Welcome to Debate Club! Please be aware that this is a space for respectful debate, and that your ideas will be challenged here. Please remember to critique the argument, not the author.
How well do BMR estimation equations work? What should you do?
Options
EvgeniZyntx
Posts: 24,208 Member
This thread is intended to discuss issues with BMR equations and raise awareness of how they were developed and their intended use.
So you went on line and evaluated your BMR or TDEE. How good is it?
Let us look at the equations and how good or bad they are and then lets look at what you can do to create a personal working estimate that is more operationally accurate.
The RMR formulas are estimators. In General, "Four prediction equations were identified as the most commonly used in clinical practice (Harris-Benedict, Mifflin-St Jeor, Owen, and World Health Organization/Food and Agriculture Organization/United Nations University [WHO/FAO/UNU]). Of these equations, the Mifflin-St Jeor equation was the most reliable, predicting RMR within 10% of measured in more non obese and obese individuals than any other equation, and it also had the narrowest error range."
These equations are:
The general convention is that these predictive BMR equations perform best for groups of people instead of individuals.(1) There are conflicting results in the literature as to which equation is best suited for a general patient population. For example, one analysis indicates that despite a clinically relevant error rate of 20%, the Mifflin–St Jeor equation has the most accuracy and lowest magnitude of error and should be used among healthy non-obese and obese adults, (1) while a more recent analysis indicates that the Harris–Benedict equation and two adaptations of the WHO/FAO/UNU equation outperform the Mifflin–St Jeor equation in this population (Brazil). (2) The WHO/FAO/UNU and Schofield equations (3) adapted for children appear to be the best estimates of resting EE in children and adolescents.(4) However, discussion cautions against using a single equation when estimating resting EE of adolescents across all racial groups, which warrants new equations be derived (5).
In theory under- or overestimations of 10% in the prediction of BMR seems insignificant for a population, but it can be much more important in an individual approach, particularly for active athletes, where 10% of BMR can stand for a 250 kcal difference or more (6).
In that study, the example was given of a subject whose BMR was underestimated in average by 150 Kcal (10% of BMR) and considering all the tested equations, his daily energy requirement would be underestimated in about 270–345 Kcal/day. They authors note, "This variation can be a major interference in this individual’s dietary prescription and would possibly result in a monthly loss of body weight of about 1,4kg, which could cause deficit in energy-protein recovery, increase the risk of fatigue and muscle damage among other factors that could interfere with the performance of a high level athlete. On the other hand, an overestimation would possibly result in body weight and body fat gain."
In underweight females these equations tend to overestimate RMR [7]. They note, "The commonly used predictive equations were not appropriate for underweight subjects... among 10 RMR predictive equations that were used in this study, Muller et al. equation gave a fairly acceptable RMR prediction, while most of the commonly used RMR predictive equations did not accurately predict RMR at both group and individual levels. Our data also showed that all of the equations except Muller and Abbreviation equations significantly overestimated RMR in underweight young females, with mean differences ranging from 42.2 to 224kcal/day."
These findings are confirmed by Müller (who developed another set of equations) "WHO prediction equations systematically overestimated REE at low REE values but underestimated REE at high REE values” [8]. These researches also developed a model using fat free mass and fat mass and found, surprisingly, that this type of measurement was not significantly more accurate. (see image below)
In young Hispanic women, it was found that the Owen equation was the only one that was predictive with some accuracy [10]....
I’ll stop here for now - there is actually a huge amount of research on the subject.
Here are my references:
1. Frankenfield D, Roth-Yousey L, Compher C. Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: a systematic review. J Am Diet Assoc. 2005;105:775–789
2. de Oliveira EP, Orsatti FL, Teixeira O, Maesta N, Burini RC. Comparison of predictive equations for resting energy expenditure in overweight and obese adults. J Obes. 2011;2011:534–714
3. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr. 1985;39(Suppl 1):5–41
4. Finan K, Larson DE, Goran MI. Cross-validation of prediction equations for resting energy expenditure in young, healthy children. J Am Diet Assoc. 1997;97:140–145.
5. Torun B, Davies PS, Livingstone MB, Paolisso M, Sackett R, Spurr GB. Energy requirements and dietary energy recommendations for children and adolescents 1 to 18 years old. Eur J Clin Nutr. 1996;50(Suppl 1):S37–S80. discussion S-1
6. Loureiro LL, Fonseca S Jr, Castro NG, Dos Passos RB, Porto CP1, Pierucci AP. Basal Metabolic Rate of Adolescent Modern Pentathlon Athletes: Agreement between Indirect Calorimetry and Predictive Equations and the Correlation with Body Parameters. PLoS One. 2015 Nov 16
7. Soghra ALIASGHARZADEH, Reza MAHDAVI, Mohammad ASGHARI JAFARABADI and Nazli NAMAZI, Comparison of Indirect Calorimetry and Predictive Equations in Estimating Resting Metabolic Rate in Underweight Females. Iran J Public Health. 2015 Jun; 44(6): 822–829. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4524307/)
8. MJ Muller et al. World Health Organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. Am J Clin Nutr November 2004 vol. 80 no. 5 1379-1390
(http://ajcn.nutrition.org/content/80/5/1379.full)
9: https://www.uni-giessen.de/fbz/fb09/institute/ernaehrungswissenschaft/ag/neuhaeuser-berthold/forschung/veroffentlichungen/artikel/2004_1.pdf
10 Miller S1, Milliron BJ, Woolf K. Common Prediction Equations Overestimate Measured Resting Metabolic Rate in Young Hispanic Women.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3779143/
So you went on line and evaluated your BMR or TDEE. How good is it?
Let us look at the equations and how good or bad they are and then lets look at what you can do to create a personal working estimate that is more operationally accurate.
The RMR formulas are estimators. In General, "Four prediction equations were identified as the most commonly used in clinical practice (Harris-Benedict, Mifflin-St Jeor, Owen, and World Health Organization/Food and Agriculture Organization/United Nations University [WHO/FAO/UNU]). Of these equations, the Mifflin-St Jeor equation was the most reliable, predicting RMR within 10% of measured in more non obese and obese individuals than any other equation, and it also had the narrowest error range."
These equations are:
Harris-Benedict
Harris-Benedict Corrected
Mifflin St Jeor
BMR = ( 10 x weight in kg ) + ( 6.25 x height in cm ) – ( 5 x age in years ) + s where s = 5 for men and -161 for women
Ketch McArdle
BMR = 370 + 21.6 * LBM in kg
as an example, there are quite a few others...
- Men: BMR = 66.5 + ( 13.75 x weight in kg ) + ( 5.003 x height in cm ) – ( 6.755 x age in years )
- Women: BMR = 655.0955 + ( 9.5634 x weight in kg ) + ( 1.8496 x height in cm ) – ( 4.6756 x age in years )
Harris-Benedict Corrected
- BMR = 88.362 + ( 13.397 x weight in kg ) + ( 4.799 x height in cm ) – ( 6.755 x age in years )
- BMR = 447.593 + ( 9.247 x weight in kg ) + ( 3.098 x height in cm ) – ( 4.330 x age in years )
Mifflin St Jeor
BMR = ( 10 x weight in kg ) + ( 6.25 x height in cm ) – ( 5 x age in years ) + s where s = 5 for men and -161 for women
Ketch McArdle
BMR = 370 + 21.6 * LBM in kg
as an example, there are quite a few others...
The general convention is that these predictive BMR equations perform best for groups of people instead of individuals.(1) There are conflicting results in the literature as to which equation is best suited for a general patient population. For example, one analysis indicates that despite a clinically relevant error rate of 20%, the Mifflin–St Jeor equation has the most accuracy and lowest magnitude of error and should be used among healthy non-obese and obese adults, (1) while a more recent analysis indicates that the Harris–Benedict equation and two adaptations of the WHO/FAO/UNU equation outperform the Mifflin–St Jeor equation in this population (Brazil). (2) The WHO/FAO/UNU and Schofield equations (3) adapted for children appear to be the best estimates of resting EE in children and adolescents.(4) However, discussion cautions against using a single equation when estimating resting EE of adolescents across all racial groups, which warrants new equations be derived (5).
In theory under- or overestimations of 10% in the prediction of BMR seems insignificant for a population, but it can be much more important in an individual approach, particularly for active athletes, where 10% of BMR can stand for a 250 kcal difference or more (6).
In that study, the example was given of a subject whose BMR was underestimated in average by 150 Kcal (10% of BMR) and considering all the tested equations, his daily energy requirement would be underestimated in about 270–345 Kcal/day. They authors note, "This variation can be a major interference in this individual’s dietary prescription and would possibly result in a monthly loss of body weight of about 1,4kg, which could cause deficit in energy-protein recovery, increase the risk of fatigue and muscle damage among other factors that could interfere with the performance of a high level athlete. On the other hand, an overestimation would possibly result in body weight and body fat gain."
In underweight females these equations tend to overestimate RMR [7]. They note, "The commonly used predictive equations were not appropriate for underweight subjects... among 10 RMR predictive equations that were used in this study, Muller et al. equation gave a fairly acceptable RMR prediction, while most of the commonly used RMR predictive equations did not accurately predict RMR at both group and individual levels. Our data also showed that all of the equations except Muller and Abbreviation equations significantly overestimated RMR in underweight young females, with mean differences ranging from 42.2 to 224kcal/day."
These findings are confirmed by Müller (who developed another set of equations) "WHO prediction equations systematically overestimated REE at low REE values but underestimated REE at high REE values” [8]. These researches also developed a model using fat free mass and fat mass and found, surprisingly, that this type of measurement was not significantly more accurate. (see image below)
In young Hispanic women, it was found that the Owen equation was the only one that was predictive with some accuracy [10]....
I’ll stop here for now - there is actually a huge amount of research on the subject.
Here are my references:
1. Frankenfield D, Roth-Yousey L, Compher C. Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: a systematic review. J Am Diet Assoc. 2005;105:775–789
2. de Oliveira EP, Orsatti FL, Teixeira O, Maesta N, Burini RC. Comparison of predictive equations for resting energy expenditure in overweight and obese adults. J Obes. 2011;2011:534–714
3. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr. 1985;39(Suppl 1):5–41
4. Finan K, Larson DE, Goran MI. Cross-validation of prediction equations for resting energy expenditure in young, healthy children. J Am Diet Assoc. 1997;97:140–145.
5. Torun B, Davies PS, Livingstone MB, Paolisso M, Sackett R, Spurr GB. Energy requirements and dietary energy recommendations for children and adolescents 1 to 18 years old. Eur J Clin Nutr. 1996;50(Suppl 1):S37–S80. discussion S-1
6. Loureiro LL, Fonseca S Jr, Castro NG, Dos Passos RB, Porto CP1, Pierucci AP. Basal Metabolic Rate of Adolescent Modern Pentathlon Athletes: Agreement between Indirect Calorimetry and Predictive Equations and the Correlation with Body Parameters. PLoS One. 2015 Nov 16
7. Soghra ALIASGHARZADEH, Reza MAHDAVI, Mohammad ASGHARI JAFARABADI and Nazli NAMAZI, Comparison of Indirect Calorimetry and Predictive Equations in Estimating Resting Metabolic Rate in Underweight Females. Iran J Public Health. 2015 Jun; 44(6): 822–829. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4524307/)
8. MJ Muller et al. World Health Organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. Am J Clin Nutr November 2004 vol. 80 no. 5 1379-1390
(http://ajcn.nutrition.org/content/80/5/1379.full)
9: https://www.uni-giessen.de/fbz/fb09/institute/ernaehrungswissenschaft/ag/neuhaeuser-berthold/forschung/veroffentlichungen/artikel/2004_1.pdf
10 Miller S1, Milliron BJ, Woolf K. Common Prediction Equations Overestimate Measured Resting Metabolic Rate in Young Hispanic Women.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3779143/
1
Replies
-
So what should you do?
1) Use any estimator - if you are at the high or low end of BMI these will be a bit more off, but at the end, if you consider them as estimators, it doesn't matter much.
2) Track your own food calorie consumption. Average the calories you are eating over a period of time.
3) Based on weight loss over time. Estimate the amount of average loss in weight and convert that to an estimate of corresponding calories.
For example, if you lost 2 lbs over 30 days:
-- calories burned over the period 2lbs x 3500 cals / lb = 7500 cals
-- 7500 cals / 30 days = 250 cals / day
4) Add the calories burned from weight loss to the average calories eaten during the period gives you a working estimate of your own TDEE.
For the example given: if you are eating 1800 cals on average over those 30 days, then your operating TDEE is the food eaten plus the cals from weight loss = 1800 + 250 or about 2050.
(Note: if you gained weight during that period subtract the weight gain cals/day from the average eaten per day)
5) Continue to calculate this during different periods - weight loss, maintenance, increased activity. You'll see that it shifts and adjust accordingly
6) Profit. (Adjust your targets based on your own data.)
Or have it measured, but given that these values change based on non-lab parameters remember that they can still shift and you might need to adjust them.6 -
Have it measured ? Or back calculate your own value from self reported food intake and body weight / composition over say 4 weeks.2
-
-
I used MFP's calorie goal (without factoring exercise calories because I wasn't exercising) and logged for several months to build up data. From those numbers, I could see that I needed to eat about 150 fewer calories than estimated to lose at the predicted rate for my stats.
I've lost over 90 pounds in 15 months and I'm still burning 150 fewer calories per day than expected. My metabolism is slower than average but it hasn't gotten any slower as I've lost weight.
All of those equations give a starting point. If a person wants their own personal numbers, the only way to find those is through experimentation.2 -
I used MFP's calorie goal (without factoring exercise calories because I wasn't exercising) and logged for several months to build up data. From those numbers, I could see that I needed to eat about 150 fewer calories than estimated to lose at the predicted rate for my stats.
I've lost over 90 pounds in 15 months and I'm still burning 150 fewer calories per day than expected. My metabolism is slower than average but it hasn't gotten any slower as I've lost weight.
All of those equations give a starting point. If a person wants their own personal numbers, the only way to find those is through experimentation.
Doing it right.
1 -
EvgeniZyntx wrote: »So what should you do?
Use results?
When I first started and researched a bit, I discovered that they (well, HB and MSJ, at least) were expected to overstate BMR in obese subjects, which is not surprising if they are based on populations of healthy weight or even lean people (I think one was for athletes -- have not followed your links yet, so might be misremembering), as the average obese person will thus have a much higher fat % and fat and lean mass won't have the same effect on BMR.
Therefore, I figured my BMR/TDEE based on MSJ, and then also a range based on guesses at my likely BF% (which I didn't know with any accuracy) as a check. Playing around like that, I figured that MFP's number made sense and since I planned on a significant deficit (because obese) figured I'd just see how it worked. It worked fine -- I lost as expected or more, and I think the more had to do with activity level.
Once I did this for a few months I switched to TDEE using my actual results, although I again looked at the formulas just to see if my own number made sense. Yes, it was in the expected range. I suppose if it had been way low I'd have checked my logging or (maybe) seen a doctor/had it tested. (I later did have RMR tested as part of a DEXA/VO2 test -- it just came with it -- and found it simply confirmed what I'd already estimated, so wasn't valuable for me. I wouldn't pay for it, given a choice.)
Obviously this might be tougher if starting at a smaller intended deficit and from a position of a lot of activity (as any issue with BMR gets aggravated by a high activity multiplier), which is why I think it makes sense to also do things like log what you are eating for a week and cut from that or just pick a number that seems reasonable after looking at the various calculators and see.
I guess I don't see the calculators not giving a perfect number for an individual to be a big deal or surprising or really creating that much difficulty. Where I have found them somewhat reassuring is with all this (annoying) panic about the BL study results or people (like some particular posters on MFP) who like to claim that if you lose a bunch of weight you will always have to eat way lower than otherwise. That's not true for me -- maybe I have to eat less than if I'd never gotten fat, who knows, but what my TDEE is seems within the standard range as estimated by the calculators (which, yes, is a pretty wide range).2 -
lemurcat12 wrote: »I guess I don't see the calculators not giving a perfect number for an individual to be a big deal or surprising or really creating that much difficulty. Where I have found them somewhat reassuring is with all this (annoying) panic about the BL study results or people (like some particular posters on MFP) who like to claim that if you lose a bunch of weight you will always have to eat way lower than otherwise. That's not true for me -- maybe I have to eat less than if I'd never gotten fat, who knows, but what my TDEE is seems within the standard range as estimated by the calculators (which, yes, is a pretty wide range).
I agree with your method - what I also wanted to create was a thread to be able to point to when people get bogged down on "what equation should I use" or "but these aren't accurate" or "but they are wrong for me".
Yes, their use is a reference tool, move forward with your own data.
With regards to the BL study (and it's a tangent but hey, MFP) why did you find something so vastly uncertain as the BMR estimations reassuring? For me, the BL study just does not correspond to a real risk unless you were losing in a way that significantly depresses your hormonal function. If you lose at rates close to 1lb a day, with massive exercise levels and under-eating something is going to give. The equations aren't what reassure me - there is a lot of reports that these TDEE drops aren't seen with reasonable weight loss with correct protein and fat quantities.
0 -
The BL study used it's own regression on a tiny sample for RMR prediction / extrapolation. A different approach may have yielded a different conclusion. There's a study somewhere showing a fairly high variability in repeatedly measuring the RMR of the same subject(s).1
-
Eh, I overstated that, in that I didn't get panicked about the BL stuff or really understand why so many were, so I guess a better way of stating it (since I've been doing maintenance for over a year) is that I was already inoculated against the panic that some experienced from how the media portrayed it. (And I agree that it's not a real risk given the differences, and that helped too, of course.)
What I was more thinking of are some MFP posters who like to claim (citing studies) that we are all doomed because if you lose lots of weight (which I did) your TDEE will be 200 or more calories (or 20% or some such number) below what it should be, so we are all going to be hungry. My response to that has been, well, my TDEE seems to be a reasonable number, and although I am quite active the calculators are in line with the current number, so if it's off, it's still not giving me some kind of unsurmountable challenge vs. people who never got fat.
I have also seen some of the studies you refer to, and find those reassuring also, but obviously seeing my own results is even more reassuring.1 -
lemurcat12 wrote: »EvgeniZyntx wrote: »So what should you do?
I guess I don't see the calculators not giving a perfect number for an individual to be a big deal or surprising or really creating that much difficulty. Where I have found them somewhat reassuring is with all this (annoying) panic about the BL study results or people (like some particular posters on MFP) who like to claim that if you lose a bunch of weight you will always have to eat way lower than otherwise. That's not true for me -- maybe I have to eat less than if I'd never gotten fat, who knows, but what my TDEE is seems within the standard range as estimated by the calculators (which, yes, is a pretty wide range).
My TDEE (whether active or not) is about 150-200 calories lower than MFP predicts (the more active the bigger the difference). It was true when I first started, and it's still true 85 pounds later. I haven't seen that astronomic decrease in my TDEE that the study has shown. Yes, it inched a bit closer to 200 than to 150, but that's a near negligible 10-15 calories difference for 85 pounds of loss.1 -
I know about what my TDEE is, but I've often been intrigued by what my BMR is. The issue is that minor variations in the activity multiplier will greatly affect BMR when we're talking about small BMR deviations from the norm.0
-
EvgeniZyntx wrote: »So what should you do?
1) Use any estimator - if you are at the high or low end of BMI these will be a bit more off, but at the end, if you consider them as estimators, it doesn't matter much.
2) Track your own food calorie consumption. Average the calories you are eating over a period of time.
3) Based on weight loss over time. Estimate the amount of average loss in weight and convert that to an estimate of corresponding calories.
For example, if you lost 2 lbs over 30 days:
-- calories burned over the period 2lbs x 3500 cals / lb = 7500 cals
-- 7500 cals / 30 days = 250 cals / day
4) Add the calories burned from weight loss to the average calories eaten during the period gives you a working estimate of your own TDEE.
For the example given: if you are eating 1800 cals on average over those 30 days, then your operating TDEE is the food eaten plus the cals from weight loss = 1800 + 250 or about 2050.
(Note: if you gained weight during that period subtract the weight gain cals/day from the average eaten per day)
5) Continue to calculate this during different periods - weight loss, maintenance, increased activity. You'll see that it shifts and adjust accordingly
6) Profit. (Adjust your targets based on your own data.)
Or have it measured, but given that these values change based on non-lab parameters remember that they can still shift and you might need to adjust them.
The part I like about the feedback loop is it accounts for inaccuracies in how you log your food, inaccuracies in food labels, as well as differences in your actual TDEE vs. calculated, as long as you are consistent.
Based on my logging, my sedentary TDEE is ~200 calories higher than predicted by MFP.2 -
tagging to read later.0
-
The_Enginerd wrote: »EvgeniZyntx wrote: »So what should you do?
1) Use any estimator - if you are at the high or low end of BMI these will be a bit more off, but at the end, if you consider them as estimators, it doesn't matter much.
2) Track your own food calorie consumption. Average the calories you are eating over a period of time.
3) Based on weight loss over time. Estimate the amount of average loss in weight and convert that to an estimate of corresponding calories.
For example, if you lost 2 lbs over 30 days:
-- calories burned over the period 2lbs x 3500 cals / lb = 7500 cals
-- 7500 cals / 30 days = 250 cals / day
4) Add the calories burned from weight loss to the average calories eaten during the period gives you a working estimate of your own TDEE.
For the example given: if you are eating 1800 cals on average over those 30 days, then your operating TDEE is the food eaten plus the cals from weight loss = 1800 + 250 or about 2050.
(Note: if you gained weight during that period subtract the weight gain cals/day from the average eaten per day)
5) Continue to calculate this during different periods - weight loss, maintenance, increased activity. You'll see that it shifts and adjust accordingly
6) Profit. (Adjust your targets based on your own data.)
Or have it measured, but given that these values change based on non-lab parameters remember that they can still shift and you might need to adjust them.
The part I like about the feedback loop is it accounts for inaccuracies in how you log your food, inaccuracies in food labels, as well as differences in your actual TDEE vs. calculated, as long as you are consistent.
Based on my logging, my sedentary TDEE is ~200 calories higher than predicted by MFP.
I agree and this is what I have implemented for quite some time.1
Categories
- All Categories
- 1.4M Health, Wellness and Goals
- 391.3K Introduce Yourself
- 43.4K Getting Started
- 259.6K Health and Weight Loss
- 175.6K Food and Nutrition
- 47.3K Recipes
- 232.3K Fitness and Exercise
- 387 Sleep, Mindfulness and Overall Wellness
- 6.4K Goal: Maintaining Weight
- 8.5K Goal: Gaining Weight and Body Building
- 152.7K Motivation and Support
- 7.8K Challenges
- 1.3K Debate Club
- 96.2K Chit-Chat
- 2.5K Fun and Games
- 3.2K MyFitnessPal Information
- 22 News and Announcements
- 911 Feature Suggestions and Ideas
- 2.3K MyFitnessPal Tech Support Questions