Maintainers with Fitbit how accurate is your data?

Just curious how accurate in maintenance you think your fitbit "calorie in and out" is?

I don't really use MFP or Fitbit but the TDEE method to maintain but I do log daily. It is interesting to me to see that while I was losing my TDEE was 1850. Fitbit generally thinks I should eat 2100 or 2200 most days depending on my activity level. I can "make" MFP match by setting it to a certain cal intake level but in general I am going to say that Fitbit overestimates by around 300 calories for me.

I know some folks have said it underestimates for them. In my case I think it is because it overestimates the time I spend riding my horse and working around the barn.

Replies

  • autumnblade75
    autumnblade75 Posts: 1,661 Member
    Most of us don't have to worry about the steps of an animal we're riding being accrued to us. That is a valid edge-case scenario. I'm not logging closely enough to verify my own data, but I spend my workday waving my arms around and not walking. I am aware that is likely to inflate Fitbit's reported burn, too. I don't worry about it, though - I don't expect the difference to be a *dangerous* number of calories to consume. If I get serious enough to log meticulously, I'll be able to adjust for the difference.
  • ExistingFish
    ExistingFish Posts: 1,259 Member
    Just curious how accurate in maintenance you think your fitbit "calorie in and out" is?

    I don't really use MFP or Fitbit but the TDEE method to maintain but I do log daily. It is interesting to me to see that while I was losing my TDEE was 1850. Fitbit generally thinks I should eat 2100 or 2200 most days depending on my activity level. I can "make" MFP match by setting it to a certain cal intake level but in general I am going to say that Fitbit overestimates by around 300 calories for me.

    I know some folks have said it underestimates for them. In my case I think it is because it overestimates the time I spend riding my horse and working around the barn.

    Does your fitbit measure HR in addition to steps or just steps?

    Horse riding is an active endeavor.

    Mine generally measures up once I adjust for breastfeeding. Some days I don't move around as much, but most days I hit my calorie goal about the time I'm laying down to go to bed.

    My maintenance TDEE without breastfeeding is around 1800, I log around 2100 after adjusting for breastfeeding (300 cal).
  • SummerSkier
    SummerSkier Posts: 5,186 Member
    So you think yours is spot on then EF? That's cool. Autumn it doesn't sound like you keep the data to really know if yours is accurate or not. Do both of you then eat per daily allocations vs TDEE method?

    Yes - I have an Fitbit HR and it has been in use for over 2 years.
  • ExistingFish
    ExistingFish Posts: 1,259 Member
    So you think yours is spot on then EF? That's cool. Autumn it doesn't sound like you keep the data to really know if yours is accurate or not. Do both of you then eat per daily allocations vs TDEE method?

    Yes - I have an Fitbit HR and it has been in use for over 2 years.

    I am currently eating mindfully, not logging food. Seeing if mindful eating can keep me at maintenance. I'm maintaining my weight, so it seems to be working.

    Regardless of what my fitbit said, I ate at 2100 for maintenance when I was logging. TDEE is an average, so why would I change my dad-to-day eating? If I were doing that, I'd be using NEAT + exercise, not TDEE.

    I mostly use the fitbit to keep me active and so I can see if I'm not being active enough over time. I usually hit my step/activity goals on exercise days, but not on regular workdays, I'm trying to fix that. I also use it to monitor my sleep quality.
  • autumnblade75
    autumnblade75 Posts: 1,661 Member
    So you think yours is spot on then EF? That's cool. Autumn it doesn't sound like you keep the data to really know if yours is accurate or not. Do both of you then eat per daily allocations vs TDEE method?

    Yes - I have an Fitbit HR and it has been in use for over 2 years.

    When I last logged seriously, I noticed that if I looked at the previous day, about 300 calories would have been subtracted overnight because I'm not as active while I'm sleeping as my activity level would require. I have since changed my Time Zone setting on MFP and no longer have that issue, but I'm just not logging seriously enough to compare the data. I suppose I should just butt out - I don't have a good contribution to your actual question. But I think the Fitbit's data is useful, even if it's not 100% accurate. I mostly use it for the HR information during running.

  • SummerSkier
    SummerSkier Posts: 5,186 Member
    autumn it's ok. I am not saying the fitbit info is not valuable but I was just wondering from maintainers if they had any particular trends as far as on target etc... Beulah, I am not sure that is quite accurate about formerly obese folks having less energy expenditure altho I will say maintenance does require attention (at least for me!). I don't want to derail the topic with an entire discussion about that tho as I think that might have already been discussed a bit elsewhere. ;)
  • krael65
    krael65 Posts: 306 Member
    edited September 2019
    Interesting question. I've been using my FitBit for over 3 years now. Since I started wearing it, I've been using the TDEE method for weight loss and maintenance. During my weight loss phase, my calories burned estimates were close enough, based on my rate of loss. I say close enough, because while I log pretty accurately, there's always going to be error. And the occasional skipped party or meal or vacation where I invoked the WTF rule and either lazy logged or just didn't log that entire day or event.

    I continued using my FitBit estimates during the next 2 years or so of maintenance. Looking back at my data, the estimate was maybe off (over) by 50-100 calories a day. Again, that could be due to user error in logging or the occasional WTF moments.

    For the past year, however, I've been struggling more. My FitBit estimates seem to be off by quite a bit more. Now in the ballpark of overestimating by 200-300 calories a day. I blame hormones! :smiley: (54 yrs old, hypothyroid, menopausal = good times!)

    Edit to add: So I basically eat to my FitBit estimate, always leaving some calories on the table (based on the info above).
  • AnnPT77
    AnnPT77 Posts: 34,598 Member
    I hope you're understanding that your Fitbit is just giving you a personalized statistical estimate, not measuring calorie needs directly?

    Any of these good quality, well-researched, well-programmed devices are going to be close for most people, further off for some, and way off for a very, very few (in either direction, high or low).

    Because of that, other people finding them to estimate high or low isn't very useful information.

    Think about it: When it comes to calorie needs, the population falls on a bell curve. It has a strong central tendency (the standard deviation is small; it's a pretty tall, narrow bell**). If you happen to be pretty close to average in calorie needs for your demographic, your Fitbit (or Garmin, or whatever) will seem pretty accurate. If you're further off from the population average, you'll find it less accurate. This is just how this sort of thing works.

    It doesn't provide much insight to ask for other people's indivicual findings: They're at different places on the bell curve.

    FWIW, my Garmin underestimates significantly for me. For other people, the exact same model is quite close, for others quite high. That's the nature of the devices, at least for now.

    ** https://examine.com/nutrition/does-metabolism-vary-between-two-people/
  • SummerSkier
    SummerSkier Posts: 5,186 Member
    Krael65, interesting that the estimate from fitbit has changed over time for you. That was not something I anticipated as an answer but I guess I should have. good point!

    AnnPT, I am just curious to hear other maintainers input on fitbit estimates for them.. It doesn't need to be important or useful information. Perhaps because you have improved your NEAT significantly is why your garmin underestimates your requirements? Just a thought. Or has it always been that way?
  • AnnPT77
    AnnPT77 Posts: 34,598 Member
    Krael65, interesting that the estimate from fitbit has changed over time for you. That was not something I anticipated as an answer but I guess I should have. good point!

    AnnPT, I am just curious to hear other maintainers input on fitbit estimates for them.. It doesn't need to be important or useful information. Perhaps because you have improved your NEAT significantly is why your garmin underestimates your requirements? Just a thought. Or has it always been that way?

    I haven't had the Garmin through the whole process of weight loss/maintenance, just for the last year or so. My NEAT, as far as I know, is unremarkable. I'm pretty sedentary, outside of intentional exercise, but maintain several hundred calories above what MFP (and some other calculators) estimate (MFP and my Garmin give similar estimates).

    I did lose weight faster than MFP predicted, and had to adjust my calorie goal upward because problems arose from losing weight a little too fast: The basic situation isn't new, it's been true since I started calorie counting (and presumably before, but I can't document it before starting to count).

    I have some guesses about why this higher calorie burn may be true for me, but daily life activity (NEAT) isn't high on that list (6,000 steps would be an above-average day for me, especially in Winter). Ill health doesn't seem to be a factor either: My blood tests have been fine, and my oncologists even ordered a full-torso scan as a precautionary check when I told them I was losing weight at higher calories than expected: Result was normal.

    There is variation between people that the normal calculator inputs don't account for. The link I put in my PP has some statistics.
  • SummerSkier
    SummerSkier Posts: 5,186 Member
    Yes - Ann, I found that an interesting link with the statistical data. Perhaps because you have always been "strong" and a rower/athlete you have more base muscle mass and it tends to not be accounted for in your garmin? I would be interested to hear what your guesses are. I suspect mine may be off because I am a older/short petite woman maintaining at the lower end of BMI but you would think Fitbit would account for those inputs.
  • LivingtheLeanDream
    LivingtheLeanDream Posts: 13,342 Member
    You have your own data to go by which speaks volumes. When I used Fitbit it over estimated my calorie burn by around 150-200 calories, it took me a while to realise that...
  • staticsplit
    staticsplit Posts: 538 Member
    I usually assume it's overestimating by about 200-300 calories. That also gives me plenty of wriggle room as I guestimate calories/take days off from logging a lot.
  • PAV8888
    PAV8888 Posts: 14,300 Member
    edited October 2019
    In general, for me, Fitbit TDEE was within 5 to <150 Cal of observed results when I was tracking food intake more accurately.

    While tracking "more accurately" and using a weight trend application to evaluate my results, my Fitbit TDEE was off:
    0.15%, 12 months, 695 deficit based on weight trend
    3.14%, 12 months, 106 deficit
    4.77%, 12 months, 26 deficit
    2.66% , 5 months, 11 deficit

    "more accurately": using food scale as much as possible and double checking entries used. during first 12 months also included not eating anything before logging it.

    However I do agree with Ann that we can't assume that results will be similar!

    What I like about fitness trackers is that they provide me with an "un-biased" calculation of activity that is less influenced by my own perceptions and can serve as a relatively consistent estimate against which I can adjust my goals based on my aims and results. It helps that my activities are well suited to the current crop of trackers.