With just this one sample, I'd guess a lab error or test miscalibration over anything else. It'll be interesting what comes out of the consultation. See if you can convince them to let you do it again!
They were as surprised as me and offered me a free retest. They say they've only ever had 1 measurement error in thousands of samples, but that it is a possibility.
Well, there are tests they can run that indicate the sample was definitely tainted. They told me the other guy (the 1 case) came back as having -33% body fat, because his water had less deuterium after drinking the deuterium drink than before.
Other things are; if they see somebody at nearly 2.5x BMR total burn, they typically see that person doing 5h of endurance training a day. They don't see essentially sedentary people with that TDEE.
Well, it does at the extremes - if you go from comatose to office worker, there's a jump. And if you go near olympic level exercise, there's also a jump. Just not in the range that most people will move in when "exercising more." It's also not clear which way the causality goes here - maybe the people who become professional athletes are the ones who CAN increase their total energy availability..
Well OK, so we're supposed to believe that all moderate levels of exercise make no difference, but once you do 5hours of endurance training every day (I have done this sort of thing occasionally and have never noticed any sort of insane phase-shift going on), suddenly you start burning three times as much energy?
I am currently rather sceptical of the thesis presented in 'Burn', pending an explanation of what is going on with your laws-of-physics defying result.
I contacted them about this, if it could be an indicator that the sample was tainted, but they say they considered this and no that's pretty normal. I think he said in the consultation that it includes the water in all foods & including "metabolic water" so it isn't that I drank 6.8l of water.
How much water/fluids do you typically drink? 6.8 liters or 7 quarts is a ridiculous amount of water and could easily cause hyponatremia if you were actually drinking that amount. The standard recommendation is eight cups/2 quarts of liquids per day. Some people insist that one gallon a day is best. I doubt you were drinking 1.8 gallons which raises a lot of questions!
I don't drink 6.8l, that's for sure, but haven't measured it. Probably 2-4l on top of food?
But they say this includes "metabolic water" from presumably fat turnover and so on. I asked specifically, and they were not concerned about my water number, that seemed pretty normal.
I'm inclined to think your last idea is correct: perhaps the double labeled water approach only applies to standard diets. It would be worth investigating whether there is research that demonstrates its accuracy for ketogenic diets and especially high fat/ low protein ketogenic diets.
Yeah, I think I am done with the low protein keto experiment and will be trying to eat a more normal keto diet for the next week or so and see if I stop feeling like absolute miserable dogshit all the time.
I think it's probably a very bad idea to do something that makes you feel awful.
For the avoidance of doubt I feel way better on ex150ish than I do on a normal diet, and this fortnight I've been playing with high-carb low-fat low-protein, although not in any systematic way, and that makes me feel absolutely stellar most of the time (a little over-stellar, truth be told) and occasionally dreadful.
Hard to say what's what, though. As much as we shit on bad science, there's a reason we have science - hard to control for all the variables in a human life.
I quit potato after trying to also incorporate honey/fruit - - had to go low-carb to recover from that debacle. For all I know, I've felt bad ever since because of after-effects of the potato, or the runaway eating-too-much-sugar.
Vegans (and exfatloss himself in 10 years on the wrong keto) fall into this - - hard to tell which part of what is making you feel which way which time.
Error's in the wrong direction for tapeworms isn't it? He's supposed to be metabolising twice as much energy as he's eating, i.e. he's somehow making more heat from food than can be obtained by complete combustion.
Any sort of inefficiency just makes the paradox worse.
This is a 'broken the laws of physics' result. At that point it's either 'measurement error' or 'the greatest discovery in a hundred years'. The last time something like this happened for real the answer to the paradox was the discovery of radioactivity.
Indeed, I haven't quite abandoned conservation of energy yet. Although as a good Bayesian my faith is ever so slightly shaken in a way that is far too hard to calculate.
I don't understand how DLW is supposed to work, although it's a cool idea. I had a look but the infallibility of the method is not leaping out at me.
I hope the problem is something interesting about metabolism and not just "Oh flaps we need to allow two years for all the isotopes to come into equilibrium before the results make sense".
Was thinking about this today; what DLW measures is "how much ATP is created from glucose/fat" I think. I think that doesn't mean I used that energy. Could it somehow be recycled back into the cycle, so every ATP kinda goes through there twice?
You know how people say "oh there was 10x more transaction volume than money exists!" But the answer is "Yea because the velocity of money is 10/yr." So maybe the same cream energy can go through the Krebs cycle twice?
I think it's just trying to measure how much carbon dioxide you exhale. You are only supposed to be doing that by burning carbon-containing molecules.
You get more energy out of a six-carbon fat than you do out of six-carbon glucose so you need to account for that somehow (that's the 'respiratory quotient', I think), and there are slight errors to do with an eighteen-carbon fat being worth slightly less than three six-carbon fats etc too. So to turn exhaled carbon into 'energy used' you need to know exactly what's being burned.
Of course you can also burn protein for energy by various pathways, and maybe that's the problem?
The only thing you can actually measure is the O/H isotope ratio going in and out (through one particular route), so that gives you two numbers, and I can't see how you can do more with those two numbers than calculate a fairly wide range of possible metabolic rates.
And even then, I don't see why you're expecting metabolism to be actually using the funny oxygen, why can't you just piss it all out unchanged, giving a zero metabolic rate for a living system?
It's not at all clear to me that drinking funny water is going to suddenly mean that all the oxygen and hydrogen in your body becomes funny at the same ratio, it seems that some molecules are going to exchange their oxygen with surrounding water fairly quickly and some are never going to do it at all, and I think that without that assumption you can get any answer you like.
But as I say, I don't understand the idea at all. Is there an accessible description of the calculations and necessary assumptions floating around?
> why can't you just piss it all out unchanged, giving a zero metabolic rate for a living system?
I suppose because the body can't tell the difference, and all the water/oxygen in your body is fairly equally mixed & fungible.
They seem pretty convinced that all the body in your water becomes funny at a similar ratio within a few hours of drinking funny water. I'll admit I haven't tested this, but it seems like they've thought about this & verified it versus other ways of measuring CO2 like metabolic chambers.
There are a bunch of papers on DLW, but I haven't read them. You could check if they contain any sketchy assumptions.
> They seem pretty convinced that all the body in your water becomes funny at a similar ratio within a few hours of drinking funny water. I'll admit I haven't tested this, but it seems like they've thought about this & verified it versus other ways of measuring CO2 like metabolic chambers.
I can sort of believe this, water splits apart and reforms constantly, so if you put a bit of funny water in a big tank of sensible water it would probably become a big tank of slightly-funny water quite quickly.
But the body is a complex thing, and there are a lot of different molecules in there that aren't water but that contain both hydrogen and oxygen in various amounts. And maybe those molecules exchange their hydrogen with the water more easily than they exchange their oxygen, who knows?
For instance I imagine that in a sugar it's a lot easier to swap the hydrogens with funny water hydrogens than it is to swap the double-bonded oxygen in the ketone group with a funny water oxygen.
And there are lots of membranes and barriers which may or may not be easy for the funniness to cross, and so on....
But I do not know much about all this so I will confine myself to being just being mildly sceptical. Maybe it has all been carefully checked out.
So apparently Kevin Hall did a follow-up study on the one you dug up, which the nice folks at Calorify sent me after I let them know of the crazy outlier one:
Unfortunately this is peak Kevin Hall: he simply decides to exclude the people who break CICO, because CICO is dogma and cannot be broken. Institutionalized disinterest lol.
Maybe? I don't understand how it's supposed to work or what assumptions need to be made. And it doesn't look like anyone bothered checking.
I think in this particular case it's probably just mismeasuring our host's metabolic rate by a factor of two., which makes me sceptical of all the other claims based on it. You can't find things out with broken tools.
I don't really think he has tape worms that was a joke. And, no, it is likely just different scales. One way of measuring calories might as well be in metric and the other imperial. Apparently 2700 calorimeter calories are 4400 doubly labeled water calories and we might as well call them two different things.
They seem pretty confident and claim (incl. Pontzer in the book) that it's validated against metabolic chambers. So this different metric might only apply in edge cases, maybe protein restriction.
The doubly labeled water (DLW) method has been successfully applied to measuring energy expenditure in humans since the 1980s and has been validated in comparison to respiratory chamber measurements in subjects consuming moderate carbohydrate diets.
The DLW has never been validated in humans consuming a very low-carbohydrate diet and there are theoretical reasons why the DLW method might result in a systematic bias when calculating energy expenditure differences between diets widely varying in carbohydrate.
Ok I just read it. Very interesting and I think this happened to me. They have an n=17 study comparing a mixed macro to a keto diet. Their keto diet is higher in carb and higher in protein each than I eat combined. I.e. I am significantly more ketarded than their subjects.
Most people had somewhat similar carolies measured in metabolic ward, DLW, and "CICO + DEXA." But there were 2 outliers who had a difference of over 1,000kcal/day, by which DLW was higher, whereas metabolic ward and CICO+DEXA roughly aligned.
They speculate that deuterium could be lost in other ways besides beta oxidation, e.g. in DNL or something like that.
So I'm not the only one, and the phenomenon seems to be in the same direction to a similar degree.
I think something's just off in the assumptions DLW makes, which is wrong in certain ketogenic conditions.
Nice, good to see someone's actually checking that DLW and metabolic ward and CICO+DEXA give the same readings.
I can't believe they've had this technique for fifty years and never tried it on 'something eating an unbalanced diet'.
------
On the possible technical explanation:
If we're assuming that the deuterium is in equilibrium throughout the body at all times, I don't think it matters what happens to it does it?
And if we're not, then how is the DLW technique supposed to get any answers at all?
Water can leave in sweat or piss or be breathed out, and there are loads of other ways out for hydrogen in general I would have thought, or it could just concentrate in some tissues somewhere as part of almost any molecule and never come out at all.
"Maybe it’s somehow not calibrated for extreme (90%+ fat) ketogenic diets?"
There does seem to be some concern that doubly labeled water may incorrectly report more calories burned when a person is on a low-carb diet. But they're talking about mere 100-200 calorie discrepancies, though, and all the authors seem deep into keto-vs-CICO-defend-my-tribe bias mode, so who knows what the real story is.
I can definitely tell that my body handles water way different when I'm keto. I wouldn't be surprised at all if infinite cream just breaks this measurement method. In any case, these are really cool results. Thanks for taking the money hit.
Thanks for this article. Maybe I missed it but how do I increase ucp-1? Keep protein intake really low? I feel like i already do this but can't seem to lose the last layer of far around stomach and waist. Maybe my protein intake isn't as low as I thought?
I had a similarly confusing result: a few years back I had my RMR measured by that CO2 method. I was a 30 year old woman, sedentary (also a programmer that works from home). It was measured as 2400 calories. I was eating 1400 calories a day at the time.
Sorry I am very late to the party here. Jaw-dropping stuff, everyone's questions make sense. The one I did not see being asked... Is your real name Michael Phelps, Experimental Fat Loss?? :-)
The next one you should try is the Lumen. I bought one, found it rather useless, but maybe because I don't do Keto, and I was already in maintenance mode, and their app keeps pushing you in the keto and weight loss direction.
OK, I said I wouldn't be that surprised by whatever result you got, but I was thinking 10% here, 10% there........
I notice that I am *extremely* surprised!
I trust you to have measured your 2850 carolies in fairly accurately, I trust you to have noticed your weight changing much. I guess I do *kind* of trust the people who do food labels to have got the numbers right to within ten per cent or so. They measure 4645kcal out. Call it 1800kcal/day difference (+/- 10% for all sorts of inaccuracies in food labelling etc.)
Fat is the most efficient way to store energy in a human body, and my sources inform me that 1kg of human fat is 7700kcal. So if the above numbers are true that means that you should be losing at least 0.23kg/day.
No way an effect like that is getting lost in the noise.
And on the face of it, the error is in the direction that can't happen. You're getting more energy out of your food than complete combustion does. That *can't* be true without a major revision to everything we know about physics and chemistry. It's physically impossible.
So unless you've somehow missed an extra half-a-pound of fat that you're eating every day and not counting, or your scales are going really badly out of calibration really quickly, I reckon this doubly-labelled water technique must be broken.
I think I'd judge these people by how seriously they take this obviously impossible result.
Would you be happy to go and spend a week at their lab, letting them control everything you eat and living in a room calorimeter so that they can actually measure exactly what you're eating and your total heat output? Because if I were them that is what I would want you to do.
I laughed out loud when I opened the PDF report :) So it's not just you.
Yea, I'm probably not creating energy out of thin air or thick water. I'll do a (free! yay) retest in a month or so. They say they don't exactly expect the sample to be tainted, since they don't see anything obvious wrong with it, but it cannot be ruled out.
If the result is confirmed, it probably just means that DLW needs to calibrate for low-protein (or something else?) as well as for respiratory quotient (glucose/fat burned). I do wonder how much that would affect all previously collected DLW samples, lol.
Half-a-pound of fat is 1.5lbs of cream, which is pretty much my total daily cream consumption (650ml). I think I would've noticed accidentally doubling my cream intake.
The people seemed quite interested and offered me a free re-test. So I think they are taking this pretty seriously.
If they offer to fly me to a lab, I'd do it in a heartbeat :) I love sitting around indoors.
With any luck you've achieved cold fusion in your mitochondria and you're using it to strip the extra neutrons out of the labelled atoms, or something. Or maybe you've got an enzyme that can catalyse radioactive decay! Have you got a Gieger counter handy?
Interest is the right response here though. It's a shame about their entire life's work, but the scientist is the man who says "Oh bugger, I take that all back, still, at least we know more about doubly-labelled water as a measurement technique now". Well done them!
I thought it was a bit odd that spending six hours every day walking around didn't increase energy expenditure.
But in fact, science has always proceeded by the invention of better ways of measuring things. So even if the central argument of Burn turns out to actually be true, sorting out DLW metabolism and turning it into a useful tool would probably be the greater contribution.
If they offer to fly me to a lab, I'd do it in a heartbeat :) I love sitting around indoors.
I thought you might be up for that. I can't imagine they'll be able to get you away from your laptop but it should be easy enough to measure how much power it draws when you're sitting in the sealed room calorimeter. I do hope the wifi is good!
Hooboy! I owe you a quick apology (and Substack owes me one).
The comment I originally wrote here was intended for Astral Codex Ten's mock-post of the presidential debate. I read that one in my email and pressed the button to reply to it but for some strange reason when it sent me to the app the comment wound up on this post. You got the comment equivalent of a pocket dial.
Fun pocket dial tho. Sorry about that.
If you go read Astral Codex Ten's post you'll get the context of the comment. It will make a lot more sense.
"it gives you a 30 minute consultation with their team that I’m looking forward to."
haha, I'm sure they'll enjoy the experience equally...
Haha yea they were just as confused/surprised as me.
watch out, they'll kidnap you to a secret black lab and do experiments on you, and not the fun "eat all the heavy cream you want" kind!
It's probably ok to take a laptop into a room calorimeter, as long as it's on a metered power supply. I doubt he'll notice.
Today I learned my lifestyle is called "metabolic ward"
With just this one sample, I'd guess a lab error or test miscalibration over anything else. It'll be interesting what comes out of the consultation. See if you can convince them to let you do it again!
They were as surprised as me and offered me a free retest. They say they've only ever had 1 measurement error in thousands of samples, but that it is a possibility.
"Our computers have literally never suffered from an undetected malfunction"
Well, there are tests they can run that indicate the sample was definitely tainted. They told me the other guy (the 1 case) came back as having -33% body fat, because his water had less deuterium after drinking the deuterium drink than before.
Other things are; if they see somebody at nearly 2.5x BMR total burn, they typically see that person doing 5h of endurance training a day. They don't see essentially sedentary people with that TDEE.
> Other things are; if they see somebody at nearly 2.5x BMR total burn, they typically see that person doing 5h of endurance training a day.
Hang on, wasn't the whole point of that book that it doesn't make a great deal of difference?
Well, it does at the extremes - if you go from comatose to office worker, there's a jump. And if you go near olympic level exercise, there's also a jump. Just not in the range that most people will move in when "exercising more." It's also not clear which way the causality goes here - maybe the people who become professional athletes are the ones who CAN increase their total energy availability..
Well OK, so we're supposed to believe that all moderate levels of exercise make no difference, but once you do 5hours of endurance training every day (I have done this sort of thing occasionally and have never noticed any sort of insane phase-shift going on), suddenly you start burning three times as much energy?
I am currently rather sceptical of the thesis presented in 'Burn', pending an explanation of what is going on with your laws-of-physics defying result.
Whatever it turns out to be, it really is cool that you're part of doing some sciencing!
Gotta science the s&(*t out of this
Just don't start a potato habit and do not blow up your habitat with rocket fuel.
Typical rookie mistake. Also don't eat the percocet.
You broke the test.
It gives your water intake as 6.8 litres. That is unrealistic at best
I contacted them about this, if it could be an indicator that the sample was tainted, but they say they considered this and no that's pretty normal. I think he said in the consultation that it includes the water in all foods & including "metabolic water" so it isn't that I drank 6.8l of water.
3.7L is recommended, 20% of which comes from food.
If we take 4kcal/g for carbs, then you consume 1kg of carbs per day on 4000kcal
1mol of glucose produces 6 mol of water when completely consumed. So if you ran just on carbs, you'd produce 600g of water per day
Similarly, 1 mol of stearic acid produces 18 mol of water. For 4000kcal you'd need 500g at 8kcal/g. That produces 570g of water.
But of course if 6.8L is what they typically see, then that's that
How much water/fluids do you typically drink? 6.8 liters or 7 quarts is a ridiculous amount of water and could easily cause hyponatremia if you were actually drinking that amount. The standard recommendation is eight cups/2 quarts of liquids per day. Some people insist that one gallon a day is best. I doubt you were drinking 1.8 gallons which raises a lot of questions!
I don't drink 6.8l, that's for sure, but haven't measured it. Probably 2-4l on top of food?
But they say this includes "metabolic water" from presumably fat turnover and so on. I asked specifically, and they were not concerned about my water number, that seemed pretty normal.
Really? What's a normal intake supposed to be? He actually went over that in the consultation and didn't seem to think it weird.
I'm inclined to think your last idea is correct: perhaps the double labeled water approach only applies to standard diets. It would be worth investigating whether there is research that demonstrates its accuracy for ketogenic diets and especially high fat/ low protein ketogenic diets.
I suspect nobody knows how much calories are in food or how much we get out of them. Also you probably got tapeworms.
Meanwhile I eat the same food you do and am too exhausted to move sometimes, though i've been on keto 2 months instead of 10 years.
> Meanwhile I eat the same food you do and am too exhausted to move sometimes
That doesn't sound right. Tell us more!
Yeah, I think I am done with the low protein keto experiment and will be trying to eat a more normal keto diet for the next week or so and see if I stop feeling like absolute miserable dogshit all the time.
I think it's probably a very bad idea to do something that makes you feel awful.
For the avoidance of doubt I feel way better on ex150ish than I do on a normal diet, and this fortnight I've been playing with high-carb low-fat low-protein, although not in any systematic way, and that makes me feel absolutely stellar most of the time (a little over-stellar, truth be told) and occasionally dreadful.
Hard to say what's what, though. As much as we shit on bad science, there's a reason we have science - hard to control for all the variables in a human life.
I quit potato after trying to also incorporate honey/fruit - - had to go low-carb to recover from that debacle. For all I know, I've felt bad ever since because of after-effects of the potato, or the runaway eating-too-much-sugar.
Vegans (and exfatloss himself in 10 years on the wrong keto) fall into this - - hard to tell which part of what is making you feel which way which time.
Oh, yes, if the badness started before you did the new thing there's not much you can read into that.
I thought you were in good health and just curious about all this for intellectual reasons. What are you trying to achieve?
It's a lot more complex than that, and I am still trying to figure it out.
A teaser, however, is that I appear to have fallen prey simultaneously to hill-climbing and goodhart's law.
But yeah I'm healthy enough, just out here climbing local optima and getting stranded on them.
> Also you probably got tapeworms.
Error's in the wrong direction for tapeworms isn't it? He's supposed to be metabolising twice as much energy as he's eating, i.e. he's somehow making more heat from food than can be obtained by complete combustion.
Any sort of inefficiency just makes the paradox worse.
This is a 'broken the laws of physics' result. At that point it's either 'measurement error' or 'the greatest discovery in a hundred years'. The last time something like this happened for real the answer to the paradox was the discovery of radioactivity.
More likely, "discovery that DLW doesn't apply at current ratios in some edge case."
Indeed, I haven't quite abandoned conservation of energy yet. Although as a good Bayesian my faith is ever so slightly shaken in a way that is far too hard to calculate.
I don't understand how DLW is supposed to work, although it's a cool idea. I had a look but the infallibility of the method is not leaping out at me.
I hope the problem is something interesting about metabolism and not just "Oh flaps we need to allow two years for all the isotopes to come into equilibrium before the results make sense".
I am cold fusion.
Was thinking about this today; what DLW measures is "how much ATP is created from glucose/fat" I think. I think that doesn't mean I used that energy. Could it somehow be recycled back into the cycle, so every ATP kinda goes through there twice?
You know how people say "oh there was 10x more transaction volume than money exists!" But the answer is "Yea because the velocity of money is 10/yr." So maybe the same cream energy can go through the Krebs cycle twice?
I think it's just trying to measure how much carbon dioxide you exhale. You are only supposed to be doing that by burning carbon-containing molecules.
You get more energy out of a six-carbon fat than you do out of six-carbon glucose so you need to account for that somehow (that's the 'respiratory quotient', I think), and there are slight errors to do with an eighteen-carbon fat being worth slightly less than three six-carbon fats etc too. So to turn exhaled carbon into 'energy used' you need to know exactly what's being burned.
Of course you can also burn protein for energy by various pathways, and maybe that's the problem?
The only thing you can actually measure is the O/H isotope ratio going in and out (through one particular route), so that gives you two numbers, and I can't see how you can do more with those two numbers than calculate a fairly wide range of possible metabolic rates.
And even then, I don't see why you're expecting metabolism to be actually using the funny oxygen, why can't you just piss it all out unchanged, giving a zero metabolic rate for a living system?
It's not at all clear to me that drinking funny water is going to suddenly mean that all the oxygen and hydrogen in your body becomes funny at the same ratio, it seems that some molecules are going to exchange their oxygen with surrounding water fairly quickly and some are never going to do it at all, and I think that without that assumption you can get any answer you like.
But as I say, I don't understand the idea at all. Is there an accessible description of the calculations and necessary assumptions floating around?
> why can't you just piss it all out unchanged, giving a zero metabolic rate for a living system?
I suppose because the body can't tell the difference, and all the water/oxygen in your body is fairly equally mixed & fungible.
They seem pretty convinced that all the body in your water becomes funny at a similar ratio within a few hours of drinking funny water. I'll admit I haven't tested this, but it seems like they've thought about this & verified it versus other ways of measuring CO2 like metabolic chambers.
There are a bunch of papers on DLW, but I haven't read them. You could check if they contain any sketchy assumptions.
> They seem pretty convinced that all the body in your water becomes funny at a similar ratio within a few hours of drinking funny water. I'll admit I haven't tested this, but it seems like they've thought about this & verified it versus other ways of measuring CO2 like metabolic chambers.
I can sort of believe this, water splits apart and reforms constantly, so if you put a bit of funny water in a big tank of sensible water it would probably become a big tank of slightly-funny water quite quickly.
But the body is a complex thing, and there are a lot of different molecules in there that aren't water but that contain both hydrogen and oxygen in various amounts. And maybe those molecules exchange their hydrogen with the water more easily than they exchange their oxygen, who knows?
For instance I imagine that in a sugar it's a lot easier to swap the hydrogens with funny water hydrogens than it is to swap the double-bonded oxygen in the ketone group with a funny water oxygen.
And there are lots of membranes and barriers which may or may not be easy for the funniness to cross, and so on....
But I do not know much about all this so I will confine myself to being just being mildly sceptical. Maybe it has all been carefully checked out.
So apparently Kevin Hall did a follow-up study on the one you dug up, which the nice folks at Calorify sent me after I let them know of the crazy outlier one:
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0225944
Unfortunately this is peak Kevin Hall: he simply decides to exclude the people who break CICO, because CICO is dogma and cannot be broken. Institutionalized disinterest lol.
DLW would hypothetically be also measuring the worms' metabolism.
Maybe? I don't understand how it's supposed to work or what assumptions need to be made. And it doesn't look like anyone bothered checking.
I think in this particular case it's probably just mismeasuring our host's metabolic rate by a factor of two., which makes me sceptical of all the other claims based on it. You can't find things out with broken tools.
I don't really think he has tape worms that was a joke. And, no, it is likely just different scales. One way of measuring calories might as well be in metric and the other imperial. Apparently 2700 calorimeter calories are 4400 doubly labeled water calories and we might as well call them two different things.
They seem pretty confident and claim (incl. Pontzer in the book) that it's validated against metabolic chambers. So this different metric might only apply in edge cases, maybe protein restriction.
From:
https://www.biorxiv.org/content/10.1101/403931v1
What is already known about this subject?
The doubly labeled water (DLW) method has been successfully applied to measuring energy expenditure in humans since the 1980s and has been validated in comparison to respiratory chamber measurements in subjects consuming moderate carbohydrate diets.
The DLW has never been validated in humans consuming a very low-carbohydrate diet and there are theoretical reasons why the DLW method might result in a systematic bias when calculating energy expenditure differences between diets widely varying in carbohydrate.
Ok I just read it. Very interesting and I think this happened to me. They have an n=17 study comparing a mixed macro to a keto diet. Their keto diet is higher in carb and higher in protein each than I eat combined. I.e. I am significantly more ketarded than their subjects.
Most people had somewhat similar carolies measured in metabolic ward, DLW, and "CICO + DEXA." But there were 2 outliers who had a difference of over 1,000kcal/day, by which DLW was higher, whereas metabolic ward and CICO+DEXA roughly aligned.
They speculate that deuterium could be lost in other ways besides beta oxidation, e.g. in DNL or something like that.
So I'm not the only one, and the phenomenon seems to be in the same direction to a similar degree.
I think something's just off in the assumptions DLW makes, which is wrong in certain ketogenic conditions.
Good find, thanks!
Nice, good to see someone's actually checking that DLW and metabolic ward and CICO+DEXA give the same readings.
I can't believe they've had this technique for fifty years and never tried it on 'something eating an unbalanced diet'.
------
On the possible technical explanation:
If we're assuming that the deuterium is in equilibrium throughout the body at all times, I don't think it matters what happens to it does it?
And if we're not, then how is the DLW technique supposed to get any answers at all?
Water can leave in sweat or piss or be breathed out, and there are loads of other ways out for hydrogen in general I would have thought, or it could just concentrate in some tissues somewhere as part of almost any molecule and never come out at all.
Interesting!
"Maybe it’s somehow not calibrated for extreme (90%+ fat) ketogenic diets?"
There does seem to be some concern that doubly labeled water may incorrectly report more calories burned when a person is on a low-carb diet. But they're talking about mere 100-200 calorie discrepancies, though, and all the authors seem deep into keto-vs-CICO-defend-my-tribe bias mode, so who knows what the real story is.
https://www.biorxiv.org/content/10.1101/403931v1
https://www.biorxiv.org/content/10.1101/403931v1
https://www.researchgate.net/publication/336982233_Methodological_error_in_measurement_of_energy_expenditure_by_the_doubly_labeled_water_method_much_ado_about_nothing
I can definitely tell that my body handles water way different when I'm keto. I wouldn't be surprised at all if infinite cream just breaks this measurement method. In any case, these are really cool results. Thanks for taking the money hit.
Thanks for this article. Maybe I missed it but how do I increase ucp-1? Keep protein intake really low? I feel like i already do this but can't seem to lose the last layer of far around stomach and waist. Maybe my protein intake isn't as low as I thought?
Yes, in the studies I've seen it's protein intake of about 7% of total kcals. Which is <50g for most people.
This might also not be the end all be all of fat loss, after all I'm plateauing again this summer :)
How much protein do you eat?
Ah ok I see. I don't a lot of protein mostly. Some days maybe about 50g and other days slightly more. I should probably start counting properly
Specifically, isoleucine
I had a similarly confusing result: a few years back I had my RMR measured by that CO2 method. I was a 30 year old woman, sedentary (also a programmer that works from home). It was measured as 2400 calories. I was eating 1400 calories a day at the time.
And presumably you weren't losing a pound of fat every 3 days?
At the time, on average, I was losing nothing.
Yea we really can' t measure anything, hah
Sorry I am very late to the party here. Jaw-dropping stuff, everyone's questions make sense. The one I did not see being asked... Is your real name Michael Phelps, Experimental Fat Loss?? :-)
The next one you should try is the Lumen. I bought one, found it rather useless, but maybe because I don't do Keto, and I was already in maintenance mode, and their app keeps pushing you in the keto and weight loss direction.
I am not Michael Phelps, shockingly ;)
I've read about the Lumen but I just think their idea of "optimize RQ to your goal" is nonsensical. AFAIK there is just nothing behind it.
OK, I said I wouldn't be that surprised by whatever result you got, but I was thinking 10% here, 10% there........
I notice that I am *extremely* surprised!
I trust you to have measured your 2850 carolies in fairly accurately, I trust you to have noticed your weight changing much. I guess I do *kind* of trust the people who do food labels to have got the numbers right to within ten per cent or so. They measure 4645kcal out. Call it 1800kcal/day difference (+/- 10% for all sorts of inaccuracies in food labelling etc.)
Fat is the most efficient way to store energy in a human body, and my sources inform me that 1kg of human fat is 7700kcal. So if the above numbers are true that means that you should be losing at least 0.23kg/day.
No way an effect like that is getting lost in the noise.
And on the face of it, the error is in the direction that can't happen. You're getting more energy out of your food than complete combustion does. That *can't* be true without a major revision to everything we know about physics and chemistry. It's physically impossible.
So unless you've somehow missed an extra half-a-pound of fat that you're eating every day and not counting, or your scales are going really badly out of calibration really quickly, I reckon this doubly-labelled water technique must be broken.
I think I'd judge these people by how seriously they take this obviously impossible result.
Would you be happy to go and spend a week at their lab, letting them control everything you eat and living in a room calorimeter so that they can actually measure exactly what you're eating and your total heat output? Because if I were them that is what I would want you to do.
> I notice that I am *extremely* surprised!
I laughed out loud when I opened the PDF report :) So it's not just you.
Yea, I'm probably not creating energy out of thin air or thick water. I'll do a (free! yay) retest in a month or so. They say they don't exactly expect the sample to be tainted, since they don't see anything obvious wrong with it, but it cannot be ruled out.
If the result is confirmed, it probably just means that DLW needs to calibrate for low-protein (or something else?) as well as for respiratory quotient (glucose/fat burned). I do wonder how much that would affect all previously collected DLW samples, lol.
Half-a-pound of fat is 1.5lbs of cream, which is pretty much my total daily cream consumption (650ml). I think I would've noticed accidentally doubling my cream intake.
The people seemed quite interested and offered me a free re-test. So I think they are taking this pretty seriously.
If they offer to fly me to a lab, I'd do it in a heartbeat :) I love sitting around indoors.
With any luck you've achieved cold fusion in your mitochondria and you're using it to strip the extra neutrons out of the labelled atoms, or something. Or maybe you've got an enzyme that can catalyse radioactive decay! Have you got a Gieger counter handy?
Interest is the right response here though. It's a shame about their entire life's work, but the scientist is the man who says "Oh bugger, I take that all back, still, at least we know more about doubly-labelled water as a measurement technique now". Well done them!
I thought it was a bit odd that spending six hours every day walking around didn't increase energy expenditure.
But in fact, science has always proceeded by the invention of better ways of measuring things. So even if the central argument of Burn turns out to actually be true, sorting out DLW metabolism and turning it into a useful tool would probably be the greater contribution.
If they offer to fly me to a lab, I'd do it in a heartbeat :) I love sitting around indoors.
I thought you might be up for that. I can't imagine they'll be able to get you away from your laptop but it should be easy enough to measure how much power it draws when you're sitting in the sealed room calorimeter. I do hope the wifi is good!
Hypothetically speaking
Hooboy! I owe you a quick apology (and Substack owes me one).
The comment I originally wrote here was intended for Astral Codex Ten's mock-post of the presidential debate. I read that one in my email and pressed the button to reply to it but for some strange reason when it sent me to the app the comment wound up on this post. You got the comment equivalent of a pocket dial.
Fun pocket dial tho. Sorry about that.
If you go read Astral Codex Ten's post you'll get the context of the comment. It will make a lot more sense.
I've noticed Substack weirdly re-ordering comments as I'm reading them. No apology necessary hah.
That seems to happen when new comments come in while you read.