Garmin watch active calories vs activity calories

I have been paying a great deal of attention to the calorie calculations provided by my Garmin watch recently.

I have noticed something interesting.
On some days, the “active” calories (from walking, weight training, whatever) before a bike workout seem to be heavily discounted after (by up to 50%) after the workout.

E.g., the watch tells me I have burned 200 active calories. I then do a 1000 kj indoor ride.
Now the watch tells me I have burned 1100 active calories.

Assuming this isn’t some sort of bug, I can see two reasons for programming it this way.

  1. Resting calories (RMR, aka desk job calories) are held constant based on your weight, but are actually several hundred calories above BMR (aka bedridden calories). Therefore, active calories might be getting adjusted to account for fluctuations in your NEET level.

E.g., you exercise more on weekends but otherwise do less other activity, so they are compensating for your RMR being lower that day.

  1. They are making adjustments for the expected decline in NEET that a high level of formal exercise causes.

E.g., you did an extra X calories of formal exercise, which science tells us means will only increase your TDEE by X / Y calories.

Does anyone have insights on why Garmin has done things this way?

My personal experience with Garmin calories is that they always end up levelling out at the end of the day.

For example, if I do very few steps/movement in the morning before a training ride, I might see an adjustment of -500 cals, however as the day goes on and I get closer to bed time this evens itself out to have less of an impact. So basically I see it as assuming I will be completely sedentary all day. And conversely, if I do a bunch of movement and training then it will give me more calories and hold those until the the end of the day.

Always evens itself out come bed time.

This is an interesting observation, and something that scientist Tim Noakes brings up in an interview he did on the FastTalk podcast. Although I don’t tend to agree with many things Noakes says, he makes an often overlooked point about the “loss” of TDEE calories that he thinks should be considered for endurance athletes with long training sessions.

For example, ponder that an athlete has a TDEE of 2,400 calories with no exercise. This is the amount the body requires just to maintain weight and homeostasis with work, walking, RMR, etc. Then add a 4 hour training ride to the equation. Logically, one would think to take whatever mechanical work was produced in kilojoules and adding this to 2,400 to get the total energy expenditure for the day. For ease of addition, the athlete does 4 hours at 166 watts which equates to 600 calories per hour, times 4 hours, giving a total expenditure of 2,400 calories on the bike. Add this to the TDEE, and the calorie allotment for the day should be 4,800.

However, this is where Noakes makes an interesting point. Since the training session was so long, the athlete may actually “lose” or “miss out” on some of the TDEE calories that would be burnt normally during this time. While biking for 4 hours, no other activity is performed. No steps are being taken, no gardening is getting done, etc. If the athlete normally takes 15,000 steps during the day, spending 4 hours on a bike may result in less walking because less time is available. Likewise, while a person may normally do 2 hours of gardening on weekends, taking part in long training sessions may “eat in” to gardening time, and less calories are burned through that activity.

In summary, long bouts of exercise may burn a lot of calories, but the time spent exercising may reduce the amount of time left for other NEAT activities during the day. Thus, TDEE may have to be adjusted, and simply adding the kilojoules from your ride to your “normal” TDEE may lead to an overestimation of calories burned during that day. I have no clue if this is what the Garmin device is accounting for, but maybe?

MASSIVE DISCLAIMER - this is just a reference to what Tim Noakes said on a podcast. Personally, I find that if I put my weight and “activity level” into some TDEE calculator online and then add my kilojoules from riding to get my total daily expenditure, I feel like crap the next day. I’m 63 kg (139 lbs), I take 15-20k steps per day, and I ride between 15-18 hours a week. On any given day, I find that to maintain my weight and capability to perform during workouts, I eat anywhere between 400-800 calories MORE than whatever my TDEE + ride kJ’s ends up being. Example - my TDEE in any calculator is roughly 2,250-2,400 calories. Add 2,000 calories for 2.5 hours of riding at 220 watts, and I “should” eat 4,400 calories at most. If I do that consistently, I feel like a soaking wet sack of potatoes. Normally, on such days, which are most days, I eat at least 5,000 calories to feel full and happy. I don’t gain weight, and my training is productive. Just wanted to make that disclaimer to prevent anyone from just taking Noakes’ thoughts as gospel and eating less because some calculator told them to. Eat when you’re hungry, people.


So far as I can work out, your Garmin resting calories just tick up through the day at a continuous rate, based on your basic physical statistics. Then for any activity you formally record it adds the calories you burned, but discounts the resting calories you would have accrued during that time. If you look at the activity record, you will see that the total calories are broken into active calories and ‘resting’ calories (which is a bit of an odd way of putting it, but still…). Then it does the same for any activity that it thinks it’s detected but you didn’t formally record.

For example, yesterday I did an activity that totalled 1646 calories, which it breaks down into 1460 active and 186 ‘resting’.
For the full day, it says I had 1633 active calories, so that’s the ones I recorded from my activity, plus 173 from walking round the house and doing a bit of gardening.

This is not what I am talking about.

The active calorie total does not match the total of active calories from activities. But only sometimes.
Today it was actually slightly higher, probably because I walked a lot, exceeding my auto goal for steps.

1 Like

The idea that your TDEE would be higher if you weren’t cycling seems silly on its face… though I can see how it would affect NEET outside of cycling.

I’m not clear what you are saying, then. Using your example:

  • You have 200 active calories; you do an activity that is 1000 calories; you now have 1100 active calories
  • your 1000 activity calories are made up of 900 active calories and 100 ‘resting’ calories
  • your 1100 active calories are made up of the 200 you’d accumulated prior to the activity + 900 from the activity.

Again, no.

To be more specific, my watch told me I did about 250 active calories from unrecorded activities (in this case, an hour of weight training) before I did a 1021 KILOJOULE (meaning 1021 ACTIVE calories, not TOTAL calories) ride, which somehow added up to 1167 total active calories at the end of the day.

I am curious where the missing 110 calories went.

I reckon its because the 250 active calories were based on if you’d have remained active for the rest of the day. Like I said, if I do a bunch of steps first thing it might raise my total calorie goal for the day but if I then am sedentary for the rest of the day this reduces/evens out a bit. So it gave you 250 based on your “activeness” but later decided you weren’t doing enough to maintain it.

Just to add, I find it gives you 100% of power recorded calories and adapts the TDEE calories around it.

Example: If I haven’t synched my watch and do a 600 calorie workout but it hasn’t “seen” my steps/activeness yet then it often will give me a -400 or so, until I synch and it realises I haven’t just been laying in bed all day :smiley:

Ah - with you now - that’s clearer than your original post - I’d missed the slide from talking about calories into talking about kilojoules, which was where the misunderstanding came in.

For what it’s worth, yesterday it gave me around 95 active calories before I did an 867 kj ride.

At the end the day, my active calories were 838! So it doesn’t count all of your activity calories.

I noticed my step counts were about 60% of my goal for that day, so I wonder if the Connect isn’t just pulling active calories out to compensate for a step deficit.

Though the Steps page actually gives a different calorie count than my Calories page.

I understand why they might be doing this, I just want to know exactly why and how it works.

Also, activity calories are not a prediction for the rest of the day. You can tell this from the way it increases fairly linearly with activity. Even resting calories is an hour by hour trickle until midnight. I’d almost prefer it wasn’t.

Interestingly, after a 800-kj ride, the app initially thinks I’ve burned about 500 calories that day (because the heart rate monitor reads low during cycling), which it then changes to just under 800 based on the data from TR.

What I think is happening is it is giving you the 867 calories for the ride, but the adjusted calories for the end of the day are still negative because your total calorie expenditure is less than the original calculated goal.

Example: Your standard daily goal is 1850, which is made up of 1650 BMR and NEAT 200. You do 800 cals riding: 2650. But your 1850 also includes “moderate” activity levels (200 cals), which you didnt reach. Therefore the final total has to be reduced to compensate for that. So your “other/NEAT” whatever it is they use to balance the equation is still negative based upon the overall daily calculation.

This is not what I see. As said, if I do 2500 steps before an activity which are not synched it will give me a -3-500 calories post workout until the steps are snyched. If i sync it prior it doesnt.

I’m not sure what you mean about the HR reading low during cycling. I haven’t experienced that either.

Additionally, I think a lot of this behaviour depends upon how you have all of the various aspects of the watch and Garmin Connect setup. Which I expect can be very different for everyone.