How to set your macros

Why protein is set first, and by body weight

Protein goes first because it is the only one of the three whose requirement is anchored to something that does not move when your calorie budget moves. The evidence base expresses it per kilogram of body weight: the Recommended Dietary Allowance at 0.8 grams per kilogram, the training meta-analysis plateau at about 1.6, the upper bound of that confidence interval near 2.2. None of those figures is a percentage of energy, and converting them into one destroys the information.

The conversion problem is easy to demonstrate. An 80 kilogram person eating 1.8 grams per kilogram needs 144 grams of protein a day, which is 576 kilocalories. On a 3,000 kilocalorie maintenance budget that is nineteen per cent of energy. On a 1,900 kilocalorie cutting budget it is thirty per cent. It is the same requirement, the same grams, the same food. Anyone who set thirty per cent as a rule and then raised their calories would silently start eating more protein than they needed; anyone who set nineteen per cent and then cut would silently start eating less at exactly the point where more was called for.

That last point is the substantive one. The protein requirement rises slightly in a deficit, because protein is doing the work of defending lean tissue while energy is short. A percentage-based target does the opposite of what is needed, falling in absolute terms exactly when the need increases. Setting grams per kilogram and letting the percentage fall where it may avoids the trap entirely.

The targets this site uses are 1.6 grams per kilogram for general health, 1.8 when training regularly, and 2.2 while cutting. Those are not arbitrary; they map onto the meta-analytic plateau and its confidence interval, with the higher figure reserved for the condition where the evidence for higher intakes is strongest. For someone carrying a large amount of fat mass, scaling to lean body mass or to a target weight is the more sensible version of the same calculation.

Why fat has a floor

Fat is the macronutrient with a genuine lower bound, and it has two independent reasons for one. The first is endocrine. A systematic review and meta-analysis of intervention studies by Whittaker and Wu examined what happened to testosterone when men moved from higher-fat to low-fat diets, and reported a small to moderate decrease in total testosterone on the low-fat diets, an effect that reached statistical significance without being dramatic. The honest summary is a real but modest signal, and it is not something a macro ratio will tell you is happening.

The second reason is absorption. Vitamins A, D, E and K are fat-soluble, as are carotenoids, and they require dietary fat present in the same meal to be absorbed efficiently. Controlled work by Brown and colleagues, measuring carotenoid absorption from salads eaten with full-fat and with fat-reduced dressings, found bioavailability higher with the full-fat dressing. A very low fat diet can therefore be nutritionally adequate on paper, in the sense that the vitamins were present in the food, while delivering less of them than the numbers suggest.

The floor used on this site is the higher of 0.8 grams per kilogram of body weight or twenty per cent of the calorie budget, and taking the higher of two candidates is deliberate. A percentage alone fails a large person eating very little, because twenty per cent of a small budget is a small absolute quantity of fat for a large body. A grams-per-kilogram figure alone fails a small person eating a great deal, because it can leave the diet at eight or ten per cent fat by energy. Whichever candidate produces more fat wins, so both failure modes are covered.

None of that argues for a high-fat diet, and the floor is a floor rather than a target. Eating well above it is fine if that is the diet you prefer and the calories permit it. The point is only that fat is the macronutrient where the downside of going too low is physiological rather than a matter of preference, and where the consequences are slow enough that they will not be obvious for months.

Carbohydrate is the remainder, and what the remainder does

Once protein and fat are set, the calories left over go to carbohydrate, divided by four. Calling it the remainder sounds like a demotion, and in a sense it is: carbohydrate is the macronutrient with no established minimum requirement in the way protein has an RDA and fat has functional roles, because the body can manufacture glucose from other substrates. That is the honest position and it is why carbohydrate takes the leftovers rather than being set first.

It is also the macronutrient that does most of the visible work in training. Glycogen is the dominant fuel for repeated high-intensity efforts, and a depleted athlete produces less work per session across weeks. If the remainder is very small because protein and fat have been set aggressively, the first symptom is not hunger, it is heavy sessions, missed repetitions and a general reluctance to train. That symptom is easy to misattribute to the deficit itself rather than to its composition.

The second thing the remainder does is carry fibre. The Institute of Medicine sets an Adequate Intake for total fibre at 14 grams per 1,000 kilocalories, which is 38 grams a day for men and 25 for women on typical intakes, and almost all of that arrives inside carbohydrate foods. A very low carbohydrate remainder makes hitting fibre a deliberate exercise rather than a by-product of eating vegetables, fruit, legumes and whole grains.

The third and least discussed function is adherence. Carbohydrate foods are, for most people in most cultures, the foods that make meals feel like meals. A plan that leaves a reasonable remainder is one that can accommodate rice with dinner and bread at lunch, and a plan that accommodates those things is one that survives past week six. That is not a physiological argument, but the physiological superiority of a diet nobody follows is zero.

The same person, in a deficit and in a surplus

The clearest way to see what the method does is to run one person through it twice. Take an 82 kilogram adult with a calibrated maintenance of 2,700 kilocalories, first cutting at a twenty per cent deficit and then gaining at a ten per cent surplus.

Cutting, the budget is 2,160 kilocalories. Protein at 2.2 grams per kilogram gives 180 grams, which is 720 kilocalories. The fat floor takes the higher of 82 times 0.8, which is 66 grams, and twenty per cent of 2,160 divided by nine, which is 48 grams; 66 grams wins, at 594 kilocalories. That leaves 846 kilocalories for carbohydrate, or about 212 grams. Expressed as a ratio, it is roughly 33 per cent protein, 27 per cent fat and 39 per cent carbohydrate.

Gaining, the budget is 2,970 kilocalories. Protein drops to 1.8 grams per kilogram, giving 148 grams and 592 kilocalories, because the deficit-specific argument for the higher figure no longer applies. The fat floor candidates are 66 grams from body weight and 66 grams from twenty per cent of 2,970, which happen to coincide, so fat is 66 grams and 594 kilocalories. The remainder is 1,784 kilocalories, or about 446 grams of carbohydrate. As a ratio, that is roughly 20 per cent protein, 20 per cent fat and 60 per cent carbohydrate.

Notice what happened. Nobody chose a ratio at any point. Two rules were applied in a fixed order, and the ratio moved from 33/27/39 to 20/20/60 as a consequence. Carbohydrate more than doubled in grams, which is exactly the change that makes a surplus feel different in training. Anyone who had instead fixed a ratio at the start would have ended up either eating far more protein than necessary in the surplus, or far less than the evidence supports in the deficit.

Why ratios matter far less than energy and protein

The argument about macronutrient ratios has consumed more attention than almost any other question in nutrition, and the controlled evidence is remarkably consistent about the answer. The POUNDS LOST trial by Sacks and colleagues randomised overweight adults to reduced-calorie diets with four different combinations of fat, protein and carbohydrate, followed them for two years, and found comparable weight loss across the groups. The composition varied substantially; the outcomes did not.

A meta-analysis by Johnston and colleagues, comparing named diet programmes against one another, reached a similar conclusion from a different direction, finding differences between low-carbohydrate and low-fat approaches that were small relative to the difference between following a diet and not following one. And tightly controlled metabolic ward work by Hall and colleagues, in which every gram of food was supplied and every variable measured, compared calorie-matched fat restriction with carbohydrate restriction and did not find the advantage for carbohydrate restriction that popular accounts predict.

The sensible reading is not that composition is irrelevant. It is that composition operates a long way down the list, below total energy, below protein, and below whether the plan is one you can actually follow for the required number of months. Once protein is adequate and fat is above its floor, the remaining split is a preference variable, and choosing it by preference is the choice most likely to produce adherence.

This has a liberating practical consequence. If you find low-carbohydrate eating easy and satisfying, the evidence does not object. If you train hard and want a large carbohydrate remainder, the evidence does not object to that either. What the evidence does object to is spending months optimising the third-order variable while the first-order one, total energy, is being estimated by eye.

What tracking accuracy really is

Any macro plan is only as good as the measurement of what was actually eaten, and that measurement is considerably worse than most people assume. It is worth separating the sources of error, because they are of quite different sizes and only some of them are fixable.

The largest and most persistent is under-reporting. Classic work by Lichtman and colleagues studied people who reported an inability to lose weight despite low reported intakes, and found actual energy intake substantially higher than reported and actual energy expenditure close to predicted. The effect was not dishonesty. It was unrecorded items, underestimated portions, and days that did not get logged, and it is a systematic bias rather than random noise, which means it does not average out.

Commercially prepared food is the second source. Urban and colleagues measured the energy content of reduced-energy commercially prepared foods against the values stated for them, and found measured energy averaged eighteen per cent above stated values for restaurant dishes and eight per cent above for supermarket frozen meals, with individual restaurant items running considerably further out than that. A restaurant meal is not a weighed meal, and the oil it was cooked in was not measured by anyone.

Then there are the smaller, structural errors. Packaged food labels carry regulatory tolerances rather than exact values, so the number on the packet is a permitted approximation of the contents. Food composition databases give averages for foods that vary in fat content, ripeness and cut. And crowd-sourced entries in tracking apps are frequently wrong in ways nobody has checked. None of these is large on its own; together they are several per cent.

The right response to all of this is proportion rather than despair. Weighing solid food on scales removes most of the fixable error at a cost of a few seconds per meal. Measuring cooking oil rather than pouring it removes one of the largest single omissions, because a generous unweighed glug is easily 150 kilocalories. And, critically, consistent error is much less damaging than random error: if you log the same way every day, your intake number may be systematically off by five per cent, but the calibration procedure against your own weight trend absorbs a constant bias entirely. What it cannot absorb is a logging habit that changes between weekdays and weekends.

Setting yours, in order

The whole method fits in five steps, and the order is the part that carries the value. Work out maintenance and apply a percentage adjustment to get the calorie budget. Set protein in grams per kilogram of body weight. Set fat at the higher of the two floor candidates. Give the rest to carbohydrate. Then check the result against the things a spreadsheet cannot see: whether the fibre target is reachable, whether the carbohydrate remainder supports the training you have planned, and whether the whole thing is a diet you would be willing to eat in eight weeks' time.

Two adjustments are worth making up front rather than later. If you are carrying a large amount of fat mass, scale protein to lean body mass or to a target weight rather than to current weight, because fat mass does not need defending and scaling to it produces impractical numbers. And if you train early in the morning or twice a day, weight the carbohydrate remainder towards the meals around training rather than distributing it evenly, which is the one piece of nutrient timing that reliably changes how sessions feel.

Finally, treat the targets as a band rather than a line. Landing within about ten grams of the protein target and within a reasonable distance of the calorie budget, most days, produces the result. Hitting three numbers exactly, every day, produces a relationship with food that very few people sustain and that nothing in the evidence requires.

• Start from a calibrated maintenance figure, not a calculator's first estimate.
• Apply the goal adjustment as a percentage: ten to twenty per cent down for loss, about ten per cent up for gain.
• Protein: 1.6 g/kg for general health, 1.8 g/kg when training regularly, 2.2 g/kg while cutting. Multiply by weight in kilograms.
• Fat: take the higher of 0.8 g/kg and twenty per cent of the calorie budget divided by nine.
• Carbohydrate: subtract protein and fat calories from the budget, then divide by four.
• Check fibre against 14 g per 1,000 kcal, and check that the carbohydrate remainder covers the training week you have actually planned.
• Review after three or four weeks of seven-day weight averages, and change one thing at a time.

Frequently asked questions

Put it into practice

Run your own numbers through the macros calculator, the protein calculator, the calorie deficit calculator and the calorie surplus calculator. Related reading: How much protein, How many calories to eat and TDEE explained.

Sources

1. Morton RW, Murphy KT, McKellar SR, et al. A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. Br J Sports Med 2018;52:376-84. doi.org/10.1136/bjsports-2017-097608
2. Sacks FM, Bray GA, Carey VJ, et al. Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates. N Engl J Med 2009;360:859-73. doi.org/10.1056/NEJMoa0804748
3. Johnston BC, Kanters S, Bandayrel K, et al. Comparison of weight loss among named diet programs in overweight and obese adults: a meta-analysis. JAMA 2014;312:923-33. doi.org/10.1001/jama.2014.10397
4. Hall KD, Bemis T, Brychta R, et al. Calorie for calorie, dietary fat restriction results in more body fat loss than carbohydrate restriction in people with obesity. Cell Metab 2015;22:427-36. doi.org/10.1016/j.cmet.2015.07.021
5. Whittaker J, Wu K. Low-fat diets and testosterone in men: systematic review and meta-analysis of intervention studies. J Steroid Biochem Mol Biol 2021;210:105878. doi.org/10.1016/j.jsbmb.2021.105878
6. Brown MJ, Ferruzzi MG, Nguyen ML, et al. Carotenoid bioavailability is higher from salads ingested with full-fat than with fat-reduced salad dressings as measured with electrochemical detection. Am J Clin Nutr 2004;80:396-403. doi.org/10.1093/ajcn/80.2.396
7. Lichtman SW, Pestone M, Dowling H, et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med 1992;327:1893-8. doi.org/10.1056/NEJM199212313272701
8. Urban LE, Dallal GE, Robinson LM, Ausman LM, Saltzman E, Roberts SB. The accuracy of stated energy contents of reduced-energy, commercially prepared foods. J Am Diet Assoc 2010;110:116-23. doi.org/10.1016/j.jada.2009.10.003
9. Institute of Medicine. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington DC: National Academies Press, 2005 - total fibre Adequate Intake of 14 g per 1,000 kcal. doi.org/10.17226/10490