I have done one proper fat-loss run: running every day, halving the rice and noodles, six weeks of it. My weight barely moved and I felt flat all day. What I had never worked out was how many calories my body burns just staying alive; my idea of eating less was pure guesswork. Only after I pinned down BMR (basal metabolic rate) and TDEE (total daily energy expenditure) did the whole thing become manageable. A note on units first: the calories in the title are the large calories on food labels, where 1 large calorie = 1 kilocalorie = 1 kcal, and this article uses kcal throughout.
1. BMR is the foundation, but its share of total burn depends on the convention
BMR is the heat you produce at rest, awake and fasted, just to keep your heart beating, your breathing, body temperature and cell repair going. Think of it as the standing charge on the electricity bill even when you lie still. How much of total expenditure it accounts for depends on which convention you use: with the 1.2-1.9 multipliers in section 3, the sedentary setting of 1.2 puts BMR at about 1 ÷ 1.2 ≈ 83%; the often-quoted "60%-70%" belongs to another convention and back-calculates to an activity multiplier of 1 ÷ 0.7 ≈ 1.43 up to 1 ÷ 0.6 ≈ 1.67. The two cannot both hold. I use my own band: 1 ÷ multiplier.
On top of BMR sits TDEE, which adds walking, work and workouts. The relationship:
- TDEE = BMR × activity multiplier: what you actually burn in a day.
- Fat loss needs intake < TDEE; muscle gain usually means intake ≥ TDEE with training. Set the deficit or surplus against TDEE, not against BMR.
2. How far apart the two formulas are, and making the numbers add up on a calculator
Mifflin-St Jeor (1990): for men, 10×weight + 6.25×height − 5×age + 5; for women the last term becomes −161. Harris-Benedict dates from 1919, and what calculators usually run today is the Roza and Shizgal revision from 1984: for women, 447.593 + 9.247×weight + 3.098×height − 4.330×age; for men, 88.362 + 13.397×weight + 4.799×height − 5.677×age. Weight in kg, height in cm, age in years. The two men's terms are often misquoted as 5.003 and 5.755, which lifts the whole line by 33 kcal; my first draft copied exactly that pair.
Here are two hypothetical examples, both columns rounded to 0.01, so the difference is the two displayed numbers subtracted directly:
| Subject (hypothetical) | Mifflin-St Jeor (1990) | Harris-Benedict revision (1984) | Difference |
|---|---|---|---|
| Woman, 28 / 165cm / 62kg | 1350.25 | 1410.84 | 60.59 |
| Man, 32 / 175cm / 80kg | 1738.75 | 1818.28 | 79.53 |
Spelling out the woman's row: 10×62 + 6.25×165 − 5×28 − 161 = 620 + 1031.25 − 140 − 161 = 1350.25; 447.593 + 573.314 + 511.17 − 121.24 = 1410.837, shown as 1410.84; 1410.84 − 1350.25 = 60.59. The man's row works the same way: 800 + 1093.75 − 160 + 5 = 1738.75; 88.362 + 1071.76 + 839.825 − 181.664 = 1818.283, shown as 1818.28; 1818.28 − 1738.75 = 79.53. The man's row diverges more because Mifflin weights body weight at 10 and the revision at 13.397.
Spread over a week, 60.59 kcal becomes 60.59 × 7 ≈ 424 kcal, enough to turn "I ate exactly the numbers and still did not lose weight" into something you cannot explain. So I take the lower of the two formulas as my reference floor. That is a rule of thumb, not something any standard requires.
I check the arithmetic with the BMR Calculator, which runs both formulas with 4.799 for the men's height term and −5.677 for age. Only when the two sides match do I carry on.
3. The step most people get wrong: two competing multiplier systems
The set 1.2 / 1.375 / 1.55 / 1.725 / 1.9 is not a national standard. It comes from the English-language habit of activity factors used alongside Harris-Benedict, and most Chinese content copies it wholesale. China's Reference Intakes for Dietary Nutrients takes another route: PAL (physical activity level) = TEE ÷ BEE, with adults banded as light 1.50, moderate 1.75, heavy 2.00. The first multiplies BMR; the second divides BEE (basal energy expenditure). The two definitions cannot be mixed, and switching conventions means switching back to the matching denominator.
| Activity level (conventional bands) | Multiplier | Typical description |
|---|---|---|
| Sedentary | 1.2 | Almost no exercise, sitting all day |
| Lightly active | 1.375 | Light exercise 1-3 days a week |
| Moderately active | 1.55 | Moderate intensity 3-5 days a week |
| Very active | 1.725 | Training 6-7 days a week |
| Extremely active | 1.9 | Manual labour or two sessions a day |
Almost everyone overestimates at this step: sedentary people often feel that training three times a week should count as moderate, but a workout burns less than they imagine. Take the same 28-year-old woman with BMR rounded to 1350: 1350 × 1.375 = 1856.25, about 1856, and that is her reference burn for a day, not 1350. Planning food around 1350 quietly leaves out 1856 − 1350 = 506 kcal.
4. Turning the number into what to eat each day
- A deficit of 300-500 kcal a day, at roughly 7700 kcal per kg of fat: 300 × 7 ÷ 7700 = 0.27 kg a week and 500 × 7 ÷ 7700 = 0.45 kg a week, so the upper end stays under 0.5 kg.
- Do not force the cut on the sedentary band: 1350 × 1.2 = 1620, only 1620 − 1350 = 270 kcal above BMR. This band can give up 270 at most; cutting 500 pushes intake to 1120, which is 230 kcal below what she burns lying still.
- Do not treat a population range as an individual floor: China's 2024 dietary guide for adults with obesity recommends 1200-1500 kcal a day for men and 1000-1200 kcal for women. 1120 sits inside the women's band and is still below this example's BMR. A range describes a population; an individual floor has to go back to your own BMR and TDEE. Such plans also need professional supervision; they are not a licence to sit at the floor long term.
- If the weight curve barely moves for two consecutive weeks (weighed at the same time and in the same state), adjust once by ±100-150 kcal. Do not cut a large amount at once.
- Strength training and protein decide whether you lose fat or muscle; weight also moves with water, sodium and the menstrual cycle, and a 1-2 kg swing within a day is normal. Watch the trend, not the single day.
5. Common questions
Q1: Do people with more muscle really have a higher BMR?
At the same body weight, lean mass is more metabolically active than fat tissue, so body composition affects basal metabolism. That is also the usual explanation for hitting a plateau after doing only cardio and losing muscle.
Q2: How accurate is the number, and how big is the error?
The empirical error is usually ±10%-15%; the result is a starting point, not a measurement. Real calibration comes from two weeks of weight and waist measurements, then adjusting the multiplier in the opposite direction.
Q3: Does long-term dieting lower BMR?
With a sustained large deficit the body adapts metabolically, lowering expenditure and becoming more efficient. That is one common mechanism behind rebound after dieting. Keep the deficit small rather than large.
Q4: What if the scale and the body-fat percentage disagree?
Both formulas estimate from height, weight, age and sex only, with no body composition. Home body-fat scales use bioelectrical impedance and are affected by drinking, eating and skin condition; read the trend, not a single absolute value. For a rough band, use the BMI Calculator.
These formulas produce statistical estimates. They are not a substitute for medical or nutrition advice; with a metabolic condition, a thyroid problem, pregnancy or breastfeeding, or medication, follow your doctor and a registered dietitian.
Looking back, those six stalled weeks were not a failure of effort. I was treating a problem that guessing cannot solve as a problem that willpower can.