Scales that report body fat percentage alongside weight have become ordinary household equipment. They do not measure fat; they measure electrical resistance and estimate the rest.
The method measures resistance to a current
A small alternating current passes between contact points, usually the feet on a bathroom scale or the hands on a handheld device.
Lean tissue contains substantial water and dissolved minerals and conducts the current readily. Adipose tissue and bone contain much less water and offer greater resistance.
The device measures how much the body impedes that current. The number it displays as body fat percentage is derived from that resistance, not observed directly.
Conversion depends on an embedded equation
Resistance is converted into an estimate of total body water, from which lean mass is inferred on the assumption that lean tissue holds a fairly constant proportion of water.
Fat mass is then calculated as whatever remains after lean mass is subtracted from measured weight. Height, age and sex enter the equation because the relationship differs across groups.
The equations are fitted to reference populations, so accuracy for an individual depends partly on how well that person resembles the group the device was calibrated against.
Hydration moves the reading directly
Because the estimate rests on body water, anything changing hydration changes the result. Dehydration raises resistance and typically inflates the fat estimate.
Recent exercise, a large meal, alcohol, and normal fluid shifts across the menstrual cycle all move readings without any change in body composition.
This is why consistent conditions matter more than the absolute number. Measurements taken at the same time of day under similar circumstances are comparable to each other.
Current path limits what is sampled
A foot-to-foot device passes current mainly through the lower body, and a hand-to-hand device mainly through the upper body. Neither samples the trunk thoroughly.
Since fat distribution differs between people, a device sampling one region may estimate poorly for someone whose distribution is unusual.
Devices using both hands and feet cover more of the body and generally agree better with laboratory methods, though they remain estimates.
How to read the output sensibly
Reference methods including imaging and displacement techniques are more accurate but require equipment available in clinical and research settings rather than at home.
The trend across weeks under consistent conditions is the informative part, since systematic error in the equation affects each reading similarly and largely cancels in a comparison.
Devices passing a current carry manufacturer cautions for people with implanted electronic devices and during pregnancy, which are questions for a physician rather than a manual.