How accurate are calorie counts in food?

AFAIK, calories in food are determined by using a bomb calorimeter. Basically, burning the food, and measuring the heat produced. But, people don’t “burn” food. If you put a chunk of wood into a calorimeter, you will get a whole lot of calories, but eating it is probably a net loss of energy. How is this difference in efficiency taken into account, or is it?

Well, some calorie tables use the Atwater convention (basically, you do bomb calorimetry again on the feces and subtract the value).

Calories are a gross value. The body processes fats, proteins, and carbohydrates and gets approximately 9, 4, and 4 calories of value per gram eaten. Some of that is used to digest the food and move the breakdown products around the body to where they are needed, but that’s part of the background basal metabolic rate (BMR), which is going to be different for each person. The net value is irrelevant to amount of energy intake.

For an analogy: the energy of a gallon of gasoline is fixed regardless of whether you burn it in city driving or highway driving and get a higher or lower MPG.

From what I’ve read, the actual end product food isn’t tested. [

](https://www.scientificamerican.com/article/how-do-food-manufacturers/)

So no one is burning a potato chip. It seems this would lead to only approximate measurements of the calories in food, since every potato or onion or steak doesn’t have the exact same makeup.

Make sure you pay attention to the “per serving” size which, in many cases, is about the size of what you would expect a POW to get.

“The FDA allows food companies wide latitude in the accuracy of the calories listed on package labels—20 percent in either direction. That means if a label says 200 calories per serving, it could be 240 calories or 160 calories or anything in between. What’s more, the FDA doesn’t do any systematic policing of labels to ensure that calorie counts meet even that lax degree of accuracy. The responsibility for label accuracy remains with the food companies, from national manufacturers to regional or local vendors. It basically works on the honor system.”

If you really cared about decimal point accuracy you’d start with the fact that 9, 4, and 4 are approximation. That 4 doesn’t even equal the other 4 exactly.

But approximations are just fine. Nobody needs to know that they’ve eaten 1022.6 calories of carbohydrates in a day rather than 1022.3 calories. Or even 1022 rather than 1000. Bodies don’t work that way, food is variable, BMR is individual, movement changes from day to day. Food can be dealt with in bulk.

Bad analogy. A better analogy would include the fact that some fuels (e.g. protein) actually cause the engine to run less efficiently, so more fuel will be burned to travel the same distance at the same speed.

LOL. I can’t count the number of times I’ve thought something like "Ok, I ate a quarter of that bag of chips, how many calories was that? Hm… “20 servings per container”?!

Protein does not cause the body to run less efficiently. That’s a nonsensical way of way at sda (aka dietary induced thermogenesis). The varying amounts of energy required to process various nutrients cannot be separated from the whole: the protein a body gets comes as part of the package with carbohydrates and fats. All three are required for health and so it is the average rate of thermogenesis that is all that matters.

The proportion of carbohydrates, fats, and proteins a body intakes varies from person to person and from day to day. No one ratio has been shown to be “better” for everybody than any other. The energy cost of eating and digesting food, and then processing and building up the pieces, is part of what the food is for. The value of that cost is not taken into account when calculating calorie counts because it is indeterminable and irrelevant.

If you work or exercise hard you will need more gross calories; if you want to lose weight you should try for fewer gross calories. By your logic, someone should eat a protein-heavy diet to lose weight. It doesn’t work that way in the real world; no one style of diet is most effective.

I’ve never found those serving sizes to be inappropriate. Remember, also, that when it comes to stuff like chips, they’re not assuming you make an entire meal of chips, but rather that it’s a side to your main dish (like a sandwich.) I’m not a skinny guy, but I find those serving sizes appropriate. The one that really gets me is people acting like a pint of ice cream should be a single serving. That should be at least four servings to me (and, for me, it’s more like 6-8).

I didn’t say that. I was working with your flawed analogy, trying to make it more accurate.

You don’t know what your’re talking about. There have been several RCT studies showing that diets higher in protein lead to increased fat loss.

Serious question - Do you have a cite for that? As far as I have seen, diet studies have a lot of systematic problems, mostly due to the fact that humans have a hard time changing their habits for a long period of time and tend to be bad at accurately reporting what they eat. Almost everything I’ve seen has said that while some diets (low fat, low carb, etc.) perform better than others in the short term, the best diet is the one that you stick to.

Also, excessively high protein diets are not good on your kidneys.

I quoted you saying that.

All diets will work on some people for some amount of time. What I said was that there wasn’t a best diet. There certainly are good diet plans that call for lower fats and sugars and correspondingly higher proportions of protein. But that has nothing to do with proteins “actually caus[ing] the engine to run less efficiently.”

So then, how do they determine the amount of those energy-containing nutrients? How do they know how much protein, etc are in a given food?

A few of the studies are cited in the article below: