was the standard for most of the twentieth century. It systematically overestimates BMR in modern populations, partly because body composition has shifted in the century since. The Mifflin-St Jeor equation, published by Mark Mifflin and Sachiko St Jeor in 1990 in the American Journal of Clinical Nutrition, replaced it as the recommended default because it more accurately reflects contemporary BMR in adults across a wider range of body sizes. Both equations remain prediction tools, however, not measurements. They predict the population average for a person of given weight, height, age, and sex. Individual deviations of 10 to 20% in either direction are common, which is why anyone making important decisions based on BMR estimates is better served by indirect calorimetry than by an equation.

How much variation is real, and how much is measurement noise

One reason BMR estimates fluctuate is that measurement itself is imperfect. The conditions required for true BMR are difficult to achieve in a clinical setting; what most people measure is RMR, which inherits some day-to-day variability from sleep quality, recent food intake, ambient temperature, hydration, menstrual cycle phase in women, and the stress of being in a measurement environment. Repeat measurements on the same person can differ by 5 to 10% across days without anything biological having changed.

Beyond noise, however, real BMR variation between similar individuals does exist. Studies measuring BMR in genetically related individuals, twins, and matched pairs consistently find residual variation that cannot be explained by body composition alone — typically on the order of 150 to 300 calories per day at the extremes of a normal distribution. This is enough to influence weight trajectory at the margin, but it is not enough to account for the dramatic weight differences sometimes attributed to "slow metabolism." Most of those differences trace to behavioural and environmental factors that the person experiencing them is not in a position to see, including NEAT, food intake estimation error, and sleep. The question of whether "slow metabolism" is real is one of the more useful entry points to the broader literature.

Energy Balance

The first law of thermodynamics says energy cannot be created or destroyed, only transformed. Applied to a human body, this means that any caloric energy taken in but not expended must be stored — usually as fat. The arithmetic is unimpeachable. The problem, as the literature on energy balance and weight regulation demonstrates, is that neither side of the equation behaves as a fixed quantity in a regulated biological system.

Start with energy in. The calories listed on a food label are derived from bomb calorimetry — burning the food in a sealed container and measuring the heat released. The human digestive system is not a bomb. It absorbs different macronutrients with different efficiencies, and some calories ingested never reach the bloodstream at all. Whole almonds, for example, deliver roughly 25% fewer absorbed calories than their labelled value because much of the fat remains trapped inside intact cell walls and is excreted. Soluble fibre alters the absorption of other macronutrients eaten alongside it. The gut microbiome ferments otherwise indigestible carbohydrates into short-chain fatty acids that contribute a small but real number of calories that varies between individuals.

Kevin Hall's group at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) has done some of the most rigorous work demonstrating that "calories in" is not as fixed as the food label implies. The food matrix — the physical structure of the food, the degree of processing, the presence of intact cell walls — affects absorption substantially. The same nutrients in a different form do not necessarily yield the same number of absorbed calories. Hall's 2019 study comparing ultra-processed and unprocessed diets in a metabolic ward, discussed at length below, showed not only that participants ate roughly 500 calories per day more on the ultra-processed diet, but that the metabolic and hormonal responses to the two diets differed in ways that affected expenditure as well as intake.

Now consider energy out. We have already seen that TDEE has at least four components — BMR, TEF, deliberate exercise, NEAT — and that each can shift in response to changes in the others. Cut intake, and NEAT tends to decline; the person spends slightly longer sitting, takes slightly fewer spontaneous steps. Increase exercise, and BMR may decline somewhat as the body adjusts its baseline downward (Pontzer's "constrained" model, discussed below). Eat more protein, and TEF rises. None of these adjustments happens consciously, and most happen on timescales that make them invisible to the person experiencing them.

The body also runs compensatory responses at the hormonal level. Reduced intake triggers hormonal shifts that increase hunger and reduce satiety — ghrelin climbs, leptin drops, peptide YY and CCK responses attenuate. These shifts produce a sustained biological pressure toward eating more, which over months and years tends to close the original deficit. The pressure is not psychological in origin; it is endocrine. Behavioural willpower is being asked to override a hormonal current.

This is why calories-in-calories-out (CICO) is often described in modern obesity medicine as "true but incomplete." The thermodynamic principle is correct. Sustained weight loss does require sustained energy deficit. What CICO leaves out is that both intake and expenditure are dynamic variables, each adjusted by feedback systems whose job, evolutionarily, is to defend stored energy. The person trying to maintain a deficit is not interacting with a passive ledger. They are interacting with a regulated system actively renegotiating the terms.

The practical consequence is that a calculated 500-calorie deficit on paper does not reliably produce a 500-calorie real-world deficit at month three. By that point, BMR has fallen further than weight loss alone would predict, NEAT has declined, TEF has fallen because intake is lower, and hunger hormones have produced enough additional eating (often through under-recording of bites and licks) to close much of what remains. The 500-calorie deficit may be functionally 100 to 200 calories, which is why the rate of weight loss slows and eventually plateaus.

None of this means CICO is "wrong." It means that treating the equation as static, when both sides are dynamic, produces predictions that diverge from reality on the timescales that matter for weight management. The accurate version of the principle is closer to: sustained changes in body weight require sustained net energy imbalance, but the body adjusts both sides of the equation in ways that resist sustained imbalance, particularly in the deficit direction.

Adaptive Thermogenesis

The term adaptive thermogenesis refers to a specific phenomenon: a reduction in metabolic rate that exceeds what changes in body composition would predict. It is the body's way of partially closing an imposed energy deficit by quietly downregulating its own expenditure. The phenomenon has been one of the most carefully documented in metabolic physiology, and one of the most consequential for understanding why diets so reliably fail to produce sustained results.

The defining paper in the modern literature is Rudolph Leibel and colleagues' 1995 study in the New England Journal of Medicine. Leibel, at Columbia University, had spent years studying the metabolic responses to imposed weight changes in tightly controlled clinical settings. In the 1995 paper, his group reported on lean and obese subjects who were maintained at a 10% reduction in body weight under metabolic-ward conditions. The finding that would shape the field was unambiguous: after a 10% weight loss, resting energy expenditure was approximately 15% lower than predicted by the change in body composition alone. The drop was not explained by smaller body size. It was an additional reduction, on top of what body size predicted. The body had adapted to conserve energy in a way the simple arithmetic of tissue loss could not account for.

The phenomenon goes in both directions. Subjects who were maintained at a 10% weight gain showed the opposite pattern: resting energy expenditure was higher than body composition predicted, by a similar magnitude. The body resisted both directions of weight change by adjusting expenditure to partially offset the imposed shift. Below baseline weight, it conserved energy; above baseline, it spent more.

Michael Rosenbaum, also at Columbia and Leibel's long-time collaborator, has spent the subsequent decades extending and refining this picture. His 2010 review in the American Journal of Clinical Nutrition synthesised the available evidence on how long adaptive thermogenesis persists, the mechanisms driving it, and its clinical implications. The summary was that the adaptation does not appear to dissipate with time. People maintained at reduced body weight for years continued to show the lower-than-predicted energy expenditure. The phenomenon was not a transient adjustment to the period of active dieting; it was a sustained recalibration.

The longest-term human evidence came in 2016, when Erin Fothergill and Kevin Hall's group at the NIDDK published a six-year follow-up of fourteen contestants from The Biggest Loser television competition. The original participants had lost extreme amounts of weight under intensive supervised diet and exercise — averaging 58 kg of weight loss across thirty weeks. Six years later, Fothergill's group brought them back for metabolic-ward measurement. Most had regained substantial weight, though all remained below their pre-competition baseline. Their resting metabolic rates were, on average, approximately 500 calories per day below what would be predicted by their current body composition. The metabolic adaptation persisted six years after the original weight loss, despite the regain. The contestants who had maintained the most weight loss tended to show the largest residual adaptations. Sustained restriction did not return the metabolism to baseline; it appeared to entrench the lowered expenditure.

The mechanisms driving adaptive thermogenesis are now reasonably well understood, though their relative contributions vary across individuals. Four major pathways stand out:

The first is reduced sympathetic nervous system (SNS) tone. Caloric restriction reduces sympathetic outflow to peripheral tissues, which lowers the energy cost of running those tissues. Heart rate falls, peripheral metabolism slows, and the body shifts toward a more energy-conserving state.

The second is a fall in triiodothyronine (T3), the more active form of thyroid hormone. T3 levels typically drop measurably during weight loss, even when TSH and total T4 remain within normal range. Lower T3 means lower cellular metabolic rate across most tissues. This is distinct from clinical hypothyroidism, and a standard thyroid panel will often miss it.

The third is the drop in leptin, which we encountered already in the appetite context. Beyond its hunger-signalling role, leptin has direct metabolic effects: low leptin reduces resting energy expenditure and biases peripheral metabolism toward fat storage. Rosenbaum demonstrated in a series of elegant studies that exogenous leptin administration to weight-reduced subjects partially reverses adaptive thermogenesis — restoring sympathetic tone, T3 levels, and resting energy expenditure toward pre-weight-loss values. The leptin drop is a major driver of the metabolic adaptation, not merely a parallel phenomenon.

The fourth is improved muscle efficiency. After weight loss, skeletal muscle requires fewer calories to perform a given amount of mechanical work. The mitochondria within muscle fibres appear to become more efficient at coupling fuel oxidation to ATP production, reducing the heat loss that would normally accompany activity. This means that not only does the person have less muscle mass burning fewer calories at rest, but each unit of activity also costs less energy than it did before the weight loss. The cumulative effect on energy expenditure is substantial.

How long adaptive thermogenesis lasts

The honest answer is that no one knows exactly. The Fothergill six-year follow-up is the longest published measurement, and the adaptation was still clearly present at that timepoint. Whether it eventually dissipates given enough years of weight stability at the reduced weight, or whether it is essentially permanent in the absence of intervention, has not been definitively answered. The available evidence leans toward the latter.

There is some indication that the adaptation can be partially mitigated by strategies that defend lean mass during weight loss — adequate protein intake, resistance training, slower rates of loss — and that adequate maintenance feeding once weight is stabilised may help. But the existence of a persistent gap between predicted and actual expenditure in people who have lost substantial weight is not seriously disputed at this point. Metabolic adaptation is one of the better-evidenced phenomena in obesity science. Whether it is reversible in any meaningful sense is the next frontier.

Metabolism and Weight Loss

Pulling the pieces together, the standard trajectory of BMR during a caloric deficit looks like this. In the first week or two, BMR falls slightly more than expected — partly due to depletion of glycogen stores (which carry water with them, producing rapid early scale-weight loss but minimal real fat loss), partly due to immediate hormonal adjustments. Through the first month, the drop continues. By month two, BMR has stabilised at a level lower than body size alone predicts. By month three, the cumulative drop in BMR plus the reduction in NEAT plus the increase in hormone-driven hunger has narrowed the deficit substantially. By month six, in most people, the loss has plateaued, and the calorie target that was producing 0.5 kg per week of loss in month one is now producing roughly zero net loss.

The plateau is not a failure of effort. It is the system finding its new equilibrium — a point at which expenditure has fallen and intake has crept up (often through unconscious behavioural drift) until the gap closes. To produce further loss from this point requires either reducing intake further (which deepens the metabolic adaptation), or increasing expenditure (which produces compensatory responses of its own).

NEAT suppression deserves particular attention because it is the most invisible component. James Levine, formerly at Mayo Clinic, has done the foundational work on NEAT and its responsiveness to caloric availability. In a 1999 study published in Science, Levine's group overfed sixteen non-obese volunteers by 1,000 calories per day for eight weeks and measured the components of energy expenditure carefully. The change in NEAT predicted resistance to fat gain better than any other variable. Subjects whose NEAT increased substantially in response to overfeeding (some by as much as 700 calories per day) gained the least weight; those whose NEAT increased little gained the most. The reverse pattern holds during caloric restriction: NEAT falls, often substantially, and people who would otherwise be moving spontaneously become slightly more sedentary in ways they do not notice.

The hormonal cascade during weight loss has its single most influential paper in Priya Sumithran and Joseph Proietto's 2011 NEJM study, conducted at the University of Melbourne. Sumithran's group followed fifty overweight or obese adults through a ten-week very-low-calorie diet, then measured appetite-regulating hormones at baseline, immediately after the diet, and at twelve months — a period over which most participants had regained substantial weight. Twelve months out, nine of the ten measured hormones remained significantly different from baseline. Ghrelin was elevated. Leptin, peptide YY, cholecystokinin, GIP, and amylin were all suppressed. The hunger driver was still pushing; the satiety signals were still attenuated. The hormonal environment had not returned to its pre-diet configuration.

Combined with adaptive thermogenesis and NEAT suppression, this hormonal pattern produces what feels, from the inside, like the dietary math suddenly stopping working. It is not a failure of arithmetic. The body has quietly renegotiated the terms on three fronts simultaneously. Why low-calorie diets backfire is, in mechanistic terms, this same story told from a different angle, and weight regain after dieting is its predictable continuation rather than a moral failure.

The weight-loss plateau seen on GLP-1 medications often catches patients off guard for the same reason. The pharmacology shifts the appetite environment, but it does not abolish adaptive thermogenesis. After substantial weight loss on semaglutide or tirzepatide, the same metabolic adaptations occur as with any other form of weight loss. The plateau, when it arrives, reflects the new equilibrium between the medication's effect on intake and the body's defensive reduction in expenditure. Some patients respond to the plateau by reducing intake further; some by increasing physical activity; some by increasing dose; and some accept the new weight as the level at which the system has settled. The choice depends on individual goals and tolerability, but the underlying physiology is the same as with any sustained weight loss. Not losing more weight on a GLP-1 medication is rarely a sign that the medication has stopped working; it is more often a sign that the metabolic adaptation has caught up.

The clinical implication is that keeping weight off is biologically harder than losing it. Loss happens during the period when the deficit is still functionally large. Maintenance requires holding the deficit closed at a steady state, against a system whose persistent setpoint pressure remains oriented toward the original weight. This is part of why long-term maintenance success rates without ongoing intervention are so low. The biology described in this section is, in obesity medicine, the explanation for those rates.

Metabolism Myths

The volume of folklore around metabolism is large, and several of the most circulated claims are worth addressing directly because they shape behavioural choices in ways that the underlying evidence does not support.

The first is the claim that eating frequently boosts metabolism. Grazing on six small meals a day to "keep the metabolic fire burning" has been tested in randomised trials repeatedly, and the result is consistent: meal frequency, at matched total calorie intake, has no meaningful effect on twenty-four-hour energy expenditure. Six small meals produce six small TEF responses; three larger meals produce three larger ones. The cumulative twenty-four-hour TEF is approximately the same. The persistence of the myth probably reflects the genuine modest benefits some people experience from regular eating in terms of hunger control, dressed up as a metabolic intervention.

The second is that cold exposure or brown fat activation can substantially raise BMR. The biology is real — brown adipose tissue dissipates energy as heat through UCP1, and cold exposure activates it — but the magnitude is small. Wouter van Marken Lichtenbelt at Maastricht has shown that prolonged daily cold exposure in adults with detectable brown fat raises expenditure by roughly 100 to 200 kcal per day at most, and requires hours of exposure at 17 to 19°C to achieve. A two-minute cold shower does not approach this. Marketed "metabolism boosters" as a class tend to claim effects larger than the underlying evidence supports.

The third is that some foods burn more calories than they contain — typically applied to celery, grapefruit, or various beverages. The thermic effect of food caps at about 30% even for high-protein meals, and is much lower for carbohydrate-dense foods. The mathematics simply do not produce a net negative-calorie food. The closest a food can come is when fibre and water content produce fullness disproportionate to calorie content, which is meaningful for appetite management but not a thermodynamic free lunch.

The fourth, and perhaps most clinically important, is that metabolism is permanently damaged by dieting. The phrase "metabolic damage" describes the experience of being unable to lose further weight despite low intake. The phenomenon — adaptive thermogenesis, NEAT suppression, hormonal adaptation — is real, but "damage" misframes it. Eric Trexler's 2014 review, drawing on the physique-sport literature, made the case that these adaptations are largely reversible with appropriate refeeding, weight stabilisation, and resistance training. People who restore maintenance intake and resume training over several months typically see substantial recovery of BMR toward predicted values. The recovery is not always complete, particularly in those who have done many cycles of severe restriction, but the framing of "damage" overstates how irreversible the adaptations are.

The fifth, and most consequential, is that slow metabolism is the main cause of obesity. The most decisive challenge came from Herman Pontzer at Duke and his collaborators, who measured TDEE in the Hadza, a hunter-gatherer population in Tanzania, using doubly-labeled water. The Hadza walk roughly 15,000 steps a day. Pontzer's 2012 paper in PLOS ONE reported a finding that surprised the field: TDEE in the Hadza was not substantially higher than in sedentary Westerners of similar body composition. His 2016 paper in Current Biology extended the model: above a moderate threshold of physical activity, total daily expenditure plateaus. The body adapts to high activity loads by reducing energy spent on other processes — immune function, reproductive physiology, stress response — partially absorbing the addition. Obesity rates differ dramatically between populations whose TDEE is similar. The obesity differences track intake patterns, not BMR. Slow metabolism is not the main driver of population-level obesity. The food environment is.

Metabolism and Aging

The conventional wisdom that "metabolism slows with age" — usually invoked to explain mid-life weight gain — turns out to be substantially overstated. The most authoritative challenge came from Herman Pontzer and a large international consortium in a 2021 paper in Science that pooled doubly-labeled water measurements from over 6,400 individuals across 29 countries, ranging in age from eight days to 95 years. The paper redrew the standard age-metabolism curve.

The findings overturned several assumptions. From birth to about age one, energy expenditure per kilogram of body weight is roughly 50% higher than predicted by adult scaling — infants are metabolically extraordinarily expensive per unit mass. From age one to about twenty, expenditure per kilogram declines gradually as the body matures. Between twenty and sixty, however, adjusted metabolic rate is essentially stable. There is no progressive slowdown across the adult decades. A forty-year-old's metabolism, corrected for body composition, runs at the same rate as a twenty-year-old's. After sixty, expenditure begins to decline by about 0.7% per year, accelerating somewhat in the eighties and beyond.

The implications for popular metabolism discourse are substantial. The widespread experience of weight gain in the thirties and forties cannot be primarily explained by metabolic decline, because there is no decline in those decades once body composition is controlled for. The apparent slowdown is largely accounted for by changes in body composition: progressive loss of lean mass, accumulation of fat mass, and reductions in spontaneous activity. The metabolism per kilogram of lean tissue stays the same. The person has less lean tissue, and so a lower absolute BMR.

Sarcopenia — the age-associated loss of skeletal muscle mass — is the mechanism most worth attention. Without resistance training and adequate protein, adults typically lose 3 to 8% of muscle mass per decade after age thirty, with the rate accelerating after sixty. Each kilogram of lost muscle removes roughly 13 to 15 kcal per day from BMR. Over twenty years, this can add up to a 100 to 200 kcal per day reduction in absolute BMR — not because metabolism per se has slowed, but because the lean tissue running it has shrunk. Reduced physical activity compounds the picture: less activity means less direct expenditure, less NEAT, and accelerated muscle loss in a self-reinforcing pattern. Add the typical slow upward drift in caloric intake, and the conditions for slow weight gain are set without any change in metabolic rate per kilogram of lean tissue.

The practical implication is that "blame aging" overstates the case substantially before age sixty. Most of the apparent slowdown people experience in middle age is a body composition story, not a metabolism-per-cell story. The interventions that work for it — resistance training, protein adequacy, sleep, and managing the food environment — are the same as for younger adults. After sixty, the picture shifts: the 0.7% per year decline becomes meaningful, sarcopenia tends to accelerate, and clinical considerations around frailty, protein adequacy, and fall prevention move to the foreground. Older adults often need more, not less, protein per kilogram of body weight to maintain lean mass.

Metabolism and Appetite

Metabolism and appetite are not separate systems. They are tightly coupled, with shared hormonal signals and reciprocal feedback that means an intervention on one almost always affects the other. Trying to "fix metabolism" without addressing appetite, or vice versa, ignores how the system was assembled.

The bidirectional link is most clearly demonstrated in the work of Éric Doucet and colleagues at the University of Ottawa. In a series of studies beginning around 2000, Doucet's group showed that the magnitude of adaptive thermogenesis during weight loss correlates with the magnitude of the increase in hunger. People whose metabolic rates drop most after dieting tend also to be those whose appetite rises most. The relationship is mechanistically sensible: both responses are driven by the same underlying signals — leptin falling, ghrelin rising, T3 dropping, sympathetic tone changing. The metabolic adaptation and the appetite adaptation are two faces of the same defensive response to perceived energy shortage. They co-vary because they share a hormonal driver.

Leptin's dual role sits at the centre of this coupling. Leptin signals long-term energy adequacy. When fat mass falls and leptin drops, the hypothalamus responds with coordinated changes on both sides of the energy ledger: hunger rises (to drive intake up) and resting expenditure falls (to conserve what is already stored). The drop in leptin is the same signal triggering both responses. This is why interventions that prevent the leptin drop — for example, leptin administration in weight-reduced subjects, as in Rosenbaum's experimental work — reduce both the adaptive thermogenesis and the appetite increase simultaneously. The biology is one system, not two.

The GLP-1 pathway intersects both arms as well. The primary clinical effect of GLP-1 receptor agonists is appetite reduction through central pathways, as discussed at length in the appetite regulation pillar and in the foundational piece on what GLP-1 is. But GLP-1 receptors are also expressed in brown adipose tissue, and some studies suggest GLP-1 agonism may modestly activate brown fat thermogenesis in some individuals. The magnitude of the metabolic effect appears small relative to the appetite effect — most of the weight loss on these medications is driven by reduced intake, not increased expenditure. The way semaglutide produces weight loss reflects the cumulative impact of changing the hormonal environment that regulates intake, with metabolic preservation as a secondary benefit rather than the primary mechanism.

The clinical implication is that "fix appetite" and "fix metabolism" are not separable interventions. Pharmacology that addresses one almost always touches the other. Lifestyle interventions that affect one almost always affect the other. The coupling means that asking which one to address first is the wrong question; the right question is which lever produces the largest favourable change in the coupled system. For most people with significant weight dysregulation, that lever is usually the appetite side, both because appetite-targeted pharmacology now exists and works well, and because intake is the larger of the two variables in absolute caloric terms. Hunger hormones and satiety hormones are the practical targets through which the coupled system is most often modulated.

Can Metabolism Be Increased?

This is the question people most often want the answer to, and the honest answer is more modest than the marketing literature suggests. Several real interventions exist; none of them produces a dramatic standalone effect, and most of the dramatic-sounding claims in popular sources do not survive scrutiny.

Caffeine raises resting energy expenditure modestly. A typical dose of 100 to 200 mg increases metabolic rate by roughly 3 to 11% for several hours — translating to perhaps 30 to 100 kcal per day of additional expenditure in a regular caffeine consumer, with substantial tolerance development over weeks. The effect is real but not transformative, and tolerance limits sustained gains.

Capsaicin (the active compound in chilli peppers) produces a similar order of effect: roughly 10 to 50 kcal per day of additional expenditure through a combination of brown fat activation and increased sympathetic tone. Sustained effects are again limited by tolerance and by the relatively small magnitude of the underlying mechanism.

Cold exposure, as discussed in the myths section, can produce 100 to 200 kcal per day of additional expenditure in adults with detectable brown fat, but only with sustained exposure (hours per day at temperatures cold enough to be uncomfortable). Brief cold exposure produces much smaller effects.

The interventions with the largest sustained effects on metabolic rate are not "metabolism boosters" in the supplement sense. They are resistance training and the maintenance of lean mass. Adding meaningful skeletal muscle through structured resistance training raises BMR by roughly 13 to 15 kcal per kilogram of additional muscle per day — modest in absolute terms, but unlike the supplement effects, the gain is sustained for as long as the muscle is maintained. More importantly, resistance training protects against the muscle loss that otherwise quietly reduces BMR through adulthood. The combined effect of building and maintaining lean mass through a sustained training programme can amount to several hundred kcal per day of preserved expenditure over decades. Strength training during weight loss is the single most effective metabolic intervention available to most people, and the one most consistently underemphasised in popular advice.

Stuart Phillips at McMaster University has done much of the foundational work on protein intake and muscle maintenance. His 2016 review synthesised the evidence on protein requirements for preserving lean mass during caloric restriction. The current consensus is that protein intake of approximately 1.6 to 2.4 g per kilogram of body weight per day — substantially higher than the standard RDA of 0.8 g/kg — supports muscle maintenance during weight loss and the metabolic capacity that comes with it. Protein also produces the highest TEF of any macronutrient. Protein targets during weight loss are particularly important for those on appetite-suppressing medications, where total intake is reduced and the risk of inadequate protein is correspondingly elevated. A practical high-protein meal plan structured around these targets makes the difference between losing primarily fat and losing meaningful amounts of muscle. Muscle preservation during loss is one of the most consequential modifiable variables in metabolic outcomes.

NEAT increases, drawing on Levine's work at Mayo Clinic, are an underappreciated lever. Standing desks, frequent short walks, deliberate increases in spontaneous movement, and other small-bore activity additions can amount to several hundred kcal per day in aggregate for people who consciously cultivate them. The effect is real and sustainable. The limit is that NEAT additions tend to be partly compensated by reductions elsewhere — Pontzer's constrained-TDEE model in action — so the net gain is smaller than the gross addition. Even so, NEAT is one of the few channels through which sustained increases in energy expenditure can be cultivated without entering the diminishing-returns territory of formal exercise.

The case of GLP-1 medications deserves separate treatment because they do not fit the "metabolism booster" frame at all. GLP-1 receptor agonists do not raise baseline metabolic rate. What they do is reduce appetite enough to produce sustained weight loss, and they appear to preserve resting metabolic rate during loss better than would otherwise be expected. The question of whether they preserve lean mass adequately during weight loss has been a particular focus of recent research. Olof Linge and colleagues at AMRA Medical in Sweden have used quantitative MRI to characterise body composition changes in GLP-1-treated patients, and the picture is mixed: muscle loss does occur, often at proportions similar to what is seen with other forms of substantial weight loss, but the muscle quality and distribution effects are still being clarified. The interpretation has evolved over the past two years — early concerns that GLP-1 medications produced disproportionate muscle loss have been tempered by evidence that, when patients engage in adequate protein intake and resistance training, lean mass preservation is comparable to other modalities. Exercise during GLP-1 treatment is, on the evidence, the single most important behavioural intervention for protecting metabolic capacity through the period of weight loss.

Evidence-Based Takeaways

Several conclusions emerge from the literature that have practical implications for how people think about their own metabolism.

The first is that BMR is rarely the main driver of weight gain or loss. Real differences in BMR between similar individuals exist, typically in the range of 150 to 300 kcal per day at the extremes of normal variation — meaningful but not large enough to account for dramatic weight differences. Most of what people experience as "slow metabolism" traces to factors they cannot directly see: changes in NEAT, under-recording of food intake, sleep, body composition shifts, and adaptive responses to repeated dieting cycles. BMR is one input to weight regulation, not the dominant one for most people. Underlying metabolic disorders like insulin resistance and significant thyroid dysfunction do affect the picture and warrant clinical evaluation when symptoms suggest them.

The second is that protein intake and resistance training are the most reliable defenses against age-related metabolic decline. The mechanism is not that protein or training "speeds up metabolism" — it is that both protect lean mass, which is the dominant determinant of absolute BMR. Sustained adequate protein (1.6 to 2.4 g/kg/day) and sustained resistance training (two to four sessions per week, progressively loaded) is the single most evidence-supported intervention for preserving metabolic capacity over decades. It has the additional benefit of preserving function — strength, balance, mobility — that becomes increasingly important after sixty.

The third is that sleep is metabolic infrastructure, not metabolic decoration. The work of Eve Van Cauter at Chicago, Karine Spiegel, and others has documented that two nights of restricted sleep (four hours per night) produces measurable shifts in appetite hormones — ghrelin up, leptin down, hunger up — in healthy young adults. Beyond appetite, chronic sleep restriction is associated with reduced insulin sensitivity, altered glucose tolerance, and lower spontaneous physical activity. The cumulative metabolic cost of chronic sleep deprivation is substantial. Sleep deprivation produces hormonal changes that bias the metabolic-appetite system unfavourably in ways no dietary intervention can compensate for. For people whose weight is not responding to other measures, sleep is often the missing variable.

The fourth is that drugs that meaningfully "boost metabolism" do not really exist. Historical attempts to engineer them — thyroid hormone administration, dinitrophenol — produced narrow benefits with serious risks, and modern stimulant-based metabolism boosters produce modest effects with substantial cardiovascular risk. The drugs that do work for weight regulation work primarily through appetite, not expenditure. GLP-1 receptor agonists, GIP/GLP-1 co-agonists, and the emerging triple agonists all act on appetite circuits. Appetite is the pharmacologically tractable side of the equation, and the magnitude of intake variation is larger than the magnitude of expenditure variation in most people.

The fifth is that pharmacology, when it is appropriate, is mechanism-matched. GLP-1 medications work because they address the hormonal environment that drives appetite — they are matched to the biology that produces sustained overeating in many people with obesity. They do not work primarily by speeding up metabolism, and patients who expect that effect are often initially surprised by the absence of a stimulant-like sensation. What they do is reduce the persistent biological pressure to eat, which produces a sustainable deficit that the body's defensive responses cannot fully close. The treatment matches the mechanism. This is what distinguishes contemporary obesity medicine from the cycles of restrictive dieting and behavioural exhortation that defined earlier decades. The framing that someone has tried every diet and nothing works often reflects a mismatch between the intervention attempted and the underlying biology — appetite was the problem, and behavioural restriction was being asked to do something the biology was unsuited to deliver.

None of this means metabolism is unimportant. It means the popular framing of metabolism as a single dial that can be turned up or down by clever interventions distorts a more interesting underlying biology. Metabolism is a regulated system that responds to inputs in patterned ways. Understanding the patterns — that BMR is dominated by lean mass, that adaptive thermogenesis defends weight from below, that the body adjusts both intake and expenditure to resist sustained deficit, that the major adaptations are reversible to varying degrees with the right inputs — is the foundation for working with the system rather than against it. The companion topics in the appetite regulation pillar and the hunger and satiety guide complete the picture: appetite, satiety, and metabolism are three views of one coupled regulatory system.