Best scale for body composition?

Equipment
1

TL;DR: What’s a good “smart scale” for athletes, that will report accurate body fat and (ideally) lean mass data? Integration with Garmin Connect or some other ecosystem would be nice, but not required.

I’m finding very little useful information on “smart scales” that will give me accurate body composition data, especially very little useful information on the accuracy of the body fat data provided.

Despite my best efforts, I find myself choosing almost blindly between a $200 Withings Body Comp and a $150 Garmin Index S2 (but worried that I’ve seen several reviews and comments about inaccurate data :exploding_head: which, WTF Garmin?)

Sigh.

The long version…

For example, Withings offers a $100 Body+ scale which reports body fat. In theory, that’s all I want. But there’s also a $190 Body Comp scale, and a $400 Body Scan scale. Is the data BETTER from the more expensive scales, or is it only about added features? How do we find out?

There’s very few comparative reviews that I’d consider trustworthy. There are very few recent reviews. (Noteworthy that @dcrainmaker has only a Garmin S2 review from 2020, a Withings review from 2016, and a Tanita review from 2010… nothing for the last 3 years.) The Withings thread from 2022 has no real data, and the DEXA scans thread is old enough that I’d be looking for updated info.

I also checked the following review articles, which all say similar things but provide little detail on the data.

Runner’s World: 6 Best Smart Scales, Jan 2023

Wirecutter: Best Smart Scales
Chicago Tribune: Best Body Fat Monitors, March 2023

TechRadar: Best smart scales 2023

CNET: Best smart scale for 2023

VeryWellFit: 14 Best Smart Scales, Aug 2023

Forbes Vetted: 8 Best Smart Scales, Jan 2023

Sports Illustrated: Best Smart Weight Scales for 2023

LiveScience: Best Smart Scales

2

can’t recommend scales, but I can say that I’ve looked at the integration part of it and If it’s w/ GC, Intervals.icu pulls data from it (but not all for some reason). Barring that, you also have 3rd party websites that will sync from A → B (but again, there will be hits and misses)

Withings does provide an API, so you can also write your own to pull and push into GC I presume.

1 Like
3

None.

11 Likes
4

Pretty sure it was Jonathan who recently shared that his (non-sponsored) Garmin scale aligns pretty accurately with Dexa scans.

1 Like
5

My Garmin Index varies by 1% bodyfat over very short periods of time, which makes me not trust it.

Like one week I’ll be 7.2% and a few weeks later I’ll be more glycogen loaded and more hydrated and I’ll be 8%+. That’s like 1.5lbs of fat with no dietary changes in a week or two. No way.

I haven’t had a dexa scan, but I also suspect I’m not quite that low bf% wise.

But it’s easy to use and syncs to GC which is super nice.

2 Likes
6

There’s no difference IME. Just get one that integrates with whatever platforms matter to you. They all should be consistent with themselves

As I understand it, impedance measured from the feet only can’t accurately assess these things

1 Like
7

I’m curious if their new scale with the handle will make a notable difference?

2 Likes
8

Interesting, wasn’t aware those were available at the consumer level yet. My uninformed guess would be that it’s better but still an estimate that isn’t provably accurate for all users. Still, :crossed_fingers: it is a step in the right direction

1 Like
9

My former Dr ran one of the WVU diabetes labs and said that they used the handle kind since it was easier than trying to get elderly and overweight patients into a bodpod, could also check at each visit to the lab more easily. He had invited me to stop by the lab anytime I wanted to test on it but unfortunately he left for a job at the VA (I think) before I could. He sort of poopood my tanita* non handle one and said that I should upgrade. I believe the one at the lab was a “lab grade” one of some sort but said the consumer ones that were coming to market were good, didn’t mention brands that I recall.

*The tanita I have is a model Nate also has/had and lined up well with his Dexa testing a few years back, unfortunately not made anymore. Assuming they didn’t get worse that seems like another decent one.

2 Likes
10

I just went through this research project and came to the conclusion that the accuracy of this kind of measurement is very flawed. Maybe it could tell you within 5% and you could look at trend lines over time but then again, one could look in the mirror and look at the scale and come to a similar conclusion.

I also tried to research the difference between a $500 Tanita, a $150 Garmin, and a $25 no-name scale from Amazon and I couldn’t find any conclusion that the more you spend the more accurate the device.

Since we already have a scale, I’ve decided to pass on the body comp feature since it’s not hugely useful. If you want to throw money at it for entertainment value, that seems to be the most you’ll get - looking at some not-so-accurate graphs and trend lines.

5 Likes
11

I have had the impedance resistance type of scales for at least 10+ years - I replaced my first non-smart unit because my girlfriend didn’t like how it looked. We ended up getting a Withings body smart smart scale and I have to admit it was a nice upgrade.
I can’t tell you how accurate it is, but I can say that as long as you weigh in consistently, you will get all the data you need to see trends over time. It syncs with my apple health apps which pulls data from the scale, and my watch which records sleep time, hrv, etc… this combined with my exercise and ride data along with my food data from loseit means that I have more than enough information.
I don’t think you can go wrong as long as all of your data syncs up regularly and it correlates with how you look and feel.

3 Likes
12

Indeed.

And many more.

2 Likes
13

Good Lord. Some very damning content in all of that. Thanks to everyone for your input. Some particular notes from reading all of what y’all posted…

  1. BIS (hands and feet) > BIA (feet only). If they’ve done a good job (no way to know for sure), the Withings Body Scan scale with the handle should provide a better estimate of fat mass than all the ones with only foot sensors since it can use more frequencies and contact points for its analysis.

  2. BIA can both over- and under- estimate fat mass. Scary thought. And just counting references makes me think underestimating is more likely. Scarier still. Worst of all, fat content of extremities influences readings more than fat content of trunk, despite lower percentage of overall body mass… and I, like many, have MOST of my fat concentrated in a large potbelly. So there’s a real chance I’m in even worse shape than I thought. :flushed:

So yeah, the actual measurement accuracy is badly flawed as @AJS914 said. But in general, I’ll take a rough/imperfect tool over no tool, and the one scale that offers at least some hope for reduced error is the Withings Body Scan, so I’ll go with that. Having a reasonable data point every day is motivationally reinforcing for me: I simply behave better when I have data.

Thanks to everyone for your input! Hugely useful in being able to make a quick decision.

1 Like
14

I have a bit of professional experience in the category and, unless something new has been developed, the handheld electrical impedance is not significantly better than going through the feet. It is still an estimate based on signal time and a number of other factors.

3 Likes
15

my old almost 25 year old OMRON handheld at least generated predictable numbers (it seemed to calculate lean weight and assumed everything else to be fat) unlike the new devices (including omron) which seem some kind of random number generators (eg when I pick up a weight I supposedly get more muscular or when I take a dumb I lose muscle…)

that said I think my samsung watch 5 is somewhat okay for at least tracking parameters

1 Like
16

OK, but predictable =/= accurate. It doesn’t even really equal reliable.

17

but the reverse is true: unpredictable==unreliable as I said going to the toilet or picking up a weight shouldn’t affect my lean weight, only my body fat

18

It seems like knowing the body fat percentage off a BIA scale really provides zero day to info that you could act on. Ok, so one day it says 25%, the next 27%, the next 26%, then 28% … What could one possibly do with that info? We know that the numbers can change with hydration and all that.

It’s easy enough to look at one of these charts (google to find many others):

750×240 32.3 KB

If one is at 30%, for example, it’s going to take a good amount of weight loss to get to 25% and it will show in the mirror and your waist size.

19

I’ve no idea if they are accurate at all (given the price I doubt they are) but out of curiosity I bought a set of these a few years back. The app is a bit of a pain, its probably compounded by my ancient smart phone. I have to let it load an add page, let it send me to a log in page, and click back to get out of it to let me get access to the weigh button. For the price paid for them I don’t really mind though.

Buy Podium BMI Wireless Digital Scale - FY-WS01 | Tweeks Cycles

20

Can only say my Renpho one from Amazon matched a dexa for stats, within 30 minutes of the dexa. And the lab scales for weight. So at that moment in time it worked, and that’s enough for me.

It’s pretty consistent - I don’t get different figures stepping off and on again, which I have had in the past with non impendence digital scales. I’ve no idea if it’s accurate - trend is your friend.

It syncs to fitbit, and from there smartscalesync (developed by patrick who was on here) pulls it, and then pushed to Garmin Connect, TrainerRoad, Strava…

1 Like