FitBit Exercise Adjustment

bjkoziara

I never thought I would be asking this question, because I thought I knew the answer, but here we are...

I've been using my FitBit Flex 2 to track my steps the last couple of days. I have MFP set to sedentary, as I work a desk job and am not really that active in my day-to-day. Currently set at 2 lbs/week loss because I started at 360 lbs (currently 339).

Today I went for maybe a 10 minute walk at a moderate pace before work, and another 15 minute moderate-to-brisk walk on my break. I've done a little walking throughout the day here and there but my FitBit estimates about 8,000 steps. While this is more active than I normally am, FitBit did a calorie adjustment of -493 calories. This seems excessive for less than 30 minutes of walking around. How accurate is this? I really don't want to eat those back if they are way off, but then I'm also eating at a 1,000 calorie deficit currently, so I guess I probably could? What are your thoughts?

PAV8888

It is accurate at midnight.

Your adjustment will decrease if you remain unmoving between the time you saw it and midnight.

Given you said your MFP actively level is set to sedentary, it will decrease at a rate of 0.25 * bmr / 1440 * m where m = minutes of being unmoving to midnight.

MFP assumes a level of activity split evenly over your 24 hours. Just the way they all do their math.

PAV8888

Of course if you remain active exceeding mfp's level of sedentary, it will keep increasing!

bjkoziara

PAV8888 wrote: »

It is accurate at midnight.

Your adjustment will decrease if you remain unmoving between the time you saw it and midnight.

Given you said your MFP actively level is set to sedentary, it will decrease at a rate of 0.25 * bmr / 1440 * m where m = minutes of being unmoving to midnight.

MFP assumes a level of activity split evenly over your 24 hours. Just the way they all do their math.

Weird, but I think I understand what you're saying. If my activity level were set to lightly active, would that adjustment have been lower?

PAV8888

Yup. The total adjustment at midnight will be less. However, you might start the day with a negative adjustment.

The purpose of the adjustment is to adjust the MFP "I've selected my own activity level" caloric burn expectation with the caloric burn detected by Fitbit.

At any point of time when you look at the adjustment, MFP assumes you will continue your day at your stated activity level.

So if you're setup as sedentary it expects you to continue moving around like a person who usually gets around 3500 to 5000 steps a day. If you're setup as lightly active it expects you to move around like a person who ends up with 5000 to ~8000 Steps. Active ~8 to 12 and a bit, Very active 12 and a bit to a bit less than 16 etc.

The levels are 1.25/1.4/1.6/1.8 x BMR. And that's exactly what MFP expects you to burn every minute (so /1440) from midnight to midnight.

By contrast Fitbit assigns 1.0 x BMR to any minute that it doesn't detect movement or exercise. Well, actually it averages the movements it detects over 5 minutes and assigns an activity factor to the five minute period... at least that's what it used to do when I looked at it years ago ;-) And it assigns BMR when it doesn't sense anything.

bjkoziara

PAV8888 wrote: »

Yup. The total adjustment at midnight will be less. However, you might start the day with a negative adjustment.

The purpose of the adjustment is to adjust the MFP "I've selected my own activity level" caloric burn expectation with the caloric burn detected by Fitbit.

At any point of time when you look at the adjustment, MFP assumes you will continue your day at your stated activity level.

So if you're setup as sedentary it expects you to continue moving around like a person who usually gets around 3500 to 5000 steps a day. If you're setup as lightly active it expects you to move around like a person who ends up with 5000 to ~8000 Steps. Active ~8 to 12 and a bit, Very active 12 and a bit to a bit less than 16 etc.

The levels are 1.25/1.4/1.6/1.8 x BMR. And that's exactly what MFP expects you to burn every minute (so /1440) from midnight to midnight.

By contrast Fitbit assigns 1.0 x BMR to any minute that it doesn't detect movement or exercise. Well, actually it averages the movements it detects over 5 minutes and assigns an activity factor to the five minute period... at least that's what it used to do when I looked at it years ago ;-) And it assigns BMR when it doesn't sense anything.

Okay, now it definitely makes sense! Thanks so much for the detailed explanation.

PAV8888

This sounds wrong. Are you synchronizing Fitbit or Apple watch?

The adjustment is just the difference between the end of day projections for Fitbit and MFP.

If the Fitbit projection to end of day is 2000 and MFP settings expect you to burn 1800, that's a +200 Cal adjustment. If mfp settings expected you to burn 2100, then you would get a -100 adjustment.

If Fitbit/MFP think you burned/will burn 2000 to end of day and they issue a 3000 Cal adjustment, then there is an error that should be explored with support!

beulah81

PAV8888 wrote: »

This sounds wrong. Are you synchronizing Fitbit or Apple watch?

The adjustment is just the difference between the end of day projections for Fitbit and MFP.

If the Fitbit projection to end of day is 2000 and MFP settings expect you to burn 1800, that's a +200 Cal adjustment. If mfp settings expected you to burn 2100, then you would get a -100 adjustment.

If Fitbit/MFP think you burned/will burn 2000 to end of day and they issue a 3000 Cal adjustment, then there is an error that should be explored with support!

Here are some screenshots of yesterday.
Fitbit says that my TDEE was 1883. MFP adjusted total burn to 1923. Am reading this all wrong? My goals are set to maintain on both apps. Since I got my Fitbit (about 2 months ago) I have been going by MFP's weekly NET calorie goal and ignoring Fitbit's #'s. Is that ok?

heybales

@beulah81 - You need to tap and hold on that Fitbit adjustment to see what the actual detailed figures are.

That will tell you if MFP is indeed dealing with the same figure as Fitbit has, or did it fail to sync. Or is time zone off.

And for sure don't try to follow 2 roads to the same destination.
MFP for eating goal and logging, Fitbit for movement goals and logging.

Of course, end of the day figures on review should match pretty close, perhaps some rounding errors.
Fitbit saw 1883, less deficit 500 eating goal 1383, MFP was told 1883, less whatever it thought you'd burn to create extra adjustment to reach same number, less deficit 500.

It may be that MFP isn't a getting a good sync to correct the end of day figures.

For instance, with trackers only syncing a new value when it's 100 higher than prior sync, that's over 1 hr of BMR rate of burn, so you could have that much off at midnight or more actually.
And if MFP didn't get a sync the next day to provide a prior day final numbers - MFP is sitting there with last sync at say 10:30 pm, with last 1.5 hrs at BMR x 1.25 sedentary say.

The details on the adjustment will provide exactly that info.

Then again another idea.
MFP 1450 eating goal, you must be set to maintenance I'm assuming with Fitbit showing almost matching CICO for days.

Fitbit says 1883 daily burn, and MFP math for a 473 adjustment would imply MFP thinks daily burn is 1410.

The 40 cal difference is probably rounding errors.

Or, you've lost weight still, and MFP never saw a 10 lb drop and auto-adjusted your eating goal.
So it stays at 1450 where it was when you weighed more.
But it actually is figuring your daily burn at 1410 because of lost weight.
That's what is used in the math with Fitbit.

Curious to see how your details picture shows that.

heybales

bjkoziara wrote: »

I never thought I would be asking this question, because I thought I knew the answer, but here we are...

I've been using my FitBit Flex 2 to track my steps the last couple of days. I have MFP set to sedentary, as I work a desk job and am not really that active in my day-to-day. Currently set at 2 lbs/week loss because I started at 360 lbs (currently 339).

Today I went for maybe a 10 minute walk at a moderate pace before work, and another 15 minute moderate-to-brisk walk on my break. I've done a little walking throughout the day here and there but my FitBit estimates about 8,000 steps. While this is more active than I normally am, FitBit did a calorie adjustment of -493 calories. This seems excessive for less than 30 minutes of walking around. How accurate is this? I really don't want to eat those back if they are way off, but then I'm also eating at a 1,000 calorie deficit currently, so I guess I probably could? What are your thoughts?

Despite the correct math @PAV8888 showed - I'm thinking that adjustment does seem high.

Since steps is only used to get a distance, it's distance that is used for calorie burn for daily activity level stuff.

How does the total distance on this day compare to other day with no adjustment and normal routine?

Your default stride-length setting which uses gender & height may not be correct for you.
This is very common when you have a lot to lose.
Your stride will length usually as you keep this up losing weight.

Ever walked a known 1/2 to 1 mile distance, and compared what Fitbit got for that chunk of time?
Should be about 1.8 mph pace, so Fitbit can dynamically adjust up and down based on impact of steps to the range you are likely to have walking.

Probably have to input it manually based on that test walk, and do that every 10 lbs lost just in case.

Because about 4K steps is generally taking people out of Sedentary and getting them an adjustment - depending on distance done of course.

So for an extra 4K this day to give an extra almost 500 cal for 25 min of walking - does seem high.

beulah81

heybales wrote: »

@beulah81 - You need to tap and hold on that Fitbit adjustment to see what the actual detailed figures are.

That will tell you if MFP is indeed dealing with the same figure as Fitbit has, or did it fail to sync. Or is time zone off.

And for sure don't try to follow 2 roads to the same destination.
MFP for eating goal and logging, Fitbit for movement goals and logging.

Of course, end of the day figures on review should match pretty close, perhaps some rounding errors.
Fitbit saw 1883, less deficit 500 eating goal 1383, MFP was told 1883, less whatever it thought you'd burn to create extra adjustment to reach same number, less deficit 500.

It may be that MFP isn't a getting a good sync to correct the end of day figures.

For instance, with trackers only syncing a new value when it's 100 higher than prior sync, that's over 1 hr of BMR rate of burn, so you could have that much off at midnight or more actually.
And if MFP didn't get a sync the next day to provide a prior day final numbers - MFP is sitting there with last sync at say 10:30 pm, with last 1.5 hrs at BMR x 1.25 sedentary say.

The details on the adjustment will provide exactly that info.

Then again another idea.
MFP 1450 eating goal, you must be set to maintenance I'm assuming with Fitbit showing almost matching CICO for days.

Fitbit says 1883 daily burn, and MFP math for a 473 adjustment would imply MFP thinks daily burn is 1410.

The 40 cal difference is probably rounding errors.

Or, you've lost weight still, and MFP never saw a 10 lb drop and auto-adjusted your eating goal.
So it stays at 1450 where it was when you weighed more.
But it actually is figuring your daily burn at 1410 because of lost weight.
That's what is used in the math with Fitbit.

Curious to see how your details picture shows that.

You are amazing! I love how you go into nitty gritty of possibilities. Thank you so much!
I did check the time zones. They are set correctly on both ends.
Is this what you mean by detailed picture?
Also, if the issue continues should I adjust my MFP goal to 1410 to get similar #'s on both ends?

beulah81

These are yesterday's figures.

StaciMarie2020

As explained a little bit of the adjustment will always go away overnight because of the different methods used MFP compared to Fitbit. And until you have a history you don't know how accurate Fitbit is for you. So feel free to leave a few hundred calories 'on the table'. That will account for the overnight adjustment and any error.

But also listen to your body. If you feel you need extra on a day where you are more active than usual, then allow extra.

heybales

beulah81 wrote: »

heybales wrote: »

@beulah81 - You need to tap and hold on that Fitbit adjustment to see what the actual detailed figures are.

That will tell you if MFP is indeed dealing with the same figure as Fitbit has, or did it fail to sync. Or is time zone off.

And for sure don't try to follow 2 roads to the same destination.
MFP for eating goal and logging, Fitbit for movement goals and logging.

Of course, end of the day figures on review should match pretty close, perhaps some rounding errors.
Fitbit saw 1883, less deficit 500 eating goal 1383, MFP was told 1883, less whatever it thought you'd burn to create extra adjustment to reach same number, less deficit 500.

It may be that MFP isn't a getting a good sync to correct the end of day figures.

For instance, with trackers only syncing a new value when it's 100 higher than prior sync, that's over 1 hr of BMR rate of burn, so you could have that much off at midnight or more actually.
And if MFP didn't get a sync the next day to provide a prior day final numbers - MFP is sitting there with last sync at say 10:30 pm, with last 1.5 hrs at BMR x 1.25 sedentary say.

The details on the adjustment will provide exactly that info.

Then again another idea.
MFP 1450 eating goal, you must be set to maintenance I'm assuming with Fitbit showing almost matching CICO for days.

Fitbit says 1883 daily burn, and MFP math for a 473 adjustment would imply MFP thinks daily burn is 1410.

The 40 cal difference is probably rounding errors.

Or, you've lost weight still, and MFP never saw a 10 lb drop and auto-adjusted your eating goal.
So it stays at 1450 where it was when you weighed more.
But it actually is figuring your daily burn at 1410 because of lost weight.
That's what is used in the math with Fitbit.

Curious to see how your details picture shows that.

You are amazing! I love how you go into nitty gritty of possibilities. Thank you so much!
I did check the time zones. They are set correctly on both ends.
Is this what you mean by detailed picture?
Also, if the issue continues should I adjust my MFP goal to 1410 to get similar #'s on both ends?

Your eating goal and what MFP thinks you burn based on BMR x 1.25 (assuming sedentary selected) are 2 different things.

You can go through setup again, change an option like activity level, put it back so it thinks you changed something, and save out with correct weight showing.
It should display the screen that shows what it's going to use.
And it should come down to whatever is correct, which may be where it's at.
My math to arrive at 1410 was based on assuming Fitbit got final sync in. But ...

Because that's not the whole problem - that details screen is exactly what you needed.
Last Fitbit sync is not at midnight as it should be.
Though I've also not seen the MFP Calories Burned jump like that, unless there was weight correction between those 2 days shown.

So that's the other issue.
MFP is assuming the time from the sync to midnight as BMR x 1.25, not the BMR level Fitbit would have allowed correction to from when you hit the couch and bed until midnight.

So another reason for higher adjustment.

Time zones can be a reason why final syncs are not happening, or other issues.
You'll have to get a case with MFP or Fitbit in.
Considering MFP is showing syncs the next morning, the ability to get them from Fitbit is occurring - so I'm wondering if Fitbit is sending correct ending info for prior day.
It may be on their end.
Or your account on MFP is slightly jacked up and that type of sync isn't happening. (mine was, and the account issue followed to Garmin syncing too)


2nd set of pictures same issue.

9:25 pm last sync, 21:25, or 2:35 left till midnight, 155 min out of 1440 in day = 10.76% of day left.
10.76% x 1449 MFP estimated daily burn = 155.9 more at MFP rate of burn.
Fitbit so far 1974 + 155.9 MFP rest of day = 2129 total estimated.
Estimated daily 2129 - 1449 MFP estimated = 680 adj

base goal 1450 + 680 adj = 2130 new eating goal

So the math is correct.
The figure from Fitbit is not.

beulah81

Wow. You blew me away once again by your insight. I will take your advice and go through the set up again. Thank you so much!

PAV8888

In any case, if you're trying to maintain your starting eating goal would be your MFP goal WITH adjustments.

So around 2k calories, not 1500.

Personally I would start by eating my full allotment while keeping an eye on my weight trend and only adjusting down if it was clear I was gaining.

I didn't go through the numbers but in the two pictures you're showing MFP at about 1450 Calories at about 9pm and Fitbit claiming you've spent more, and thus MFP giving you a positive adjustment of extra calories to consume.

I would consume them.

I would also keep in mind that if inactive to midnight I would lose about BMR * 0.25 Calories per minute of inactivity. So about 40 to 50 Calories of that adjustment.

Eat all or almost all of the higher figure and guage whether the 40-50 Calories are an issue based on weight trend over time...

beulah81

PAV8888 wrote: »

In any case, if you're trying to maintain your starting eating goal would be your MFP goal WITH adjustments.

So around 2k calories, not 1500.

Personally I would start by eating my full allotment while keeping an eye on my weight trend and only adjusting down if it was clear I was gaining.

I didn't go through the numbers but in the two pictures you're showing MFP at about 1450 Calories at about 9pm and Fitbit claiming you've spent more, and thus MFP giving you a positive adjustment of extra calories to consume.

I would consume them.

I would also keep in mind that if inactive to midnight I would lose about BMR * 0.25 Calories per minute of inactivity. So about 40 to 50 Calories of that adjustment.

Eat all or almost all of the higher figure and guage whether the 40-50 Calories are an issue based on weight trend over time...

Thank you PAV once again for sharing your wisdom and knowledge! You have been a great help in my journey!

nooshi713

bjkoziara wrote: »

PAV8888 wrote: »

It is accurate at midnight.

Your adjustment will decrease if you remain unmoving between the time you saw it and midnight.

Given you said your MFP actively level is set to sedentary, it will decrease at a rate of 0.25 * bmr / 1440 * m where m = minutes of being unmoving to midnight.

MFP assumes a level of activity split evenly over your 24 hours. Just the way they all do their math.

Weird, but I think I understand what you're saying. If my activity level were set to lightly active, would that adjustment have been lower?

Yes because 8000 steps in a day is not sedentary.

heybales

beulah81 wrote: »

Wow. You blew me away once again by your insight. I will take your advice and go through the set up again. Thank you so much!

I want to confirm your screen shots.

It says 6:21 am & 8:30 am on phone time, so those were the last sync screens for the prior day at about 9 pm, right?

And for that first view in the morning, if you enabled negative adjustments you should have a negative to view too.
And then the time of last sync for that same day is ... ?

beulah81

heybales wrote: »

beulah81 wrote: »

Wow. You blew me away once again by your insight. I will take your advice and go through the set up again. Thank you so much!

I want to confirm your screen shots.

It says 6:21 am & 8:30 am on phone time, so those were the last sync screens for the prior day at about 9 pm, right?

And for that first view in the morning, if you enabled negative adjustments you should have a negative to view too.
And then the time of last sync for that same day is ... ?

Yes, the am times that you see are screenshots I took the following day. The 9 pm sync screenshots are also taken the following day. I did not enable negative adjustments thinking I didn't need to do that because I regularly take more steps than my chosen sedentary activity setting and by synching Fitbit in the evening around 9 ish was fine. I thought I can enable them on days when I am truly sedentary. So Fitbit and MFP don't "talk" to each other automatically at midnight unless I manually sync at 11:59 pm or have negative adjustments enabled?

beulah81

I went ahead and enabled negative adjustments. Thank you for bringing it up!

heybales

The negative setting would merely allow you to do that morning look and have an adjustment there to see details on. Has nothing to do with time of syncs.
You can set it, it's not involved in any math unless you do less than your activity level.
If you always do more - it'll never matter.

Those details might spell out your last sync was still the prior day at 9:30 pm or such.
So no wonder the figures are off.

Not sure why - Fitbit used to still send updates every 100 calorie higher than last sync.
And if the account had no data from the device to see that go up, it still used basically a barely above BMR rate of burn.

For you, 100 calories at BMR rate would take 124 min.

So even if you went to bed @ 9:30pm, and device didn't sync with phone app (do you sync with phone?), there should have been another account sync about 11:30 pm, and 1:30, ect.

Perhaps they stopped doing that, figuring everyone syncs their device to phone often enough to not need those types of syncs.

Are you manually syncing your device?

beulah81

heybales wrote: »

The negative setting would merely allow you to do that morning look and have an adjustment there to see details on. Has nothing to do with time of syncs.
You can set it, it's not involved in any math unless you do less than your activity level.
If you always do more - it'll never matter.

Those details might spell out your last sync was still the prior day at 9:30 pm or such.
So no wonder the figures are off.

Not sure why - Fitbit used to still send updates every 100 calorie higher than last sync.
And if the account had no data from the device to see that go up, it still used basically a barely above BMR rate of burn.

For you, 100 calories at BMR rate would take 124 min.

So even if you went to bed @ 9:30pm, and device didn't sync with phone app (do you sync with phone?), there should have been another account sync about 11:30 pm, and 1:30, ect.

Perhaps they stopped doing that, figuring everyone syncs their device to phone often enough to not need those types of syncs.

Are you manually syncing your device?

Thank you for clarifying the negative adjustment function.
I am manually synching Fitbit on the phone app, sometimes during the day and always in the evening around 9.

heybales

Ah, they must have changed their process then, when I manually synced my Zip years ago, they still did the syncs at 100 cal higher at their rate of burn.

I don't recall an account setting for that manual sync, but I know I didn't enable bluetooth or use the app, nor used since, to know if it's an app setting that says no auto-sync. Just synced to BT dongle on computer.

But they must know now if you are manual syncing and not doing those account to account syncs.

You'll always have this issue then, days never being updated.
The records on MFP for viewing amounts and differences won't be right.

You can still do the math pav mentioned about how much you'll lose in the evening so at least your eating goal is correct.

Merely your history will be wrong on MFP.
I think Fitbit has better data views anyway.

If you want Fitbit records to be complete, probably have to figure out how to have auto-sync on. And then just not use it if not desired all day.

beulah81

heybales wrote: »

Ah, they must have changed their process then, when I manually synced my Zip years ago, they still did the syncs at 100 cal higher at their rate of burn.

I don't recall an account setting for that manual sync, but I know I didn't enable bluetooth or use the app, nor used since, to know if it's an app setting that says no auto-sync. Just synced to BT dongle on computer.

But they must know now if you are manual syncing and not doing those account to account syncs.

You'll always have this issue then, days never being updated.
The records on MFP for viewing amounts and differences won't be right.

You can still do the math pav mentioned about how much you'll lose in the evening so at least your eating goal is correct.

Merely your history will be wrong on MFP.
I think Fitbit has better data views anyway.

If you want Fitbit records to be complete, probably have to figure out how to have auto-sync on. And then just not use it if not desired all day.

Heybales, thank you so much for taking the time to help me out with this issue! I do have auto-sync enabled on the Fitbit app but I don't think it's making any difference on the MFP's end.

heybales

Only difference it would make to MFP is getting newer figures from Fitbit.
The setting for auto-sync is between the Fitbit & the phone.

But if that is enabled - you made it sound like you have to manually do it, doing it in morning, and then in evening you mentioned.
Does that mean Fitbit app isn't syncing to your device unless you manually do something?
Or that's when you care to notice to see the goals at that time?

Now I'm in troubleshooting mode.
Either the MFP account is hosed in a way not to get the updates from Fitbit.
Or Fitbit is hosed in some way not syncing device info frequently.

More options to try if it appears Fitbit only issue.

beulah81

heybales wrote: »

Only difference it would make to MFP is getting newer figures from Fitbit.
The setting for auto-sync is between the Fitbit & the phone.

But if that is enabled - you made it sound like you have to manually do it, doing it in morning, and then in evening you mentioned.
Does that mean Fitbit app isn't syncing to your device unless you manually do something?
Or that's when you care to notice to see the goals at that time?

Now I'm in troubleshooting mode.
Either the MFP account is hosed in a way not to get the updates from Fitbit.
Or Fitbit is hosed in some way not syncing device info frequently.

More options to try if it appears Fitbit only issue.

It does sync if I wait long enough but it does glitch more often than not, hence why I do it manually. I admire and appreciate your drive to figure this out!