26 April 2010

Plants & Plausibility

Are plants protective for cancer? The evidence has been mixed, to say the least. Evidence, that is, in the form of specific human interventions. And it's fairly ambiguous observationally. Recently yet another result from EPIC was published, looking at correlation between fruit and vegetable intake and cancer. Basically there was no difference. The best-case hazard ratio was only down to 0.97 (compared with a more encouraging HR of 0.72 for, um, eating lots of cheese ... oops!). Given uncertainty in figuring out what people actually ate over the period, and given uncertainty around "correcting" for variables (these corrections are built upon pyramids of observational studies, it's not an exact science, although it pretends to be), this is probably meaningless. The biggest worry for the pro-plant brigade should be that even with the flaws, we could expect a more dramatic difference in hazard ratio simply (or solely) from healthy user bias. So it's possible plants are causing cancer. Who knows. If you're choking down foliage by the bucket in the belief you're saving yourself from cancer, there's not much evidence for that. Eat them if you like them. Or maybe to foster a diverse ecology of gut flora, or to poop a lot.

In any case, lots of smart people are absolutely convinced that plants have magical health benefits. I think this stems from a problem that Pauling succumbed to -- the plausibility peril. In the lab, we can see specific interplay at the molecular level, and this builds up a promising avenue of research. The plausibility peril: once we come up with a plausible explanation for how something could or might work, the plausibility itself acts as evidence in support of the theory. Not even deliberately or intentionally, but nonetheless it becomes a bit of a trap. E.g. you sneak up on some cancer cells in a beaker, minding their own business, and hose them down with tocotrienols. They stop growing. Well, that's certainly promising. Tocotrienols are in plants, thus eating plants might slow or stop cancer growth. And freebasing vitamin C might cause immortality (sorry, Linus). But when you try it out in humans it actually has to work. And often it doesn't, no matter how plausible the story.

6 comments:

JustJoeP said...

My take on this, is that the "plants" that most of the "people" are eating, are crap, from the mass industrialized, growth hormone & antibiotic fed, factory based system - NOT the "I know my green grocer by first name" database, which is a wonderful little microcosm, but not repeated by +99% of the population en masse.

I subscribe to the micro-nutrient systematic theory. It's not the Tocotrienols themselves, but the combination of those and the fiber and cellulose they come with, released slowly, naturally, in the human body. What's most interesting are the studies on isolated populations (Asians, Pacific Islanders, Amazonian tribes, Japanese, Inuits) that show specific maladies they typically have (like stomach cancer in Japanese) and other maladies that were unheard of, before the introduction of a "western diet" (like diabetes in the Indian sub continent).

Looking at "single ingredient" case studies, in my layman's perspective, is waste of time, like studying one tree's bark in a whole forest and then making theories about the entire forest. Studies that rely heavily upon self-reporting surveys are also very dubious, and don't get me started on industry funded studies, which are nearly all inherently evil.

Ame said...

Actually, the indian subcontinent has had a long history with diabetes. Lucky for me, it's just gotten wider spread and earlier in the Indian life-span. That being said, the rapid rise of diabetes in India is the growth of a "western diet" and faster foods. People just aren't cooking any more.

An interesting study was done up on the current nut. values of veg - and how due to over farming the value is lower. I can't find it off hand - but I'll look again tomorrow and post back.

Ame said...

Articles on how fuits and veg have a declining nutient comp:
1) http://hortsci.ashspublications.org/cgi/content/abstract/44/1/15

2)http://www.grist.org/article/Less-tasty-and-not-as-good-for-you


and apparently - size does matter:
http://www.seattlepi.com/national/331421_bigfood13.html

this one has a agenda -- but i liked the graphs. :)

http://www.soilandhealth.org/06clipfile/0601.LEMag/LE%20Magazine,%20March%202001%20-%20Report%20Vegetables%20Without%20Vitamins.htm

pyker said...

Nice.
And another one, apropos of India & diabetes.

JustJoeP said...

"http://www.soilandhealth.org/06clipfile/0601.LEMag/LE%20Magazine,%20March%202001%20-%20Report%20Veg" got cut off... what's the full link?

Ame said...

JJP - I wasn't able to find your full link. :( But, I am completely fascinated by the info on the site. Which is taking me ages and ages to read all that info. So, thanks for the partial link. :) I will have to do a full up posting on soil soon.