Calories strength training
yirara
Posts: 9,928 Member
For my rehab I'm doing strength training. Of course I need to nourish my body sufficiently, but I also don't want to gain weight. I know strength training doesn't burn a ton of calories, but if you do something everyday it can easily add up. Anyone fancy a guess? Warning: High volume, low weight
Pull day:
Bendover barbell row 13.5kg 15x5
Lateral raise 1.25kgx2 10x5
Biceps barbell curl 11kg 10x5
Front raise 1.25/2.5kg 10x5
Kettlebell row 8kg 15x4
External banded rotation 7kg 10x3
Internal banded rotation 15kg 10x3
Banded row from top of door 30kg 10x3
Banded row with twist and arm up behind 7kg 10x3
My guess is 150kcal for a nearly 50yr female, not super tall, normal weight. Anyone want to go higher or lower? Garmin, btw gave me 277 active calories, for a volume of just over 4100kg
(I tried raises with higher weights and did an additional set each)
Pull day:
Bendover barbell row 13.5kg 15x5
Lateral raise 1.25kgx2 10x5
Biceps barbell curl 11kg 10x5
Front raise 1.25/2.5kg 10x5
Kettlebell row 8kg 15x4
External banded rotation 7kg 10x3
Internal banded rotation 15kg 10x3
Banded row from top of door 30kg 10x3
Banded row with twist and arm up behind 7kg 10x3
My guess is 150kcal for a nearly 50yr female, not super tall, normal weight. Anyone want to go higher or lower? Garmin, btw gave me 277 active calories, for a volume of just over 4100kg
(I tried raises with higher weights and did an additional set each) 0
Replies
-
Believe it or not, the MFP exercise database estimate is probably one of the better (relatively conservative) estimates of strength training calories.
For regular reps/sets work like that, use the entry "Strength training (weight lifting, weight training)" in the Cardiovascular Exercise part of the MFP database.
Use the total workout wall clock time, including normal short between-set rests. (Obviously, if you take a longer break to chat or refill your water bottle or something, subtract those minutes if they amount to much. But the research on which that estimate is based included the expectation of short inter-set rests.)1 -
Hmm.. actually, I've never tried this one. I usually had a break of 1-1.5min between sets, near the end basically none or just a few secs as I rotated exercises. 3 mins between the non-banded exercises. So... hmm.. about 155kcal So that's not too bad a guess then 1
-
maybe worth a read:
A heuristic for estimating energy expenditure during resistance training
https://www.strongerbyscience.com/research-spotlight-expenditure-resistance/0 -
Hmm.. actually, I've never tried this one. I usually had a break of 1-1.5min between sets, near the end basically none or just a few secs as I rotated exercises. 3 mins between the non-banded exercises. So... hmm.. about 155kcal So that's not too bad a guess then
Some fitness trackers now (IMU) now use a METS-based algorithm to estimate calories if they know the user is doing strength training, and that's what MFP is doing under the covers. Heart rate is a poor proxy for calorie burn during strength training. MFP (as I think you, yirara, know) has IMU a conceptual flaw in their METS implementation, but it's not arithmetically very important for the typical strength workout of half an hour to an hour-ish.0 -
Hmm.. actually, I've never tried this one. I usually had a break of 1-1.5min between sets, near the end basically none or just a few secs as I rotated exercises. 3 mins between the non-banded exercises. So... hmm.. about 155kcal So that's not too bad a guess then
Some fitness trackers now (IMU) now use a METS-based algorithm to estimate calories if they know the user is doing strength training, and that's what MFP is doing under the covers. Heart rate is a poor proxy for calorie burn during strength training. MFP (as I think you, yirara, know) has IMU a conceptual flaw in their METS implementation, but it's not arithmetically very important for the typical strength workout of half an hour to an hour-ish.
Yeah, I know. I'm a bit miffed at how Garmin handles this. I start a set, then finish it and go into resting mode. In the detailed analysis I see that my workout took a total of.. say 60 minutes and 30 of those were actual workout sets and the other 30 minutes rest. But just a bit below I see that 45 minutes actually were moderate intensity minutes and 2 vigorous ones. Whut? Makes no sense! Why count intensity minutes when you're just sitting about, resting? Yeah, HR is still high enough for this to count as intensity minutes, but this would be something to easily adjust for Garmin. Just count intensity minutes in strength training mode when lifting.1 -
maybe worth a read:
A heuristic for estimating energy expenditure during resistance training
https://www.strongerbyscience.com/research-spotlight-expenditure-resistance/
I looked at this. My workout would probably give me close to 400 net calories. It's possible that this equation doesn't work for women, not for people with a higher bodyfat, or any of many other reasons, including it only working for people that do spend an awful lot of time building their body. No idea. But 400 certainly is far above what I would expect.1 -
maybe worth a read:
A heuristic for estimating energy expenditure during resistance training
https://www.strongerbyscience.com/research-spotlight-expenditure-resistance/
Good stuff from you as usual, Banx. Bookmarked it.
I downloaded the spreadsheet linked in the article. It gives me a range of 91-211 calories for half an hour (the additive calories). MFP's MET-based estimate (complete with its theoretical flaw, which for a lot of cases will overestimate calories) would give me 82 calories for the same half hour at the same bodyweight. I can see the spreadsheet as potentially more accurate (METS is weird for strength training, seems like), but I don't (as you probably know) do enough strength training to worry about it myself. Others are going to get different results, maybe a bigger difference, at different body weights and BF%s.0 -
tldr Use the MFP estimate.
I use a MET estimate. I've seen MET estimates for weights as anywhere from 3 (moderate intensity) to 6 (high intensity).
I worked out my MET for being awake as being 4.426% of maintenance calories, which is a little more than 1/24, because sleeping uses less. That's about 1.06 MET. Then I assume for myself a MET for my weights workout of about 3.5 which is probably conservative, but also I'm sure 5 or 6 would be too high. Subtract the waking MET to avoid double counting it. So that's 2.5*MET as additional calories entered.
The MFP entry for Cardio Strength Training looks quite accurate, about 3.75 MET afaict. Of course if you're working really hard maybe you're at 5 MET.
I should actually return to this math one day to better take into account that MFP uses BMR * 1.25 as the baseline calc for Sedentary, which is probably not quite the same as 1 MET.1 -
Btw, I see you're doing 5 sets of curls, lateral raises and front raises.
YMMV, I don't personally find the utility in front raises, as they seem to get plenty of work doing bench press and shoulder press. But anyway, I encourage you to try Myo-reps for these three exercises.
e.g. with curls instead of 10x5, start with a weight you can do 10-15 reps with. Do a set with 1-2 RIR. Put the weight down for about 15 seconds. Repeat. And keep repeating until you can't do 5 reps with your last set, which is to failure. You'll get a lot of effective reps this way, you'll get a hell of a burn, and it'll be faster than 5 regular sets with normal breaks.0 -
Deleted. Forum is doing odd quoting stuff here.0
-
maybe worth a read:
A heuristic for estimating energy expenditure during resistance training
https://www.strongerbyscience.com/research-spotlight-expenditure-resistance/
Good stuff from you as usual, Banx. Bookmarked it.
I downloaded the spreadsheet linked in the article. It gives me a range of 91-211 calories for half an hour (the additive calories). MFP's MET-based estimate (complete with its theoretical flaw, which for a lot of cases will overestimate calories) would give me 82 calories for the same half hour at the same bodyweight. I can see the spreadsheet as potentially more accurate (METS is weird for strength training, seems like), but I don't (as you probably know) do enough strength training to worry about it myself. Others are going to get different results, maybe a bigger difference, at different body weights and BF%s.
Ah ok. I played with the spreadsheet a bit more. Seems like the higher the bodyfat % the higher the calorie expenditure for the exercise. I'm not sure whether that makes sense. I also have no idea what my bodyfat % is. I don't think this website specifically mentions what the definition of training duration is. is it active work plus short breaks between sets? Or including longer breaks between exercises?
So here I'm getting upper end - lower end for two potential body fat estimates
high %: 181 - 421
low %: 176 - 416
If I use the intensity minutes garmin gives me (without double counting for vigorious ones) I get a bit less. This assumes it's a good estimate to eliminate the longer breaks between different exercises.Retroguy2000 wrote: »Btw, I see you're doing 5 sets of curls, lateral raises and front raises.
YMMV, I don't personally find the utility in front raises, as they seem to get plenty of work doing bench press and shoulder press. But anyway, I encourage you to try Myo-reps for these three exercises.
e.g. with curls instead of 10x5, start with a weight you can do 10-15 reps with. Do a set with 1-2 RIR. Put the weight down for about 15 seconds. Repeat. And keep repeating until you can't do 5 reps with your last set, which is to failure. You'll get a lot of effective reps this way, you'll get a hell of a burn, and it'll be faster than 5 regular sets with normal breaks.
Thanks for your inside, would normally appreciate them
However, this is for rehab after a complex humeral head fracture, and for learning to do certain movements again. The biceps curls are in isolation to give my atrophied biceps something to do and to potentially re-stabilize the biceps tendon that was likely subluxed for over 3 months until I put it back in it's place (yeah, Ehlers Danlos is fun: You can end up with soft tissue, or even bones in the wrong place with limited damage to other stuff). The two raises also have a specific aim: I'm in the kitchen, want to grab tea from the top shelf but arm is to weak to lift to the front, and back shoulder is unstable. I can focus on every muscle doing these frost raises, build some front lifting strength and hopefully automate the right response in the back to this again. Or I walk between two laundry lines, to get laundry left and right of me. The injured arm is too weak to move sideways up repeatedly and the back shoulder is also doing odd things. I'm currently only going to just over 90 degrees, but it makes a ton of a difference in everyday life.
Thus none of this is about going to failure or doing as many as possible, but the aim is to go slow, focus on every muscle involved, regain strength and mass again, but also retrain my brain to make use of the right muscles again for the most common situations. Exercise weight is slowly going up, but I need to stick with higher reps for my stupid brain to remember how to move. Sometimes I only start to get it right at rep 8 in set 3. I also can't do much higher reps than 15 because of an inborn muscle condition that still waits for answers (the accident interrupted diagnosis). I only know that every exercise I do is at or above the anaerobic threshold, including running at walking pace with 80-100km/month over many months. And that my muscles just give up once lactate reaches a certain, not even that high level. Thus there's only that much I can do with this rehab.
Normally I'd probably just have fun with bodweight exercises without any specific programming in mind, or work on the top 5 lifts if I had a gym with squat rack and enough and safe equipment. But that's just not my aim here. Today is push day, btw. And I also throw in a lower body day every now and then.1 -
maybe worth a read:
A heuristic for estimating energy expenditure during resistance training
https://www.strongerbyscience.com/research-spotlight-expenditure-resistance/
Good stuff from you as usual, Banx. Bookmarked it.
I downloaded the spreadsheet linked in the article. It gives me a range of 91-211 calories for half an hour (the additive calories). MFP's MET-based estimate (complete with its theoretical flaw, which for a lot of cases will overestimate calories) would give me 82 calories for the same half hour at the same bodyweight. I can see the spreadsheet as potentially more accurate (METS is weird for strength training, seems like), but I don't (as you probably know) do enough strength training to worry about it myself. Others are going to get different results, maybe a bigger difference, at different body weights and BF%s.
Ah ok. I played with the spreadsheet a bit more. Seems like the higher the bodyfat % the higher the calorie expenditure for the exercise. I'm not sure whether that makes sense. I also have no idea what my bodyfat % is. I don't think this website specifically mentions what the definition of training duration is. is it active work plus short breaks between sets? Or including longer breaks between exercises?
So here I'm getting upper end - lower end for two potential body fat estimates
high %: 181 - 421
low %: 176 - 416
If I use the intensity minutes garmin gives me (without double counting for vigorious ones) I get a bit less. This assumes it's a good estimate to eliminate the longer breaks between different exercises.
(snip)
I'm not sure exactly what info you're looking for, but I thought the article (Banx's link) was reasonably clear about what they were testing:João and colleagues recruited 15 trained men for a crossover study investigating the impact of training intensity on energy expenditure during resistance training. After assessing 1RMs, the subjects performed three different sessions of eight exercises (chest press, pec deck, squat, pull-down, biceps curl, triceps extension, hamstrings curl, and machine crunch). The low-intensity session consisted of 2 sets of 15 reps of each exercise at 60% of 1RM, the moderate-intensity session involved 3 sets of 10 reps at 75% of 1RM, and the high-intensity session required 6 sets of 5 rest mps at 90% of 1RM (Figure 1). Subjects rested two minutes between sets. During each session, subjects wore a portable spirometer (to measure gas exchange) as a means to estimate energy expenditure.
That doesn't specifically talk about rests between exercises (just between sets), though.
If that's not enough the article includes a link to the actual study article, and (at least from here in the US), I can get a full text version at this link:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8714826/
I usually recommend the MFP estimator for strength training because:
* It's relatively lower,
* I know heart rate is stupid for strength training,
* In general strength training is hard to estimate, and
* There's a certain amount of physiological weirdness around estimating it. (I'm thinking not just about EPOC, but . . . how much calorie cost is there to the higher level of muscle repair? Seems like the regular heavy lifters here often have high TDEEs (anecdotal), and the research suggests that's not just about muscle mass at rest.)
* On top of that, there have been a couple of people here over the years (Azdak was one) much more expert than I in exercise physiology, who thought that the MFP METS estimate was probably the best of a bad lot of options.
TBH, beyond that, I haven't put a lot of thought into this. For one, I don't do much strength training. For two, my perspective is to find practical exercise calorie estimating methods that are likely to be in the ballpark, with a bias toward avoiding overestimating vs. underestimating. I don't need perfection, just a workable, consistent estimate.
That's just me.
I'm not saying this about you, but I see people here get all tied up in knots about how hard it is to estimate exercise calories accurately. Sometimes they're getting anxious over differences between things like 200 vs. 300 calories for something they do 3 times a week. "Yikes, a whole hundred calories! Yikes, one is 50% higher!").
From the perspective of arithmetic magnitude, and in a context where every blippin' thing is an estimate (food, BMR/RMR, an activity level multiplier, for pity's sake!) . . . it just doesn't matter (IMO).
Couldn't resist (sigh) the usual essay/digression, so spoiler:If someone's eating 1500 calories x 7 days, being off by 3 x 100 calories a week on exercise . . . is that 2.8% of calories? There's a limited amount of angst and drama I'm willing to spend on 2.8% of weekly calories, personally. I probably make more logging errors than that in a week, even though I try quite hard to avoid them. Some will be over, some will be under, I'm watching the weight trend, so . . . meh.Even as a happy Garmin user, I wouldn't think Garmin intensity minutes would be great for estimating strength training in any capacity, even just to figure out how much of the session was work vs. rest. Don't you have heart rate idiosyncrasies anyway (seems like you may've mentioned that in other posts, but I may be mistaken)?
When I start doing some new exercise, I look into methods for estimating the exercise calories, and compare multiple methods (if they exist). I try to think about how well the method fits the exercise. (Examples: METS is iffy for things where the work (in the physics sense) isn't very influenced by bodyweight. Heart rate is iffy for strength training because of the influence of physical strain on heart rate independent of oxygen consumption.) Given two methods that seem equally probable to be semi-sensible, I'd pick the one that yields lower estimates. That's good enough for me. All I need is workable estimates.
After that, I use the method consistently, and don't worry about it, unless I come across a new method that looks better (could be more accurate, easier but similarly accurate, whatever). Then I may switch, stick to the new one. (Example: I switched to the ExRX calculator for walking because it's more granular than some, and automatically calculates net, which matters more for walking as I do it than gross vs. net matters for some other things.)
I'm honestly not sure how Garmin estimates strength training, either: IMU some trackers are starting to use METS not HR for strength training. I get low estimates from Garmin (usually close the MFP estimate), so was guessing it might use METS. However, I also tend to do higher rep/lower weight (muscle imbalances/injury risk) which won't bring strain into the picture as much. So who knows - it's probably published somewhere, but I haven't even looked.
0 -
@yirara I played with the spreadsheet too. It gave nearly identical results to my own MET estimate I outlined before. For me it says 267 cal for an hour (additive), and I typically enter 250-300 into MFP. MFP's estimate says 299. I wonder if that would change if I changed my MFP Goal settings from Sedentary.
They're saying in that article that the total calories is about 360 per hour, i.e. 6 per minute. What changes when you adjust the body fat % is the BMR, which when I plug these numbers into sailrabbit, looks like they're using the Cunningham model. Adjusting my body fat % lower in the spreadsheet without changing my weight adds a little to the BMR. This makes sense. In this scenario I'd have more muscle mass, which burns more calories at rest than fat, so at the same weight with less body fat my BMR would be higher. Since the BMR is higher, and the total expended is assumed to be the same, it follows that the additive amount (360-BMR) is lower. But only by a few calories, which is basically noise.
The subjects were all trained men. I'm a bit surprised that total energy expended is the same regardless of body mass. If I halve the body mass here, which would be in line with a petite woman, they still have the same estimate of 360 per hour, and she'd have burned more additive calories because her BMR is much lower. I know that intensity is specific to the person, so e.g. me doing 10 reps bench press with some weight and that petite woman half my mass doing 10 reps bench press with much lower weight can feel similarly intense to both of us, but does a 100kg man and 50kg woman really both use the same amount of energy to move different volumes of weight through space? What would Sir Isaac Newton say?
@AnnPT77 What is the MET flaw in MFP you refer to?0 -
Thanks a lot guys. I wasn't quite awake yet this morning. But some data geekery is always good, no?
No workout today. I did stretching and mobility stuff all day and just didn't have any strength left. But hey, tomorrow is another day.1 -
Retroguy2000 wrote: »@yirara I played with the spreadsheet too. It gave nearly identical results to my own MET estimate I outlined before. For me it says 267 cal for an hour (additive), and I typically enter 250-300 into MFP. MFP's estimate says 299. I wonder if that would change if I changed my MFP Goal settings from Sedentary.
They're saying in that article that the total calories is about 360 per hour, i.e. 6 per minute. What changes when you adjust the body fat % is the BMR, which when I plug these numbers into sailrabbit, looks like they're using the Cunningham model. Adjusting my body fat % lower in the spreadsheet without changing my weight adds a little to the BMR. This makes sense. In this scenario I'd have more muscle mass, which burns more calories at rest than fat, so at the same weight with less body fat my BMR would be higher. Since the BMR is higher, and the total expended is assumed to be the same, it follows that the additive amount (360-BMR) is lower. But only by a few calories, which is basically noise.
The subjects were all trained men. I'm a bit surprised that total energy expended is the same regardless of body mass. If I halve the body mass here, which would be in line with a petite woman, they still have the same estimate of 360 per hour, and she'd have burned more additive calories because her BMR is much lower. I know that intensity is specific to the person, so e.g. me doing 10 reps bench press with some weight and that petite woman half my mass doing 10 reps bench press with much lower weight can feel similarly intense to both of us, but does a 100kg man and 50kg woman really both use the same amount of energy to move different volumes of weight through space? What would Sir Isaac Newton say?
@AnnPT77 What is the MET flaw in MFP you refer to?
IMU it's a gross figure rather than net (not adjusted for BMR/RMR or activity level that would normally be the calorie burn during the time period). I don't mostly consider that a big deal arithmetically for most activities, i.e., somewhat vigorous things done for half an hour or an hour. It can be a bigger deal for something like long, slow walking. I may be biased by the fact that my estimated (not measured) BMR/RMR is pretty low, like 50-ish calories per hour.
There are IMO some issues with particular other entries in the MFP database (like using speed ranges to specify the METS for outdoor biking, when things like bike style, terrain, etc. can make a difference; or the combination of rowing/canoeing in a single entry with the implication that calorie expenditure is the same at any given speed for either of those, irrespective of boat type among other variables; but those aren't the only head-scratchers, just a couple of examples). It's research-based, not always the worst choice of methods.
People on average IMO tend to give fitness trackers/heart rate monitors too much credit for accuracy, and over-criticize the METS approach.
But all we need is workable estimates, and I'd guess that consistency of estimating is a bigger deal in the real-life context than exact technical accuracy, at least once there's reason to think the estimate's somewhere near the ballpark.
0 -
@AnnPT77 I did a bit of testing.
As mentioned, the site with the spreadsheet appears to use the Cunningham model to calculate BMR. It's quite a bit higher than the Miffin St Jeor estimate. The sailrabbit page says this:
"I don't have much on the Cunningham methodology, but it appears to be popular in some bodybuilding and fitness circles. I did run across a study (see below reference #2) however that tested 7 different equations on 51 male athletes and while 6 significantly underestimated Resting Metabolic Rate (RMR), the Cunningham method was the only one to overestimate (at +59 kcal per day). That's pretty darn close, so this may be a good choice if you are heavily active or looking to add some weight. It works pretty much like the Katch-McArdle formula so resting metabolism for fat tissue doesn't appear to be represented here either."
I changed my settings here in Goals to test what the MFP estimate for strength training is whether I choose Sedentary (BMR * 1.25) or the most active (BMR * 1.8). That pushed my TDEE up by about 1,000, however neither setting changed the calorie estimate for strength training.
Adding the MFP estimate for me for 60 minutes (299) plus a single hour of MFP's BMR estimate (Miffin formula) gets to 379. The spreadsheet above estimate for total calories is 360. That's very close. So it may be that MFP's exercise estimate is additive and not including BMR * 1.0.
Therefore, both spreadsheet and MFP are slightly inaccurate for additive calories. The amount of additive calories would be the total used in the workout minus a typical waking hour, not an hour of laying still and not moving a muscle.
The MFP estimate would be over-estimating by the multiplier you've chosen minus one, multiplied by BMR. So in the example of Sedentary, it would be over-estimating by 0.25 * BMR.
Earlier, I said I used my own MET estimate of 250-300 depending on my workout, and MFP said 299. Well, if I subtract 0.25 * BMR from 299, I get 279. That's basically identical.
Btw, MFP has a BMR calculator in the Apps tab at the top of the page on desktop.
@yirara Try the Navy method to estimate your body fat.
0 -
Retroguy2000 wrote: »@AnnPT77 I did a bit of testing.
As mentioned, the site with the spreadsheet appears to use the Cunningham model to calculate BMR. It's quite a bit higher than the Miffin St Jeor estimate. The sailrabbit page says this:
"I don't have much on the Cunningham methodology, but it appears to be popular in some bodybuilding and fitness circles. I did run across a study (see below reference #2) however that tested 7 different equations on 51 male athletes and while 6 significantly underestimated Resting Metabolic Rate (RMR), the Cunningham method was the only one to overestimate (at +59 kcal per day). That's pretty darn close, so this may be a good choice if you are heavily active or looking to add some weight. It works pretty much like the Katch-McArdle formula so resting metabolism for fat tissue doesn't appear to be represented here either."
I changed my settings here in Goals to test what the MFP estimate for strength training is whether I choose Sedentary (BMR * 1.25) or the most active (BMR * 1.8). That pushed my TDEE up by about 1,000, however neither setting changed the calorie estimate for strength training.
Adding the MFP estimate for me for 60 minutes (299) plus a single hour of MFP's BMR estimate (Miffin formula) gets to 379. The spreadsheet above estimate for total calories is 360. That's very close. So it may be that MFP's exercise estimate is additive and not including BMR * 1.0.
Therefore, both spreadsheet and MFP are slightly inaccurate for additive calories. The amount of additive calories would be the total used in the workout minus a typical waking hour, not an hour of laying still and not moving a muscle.
The MFP estimate would be over-estimating by the multiplier you've chosen minus one, multiplied by BMR. So in the example of Sedentary, it would be over-estimating by 0.25 * BMR.
Earlier, I said I used my own MET estimate of 250-300 depending on my workout, and MFP said 299. Well, if I subtract 0.25 * BMR from 299, I get 279. That's basically identical.
Btw, MFP has a BMR calculator in the Apps tab at the top of the page on desktop.
@yirara Try the Navy method to estimate your body fat.
What METS value did you assume MFP was using for strength exercise?
I appreciate your sense of curiosity, but like I said, I'm not super invested in the gross vs. net issue: I don't use the MFP estimates much (other than for strength training), and as I also mentioned, I don't think gross vs. net is a huge deal most of the time. I don't think it is for typical strength training for most people. For sure, I don't worry about it for me. My estimated BMR/RMR is 46-51 calories/hour by Harris-Benedict or Mifflin-St. Jeor, never been tested, but my guess is that that's a little low (Cunningham, considering estimated BF%, would give me 62).
I think @PAV8888 has looked into gross vs. net issues with the MFP values more than I have, so he might have some insights. (@Yirara, I hope you're OK with this getting off into details on your thread.)
At some point, I thought MFP had said they were getting rid of their BMR calculator, so I didn't realize it was still here. I always use Sailrabbit for reference, might as well see a range of estimating methods.
0 -
Sorry guys. Not much into it at this moment.
But back when when I looked at it and randomly sampled MFP appeared to be just straight MET values from the tables.
This means MFP INCLUDES/DOUBLE counts at least 1.25 METS for the duration of the exercise. And if you're setup at anything above sedentary, then you would be subtracting an even higher value to find the net additional CALS using MFP, i.e. 1.4 or 1.6 or 1.8)
If you introduce a tracker and you've got the tracker connected to MFP, I would suggest you make all the adjustments on the tracker side as MFP and trackers have a hard enough time keeping things straight.
The flow used to be that MFP manual activity over-writes tracker detection on the tracker and then feeds-back to MFP, thus negating the tracker doing the work.
Are you trying to gain weight or lose? An easy sanity check would be "moderate" (3 MET = walking level) if you're trying to lose. Higher 5/6 MET if you're trying to gain.
1 MET minute = 1 RMR minute (i.e. MFP miflin BMR) approx. There exists a "corrected" MET which if you think MET values are generally speaking already high (and @yirara in particular generally finds her actual burns to be much less than expected), well this correction would definitely head in the wrong direction.... https://www.sites.google.com/site/compendiumofphysicalactivities/corrected-mets
In another thread I had a beautiful gif of a puppy digging at the beach... but I didn't post it!
1 -
The 150 calories is probably pretty close. The true way to see if your equations are correct for your goal is to run it for a month to 6 weeks and see what your weight is doing. You may see indications at the 3 week point that you need to adjust that is fine also. Adjust calorie intake accordingly.For my rehab I'm doing strength training. Of course I need to nourish my body sufficiently, but I also don't want to gain weight. I know strength training doesn't burn a ton of calories, but if you do something everyday it can easily add up. Anyone fancy a guess? Warning: High volume, low weight
Pull day:
Bendover barbell row 13.5kg 15x5
Lateral raise 1.25kgx2 10x5
Biceps barbell curl 11kg 10x5
Front raise 1.25/2.5kg 10x5
Kettlebell row 8kg 15x4
External banded rotation 7kg 10x3
Internal banded rotation 15kg 10x3
Banded row from top of door 30kg 10x3
Banded row with twist and arm up behind 7kg 10x3
My guess is 150kcal for a nearly 50yr female, not super tall, normal weight. Anyone want to go higher or lower? Garmin, btw gave me 277 active calories, for a volume of just over 4100kg (I tried raises with higher weights and did an additional set each)
You’ll never be exact on your calorie burning estimates and your calorie intake estimates so the only true way to see how you’re doing with things is by actual results after a short period of time.
0 -
I didn't. I just found it interesting that the MFP estimate, and my own estimate, and that spreadsheet linked earlier, were all so close. For my own estimate, I assumed 3.5 MET, minus the 1 for sitting idle instead.What METS value did you assume MFP was using for strength exercise?
For me, the issue of double counting is significant. At 220 pounds, 100kg, I get the following numbers:
1 MET cals per hour = 105 (that's 3.5 * weight in kg / 200 * 60)
Cunningham BMR per hour = 93
MFP BMR per hour = 80
I see MET is often used synonymously with RMR, but it is not the same as BMR. Some sites use all these terms interchangeably, which is wrong.
At an average 3.5 workouts per week, if I didn't account for the double counting of 1 MET, that works out to nearly 350 calories difference per week, potentially 0.1 pounds per week difference, nearly 0.5 pounds per month.
MFP one hour strength training = 299
299/105 = 2.85 MET, which is too low for weights.
However, if we instead treat BMR as RMR, as some sites incorrectly do, then it would be:
299/80 = 3.7, which is a more realistic multiplier. If MFP is simply using MET tables as is, then it's also basing them on the wrong multiplier, using BMR instead of MET.0 -
Do discuss on, please! I love those data-nerdery threads.1
-
Fair warning: My answer is going to be more about issues of estimating, less about comparative applied math.Retroguy2000 wrote: »
I didn't. I just found it interesting that the MFP estimate, and my own estimate, and that spreadsheet linked earlier, were all so close. For my own estimate, I assumed 3.5 MET, minus the 1 for sitting idle instead.What METS value did you assume MFP was using for strength exercise?
Without great thought about it, I was thinking the MFP database entry represented something around 3 METS.For me, the issue of double counting is significant. At 220 pounds, 100kg, I get the following numbers:
1 MET cals per hour = 105 (that's 3.5 * weight in kg / 200 * 60)
Cunningham BMR per hour = 93
MFP BMR per hour = 80
I see MET is often used synonymously with RMR, but it is not the same as BMR. Some sites use all these terms interchangeably, which is wrong.
At an average 3.5 workouts per week, if I didn't account for the double counting of 1 MET, that works out to nearly 350 calories difference per week, potentially 0.1 pounds per week difference, nearly 0.5 pounds per month.
Well, clearly you and I have different wishes/hopes about accuracy. That's not a criticism, it's just about preferences and context.
Even as a much smaller, older woman (5'5", 133 pounds this morning, age 67), I'd be fine with a 0.1 pound per week error. My TDEE is around 14,700 calories/week, so that's around 2.4% of the big picture. Meh. If this is repeat exercise, that'll show up in the trend, and I can adjust (base calories is what I'd adjust). If it's not repeated exercise, somewhere within 6 or 8 blocks of the ballpark is good enough - drop in the ocean.
All of this stuff is estimates - food logging, base calorie needs (BMR and activity), exercise. They need to be workable. They're not going to be perfect. Some will be under, some over.
Research suggests that activity of all types is more likely to be overestimated, food intake underestimated. When I don't know a solid estimate, I try to err - if at all - in the opposite direction. The MFP strength training estimate seems like it's more likely a lowball, to me, and that's OK with me. Strength training is really hard to estimate, IMO. I'm willing to invest some time in estimating it reasonably, but there are limits on that investment.
My individual context is that if I set my MFP profile values accurately, its base calorie estimate for me is off by around 500 calories daily (more or less), which is 25-30%. (I've had stretches of a month where I couldn't much exercise, but that ratio's held in times of quite variable exercise, too.) If I set MFP at "active", it's still a little low, and I'm actually around the border (steps-wise) between sedentary and lightly active (pre-exercise), usually on the sedentary side.
TDEE calculators do a little better with estimating for me, especially formulas that consider BF%, but they're generally still meaningfully low, couple of hundred calories. (I think it's unlikely that I'm eating that much less food than I log for the last 7+ years.) But TDEE doesn't work for me as a guide (vs. the MFP method) because my exercise load isn't consistent (weather dependent, seasonal). So I set base calories manually, estimate exercise in some way that I think is reasonable, call it good.
Given all of that, which probably creates cognitive bias, I'm not worried about 2.4% . . . though if I had confidence in an algorithm that adjusted my estimates by that much, I'd adopt it.
There's only so much energy I'm willing to invest, though - a reasonable estimate that's more likely to be low than high is fine, for me, for exercise. Exercise is variable, but usually only 25% or so of my TDEE, more or less. Close is good enough.MFP one hour strength training = 299
299/105 = 2.85 MET, which is too low for weights.
Curiosity: How do you know it's too low? Is it too low for you, doing the weight routine you do, or are you asserting that it's too low for everyone in general? (I understand what's in the Compendium of Physical Activities. If that's the basis for saying 2.85 is too low, OK by me.)
I feel like you may have more confidence in the METS values than I do. The methodology is research-based, which is great, but it's got to be that the underlying research results yield a range (probably but not necessarily a normal distribution) that's boiled down into some kind of average to get a single number. It's truly and literally a SWAG (scientific wild-a** guess. emphasis on "scientific"). It's workable. More so in some cases than others, but still an estimate with a lot of assumptions.However, if we instead treat BMR as RMR, as some sites incorrectly do, then it would be:
299/80 = 3.7, which is a more realistic multiplier. If MFP is simply using MET tables as is, then it's also basing them on the wrong multiplier, using BMR instead of MET.
I assume you're using the weight you have recorded in MFP (not your current weight if different) to evaluate the results it gives you with the exercise? (I'm not doubting you, but when I started playing with this myself this time, I forgot that MFP didn't have my most current weight, since I don't much update it in maintenance, got some odd results. IOW, I'm sharing a mistake I made, not assuming you made one.)
Sure, BMR and RMR are different, but they're used somewhat interchangeably for decent reasons: They tend to be quite close, and RMR is somewhat easier (logistically) to measure. Overwhelming, most of us are relying on estimates of BMR/RMR anyway. IMU, one standard deviation for RMR is around 5-8%**. (For most of us, that variability is a bigger number than 350 calories per week, especially once the activity multiplier enters the picture. Over 30% of us are presumed to be more than one standard deviation away from the RMR estimate besides (basic statistics), i.e., more than 5-8% away from the estimate.)
** https://examine.com/articles/does-metabolism-vary-between-two-people/ which is a readable explanation based on https://pubmed.ncbi.nlm.nih.gov/15534426/
I get the appeal of digging into the numbers, kind of, even though there's only so far I personally want to take it. If you want to go further, I'll be over here cheering you on, encouraging you to share your analysis and results . . . but also encouraging you to consider the error ranges around all of the other moving parts, and think about what constitutes meaningful variation in context. (Individuals' definitions of "meaningful" will vary, which is fine - part of what makes discussion interesting.)0 -
We're probably in agreement about the relative merits of estimation accuracy. I know there are so many estimates upon estimates, a big one obviously being a blanket multiplier of BMR to estimate for general daily activity. Food labels can be inaccurate. Portions may be eye-balled not weighed. Exercise estimates can be off. And so on.
My thinking on the strength training estimate is this is something I do a lot. It's generally an every other day addition here. It's not a new meal from a restaurant I may never have again, in which case expending a lot of time trying to get an accurate number for a single entry would seem pointless. However, if I'm going to be entering the same number in manually with such high frequency, why not try to make it accurate? Just because many other parts in the system have potential errors is not sufficient reason to avoid being more accurate where one can, when the effort required is so minimal. On an ongoing basis, that effort is zero.
It's not that I have a high confidence in MET estimates. It's a useful starting point. I don't have a fitness watch, so any number I enter for an activity is likely going to be based on MET, whether that's MFP's number, or a number from some other internet calculator, or the one I applied to myself (assuming 3.5 MET), and I enter that every other day.
Yes, my weight is accurate in Goals here, and I suggested 2.85 MET was too low based on Google results about the MET range for weights (3-6 seems to be the consensus), and that spreadsheet used in this thread is much higher than 2.85 MET, and my own estimation of my effort.
One thing I don't understand, is why in that earlier spreadsheet does it estimate the same amount of calories if e.g. you halve or double the body mass? Surely someone with half or double the body mass is lifting lighter or heavier weights respectively to match relative intensity with each other, but it apparently takes the same amount of energy in both cases?0 -
MFP (and fitbit and ??), based on observation, use Mifflin. Mifflin is an RMR formula. Other than the original Harris... which estimate/formula provides BMR values?0
-
The spreadsheet linked via the article earlier in this thread computes additive calories as being total during the exercise session minus "basal metabolism for the duration".MFP (and fitbit and ??), based on observation, use Mifflin. Mifflin is an RMR formula. Other than the original Harris... which estimate/formula provides BMR values?
And in MFP, the strength training estimate given for me for an hour may be 2.85 * my MET, or 3.7 * my BMR (as taken from MFP's BMR calculator at the top of this page). As mentioned earlier, their estimate does not change whether I've chosen sedentary or very active in my Goals here.0 -
Retroguy2000 wrote: »
The spreadsheet linked via the article earlier in this thread computes additive calories as being total during the exercise session minus "basal metabolism for the duration".MFP (and fitbit and ??), based on observation, use Mifflin. Mifflin is an RMR formula. Other than the original Harris... which estimate/formula provides BMR values?
And in MFP, the strength training estimate given for me for an hour may be 2.85 * my MET, or 3.7 * my BMR (as taken from MFP's BMR calculator at the top of this page). As mentioned earlier, their estimate does not change whether I've chosen sedentary or very active in my Goals here.
FWIW, the underlying study seems to have measured RMR (by gas analysis, 30' supine). Cunningham **, which he references in the article, does seem to estimate BMR. From that, I'd assume the spreadsheet is using the Cunningham estimate of BMR.
** https://pubmed.ncbi.nlm.nih.gov/7435418/0 -
Yes, I said that was what was used in the spreadsheet, earlier in the thread.0
-
Retroguy2000 wrote: »Yes, I said that was what was used in the spreadsheet, earlier in the thread.
I meant to be reacting primarily to PAV's ". . . what other than the original Harris . . .", in context of your comment about the spreadsheet referring to "basal metabolism", but managed the quoting unclearly for that - apologies.0
Categories
- All Categories
- 1.4M Health, Wellness and Goals
- 393.3K Introduce Yourself
- 43.8K Getting Started
- 260.2K Health and Weight Loss
- 175.9K Food and Nutrition
- 47.4K Recipes
- 232.5K Fitness and Exercise
- 424 Sleep, Mindfulness and Overall Wellness
- 6.5K Goal: Maintaining Weight
- 8.5K Goal: Gaining Weight and Body Building
- 153K Motivation and Support
- 8K Challenges
- 1.3K Debate Club
- 96.3K Chit-Chat
- 2.5K Fun and Games
- 3.7K MyFitnessPal Information
- 24 News and Announcements
- 1.1K Feature Suggestions and Ideas
- 2.6K MyFitnessPal Tech Support Questions