How accurate are scales that use BIA to measure bodyfat?

Options
sarahbums
sarahbums Posts: 1,061 Member
edited April 2018 in Health and Weight Loss
I just got a new scale (this one) today. It supposedly measures body composition and BF% using Biological Impedance Analysis. For a long time I've read that this method isn't particularly accurate and is really just an estimate.

But, how rough of an estimate is it? Is it off by just a couple of percentage points, or can it be way, way off? As far as measuring weight, the scale is very accurate and reliable. But it's telling me i'm 15.7% bodyfat, which seems kind of low to me (online BF% calculators usually have me at 18-22% or so). idk, I'm skeptical.

[ETA: for reference my measurements are 5'4, 101lb, 33" bust, 25" waist, 33" hips]

Replies

  • kimny72
    kimny72 Posts: 16,013 Member
    Options
    15% is really low for a woman, so yeah I would question that!

    The scales are very affected by water weight and if I remember correctly only really measure your lower half so if you don't carry your weight evenly will be even more off. What I've heard is the number is useless but the trend can be kind of accurate.

    Having said that, I have one that has always worked great as a scale, but has had me at about the same bf% for 2 years through a 15 lb loss, so now I ignore it.
  • AnvilHead
    AnvilHead Posts: 18,344 Member
    edited April 2018
    Options
    Here's a good read which explains the accuracy (or lack thereof) of Bioelectrical Impedance Analysis (BIA) very thoroughly: https://weightology.net/the-pitfalls-of-bodyfat-measurement-part-4-bioelectrical-impedance-bia/
  • PAV8888
    PAV8888 Posts: 13,730 Member
    edited April 2018
    Options
    In contrast to others (who may well have responded before the edit that disclosed the OP's measurements) I am not too surprised to hear that someone who is appreciably under-weight has less than an average % of body fat.

    That said BMI scales are far from a gold standard and a trained eyeball will do as good (or better) a job estimating.
  • sijomial
    sijomial Posts: 19,811 Member
    Options
    They have a wide range of accuracy, don't all measure the same (upper body only, lower only, upper and lower). Technology is also improving.

    Because they have a wide range of accuracy some will be high some low some pretty accurate.
    I've had good results with a hand and foot sensor domestic model and also a commercial grade one. I've also had completely comical results with foot sensor only domestic ones. If you can validate against another source and use them sensibly and consistently then some can provide an economical and believable trend.
  • Johnd2000
    Johnd2000 Posts: 198 Member
    Options
    In my experience, my scales always measure about half a % more BF than I get from measuring with calipers (7 points). On that basis, I use the scales as a general indicator only, and use the calipers whenever the scales indicate a significant change.
  • kimny72
    kimny72 Posts: 16,013 Member
    Options
    PAV8888 wrote: »
    In contrast to others (who may well have responded before the edit that disclosed the OP's measurements) I am not too surprised to hear that someone who is appreciably under-weight has less than an average % of body fat.

    Yeah, whoops :lol:
  • collectingblues
    collectingblues Posts: 2,541 Member
    edited April 2018
    Options
    kimny72 wrote: »
    15% is really low for a woman, so yeah I would question that!

    The scales are very affected by water weight and if I remember correctly only really measure your lower half so if you don't carry your weight evenly will be even more off. What I've heard is the number is useless but the trend can be kind of accurate.

    Having said that, I have one that has always worked great as a scale, but has had me at about the same bf% for 2 years through a 15 lb loss, so now I ignore it.

    15.7 is low, but 101 pounds is seven-pounds underweight for a woman at 5'4" (so a BMI of 17.3, which, I'd add, is also below the generally accepted BMI required for an anorexia diagnosis), so it doesn't seem out of the realm of possibility.
  • mom23mangos
    mom23mangos Posts: 3,070 Member
    Options
    I found my scale was generally 1-2% lower than DEXA or BodPod. So with your measurements, I'd say that yeah, 16ish % BF is not unreasonable. It's definitely something to keep an eye on as you don't want to dip down below 15% for long. Competitive bodybuilders will do it short term for a show, but there's no reason for a non-competitive healthy female to be below that.
  • iyobosaaiyanyo
    iyobosaaiyanyo Posts: 22 Member
    Options
    Personally I use the one in my local gym .They have two there but then to stick with a particular machine as I was told .
    I also have a body fat caliper at home which I use then compare both results.
    They are accurate within less than 1 percent of each other
  • mabelgrex
    mabelgrex Posts: 24 Member
    Options
    as the only reliable and accurate measure is dunking oneself in a water tank, which is not practical, I look at those scales as just one tool to keep an eye on what is happening re: loosing or gaining
  • PapillonNoire
    PapillonNoire Posts: 76 Member
    Options
    I have two BIA scales. One consistently put me around 18% and the other consistently put me around 30%. So no, I don't have much faith in their accuracy. At this point I just go by what I see in the mirror. If I'm happy with my reflection, than the actual percentage is not important to me.
  • CSARdiver
    CSARdiver Posts: 6,252 Member
    Options
    All BIA devices carry a high level of inaccuracy; however you can minimize this by remaining consistent in measuring. I have a handheld Omron and measure my body fat % each morning after voiding. I also have a DEXA at work and scan myself from time to time with a result within ~1%.
  • mph323
    mph323 Posts: 3,565 Member
    Options
    My scale (Aria) puts me at around 20% body fat, compared to a DXA scan at about 25%. In my case, I'm guessing the discrepancy is because I have very muscular legs, and carry most of my fat around my middle, which probably skews the results.
  • PAV8888
    PAV8888 Posts: 13,730 Member
    edited April 2018
    Options
    mph323 wrote: »
    My scale (Aria) puts me at around 20% body fat, compared to a DXA scan at about 25%. In my case, I'm guessing the discrepancy is because I have very muscular legs, and carry most of my fat around my middle, which probably skews the results.

    I found zero correspondence between both the % changes and the numbers of Aria vs DEXA for me.
    Even when looking at legs only.
    Even when looking at just % difference between scans for either whole body or legs only or legs+pelvis.

    Now my stats abilities are not very good.
    Perhaps there is a relationship that I cannot see with the naked eye.
    I mean every single number did go down while I was losing weight... at varying amounts, levels and percentages :wink:

    <currently Aria comes up with anywhere between 15.6 and 16.0%. My last DEXA was 23.4% overall. Legs were 19.1% and 21.4%. Trunk 25.5%. My weight is 0.5lbs lower than at the time of the DEXA scan 8 months ago. I have zero reason to believe I am positively recompositioning.>
  • mph323
    mph323 Posts: 3,565 Member
    Options
    PAV8888 wrote: »
    mph323 wrote: »
    My scale (Aria) puts me at around 20% body fat, compared to a DXA scan at about 25%. In my case, I'm guessing the discrepancy is because I have very muscular legs, and carry most of my fat around my middle, which probably skews the results.

    I found zero correspondence between both the % changes and the numbers of Aria vs DEXA for me.
    Even when looking at legs only.
    Even when looking at just % difference between scans for either whole body or legs only or legs+pelvis.

    Now my stats abilities are not very good.
    Perhaps there is a relationship that I cannot see with the naked eye.
    I mean every single number did go down while I was losing weight... at varying amounts, levels and percentages :wink:

    Yeah, my numbers went down too, from close to 40% body fat to around 20%, but it seemed to me to be more weight-related than fat-related. It seemed to go down at the same rate when I was completely sedentary and when I was well into resistance training. I wish I'd had a DXA when I first started losing weight for comparison.
  • PAV8888
    PAV8888 Posts: 13,730 Member
    edited April 2018
    Options
    mph323 wrote: »
    PAV8888 wrote: »
    mph323 wrote: »
    My scale (Aria) puts me at around 20% body fat, compared to a DXA scan at about 25%. In my case, I'm guessing the discrepancy is because I have very muscular legs, and carry most of my fat around my middle, which probably skews the results.

    I found zero correspondence between both the % changes and the numbers of Aria vs DEXA for me.
    Even when looking at legs only.
    Even when looking at just % difference between scans for either whole body or legs only or legs+pelvis.

    Now my stats abilities are not very good.
    Perhaps there is a relationship that I cannot see with the naked eye.
    I mean every single number did go down while I was losing weight... at varying amounts, levels and percentages :wink:

    Yeah, my numbers went down too, from close to 40% body fat to around 20%, but it seemed to me to be more weight-related than fat-related. It seemed to go down at the same rate when I was completely sedentary and when I was well into resistance training. I wish I'd had a DXA when I first started losing weight for comparison.

    Me too! My first was ~55lbs down at 225lbs and it was 38.8%


    It also doesn't change the fact that for the OP being substantially under-weight and having a low fat percentage are entirely consistent states of being!

    Thus there exists no real reason to doubt the accuracy of her scan at this point of time...
  • svel713
    svel713 Posts: 141 Member
    Options
    At a 18.1-18.3 BMI I had registered anywhere from 16.5-19% in body fat percentage, so I can see someone being a 17.3 BMI and 15.7% BF. Probably closer to 16-17% if you don't lift heavy.

Categories