Protein Maxing

The wall

Let me give you the picture I use, because once you have it the whole argument becomes obvious.

Your muscles are walls. You are building them.

To build a wall you need bricks. Those are the amino acids: the protein you eat, broken down into parts. There are different kinds of brick, some better cut than others, and people will argue about that forever.

But a pile of bricks is not a wall.

You need cement, or nothing holds together. That is the rest of what you eat: the vitamins, the minerals, the water. Unglamorous, invisible in the finished wall, and without it you have a heap.

You need workers. Someone has to lift each brick and place it. That is your recovery, and mostly that is sleep. Tired workers lay fewer bricks and lay them badly.

And above all you need a project. A drawing. How tall, how long, how thick. That is your training: the sets, the loads, the weeks.

Here is the part that matters. If the project says a two-metre by two-metre wall, then the wall needs a certain number of bricks and no more. You can have infinity bricks delivered. They will sit in the yard. They will not become wall. Eventually they are in the way, and some of them go to waste.

That is protein maxing. Ordering more and more bricks for a project nobody has drawn.

Why I call it a hallucination

I build with AI every day, and it gave me the only word that fits what has happened here.

When a language model hallucinates it does not produce nonsense. Nonsense would be easy to catch. It produces something fluent, confident, specific and correctly formatted, with nothing underneath. It will hand you a citation with a real-sounding author, a real-sounding journal, a volume, a page range and a DOI that resolves to nothing at all. Every surface feature of authority is present. The ground is missing.

The tell is always the same: precision and confidence rise together while the evidence stays flat. The more exact the claim gets, the less is holding it up.

Now read a protein post with that shape in mind.

1.6 grams per kilogram. Thirty grams per meal, four meals a day. Forty within the anabolic window. A leucine threshold to one decimal place. Complete versus incomplete. Quality scores ranked like league tables.

Fluent. Confident. Specific. Correctly formatted.

I am not saying anyone invented those numbers. Most trace back to a real study. I am saying the structure came loose from the evidence the way a hallucination does: each round added a decimal place, the certainty hardened, and now there is an entire vocabulary of precision sitting on top of a fairly simple finding.

And like a model hallucination, it is hard to catch precisely because it is not wrong at the edges. It is over-specified in the middle. You cannot refute 1.6 grams per kilogram. There is a study with that number in it. What you can notice is that nobody quoting it can tell you the interval around it, or what happens at 1.4, or whether it was measured in anyone like you.

That comparison is mine. It is an argument, not a finding, and no study below makes it. What the studies do is show where the ground actually is.

My own equation, which never had protein as the variable

I am lacto-vegetarian. Dairy, yes. Meat, fish and eggs, no.

That combination is genuinely awkward to count. Take away meat and fish and eggs and you have removed most of what protein tables are built around. What is left is dairy, beans, lentils, grains, nuts. It is perfectly adequate and it is a nuisance to tally, which is exactly why I never tallied it.

So protein was never a lever I pulled. It was a floor I stood on. I ate a balanced diet, I got a normal amount, and I stopped thinking about it.

What I did think about, constantly, was the rest: drinking enough water, sleeping properly, eating actual food with vitamins in it, showing up week after week, and lifting heavy.

And my body grew.

One person proves nothing, and I am not asking you to take my word for it. I am telling you what my equation looked like because for years the internet told me it could not work, and it did. That is not evidence. It is the reason I went and found some.

For what it is worth on the specific fear I was sold: a meta-analysis of eight randomised trials in 188 adults compared plant-based against omnivorous diets on strength and found no significant difference in upper body, lower body or overall strength, with every confidence interval crossing zero. That is a small evidence base and a crossed interval means no difference was detected rather than proven absent. But the confident claim was that my diet costs strength, and the randomised evidence gathered to test it did not find that.

The bricks: graded obsessively, and the grade does not predict much

Start with how much energy goes into ranking bricks.

A review examined the role and limitations of protein quality metrics, focusing on the Digestible Indispensable Amino Acid Score, when applied beyond assessing amino acid adequacy.

DIAAS is the score behind every comparison table you have seen: how well a protein's amino acid profile matches human requirements, corrected for digestion. It is a real measurement of a real thing.

And here is what the review reports. Evidence from randomised trials and large prospective cohorts indicates that although animal-derived sources typically achieve higher amino acid-based quality scores, these differences do not consistently translate into superior cardiometabolic or mortality-related outcomes.

Higher grade of brick. No better building.

It goes further: substitution analyses and intervention studies suggest replacing animal-based protein with plant-based sources is associated with favourable changes in cardiovascular risk markers and reduced chronic disease risk. The authors stress the food matrix and its bioactive compounds, and note that in Western populations total protein intake often already exceeds physiological requirements.

The paper's own title is the summary: protein quality based on amino acid composition is poorly associated with health outcomes.

This is the hallucination in one object. A number that is precise, correctly calculated, genuinely measuring something, and carrying an implication it was never shown to support.

This is a narrative review situating protein within whole dietary patterns. It reports the association between quality scores and outcomes as weak, not that protein quality is meaningless, and it is explicit that it concerns populations already eating enough.

The project: what actually decides the size of the wall

Now the drawing, and this is the biggest evidence base in the piece by a distance.

The American College of Sports Medicine published a position stand updating its 2009 guidance, synthesising 137 systematic reviews covering more than 30,000 participants, restricted to randomised trials of healthy adults training at least six weeks.

One hundred and thirty-seven reviews. Thirty thousand people.

Against no exercise, resistance training significantly improved strength, size, power, endurance, contraction velocity, gait speed, balance and multiple physical function outcomes.

Then the line I would frame: few resistance training prescription variables affected primary adaptations.

Few. After decades of argument, most of the knobs people turn did not change the result.

The ones that did are short and specific. Strength: loads at or above 80 percent of one-rep max, full range of motion, 2 to 3 sets, early in the session, at least twice a week. Hypertrophy: higher volumes, at or above 10 sets per week, and eccentric overload.

The list that did not reliably matter is longer, and includes training to momentary muscle fatigue, equipment type, exercise complexity, set structure, time under tension and blood flow restriction.

A companion meta-regression across 67 studies and 2,058 participants quantifies the volume finding: the posterior probability that gains rise with volume was 100 percent for both size and strength, with diminishing returns in both, considerably more pronounced for strength.

One hundred percent probability that more sets means more wall, on a curve that flattens.

That is your project drawing. Ten hard sets a week per muscle group is a specification. It tells you how many bricks the wall will consume. Nothing about your protein intake changes that number, and no amount of protein builds a wall that was never drawn.

The workers: why tired crews build nothing

Bricks and a drawing still need someone to lay them, and this is the part I actually protected.

A study using the China Health and Retirement Longitudinal Study across the 2011, 2013 and 2015 waves assessed the risk sleep duration carries for sarcopenia in older adults, using logistic and Cox regression with subgroup analyses.

Sarcopenia is the loss of muscle mass and function. It is the wall coming down.

The finding: four abnormal types of sleep duration were significantly associated with sarcopenia, and the authors conclude that abnormal sleep durations, including both sleep deprivation and oversleeping, are linked to heightened sarcopenia risk.

Both ends. Too little and too much.

I want to be careful here, because this is observational, in older Chinese adults, and it measures the loss of muscle rather than the building of it. Illness causes bad sleep as readily as bad sleep causes illness, and this design cannot separate those.

But set it beside the mechanism. A review examining ribosome biogenesis argues the primary structural bottleneck on muscle growth is not the acute signal that protein triggers, but the physical expansion of the cell's protein-building machinery through activation of RNA Polymerase I and the synthesis of new ribosomes.

Ribosomes are the crew. Protein is the brick delivery. If your constraint is how many workers you have and how well they are working, another lorry of bricks does not raise the wall. And you cannot hire ribosomes by eating. You build that capacity by training over months, and you protect it by recovering.

That review is mechanistic and proposes a hypothesis under active investigation rather than a settled fact. It tests no protein intake against another, and its authors are candid about methodological problems in their own field.

The cement: the boring nutrients nobody posts about

Between the bricks there has to be something holding them, and it is the least discussed part of eating well.

A narrative review synthesised evidence from the past five years on micronutrient status in adults aged 65 and over, comparing people in care settings with those living in the community.

Community-dwelling older adults showed high prevalence of deficiency, particularly vitamin D, calcium, magnesium, folate and zinc. Institutionalised adults fared worse: vitamin D deficiency in 70 to 94 percent, roughly 6.3-fold higher odds, zinc deficiency in 50 to 66 percent against 31 to 49 percent in the community, iodine deficiency in 67 to 78 percent against 22 percent.

The consequences the review lists are the ones that matter for a wall: frailty and falls.

Notice which nutrients those are. Vitamin D, calcium, magnesium, zinc. Not one of them is protein, and they are the difference between tissue that works and tissue that does not.

This review covers older adults, with the worst numbers in institutional care, so it does not describe a healthy adult eating a varied diet. I am citing it to make a structural point rather than a personal one: the nutrients that turn out to be commonly deficient, and consequential, are not the one the internet counts.

Eating a balanced diet is how you get the cement. There is no post about it because there is nothing to optimise, and that is precisely why it goes missing.

The thing that actually decides it: still being there next year

And now the finding that should end most protein arguments.

A retrospective cohort followed 587 adults enrolled between 2018 and 2024 at a supervised small-group training studio in São Paulo, analysing participation duration and dropout with Kaplan-Meier survival analysis, Cox regression and a decision-tree model.

Mean participation was 13.8 months. The 12-month survival rate was 51.7 percent.

Read that again. In a supervised programme, with a coach, in a small group, which is close to the best-case scenario for adherence, nearly half the people were gone inside a year.

That is the real competition for your attention. Not whether you hit 1.6 or 1.8 grams per kilogram. Whether you are still training in fourteen months.

Because the wall is not built by any single delivery. It is built by a crew turning up, week after week, working from a drawing, with the bricks already in the yard. Consistency is not a virtue I am recommending at you. It is the variable with the largest effect size in the whole equation, and it is the one nobody sells a supplement for.

This is a retrospective cohort at a single studio in one city, with dropout predictors specific to that population, and it measures attendance rather than muscle. It tells you how hard the human part is, not how much muscle anyone built.

What I would tell you, plainly

Get the base right and stop auditing it. Enough protein, from a balanced diet. If you are older, ill, dieting hard or recovering, raise it deliberately, because that is where the original real finding lives. Otherwise it is a floor, not a lever.

Draw the project. Ten or more hard sets per week per muscle group, heavy work above 80 percent for strength, full range, at least twice weekly. That is the ACSM list and it is short because most other variables did not matter.

Pay the workers. Sleep is not a recovery hack, it is the crew. Both too little and too much tracked with muscle loss in that cohort.

Buy the cement. Vitamins and minerals and water, from actual varied food. The commonly deficient nutrients are not protein.

And then keep turning up, because half of people in a supervised programme were gone within a year, and that statistic beats every macro argument ever had.

If you eat like I do, none of this is a handicap. I never optimised protein. I got the base right and spent my attention on the project, the crew and the cement. The wall went up.

These are my conclusions from the evidence and my own experience, not medical advice. Nothing here describes you specifically, and anything involving kidney disease, pregnancy, disordered eating or a diagnosed condition is a conversation with a clinician who knows your history.