You read the packages on foodstuffs, or maybe in books, and they give breakdowns of what part or % of the food is what. How? How do they tell that a chicken breast or a bag of chips is x% protein, x% fat, etc? Do they dehydrate it and crumble it up, incinerate it and look at the gas spectrum, or ???
Just curious as to how they come up with those labels on packages and how accurate they are. Who does this (FDA?)?
Thanks-
-Tomcat
Cecil Himself has answered the part about determining calorie content
Apart from muttering something about “the oxidation of sugars by an alkaline solution of trivalent bismuth in the presence of potassium-sodium-tartrate” Cecil doesn’t provide much guidance on how one would go about determining the amount of protein, fat, and carbohydrate in food to begin with, but perhaps some Friendly Food Chemist will come along and help you out.
Although the producers are the ones who are required to put the labels on their foods, the FDA tests to make sure that the labels are accurate using standard scientific and laboratory techniques. Each individual item has specific testing and labeling requirements.
Here’s a link to the Food Analysis section of the FDA’s Training Manual, which contains stuff like this:
My first job out of college was analyzing food to generate those little labels (I even did the K-Na-tartrate assay!)
The components are measured by various analytical chemistry techniques depending on what you are measuring. In our lab (an independent contract lab), fat soluble vitamins were done by HPLC, water solubles by a microbiological method, minerals by ICP, amino acids by an amino acid analyzer (basically a glorified HPLC), fats by GC, protein by a variety of methods, and so on. After that, calories are simply a calculation based on protein, carb, and fat. You do need to be certified by the FDA to do this kind of work, and for that you need to use only certain approved methods. That is why it was a pain in the ass to work there. The microbiological assays were vintage 1940’s technology but to change them would mean a horrific revalidation of your new methods (VERY costly).
Kind of an interesting job, but glad I’m not there anymore.
Don’t get me started on how well enforced some FDA requirements are either. Suffice to say, you know those supplements you get at the health food store? Don’t waste your $$$ (in most cases at least).
oops, forgot part of the OP.
To get a homogenous sample, everything is mixed in what amounts to a big blender. Our lab had an entire (very well-supplied) kitchen to prepare things too. TV dinners, pizzas (and anything else that was supposed to be eaten cooked) was cooked, then blended. The idea was to analyze it as it would be eaten, as cooking can really change the nutritional value of food.
We also did environmental analysis too (pesticides, mercury and whatnot). You haven’t lived til you’ve gone to work hungover and watched 6 big carp get homogenized, then had to analyze the resulting carp puree. :eek:
Wait!
I want the specifics!!!
OK, so you blended them. What THEN? What exactly is going on? Help a clueless person out! Prepare the speech for a 6 year-old. What did you do, what happened? I want details
!!!
oh, Thank you mister food blender-cum-analyzer!
-tcat
Just for the record, I am not, nor have I ever been, a cum analyzer.
I think they did that on 3rd floor.
First, thanks to Tomcat for starting this thread. I’ve been wondering about this for years. Thanks also to jk1245 for posting all the information. I put myself through college working in a US Geological Survey water quality lab, so I can relate to a lot of what you’re saying!
The only thing I’m unsure of at this point is if there are any adjustments made for the actual utilization efficiency of the nutrients in the human body? A couple of cases, in the chronological order that they occurred to me:
Parts of certain foods will pass completely through the body undigested. I always knew that if you just put the stuff in the bomb calorimeter, you’d get an artificially high result. OK, you don’t use a calorimeter, you analyze for the component nutrients and calculate from there, and you use a sample digestion method that simulates human digestion (see MEBuckner’s post.) Great. But it’s the mechanical step that I’m wondering about. Take strawberries, for example. When you eat them, I’m pretty sure the little seeds pass through undigested. But if you homogenize the sample in a blender, the oils in the seeds become available to the digestion step, and will show up in your fats analysis (which, I’m betting, is a lot like the Oil and Grease analysis we used to run on water samples.) Strawberries aren’t a great example, but the only other one that comes to mind right now is (WARNING! Possible TMI ahead!) - ahem - corn.
A more recent example would be foods using Olestra. That will show up in the fats analysis, but of course, it’s not digestible by humans. Now that I know the values are calculated, I guess you would determine the percentage of total fats contributed by the Olestra, and subtract that out to get to the final result - am I on the right track here?
There was a report about how they determined the calories in a Thanksgiving day dinner on NPR yesterday.
Excellent questions Dono.
You are right on Olestra. IIRC, it is a sugar attached to the normal vegetable fats you might see (It’s been a while, so if I’m wrong please correct). That is easy to ID on a chromatograph and adjust accordingly.
The other part (seeds & whatnot) is a little more interesting, and frankly, I hadn’t considered it before you brought it up.In my experience, no adjustment is made for rates of absorption or digestion beyond the obvious (removing an orange peel before analysis for example). We, in fact, did analyze strawberries, and they were ground up seeds and all (smelled great by the way. MUCH better than the carp).
This is most likely due to the fact that it is impossible to say how efficient each person’s digestion is. The labels merely tell you what is in the normally edible portion and how much. Digestion/absorption rates vary from person to person, so can’t be predicted.
Hope this helps!
Thanks for the link, phartizan, and thanks for the follow-up, jk1245!
Oh, right, GC. Uhh, I knew that. They wuz just gettin’ them new-fangled contraptions back in my day!
Kind of like the EPA MPG figures - useful for comparison purposes, but YMMV!
Cool - the frontiers of my ignorance have been pushed back a bit today!
I used to work for a company called Covance that does this testing on a contract basis with food manufacturers.
Covance is an FDA Certified lab, which means they take our word for test results, though they can audit any tests they want, and the FDA audited our lab procedures or particular studies several times every year.
The manufacturers sent us samples of their production lots. We tested not only for nutritional content, but pesticides/contaminents or their residues. We tested for whatever they wanted to us to test for, based on what the product was and how it would be used.
We sometimes got unpleasant (results exceeding FDA limits on pesticide, sometimes too high or low a vitamin content for the product for which it was intended) results, and sometimes the clients asked us to retest things. As far as I know, our retests never came up with a reversal of results.
However, it was up to the client what to do with the information. If lot A of whatever had 50% more of some pesticide than allowed, and lot B had 50% less, they might mix the two for a result within parameters.
If you want particulars on test details, go to http://www.fda.gov/ and search for “method” and “vitamin” or “mineral” or something like that. Testing methods, or the parameters within which proprietary methods must operate, are pretty widely published.
One more thing about independent testing like Covance does.
For all we know, the manufacturer might take a load of poison laden wheat and bake it into bread. Perhaps their QA person goes nuts that day. We provide a clear audit trail so if there is a problem, culpability can be traced.
And I know of at least one case where we got a manufacturer off the hook on a product liability case. Someone on a special medical treatment died due to a salt imbalance, and the spouse sued a food maker claiming the salt content was too high. Turned out the salt in the “fatal” product was iodized, and the manufacturer could document that the product was made with salt not containing iodine. So it was product tampering, and the spouse was later charged with murder. Never heard how it ended.
Hey ** Yojimboguy**. I used to work at Covance also!
I see from your profile, you are from from Wisconsin too. When did you work there? I was from 94-97.
from 99 til a couple of months ago.