Chapter 5 — Metabolism: BMR, TDEE, the Thermic Effect of Food, and Why Your Metabolism Is Not "Broken"

The Hook: The most common sentence in my clinic

Renata Almeida is twenty-nine, works in insurance claims, and said it in the first ninety seconds.

"I think my metabolism is broken."

I hear that sentence, or a close variant of it, more than any other. Some weeks I hear it three or four times. It arrives in a particular tone — not dramatic, usually. Flat. The tone of someone reporting a fact they've made peace with.

And it is nearly always said by someone who has been told, implicitly or explicitly, that their body is the problem and their effort is insufficient. The "broken metabolism" theory is, among other things, a more merciful explanation than the one they've been handed everywhere else. I understand completely why people reach for it.

So I asked Renata what made her think so.

She'd used an online calculator that told her she needed 1,850 calories a day. She'd eaten "about 1,500" for four months. She'd lost seven pounds in the first three weeks and nothing since. Her friend, who is taller and eats "way more," is thinner. Her mother has always said the women in their family have slow metabolisms.

Four pieces of evidence. Let me tell you what each one is actually worth, because the answers are more interesting than "your metabolism is fine."

The calculator's 1,850 has an error bar of roughly ±10% — so somewhere between about 1,665 and 2,035, and that's before the activity multiplier, which is the least reliable part of the whole calculation. Her true requirement might be 1,700. It might be 2,100.

The "about 1,500" is an estimate produced by the same cognitive machinery that produced Theo's 2,000 when the answer was 3,160. Not a lie. Just an estimate, and estimates in this domain run low, systematically, in everyone.

The seven pounds then nothing is the single most predictable pattern in weight management, and §4.9 already explained most of it: the early loss was substantially glycogen and water, and then the actual fat-loss rate — slow, small, easily buried in daily fluctuation — became the only thing left.

The taller friend who eats more almost certainly does need more, because she's bigger. Body size is the dominant determinant of metabolic rate, and it works in exactly the direction that feels unfair.

And her mother's theory is the interesting one, because it's not entirely wrong. There is genuine between-person variation in metabolic rate that we cannot explain by size, age, sex, or body composition. It's real. It's also — and this is the number that matters — roughly ±10%, which for Renata is about 180 calories a day. Not nothing. Also not the explanation for anything she's experiencing.

Here's what I told her, and it's the thesis of this chapter:

Your metabolism is not broken. It is responsive — and it is responding to what you're doing, which is a completely different and far more useful fact.

🏃 Fast Track: §5.2 (what your metabolism is actually spent on — this will surprise you), §5.3 (calculate your own), and §5.6 (metabolic adaptation, honestly) are the core.

🔬 Deep Dive: §5.5 (what genuinely varies between people) and §5.6's treatment of the weight-loss competition follow-up study are where students and clinicians should spend time. §5.8 is the myth-clearing section and is unusually dense with 🔬 verdicts.

5.1 The vocabulary, since everyone uses it wrong

Four terms, used interchangeably in popular writing and meaning different things.

In practice, RMR is what calculators estimate, and TDEE is the number people actually want. The rest of this chapter uses RMR.

One more piece of vocabulary that matters: "metabolism," in casual use, almost always means TDEE. When someone says their metabolism is slow, they mean they burn fewer calories per day than they think they should. That's a statement about TDEE — and as we'll see, the component of TDEE most likely to actually differ between two people is not their RMR at all.

5.2 What your metabolism is actually spent on

Here's the table that changes how most people think about this, and almost nobody has seen it.

Your organs don't cost the same to run per kilogram. Not remotely.

(Figures approximate and widely cited from organ-level metabolic studies; they vary with body composition, age, and method.)

Read that carefully, because there are two genuinely startling facts in it.

Fact one: your brain, liver, heart and kidneys are about 5–6% of your body mass and roughly 55–60% of your resting energy expenditure. A handful of organs, most of which you never think about, are doing the overwhelming majority of the burning. Your liver alone costs more per day than all forty percent of you that is muscle.

Fact two — and this one has consequences — skeletal muscle costs about 13 kcal per kilogram per day. Not 100. Not 50. Thirteen.

🔬 Claim → Evidence → Verdict

The claim: "Every pound of muscle you gain burns an extra 50 calories a day. Build muscle and you'll boost your metabolism dramatically."

Where it comes from: A real principle — muscle is metabolically more active than fat, and resistance training does increase RMR — stretched by roughly a factor of ten somewhere in the fitness literature, then repeated for thirty years.

What the evidence actually shows: Skeletal muscle at rest costs on the order of 13 kcal/kg/day — that's roughly 6 kcal per pound per day. Gaining 10 lb of muscle, which is a great deal of muscle and would take most people a year or more of serious training, raises RMR by something like 60 kcal/day. Real, and about the calories in a small apple. The "50 kcal per pound" figure would imply that 10 lb of muscle adds 500 kcal/day, which would be a metabolic transformation and is not what happens.

📉 Evidence quality: Organ-level metabolic measurements are well established. The 50 kcal figure has no identifiable source in the primary literature.

Verdict: 🟠 Probably false — the direction is right, the magnitude is off by roughly an order of magnitude.

And here is why you should lift anyway, because this verdict is routinely misused: resistance training preserves lean mass during weight loss (which is most of what determines whether you end up lighter-and-stronger or lighter-and-weaker), improves insulin sensitivity, protects bone, substantially improves function and independence in later life, and burns energy during the training and for some hours afterward. None of those reasons is the RMR increase, and the fact that the RMR argument was oversold doesn't touch any of them.

🔍 Why this works. Why is muscle so cheap at rest? Because at rest it is doing almost nothing. Muscle is a tissue with an enormous dynamic range — its energy use can rise more than twenty-fold during hard exercise — but its idle cost is low, which is exactly what you'd want from an engine you only sometimes need. Your liver, by contrast, is running continuously: synthesizing proteins, handling glucose, processing everything absorbed in §3.5, detoxifying, making bile. It never idles. The tissues that cost the most are the ones that never stop, not the ones that can do the most.

📊 Diagram (described). Picture a human silhouette, and imagine shading it not by size but by energy cost, so that the parts doing the most burning are drawn largest.

The body you'd draw is grotesque and instructive. The head would be enormous — the brain is 2% of mass and about a fifth of resting expenditure, so it dominates the picture. The liver would be nearly as large, a great heavy shape filling most of the torso. The heart and kidneys, which are physically tiny, would each be roughly the size of a limb. And the arms and legs — forty percent of your actual mass — would shrink to thin lines, because at rest muscle costs almost nothing. The adipose tissue would be barely visible.

That grotesque figure is your metabolism. It is mostly a brain and a liver, kept alive by a heart and kidneys, being carried around by some very cheap limbs. And it explains at a glance why body composition changes move RMR so much less than fitness culture claims — you can add or remove a great deal from the thin lines without changing the big shapes at all.

🔄 Check your understanding. Two adults have identical height, weight, age and sex. One has substantially more muscle and less fat than the other. How different are their resting metabolic rates likely to be — dramatically, moderately, or slightly?

Answer

Moderately at most — probably somewhere in the range of 50–150 kcal/day. Muscle costs ~13 kcal/kg/day and fat costs ~4.5, so swapping 10 kg of fat for 10 kg of muscle changes RMR by roughly 85 kcal/day. Real, worth having, and far less than fitness culture implies.

The reason body composition matters less than expected for RMR is Fact One: the organs doing most of the burning are roughly the same in both people. Body size dominates RMR; body composition modifies it.

5.3 Calculating your own: three worked examples

The most widely used equation for estimating RMR is Mifflin-St Jeor. It's not perfect. It's the best of the simple ones for general populations.

Men:    RMR = (10 × weight kg) + (6.25 × height cm) − (5 × age) + 5
Women:  RMR = (10 × weight kg) + (6.25 × height cm) − (5 × age) − 161

Let's run three real cases from this book, step by step.

Theo — 34, male, 178 cm, 96 kg

10 × 96          =   960
6.25 × 178       = + 1,112.5
5 × 34           = −   170
constant (male)  = +     5
                   ─────────
RMR              =  1,907.5  ≈  1,908 kcal/day

Devi — 21, female, 168 cm, 55 kg

10 × 55            =   550
6.25 × 168         = + 1,050
5 × 21             = −   105
constant (female)  = −   161
                     ─────────
RMR                =  1,334 kcal/day

Walt — 68, male, 175 cm, 88 kg

10 × 88          =   880
6.25 × 175       = + 1,093.75
5 × 68           = −   340
constant (male)  = +     5
                   ─────────
RMR              =  1,638.75  ≈  1,639 kcal/day

Now the part every calculator omits.

⚠️ Every one of those numbers is ±10%, at best.

• Theo: 1,908 → realistically 1,717–2,099
• Devi: 1,334 → realistically 1,201–1,467
• Walt: 1,639 → realistically 1,475–1,803

That's a spread of roughly 380, 270, and 330 calories respectively. The uncertainty in the starting number is comparable to the deficit most people are trying to create.

The equations also perform worse at the extremes — in people with very high or very low body fat, in very muscular people, in the very old, and in children. Mifflin-St Jeor was derived from a general adult population; the further you are from that, the worse it does. Athletes with high lean mass are frequently underestimated; the Katch-McArdle equation, which uses lean body mass instead of total weight, does better if you have an accurate body composition measurement, which most people don't.

5.4 From RMR to TDEE — and the weakest link

Multiply RMR by an activity factor:

Theo: 1,908 × 1.4 (desk job, two 30-minute walks a week) = 2,671 ≈ 2,670 kcal/day.

Now compare that to his measured intake of 3,160, and his surplus of about +490 kcal/day falls straight out — which is where Chapter 4's arithmetic came from.

But be honest about what just happened. We took a number with ±10% uncertainty and multiplied it by a factor selected from a dropdown menu, on the basis of a subjective self-assessment, in a category system where the gap between adjacent options is 10–15%.

💡 Aha moment. The activity factor is the least reliable part of the entire calculation, and it is the part people agonize over least. Two honest people with identical bodies and identical lives can select 1.375 and 1.55 and produce estimates 130 calories apart. And the biggest source of real variation between them — NEAT (§4.5) — is unconscious, which means neither of them can report it accurately, because they don't know how much they fidget.

So what is the number for?

It's a starting hypothesis, not a measurement. Here is the honest protocol, and it is the only one that actually works:

1. Calculate your estimate.
2. Eat at approximately that level for two to three weeks, tracking however loosely you like.
3. Watch your weekly average weight (§4.9), not daily.
4. Adjust the number based on what actually happened. If weight is stable, your TDEE is approximately your intake — and that observation is worth more than any equation.

Your body is a better calorimeter than the formula. The formula's job is to stop you starting at 1,200 or 3,500. It is not to tell you the truth.

5.5 What actually varies between people

Renata's mother said the women in their family have slow metabolisms. Let's take that seriously and see what's real.

On age

The folklore says metabolism collapses in your thirties. The better evidence suggests something more specific and less dramatic: energy expenditure adjusted for body composition is relatively stable through most of adult life, with meaningful decline arriving considerably later than commonly believed — into the sixties and beyond.

What actually happens to most people in their thirties and forties is a slow loss of muscle, an increase in fat mass, and a substantial decline in physical activity — a smaller engine and less driving, not a broken engine. That distinction matters, because one of those is largely modifiable and one isn't.

On genetics

Between-person variation in RMR unexplained by size, composition, age, and sex is genuinely around ±10%. For Renata that's roughly ±180 kcal/day.

Take that seriously in both directions. It means two people the same size can genuinely have RMRs differing by 300+ calories — which is a real disadvantage and it is unfair. It also means it is not the explanation for a 1,000-calorie discrepancy, and telling someone their genetics explain everything is as dishonest as telling them genetics explain nothing.

⚠️ When to see a professional. Genuine metabolic conditions exist and should be ruled out rather than assumed. Hypothyroidism is the most common — it lowers RMR and causes fatigue, cold intolerance, constipation, dry skin, and hair changes; it's diagnosed with a straightforward blood test and it's treatable. PCOS affects insulin sensitivity and body composition in ways relevant here. Cushing's syndrome is rare but real. A number of medications — corticosteroids, some antipsychotics, some antidepressants, some hormonal treatments — cause genuine weight gain.

If you have symptoms, get tested. "My metabolism is broken" deserves a blood test before it deserves a rebuttal, and a clinician who skips that step is doing it wrong.

5.6 🚪 Metabolic adaptation: real, and smaller than you think

Now the threshold concept, and it requires holding two facts that most people treat as incompatible.

🚪 Threshold concept.

Metabolic adaptation is real, measurable, and persistent. It is also modest — measured in hundreds of calories, not thousands. Your metabolism is not broken; it is responsive.

The real part

When you lose weight, your energy expenditure falls for two reasons.

The expected part: a smaller body costs less to run. Lose 10 kg and your RMR falls by roughly 100–150 kcal/day just from being smaller. Entirely predictable, and not adaptation.

The adaptive part: expenditure falls more than body size alone predicts. This is adaptive thermogenesis, and it comprises a modest reduction in RMR beyond prediction, a larger unconscious reduction in NEAT, and improved muscular efficiency (you use slightly less energy to do the same movement).

Estimates of its size vary — this is a genuinely contested area — but the range typically discussed is on the order of tens to a few hundred kcal/day.

And the hormonal changes persist. Ghrelin stays elevated and leptin stays suppressed after weight loss, in some studies for a year or more. That is not a metabolic rate change; it's an appetite change, and it may matter more than the metabolic one.

The famous study, handled honestly

The most widely cited evidence here is a follow-up study of contestants from a televised weight-loss competition, examining them six years after the show. It found substantial persistent metabolic adaptation — resting metabolic rates well below what body composition predicted, still present years later — alongside substantial weight regain.

It's a genuinely important study and it deserves to be read carefully, in both directions.

What it establishes: that severe, rapid weight loss achieved through extreme caloric restriction and enormous exercise volumes can be followed by persistent metabolic adaptation and regain. That's real, it's important, and it should change how aggressive weight-loss interventions are designed.

What it does not establish: that this is what happens to everyone who loses weight. The sample was small. There was no control group. The intervention was extraordinary — far outside anything a person would do voluntarily at home, involving daily exercise volumes that most people could not sustain for a week. And people who maintain substantial weight loss through more moderate means — the National Weight Control Registry population, for instance — do not consistently show adaptation of that magnitude.

📉 Evidence quality: A small, uncontrolled, longitudinal follow-up of an extreme intervention. Genuinely informative about the extreme; weak as a general model.

The honest summary: metabolic adaptation is real and scales with the severity of the deficit. Aggressive deficits produce more of it. This is an argument for moderate, sustainable deficits, not an argument that weight loss is futile.

🧩 Productive struggle. Four minutes on this before reading on.

Someone loses 15 kg. Their RMR is now 250 kcal/day lower than it was before.

How much of that 250 is "adaptation," and how much is just being smaller? And what would you need to know to separate them?

What I'd say

Most of it is just being smaller. Fifteen kilograms less body mass means less tissue to maintain — expect roughly 150–200 kcal/day of the drop from size alone. The genuinely adaptive component is the remainder — perhaps 50–100 kcal/day — and even that estimate is uncertain.

To separate them you'd need to predict what their RMR should be at their new size and composition, and compare it to what it actually is. That requires measuring body composition (not just weight — because losing 15 kg of mostly-fat and losing 15 kg with a lot of lean tissue predict different RMRs) and measuring RMR by indirect calorimetry rather than estimating it. The gap between predicted and measured is the adaptation.

This is exactly what the weight-loss competition follow-up did, and it's why that study matters despite its limitations — it did the measurement rather than the estimate.

Notice the practical implication: a person whose expenditure fell 250 kcal has not had their metabolism broken. They've had it reduced, mostly by physics, partly by adaptation — and the adaptive part is roughly the size of a slice of toast.

🔄 Check your understanding. A coach tells a client: "Your metabolism has adapted, so we need to do a two-week 'reverse diet' at maintenance to repair it before we can lose more fat." Which parts of that are defensible and which aren't?

Answer

Defensible: adaptation is real; a period at maintenance genuinely does allow appetite hormones to normalize, NEAT to recover, and — importantly — gives the person a psychological break from restriction, which is often the actual benefit. Deficits are more effective when they're intermittent than when they're endless, largely for adherence reasons.

Not defensible: the word "repair." Nothing is damaged. Expenditure fell partly because the body is smaller (physics, permanent, and appropriate) and partly through adaptation (modest, and it reverses as intake and body mass recover). Framing it as repair implies a broken system and sells a protocol; framing it as recovery of appetite regulation and activity describes what's actually happening.

Also worth noticing: the claim is unfalsifiable as stated. If the client loses weight afterward, the reverse diet worked; if they don't, they need more of it. Chapter 2's tools apply to coaching claims exactly as they apply to studies.

5.7 "Starvation mode," and what's actually true

The folk version: eat too little and your body "goes into starvation mode," holds onto fat, and you stop losing weight.

What's false: you cannot stop losing weight in a genuine, substantial energy deficit. Sustained severe restriction produces weight loss — this has been observed, repeatedly and grimly, in famine, in illness, and in the Minnesota Starvation Experiment. Nobody in a real deficit maintains their weight.

What's true, and worth knowing:

• Expenditure does fall — RMR modestly, NEAT substantially, and the larger the deficit the larger the fall. So a severe deficit delivers proportionally less than a moderate one.
• Hunger becomes physiologically overwhelming. The Minnesota participants — healthy young men with no prior eating problems — developed food obsession, ritualized eating, binge episodes on refeeding, and lasting psychological effects. Restriction produced disordered eating in men who had none.
• Lean tissue loss accelerates at severe deficits, particularly without adequate protein and resistance training. You get lighter and weaker.
• Adherence collapses. This is the real mechanism behind most "plateaus" at very low intakes: people eat more than they report, then more than they intend, then abandon it — a sequence driven by physiology, not character.

So "starvation mode" is a garbled version of something real. The garbling matters, though, because it leads people to the wrong conclusion. The correct conclusion isn't "eat more to lose weight." It's "a moderate deficit outperforms a severe one on every axis that matters" — retention of lean mass, sustainability, psychological cost, and total weight lost at twelve months.

5.8 Can you "boost" your metabolism?

The short answer is: a little, temporarily, in ways too small to matter for body weight. Here are the claims you'll actually encounter.

🔬 Claim → Evidence → Verdict

The claim: "Eating six small meals a day keeps your metabolism stoked. Eating less frequently slows it down."

Where it comes from: The thermic effect of food is real (§4.5) — eating does raise expenditure. It seems to follow that eating more often means more raising.

What the evidence actually shows: TEF is proportional to how much you eat, not how often. Three 800-kcal meals and six 400-kcal meals produce essentially the same total thermic effect, because it's the same 2,400 kcal being processed. Controlled studies comparing meal frequencies at matched total intake generally find no meaningful difference in total energy expenditure or fat loss. Meal frequency may matter for individual appetite control, adherence, and training schedules — in both directions, since some people eat less on fewer meals and some on more — but that is a behavioral effect, not a metabolic one.

📉 Evidence quality: Multiple controlled trials, consistent null.

Verdict: ❌ Not supported as a metabolic claim. Eat on whatever schedule suits your appetite and your life.

🔬 Claim → Evidence → Verdict

The claim: "Green tea / capsaicin / caffeine / cold showers boost your metabolism and help you burn fat."

Where it comes from: Genuine measured effects. Caffeine really does raise energy expenditure acutely. Green tea catechins and capsaicin have measurable thermogenic effects in controlled studies. Cold exposure activates brown adipose tissue, which is genuinely present in adult humans — a finding that surprised the field when confirmed by imaging — and which does generate heat.

What the evidence actually shows: Every one of these is real and every one is small. The measured effects are typically on the order of tens of kilocalories per day — commonly quoted in the range of roughly 50–100 kcal/day for caffeine at meaningful doses, less for the others. Tolerance develops to some of them. And in trials measuring actual weight outcomes rather than short-term expenditure, effects on body weight are small at best and frequently null.

For scale: the entire measurable thermogenic effect of a green tea supplement is comfortably smaller than the uncertainty in your own RMR estimate (§5.3), and smaller than one biscuit.

📉 Evidence quality: Good short-term mechanistic and expenditure data; weak and inconsistent long-term weight outcome data.

Verdict: 🟡 Unclear / it depends — the effects are real and the practical significance is near zero. Drink coffee and eat chilli because you enjoy them.

🧾 Cost check. A "metabolism support" supplement stack — green tea extract, capsaicin, L-carnitine, chromium — typically runs $35–$70 a month, or $420–$840 a year, to purchase an effect of perhaps 50 kcal/day, if it works as well as the best short-term studies suggest. That's roughly $10 per kilocalorie per day, per year. Also note: high-dose green tea extract has been associated with rare cases of liver injury, which is a genuinely poor trade for 50 calories.

🔬 Claim → Evidence → Verdict

The claim: "Skipping breakfast slows your metabolism."

Where it comes from: The same observational data as Chapter 2's breakfast verdict, plus a conflation of TEF with metabolic rate.

What the evidence actually shows: Total daily TEF depends on total daily intake, not on when it starts. Controlled studies do not find that skipping breakfast reduces total energy expenditure in any meaningful way. Chapter 2 covered the associated weight claims; this is the metabolic version of the same error.

Verdict: ❌ Not supported. Eat breakfast if you're hungry in the morning. That's the whole decision rule.

🔄 Check your understanding. A supplement is advertised as increasing metabolic rate by "up to 8%." Before checking anything else, estimate what that would mean in calories for a person with an RMR of 1,600 — and then say whether the claim, even if entirely true, would matter.

Answer

8% of 1,600 is 128 kcal/day — and note that "up to" means this is the ceiling, typically drawn from the best result in the best short-term study, often at a dose higher than the product contains. A realistic average effect might be half that or less, and tolerance may reduce it further.

Does 128 kcal/day matter? Compare it to the ±10% error bar on the RMR estimate itself: ±160 kcal. The claimed effect is smaller than the uncertainty in the number it's a percentage of. It's also smaller than the day-to-day variation in NEAT, smaller than the error in one estimated dinner portion, and smaller than one beer.

This calculation takes fifteen seconds and it defeats essentially every metabolic supplement claim you will ever encounter. Convert the percentage to calories, then compare it to something real.

What does meaningfully change your TDEE, in descending order of size:

1. Body size. Being bigger costs more. (Not a recommendation — an explanation.)
2. NEAT. The single largest modifiable component. Walking, standing, moving. This is why "get more steps" outperforms almost every metabolic intervention ever marketed.
3. Exercise, both during and — modestly — after.
4. Lean mass, slowly, at ~13 kcal/kg/day.
5. Protein intake, via TEF, worth a few tens of kcal/day at realistic intakes.

Notice that the top two are free and the marketed ones aren't on the list.

5.9 What to actually do with your number

Four uses, and one non-use.

Use 1 — as a starting hypothesis. Calculate it, eat there for two to three weeks, watch the weekly average, adjust. The formula's job is to prevent absurd starting points, not to be right.

Use 2 — as a sanity check on a plan. If a diet recommends 1,200 kcal/day for a 96 kg man with an RMR of 1,908, you now know that's a deficit of over 1,400 calories against TDEE — deep enough to predict NEAT collapse, overwhelming hunger, lean tissue loss, and abandonment. You can identify a bad plan by arithmetic, before you waste four months on it.

Use 3 — as a floor. Very few adults should be eating below their RMR for extended periods without supervision, and Devi's case in Chapter 4 shows what happens when intake is too low relative to demand.

Use 4 — as an explanation, offered gently. When Renata's friend eats more and stays thinner, the answer is usually size, NEAT, and ±10% of genuine unexplained variation. That's a real answer, and it's kinder and more accurate than either "you're eating more than you think" or "your metabolism is broken."

The non-use: do not treat it as a measurement. It is an estimate with an error bar of hundreds of calories, multiplied by a dropdown selection. Apps that display it as a whole number, and then track your day against it to the calorie, are performing precision theatre.

🍽️ On your plate. Three things to actually do with this chapter, in order of how much they'll change your life. (1) Stop treating your calculated number as a fact. Write the range next to it, permanently, and let the two-to-three-week trial tell you the truth. (2) If you want to change your TDEE, walk more. NEAT is the largest modifiable component, it's free, it doesn't require a gym, and it dwarfs every thermogenic supplement ever sold. (3) If your deficit plan puts you below your RMR, redesign it. Not because of "starvation mode," but because §5.7 predicts what will actually happen: lean tissue loss, unmanageable hunger, NEAT collapse, and abandonment at around week eleven.

5.10 How it's measured properly — and what a gym metabolic test is worth

Everything above has been estimation. It's worth knowing what actual measurement looks like, partly because you may be offered it and partly because the method previews the next chapter.

Indirect calorimetry

You don't measure energy expenditure directly — you'd need to put someone in a sealed chamber and measure the heat they give off, which exists (whole-room calorimeters) and costs a fortune.

Instead, indirect calorimetry measures what you breathe. A mask or hood collects your expired air, and the device measures oxygen consumed (VO₂) and carbon dioxide produced (VCO₂).

The logic is elegant: burning fuel consumes oxygen and produces carbon dioxide in known ratios. Measure the gases and you can calculate the energy released. A test takes twenty to thirty minutes of lying still, and it produces a genuine measurement rather than an equation's guess.

The respiratory quotient — a preview of Chapter 6

Here's the part that matters beyond the number. The respiratory quotient (RQ) is the ratio of CO₂ produced to O₂ consumed, and it tells you what fuel you're burning:

Look at that middle row, because it contains the whole of Chapter 6 in advance. A resting human sits at around 0.8 — which is not 0.7 and not 1.0. You are not burning fat or carbohydrate. You are burning both, continuously, in a ratio that shifts with what you last ate, how hard you're working, and how long since your last meal.

There is no switch. There has never been a switch.

So is a $150 gym metabolic test worth it?

It's a real measurement, so this is a genuine question rather than a dismissal.

What you get: a measured RMR rather than an estimated one, which removes the ±10% equation error and replaces it with the test's own error.

What the caveats are:

• Test conditions matter enormously. Proper RMR measurement requires an overnight fast, no exercise beforehand, no caffeine or nicotine, a rested state, and a thermoneutral room. A test done at 2 p.m. after you drove across town and had coffee is measuring something, but it isn't your RMR.
• Test–retest variability is real. Measure the same person on two different mornings and you will not get the same number. The spread is smaller than the equation's ±10%, but it is not zero, and a single test reported to the calorie hides it.
• It measures RMR, not TDEE — and TDEE is what you actually wanted. You still have to multiply by an activity factor from a dropdown, which reintroduces the largest source of error in the whole calculation (§5.4). You have paid $150 to improve the more accurate half of a two-part estimate.
• It doesn't tell you what to do. Knowing your RMR is 1,640 rather than 1,720 changes almost nothing about the intervention.

🧾 Cost check. Gym or clinic RMR testing: $100–$250, sometimes packaged as a repeatable monthly service. The information is genuine. Its practical value over the free two-to-three-week trial in §5.4 is close to zero for most people — because the trial measures your actual TDEE under your actual conditions, which is the number you wanted, and it costs nothing but patience.

Where testing genuinely earns its money: clinical situations. Critical care, where feeding targets matter and equations perform badly. Post-bariatric patients. People with very high or very low body fat, where the equations are least reliable. Athletes with unusual body composition. Cases where a genuine metabolic disorder is suspected and the measurement will inform treatment. In those settings it's not a gadget — it's the right tool.

5.11 When it actually is medical

I've spent this chapter arguing that "my metabolism is broken" is usually not what's happening. Now the counterweight, because the failure mode in the other direction is real and I've watched it hurt people.

Some people saying that sentence have an undiagnosed medical condition, and they have been told for years that they're eating more than they think. A clinician who reflexively reaches for the underreporting explanation will eventually miss a treatable illness in someone who has been telling the truth the whole time.

So: the differential, briefly.

Two rules that fall out of this table.

Rule one: symptoms are the signal, not the weight. Weight change alone is a weak diagnostic sign, because it has a hundred causes. Weight change plus cold intolerance and hair thinning, or plus purple striae, or plus a new medication, is a different conversation entirely. Ask what else is happening.

Rule two: a blood test costs less than four months. Renata spent four months at "about 1,500 calories," frustrated and self-blaming. A TSH test would have taken a week and cost very little. Even if it comes back normal — and it usually does — ruling it out is worth doing, because it removes the possibility that has been sitting underneath the frustration the whole time.

⚠️ When to see a professional. See a physician, rather than adjusting your diet further, if you have: unexplained weight gain or loss alongside fatigue, cold or heat intolerance, hair or skin changes, or bowel changes · rapid weight change over days (suggesting fluid) · swelling of legs or abdomen · irregular or absent periods · new weight change after starting a medication · loud snoring with daytime sleepiness · or any weight change that genuinely doesn't fit what you're eating.

And bring the food diary. A physician with three days of real data in front of them can do far more than one working from "I don't eat much." That's the most useful thing Chapter 4's checkpoint produces, and almost nobody thinks to take it to an appointment.

What we don't know

We can't predict individual metabolic adaptation. Two people running identical deficits will adapt to different degrees, and we currently cannot say in advance who will be which. We don't know how much of the variation is genetic, how much is prior dieting history, and how much is behavioral change we failed to measure.

We also genuinely don't know how durable adaptation is in typical weight loss — most of the long-follow-up data comes from extreme interventions, and extrapolating from those to a person who lost 8 kg over a year is exactly the kind of move Chapter 2 warns against.

Spaced Review

Try to answer before reading on.

1. (Chapter 4) Which component of energy expenditure falls most in a deficit, and why is that particularly hard to counteract?

NEAT. And it's hard to counteract because it's unconscious — you cannot decide to stop fidgeting less. The countermeasure isn't willpower; it's making movement structural and deliberate (a scheduled walk, a standing desk) so it doesn't rely on an unconscious process that's being suppressed.

2. (Chapter 4) Why can't you conclude from "my weight hasn't moved in three weeks" that your metabolism has adapted?

Because of §4.9: daily weight noise is 1–2 kg while a good rate of fat loss is ~70 g/day, so three weeks of a real deficit can easily be invisible; and §4.7: your estimate of intake carries ±20–30% error skewed toward underestimation, so the deficit may be smaller than you think or absent. Both explanations are far more likely than adaptation, and both are checkable.

3. (Chapter 2) The weight-loss competition follow-up study is small, uncontrolled, and studies an extreme intervention. Why is it still worth citing?

Because it measured rather than estimated — indirect calorimetry and body composition, six years out — which almost nothing else in this literature does. Chapter 2's principle: a study's limitations tell you what it can support, not whether to discard it. It supports strong claims about extreme interventions and weak claims about everyone else, and the error is using it for the second.

Project Checkpoint: Calculate Your Own Numbers — With Error Bars

Component five of Your Nutrition Framework. Fifteen minutes with a calculator.

Step 1 — RMR. Use Mifflin-St Jeor from §5.3. Write out every line of the arithmetic rather than using a website, once, so you can see what the equation is actually doing.

Men:    (10 × kg) + (6.25 × cm) − (5 × age) + 5
Women:  (10 × kg) + (6.25 × cm) − (5 × age) − 161

Step 2 — the error bar. Multiply your answer by 0.9 and by 1.1. Write down all three numbers. This is the step nobody does, and it's the point of the exercise.

My RMR estimate: ______ kcal/day · Realistic range: ______ to ______

Step 3 — activity factor. Choose from §5.4 and be conservative — most people overestimate. If you have a desk job and exercise three times a week, you are probably closer to 1.4 than 1.55.

Step 4 — TDEE, with its error bar. Multiply all three RMR figures by your factor.

My TDEE estimate: ______ kcal/day · Realistic range: ______ to ______

Step 5 — the comparison that matters. Put this next to your three-day diary total from Chapter 4.

That last row is the whole exercise. If your surplus or deficit is smaller than the uncertainty in your own numbers, then you have not established anything — which is genuinely useful to know, and is the situation for a great many people. If it's substantially larger, as Theo's +490 was, you've learned something real.

Step 6 — the honest protocol. Write this down and follow it instead of trusting the number:

Eat at approximately my estimate for 2–3 weeks. Track weekly average weight, not daily. Adjust the number based on what actually happened. My body is a better calorimeter than the formula.

Non-tracking alternative. Calculate the numbers and write them down as context — understanding that your body has an approximate energy requirement, and roughly what it is, is valuable in itself. Then skip the comparison to intake, and instead record: am I eating in a way I could sustain for a year? That question does more work than most numbers. Appendix B has the full toolkit; Appendix F has the worksheet.

Next checkpoint (Chapter 6): your fuel-mix reflection — mapping a day's energy demands onto the substrates your cells actually use.

Chapter Summary

The vocabulary: BMR (lab conditions) · RMR (realistic, ~5–10% higher) · TEF (~8–12% of intake) · TDEE (everything). "Metabolism," casually, means TDEE.

What it's spent on — the table that reframes everything:

Mifflin-St Jeor, worked: Theo 1,908 · Devi 1,334 · Walt 1,639 — each ±10%, a spread of 270–380 kcal. Multiply by an activity factor chosen from a dropdown, and the compounded uncertainty exceeds most people's target deficit.

🚪 The threshold: metabolic adaptation is real, measurable, persistent — and modest. A smaller body explains most of a post-loss drop in expenditure; adaptation is the remainder, on the order of tens to a few hundred kcal/day, larger with more aggressive deficits. Your metabolism is not broken; it is responsive.

What genuinely varies: body size (dominant) · NEAT (largest modifiable) · lean mass · age (later and less than folklore) · genetics (~±10%, real, not an explanation for 1,000 kcal) · thyroid (testable, treatable).

This chapter's verdicts:

The one thing to remember: your metabolism is not broken — it is responsive, and it is responding to what you're doing. That's a far more useful fact, because responsive systems can be worked with and broken ones can only be mourned.

What's Next

Chapter 6 goes one level down — inside the cell — and does it without requiring a chemistry background. Glycolysis, beta-oxidation, gluconeogenesis, ketosis, mTOR.

It exists because you cannot properly evaluate the keto argument, the fasting argument, or the protein argument without knowing roughly what's happening in there. And it dismantles the single most abused idea in fitness content: the notion that you burn one fuel at a time.

You don't. You never have. And the "fat-burning zone" on the treadmill is about to make a great deal more sense — and a great deal less difference.