Negative calorie adjustments too much?

simcon1

I just switched from a Fitbit Ionic this weekend, and don’t understand why there’s such a discrepancy between what I got in the way of calories with the Fitbit vs. Garmin. I think my activity level is set to active, and I often run. Fitbit would send the exercise calories over all in one figure, and it made sense and worked for me to lose 25 lbs fairly predictably. The Garmin sends over a separate exercise activity, and then the calorie adjustment, which is often negative, even with 20K+ steps. I know something happens so it doesn’t double count my runs, but it still seems pretty low for exercise calories even taking that into account. Has anyone else experienced this?

heybales

Fitbit never sent exercise calories over.
It sent total daily burn calories over - MFP did the math for difference between that and what it expected you to burn with no workouts.

That Calorie Adjustment could have been totally increased daily activity, no workouts, or workouts and less daily activity than it expected.

Same with Garmin, except Garmin does send the optional workouts over, along with daily burn which is the workouts and everything else you burn each day.

MFP is doing the math and creating the adjustment. Not the tracker.

Since MFP knows about the workout separately - the adjustment is going to be less, perhaps even negative, since both are added to your eating goal.

So if you have your level set to Active because you run - that's incorrect.
MFP activity level is your non-workout level - workouts added.
You can do what you are doing - but expect negatives when you don't run as much.

Click on the Calorie adjustment for more info and report all the actual figures, will help to see if anything actually wrong.

Are the actual eating goals about the same on equal days?

Did you manually tweak your stride length on the Fitbit, but haven't on the Garmin yet?
There's one potential.

Have you compared the calorie burn for the runs between the devices?
There's one potential.

Have you compared the steps and more importantly distance for average day between them?
Distance is what causes the daily calorie burn estimate.

simcon1

heybales wrote: »

Fitbit never sent exercise calories over.
It sent total daily burn calories over - MFP did the math for difference between that and what it expected you to burn with no workouts.

That Calorie Adjustment could have been totally increased daily activity, no workouts, or workouts and less daily activity than it expected.

Same with Garmin, except Garmin does send the optional workouts over, along with daily burn which is the workouts and everything else you burn each day.

MFP is doing the math and creating the adjustment. Not the tracker.

Since MFP knows about the workout separately - the adjustment is going to be less, perhaps even negative, since both are added to your eating goal.

So if you have your level set to Active because you run - that's incorrect.
MFP activity level is your non-workout level - workouts added.
You can do what you are doing - but expect negatives when you don't run as much.

Click on the Calorie adjustment for more info and report all the actual figures, will help to see if anything actually wrong.

Are the actual eating goals about the same on equal days?

Did you manually tweak your stride length on the Fitbit, but haven't on the Garmin yet?
There's one potential.

Have you compared the calorie burn for the runs between the devices?
There's one potential.

Have you compared the steps and more importantly distance for average day between them?
Distance is what causes the daily calorie burn estimate.

This is all really helpful, thanks! It gives me a lot to troubleshoot, although I still find it confusing. I set my activity level to “active” because I still get about 10K-15K steps in addition to my running, and with the Fitbit, even when I didn’t run, if I got 10K steps I wouldn’t get a negative adjustment. Today I did the same (no run) and ended up with about -200 cal. The distances should be all pretty similar in the steps—I’ve been doing the same kind of walking. I didn’t manually adjust stride with either the Fitbit or Garmin, but maybe I should with the Garmin? The Fitbit did that with my GPS runs, I think. Does the Garmin also do the same?

I could set my activity level lower, so my tracked calories would import without an estimate, but I’m worried that will drop my calorie goal even lower. I was pretty steady with losing what I was supposed to (1 lb/week) so I didn’t want to eat less because I don’t want to lose faster. I’m training for a half marathon, so I don’t want to underfuel (I’ve dropped 4 lbs since I switched to Garmin on Sat, and while I know that is a variety of things, undereating can trigger a “whoosh” for me.)

Thanks again for taking the time to point me to these possibilities!

heybales

Changing your activity level has no effect on the daily goal - in the end.
Lowering it doesn't stop an estimate from being done.

It does bear on how big the adjustment is compared to the starting calorie goal.

Lower activity level - smaller base goal - bigger adjustment.
Higher activity level - higher base goal - smaller adj.

The day ends the same either way.

The problem with the latter - which you may not have had - if you go to be early, then MFP assumes rest of the day is at the high activity level of calorie burn.
The next day on syncing the Fitbit will inform MFP it was actually BMR rate of burn, not what MFP used of BMRx1.6 for active.
If you go to bed near midnight and don't finish up meeting your goal till later - no big deal.
If you hit the couch at 6pm and eat to goal, and bed at 8pm - that 6 hrs of BMRx 0.6 could be rather high number you've eaten to that will be taken away the next morning.

Either way works - either way really works better if you have an idea of what the totals are going to be anyway - and after a few days, probably do.


The 2 devices may use one of the common formula's for calculating a stride length - then again maybe not.

If steps are similar - what is the distance for the similar days - since that's what calories is based on?

The Fitbit, and Garmin, will not adjust your stride length setting based on runs (some Garmin models have the option, perhaps Fitbit does too - if GPS can really be that accurate, and it's not).
But really doesn't matter, because running stride is not walking stride - 2 settings for reason.
Also, with HR-based calorie burn being used during workouts, like running, the distance doesn't matter for the burn - though Garmin I know uses HR and distance based calorie burn combo. Never heard Fitbit claim that, though it would be smart - because for actual step-based activities - distance and pace and weight is more accurate formula than HR calorie burn.

If you like the math, the 2nd section here explains what happens on all the fitness trackers, though it's about the Fitbit.
Only difference is the Garmin took the option to send the workouts over to MFP, so tad difference in math ending in same result.

http://community.myfitnesspal.com/en/discussion/10098937/faq-syncing-logging-food-exercise-calorie-adjustments-activity-levels-accuracy

simcon1

Ok, thanks so much for this and for the link—I think I’m getting closer to figuring this out. Garmin seems to be calculating my total burn lower than Fitbit: on otherwise comparable days, they both sent about 2600 calories to MFP where the Fitbit day was 18K steps including a 4 mile run, and the Garmin was 20K steps including a 4.5 mile run, which should probably have been more. I know I’m being very imprecise, but I don’t have my Fitbit distances handy at the moment because I deleted the app (though they should still be there if I log in on the site, but it’s harder to get there and navigate on my phone) It seems a first step would be manually entering my stride length, so I’ll find a treadmill to do that soon.

I didn’t really get into the math with the Fitbit because it was pretty reliable over a couple of years (not all in deficit) of similar kinds of activity. But I was down another pound this morning and have my long run today, and would rather not arbitrarily eat more day-to-day when I really found the previous device/system so helpful! I could change my deficit to .5 lbs I guess, but I’d just rather more precision. And not too much math if possible.

The other confusing thing to me is that Garmin Connect takes my goal from MFP (with deficit already included), but then the +activity is a different number and so it gives me a different (higher) number of calories remaining on any day. There must be a basic reason for this, but I am feeling a bit thick at the moment—any tips on that?

I sync a lot and am reasonably active in the evenings (not necessarily workouts, but otherwise) and don’t go to bed particularly early (usually 10-11 pm) so I don’t think the timing of syncing is particularly likely to be the issue.

heybales

I could never figure out the Garmin side - their multiplier was all over the place for the activity extra.

But even Fitbit eating goal was different - until the end of the day.

So MFP for eating goals & tracking, Garmin (or Fitbit) for activity goals & tracking.

I don't think it's the timing - it may be the way they both estimate to the end of the day, Fitbit had it's own method too that didn't match MFP.

I never thought to figure out that angle by looking at time of sync, seeing the figure all over the place I decided it didn't really matter.

Sounds like the Garmin saw more steps, but about equal distance, or gave the greater distance smaller calorie burn.