Your Metabolism Doesn’t Slow Down at 40

Every birthday after about thirty-five arrives with an unsolicited theory attached. The trousers fit differently, the scale has drifted up two kilos since the last time anyone checked, and somebody at the table explains it in four words: your metabolism slowed down.

It is the rare belief that is universal, intuitive, and — as far as the best measurement humans have ever built can tell — mostly wrong.

The only honest way to measure a day

You cannot know how much energy you burned yesterday. Not really. A treadmill console guesses from your weight and a formula. A wearable guesses from your heart rate and a formula. A food diary is a work of autobiography.

There is one method that does not guess, and it is beautiful. You drink a measured glass of water in which some of the hydrogen and oxygen have been swapped for heavy stable isotopes. The labelled hydrogen leaves your body only as water. The labelled oxygen leaves as water and as carbon dioxide. Track both in urine over a week or two, and the gap between the two disappearance rates is the carbon dioxide you produced — which is, to a very good approximation, the energy you spent. No self-report. No formula. Just isotopes doing bookkeeping while you live your life.

It is called doubly labelled water, it is expensive, and for forty years researchers have been quietly pooling every measurement ever taken into a shared database. In 2021 someone finally asked that database the obvious question.

Six thousand four hundred people, twenty-nine countries, one flat line

Herman Pontzer and a large collaboration published the answer in Science: 6,421 people, 64% of them female, from 29 countries, aged eight days to ninety-five years. It is the closest thing we have to a portrait of human energy use across a whole life.

The portrait has four phases, and the phases do not fall where anyone expected. Infancy is the metabolic riot — adjusted expenditure climbs roughly 85% in the first year and peaks around nine to fifteen months at about 150% of an adult's rate, pound for pound. Childhood and adolescence spend two decades coming down from that, reaching adult levels at a statistical breakpoint of 20.5 years.

Then nothing happens. For forty years, nothing happens.

The flat stretch runs straight through the years we blame

From twenty to sixty, total daily expenditure, resting expenditure and fat-free mass were all stable. No sex difference survived the adjustment for body size. The line runs level through the thirties, through the forties, through pregnancy, through the ages at which menopause typically arrives.

If human metabolism fell off a shelf at forty, a dataset of 6,421 people ought to show a bend at forty. It shows one at sixty-three, with a confidence interval of 60.1 to 65.9 years. And what follows the bend is not a collapse either: about 0.7% per year, which compounds to roughly 26% below middle-aged levels by the nineties. Real, worth planning for, and nothing like the folk model.

Eighty-three per cent of it is just how much of you is not fat

Here is the sentence in the paper that does the most work. Across the entire dataset, fat-free mass alone — muscle, organs, bone, everything that is not adipose tissue — accounted for 83% of the variation in how much energy a body spent per day.

That number reframes the whole argument. The flat line is what remains after you account for body composition. The decline people feel in their fifties is largely the unadjusted version of the same curve: not a slower engine, a smaller one. Tissue that costs energy to run gets quietly traded for tissue that does not, a little each year, and the bill arrives as a number on a scale.

Which is the useful half of this whole story, because fat-free mass is the one term in the equation you can actually move. Resistance training after forty and enough protein are not vanity projects; they are maintenance on the denominator. Be honest about the size of the effect, though: resting muscle is metabolically cheap, so adding a couple of kilos of it changes your daily burn by tens of calories, not hundreds. The reason to keep it is what muscle does, not what it subtracts from dinner.

So why does everyone burn less than their parents did?

Because they do — and for a reason nobody predicted. In 2023, John Speakman and colleagues went back to the same isotope database with a different question: has human energy expenditure changed over the decades? Using around 4,000 measurements from adults in Europe and the United States, they found that size-adjusted total expenditure has fallen since the early 1990s by roughly 7.7% in men and 5.4% in women.

Now the twist. The part that fell was not activity. Activity expenditure held steady or edged slightly upward. What declined was basal expenditure — the cost of simply being alive at rest — and the same team replicated that decline in a much larger set of 9,912 basal-rate measurements drawn from 163 studies spanning a century.

So the tidy story we all tell, that we are heavier than our parents because we sit more than they did, is not what the isotopes say. Something has changed in the resting bill, and the honest position is that nobody has pinned down what. Body composition, ambient temperature, sleep, the composition of the food supply and drift in measurement practice are all live candidates. It is also a comparison between cohorts, not a film of the same people ageing — which is exactly the kind of caveat that gets dropped when a finding is this quotable.

Exercise buys less than the treadmill console claims

There is a third strand, and it is the one people find hardest to hear. In 2016 Pontzer's group measured 332 adults across five populations with doubly labelled water and accelerometers at the same time. Physical activity explained less than 9% of the variation in daily energy expenditure. Above roughly 219 accelerometer counts per minute per day, each additional 100 counts was associated with under 50 extra kilocalories a day.

Expenditure, in other words, behaves less like a bank account and more like a thermostat. A February 2026 review in Current Biology by Pontzer and Trexler gathers the evidence across humans and other animals and argues that when activity rises, the body pays for it by spending less elsewhere — reallocating a finite budget rather than enlarging it. That model is genuinely contested; additive assumptions still run most nutrition software and most animal ecology. Take it as a strong argument in a live debate, not a verdict.

But read it the right way round. If the daily total is roughly defended, then exercise was never a subtraction exercise to begin with. What training changes is what the budget is spent on — which is a physiological argument, and a far better one than the arithmetic ever was. It also explains why the dose that shows up in outcome studies has so little to do with the numbers glowing on a machine.

What to do with a flat line

Four things, all of them small.

Stop budgeting your life around a decline that starts near sixty-three. If something genuinely changed for you at forty-two, the suspects are behaviour, sleep, alcohol and body composition — not your birth certificate. That is a much better list, because every item on it is negotiable.

Defend fat-free mass, since it is 83% of the equation. Two or three resistance sessions a week, held for years, plus protein spread across the day. There is a dose worth knowing, and it is smaller than the internet implies.

Read exercise as physiology, not currency. Never cancel a session because the console only offered you 240 kilocalories. That number was always the least interesting thing the session did.

Look at the intake side with fresh suspicion. If expenditure is this stable and this defended, then drift in weight is more likely to be arriving through the door than leaking out of the engine — which is why what the food is made of and how fast you eat it turn out to matter more than another recalculation of your basal rate.

And if you are past sixty, the decline is real: 0.7% a year is about 7% over a decade. Worth adjusting for. Not worth dreading.

The honest limits

Doubly labelled water is the reference method, not a perfect one. It measures a week or two, not a year. The life-course curve is assembled from thousands of people measured once each, not from individuals followed for ninety years, so it describes a population's shape rather than your personal trajectory. Isotope behaviour itself may shift with age, which could bias estimates in the oldest groups. And the adjusted flat line is a statistical construct: adjusting for fat-free mass is the right way to ask whether cells got slower, but you live in the unadjusted world, where composition changes and the bill changes with it.

None of that rescues the folk model. Every reasonable reading of the best data puts the metabolic bend two decades later than the conversation at the dinner table does.

The bottom line

The largest measurement of human energy use ever assembled says your metabolism does not slow down at forty. It holds flat from about twenty to about sixty-three, then declines by roughly 0.7% a year. Most of what varies between two adults is simply how much of them is not fat — which is the part you can still argue with. Use the numbers to correct the fantasy: the birthday is not the problem, and the console is not the point.