Ben Krasnow Measures Human Calorie Consumption By Collecting The “Output”

It’s a bit icky reading between the lines on this one… but it’s a fascinating experiment! In his latest Applied Science video, [Ben Krasnow] tries to measure how efficient the human body is at getting energy from food by accurately measuring what he put in and what comes out of his body.

The jumping off point for this experiment is the calorie count on the back of food packaging. [Ben] touches on “bomb calorimetry” — the process of burning foodstuff in an oxygen-rich environment and measuring the heat given off to establish how much energy was present in the sample. But our bodies are flameless… can we really extract similar amounts of energy as these highly controlled combustion chambers? His solution is to measure his body’s intake by eating nothing but Soylent for a week, then subjects his body’s waste to the bomb calorimetry treatment to calculate how much energy was not absorbed during digestion. (He burned his poop for science, and made fun of some YouTubers at the same time.)

The test apparatus is a cool build — a chunk of pipe with an acrylic/glass laminated window that has a bicycle tire value for pressurization, a pressure gauge, and electrodes to spark the combustion using nichrome wire and cotton string. It’s shown above, burning a Goldfish® cracker but it’s not actually measuring the energy output as this is just a test run. The actual measurements call for the combustion chamber to be submerged in an insulated water bath so that the temperature change can be measured.

Now to the dirty bits. [Ben] collected fecal matter and freeze-dried it to ready it for the calorimeter. His preparation for the experiment included eating nothing but Soylent (a powdered foodstuff) to achieve an input baseline. The problem is that he measures the fecal matter to have about 75% of the calories per gram compared to the Soylent. Thinking on it, that’s not surprising as we know that dung must have a high caloric level — it burns and has been used throughout history as a source of warmth among other things. But the numbers don’t lead to an obvious conclusion and [Ben] doesn’t have the answer on why the measurements came out this way. In the YouTube comments [Bitluni] asks the question that was on our minds: how do you correlate the volume of the input and output? Is comparing 1g of Soylent to 1g of fecal matter a correct equivalency? Let us know what you think the comments below.

The science of poop is one of those 8th-grade giggle topics, but still totally fascinating. Two other examples that poop to mind are our recent sewage maceration infrastructure article and the science of teaching robot vacuums to detect pet waste.

37 thoughts on “Ben Krasnow Measures Human Calorie Consumption By Collecting The “Output”

  1. The reason is, I suspect, that his calorimetry is burning the dietary fiber and counting that as well, while the caloric value of the soylent doesn’t include it. There are many things we can’t digest, that burn and produce energy

      1. sure, but leaving aside the matter of equivalencies, if 75% of the caloric content (by burning) in soylent is indigestible cellulose, then the efficiency of the digestive tract in absorbing the nutritional content would be 100% as measured. What you’d have to do is determine the burny-energy of non-digestible materials in the soylent, and THEN compare. An approach might be to just measure the result of burning the weight of paper equivalent to the dietary fiber listed, but to be rigorous you’d have to take a look at the full ingredient list.

    1. I’ve always thought that the testing calories by burning and applying it to human physiology was a pretty poor comparison, we aren’t internal combustion engines. Also what about the waste products humans produce? How many unusable calories are in, for instance, bilirubin?

        1. What I’ve always wondered, since calories are measured by the amount of enrgy it takes to burn them, then why is water 0 calories? in order to “burn” it, you still have to insert a lot of energy

  2. He is using the wrong units, human energy should be measured in $$$ so we can compare transport costs for bicycling compared to driving.

    I think that humans are extremely inefficient because it is cheaper to put 3 gallons of gas in a 4000 lb car than it is to eat 6 servings of pasta for a century ride.

    1. I think cars still lose. People are super efficient when you *really* look at it. To go 100km a car needs 10L of gas, that’s roughly 90,000 capital c Calories. A bicyclist needs about 3,500 to go the same distance. 3,500 Cal is a pizza and arguably cheaper than 10L of gas.

      1. Lose how? What is the calories a bicyclist needs to go 80mph whilst fighting wind drag and friction?
        How much body heat would it create to get your metabolism that high?
        Efficiency would drop as well, maybe not at ineffcient as a car, but it is fair to factor this in.

        1. I think the human beast is very efficient. Carbohydrate and oxygen in, CO2 and H2O out. Once glucose is in the blood, the glucose-ADP-ATP machines converts ADP to ATP at the rate of about 65kg per day. And muscles, which are ATP powered molecular machines walking on bundles of microtubules, are over 95% efficient. People are comfortable in about 70 def F. Think what it takes to keep 14 gallons of water 30 deg warmer all the time as a baseline. It is all incredibly complicated and incredibly efficient. I can imagine the whole thing redesigned and optimized some day. https://www.youtube.com/watch?v=S5uFaqpEPMI

      2. 10L/100km is beyond inefficient for a petrol engine, at least at reasonable speeds.

        A more reasonable estimate would be 5L/100km, still less efficient than the bicycle, but also still at approximately double speed compared to the bicycle.
        Stuff 4 extra people in the car and we’re down to approximately double energy usage per person of the bicycle, as each extra person in the car doesn’t proportionately increase energy consumption.

        The economic aspect will depend on your location as pizza cost is mostly tied to local wage and petrol cost is mostly tied to raw oil price and local taxes.
        At my current local prices I would get approximately 13L of petrol for the cost of a pizza. The bicycle engine also needs some type of liquid, and if you would buy 0.5L of liquid from the pizza place, you could get a bit more than 2L of petrol for the price.

  3. He is also calculating the effects of the dead cells and other bodily waste. There is no way to explicitly exclude things in his data. If you refrain from eating, the body doesn’t discard as much dead cell material and switches into energy hoarding mode.

  4. No, of course, with all due respect to Ben Krasnow, the comparison doesn’t make much sense.

    Suppose that, instead of Soylent, you eat a mixture of a fully digestible carbohydrate with a (calorimeter measured) energy density A, and a completely non digestible one with a calorie density of B, such that B>A.

    In that case, the digestible part will be excreted in other ways as CO2 and water, while the fecal matter will contain only the non digestible carbohydrate (and some other constituents, which we can ignore if B/A is large enough). That way, it will have a larger caloric density than the food you ate, and by this reasoning point to a negative digestion efficiency

    1. Normal glucose content of urine is zero.

      Anything higher is pathologic, the term diabetes mellitus being Latin or Greek based (can’t remember which) for “honey.” Once the glucose content of the blood is high enough the kidney is not able to resorb it all as it filters blood and it “spills” into the urine.

  5. I’m glad to see this because I always suspected that the “calories in, calories out” theory was wrong. If I eat 1000 caloriea and go for a run and burn off 800 cal, I susoect that I would be in a deficit because crap is also a “calories out”. If there was no energy left in crap, then it wouldn’t very useful to compost right?

    Also, I suspecy that depending on one’s metabolism, some people shit more caloriea out (call it inneficient digestion I guess) and that might be why some people are leaner than others

    1. If you don’t eat, you typically loose about 200g daily just for breathing (oxygen from air + carbon from your body => CO2). “calories in, calories out” theory is not wrong, but blindly using it just leads to yoyo effect. I’ve tried, I’ve made neat calculator which essentially treated my body as a capacitor, food as charge and gradually linearly decreased my weight. It worked better than any diet before in my life, up until my body thought “oops, food is scarce, better conserve” and I was somehow not losing weight anymore despite 1/10 of normal input and just couldn’t continue, but not because of lack of motivation. I knew what REAL hunger feels like, when your stomach doesn’t send you any normal hunger signals, but every cell of your body craves for energy. In such state, when you see some food, you don’t think rationally anymore. Now I’m back exactly where I started.

  6. One thing to consider is that the body adds a lot or other waste to the poop so it isn’t just digested food. It will include protein, cells, and bacteria from the GI tract, wasted blood cells (e.g. dead blood cells stuff from the liver), bilirubin, maybe tape worms too. This is an interesting experiment though, they need to wrap their hands around the mass balance better and then I expect they will get more consistent results.

  7. One would need to account for the amount of material going into the body compared to what comes out.
    Measuring 0.5 grams of powder to 0.5 grams of excrement isn’t a correct comparison.

    Water we can largely ignore since it has no major nutritional value in regards to energy. (He even freeze dried the excrement to get rid of the impact of water.)

    But the amount of powered going in compared to dried excrement coming out the other end needs to be accounted for. Since a healthy human is expected to breath out more carbon than one breaths in. (And breaking down carbohydrates results in a lot of carbon dioxide and water that leaves the body in other ways than the excrement does. And if one really wants to increase the accuracy, then even the urine needs to be accounted for.)

    In the end, the human digestive track is likely more efficient than 80%, but how much more is though a better question.

  8. The error bars on calculated efficiency will be large, as he can’t really account for the bile salts, pancreatic enzymes and mucosal lining lost every day in the gastrointestinal tract with normal functioning, and he cannot account for micro-organisms metabolising non absorbable carbohydrates in the large bowel, i.e. producing methane from sorbitol or fibre, which will reduce the calorific content of what he is burning and also leave a biological residue.

    1. I think this comment is approaching the right answer, though mine will also not be complete. The bacteria in our guts don’t get digested, but I’ve read that most of our poop is bacteria. Add in all the processes external to the intestine which also use energy and expell waste, and I think we’d need a better baseline.

  9. The human body is now oven. There a tons of examples of undigestibel things like wood that burn and deliver energy.

    It is no trivial task to measure the calorie intake of the human body.

  10. He’s measuring the calorific value per gram of the turd, but not the difference in mass between the amount of (dry) soylent consumed versus the amount of (dried) poo excreted.

    As stated above, much of the food carbon and hydrogen (->H20) is exhaled, some may be stored, some farted out, pissed out, etc etc.

  11. > how do you correlate the volume of the input and output?

    Do the measurement of input and output volume over an extended period. If you on average eat three cups of soylent per day and excrete on average one cup of fecal matter per day, then that’s what you need to correlate.

  12. “tries to measure how efficient the human body is at getting energy from food”

    You just made a gargantuan leap from the idea of the experiment, and the authors own words. Fix that text to not say “getting energy from food” as that is not even close to what the experiment is measuring.

    Energy from food would involve him living his life in a tightly controlled environment measuring the energy output of his body vs the intake of food. Even that we know to be an imperfect measure, as too little food would mean using energy stores, too much involves storage. We also know about things like uncoupling proteins which can drastically impact the efficiency of the mitochondria, so each human body, depending on the presence or lack of uncoupling proteins will have different efficiencies, and those differences can be drastic. So to say “measuring how efficient the human body is at getting energy from food is at worst ambiguous, at worst, an intentional misdirection and oversimplification of a complex system. Change it to remove the ambiguity. I was quite negative towards the author until I realized these were your words not the author’s. At worst, your interpeted line should read “tries to measure how efficient the human body is at absorbing macronutrients from food”. Huge difference in meaning.

  13. The procedure simply does not allow for any conclusion to be drawn.

    In order to get a meaningful result you need to compare the dry mass that enters the body *in its entirety* to the dry mass that leaves the body *in its entirety* as opposed to a gram versus a gram.

    The Energy that the body has used will simply be missing in the “output”. But the energy density in a sample of output tells absolutely nothing about the body’s efficiency.

    Depending on the diet, one dry gram of output could even contain *more* calories than one dry gram of input.

    Bottom line: Nice experiment but no sensible conclusion to be drawn, unfortunately.

Leave a Reply to chrowensCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.