Thursday, June 6, 2013

Fairy stories and evolutionary medicine

Plausible does not mean probable
Nothing in medicine makes sense except in the light of evolution!  So goes a modification of the great 20th century evolutionary biologist Theodosius Dobzhansky's statement about biology; Dobzhansky was advocating the teaching of evolution in the schools, but his quip is now tailored as the dictum of the field of evolutionary medicine (EM).

EM is the idea that evolutionary theory can be used to explain why we get the diseases we do.  That is, when something goes wrong with the body, it's because our genome evolved long ago, in a very different ecological context, and thus we're not adapted to life in the modern world.  We're out of sync with our evolutionary history, and this is what leads to disease.  And should, as a consequence, inform clinical practice.

A very nicely argued paper by Michael Cournoyea in the latest Perspectives in Biology and Medicine ("Ancestral Assumptions and the Clinical Uncertainty of Evolutionary Medicine") questions the theoretical basis of this idea, as well as its clinical relevance.  He refers to two recent collections of writings on EM, noting that they include chapters on
...nutrition, type 2 diabetes, childbirth, menstruation, stress, altitude sickness, sleep, heart disease, obesity, addiction, delusions, and more. For example, contributors suggest that addiction is just the malfunctioning of ancestral wanting and seeking mechanisms in modern lifestyles; sleep problems reflect an ancestral tension between the need to remain vigilant and the need to rest; and the life of those with chronic heart failure may be prolonged by blocking ancient neurohormonal adaptations.
These kinds of arguments are not uncommon even in everyday analyses of modern society: it's often said, e.g., that severe anxiety disorder is common in today's stressful society, when we have to do things we did not evolve to do like public speaking or driving across bridges.  But anxiety evolved as a normal defensive response to fearful situations, and at the time was adaptive rather than debilitating.  Too little anxiety would have been selected against, because, well, it's obvious -- he who had no fear of lions was dinner. 

Cournoyea describes the classic example of the practical uses of EM.  Fever, it is thought, must have evolved for a reason.  Because it accompanies illness, it must somehow help the immune system combat infection.  Therefore, it's probably best to leave fever untreated.  But, as he says, while this idea might pinpoint a way for researchers to study fever, it isn't clear that this is relevant to clinical practice, and the conclusion warranted. 

Cournoyea stresses the difference between evolutionary explanations on the micro scale (antibiotic resistance, or the evolution of influenza viruses, e.g.) and the macro scale ("the origin and adaptive function of physiological processes" such as breast-feeding), noting that the application of evolutionary theory to the first is much more informative than the speculative kinds of applications that are considered in the longer time scale, and indeed the only kinds of applications that can be considered since these traits evolved so long ago.  Why did fever or menopause evolve?  These questions can't be answered other than speculatively.  Yet, much of evolutionary medicine is dedicated to answering such questions.  And, as Cournoyea points out, generally from a strong adaptationist perspective.
I conclude that faults in these fundamental assumptions undermine EM’s central objective of providing clinically relevant medical knowledge. I maintain that evolution’s place in the clinic is controversial at best, and that these conceptual blind spots lead to misguided conclusions about our supposed human nature.
As he also points out, many evolutionary theorists by now have questioned the assumption that all traits have been molded by natural selection to be highly adaptive, but EM is built on this assumption.  "Developmental constraints and the nonselective mechanisms of evolution are given a backseat to adaptive functionality..."  But, he writes, explanations of ultimate causation, the adaptive hypotheses put forward by EM, are not relevant to clinical practice. And we think it is defensible to assert that many who make evolutionary arguments of this sort have only a cartoon understanding of how evolution works.

And there's more
As regular readers of MT know, we harp on the problem of assuming adaptation and the all-encompassing role of natural selection in evolution, all the time, so we are in complete agreement with Cournoyea's critique of EM in this regard.  But there's another fundamental problem with EM that Cournoyea's spot-on paper doesn't address -- not that it should -- and that is the idea that we evolved in a given environment to be highly adapted to that environment, and that one alone.

Organisms that are highly adapted to a single environment, unable to adapt to change, are very vulnerable.  Or, put another way, extinct.  One of the most ubiquitous characteristics of life, so ubiquitous that we've called it a principle of life, is facultativeness, the ability to adapt to change.  The "paleo diet" craze is another 'philosophy' founded on the idea that we evolved to eat the raw foods that our ancestors ate as they were becoming modern, and thus that we only achieve optimal health following that diet.

But by this logic, we're in danger of going down Alfred Russel Wallace's slippery path to excepting humans from evolution.  And he was by no means alone in invoking human exceptionalism in such contexts (Darwin did it in his own way, too).  Indeed, we might even find ourselves attending seances as Wallace did late in life.  He thought that the human ability to do calculus was evidence that it was impossible for us to have evolved by natural selection alone, because the human brain didn't evolve doing calculus.  Similarly, because we didn't evolve eating Twinkies, they must be lethal.

But we can do calculus, because what evolved was our brain's ability to figure things out.  We didn't evolve knowing how to cross busy city streets either, but we can do that, too, and we can build rockets and go to the moon.  And, people eat a huge range of foods now, and have done as we evolved -- there are 7 billion of us on Earth, speaking thousands of languages and still eating varied diets, so clearly we're adaptable.  There's no one way to be human -- nor are there single pathways to disease.

Fairy stories can't always be refuted, but that doesn't make them true
Just because we can invoke -- or, basically, assume -- a principle of nature, like natural selection, this doesn't mean that the principle works without exception or is invoked in an empirically correct way.   But if a tale is being told that rests on that assumption, there is a very widespread tendency to let the assumption slip by without justification, and accept the fable and its moral.  Fairy stories can't be refuted if the existence of fairies is assumed, not something to prove before we judge the story itself.

Likewise, despite Darwin's venerable insights, natural selection is a possible aspect of evolution -- nobody can doubt that -- but it is an aspect, not a law of nature.  Other factors, notably chance and implications of organizational complexity, greatly modify what one can legitimately say about selection.  This is so widely understood by those who care to look at things carefully that there is no excuse for the widespread uncritical selectionism so widely assumed about Nature by so many, in evolutionary medicine but also in evolutionary psychology, too much of anthropology, genetics, and beyond.

If you assume that some trait is the result of natural selection, then you can make up stories as to how it worked to produce your particular trait.  If your first guess doesn't fit, then you feel justified to make up counter-adaptationist stories.  But if selection is assumed, it can't be refuted since it's not open to test.
Just because a tale is possible, or seems plausible, this does not in any way make it true, especially if its basic premise is not open to question, as is so often the case in biomedical research or EM.

And Dobzhansky's quip?
The co-opting of Dobzhanksy's quip is very telling.  As we also noted in our book MT, even in its original form with respect to biology, it is manifestly untrue. Did nothing in biological research make sense before Darwin?  While nothing in biology (or medicine, for that matter) is inconsistent with our understanding of evolution, it is just not true that nothing 'makes sense' except in that context.

The blatantly obvious truth of this is that much of biomedical research has nothing to do with evolution, as well as the fact that many if not most people in the field know very little about evolution beyond a caricature (and, of course, many creationists of various religions work in biology and medicine).  But the slogan can be routinely invoked without any critical assessment by almost anybody, even in biology, including a cult -- and that's the right way to characterize it -- of a few leaders of EM whose words seem rarely to be questioned, much less critiqued properly.  This fact shows that invoking the quote isn't that different from saying a rosary: a comfortable, feel-good reinforcement of what often amounts to little more than a mantra of an ideology.  That's not the same as science!

It's a shame people have to be this way, because the scientific challenges are serious, and there is much that we don't know about evolution; the subject deserves better for its own sake, and for the public that pays for results.


michael.cournoyea said...

Anne, thank you for your thoughtful commentary on Evolutionary Medicine. It's not easy being critical when there continues to be such enthusiasm for EM. Just yesterday, I was reading an editorial in BMC Medicine about why the time is ripe for translational EM research. In the article, Brüne and Hochberg contend that

"questions concerning the causes of sickness and disease cannot be answered without an evolutionary approach. Nor can medicine promote preventive action to help people maintain health without a profound knowledge of gene-environment interactions that were shaped in our distant past." (April 2013, p.1)

The editorial serves to preface a small series of articles on EM that "demonstrate how insights from evolutionary theory not only improve our understanding of disease, but offer new ways to diagnose, manage and prevent human ailment." (p.2) After reading these articles, I remain a skeptic about the real value of EM.

(You might be interested that a couple of these authors suggest that EvoDevo might influence medicine.)

My interest in EM began a few years ago when I overheard a CBC radio program discussing evolutionary exercise (or CrossFit, side dish to the paleo-diet) wherein participants aim to replicate exercises from the Pleistocene. Such models rely on notions of a static or minimally dynamic EEA, emphasizing the mismatch to current lifestyles. Listening to that radio snippet, I couldn't help but think about Buller's critique of EEA (which you mention). It is a critique that I now take for granted. Little did I know that those cursory thoughts about caveman exercise would lead me down a rabbit hole of research.


Anne Buchanan said...

Thanks, Michael. It's so nice to have your comments. I haven't seen the BMC Medicine pieces. Like most adaptationist Just-So stories, belief that we know what explains traits is akin to religion. Funny that we need to follow a paleo-diet and paleo-exercises, but it's okay to fly in planes.

Randolph Nesse said...

Those of us who are working to bring evolutionary biology to bear on problems in medicine and public health are unanimous in recognizing that efforts to understand evolutionary reasons for vulnerability to disease are just one of many applications; see my article with Stearns to see where this fits in all 10 categories of applications. Evolutionary medicine is the application of a basic science to medicine in the same way that genetic medicine applies the basic science of genetics. Most of us in evolutionary medicine are all too aware that testing hypotheses about why natural selection has left bodies with traits that make them vulnerable to disease is especially difficult; see my article "Ten Questions for Evolutionary Studies of Disease Vulnerability" for my attempt to help students and others avoid elementary mistakes. Cournoyea ignores this entire literature. He also seems to think that evolutionary medicine claims to offer specific direct clinical recommendations, when most of us in the field have repeatedly emphasized that evolution does not make direct clinical recommendations. It is useful the way other basic sciences are a foundation for understanding the body and for deriving hypotheses whose testing provides the evidence for making clinical decisions. Of course, there are people who advocate for unsubstantiated ideas, and some who try to derive clinical recommendations directly from theory, but such problems are unfortunately common in many areas of medicine. Examples from evolutionary medicine do not make an argument that evolutionary biology should not be applied to medicine. Instead, they suggest the need for better evolution education for physicians, medical researchers, and philosophers.

Global generalizations about evolutionary medicine are not very useful. What is needed, as in all science, is the hard work of considering specific questions and hypotheses, one by one. As for the generalization that evolutionary medicine assumes all traits are adaptations, the very opposite is closer to the truth; an evolutionary view explains why nothing in the body can be perfect and why bodies have plenty of maladaptations (see "Maladaptation and Natural Selection" for details).

Those interested in learning more about evolutionary medicine will want to read more at The Evolution and Medicine Review

Ken Weiss said...

We'll stick with our original post.

Readers who want to see the marketing propaganda for evolutionary medicine can find it all over the web, including the site provided in the foregoing comment.

There are also countless articles and a few recent books that provide a responsible, properly circumspect understanding of is known about how evolution actually works. They, however, are not as simplistic, and require more careful attention.

Anne Buchanan said...

Thanks for weighing in. Perhaps it's a gestalt thing, but after reviewing the publications you point to in your comment, I stand by our post.

Randolph Nesse said...

How odd. I have been using Ken's article on Quixotic trait loci as one of the best examples of the utility of applying evolution in medicine. If more geneticists took this kind of evolutionary sophisticated view, genomic medicine would be much further along now. Providing physicians with education about evolution is a goal I hope we share. If you see a better way to do it, please share your suggestions so we can learn from you.

Anne Buchanan said...

As we say in the post, much biomedical research has nothing to do with evolution. Which doesn't mean that evolution has nothing to do with biomedicine but that evolutionary theory usually isn't much help. I don't presume to speak for him, but based on his article, it seems that Cournoyea would concur.

Ken Weiss said...

The problem to me is the plethora of, I'm sorry to say, simplistic panselectionist stuff being published under this catch-phrase, as if it in some way specifically relates to adaptative evolution.

The issue is that causation today is so often so genomically complex that referring it to evolution or Darwin or adaptation adds little to our understanding of the biology. It is far too deterministic relative to what we actually know about life.

Of course our traits evolved, but we rarely have any way to know why or how (or, often, even when). Many of our traits long antedate humans, of course. Extrapolating from today's experience is generally groundless.

I don't practice medicine, of course, but understanding the biology of the effects of variation today is difficult enough without trying to relate it to any serious level patterns of past fitness, etc.

To relate genomic variation among people and continents (and species) to variation in outcomes (like diabetes or menopause or fever) rarely requires explicit invocations of past selection.

It may be true that examples like lactase persistence and malaria resistance evolved by some plausible (and, on rare occasions, actually demonstrable) selective mechanism. But it gains virtually nothing to know that in terms of therapy. Evolution involved past environments and so on, but health issues relate to present ones.

The major exception is history and evolutionary concepts in the sense of their role in generating the amount and general pattern of variation that may (or may not) be functional. Or whether or not meta-analysis is likely to be useful if the samples are from different populations, and things like that.

Knowing the role of variation is different from evolutionary reconstructions, except in terms of reconstructing history to understand geographic variation (and why some thalassemia genes are found in Asia, others in Africa, etc.).

For short-term evolution, such as strategies to combat parasites and infectio, antibiotic and pesticide resistance and the like, evolution may really be relevant in every respect. There, selection is strong, movement of people and pathogens rapid and on a large scale, etc. We can see it in action on a relevant scale.

But this doesn't take away from what I believe is a large, facile amount of untrammeled speculation in the name of Darwin as if it was actually explanatory and relevant to medicine or epidemiology.

For example, there is a plethora of studies purporting to relate to reproductive fitness, in the 'evolutionary' context, by analyzing determinants of lifespan and reproduction and grandparenting and so on, in 21st century America. This implied extrapolation really has only to do with family and health structures today, that are demonstrably ephemeral, and has essentially zero relevance to future genomic variation or how today's variation got here. But examples are all over the journal and blogosphere.

We have had direct personal experience of blatant censorship in evolutionary medicine, when even responsibly trying to challenge the widespread simplism, as we see it, of this area.

Evolution is important in understanding variation, and was important in molding what's here to day. But from a functional point of view, we rarely can know the responsible forces or processes. We may need evolutionary awareness but we don't need a catch-phrase that to me is more about marketing than substance, as it is widely used.

But circumspection and tempering of speculation are not good for business these days, and I think we're unlikely to agree on what's simplistic, ideological, or excessive in this area, or what is responsible and cogent.

Holly Dunsworth said...

Thank you so so much for this post.

Anne Buchanan said...

I in turn thank Michael Cournoyea for his excellent Perspectives paper.