What Is the Thermic Effect of Food, and How Much Does Protein Change It?
The thermic effect of food is the energy spent digesting a meal, about 10% of daily expenditure. Protein costs 20 to 30% of its calories, so 30% protein instead of 15% on 2,000 kcal adds roughly 50 to 67 kcal.
Key takeaways
- NASEM puts the thermic effect of food at about 10% of daily energy expenditure, and Westerterp's review gives 5 to 15% for a mixed diet.
- Reported values are 20 to 30% of calories for protein, 5 to 10% for carbohydrate, 0 to 3% for fat and 10 to 30% for alcohol.
- Moving protein from 15% to 30% of a 2,000 kcal day adds about 50 to 67 kcal of digestion cost, using the ends of those ranges.
- NASEM's total energy expenditure already includes the thermic effect of food, so adding another 10% to a calculator result double counts about 200 kcal on a 2,000 kcal day.
- In a trial of 17 adults, a whole-food sandwich meal used 137 kcal in the following 5 to 6 hours against 73.1 kcal for a processed one with the same energy.
The thermic effect of food (TEF) is the energy your body spends digesting and absorbing what you eat, and the National Academies put it at about 10% of daily energy expenditure. Protein costs the most. Reported values are 20 to 30% of its own calories, against 5 to 10% for carbohydrate and 0 to 3% for fat. On a 2,000 kcal day, moving protein from 15% to 30% of calories changes that cost by tens of calories, not hundreds.
What is the thermic effect of food?
It is the rise in energy use after a meal, counted above your fasting rate and expressed as a share of the meal’s calories. Westerterp’s 2004 review defines it exactly that way. NASEM, the 2023 energy report from the US National Academies, lists it as one of three parts of daily energy expenditure. The other two are resting metabolism and physical activity. For a whole day on mixed food, with intake matching expenditure, the review puts it at about 10% of the energy you eat. Its abstract gives a wider span, 5 to 15%. On a 2,000 kcal day, 10% is 200 kcal.
You can’t measure your own. Within the same person, repeat tests with a ventilated hood vary by 6 to 30%, and 24-hour chamber calculations by 43 to 48%. If a lab can’t pin the number down for one volunteer, a kitchen won’t do better. The wider picture of resting rate, digestion and movement sits in what is metabolism.
Try this: multiply your usual daily calories by 0.10. That’s the ballpark, and the yardstick for everything below.
How many calories do you burn digesting protein, carbs and fat?
Protein uses 20 to 30% of its own calories, carbohydrate 5 to 10%, fat 0 to 3% and alcohol 10 to 30%. NASEM prints those ranges, taken from Westerterp’s review, and the table turns them into amounts per 100 kcal eaten. His explanation for the order is the energy needed for the first steps of metabolism and for storing what isn’t used straight away.
| Nutrient | TEF, % of its own calories | Used per 100 kcal eaten | Left per 100 kcal eaten |
|---|---|---|---|
| Protein | 20–30% | 20–30 kcal | 70–80 kcal |
| Carbohydrate | 5–10% | 5–10 kcal | 90–95 kcal |
| Fat | 0–3% | 0–3 kcal | 97–100 kcal |
| Alcohol | 10–30% | 10–30 kcal | 70–90 kcal |
Read the middle column as a price list. Eat 100 kcal of protein and digesting it uses 20 to 30 kcal. Eat 100 kcal of fat and the bill is 0 to 3 kcal. A serving with 30 g of protein is 120 kcal at the 4 kcal per gram explained in protein calories per gram. Its digestion cost is 24 to 36 kcal.
Measured meals tend to land lower. In Westerterp’s table, most studies found that less than 10% of a meal’s energy was used, measured over 4 to 5.5 hours after eating. In one of them, swapping 20% of a meal’s energy for protein showed a tendency (not a significant rise) for five-hour TEF to go from 7.1 to 8.3%. That’s 1.2 points. At the textbook ranges, protein replacing carbohydrate on a 20% swap would add 2 to 5 points. The whole-day chamber studies came out near 10% once their baseline was matched to the others. Westerterp ties the difference to how long, and against what baseline, each study measured.
Alcohol’s row is there for completeness. Westerterp notes that alcohol energy is largely additive to the usual diet instead of replacing food.
Pick a meal you eat often and multiply its calories by 0.10. For a 600 kcal plate, that’s about 60 kcal, and the true figure for you sits somewhere around it.
Does protein burn more calories? A 2,000 kcal day at 15% and 30%
Slightly. Doubling protein from 15% to 30% of a 2,000 kcal day raises the digestion cost by about 50 to 67 kcal, using the ranges NASEM reports.
I ran the day through the ends of every range, since one pair of percentages hides how loose they are. Day A has 15% protein, 50% carbohydrate and 35% fat. Day B: 30% protein, 40% carbohydrate and 30% fat. Both total 2,000 kcal. The formula is:
TEF = protein kcal × (0.20 to 0.30) + carbohydrate kcal × (0.05 to 0.10) + fat kcal × (0 to 0.03)
| Day A: 15% protein | Day B: 30% protein | |
|---|---|---|
| Protein | 300 kcal (75 g) | 600 kcal (150 g) |
| Carbohydrate | 1,000 kcal | 800 kcal |
| Fat | 700 kcal | 600 kcal |
| TEF at the low end of each range | 110 kcal | 160 kcal |
| TEF at the high end of each range | 211 kcal | 278 kcal |
| TEF as a share of 2,000 kcal | 5.5–10.6% | 8.0–13.9% |
The gap is 50 kcal at the low ends (160 minus 110) and 67 kcal at the high ends (278 minus 211). That’s 2.5 to 3.4% of the day. Both days sit inside the 5 to 15% span from the review, which is a decent check that the method isn’t off.
Look at the spread, too. Day A alone runs from 110 to 211 kcal, a range of 101, and Day B from 160 to 278, a range of 118. Each day’s own uncertainty is larger than the 50 to 67 kcal gap between the days. The ranges are looser than the effect you’d be chasing.
Day B holds 150 g of protein, double Day A’s 75 g and well beyond what many people plan for. Treat it as a stress test of the arithmetic, not a target. How much protein should I eat a day covers the amounts.
A measured version exists. In a 36-hour chamber study of eight women, a diet with 30% of energy from protein produced 1,295 kJ of TEF over 24 hours, against 931 kJ on a diet with 10% protein. Divide by 4.184 and that’s 310 and 223 kcal, a gap of 87 kcal. Total 24-hour energy expenditure did not differ significantly. The high-protein diet was 60% carbohydrate, 10% fat and 30% protein, and the other was 30% carbohydrate, 60% fat and 10% protein, so the 87 kcal isn’t protein alone, and eight people is a small group.
Now set those gaps against the error of the estimate itself. NASEM’s adult equations miss by a typical 246 kcal a day for women and 339 for men, as root mean squared error. A difference of 50 to 87 kcal sits well inside that miss.
Write down yesterday’s protein in grams, multiply by 4, and divide by your total calories. That’s your protein share. Compare it with 15 and 30%, and notice how small the gap between them is next to a margin of error.
Why can’t you add the thermic effect of food to your daily calories burned?
Because it’s already in the number. NASEM defines total energy expenditure as resting metabolic rate plus the thermic effect of food plus physical activity, so any estimate of the total contains TEF.
The equations behind those estimates were fit to doubly labeled water data. The method follows carbon dioxide output from the whole body over days or weeks, and NASEM describes it as the benchmark standard for measuring total energy expenditure in free-living people. It can’t pick digestion out of the total, so digestion stays inside. The adult energy requirement is built on those total-expenditure equations. Activity is what remains after subtracting resting rate and TEF, and TEF is generally taken as about 10% of the total.
Add it again and you count it twice. Say your estimate is 2,000 kcal. Add 10% for TEF and you’re working from 2,200, which is 200 kcal a day of burn counted twice. That is not far below the 246 kcal typical miss for women. The same goes for a protein bonus: don’t tack the 50 to 67 kcal from the table onto your total because you ate more protein.
The mirror mistake happens at the food log. Don’t subtract 24 to 36 kcal from a 120 kcal protein serving because digestion will use it. Log the label as it is. The cost sits on the expenditure side, and it’s inside the estimate already. The full build-up of that estimate is in how many calories you burn a day.
What doesn’t work is a calculator box for TEF, or a plan that treats a high-protein day as a bigger budget. To see the part of your day you can actually change, the steps to calories calculator estimates what a walk adds on top of your usual day.
Does food processing change the thermic effect of food?
In one small trial, by a lot. The authors knew of no earlier test of the same kind, and they warn against stretching the result.
Sadie Barr and Jonathan Wright (2010) gave 17 healthy adults two cheese sandwich meals on separate days, in random order. One was multigrain bread with cheddar, the other white bread with a processed cheese product. Each person ate the same size meal both times, 600 or 800 kcal, with the same share of bread and cheese. Over the next 5 to 6 hours, the whole-food meal used 137 kcal, which is 19.9% of its energy. The processed meal used 73.1 kcal, or 10.7%. The difference is 64 kcal (137 minus 73.1 is 63.9) on one meal.
Three things limit it. The whole-food meal had 20% of its energy from protein against 15%, and about three times the fiber, so processing isn’t the only difference. Ratings of fullness did not differ between the meals. And it was one meal pair, so the authors say the finding applies to that type of meal and needs follow-up.
I’d add a fourth. A 19.9% reading is well above the under-10% values most studies in Westerterp’s table found over similar windows. Read this as one small result: a direction worth watching, a size still open. If you log a sandwich, log the label.
What should you do about the thermic effect of food?
Leave it alone and count protein in grams.
Protein is the one part of this worth writing down, and the reason is fullness more than digestion. Halton and Hu’s systematic review of randomized studies found convincing evidence that higher protein intake raises thermogenesis and satiety compared with lower protein diets. The weight of evidence, they wrote, suggests high-protein meals lead to less energy eaten later. Results on weight loss and fat loss haven’t been consistent, and they called for longer, more rigorous studies. Westerterp’s review points the same way: he describes protein’s main effect on energy balance as satiety related to TEF.
The useful number is the one on your plate: the protein grams you already wrote down, set against your own target. That’s a decision about food, and no percentage in this piece changes it.
Two limits apply. NASEM lists age, activity, and a meal’s size and composition as factors in TEF, and a 2019 review it cites found one study in which active men had about 45% higher TEF than inactive men. Your figure won’t sit at the middle of any table. If a scale gave you a metabolic age, that label is about resting burn, not digestion. And with kidney disease or a medical diet, your protein target is a conversation for your doctor, not a calculator.
TEF is real and small, and it’s inside the number you started with. A 2,000 kcal day at 15% or 30% protein differs by tens of calories, less than the typical miss in the equation that gave you 2,000. Keep the percentages for curiosity. For the plan, log the food and the protein grams, and spend your attention on choices that move a day by hundreds. I find that easier than it sounds, because the grams are on the label and the TEF isn’t.
Questions and answers
Does protein burn more calories than carbs or fat?
Slightly. Reported thermic effect values are 20 to 30% of calories for protein, 5 to 10% for carbohydrate and 0 to 3% for fat. On a 2,000 kcal day, raising protein from 15% to 30% of calories adds about 50 to 67 kcal of digestion cost at the ends of those ranges.
How many calories do you burn digesting protein?
About 20 to 30 kcal for every 100 kcal of protein you eat, according to the ranges NASEM reports from Westerterp's 2004 review. A serving with 30 g of protein is 120 kcal, so its digestion cost is roughly 24 to 36 kcal. Measured meals often come out lower than the textbook range.
Should I add the thermic effect of food to my daily calories burned?
No. NASEM defines total daily energy expenditure as resting metabolism plus the thermic effect of food plus physical activity, so an estimate of the total already contains it. Adding 10% on top of a 2,000 kcal estimate would count about 200 kcal twice.
Is the thermic effect of food 10% of the calories I eat?
About 10% for a whole day of mixed food when intake matches expenditure, per Westerterp's 2004 review. Single meals measured over 4 to 5.5 hours often show less than 10%. Age, activity, meal size and meal composition all move the figure.
Sources
- Dietary Reference Intakes for Energy (2023), Chapter 4: Factors Affecting Energy Expenditure and Requirements, National Academies of Sciences, Engineering, and Medicine
- Calories on the Nutrition Facts label (2,000 calories a day as a general guide), U.S. Food and Drug Administration
- Westerterp, Diet induced thermogenesis, Nutrition & Metabolism 2004, Nutrition & Metabolism
- Westerterp, Wilson and Rolland, Diet induced thermogenesis measured over 24 h in a respiration chamber: effect of diet composition, International Journal of Obesity 1999, International Journal of Obesity
- Dietary Reference Intakes for Energy (2023), Chapter 5: Development of Prediction Equations for Estimated Energy Requirements, National Academies of Sciences, Engineering, and Medicine
- Barr and Wright, Postprandial energy expenditure in whole-food and processed-food meals: implications for daily energy expenditure, Food & Nutrition Research 2010, Food & Nutrition Research
- Halton and Hu, The effects of high protein diets on thermogenesis, satiety and weight loss: a critical review, Journal of the American College of Nutrition 2004, Journal of the American College of Nutrition



