← Longevity & Supplement Guides
Who Decided How Much Protein You Need?
In November 1902 a Yale professor of physiological chemistry began to eat less on purpose. Russell Chittenden was nearly 47 and weighed 65 kilos. Over a month or two he cut his breakfast down to a small cup of coffee. He kept a light lunch and a proper dinner, and he collected his own urine so he could measure the nitrogen in it. By June 1903 he weighed 58 kilos. He felt, he wrote, better than he had in years.
The experiment was not a diet. It was an attack on a number. In 1902 every serious physiologist knew roughly how much protein a working man needed, and the figure had come from Munich: about 118 grams a day. Chittenden thought the figure was wrong by a factor of two or three. He spent the next two years, and a long book, trying to prove it.
He was mostly right. That should surprise anyone who has watched protein become the nutrient of the decade. The number on today’s food labels has a strange history. It was set too high, then too low, then too high again. Each time, it reflected what that era was afraid of. This is the story of how we got to 0.83 grams per kilo, and why that figure answers a narrower question than most people think.
The number from Munich
In the late nineteenth century, Carl von Voit of Munich was the most respected nutrition scientist in Europe. He gave a standard for an adult man of 70 to 75 kilos doing moderate muscular work: 118 grams of protein a day, of which 105 should be absorbable, plus 56 grams of fat and 500 of carbohydrate. His student Max Rubner put the figure at 127 grams. In the United States, Wilbur Atwater, the man who gave us the calorie, suggested about 125 grams for a man doing hard labour. For soldiers, the standards went higher still.
These numbers were not foolish. Protein was the stuff of muscle, and muscle was the stuff of work. Rich industrial nations ate a lot of meat, and they also had the strongest armies and the fastest-growing economies. It seemed natural to read the first fact as a cause of the second. The standards also came largely from watching what working men who could afford meat actually ate. Then they were written down as what every man needed.
Two British nutritionists, D. S. Miller and P. R. Payne, looked back at this period in 1969. They noted that many early estimates were simply a reflection of the eating habits of the people studied. The turn-of-the-century figures of 118 to 150 grams, they wrote, came to be defended as irreducible minimums, not merely as feeding standards. That shift, from what people ate to what people need, is the hinge of this whole story. It has happened more than once since.
Three groups and a urine flask
Chittenden’s method was nitrogen balance. Protein is the only major nutrient that carries nitrogen. If you measure the nitrogen that goes into a person as food and the nitrogen that comes out, you can tell whether the body is holding its protein, losing it, or gaining it. If input equals output, the person is “in equilibrium”. Chittenden’s question was simple: how low can protein intake go while a healthy person stays in equilibrium, keeps their weight, and keeps their strength?
He did not stop at himself. His 1904 book, Physiological Economy in Nutrition, describes three groups:
- Five professors and instructors at Yale, chosen to represent the “mental worker”. One was his colleague Lafayette Mendel, who would later become famous for vitamin research.
- Thirteen volunteers from the Hospital Corps of the US Army. They lived in New Haven for six months, until April 1904, ate a controlled diet, and trained in the Yale gymnasium every weekday.
- Eight student athletes, some with exceptional competition records.
His own figures are the clearest. For nearly nine months, at 57 kilos, he broke down an average of 5.7 grams of nitrogen a day. That is about 0.1 gram of nitrogen per kilo, or 0.625 grams of protein per kilo of body weight. In his last two months, the figure dropped to about 34 grams of protein a day. That is roughly a third of Voit’s standard, with about 2,000 calories of food in total. Mendel, at 70 kilos, held equilibrium on about 41 grams a day for seven months.
Keep the number 0.625 in mind. It comes back about a hundred years later.
The original data, then and now
The protein number, in grams per kilo of body weight per day
What the soldiers really showed
The soldiers were Chittenden’s showpiece. Over six months on a reduced-protein diet they did not weaken. They got much stronger, and the dynamometer tests rose steadily. He printed photographs of the men to prove it.
Here the book claims more than its design can carry. The men were also training in a gymnasium every weekday, many of them for the first time in a structured way. Chittenden saw the objection. He wrote that the daily training must count for something, but that increased skill alone could not explain the great gain in muscular power. Then he credited the gain to the diet. No group did the same training on Voit’s 118 grams, so the experiment cannot separate the two causes. Six months of daily gym work will make most untrained men stronger on almost any adequate diet.
He made a similar leap about himself. He believed several old minor complaints went away when he ate less protein. That is one man’s notebook, and he was cutting total food at the same time. German critics such as Carl von Noorden raised a different objection: a few months of equilibrium do not show that a person can stay healthy on that intake for decades.
Those are fair points against his conclusions. They are not points against his measurement. The nitrogen numbers were careful, repeated daily, and they matched across very different people. Chittenden showed that healthy adults can stay in balance on far less protein than the Munich standard. He did not show that less was better. On the first claim, history has proved him right. The second claim is where he slipped from data into belief.
Why the world kept eating more
The book was widely read and widely argued over. It did not change how people ate. The record does not give one clean reason. A few plausible ones overlap. Voit’s standard was already built into institutional menus. Meat was a sign of prosperity. And a scientist telling well-fed people to eat less of the food they valued most was not a message that spread easily.
The next big step came from the state, not from the laboratory. During the Second World War, the US National Academy of Sciences set up a committee to advise on nutrition for national defence. It became the Food and Nutrition Board in 1941, and it published the first Recommended Dietary Allowances. The nutritionist Lydia Roberts helped to lead the work. From that point, protein was a policy number with a committee, a revision cycle and a table. Such numbers do not stay in the table. They go into school meals, military rations, food aid and, eventually, the side of a cereal box.
The protein gap
After the war, protein became a global emergency, or seemed to. Doctors in Africa had described a severe childhood malnutrition syndrome, kwashiorkor, and it was widely read as a sign of protein shortage. Requirement figures for children were set high, partly from animal studies. When economists multiplied those figures by the world’s population and compared the result with the food supply, they found a large and growing “protein gap”.
The United Nations set up a Protein Advisory Group in 1955. Over the next two decades, aid money and research effort went into high-protein cereal strains, amino-acid fortification, fish meal, oilseed flours and protein grown from microbes. The nutrition writer Geoff Webb has estimated the total at billions of pounds in today’s money.
In July 1974 Donald McLaren, a nutritionist at the American University of Beirut, published a paper in The Lancet called “The great protein fiasco”. His argument was that the protein gap rested on a false premise. Most malnourished children did not lack protein specifically. They lacked food. If children ate enough of almost any staple diet to meet their energy needs, the protein would usually follow. A year later, John Waterlow and Peter Payne made a similar case in Nature under the plain title “The protein gap”. Their analysis of children’s diets found that diets short of protein but adequate in energy were rare.
Requirement figures came down. In Webb’s account, the US allowance for a two-year-old fell from 40 grams a day in 1943 to 18 grams. The protein gap mostly disappeared, not because the world grew more protein, but because the arithmetic changed. Later researchers have argued that protein quality still matters for children’s growth in some settings, so the debate is not fully closed. But the main lesson stands: a requirement number that is set too high does not stay a technical error. It shapes policy, budgets and what goes into children’s bowls.
Where 0.83 actually comes from
The figure most adults meet today comes from a 2003 paper by William Rand, Peter Pellett and Vernon Young in the American Journal of Clinical Nutrition. They collected every suitable nitrogen-balance study they could find: 19 studies covering 235 healthy adults, each tested at three or more protein intakes. For each person, they estimated the intake at which nitrogen in matched nitrogen out.
The median requirement was 105 milligrams of nitrogen per kilo per day, which is about 0.66 grams of protein per kilo. People differed, and the spread was log-normal. To cover almost everyone, committees take a high point on that distribution rather than the middle. That gives about 0.83 grams per kilo. The WHO/FAO/UNU expert consultation used this analysis. The European Food Safety Authority’s 2012 opinion uses the same logic: an average requirement of 0.66, and a population reference intake of 0.83 g/kg for adults of all ages.
Now put the two numbers side by side. Chittenden’s own nine-month average was 0.625 grams per kilo. The modern median is 0.66. A man with a urine flask and a notebook in 1903 was within about five percent of a meta-analysis published a century later. On the narrow question he asked, he was right, and the Munich standard was not.
What nitrogen balance cannot see
That narrow question matters. Nitrogen balance tells you whether a body is losing protein. It does not tell you whether muscle is being built, whether strength holds as you age, or how well you recover from training. The 0.83 figure is a floor for healthy, mostly sedentary adults. It is not an estimate of the best intake for any one goal.
The floor itself is also disputed. In 2010 Rajavel Elango and colleagues re-analysed the same nitrogen-balance data with a two-phase model and got a mean requirement of 0.91 and a safe intake of 0.99 g/kg. With a different method, which tracks how the body oxidises a marker amino acid, they estimated 0.93 and 1.2 g/kg for adult men. Their conclusion was that the requirement may have been underestimated.
Above the floor, the questions change. The PROT-AGE study group of European geriatric societies recommended in 2013 that adults over 65 eat at least 1.0 to 1.2 g/kg, and 1.2 or more if they are active. They based this partly on how the ageing body responds less strongly to a given dose of protein. In 2018 Robert Morton and colleagues pooled 49 randomised trials with 1,863 participants. Protein supplements during resistance training added modest extra gains: about 0.3 kilos of lean mass and 2.5 kilos on a one-repetition maximum, on average. Above about 1.6 g/kg of total daily protein, extra protein added no further lean mass.
Look at that last number again. 1.6 grams per kilo is almost exactly what Voit’s 118 grams works out to for a 70 to 75 kilo working man. The Munich standard was wrong as a minimum for everyone. As a rough description of what a hard-working, muscle-building body can use, it was not far off. Voit and Chittenden were each right about a different question, and each treated their question as the only one.
Read the number by its question
Four protein figures, four different questions
About 0.66 g/kg: the median requirement
The intake at which half of healthy adults in the nitrogen-balance studies stopped losing nitrogen. It is a midpoint, not a target. Rand, Pellett and Young, 2003.
0.83 g/kg: the reference intake on labels
Set high on the distribution so that almost all healthy adults are covered for balance. EFSA uses this figure for adults of all ages. It is a floor for not losing protein, not an estimate of the best intake.
1.0–1.2 g/kg: an expert view for adults over 65
The PROT-AGE group’s recommendation, aimed at keeping muscle and function with age; 1.2 or more if active. It is a position paper based on mixed evidence, not a single trial result.
About 1.6 g/kg: where added gains stopped in lifters
In a pooled analysis of 49 trials, more protein above this level added no further lean mass during resistance training. It is a ceiling for that one outcome, not a daily goal for everyone. Morton et al., 2018.
How the pendulum reached your kitchen
Each era has had a protein number that matched its anxieties. The industrial era wanted strong workers and soldiers, so it chose 118 grams. Chittenden’s moment cared about efficiency and moderation, so it chose economy. The post-war development years feared a hungry world, so they found a gap. The current market finds something else. In the most recent swing, protein has moved from a fear of shortage to a marketing claim. It is printed on cereal, coffee and crisps, as we covered in the fibre-versus-protein moment.
The current enthusiasm is not baseless. The evidence for more protein in older adults and in people who train is real, if modest. But the pattern is the same one Miller and Payne described in 1969. A number that describes a group with a particular goal turns into a number that everyone is told they must reach. The difference between “a level that a strength-training 30-year-old can use” and “a level that everyone needs” is exactly what Voit lost when his observation became a minimum.
What the regulators allow us to say is modest and precise. Protein contributes to a growth in muscle mass and to the maintenance of muscle mass, and it contributes to the maintenance of normal bones. EFSA did not set an upper limit. It noted that intakes up to about twice the reference intake are regularly eaten from mixed diets by some active, healthy European adults and are considered safe. That is a wide band. Most people in rich countries already sit inside it.
What the story is really evidence for
None of this history is evidence that you need more protein, or less. A 1904 nitrogen study tells us about balance in 26 men in New Haven, most of them young. A 1974 argument about child nutrition tells us about policy, not about your breakfast. The useful lesson is about how to read a number.
Every protein figure you meet answers a specific question for a specific group. First ask what the question was: not losing protein, keeping muscle as you age, or building muscle under training. Then ask who was studied. In our earlier piece on who extra protein actually helps, the honest pattern was that protein acts more like a floor than a dial. Getting enough matters, and going far above enough adds little for most people. That fits this history well. The floor is real and well measured. The dial is where most of the selling happens.
If you have a kidney condition or any medical reason to watch protein, talk to your doctor before you change your intake. These general figures were built from healthy adults.
The bottom line
The protein number on your label is about 120 years old, and it has been argued over the whole time. Voit described what hard-working men ate and called it a need. Chittenden measured what healthy men needed in order to not lose protein, then claimed more for it than his data showed. The post-war world multiplied a too-high number by a population and invented a gap. Today’s 0.83 g/kg is a careful answer to one narrow question. This week, try one small experiment of your own. For three ordinary days, write down roughly how much protein you eat, then divide by your weight in kilos. Most people will find they are well above the floor. Then decide which question you are actually asking before you buy anything with a bigger number on the front. If you train hard or you are over 65, the evidence points higher. And if timing worries you, the total for the day matters more.
