I burn 4,600kcal/day being sedentary

In which I break the first law of thermodynamics (cops HATE him!)

As alluded to in a recent post, I purchased the $1,000 doubly-labeled water test from Calorify.com. Doubly-labeled water is the “gold standard” way of testing how many calories someone actually burns during the day. It’s not just a guesstimate, it claims to measure the actual, real, physical, true “CO” part of “CICO.”

I say “claims” because…

Lol, ok then. When I had finished counting calories for a week and posted about it, I had announced my expectation that my TDEE would be either exactly the calories I’d consumed (~2,850kcal/day), which I thought was a bit low, or around 3,300kcal/day, an estimate based on my previously measured Resting Metabolic Rate (RMR).

So imagine my surprise when the number came back… over 4,600kcal/day. Literally maxing out their scale for burned calories.

You can trust the computer. The computer is your friend.

How is that possible? I am a sedentary, introverted, night owl programmer. My day job is sitting in front of a computer typing, and my hobby is also sitting in front of a computer (and typing this nonsense).

I hate cardio. I hate the outdoors. I exercise about 60 seconds per day with my x3 bands, when I don’t have a rest day. I take a lot of rest days.

Do you even uncouple, bro?

The Lamming Lab mouse studies about protein/BCAA/isoleucine restriction have shown that low enough isoleucine intake can activate the “uncoupling protein” UCP-1, also called thermogenin. This somehow allows these mice to eat more calories than their normally-fed counterparts, yet lose weight by burning off even more.

In gym bro parlance, these mice “ate more, moved more” and also burned more energy at rest. There’s an idea that UCP-1 is used to “flare off” excess energy by increasing thermogenesis in brown adipose tissue, which is apparently a special form of body fat. Maybe similar to oil wells flaring off excess gas.

Maybe this is happening in me? In a sense, this is a confirmation of what I’d suspected ex150 does via such aggressive protein restriction: Lamming Lab used 7% protein/BCAA/isoleucine diets to trigger uncoupling in the mice. ex150 is 6-7% protein, depending on how much fat I eat.

So I might’ve just gotten “lucky” with the amount of protein restriction, ending up at the exact same level that the science folks found.

As a reminder, here are my food logs for the week I tracked.

I ate 2,859kcal/day on average, for a deficit of 1,786kcal/day. Yet I didn’t lose nearly any weight, which is why Calorify believes that I ate 4,614kcal/day. (I sent them some more weight data to clarify, that’s why the data below data goes from May all the way through June.)

So, uh, how do you explain that?

I don’t. I have no clue.

Yea, that was pretty much my prediction - 2,850-3,300kcal.

If I had an unexpected, insanely high TDEE and lost a lot of weight over that time, that would make sense. If I had an unexpected, insanely high TDEE and also ate an insanely high amount of food, that would be interesting due to the UCP-1 reason Lamming Lab discovered in mice, but it would still make sense.

But having an unexpected, insanely high TDEE and eating at nearly a 1,800kcal/day deficit and being weight stable for 2 months now… that’s just… supposed to be impossible?

Other Calorify metrics

Besides just the caloric burn, Calorify gives you some more metrics and comparisons.

This one’s curious. Apparently I have the “activity level” of an athlete about to hit the “overreaching” level. As you might know, I’m an introverted programmer who works from home. My idea of activity is getting dressed for a Zoom meeting with my team. Or getting up to make another cup of coffee. Ugh, how I HATE getting up.

I work out most days doing my x3 resistance bands, but that’s resistance exercise and often less than 60 seconds per day. Hardly burns 2,677kcal/day.

My energy availability is at the upper limit of the range. Which is kind of what it feels like. I feel extremely energized most times of most days. I’ve previously written how eating cream gives me a sort of “energy burst” that pretty much forces me to go outside and go for a walk sometimes, or makes me write a blog post or code something.

My training sustainability is kinda nonsense I think, since my high calories out are not due to training. I think they just divide the theoretical maximum you can burn by what you actually burn. The upper limit is commonly assumed to be 2.5x your BMR, and that’s exactly what that number (4,920kcal) is.

This is very interesting, showing where the energy burn is actually coming from. I think they do this just by calculation, as I gave them an estimate of how much exercise I was doing and I wrote “60 seconds per day” so they just put 100kcal. That’s probably generous, lol. I don’t think 60 seconds of resistance exercise burns 100kcal.

That means all the energy burn is coming from NEAT, or Non-Exercise Activity/Thermogenesis. This is body temperature, fidgeting, brain activity, walking around during the day… all those things we do that aren’t proactive “exercise” yet are above the bare “maintenance/coma” level of the Base Metabolic Rate.

In other words, my body is a metabolic furnace: I somehow burn the TDEE of an adult woman, 2,400kcal/day, just as NEAT doing essentially nothing.

This is cool, they estimated my body composition extremely precisely compared to the DEXA scans. They do this using a technique called “deuterium dilution.” My last DEXA had me at 150lbs lean mass and 68lbs fat mass, and just over 30% body fat. So they’re within <1% body fat and within 3-4lbs on fat mass/fat free mass.

This shows that I’m a pretty extreme outlier in terms of Carolies Out for my lean body mass, but not the craziest they’ve ever measured. There are people burning 6kkcal/day with my LBM. But those might be endurance athletes training 5h a day. Not indoor programmers sitting around and pecking at a keyboard 16h a day.

Should you get the Calorify DLW test?

Probably not. While it’s fun for the nerd in me, the price is pretty steep. Their introductory offer is $800, and the “PRO” one I did was $1,000. I found a $100 off coupon somewhere but still.

If you’re wondering, the PRO just adds the little pie chart w/ NEAT vs. exercise, and it gives you a 30 minute consultation with their team that I’m looking forward to.

The results open up more questions than they answer, I think. In a sense it’s nice to verify that they basically verify my protein-restricted approach, and show Lamming Lab’s mouse studies seem to translate to humans.

On the other hand, there seems a clear disconnect. According to even my understanding of CICO as a “trivial accounting entity” this result should still be impossible. Their team and all the Big Name Scientists like Pontzer and Speakman seem to think the same. I talked to both of Pontzer and Speakman on Twitter about my result, and neither had an answer.

Does the result change anything in what I’m doing? No. If you want to know if protein restriction works for you, just try it for 30 days. No need to spend $1,000 on a test that will only look broken/impossible lol.

Cheaper options

If you’re interested in your metabolic rate, e.g. because you’ve done a lot of longer term caloric restriction or fasting, have low body temperatures or energy levels, you can always get a much cheaper test called Resting Metabolic Rate via Indirect CO2 Calorimetry. I’ve written about that before.

RMR testing is usually <$100. It will only give you your “resting” metabolic rate, not whatever comes from exercise or maybe NEAT. But it would definitely be good enough to show you if your metabolism is severely down-regulated.

A few readers have done it and sent me their results, and yes, the RMR showed severely depressed metabolic rates after years of restricting calories to lose weight. One’s RMR was just over 1,000kcal/day, or about half of what it should be for his lean mass.

On the other hand, if you have $1,000 burning a hole in your pocket, you might as well spend it on a small bottle of deuterium water with Calorify.

It’ll be interesting to see if these DLW tests come down in price in the future. If you read Pontzer’s book and see the history of the test, this used to cost hundreds of thousands per person. It’s only cheap enough for even nerds like me to purchase as of a few years ago.

Maybe it’ll be down in the $100-250 range in 5 years, cheap enough for many fitness/nutrition nerds to purchase annually? In which case we should also see a rapidly growing database of numbers, and maybe they will find edge cases like mine and address them.

Conclusion

Is physics not real? Did I violate the fIrSt LaW of tHerMoDynaMicS? Is CICO a lie?

Doubly-labeled water is allegedly the Gold Standard (tm) test for energy expenditure. It’s what Herman Pontzer and John Speakman have built careers on. It’s apparently so precise it’s used to calibrate all the other tests, basically gospel.

So what gives?

Maybe it’s somehow not calibrated for extreme (90%+ fat) ketogenic diets? Or for UCP-1 diets that are super low in protein? Maybe thermogenesis via brown adipose tissue isn’t using the same amount of oxygen/hydrogen, and their assumed ratio is wrong?

Who knows. Science is hard, people.

Update: After my consultation with Calorify, they offered me a free retest. They say they’ve only ever had 1 bad result in thousands of samples, but that it is a possibility.