Food extraction... a dilemma


Hi all… so we’re evaluating what extraction techniques we want to use on incoming food for the Real Food Campaign Lab.

We are trying to create a common extraction to run 3 different tests - Bradford protein, polyphenols, and antioxidents (FRAP). Details on each method below for those so inclined.

Polyphenols (using liquid part (aka supernatant) after centrifugation)
There are two possible methods - one using Folin-Colteau reagent, another using Fast BB reagent. Both are described in the paper below.
Method -

Antioxidents (using liquid part (aka supernatant) after centrifugation)
Ordering page -
Full directions -

Bradford Proteins (using solid part (aka pellet) after centrifugation)
Ordeing page -
Full directions -

There are many options - extractants - ethanol, methanol, ultra-pure water, how to macerate the sample - blender, food processor, freeze dryer, … every option produces different results depending on the nutrient extracted, and there is no perfect cost/ease/efficacy answer.

I’m struggling with a few questions, but they boil down to the following:

As the RFC… what is the nutrition we’re attempting to measure? The total nutrition in the sample? The nutrition in the sample that’s able to be absorbed by a human? Something else which, so long as it’s comparable from one sample to the next? I’ve been struggling to create a consistent, easy, fast way to fully macerate (smash all the cells) in order to measure the total nutrition as accurately as possible… but I’m wondering if that’s even the right number to report if we could get it. If there’s 1000 mg/g beta carotene, but the normal process of chewing and digesting raw food only absorbs 200 mg/g, should we be reporting 1000? Then… if we’re off by a factor of 5 all the time, does that matter as long as it’s always a factor of 5… or is there differential absorbance of beta carotene depending on it’s location in the cell… like perhaps different samples will contain beta carotene in different locations, which may be accessible to me in the lab using super maceration, but inaccessible to you in your gut. This has left my quite confused, and wondering if I’m going through all this trouble just to get a more accurate but less useful result.

I know there are many more questions than this - different compounds will extract in different ways, do different extraction methods differentially oxidize or otherwise make unmeasurable compounds impacted by heat/oxygen/time etc… but the above question feels much more central to what we do next.

So - in the end… I think we can macerate the crap out of a wide range of samples quickly and efficiently, better than juicers or blenders or whatever and faster… but I’m not sure if that’s even the right thing to be doing.

I’d love any thoughts on this @DanT @burton Jill (I’ll email you, you need to get on the forum!) and anyone else. Thanks!



Human capacity for absorption sounds right to me, but even that number varies by the eater’s recent diet, no? The human body might absorb 200mg/g beta carotene under normal circumstances. But what is a “normal circumstance” for a human eater? Wouldn’t your absorption capacity be higher or lower depending on the mineral content of the other foods you’ve eaten recently? (With no background in human nutrition, I’m speaking here based on what I know about soil nutrition.)

This may be an unnecessary complication for the measurement; it may be sufficient to assume a balanced diet, if such a diet has a relatively uniform absorption effect across all nutrients.

What standard does the USDA use for nutrition labels?

BUT WAIT! I just read more about the problem you’re trying to solve. RFC is about food transparency,
but your method for achieving it is a hardware device that will be used for food as well as non-food purposes. So maybe human absorption should not enter into the mix unless you are certain of the application.


Herbalists say, “you are what you absorb” As you mention, each person absorbs at a different rate based on how they chew, gut flora, what they eat in a combined meal (think glycemic index), etc. You can’t bother with any of that–those issues will have to be addressed later with education.

On the other hand, the solvents you choose definitely matter. Different solvents will extract different constituents of the plant. I’m a traditional herbalist, so we use drying (removes formic acid), water (teas, infusions), and alcohol. We also use annual cycles. When a plant part is picked can make all the difference to what is in the plant. It’s a dynamic system.

Given that you are trying to compare nutrients against the USDA standard, if you do decide to use a method other than what is used by the USDA, you will have to correlate the results against the standard in order to have that conversation. There may be a very good reason to deviate from the standard, but you will have to make a strong case for doing so.

The simplest solution is to use the USDA standard for now, and over time develop new methods as appropriate, keeping a record of your reasoning and methods.

(First post! Glad to be part of the conversation!)


Thank you all for your thoughts… I’m assumed there was no perfect answer, but I’m glad we had a chance to lay out the pros and cons.

I think the consensus sounds like while sometime in the future some more accurate estimate of human absorbable nutrients would be best (or some range… or whatever) for now we need to do what’s practical.

I actually don’t know what the USDA used during it’s two sampling points in the early and mid 2000s, but it was likely methanol and water and some form of shaking just like everyone else. The key for us is to use a method which is relatively simple (low capital cost, so no 44kz lab grade sonicators) but still effective. I’ll post back here once we’ve got something down to get additional feedback.

Thanks for processing this and for your input! PS pretty soon we’ll have pictures up of the lab, with a more concrete initial listing of our processes and methods which folks can tear apart / improve / iterate on etc. via this working group :slight_smile: