Macronutrient Splits: Protein, Fat, Carbohydrate

A macronutrient split is a set of three numbers: grams of protein, fat, and carbohydrate. The split is not a property of a food or a diet. It is a property of a person’s body weight, energy target, and the order in which the three are assigned. Fixed percentage splits — 40/30/30, 50/30/20 — are common in print because they are easy to typeset, not because they match the arithmetic of human bodies at different sizes.

The arithmetic that does match is sequential. Protein is set first from body weight. Fat is set second, to a floor. Carbohydrate takes the remainder. That ordering is the page. Everything below is the arithmetic, the sources of the numbers, and the places where the numbers are softer than they look.

Where the numbers come from

Energy figures on food labels and in databases are not measured for every food. They are calculated from Atwater factors, published by W.O. Atwater in 1899 and refined in USDA Handbook 74 (1973). The general factors are 4 kcal/g for protein, 9 kcal/g for fat, and 4 kcal/g for carbohydrate. These are averages across mixed diets, not constants. For specific foods the factors differ: protein in meat averages nearer 4.3 kcal/g, carbohydrate in vegetables nearer 3.5 kcal/g because of fiber. The FDA permits the general factors on labels.

The label itself carries tolerances. FDA guidance allows a food to be labeled as having 0 g fat if it contains less than 0.5 g per serving, and 0 kcal if it contains fewer than 5 kcal per serving. A cooking spray can print “0 calories” while delivering 1–2 kcal per spray because the serving size rounds to zero. A label can also round grams to the nearest 0.5 g and calories to the nearest 10 kcal. The arithmetic to recover the real figure is to multiply grams by Atwater factors and compare to the printed calories.

Protein requirements are set from body weight, not from a percentage of energy. The RDA is 0.8 g/kg for adults (Institute of Medicine, 2005). That is a minimum to prevent deficiency in sedentary people. For active adults, position stands from the American Dietetic Association, Dietitians of Canada, and the American College of Sports Medicine (2009) give 1.2–1.7 g/kg. For resistance-trained individuals, 1.6 g/kg is a common target, with a range of 1.4–2.0 g/kg.

Fat has a floor. The Institute of Medicine’s Acceptable Macronutrient Distribution Range for fat is 20–35% of energy. In grams per kilogram, that floor is roughly 0.6–0.8 g/kg. Below that, fat-soluble vitamin absorption and hormonal function become harder to maintain. The floor is not a target; it is the point below which the arithmetic stops being safe to run.

Carbohydrate is the remainder. The RDA for carbohydrate is 130 g/day for adults, set to fuel the brain. That is a floor for the diet, not a target for the split.

The table: what changes the figure

The figure that changes most is the percentage that results from a fixed gram assignment. The table below shows the condition under which the percentage changes.

The last two rows are the same because a percentage split does not know body weight. At 100 kg, 200 g protein is 2.0 g/kg, above the range. At 50 kg, 200 g protein is 4.0 g/kg, far above any evidence-based target and a dose that displaces fat and carbohydrate. The 40/30/30 split is not wrong as arithmetic; it is wrong as a default because it assigns grams without reference to the person.

Worked calculation: two body weights

The calculation below uses Mifflin-St Jeor for resting metabolic rate (Mifflin et al., 1990), an activity multiplier, and the sequential assignment. The standard error of Mifflin-St Jeor is about 10% in adults, so the energy target carries that band.

Step 1: Resting metabolic rate (Mifflin-St Jeor)
  Men:   RMR = 10*kg + 6.25*cm - 5*age + 5
  Women: RMR = 10*kg + 6.25*cm - 5*age - 161

Step 2: Maintenance energy
  Sedentary: RMR * 1.2
  Light activity: RMR * 1.375
  Moderate: RMR * 1.55
  Very active: RMR * 1.725

Step 3: Protein
  Protein_g = kg * g_per_kg

Step 4: Fat floor
  Fat_g = kg * 0.6 to 0.8

Step 5: Carbohydrate remainder
  Carb_g = (kcal - protein_g*4 - fat_g*9) / 4

Example A: 50 kg woman, 160 cm, 30 years, moderate activity
  RMR = 10*50 + 6.25*160 - 5*30 - 161 = 500 + 1000 - 150 - 161 = 1189 kcal
  Maintenance = 1189 * 1.55 = 1843 kcal
  Protein = 50 * 1.6 = 80 g (320 kcal)
  Fat = 50 * 0.8 = 40 g (360 kcal)
  Carb = (1843 - 320 - 360) / 4 = 1163 / 4 = 291 g
  Split: 80 g protein (17%), 40 g fat (20%), 291 g carb (63%)

Example B: 100 kg man, 180 cm, 40 years, moderate activity
  RMR = 10*100 + 6.25*180 - 5*40 + 5 = 1000 + 1125 - 200 + 5 = 1930 kcal
  Maintenance = 1930 * 1.55 = 2992 kcal
  Protein = 100 * 1.6 = 160 g (640 kcal)
  Fat = 100 * 0.8 = 80 g (720 kcal)
  Carb = (2992 - 640 - 720) / 4 = 1632 / 4 = 408 g
  Split: 160 g protein (21%), 80 g fat (24%), 408 g carb (55%)

The two splits are close in percentage terms — 17/20/63 and 21/24/55 — but the gram amounts differ by a factor of two. A fixed 40/30/30 split would give both people 200 g protein. For the 50 kg woman that is 4.0 g/kg, a dose with no evidence base and one that leaves only 67 g fat and 150 g carbohydrate at 2,000 kcal. The percentage looks the same; the biology does not.

What people get wrong

The common error is to treat a macro split as a property of a diet rather than a property of a person. The mistake is natural because nutrition communication often presents splits as ratios — 40/30/30, 50/25/25 — and ratios are scale-free. A ratio of 40:30:30 is the same whether the plate is small or large. But the body is not scale-free. Protein needs scale with body weight, fat needs scale with body weight, and energy needs scale with body weight and height. The ratio is the output of the calculation, not the input.

A second error is to read a label’s zero as zero. A food with 0.4 g fat per serving can print 0 g fat. A food with 4 kcal per serving can print 0 kcal. The arithmetic to recover the real figure is to multiply the grams by the Atwater factors and compare to the printed calories. If the printed calories are lower than the sum of the macros times their factors, the label is rounding down. Weighing the food and using a database with per-100-g values avoids the serving-size rounding entirely.

A third error is to treat the fat floor as a target. The floor is 0.6–0.8 g/kg. Going below it is not a way to create a larger carbohydrate remainder; it is the point at which the arithmetic stops being safe to run. The IOM’s AMDR for fat is 20–35% of energy, and the floor in grams per kilogram is a separate check.

What to do differently

Weigh food rather than distrust labels. A kitchen scale and a per-100-g database give numbers that do not depend on the manufacturer’s rounding. Set protein from body weight using 1.2–1.7 g/kg for active adults and 0.8 g/kg as the sedentary RDA. Set fat to 0.6–0.8 g/kg. Let carbohydrate take the remainder. Recalculate when body weight changes by more than a few kilograms, because the gram assignments move with it. The energy target carries a 10% standard error from Mifflin-St Jeor, so treat the final carbohydrate number as a band, not a point.

This page is not medical advice. For disordered eating support, contact the National Eating Disorders Association Helpline at 1-800-931-2237 (US) or Beat at 0808 801 0677 (UK).