20 January 2010

n-3:n-6 ratios of cooking fats

damn, this info is kind of hard to come by. I had found a chart somewhere online that included pasture butter, lard and others, but for the life of me I can't find it. I did see this page today, which is from Cordain's site. they link back to the USDA site for where they got their data, which is a horribly organized site with a terrible search engine. so i'm grateful for Cordain's summary data, which i've re-represented here for convenience.

if anyone has similar data for butter, lard, bacon fat, red palm oil, or the like, i'd be happy to update the table to include them. until then:

I went to the USDA site and, based on the Cordain methodolgy, calc'ed values for butter, lard, and "Meat drippings (lard, beef tallow, mutton tallow)". These values aren't exactly in line with what I'd seen before, which makes the USDA values a bit suspect. Then again, their butter values are probably for corn-fed cows, not pastured ones, etc etc.

n-3:n-6 ratio
flaxseed 1:0.2
canola 1:2.2
mustard 1:2.6
Meat drippings (lard, beef tallow, mutton tallow) 1:4.8
walnut 1:5.1
butter 1:6.9
soybean 1:7.5
wheat germ 1:7.9
lard 1:10.2
olive 1:11.7
avocado 1:13.1
rice bran 1:20.9
oat 1:21.8
palm 1:45.5
corn 1:46.1
sesame 1:137.7
cottonseed 1:258
grapeseed 1:696

in the household here, we shoot for cooking fats no worse than 1:4 n-3:n-6 ratio. i've used a fairly obvious green/yellow/red color scheme to indicate what i think of the desirability of the ratios, though that's not to say that we're using canola oil just because it's "in the green." And we do use olive oil, but pretty much for salads exclusively and not for cooking.

Mark's Daily Apple did a piece today on oils, though it didn't include all the info i would have liked. Still, it's a decent reference.

One final note: hazelnut, peanut, safflower, sunflower, almond and coconut oil all rated zero n-3 fatty acids, so i left them off the chart.


JustJoeP said...

1 to 137? Leapin Lizards! I have to re-think some of my wok cooking flavorings.

Thanks for posting this handy reference Zim!

pyker said...

nutritiondata uses the same data, I think, but with a slightly better interface.

You have to take in absolute amounts, not just ratio. E.g. canola oil has triple the o-6 of butter, even though the canola ratio looks good. Rather than just focusing on ratios of individual ingredients, one tactic is to limit total o-6 intake and add a good source of o-3 (ideally food, but fish oil works, too).

zim said...

yes, it does look like nutritiondata is using the same data. much better presentation, too, though the search mechanism is still tedious.

good point on the amounts. i wish i could find that original table, and figure out from where they got their data.