Measuring body composition: what the methods can do
Body weight sums up very different things in one number — muscle, fat, water, bone, stomach contents. There are many methods that try to separate these. They differ less in whether they are accurate than in what their inaccuracy looks like.
Why weight alone says little
Two people can weigh the same and be the same height and still be built completely differently. The scale adds up everything that belongs to the body and states a sum. Whether the last two kilograms are muscle tissue, retained water or fat does not follow from it.
For many questions that is still enough. Anyone gaining or losing over months sees the direction without a breakdown. Composition becomes interesting where weight and appearance decouple: when returning to training, in phases with a calorie deficit, in older people.
A common case is the start of strength training. In the first months little may happen on the scale while circumferences and appearance change — muscle tissue is added, fat tissue goes down, and the two partly cancel out in weight. Anyone looking only at the sum during this time concludes that nothing is happening, although the opposite is true.
Scales with bioimpedance
These devices send a weak alternating current through the body and measure the resistance. Fat tissue conducts less well than muscle and water tissue; from the resistance and stored equations the device estimates a fat percentage.
The method is cheap, fast and repeatable — and it reacts sensitively to fluid balance. After drinking, after training, after a salty meal or at different times of day the displayed values can differ noticeably without anything having changed in the body. Household devices that measure only through the feet also capture the upper body only indirectly.
The practical consequence: the absolute value from a bioimpedance scale carries little. The change over weeks carries more — provided you always measure under comparable conditions.
Calipers and circumferences
With skinfold calipers the thickness of a lifted fold of skin is measured at defined points. From several measuring points, equations estimate an overall value. The method is independent of fluid balance but depends heavily on the person performing it: grip the fold slightly differently and you measure differently. On yourself this is hard to keep consistent.
Circumference measurements with a tape measure are simpler still and capture not composition but shape. Precisely for that reason they are often more informative in everyday life than an estimated percentage: if waist circumference falls and upper-arm circumference stays the same, that is information no estimating equation has to supply.
The elaborate methods
Research and medicine use methods that require more effort.
- DEXA. An X-ray measurement at two energy levels that separates bone, fat and fat-free mass and also breaks it down regionally. In many studies it serves as a reference, but it is not error-free itself and gives different values depending on device and settings.
- Air displacement. In a sealed chamber, body volume is determined via air pressure and a density calculated from it.
- Underwater weighing. The same principle via buoyancy in water; long the reference standard, rare today.
- Multi-component models. Combine several measurements and are considered the closest approximation — correspondingly elaborate.
These methods are not accessible to most people, and for the question of whether training is working they are not needed.
What all methods share
None of these methods measures body fat directly. All measure something else — a resistance, a fold thickness, a density, an attenuation of radiation — and calculate a fat percentage from it via equations. These equations come from samples and fit an individual person to varying degrees.
From this follows a rule that applies regardless of the device: values from different methods cannot be compared. Measuring 22 % by scale today and 18 % by caliper next month tells you nothing about a change. Only measurements by the same method under similar conditions are comparable.
What works in everyday life
For most people a simple combination is enough: weight regularly and always under the same conditions, plus two or three circumferences several weeks apart, and photos in the same light and the same posture. These three together show a development more reliably than any single percentage.
An estimated fat percentage can run alongside as an additional line. It just should not carry the decision — its uncertainty is too large and too hard to judge for that.
It also helps to consider beforehand what a measurement is actually meant to settle. Anyone wanting to know whether training is working gets the answer faster from training performance: if the same exercise works with more weight or more repetitions, something has changed — demonstrably, without an estimating equation.
The key points
- No common method measures body fat directly; all estimate via equations.
- Bioimpedance reacts strongly to fluid balance — time of day and fluid intake change the value.
- Caliper measurements depend on the person performing them and are hard to keep consistent on yourself.
- Values from different methods cannot be compared with each other.
- The change over weeks is more meaningful than any single measurement.
- Weight, circumferences and photos together show more than an estimated percentage alone.
Last reviewed: 14.08.2026
Created by AI