Body Fat - The Grand Experiment - Some results

As I posted last week, I have completed Body Fat analysis by 4 methods.  I don't have the experimental MRI results yet, but I have Bioelectrical Impedence (Tanita Scale), BodPod and DEXA scan.

Reading about the various measures, the BodPod and Dexa have a probable 5% error - and BIA is worse...  But there is some correlation in the methods.

DEXA      5/20/14     9/5/13  

BF%       11.2%        10.9%    UP .3%

Wt          188.7         181.4    UP 7.3 (NOTE - not the same machine - different facility, same brand)

DEXA scans imply 0.3% increase in BF%, not bad considering last year was at the very end of the cycling season and this time at the beginning.  

DEXA also implies gain of 7.3 lbs, but the vast majority is Lean tissue

DEXA does show BF by limb/torso - legs are 7%... and still shapely!

BodPod    5/20/14     12/6/13  

BF%       14.9%        14.5%    UP .4%

Wt          184.8         181.9    UP 2.9lbs (NOTE - not the same machine - same facility)

BodPod imply 0.4% increase in BF%, While the absolute number doesn't match the DEXA, the trend matches very well.  

BodPod also implies gain of 2.9 lbs, but the majority is Lean tissue.

Tanita - Sorry, I don't have a history as I gave up on it long ago, but today it reads

BF%: 16.3% (within an hour of BodPod)

Weight: 184.1

 Overall impression:

Dexa and Bodpod are seemingly pretty well correlated, with BF% up about the same. Both have weight up, but the new DEXA scan weight seems like an outlier - main difference was I was wearing a hospital gown during dexa, so that would account for some of the weight.  With BodPod you wear a pair of compression shorts.  I had those on during DEXA also.

But there's clearly a big error somewhere between BodPod and DEXA (we know there's lots of error in the Impedence devices already).

I really think BodPod overestimates my body fat due to loose skin.  You have to wear a bathing suit cap in the machine so your hair doesn't through it off, so I can only imagine what loose skin would do!  It makes you seem like you have more "volume" which means you have lower density, which implies higher body fat...

Getting an absolute number without being dissected and rendered is probably impossible... each method has 5% or more error, and trying to compare two methods to get an actual BF% number is difficult. But at least for a small sample size the two main methods are indicating the same thing... off-season weight gain that was mostly lean tissue from resistance training. No big jump in BF (which is what i was concerned about with the weight gain).

I'll report back on the MRI when I hear.

Very interesting results and comparison Tom. Nothing definitive but the trends are there.

Regarding your thoughts about loose skin throwing off results, when I recently had DEXA I was instructed to leave my 'support garments' on. I was 'supported' from pubis bone to below shoulder blades. 

Good info!

Great experiment,

I had the same concern regarding loose skin and my Bod Pod results. I had asked that same question to the technician. Who also thought it was a great question. SO he in turned reached out to the manufacture. Here was the response from CosMed USA: 

Loose Skin after Bariatric Surgery

It has been found that the loose skin does not affect the accuracy of the BOD POD results. Body Fat percentage and Fat Free Mass percentage should all be within the correct range.

The makeup of the extra skin can only be broken down via chemical analysis, so we can't say the loose skin is only in the fat free mass category. But a study done with patients before and after gastric bypass surgery showed that excess skin did not skew the results obtained from the BOD POD.

Manoj Raghuraman

Clinical Research Manager

Thanks Michael!

"I really think BodPod overestimates my body fat due to loose skin."

I think, in spite of what the manufacturers say, that it could. Because you have to get all the air out. I wore a one piece bathing suit, not a 2 piece and I think that's why my body fat percentage was so high (much higher than common sense would dictate). I think air was trapped in the suit. They actually warn you about that in the instructions though I did not get that warning and did not read about it until after my test.

Yep... I saw your post on another website discussing the same issue when I was researching... I can't see how it could be quite as accurate - i mean if you need to wear a beanie to compress your hair... a blob of skin that is neither fat nor muscle could through it off.

That said, BodPod came in the middle between the DEXA and the Electrical Impedance scale (Tanita).  Now I'm pretty sure the Impedance is off partly because my scale does not have an "athlete" mode - which I have read changes the equation for those with lower bodyfat.... so maybe that would bring the BIA method closer...

However, the MRI radiologist seemed to confirm that the DEXA was pretty much on the money with respect visceral fat, so I'm going to tend to believe its BF% estimate of 11.2% (and it's the best number as well of course !).

I think the volumetric and impedance methods are going to be off for people at the extremes of BF% AND for odd situations like extra skin.  They are estimates based on equations which are based on "normal" populations, typically.  My tanita was pretty close until I dropped below 20%... then it started to hang tough and even though other methods have typically reported 10 - 11 - 12 and 14% the Tanita refuses to report anything below 16.3%

According to the DEXA printout i'm in the 0th percentile for BF% for men my age - so I'm in an extreme situation which is probably outside the ability of a sub-100 dollar scale to figure out.

"the Tanita refuses to report anything below 16.3%"

I had a similar issue at one point. It's like it just couldn't believe that a woman my age could have the kind of muscle I had. Too bad I still don't have it! LOL