I have a fancy scale that in addition to showing my weight, shows me body fat percentage and body water percentage.

It works by sending a small electric current up one leg and down the other, and by some magical means purports to tell my my body fat percentage. I make no assertions about its accuracy, other than to say it is very consistent. It will give me exactly the same reading to the nearest one-half percent ten times in a row. According to the scale, my body fat percentage varies day to day from 14% to 16%.

However, there is one thing about it that makes me doubt its veracity. If I am careless and eat an extra piece of chocolate cake or engage in other evil behavior enough to run my body fat readout to 16%, if I then go out and exercise with a hard two-hour kayak trip, or maybe a four hour hike in the mountains, when I get back the body fat readout will have dropped to 15% or maybe even 14%. It doesn't seem reasonable to me that I could have lost more than two pounds of fat in those few hours, particularly since my overall weight won't have gone down that much, or possibly even increased depending on how what I ate during the exercise and how much I drank.

This behavior from the scale is absolutely consistent.

So, how do these scales measure body fat, and why does it give me these very consistent results that cannot be true?

tanstaafl.

Based on your description, I too, am skeptical. What is it measuring in the current? Resistance? Voltage? Pulse timing? Something else? I've never seen one of these things before, so I don't know how it's supposed to work.

I could imagine that measuring how well electric current travels up one leg and down the other can measure certain things, but I really doubt that body fat percentage is truly one of them. For instance, I would guess that the current would travel better if you were better hydrated, and that's something which can vary greatly. Maybe a difference in hydration accounts for the different readings you're getting.

I could also imagine it reading different values for people of different heights, or for people of different leg-to-height ratios. For that matter, does it attempt to control whether your thighs are touching? Someone with a space between their thighs would probably read differently than someone whose thighs were touching. That might be a fun test: Try deliberately standing in different positions, either with your knees touching, or your knees far apart, to see if that alters the reading. See if being wet reads differently than being dry, that sort of thing.

Conductivity, which is the primary thing measured, changes with hydration level and will affect the readout.
Use it for trending, trying to keep circumstances before measuring as like as possible, rather than looking at the absolute value.

Went through a big weight loss/diet/exercise regime 11 years ago and graphed the various values the scale spat out. %Fat was a quite jagged one depending on how much I'd been drinking etc.

None of the above are applicable.

The scale "knows" my age and my height as it is personalized with that data. It also "knows" my weight and my hydration, as those are some of the data it records and displays. There are marked areas on the scale platform where I put my feet. My hydration is almost always at 57--59%. It makes no difference if my feet are wet from the shower or if they are dry. I take the readings first thing in the morning.

The only time anything changes inexplicably is when I take a morning reading and see 16% fat, then go out and exercise for a few hours and come back and it's 14%, with weight and hydration virtually unchanged. If I go the rest of the day without eating, the fat percentage will creep back up. Somehow exercise seems to be misleading the scale.

tanstaafl.

Mine stays in a very narrow range when looked at only in the morning before heavy exercise. I can actually predict what it will be based on what and how much I ate the day before. Then, when I exercise (I'm talking two to six hours of aerobic exercise at a high physical output) it drops, only to gradually come back up during the rest of the day. Never more than 16%, almost never less than 14%.

It is almost certain that my actual, true body fat percentage is NOT 14% or 15% or 16%. But the readings are so consistent with the exception of the post-exercise temporary drop that the scale is nonetheless a useful tool for day-to-day monitoring.

tanstaafl.

From what little I've read about them, exercise does that. My theory (grain of salt -- I'm not a bio-anything, much less bio-electrical engineer) is that it's due to the pump you get from exercise. The extra blood flow increases the size of your muscle cells, which in turn changes the electrical resistance ratios enough to throw off the body fat calculation from the norm. As you lose your pump, you drift back to your baseline.

Having formerly owned a scale like this and tracked trends over time, I found that the fat percentage was largely unhelpful. When I went through a diet thing to reduce my overweight saggy self, the % fat declined alongside my weight, telling me precisely nothing of value.

Apparently, if you're seriously about knowing your fat percentage, you can get weighed while underwater. Compared to a regular scale, the change in your, umm, buoyancy says something about your fat.
http://en.wikipedia.org/wiki/Hydrostatic_weighing

While I doubt that the number my scale tells me is my true real body fat percentage, it is consistent and repeatable and does allow me to accurately monitor changes and adjust my diet and exercise regimes accordingly.

Whatever my true body fat percentage is, I am satisfied with what is is now. I think at the age of 70 years I am more fit now than I have ever been. So as long as my scale confirms that my fat percentage is not changing, I'm happy.

tanstaafl.

That sounds entirely plausible. It is the best explanation I have come across.

tanstaafl.

Heh. I always forget you're up there...

I have a scale like that, but I don't know how far I would trust the %fat reading since it can't even calculate my age correctly... I'm born in -69 and according to the scale I'm 51 years old. The weight reading corresponds to the reading from a different scale though.

But... doing OK anyway.

tanstaafl.

It's obvious: you've worn too many pointy boots. Damn! Alaska, anyway...



-jk

I hope I am in such good shape at 70

I'd be impressed if I make it to 70, in any shape.

Heck, yeah. Better than I am, anyway.