The fitness world treats "eat X grams of protein per kilogram of body weight" as scripture. It is repeated by every coach, every supplement label, every gym poster. But trace the number back to its source, and the story is surprisingly shaky.
The Official Endorsement First
China's official dietary guidelines recommend about 1.0 to 1.2 grams of protein per kilogram for ordinary adults (65 g/day for men, 55 g for women). But for athletes and gym-goers, the same official channels suddenly quote "1.2 to 2.0 g/kg, up to 1.6 to 2.2 for strength training."
Where did that come from? The guidelines' editorial committee — 27 people — included exactly one sports-nutrition specialist. Her core knowledge base comes from translating an Australian textbook, Clinical Sports Nutrition (Burke & Deakin). Search her published papers: more than thirty, mostly on exercise intervention in children with obesity, athlete dietary surveys, and endothelial protection. Not one measures Chinese athletes' protein requirements. A 2005 survey of elite Chinese athletes measured what they ate — not what they should eat. Those are different questions.
So the official "athlete" numbers in the guidelines were not produced by local research. They were copied.
The "1.6" Comes From One Meta-Analysis
Copied from where? First: a 2016 joint position statement by the American Dietetic Association, Dietitians of Canada, and the American College of Sports Medicine (1.2–2.0 g/kg for athletes). And second — the number gym-goers chant, 1.6–2.2 — comes from the 2017 ISSN position stand on protein and exercise.
Dig one layer deeper and the core evidence for 1.6 turns out to be a single 2018 meta-analysis by Morton et al. in the British Journal of Sports Medicine: 49 randomized controlled trials, 1,863 participants, conclusion that muscle gain stops increasing significantly beyond about 1.62 g/kg. Round it and you get 1.6.
Now look at who was actually in those 49 trials.
The average age was 35 ± 20 years — a huge spread — but the core was twenty-something young men. Fewer than 20% of participants were women. Most were beginners with under a year of training; people with years of real lifting experience made up under 10% of the sample. These are young American male novices, training for 6 to 12 weeks.
The confounding factors are too many to count: training plans, sleep, stress, total calories, carbs, absorption, muscle fiber type, testosterone — none of it strictly controlled. Just people training, and a muscle measurement at the end.
That measurement, mostly DXA, has an error of ±1 to 2 kg of muscle. The study's headline difference — 0.5 kg more muscle for the high-protein group — is smaller than the measurement error itself.
And the most basic data point, how much protein each participant actually ate? Measured three ways: 24-hour recall (asking people what they ate yesterday, from memory), food-frequency questionnaires (asking how often they ate beef or eggs over the past month), and weighed food records (accurate but rarely completed — used in fewer than a third of the trials). Most of the intake data in the meta-analysis was estimated from memory and questionnaires.
A conclusion built on that foundation — "1.62 g/kg is the ceiling" — rests on very soft ground.
The "2.2" Comes From Nitrogen Balance
The 2017 ISSN statement pushed the upper end to 2.2 g/kg. Its own basis, written in the paper: nitrogen balance assessment and amino-acid tracer studies.
Nitrogen balance works like a ledger. Weigh the nitrogen eaten (protein contains nitrogen) against the nitrogen excreted (urine and feces). If output is less than intake, the body is building protein. Find the intake where they balance, and that is "the requirement."
The method is crude, and even its users admit it. First, it measures whole-body nitrogen balance, not muscle — other organs also synthesize and break down protein. Second, the measurement window is days to two weeks, and balance swings wildly day to day. Third, it ignores protein quality, amino-acid composition, digestion and absorption, training stimulus, and hormone levels. Feed in X nitrogen, excrete Y nitrogen, get a recommendation — that simple.
Amino-acid tracer studies are a step up: label amino acids with stable isotopes, inject them, and track where they go — how much reaches muscle, how much is broken down. But they measure muscle-protein synthesis over hours, not long-term growth; they run in a lab with subjects lying or doing standard motions, not real training; and sample sizes are tiny — often a dozen people, mostly young male beginners again.
