Aria seems off with calculations

[double0kevin34]

So I have started to track body fat as well as weight lost, and have started taking measurements as well.  Out of curiosity, I logged in to look at the data that has been sent to Fitbit through my Aria scale.

Bear with me here, and tell me if this remotely makes any sense to you: I am eating around 1,500 calories a day with a very high protein intake.  I am lifting weights 3 times a week, and doing martial arts or cardio on non-weight training days.  Needless to say, I have been very active.  I hit a wall as far as weight loss goes the past few weeks, and I assumed it is because I am gaining lean mass.  My body fat was going down, and still is.

Today, I noticed that I am back to losing weight again.  Interested in the stats, I checked the graph.

Riddle me this: If I have not lost 10 pounds since September 13, and today is September 29; if my body fat has been doing NOTHING but decreasing since I started working out, then -

How it is even remotely possible that Fitbit has calculated that my lean body mass on September 13 was 163lbs, and somehow today it is 153lbs?  I am noticing changes, losing fat, but I certainly have not lost 10 pounds of muscle in the past 2 weeks.

How accurate is this **ahem** thing??

[Dominique]

Welcome to the community, @double0kevin34! I don’t know which of the two guys in your profile photo you are, but both look pretty jacked to me. Which body type have you selected for yourself under the People tab of your Aria: Regular or Lean? Where would you put yourself if you look at the sample pictures in this link? If you have Regular, but look like the 10-12% guy or leaner, you may want to try Lean for body type.

BIA scales are very sensitive to hydration. For optimal results, you should use it in the same conditions, ideally first thing in the morning.

Your Aria doesn’t care (or even know) what diet you are eating and what exercise you are doing. It just sends a weak current through your feet, gets an impedance value in return, compares it to a database of people who had their BF% determined with a more reliable method and try to find some that are close enough to you (in terms of age, height, weight, body type etc.). If you have an unusual body type, it may not find enough subjects close to you. 

A loss of 10 pounds in 16 days (4.3 lbs per week) is extremely high. Chances are a large part of it is water (no way you can lose so much fat in such a short time), which would throw off BF% estimates (because hydration is so critical). No way you would have gained so much muscle (again, in such a short period of time) this would explain why your weight loss stalled. Not sure what your expectations are, but it’s unlikely you could sustained a 4+ lbs weekly loss for very long. Have you cut your carbs drastically?  

[double0kevin34]

I have not cut carbs on purpose, I am more counting calories and ensuring that I get 120+ grams of protein in daily within my limits.  I am not questioning the accuracy of the weight, and really not even the accuracy of the BF%.  I went into the graphs on this site and looked at the lean mass chart.  That showed that I had lost 10 pounds in lean body mass since September 13, which I find to be impossible.  Because it makes sense that my weight loss stalled with my body fat going down consistently, as that would indicate I am actually gaining lean mass as a result of my training.  Does my logic make sense?  I just don't see it possible that I lost 10 pounds in lean muscle since September 13, or even lost any lean mass at all based on how I am training.

[Dominique]

The main problem here is you most likely have two very different BF% numbers on Sep-13 and Sep-29. Since fat mass is [bodyweight x BF%] and lean mass is [bodyweight - fat mass], lean mass and fat mass are completely off and make no sense. Your scale clearly doesn’t work for you (for BF%, and therefore fat vs. lean mass) during the very short period of time you’ve been using it.

Your real BF% can’t change dramatically in two weeks, because there’s only so much fat & muscle you can lose or gain during such a short period of time.

This is my own BF% for the past four weeks, as estimated by the Aria scale:

Highest value in that period was 14.5%, lowest 13.9%. May not be accurate either, but the magnitude of the change is plausible.

[double0kevin34]

My body fat looks fine, as does my weight.  I think that I should maybe simply pay attention to only these two parameters, and just never look at the lean body mass chart.  Even my BMI looks just fine.  The only thing that makes no sense is the lean body mass indicator as shown here.  And actually, now that I look even closer, it was WAYYYY off just 1 day apart.  Perhaps it was calibrating to me?  Not sure, but it shows I have gained body fat and lost muscle, which is impossible:

Sept 13: 233.8lbs, 29.9%BF, 163.8lbs lean, 70lbs fat

Sept 14: 230.6lbs, 35.3%BF, 149.2lbs lean, 81.7lbs fat

Sept 29: 227.9lbs, 32.0%BF, 154.9lbs lean, 73lbs fat 

I have no idea what could be causing such data discrepancies.  I do take my scale back upstairs to my room with me after using it in the bathroom on the same part of floor.  I also make it a point to step on the scale a few times, as I read that it will recalibrate itself if moved.

Just is not making sense to me at all.

[Dominique]

OK, the problem is with the huge variation in BF% (more than 5% difference in two consecutive days). The big variation of weight (3.2 lbs) between the same two days is also what gives crazy results when looking at changes in lean mass and fat mass. Obviously, any real change in your bodyweight (actual tissue, like fat and muscle, as opposed to water) can’t be that big one day to the next. Same with any real change in BF%.

What happened between Sep-14 and Sep-29: didn’t you weigh at all? Here are a few tips that should get you more sensible numbers: 1) weigh everyday at the same time of the day and in the same conditions (first thing in the morning, after trip to bathroom), 2) link your Fitbit account to TrendWeight (see link in my signature), 3) look at the numbers (both weight and BF%) over longer period of times (4 weeks, 3 months, 6 months etc.) as opposed to daily fluctuations or changes within a couple of weeks. 

The rationale in 1) is both to minimize the impact of hydration (there will still be some, for instance depending on what you have eaten the day before, like saltier food than usual etc.) and to have enough individual data points for TrendWeight to spit out meaningful numbers. For 2), TrendWeight calculates a moving average (technically speaking: exponentially smoothed moving average).

Regarding the placement of your scale: for optimal results, I would suggest leaving it in the same place and same position (e.g. not store it vertically after use, like many people are doing). You are correct that your Aria needs to be recalibrated everytime you move it. Since it can take up to five weigh-ins for the recalibration to be complete, and each weigh-in takes quite some time (the Aria is no speed demon), you’re wasting valuable time unnecessarily IMO.