How much protein do you really need?

You turn the packet around. Protein: 12 grams per portion. And then you stand there. Is that a lot? Is that enough for a day, for a meal, for you?

Almost everyone carries one number in their head: zero point eight grams per kilogram of body weight. In my practice I hear almost daily how it gets read: as an upper limit, beyond which things turn dangerous. I will show you where it comes from and how its history changes the way you read it.

The most famous number in nutritional medicine, and what it actually says

Ask three people how much protein an adult needs. Two will name the 0.8. Then ask what that number measures. And it goes quiet. No blame here: I learned it in medical school myself, without ever hearing how it came about.

It comes from nitrogen balance. Protein is practically the only nutrient that supplies nitrogen in any relevant amount. You measure how much someone eats and how much they lose through urine, stool and skin. Picture a bank account: the point where deposits and withdrawals cancel out counts as the requirement.

Robert Wolfe and Sharon Miller named the conceptual problem in JAMA in 2008: a recommended allowance of this kind is the amount that prevents a deficiency in almost everyone, not an optimum for other endpoints. They called it a misunderstood concept.

And official nutritional medicine is moving too. The German Nutrition Society names 0.8 grams per kilogram for people aged 19 to 64, and from age 65 now 1.0. That is science at work.

And now you know why the same number can be either entirely right or entirely unsuitable, depending on the question.

Two methods, two numbers: why the field is still arguing

Imagine weighing the same person on two scales. One reads 66 kilograms, the other 90. You would start doubting the scales. That is exactly what happens with protein.

The criticism of nitrogen balance sits in the specialist literature, not in blogs. Ronald Elango and Paul Pencharz have gathered it: losses through skin, hair and sweat are hard to capture, which makes the balance look better than it is. And the analysis fits a straight line through data that show a bend.

The alternative is called indicator amino acid oxidation, IAAO for short. Back to the building site: if one component is missing, construction stalls, and the rest get burned. You label one component and measure how much of it is breathed out. From the point where nothing is missing any more, the burning stops falling. That bend counts as the requirement.

In 2007 a team around Mohammad Humayun gave eight young men seven protein levels from 0.10 to 1.8 grams per kilogram: mean requirement 0.93, population-safe value 1.2. In twelve women over 65 the same researchers found 0.96 and 1.29, in six men over 65 then 0.94 and 1.24. A Chinese group arrived at 0.94 as well in 16 older adults in 2023.

And now you know why the recommendations diverge even though everyone reads the same literature.

What more protein delivers in muscle, in numbers

You may know the sentence from the gym: more protein, more muscle, on and on. The data say something more precise.

More interesting than the boundary is the shape of the curve. The same Japanese group laid a spline model over 105 studies with 5402 participants in 2021. Below 1.3 grams per kilogram, every additional tenth brought on average 0.39 kilograms more fat-free mass, above that only 0.12.

A third analysis across 74 randomised trials found gains in fat-free mass only with resistance training. From age 65, effects showed up already between 1.2 and 1.59 grams per kilogram, in younger people only from 1.6. The authors explicitly call the effects small.

A second route runs not through muscle but through how much you eat overall: Full is a signal. Related to that are the calorie deficit and its limits and weight loss injections and muscle loss.

And now you know why the answer is not more protein equals more muscle, but rather: it depends on where you start.

Not only how much, but how good: from biological value to DIAAS

Up to here it was only about grams. As if protein were protein. It is not. Two portions each carrying 20 grams of protein on the label can deliver very different amounts of usable building blocks.

For a long time the field worked with PDCAAS. That value rates the digestibility of the whole protein and caps the result at 1.0. Since 2013 the Food and Agriculture Organization of the United Nations has recommended the successor DIAAS, the Digestible Indispensable Amino Acid Score. Picture a building site that needs nine sorts of screws: PDCAAS counts the lorries at the gate, DIAAS opens them and counts how many screws of each sort are still there at the end of the small intestine.

In 2020 a Dutch group calculated DIAAS values for 17 protein sources using the FAO age reference patterns. From that you can build an overview that speaks for itself.

And now you know why two labels with the same gram figure do not promise the same thing.

Lysine, methionine and the art of combining

An old rule of thumb says: grains lack lysine, legumes lack methionine, so combine the two. Rice with beans, bread with lentils, hummus with flatbread. Almost every food culture found this independently.

The counterpoint sits in the same dataset. Leucine ranges from 5.1 percent in hemp to 13.5 percent in corn. Corn thereby sits above milk with its 9.0 percent. Anyone claiming across the board that plant protein has too little leucine is not doing this dataset justice.

Why animal sources still often perform more strongly in muscle experiments is sorted out by a review from Stephan van Vliet and Luc van Loon: lower digestibility, stronger splanchnic extraction, meaning more diverted into gut and liver, and the relative shortage of individual amino acids. Their four ways out: fortification, breeding, larger amounts, mixture.

And now you know why amount and quality can answer for each other.

The leucine threshold and anabolic resistance in later life

Two people eat the same omelette. One is thirty, the other seventy-two. In the muscle, the same thing does not happen. That is not a feeling, that is measured. The technical term is anabolic resistance: ageing muscle responds on average more weakly to the same amount of amino acids.

In 2005 a British group around Daniel Cuthbertson and Michael Rennie gave a total of 44 healthy young and old men amino acids while insulin was held constant. The resting rate was the same, yet the older men responded more weakly, with less activation of mTOR, p70 S6 kinase, eIF4BP-1 and eIF2B. At the same time they showed clearly raised levels of the inflammatory transcription factor NFkappaB.

Here I like to add the lens of Clinical Psychoneuroimmunology. NFkappaB counts as one of the central switches of silent inflammation. If it is upregulated in ageing muscle while the building signals grow weaker, anabolic resistance might also have an immunological side. That is an association, not a causal chain: mechanistically plausible, human studies thin. More on this in silent inflammation and weight.

Picture a doorbell. In young muscle a short tap is enough. In older muscle the bell is set quieter, you have to press longer. Leucine is the finger on the button. A group around Daniel Moore calculated from pooled laboratory studies how much is needed per meal: the older participants reached the plateau only at 0.40 grams of protein per kilogram, the younger ones already at 0.24.

As pure arithmetic to think along with, not as a recommendation: 0.40 grams per kilogram comes to roughly 28 grams per meal at 70 kilograms, the 0.24 of the younger group to roughly 17 grams. That conversion is mine, not from the study.

In 2013 PROT-AGE named 1.0 to 1.2 grams per kilogram per day for healthy older people, the European society for clinical nutrition in 2014 the same order of magnitude, 1.2 to 1.5 during illness. Both name physical activity in the same breath.

And now you know why the same portion can trigger a different answer at seventy than at thirty.

Sarcopenia: why protein alone is only half the answer

There is a moment many people know around the age of sixty. The shopping bag is heavier than it used to be. The armchair sits lower. The staircase has more steps.

This has a name: sarcopenia. In 2019 the European working group clarified two things. Sarcopenia counts as a muscle disease, not as a normal sign of ageing. And muscle strength stands at the centre, not mass. How common it is depends heavily on the definition, which is why I name no percentage.

The third number comes from the Health ABC cohort with 2066 people between 70 and 79 years old: the highest protein fifth lost 0.501 kilograms of lean mass, the lowest 0.883. Everyone lost, but at different speeds. A cohort proves no cause.

What that stimulus can look like is in Resistance training from 40 and longevity. And now you know why a protein shake without movement stays an incomplete strategy.

Distribution across the day: important, but less than you think

Hardly any rule gets repeated as often: spread your protein evenly, three times roughly the same. I passed it on that way for years myself. Then I read the counter study.

First the paper almost everyone cites. In 2014 a group around Madonna Mamerow gave eight adults the same amount of protein for seven days, once spread evenly, once heavily skewed. The 24 hour muscle protein synthesis was 25 percent higher with even distribution. Except: eight people, and the contrast was extreme, 10.7 against 63.4 grams.

A third data point: in 2013 an Australian group around José Areta compared three patterns for the same 80 grams of whey protein in 24 trained men. Four times 20 grams sat 31 to 48 percent above the other two patterns.

And in 2023 Jorn Trommelen and Luc van Loon gave either 100 or 25 grams of protein after exercise, tracked with four simultaneous isotope labels. The large portion produced a larger and longer lasting anabolic response, dose dependent, with no visible ceiling. The claim that more than 30 grams per meal cannot be used no longer holds as a hard upper limit on these data.

And now you know why I have become more careful about the distribution question than I used to be.

The three worries: kidney, bone, cancer

Three sentences reach me again and again. Protein strains the kidneys. Protein makes the bone acidic. Protein feeds cancer. All three deserve numbers instead of reassurance.

The kidney

The core is true: a higher protein intake can raise the filtration work of the kidney. Is a higher filtration rate damage or an adaptation?

At the top end there is one signal: fourteen resistance-trained men ate over 3 grams per kilogram for parts of a year, without conspicuous liver or kidney values. Young men with self-reported logs are, however, no basis for general statements.

The bone

The acid-ash hypothesis sounds logical. Protein supplies sulphur-containing amino acids, acid arises from them, and the body buffers it with calcium from bone. And indeed, in balance studies calcium in the urine rises with higher protein intake.

In 2009 a Canadian group around Tanis Fenton did the arithmetic and admitted only the methodologically strictest balance studies: of 16, five remained. The acid load clearly raised calcium in the urine. But there was no link to calcium balance, at 94 percent statistical power, and none to the bone resorption marker N-telopeptide. In 2017 the National Osteoporosis Foundation, across 16 randomised trials and 20 cohorts, even found a small positive effect on lumbar spine bone density, plus 0.52 percent.

Cancer

Here real uncertainty remains, and I do not want to argue it away. The worry usually goes back to a 2014 paper by Morgan Levine and Valter Longo. In the observational part, high protein intake in 50 to 65 year olds was linked to clearly higher overall and cancer mortality, while above 65 the picture reversed. The second part consists of mouse experiments supporting the IGF-1 mechanism. Anyone citing the study should cite both halves and name the animal part as an animal part.

Two larger analyses draw a different picture. A meta-analysis of 32 cohorts with more than 715,000 people found lower overall mortality with higher total protein intake.

That is the integrative lens in its purest form: it is not the isolated substance that decides, it is the context. The question of meat and processing therefore has its own text: Meat and cancer.

And now you know why I give neither an all-clear nor a warning here, but a separation: healthy kidneys against diseased kidneys, amount against source, substance against context.

If you would like to place this in your own situation rather than only read it: below this article there is the option to arrange an appointment.

Frequently asked questions about protein requirement

Protein in the bigger picture

The protein question never stands on its own. It hangs on muscle, on satiety and on how you move.

Why insulin resistance shifts the calculation further is covered in Insulin resistance and weight loss.