Food Calorimetry

I've always wondered how scientists determine how many calories are in what foods, and the derivation of the source of calories.  From an article on Slate, laboratories freeze samples in liquid nitrogen then blend it into a fine monochromatic powder from which to test.  Removal of the nitrogen via the Kjeldahl process allows scientists to calculate the amount of protein.  Fats are determined through a hexane extraction, and carbohydates are what remains after fats and proteins are isolated.

Of course, one could always burn the sample and measure all the heat released as calories (in the chemical sense) and convert that to kilocalories, which is what we all call "calories."  All that is given off during the burn process, however, doesn't accurately describe the amounts that is taken up by the body.

While the FDA requires that companies publish nutritional facts on foods, it doesn't specify how they are to obtain those figures.  Companies may even guesstimate values based on the USDA's published tome of nutritional data, which is available online.

Wikipedia also shows an energy density table which standardizes food energy into values actually absorbed by the body.

Food component	Energy Density
	kJ/g	kcal/g
Fat	37	9
Ethanol (alcohol)	29	7
Proteins	17	4
Carbohydrates	17	4
Organic acids	13	3
Polyols (sugar alcohols, sweeteners)	10	2.4
Salatrims (reduced energy fat) [1][2][3][4]	25	6
Fibre	8	2
Erythritol	0	0

Pretty neat, huh?

devyn

USDA Nutrient Data Laboratory