Report on accuracy of fitbit

Options
Live_life_well
Live_life_well Posts: 86 Member
edited March 2018 in Health and Weight Loss
Hi,

I have been using Fitbit to track excercise and MFP to track calories since January and now have 2.5 months of data. I know that people have said to cut the Fitbit excercise adjustment by 50%, but I have found the opposite.

I have analyzed the data and with the Fitbit adjustment i should have lost 9 pounds based on a deficit of 30,263 calories over the past 2.5 months. However, I have lost 14 pounds which means that I ran a true deficit of 49,000 calories over that period.

This is just one data point but thought to share to encourage others to do their own math instead of just assuming that Fitbit overstates excercise calories. In my case, either my base calories in MFP are understated and my metabolism is higher than average or Fitbit understated my excercise calories. I lean towards Fitbit understating my excercise calories as I would not have needed to lose weight if my metabolism was that abnormal lol

Replies

  • Poisonedpawn78
    Poisonedpawn78 Posts: 1,145 Member
    Options
    do you know that all 14 pounds lost were fat and nothing else?
  • Zodikosis
    Zodikosis Posts: 149 Member
    Options
    I have found my Fitbit Alta HR to be incredibly accurate, down to the 0.1 lb lost. However, I wear it almost 24/7 (except for in the shower) and am incredibly meticulous with my food log -- everything gets recorded, preferably by weight in grams. I am also probably considered your "average" person losing weight -- I am not morbidly obese, nor am I incredibly athletic/active.

    The Fitbit does need time to gather data on you before its predictions can become very accurate, and you must wear it almost all the time. Also, if yours does not have a heart rate monitor, it is probably less accurate.
  • Live_life_well
    Live_life_well Posts: 86 Member
    Options
    do you know that all 14 pounds lost were fat and nothing else?

    No. But I checked whether the understatement of my CO in CICO was different in January vs February vs March and the understatement varied but is of a similar size. I would assume that the water weight loss would have happened in late December as that’s when I stopped eating out all together and started to eat healthy and upped my exercise.

    Is there other analysis I could do to rule out water weight losses consistently happening throughout 2.5 months of fairly steady weight decline?
  • Zodikosis
    Zodikosis Posts: 149 Member
    edited March 2018
    Options
  • Live_life_well
    Live_life_well Posts: 86 Member
    Options
    Zodikosis wrote: »
    I have found my Fitbit Alta HR to be incredibly accurate, down to the 0.1 lb lost. However, I wear it almost 24/7 (except for in the shower) and am incredibly meticulous with my food log -- everything gets recorded, preferably by weight in grams. I am also probably considered your "average" person losing weight -- I am not morbidly obese, nor am I incredibly athletic/active.
    I have Fitbit blaze which has HR and wear it religiously (only take it off for the shower). I also weigh and log all of my food - however, if there’s errors in my logging it is likely understating what I eat vs the inverse - I didn’t gain weight by overstatement what I ate lol.

    The Fitbit does need time to gather data on you before its predictions can become very accurate, and you must wear it almost all the time. Also, if yours does not have a heart rate monitor, it is probably less accurate.

    Good to hear that yours doesn’t overstate. To be honest, I am just glad that I no longer have to eat only 50% of my exercise as that’s what I was advised previously. If anything, I will eat more than recommended to find out what my true maintenance calorie level is.

  • PAV8888
    PAV8888 Posts: 13,757 Member
    edited March 2018
    Options
    Changes in body composition cannot really be captured. Some general assumptions could be made based on degree of available fat mass and level of deficit but it gets very complex for my little brain very fast!

    I think you will find a lot of people suggesting you trust but verify by looking at your results over 4-6 weeks. I am certainly one of them.

    Remember that the size of your adjustment actually depends on your mfp activity setting.

    A person who is set up a sedentary and a person who is set up as very active will get different adjustments for the exact same TDEE detection.

    Therefore the percentage of the adjustment is not the right thing to look at.

    The total daily Fitbit TDEE and the divergence of your expected logging results from that is the right thing to look at in a general sense.

    My 3 year results have not exceed a 5% divergance from trending weight in any multi month period.
  • CarvedTones
    CarvedTones Posts: 2,340 Member
    Options
    A couple of people I have walked with have FitBits that record more steps and distance than we actually walked (known distance routes). Oddly enough, my cheap Asian knock off is extremely accurate on distance (but horrible at heart beat).
  • psychod787
    psychod787 Posts: 4,088 Member
    Options
    As far as calories burned. I have no idea how accurate mine is. Pretty close I think. As far as distance walked... pretty damn accurate!
  • Angierae75
    Angierae75 Posts: 417 Member
    Options
    If I’m hungry, I eat all my Fitbit calories. If I’m not, I don’t. I’ve lost 45 pounds.
  • PAV8888
    PAV8888 Posts: 13,757 Member
    Options
    Distance walked, without GPS, is a function of your stride length. Your stride length varies based on your height and individual physiology and based on your speed and terrain. It is also something that can be edited (on Fitbit's settings) separately for walking and running.
  • Kst76
    Kst76 Posts: 935 Member
    Options
    My fitbit is also very accurate.
  • Francl27
    Francl27 Posts: 26,371 Member
    Options
    I've been using a Fitbit for over a year now (Charge 2). If I believe its numbers, I would have gained 20 lbs back. I only gained back 10 lbs in that time... so it underestimates calories for me... which is a good thing!
  • zyxst
    zyxst Posts: 9,134 Member
    Options
    The only Fitbit that wasn't accurate for me was the Zip. Added about 1,000 calories to my TDEE. The Ultra and One are spot on.
  • BattyKnitter
    BattyKnitter Posts: 503 Member
    edited March 2018
    Options
    Mine (Charge HR) seems to be spot on, I'm losing at the rate I should be.
  • af240
    af240 Posts: 37 Member
    Options
    Mine is hard to say, I have the alta HR. I think it overstates my TDEE, but not as much as I originally thought.

    I've been logging my daily calories and TDEE in excel and according to that I should have lost 6.5lbs since Jan 8. But the reality is I've only lost 3.5lbs (5'2, SW: 127, CW: 124, GW: 115-110).

    However that being said, I also started a progressive weightlifting program and I've NEVER seriously lifted weights before. And, from Jan 8 to mid Feb I didn't lose a pound (even went up 2 pounds). My weight didn't start to drop until the end of Feb and has been drastically dropping since, so all the weight I've lost has been over the last 3-4 weeks. So I know I'm holding onto excess water weight with my new exercise.

    I'm also convinced I've gained a tiny amount of muscle, (I'm in a deficit, I know, but newbie gainz ftw), since I can DEFINITELY see changes in my muscle definition even though I haven't lost very much weight. Especially compared to how I've looked in the past at similar weights.

    So with all that to go off of, I think it slightly overestimates my TDEE but not by too much.
  • GottaBurnEmAll
    GottaBurnEmAll Posts: 7,722 Member
    Options
    Both my Flex and my Charge2 have been accurate for me.
  • yirara
    yirara Posts: 9,467 Member
    Options
    It's completely individual if it works for you or not. This also has a lot to do with how well your maximum heartrate fits with this dreaded equation 220-age, whether you set custom heartrate, and how fit/unfit you are and your heartrate changes throughout the day. If your maxHR is more than 2 standard deviations higher than the equation and your resting HR is fairly low because you're fit you're more likely to get inflated calorie burns for everything.
  • inertiastrength
    inertiastrength Posts: 2,343 Member
    Options
    I find mine is incredibly accurate as well; most of my cardio is just walking so that makes sense. My data is pretty solid logging wise and I'm fairly confident using my fitbit numbers. If that changes that's the first place I'll tighten up but for now I'm good!