Exercise/calorie accuracy

[azfalcons]

Ok so tonight I exercised and I used both my Apple Watch and my fitness from the iPhone to measure what I did. I set the Apple Watch to outdoor walk and the my fitness was set to run. I walked a pretty fast walk then occasionally would do a good steady jog as far as could...did this off and on. Here are the end results. I'm curious which is more accurate and what should I rely on when counting calories and exercise for dieting.

Apple Watch
Time 31.56
Distance 2.26 miles
Active calories 219
Total calories 284

Apple Watch only syncs the 219 calories to my fitness...why?

My fitness app
Time 32.16
Distance 2.31miles
Calories 373

Which is more accurate and better to use?

THANKS

 

[exxxviii]

I have analyzed the AW and its calorie calculations many different ways and in a number of its different workout options. I have found it completely untrustworthy. If you search my user ID and the they keyword calorie, you should be able to find them and other forum posts that align. I tried to keep it based on objective data and analytics, but there is a lot of typical forum AW-defensive bickering as well.

The nut is, until someone can demonstrate with objective analytics that the AW calorie numbers are reliable, then I count them as junk. (So far, I have shredded any attempt to defend the AW calorie algorithms.)

Trust the MFP numbers over the AW. Most folks are in the range of 150 calories per mile. You walked 2.3 miles. Simple math = MFP wins.

 

[azfalcons]

Ok thanks. I thought it was odd everything was off. From distance, time etc. I'll stick with MFP and keep the 373 burned. Thanks

 

[BarracksSi]

Ok thanks. I thought it was odd everything was off. From distance, time etc. I'll stick with MFP and keep the 373 burned. Thanks

If you're counting calories to lose weight, use the lower number, no matter which app it comes from.

If you use multiple fitness devices at the same time and compare their results, you would be lucky if they agreed, especially across brands and even more so across different measuring methods. The best way to learn how your body reacts to exercise is, still, getting yourself tested at a sports lab.

As far as the AW is concerned, I don't really care these days. Back when I was working within a calorie budget, it was in 2009/2010, and I was using an app on my iPhone 3G. The Fitbit and AW hadn't been invented yet, and I didn't pick up my Garmin until a year later.

The app had pretty decent estimates for calories and exercises, I figured, but I still purposely underestimated my exercise figures -- and overestimated my food intake -- to ensure I stayed well within my budget. Ka-ching: 30 lbs in six months, right on schedule.

 

[azfalcons]

I'm not worried about the calories. I've lost over 39 pounds with MFP and just started playing with be AW this week. If MFP says I lost 373 in sticking with it. It still shows I have 840 remaining for the day, just over 700 if I use AW. Thanks for your input though.

So I decided to go with the AW total burned of the 284. I've read the Apple Watch will get more accurate be more you use it. Seeing as I'll wear it daily I'll see how things go

 

[exxxviii]

Ok thanks. I thought it was odd everything was off. From distance, time etc. I'll stick with MFP and keep the 373 burned. Thanks

Distance will always be a little off with all devices and apps. But yours got pretty close and are within expected tolerances. The problem is that Apple's calorie calculation algorithms are incorrect. Apple will likely fix them in future watch OS releases, but for now, they are wrong.

 

[zhenya]

It's highly unlikely you burned 373 calories from 32 minutes of walking/light running. The somewhat high end of the range accepted in the running community is about 100 calories per mile. I would take the Apple number which sounds completely reasonable.

Note that some apps calculate your calorie burn during exercise and include your basal metabolic rate (the calories you'd burn doing nothing over that time) while others ignore those calories and only count the ones you burn over and above your BMR. (You can see this difference in the numbers Apple reports; it's not clear how the other app is calculating). Either way though, 373 calories is way too high for 30 minutes of light exercise.

 

[exxxviii]

I've lost over 39 pounds with MFP and just started playing with be AW this week.

This made me think of one more thing... double-check your biometric data (height, weight, age, sex) in both the AW and MFP. These are critical inputs into the calorie calculation algorithms.

Second, here is a simple little illustration to show some of the internal problems with the AW calorie numbers. According to your data, you burned 65 BMR calories in about 32 minutes. That extrapolates to a daily BMR of 2,931 calories. That is highly unlikely, unless you are a hummingbird. (Apple used to show the daily BMR calories in the Activity app, but they removed it. You can still calculate it, and it appears reasonable. But the higher BMR numbers are still present in the workout calculations.)

 

[azfalcons]

My numbers are set up just fine. You don't think someone can burn 300 plus calories by jogging for 39 minutes. I've done it at the gym so how is running outside different

 

[exxxviii]

You don't think someone can burn 300 plus calories by jogging for 39 minutes. I've done it at the gym so how is running outside different

Yes, though it depends on your age, weight, and pace. There are tons of calorie calculators on the web and numerous scientific articles and studies published on the web. So, check out a few and enter your information to get a collection of input metrics. Running burns a little more than walking. Men burn more than women. Heavy people burn more than light people. To normalize the information, make sure that everything is either incremental burn (just the additional from the activity) or total burn (activity plus the BMR calories). AW likes to give net burn, but most other devices give numbers in total burn.

I tend to look at everything in terms of total burn, because that is the number that matters most to me when I am managing calorie intake during an extended event or for the day as a whole.

Plus, just based on my simple calculation above about AW's active calorie burn and BMR inconsistency, I am pretty sure that Apple's active calorie numbers are wrong. That elevated BMR factor is present across all activities. So, you either have to choose to believe that the AW total number is correct, but the active calories are incorrect. Or, you can believe the AW active calories but then know that the total calories must be decreased to be accurate.

 

[azfalcons]

Yea I know they aren't right. They seemed a little low to me....AW calories. I'll just keep using it and see what results occur and if they ever improve as Apple says they will. Like I said I've lost weight and had great success with MFP so I'll probably just stick with that

 

[zhenya]

Every calorie calculator is an estimate. 100 calories +/- about 25% per mile is about as good as you are going to get.

 

[BarracksSi]

My numbers are set up just fine. You don't think someone can burn 300 plus calories by jogging for 39 minutes. I've done it at the gym so how is running outside different

What gave you the calorie number at the gym?

 

[exxxviii]

Every calorie calculator is an estimate. 100 calories +/- about 25% per mile is about as good as you are going to get.

I would recenter and open the range. Age and weight play a huge factor. A middle-aged 220 lb. man could be well above 150 calories/mile. On the other end, a lean guy in his 20s may be around 100 calories/mile.

 

[zhenya]

I would recenter and open the range. Age and weight play a huge factor. A middle-aged 220 lb. man could be well above 150 calories/mile. On the other end, a lean guy in his 20s may be around 100 calories/mile.

The 100 calorie estimate is about the center of the range for someone fit enough they can run a couple of miles, even slowly. There are certainly people on the higher and lower ends of that, (Garmin, using paid, proprietary algorithms that are considered the best available typically has me under 70 calories per mile when I run easy), but the quoted number should cover the heart of the curve. Although we all think we are an edge-case.

 

[exxxviii]

Garmin, using paid, proprietary algorithms that are considered the best available typically has me under 70 calories per mile when I run easy

I agree that Garmin's licensed Firstbeat algorithm is probably the best. I am recovering from an injury and ran an easy 3.3 miles today (avg. HR was 137). My Garmin 910XT with HR gave me 155 calories/mile. 70 calories/mile seems incredibly low, unless you are female and weigh around 100 lbs. That is lower than a study by Syracuse early last decade that produced some of the lowest numbers I have read. (Average male burned 124 calories/mile running & 88 walking, and their average female burned 105 and 74.)

Age and weight drive the range very broadly, and people at the higher end of the calorie consumption spectrum are still well within the capabilities of someone who walks and jogs for exercise. (For example, Map My Walk's calculator estimates 230 calories/mile walking for someone at 300 lbs. and 6' tall-- just like a friend of mine.) 100 calories/mile is toward the low end: people who are younger, weigh less, and are more fit. That probably represents a solid subset of the running community, but it is not representative of the population engaged in calorie tracking.

The subjects in the Syracuse study above were age 18-30. All were recreationally active with no orthopedic limitations. The study goes into detail about their body fat and V02 Peak, etc. But in summary, they were all relatively fit. You might want to double-check the biometrics in your Garmin.

 

[x-evil-x]

my numbers are way off. shows my workout average at around 80-87 usually during weight lifting. its more like 110-160 average should be 135 or so which throws off total calories burned I'm assuming(on the watch isn't this how it calculates it? hr)?
If I'm on the stair stepper or doing somethings with my arms up and not moving much its accurate and reads 128 average for 30 minute stair stepper.
2 hr weights workout is 522 cals
30 min stair stepper is 290.
not the most accurate at all.
if i look at the the cals burned from the stair stepper the workout should be closer to 1160 cals burned

 

[zhenya]

I agree that Garmin's licensed Firstbeat algorithm is probably the best. I am recovering from an injury and ran an easy 3.3 miles today (avg. HR was 137). My Garmin 910XT with HR gave me 155 calories/mile. 70 calories/mile seems incredibly low, unless you are female and weigh around 100 lbs. That is lower than a study by Syracuse early last decade that produced some of the lowest numbers I have read. (Average male burned 124 calories/mile running & 88 walking, and their average female burned 105 and 74.)

Age and weight drive the range very broadly, and people at the higher end of the calorie consumption spectrum are still well within the capabilities of someone who walks and jogs for exercise. (For example, Map My Walk's calculator estimates 230 calories/mile walking for someone at 300 lbs. and 6' tall-- just like a friend of mine.) 100 calories/mile is toward the low end: people who are younger, weigh less, and are more fit. That probably represents a solid subset of the running community, but it is not representative of the population engaged in calorie tracking.

The subjects in the Syracuse study above were age 18-30. All were recreationally active with no orthopedic limitations. The study goes into detail about their body fat and V02 Peak, etc. But in summary, they were all relatively fit. You might want to double-check the biometrics in your Garmin.

My biometrics are correct and I've talked to Garmin on numerous occasions about it. They believe it to be correct. I personally add about 10-20% which splits the difference with the other system I use to track; although both are under 100 calories/mile for me. But yes, I'm on the thin/fit end of the curve. It seems to be about right as I can hold/lose/gain weight using these systems and small adjustments to my calorie intake.

 

[exxxviii]

shows my workout average at around 80-87 usually during weight lifting. its more like 110-160 average should be 135 or so which throws off total calories burned I'm assuming(on the watch isn't this how it calculates it? hr)?

You have two separate issues in this post. First is heart rate accuracy during weight lifting. There are a number of threads on this. The consensus is that the AW does not get good HR during weight lifting because gripping tightly may interrupt the watch's ability to see blood flow. The fix is to use a chest strap during weight lifting.
https://forums.macrumors.com/threads/apple-watch-hr-not-accurate.1959051/
https://forums.macrumors.com/threads/weight-lifting-and-apple-watch.1881591/
https://forums.macrumors.com/threads/people-that-workout-with-the-apple-watch.1892723/#post-21465123

30 min stair stepper is 290...
if i look at the the cals burned from the stair stepper the workout should be closer to 1160 cals burned

The AW's calories may be low for a 30 minute stair stepper workout, but 1,160 is extremely high. Your stair stepper numbers are probably incorrect. You might want to double-check that it has your correct biometric information and is capturing your HR accurately.

If you really hit 1,160 calories, then you were burning calories at the extraordinary rate of 39 calories per minute. It takes a moderate aerobic level of effort to hit 10 calories/minute. When I ride an indoor bike trainer at a pretty intense level of effort (250 watts, average), I am only burning about 17 calories per minute. My highest sustained effort, when I was riding a 20 minute FTP test interval on the bike, was 20 calories/minute. I cannot fathom what it would feel like to burn 39 calories/minute for 30 minutes.

 

[BlueMoon63]

I agree that Garmin's licensed Firstbeat algorithm is probably the best. I am recovering from an injury and ran an easy 3.3 miles today (avg. HR was 137). My Garmin 910XT with HR gave me 155 calories/mile. 70 calories/mile seems incredibly low, unless you are female and weigh around 100 lbs. That is lower than a study by Syracuse early last decade that produced some of the lowest numbers I have read. (Average male burned 124 calories/mile running & 88 walking, and their average female burned 105 and 74.)

Age and weight drive the range very broadly, and people at the higher end of the calorie consumption spectrum are still well within the capabilities of someone who walks and jogs for exercise. (For example, Map My Walk's calculator estimates 230 calories/mile walking for someone at 300 lbs. and 6' tall-- just like a friend of mine.) 100 calories/mile is toward the low end: people who are younger, weigh less, and are more fit. That probably represents a solid subset of the running community, but it is not representative of the population engaged in calorie tracking.

The subjects in the Syracuse study above were age 18-30. All were recreationally active with no orthopedic limitations. The study goes into detail about their body fat and V02 Peak, etc. But in summary, they were all relatively fit. You might want to double-check the biometrics in your Garmin.

I'm just now believing everything you write

You have two separate issues in this post. First is heart rate accuracy during weight lifting. There are a number of threads on this. The consensus is that the AW does not get good HR during weight lifting because gripping tightly may interrupt the watch's ability to see blood flow. The fix is to use a chest strap during weight lifting.
https://forums.macrumors.com/threads/apple-watch-hr-not-accurate.1959051/
https://forums.macrumors.com/threads/weight-lifting-and-apple-watch.1881591/
https://forums.macrumors.com/threads/people-that-workout-with-the-apple-watch.1892723/#post-21465123

The AW's calories may be low for a 30 minute stair stepper workout, but 1,160 is extremely high. Your stair stepper numbers are probably incorrect. You might want to double-check that it has your correct biometric information and is capturing your HR accurately.

If you really hit 1,160 calories, then you were burning calories at the extraordinary rate of 39 calories per minute. It takes a moderate aerobic level of effort to hit 10 calories/minute. When I ride an indoor bike trainer at a pretty intense level of effort (250 watts, average), I am only burning about 17 calories per minute. My highest sustained effort, when I was riding a 20 minute FTP test interval on the bike, was 20 calories/minute. I cannot fathom what it would feel like to burn 39 calories/minute for 30 minutes.

There is definitely some truth to weightlifting and losing HR but I don't buy that the consensus is that it doesn't do a good job. I've had some issues when doing severe wrist bending but the number of losses per hour might be 2-3 times.

Today, I did a heavy (5x5) and (4x8) chest and tricep workout with severe bending on dips and hex presses and everything worked as expected. Attaching the file to show how often a pulse was lost.

Can't disagree with you that a heart rate strap might be more accurate but I'm not a fan of straps around my chest and when I did use one, it would slide or spike my HR to 200 at random times or drop to zero.

If I am good at anything, it is weightlifting with the Apple Watch and all my charts are similar to the attachment. The heart rate losses in this hour plus workout is insignificant when compared to the full workout.

I know by now that I must be the luckiest Apple Watch wearer.

 

[exxxviii]

If I am good at anything, it is weightlifting with the Apple Watch and all my charts are similar to the attachment. The heart rate losses in this hour plus workout is insignificant when compared to the full workout.

Very cool. I have not weight lifted with my AW, so I just take the forum complaints at face value. I switched my weight lifting workouts to swimming a couple years ago when I had an injury that prevented me from doing pretty much anything else. Since then, I have kind of taken to swimming and kept it up.

 

[x-evil-x]

You have two separate issues in this post. First is heart rate accuracy during weight lifting. There are a number of threads on this. The consensus is that the AW does not get good HR during weight lifting because gripping tightly may interrupt the watch's ability to see blood flow. The fix is to use a chest strap during weight lifting.
https://forums.macrumors.com/threads/apple-watch-hr-not-accurate.1959051/
https://forums.macrumors.com/threads/weight-lifting-and-apple-watch.1881591/
https://forums.macrumors.com/threads/people-that-workout-with-the-apple-watch.1892723/#post-21465123

The AW's calories may be low for a 30 minute stair stepper workout, but 1,160 is extremely high. Your stair stepper numbers are probably incorrect. You might want to double-check that it has your correct biometric information and is capturing your HR accurately.

If you really hit 1,160 calories, then you were burning calories at the extraordinary rate of 39 calories per minute. It takes a moderate aerobic level of effort to hit 10 calories/minute. When I ride an indoor bike trainer at a pretty intense level of effort (250 watts, average), I am only burning about 17 calories per minute. My highest sustained effort, when I was riding a 20 minute FTP test interval on the bike, was 20 calories/minute. I cannot fathom what it would feel like to burn 39 calories/minute for 30 minutes.

All I did to get the 1160 reading was figure out the math for the stair stepper which was accurate based on the heart rate and multiplied it for 120 minutes what my workout was. Not saying 1160 could be more accurate but definitely closer than 522 for a 2 hour brutal weights workout.

 

[Blujelly]

Thing is surely MFP is just calculating what it could be, its not taken in account of your heart rate or anything like that? I'm pretty certain MFP states it takes averages for activities and can't give accurate measurements. I use MFP purely just for food basis, I use my AW to track my health I always find MPF was running to high.

 

[x-evil-x]

Exxxvii
All I did to get the 1160 reading was figure out the math for the stair stepper which was accurate based on the heart rate and multiplied it for 120 minutes what my workout was. Not saying 1160 could be more accurate but definitely closer than 522 for a 2 hour brutal weights workout.

I'm assuming you thought I was talking about my stair stepper 30 minute session being 1160 I meant my weights workout that was 2 hours. 290x4 = 1160.
My stair stepper workout was more accurate being my average heart rate was in the 120's.
Usually when I look down at my watch during weight training it's in the 50's- 90's. Sometimes it reads right for a little while at 130-150. Usually most my workouts over the last 6 months average heart rate for a weights workout is in the 80's. Not accurate at all.

 

[exxxviii]

All I did to get the 1160 reading was figure out the math for the stair stepper which was accurate based on the heart rate and multiplied it for 120 minutes what my workout was. Not saying 1160 could be more accurate but definitely closer than 522 for a 2 hour brutal weights workout.

Aaaah. Yes, 1,160 / 120 minutes is 10 calories per hour. That makes sense, assuming you were at a cadence that put you at a high-moderate (zone 3) heart rate level. But, I could not follow your math about how calculation relates between the calorie estimate on the Apple Watch during the stair stepper workout and the calorie estimate from the stepper machine itself.