Why the Scale Moves When Fat Does Not

A scale reading is a single measurement of total body mass at one moment. It includes fat, lean tissue, bone, and water. The water component moves by one to two kilograms within a day, and almost none of that movement is fat. This page sets out the arithmetic behind those swings, the conditions that change the figure, and how to read a weekly average when a single reading cannot detect a week’s real fat loss.

What moves the number within a day

Body mass is the sum of fat mass, fat-free mass, and the water bound to both. Fat-free mass is roughly 70–75% water by weight. Glycogen, the storage form of carbohydrate in muscle and liver, binds approximately three grams of water per gram of glycogen. A 500 g change in glycogen stores therefore carries about 1.5 kg of associated water. Sodium intake shifts extracellular fluid: a high-sodium meal can add 0.5–1.5 kg of water retention within 24 hours. The menstrual cycle adds 0.5–2 kg in the luteal phase for many people. Gut contents from a large meal or high-fibre intake can add 0.3–1 kg before elimination. Training-induced inflammation after unaccustomed exercise can add 0.5–1 kg of fluid for 24–72 hours.

These figures come from controlled feeding studies and body-composition literature. The glycogen-water ratio of roughly 1:3 is reported in classic biopsy work and is consistent across several reviews. Sodium-related water retention is documented in balance studies. Menstrual-cycle fluid shifts are measured by bioelectrical impedance and scale weight across phases. None of these are fat.

The arithmetic of a weekly average

A single scale reading cannot detect a week’s real fat loss because the noise is larger than the signal. Suppose a person’s true fat loss over a week is 300 g. That is 300 g of fat, or about 2,700 kcal at 9 kcal per gram. The daily water fluctuation is 1–2 kg. The signal-to-noise ratio is roughly 1:5. A single reading is therefore uninformative about fat loss.

A weekly average of daily weights reduces the noise. If daily weights are 70.0, 70.5, 70.2, 70.8, 70.3, 70.1, and 70.4 kg, the average is 70.33 kg. The standard deviation of those seven readings is about 0.26 kg. The standard error of the mean is 0.26 / sqrt(7) = 0.10 kg. That means the weekly average has an uncertainty of about ±0.10 kg, or 100 g. A true fat loss of 300 g is therefore detectable in the weekly average, but not in any single day.

Daily weights (kg): 70.0, 70.5, 70.2, 70.8, 70.3, 70.1, 70.4
Sum = 70.0 + 70.5 + 70.2 + 70.8 + 70.3 + 70.1 + 70.4 = 492.3
Mean = 492.3 / 7 = 70.33 kg
Deviations from mean: -0.33, 0.17, -0.13, 0.47, -0.03, -0.23, 0.07
Squared deviations: 0.109, 0.029, 0.017, 0.221, 0.001, 0.053, 0.005
Sum of squares = 0.435
Variance = 0.435 / 6 = 0.0725
Standard deviation = sqrt(0.0725) = 0.269 kg
Standard error = 0.269 / sqrt(7) = 0.102 kg

What people get wrong about water weight

The natural mistake is to treat the scale as a fat meter. It is not. The scale measures total mass, and water is the largest rapidly changing component. A person who gains 2 kg overnight has not gained 2 kg of fat. To gain 2 kg of fat, they would need to consume about 18,000 kcal above maintenance (2,000 g × 9 kcal/g). That is not a single day’s eating. The 2 kg is water, glycogen, sodium, and gut contents. The mistake is natural because the scale is the only feedback available, and it responds to water within hours. The correction is to weigh daily and average weekly.

Conditions that change the figure

The table below lists the condition, the typical change in scale weight, and the source or mechanism.

Why maintenance falls as mass falls

Energy balance is not a straight line. Maintenance energy expenditure falls as body mass falls. The Mifflin-St Jeor equation gives resting metabolic rate (RMR) in kcal/day:

Men: RMR = 10 × weight (kg) + 6.25 × height (cm) - 5 × age (y) + 5 Women: RMR = 10 × weight (kg) + 6.25 × height (cm) - 5 × age (y) - 161

Total daily energy expenditure (TDEE) is RMR multiplied by an activity factor, typically 1.2–1.9. The standard error of Mifflin-St Jeor is about 10% in adults. So a predicted TDEE of 2,300 kcal has a range of roughly 2,070–2,530 kcal.

As weight falls, RMR falls. A 10 kg loss reduces RMR by about 100 kcal/day in the Mifflin-St Jeor equation (10 kcal per kg). That means a diet that starts at a 500 kcal deficit may end at a 400 kcal deficit if intake is unchanged. The line is not straight.

Worked example: weekly fat loss

A person weighs 80 kg and wants to lose fat. Their estimated TDEE is 2,400 kcal. They eat 1,900 kcal/day, a 500 kcal deficit. Over seven days, the expected fat loss is 500 × 7 / 9,000 = 0.39 kg, or 390 g. But daily water swings are 1–2 kg. The weekly average of daily weights is needed to see the 390 g.

Estimated TDEE = 2,400 kcal/day
Intake = 1,900 kcal/day
Daily deficit = 500 kcal
Weekly deficit = 500 × 7 = 3,500 kcal
Energy per kg fat = 9,000 kcal (Atwater factor for fat: 9 kcal/g)
Expected fat loss = 3,500 / 9,000 = 0.389 kg

The 7,700 kcal per kilogram figure often quoted describes adipose tissue, not pure fat. Adipose tissue is about 87% fat, so 1 kg of adipose tissue contains about 870 g fat and 130 g water and protein. The energy density is therefore about 7,700 kcal/kg. This figure is wrong early in a diet because water loss dominates, and wrong late because maintenance falls. It is a simplification in two directions.

The practical answer

Weigh daily at the same time, after waking and using the bathroom, before eating or drinking. Record the number. At the end of the week, average the seven readings. Compare weekly averages, not daily readings. A change of 300–500 g per week is within the range that weekly averages can detect. A single reading cannot.

This page is not medical advice. If you have a history of disordered eating or are concerned about your relationship with food or weight, contact a healthcare professional. In the United States, the National Eating Disorders Association (NEDA) helpline is available at 1-800-931-2237. In the UK, Beat offers support at 0808 801 0677.

FAQ

Q: Why did I gain 2 kg overnight? A: A 2 kg overnight gain is almost entirely water, glycogen, sodium, and gut contents. To gain 2 kg of fat, you would need to eat about 18,000 kcal above maintenance. That is not a single day’s food. The scale measures total mass, not fat.

Q: How much does water weight fluctuate daily? A: Daily swings of 1–2 kg are routine. Glycogen binds about three grams of water per gram, sodium shifts 0.5–1.5 kg, the menstrual cycle adds 0.5–2 kg, and gut contents add 0.3–1 kg. Training inflammation adds 0.5–1 kg for 24–72 hours.

Q: Can I detect fat loss with a single weigh-in? A: No. A week’s fat loss is often under 300 g, while daily water noise is 1–2 kg. The signal-to-noise ratio is about 1:5. A weekly average of daily weights reduces the noise to about ±100 g, making 300 g detectable.

Q: What is the 7,700 kcal per kilogram rule? A: It describes adipose tissue, not pure fat. Adipose tissue is about 87% fat, so 1 kg contains about 870 g fat. The energy density is about 7,700 kcal/kg. It is wrong early in a diet because water loss dominates, and wrong late because maintenance falls.

Q: When should I weigh myself? A: Weigh daily at the same time, after waking and using the bathroom, before eating or drinking. Record the number and average the week. Compare weekly averages. A single reading is not informative about fat loss.