Wednesday 16 October 2013

Does a body composition scale provide independent information? Two experiments: weight and age

Since the beginning of the year I have gained about 8 kg. Was that fat, muscle, water? Probably some of all. My body fat scale estimates that my body fat percentage has increased by about 3%, which would be 2 to 3 kg of fat.

The problem

My problem? My scale knows my weight gain and likely uses my weight to estimate my body fat percentage. If it would not use my weight to estimate body fat, why would it want to know my height (body mass index, BMI), age and sex? Part of the information on body fat likely comes from the currents send through my body by the scale, but another part from my weight or BMI. Thus my problem is, does the increase in body fat percentage tell me more than what I already knew, that I have gained 8 kg?

The experiment

Now luckily, my fitness studio has another instrument to measure body fat. It is attached to a computer and also wants to know my weight. But here, you have to type in. Thus with this instrument you can experiment.

The instrument has a infra-red sensor that has to be placed on your biceps. To compute your body fat, and a colorful page full of other probably highly reliable information, it uses your weight, height, age and fitness.

We made one measurement stating my real current weight and one stating my weight at the beginning of the year. The difference in computed body fat percentage: 3,3%! Almost to similar to the increase estimated by my scale to be true.

The interpretation

It is possible, I gained some fat and my waist did increase some. It is also possible that nothing happened fatwise, other places look more muscular nowadays. I guess all I know is that I gained 8 kg.

I guess this does not surprise scientists and engineers working on these measurement devises. The German Wikipedia even mentions this effect qualitatively. But I am not sure if all users are aware of this problem and we can now put a number on it. If your fat gain is less than a third of your weight gain, the measurement is too uncertain to determine whether there really was a fat gain.

Post Scriptum. We forgot one experiment. I would love to know what the estimate for my body fat percentage would be if I tell the computer I am a young man. I expect some age discrimination by the multiple linear regression equations used.

Post Post Scriptum. Thanks to Caro of Stangenliebe (Pole art fitness Bonn) for the measurement.

UPDATE. I realised I could do the experiment with my age using my own scale. My Soehnle scale has two presets. I set person 1 to my age (rounded to 42) and person 2 to an Adonis of 18 years. In all four measurements the Adonis has exactly 3.7% fat less. If only my age group would not be so overweight, my readings would be a lot better.


  1. What you need Victor is a bicycle! This would allow you to eat whatever you want and still not get fat.

    Although I imagine the pole dancing (looks interesting!) would have much the same effect.

  2. Welcome.

    I have a bicycle. I am Dutch, I could cycle before I could walk! Well almost. Even more, I am from Groningen, which was often elected to be the World Cycling Capital.

    Now that I life in Bonn, I mainly go by foot, though. For Germany, cycling in Bonn is very good and the locals are very enthusiastic about how good the situation is for cyclists, but I am a little spoiled and you still have to concentrate not to be run over by a car. Walking lets you day dream and is nicely relaxing.

    I can already eat whatever I want. I used to be overweight, had to watch what I ate and was hungry most of the time. Two year ago I made a few simple life style changes, now those times are over for me. More walking was one of these changes.

  3. Trust me to tell a Dutchman to get a bicycle! Sorry! I didn't know you were Dutch. I love, love, love my Dutch bakfiets. I think the Dutch are very clever to have invented this contraption. I get positive comments and smiles from people in York almost everyday I ride it.

  4. Hi, Victor!

    I just wanted to say I liked this post and found it helpful. My parents (age 60 or so) are on a fitness kick and my dad, a retired engineer, thought maybe he should get one of these smart, body-composition-sensing scales.

    Based on the price of one he was looking at, I doubted it could have any actual, DEXA-type technology that really *could* tell your body composition, so I figured it must give you the numbers based on guesswork.

    Your post has confirmed that for me, so we'll just be keeping our old-school scale.

  5. Thanks for your comment. I did put a question mark in the title on purpose. If the change is fat percentage is large enough, it will give you useful information.

    To track progress, photos, a measure tape (especially, waist and hips) and caliper seem worthwhile. Maybe even more important is your health and digestion, but that is more subjective and thus harder to track.

  6. Lindsay, maybe I should add that an old-school body weight scale would have performed even worse in my case. If I had only known that my weight had increased by 10 kilogram, I might have thought I had become more fat.

  7. And here comes the problem of scales' accuracy. Are results really 100% or even about 80-90% correct? We probably can't rely on it, because the nubers are inexact.


Comments are welcome, but comments without arguments may be deleted. Please try to remain on topic. (See also moderation page.)

I read every comment before publishing it. Spam comments are useless.

This comment box can be stretched for more space.