Sunday, January 5, 2014

Biochemical Warfare

In response to the 30-day trial that Zooko and I recently recommended, one correspondent argued about the importance of plant foods. He made several points I'd like to address, but the one I want to talk about here is one I've heard many times before. The statement is to the effect that plants are full of a variety of healthful compounds, many of which we have surely not even yet discovered.

This idea has an assumption behind it that I strongly disagree with: that we evolved to eat a significant amount of plant matter, and therefore we are likely to have optimised our functioning on the biochemical compounds in those plants. I disagree for the following reasons:

  • Whether we evolved eating a lot of plants is contentious. At the very least, there have been times and places that we had little to none.
  • What we do know about plants is that their survival strategy is biochemical. They generate many chemicals (many of which we surely have not even yet discovered) with which to poison their would-be eaters!

My correspondent went on about his diet, listing biochemicals in the foods he eats, and lining them up with diseases those chemicals have been shown to have promise in fighting.

I agree that the biochemicals in plants often turn out to have medicinal properties that we can make use of. For those properties, the chemical typically has to be extracted to get a high enough concentration to have any effect. These compounds are much like drugs we make ourselves, in that they usually have unwanted side-effects. Commonly, they are double edged swords. For example, some plant chemicals are widely touted because they harm cancer cells. The bad news is that they also harm your healthy cells. So like chemotherapy, it is a matter of hoping the healthy cells survive better than the cancerous ones.

In a recent article by Michael Pollan, whose brilliant slogan we bastardised without even a nod (we hope he takes it in stride) [1], he describes how plants produce chemicals, not just as a matter of growing, but in a real-time response to predators. For example:

"One of the most productive areas of plant research in recent years has been plant signalling. Since the early nineteen-eighties, it has been known that when a plant’s leaves are infected or chewed by insects they emit volatile chemicals that signal other leaves to mount a defense. Sometimes this warning signal contains information about the identity of the insect, gleaned from the taste of its saliva. Depending on the plant and the attacker, the defense might involve altering the leaf’s flavor or texture, or producing toxins or other compounds that render the plant’s flesh less digestible to herbivores. When antelopes browse acacia trees, the leaves produce tannins that make them unappetizing and difficult to digest. When food is scarce and acacias are overbrowsed, it has been reported, the trees produce sufficient amounts of toxin to kill the animals."

To think that we co-evolved with plants in symbiosis, them providing us with countless, needed medicinal concoctions, while we selectively kill them and eat them, seems a bit naive. While it's true that some species rely on having their seeds carried, undigested, for better distribution, or their pollen spread, this is does not imply they get value from having their very bodies eaten. Even that fruit and nectar need only be helpful to some species; It didn't evolve dependent on humans per se.

Finally, I find it a little sad that my correspondent is seeking to avoid disease through constant miniscule doses of medicine that are likely to be accompanied by as much toxin. There is no evidence that such a strategy has any beneficial effect. On the other hand, a ketogenic diet has increasingly stronger types of evidence suggesting it will protect against those diseases [2]. As we speak, randomised clinical trials are underway to help clarify whether this is true. Such is not the case for vegetable eating.

[1]For those of you who didn't recognise it, Michael Pollan famously recommended: "Eat food. Not too much. Mostly plants."
[2]See, for example, Zooko's and my post, The medical-grade diet.


  1. Hi Amber~
    I assume you've read the "vegetables" posts on Hyperlipid? Great stuff.

    And thank you for all the wonderful posts! I am forever linking to your site.

    1. Hi, Sondra, thank you.

      I've read some of his fruits and vegetables posts, like this one, and Vegetables damage your DNA, latest study headline!, but there are newer ones I hadn't seen, like Fruits and Vegetables.

      I'm a big fan of Petro Dobromylskyj.

    2. Peter rocks!

      Oh and I also love Dr. Ede's talk at AHS 2012:

  2. Hi Amber,
    Another great post! My thinking about vegetables along these lines was first influenced by Jeffrey Steingarten's 1988 article "Salad, the Silent Killer". So true.

  3. Have you seen the Italian who promotes plant extracts to accompany a keto diet

    1. Thank you for posting it. It is a shame they didn't have the resources to use a control group. As it is, it is hard to conclude much about their added intervention. There are certainly extracts that have been shown to help certain conditions, like high blood sugar. I have no objection to plant-derived medicine (nor man-made, for that matter) if it is the alternative that maximises benefit to cost.

      If it is true, as they suggest, that adding the "carbohydrate-like" foods increases compliance, then that would be useful to know. It is not uncommon in what I have read for people to start out on this diet by eating a lot of such concoctions. However, I also see a lot of people repeatedly failing to stay on the diet long term until they release their desire to mimic their old diets.

  4. Hi Amber,
    I don't quite see how the correctness of the mentioned assumption (that we evolved to eat a significant amount of plant matter..) would be a, let alone necessary, condition for certain plants to contain one or more healthful compounds.
    Also I think it would be helpful to distinguish more precisely between food and medicinal substances, to make clear which of these uses of plant matter you're ultimately targeting in your article.
    Then you're arguing the naivety of the idea of some general human symbiosis with plants, which I agree with, but that still doesn't rule out a potential nutritional value of certain plants, especially after some adequate form of processing. You're also not applying the same reasoning to meat consumption - and why would you - a 'missing symbiosis' with the origin life form just isn't an indicator for healthiness of food. The same goes for your final argument, which seems to imply that the ketogenic diet - meat - is somehow void of toxins.
    I understand you're doing your best to stick to the rules of science, cite your sources and ask your readers to be skeptical, which is great. However, science is just starting to begin to understand how the human body works and reacts to food, influenced by an immense diversity of health conditions, genetic predispositions, environmental circumstances, exposure to toxins etc., and I can't help but reading this article and the referenced "how to try a keto diet"[2] as the assumption of the potential existence of some kind of silver-bullet diet ("[...] suggesting it _will_ protect against those diseases" , which are 'those', btw?).
    And finally, as you explicitly mention practicality on your science blog - let me say that the ketogenic diet is an absolute luxury in terms of resources, at the very least on a global scale. The combined per capita beef and sheep meat consumption in China in 2010 was 7.87 kg, and according to is estimated to grow up to roughly 9 kg until 2020. Now, put China on a ketogenic diet instead..
    If we are looking for diets with such desirable properties as you seem to expect the ketogenic diet to have, the research done on the keto diet might help, but it in itself won't be of any practical value for feeding the masses. Which makes me conclude that we have no choice but to (continue to) consider plant matter as a potential ingredient of hopefully sufficiently many healthy diets.

    regards, Johann

    1. Dear Johann,

      Thank you for writing.

      First let me acknowledge that this article is not one of the kind you should expect to find on my science blog (The Ketogenic Diet for Health. This is my personal blog, where I only intend to banter ideas around. I'm glad if you've come to expect that, though.

      Let me see if I can address your specific points.

      First, I think you are disagreeing with my assumption about my reader's motivation in making the statement that we should eat plants because they have so many good things in them, and that we are still discovering what those good things are. I perceive this as an implicit statement that there is a reason we have been eating them, or that our eating them has a function, to be more precise.

      So one of my responses is that those things in plants aren't necessarily good at all, but ambivalent at best. The other was that we haven't, in fact, necessarily been eating them all along. Insofar as we do eat plants, they probably bear little resemblance to any specific profile during evolution.

      I agree that preying on animals isn't symbiotic either (modulo arguments about ecological balance, which are taken quite seriously by wildlife ecologists who stress the importance of hunting, and this argument, of course, also applies to wild plants).

      I actually do think that meat is less rife with toxins (not necessarily counting those coming from our farming practices—that is another issue, and one that applies to plant farming as well). The reason I think so is that for the animals we routinely eat, it is not part of their survival strategy to be toxic, as it is in plants, and some insects, amphibians, and reptiles.

      It is a shame that my enthusiasm does sometimes come across as sounding like I believe in a silver bullet. The range of conditions that appear to be relieved by a ketogenic diet is surprising, even to me, or at least intially was. After several such surprises, it is difficult not to begin to suspect that this is because glycolytic diets are interfering with function. Perhaps the variety of conditions reflects genetic variation in the way different human bodies cope with such interference. In other words, I suspect that the ketogenic metabolism, which is, after all, a normal mode of functioning common to humans, may be the healthy default.

      I take your point about sustainability of resources quite seriously. I sincerely hope that raising meat turns out to be more sustainable than some people believe, or that we figure out how to make it so. I have heard points to both sides. However, it isn't really relevant to the question at hand, which is whether it is important for health to eat plants.

      I don't think it makes sense to ask people who can afford a resource that drastically improves their lives to choose to forgo it because not everyone has access (not to imply that you are suggesting so, but others have).
      If we are concerned with how best to feed the poor, it would be better if we knew what is essential and what is not.


  5. This comment has been removed by the author.

  6. Sort of on topic:

    We present early evidence linking a high prevalence of caries to a reliance on highly cariogenic wild plant foods in Pleistocene hunter-gatherers from North Africa. This evidence predates other high caries populations and the first signs of food production by several thousand years. We infer that increased reliance on wild plants rich in fermentable carbohydrates caused an early shift toward a disease-associated oral microbiota. Systematic harvesting and processing of wild food resources supported a more sedentary lifestyle during the Iberomaurusian than previously recognized. This research challenges commonly held assumptions that high rates of caries are indicative of agricultural societies.

  7. Full post by Dr Ede:

  8. Another version of Pollan's slogan:

    Eat food. Too much. Mostly animals.