Nutrition

The question this branch answers

Nutrition advice arrives with more precision than nutrition evidence supports. A target of 1.6 grams of protein per kilogram. A calorie figure calculated to the digit. A creatine loading protocol with a schedule. A gain rate stated as though it were measured.

Read the sources and the pattern is consistent. The direction of each recommendation is usually well supported. The precision attached to it is usually not, and the precision is what gets sold.

This branch asks what each number is resting on, and what to do given that the answer is often a range.

What is settled

Protein supplementation works, and the effect is modest in absolute terms. Pooled across 49 studies it added roughly 2.5 kg to a one-repetition maximum and 0.3 kg of fat-free mass alongside resistance training (Morton et al., 2018). The effect is real and it is not the difference between progress and no progress.

Creatine monohydrate works. Pooled across 23 studies in adults under 50, it significantly increased upper-body strength against placebo (Wang et al., 2024). It is the supplement in this publication with the clearest yes.

Energy balance determines weight change. Nothing in this branch disputes the arithmetic. What it disputes is whether you can know your side of the equation precisely enough to act on it without measuring.

Rate of change affects composition during a deficit. Losing weight slowly produced better lean mass outcomes than losing it quickly (Garthe et al., 2011), and the difference had disappeared at twelve months.

And the well-known numbers are less solid than their fame suggests. The 1.6 g/kg protein figure came from a breakpoint that did not reach statistical significance, with a confidence interval (the range of values the data are compatible with) spanning more than a two-fold range (Morton et al., 2018). The best resting metabolic rate equation misses by more than 10% for about a third of people (Frankenfield et al., 2005). The creatine dosing protocols people argue about did not determine lean mass gains in the analysis that compared them, although strength gains depended on a loading phase or a higher dose. Each of these numbers is softer than the way it is quoted.

Where it splits

Protein intake. The most-cited threshold in sports nutrition and where it came from. Three meta-analyses used three analytical methods and the method chose the answer. Our reading is that 1.6 g/kg is a reasonable floor rather than a target, and that the range above it is uncertain, not known to be wasted. (→ Nutrition 01)

Creatine. The evidence is strong enough that the interesting question is which of the many protocol arguments are load-bearing. Timing appears not to be. In older adults, loading and dose did not change lean mass gains, although strength gains depended on them. The evidence base is also heavily male, which is the largest limitation and the one most often omitted. (→ Nutrition 02)

Basal metabolic rate (BMR), total daily energy expenditure (TDEE) and the error around them. Two estimation steps, each carrying error, producing a total plausible error of a few hundred calories per day, which is the entire size of a normal deficit or surplus. No equipment removes this, because expenditure is not a fixed property. The estimate is a starting hypothesis and your weight trend is the measurement. (→ Nutrition 03)

How fast to lose, how fast to gain. Slow weight loss preserved lean mass better in the reference trial and the advantage was gone a year later. We give both numbers because the second bounds what the first means. On gaining, the trial evidence does not exist and we give our position anyway rather than refuse. (→ Nutrition 04)

What this branch does not cover

Meal timing and distribution. Per-meal protein thresholds, the anabolic window and eating frequency each have their own literature. Some researchers consider distribution more consequential than daily totals.

Carbohydrate and fat. Once protein and total energy are set, the remaining split has its own body of work, largely concerning performance and preference rather than muscle.

Supplements other than creatine. Caffeine, beta-alanine, citrulline and the rest. Each needs the same treatment we gave creatine, which is more than a paragraph each.

Micronutrients and deficiency. A different question with clinical dimensions we are not equipped to cover. The exception we have addressed is high-dose antioxidant supplementation, which sits in the inflammation branch because the argument about it concerns adaptation rather than nutrition. (→ Inflammation sub-hub)

Anything sold as accelerating fat loss. The base rate of supported claims in that category is low enough that it is not where we would start.

Clinical nutrition of any kind. Nothing here is written for anyone managing a medical condition through diet.

A note on what counts as nutrition here

We have placed this branch under Recovery, which is a judgment rather than a taxonomy. Protein and creatine act on the same processes the training branch is about, and energy balance sets the conditions everything else operates under. None of it is recovery in the sense of something you do after a session.

The reason it sits here is that these are the levers outside the gym that determine what the work inside it produces. Read alongside sleep, which is the other one. (→ Sleep 01: Sleep and muscle)

How to use this branch

The practical summary is short and it is the same in all four entries.

Set the direction from the evidence and the precision from your own measurements. Enough protein, creatine if you want it, an intake derived from a calculator, and a rate of change chosen deliberately rather than by impatience. Then read your own trend over weeks and correct from what it tells you, because your trend is the only number in this branch that was measured on you.

Everything else here is a population estimate wearing a decimal point.

The corollary is worth stating because it runs against how nutrition is usually discussed. The effort most people put into refining these numbers would be better spent making them consistent. A protein target hit five days in seven is not a protein target. A calorie figure that changes every time you recalculate is not a measurement. The branch's evidence supports a small number of unglamorous decisions held steady for months, and offers very little to anyone looking for a decisive one.