The evolutionary value of cancer

A recent article in Nature Reviews documents the rarity of cancer in antiquity, and speculates that "carcinogenic environmental factors in modern societies" may explain today's much higher cancer rates. Without disputing the authors' basic premise or conclusions, I wish to posit an alternative explanation: old people are useless. 

Women typically finish child bearing by 40 and child raising by 60 - sometimes much earlier. Sure, every so often someone tries to make middle-aged ladies feel good by writing some nice paper about the evolutionary value of grandmothers, but basically living past middle age contributes minimally to the number of surviving children and grandchildren one has, and from an evolutionary point of view, that's all that matters.

And, lest male readers feel too smug about their continued reproductive capabilities:

An old male is competition for scarce reproductive resources. Once he has lost his pride to a younger male, he's not likely to last very long.

From an evolutionary point of view, therefore, cancer has distinct advantages. It is primarily a disease of middle and old age, so it reduces the number of elderly (and less productive) mouths to feed. This frees up resources for more productive younger workers and their families hence increases their chances of survival. Cancer deaths also, potentially, increase the productivity and survival possibilities of the young through the reallocation of capital: a person who inherits the family farm at 30 can do better for his or her children than a person who does not inherit until 50 or 60.

The theory that evolution accounts for the increasing prevalence of cancer is vulnerable to a number of criticisms. First, if cancer enhances survival of the young, why didn't it appear earlier in human evolution? That can be explained by the relatively late emergence of agricultural societies - in a hunter-gather environment the old are vulnerable to other forms of mortality, such as not being able to keep up with the group. Second, one could argue that evolutionary change happens so slowly that it cannot possibly explain the differences between ancient Egyptians and modern peoples. Third, any evolutionary effects are bound to be small relative to the large and well-documented effects of carcinogens such as asbestos, tobacco, etc. Moreover, I have offered no empirical support whatsoever for the hypothesis that cancer has evolutionary value.

Indeed, this is the fatal attraction of socio-biology - it's too easy to come up with untestable explanations for just about anything.

 

Comments

[david]

The naive explanation as presented here would also suffer from the flaws of older group-selection ideas...

[Frances Woolley]

David, not if people live with their children. So in a pre-agricultural society, when a tribe lives together as a single group, people whose parents die of cancer would have no particular survival advantage over people who don't. This would explain the lack of cancer in the Egyptians - at that time, people hadn't been living in agricultural societies for all that long.

But in an agricultural society, when people live in family units and there is significant inherited property, those who inherit young should have a survival advantage over people who don't, because they have more capital, more land, and thus can produce more.

I'm not convinced, honestly, that the theory is right. But the group selection problem isn't the reason that it's wrong.

[Maxine Udall (girl economist)]

Two observations: 1) you seem to be assuming that only market productivity counts and 2)implicit in your conjecture is the idea that we should allow CA to "do it's thing" since we no longer live in a society where the elderly can simply wander out the cave door and be eaten by the nearest grizzly or sabre tooth.

It's not clear to me that the elderly are less productive (or useless as you put it) and therefore expendable. Many care for and provide role models for their grandchildren and great grandchildren, enabling middle aged children to participate in the labor force. Many of the currently elderly in the the US made extreme sacrifices in WW II. In a utilitarian world, how should we counterbalance that productivity that benefited us all as well as other countries against late in life waning of productive capacity? What about all the own time and effort that went into child-rearing, does that not offset to some extent the larger consumption of resources at the end of life? When we add up the benefits and costs of an individual's life, is it over the entire life cycle retrospectively, their current life stage, or the unknown, but theoretically longer expected life of the very young? And if CA is an evolutionary response to our inability to wander in front of speeding mastodons or Mack trucks at the optimal point in our "productivity" and we adopt a public policy that allows it to fulfill this purpose, then how shall we compensate for the loss of profits and tech innovation that have resulted from all the (insured) old people getting this disease; innovation that benefits the young and middle aged who also get it? Like everything else, there are probably economies of scale and incentives of scale for cancer research and treatment. I suspect that some of it would not have occurred or would not have occurred so quickly had there not been the larger "market" of the insured elderly. Should that be considered?

I don't know the answers to any of these questions although I have some ideas about how we should think about them. At some point, we are also going to have to think about how we allow people to wander out of the cave to be eaten by a sabre-tooth, so to speak, when they wish to do so. There are some among us who do appreciate resource costs and who do evaluate the benefits and costs of our medical treatment and who would be happy to check out earlier rather than later when the costs start outweighing the likely benefits. For us, the challenge will be to persuade our families and friends to step away from the cave door.

[Frances Woolley]

Maxine: "At some point, we are also going to have to think about how we allow people to wander out of the cave to be eaten by a sabre-tooth, so to speak, when they wish to do so." Yup. And you thought abortion was a hot-button social issue...

There's actually a fairly extensive literature on the relationship between age and productivity, and most measures of productivity decline with age - in my own research (hopefully to appear on this blog in the new year) teaching evaluations, for example, do.

As someone who's staring biological uselessness in the face, I certainly would like to buy into your ideas about the value of older people. But when I'm fumbling with my credit card, trying to figure out whether it's a chip and pin or a swipe machine and then typing in the wrong PIN number, and the cashier gives me a "geez you're stupid" look, I don't feel it.

[Bob Smith]

In practice, for most of human history most people wouldn't have lived long enough to die of cancer, so its evolutionary effects would have been miniscule (for the same reason, there would have been little evolutionary pressures which would have made humans less susceptible to cancer, whereas we have seen evolutionary adaptations to other diseases - for example the protection against maleria provided to carriers of the genes for sickle cell anemia). Cancer is, to a large degree, an illness of otherwise wealthy and healthy societies.

The modern prevalence of most cancers probably has less to do with "carcinogenic environmental factors in modern societies" than to do with the fact that, for the first time, people are living long enough to die from cancer, rather than from other diseases, hunger, and sabre-tooth tigers (and the fact that we have the technology to detect otherwise benign occurences of the disease).

Moreover, query whether the modern environment is neccesarily more carcinogenic than in the past. For example, I gather that the use of solid fuel for indoor cooking is a major health issue in the developing world and smoke from indoor cooking has been linked with higher incidences of lung cancer. I'm not equipped to go to bat on that science, but it certainly sounds plausible. Living in a smoke-filled house would have been the norm for most of our ancestors (certainly for all of us whose ancestors lived in colder environments) since the first caveman (or, more likely, cavewoman) moved a fire inside.

[Maxine Udall (girl economist)]

"There's actually a fairly extensive literature on the relationship between age and productivity, and most measures of productivity decline with age - in my own research (hopefully to appear on this blog in the new year) teaching evaluations, for example, do."

Frances, I'm familiar with some of it. I don't see the lit measuring non-market output very accurately especially among the elderly. Also, I wonder what the incentive will be on the non-elderly when they realize that all that matters is market productivity and that once they become "useless" in the eyes of the young (yikes! there's a scary thought...the young judging anyone's usefulness), it's effectively "curtains." I suppose they might work longer and harder, but they might also decide to eat, drink, and be merry since tomorrow's death may be sooner than we think.

I hope wisdom is not measured by how adept we are at swiping our credit cards. In fact, I'd wager we'd be a tad better off now if more of us had been less adept at swiping them.

[Frances Woolley]

Maxine: "I don't see the lit measuring non-market output very accurately especially among the elderly." Particularly an issue when it comes to care-giving work.

"they become "useless" in the eyes of the young"

The danger of socio-biology is that it can be used to argue "Whatever is, is right."

I don't see it like that. Socio-biology helps me understand why the homemade peanut butter and chocolate chip cookie I just ate tasted so good - in a calorie-scarce environment there's an evolutionary advantage to craving sugar and fat because these are the best sources of calories. Understanding that craving helps me overcome it (sometimes).

We can pretend that everyone loves and adores old folks, and especially older women, but that's not what I see around me - check out the women over 40 on TSN or Spike TV or Space sometime - it won't take you very long.

Confronting our feelings is more useful, and more likely to lead to real policy insights.

Bob Smith, yes, indoor stoves can produce fine particulate levels comparable to Beijing on a bad day. Other natural nasties: toxic molds, arsenic, cyanide, micro-nutrient deficiencies. On the other hand: microwaves? cell phones? artificial light? pesticides? synthetic estrogen-type compounds?

[Jacques Krzepkowski]

Indeed, this is the fatal attraction of social economics - it's too easy to come up with untestable explanations for just about anything.

[Frances Woolley]

Terry McGarty thinks I've lost it (but he says some nice things about Nick Rowe):

http://terrymcgarty.blogspot.com/2010/11/cancer-and-evolution.html

Terry writes: "How in God's name one can state this without any basis goes beyond reason. It is nice to state that when a reproductive entity reaches the time at which it can no longer reproduce it should be expendable, and even more so expended, then we males may just keep going till our 90s and somehow extending Woolley's argument just rid the world of postmenopausal women! This is akin to the Modest Proposal suggesting the use of Irish children for food! Do the English really think this way."

I take the comparison to Swift as a compliment.

Swift's Modest Proposal was intended as a devastating satire of the attitude of the English towards the natives of his home country, Ireland. I'd be thrilled if people felt a similar level of outrage towards my suggestions about the useless of postmenopausal women, and were spurred to do things like respect the suggestions of older women in committee meetings, or make job offers to older people.

Interestingly enough, in one of the lesser known parts of Swift's Gulliver's Travels, he describes a group of people who were born immortal. The birth of ones of these individuals was seen as a curse - they had eternal life, but not eternal youth.

I'd actually been thinking about writing about that part of Swift when I came up with the cancer idea.

Terry, by the way, I'm a "she" not a "he."

[Terry McGarty]

Frances
"
I am sorry if any afront was taken, and I believe I corrected the "he" to "she", a bit too early in the am and now I note the "es" not "is", snow falling and the like is a distraction. My Irish humor at times is a bit biting and being an engineer by training, along with a bit of medicine, one gets to become blunt at times. My point is that we know a great deal of cancer and that the cellular basis does not change so quickly, we just seem to be getting older, and cells were meant to replicate just so many times, then off to the dump with them.As to my words, Swiftian satire with an American edge, New York for specifics.

Interesting thoughts but cancer is a challenge for which those of us in genomics still find a bit difficult. Regards from south of the border!

Terry McGarty

[Linda Welling]

Fran, I took your post, in part, as an attack on evolutionary explanations - which I have a hard time taking seriously, for the most part. Interesting thoughts about the chip card et al, though: do the old become even more useless, more quickly, in times of rapid technological change? Did something akin to cancer (in effects) arise in the past when technological change made the older generations' skills obsolete? Then again, in the build-up to 2000, the older generation of computer programmers had windfall gains as they were brought in to rewrite older programs.

[Patrick]

A few loosely points:

I think you're conflating ageing and death with cancer. Young people get cancer too.

@Frances: " living past middle age contributes minimally to the number of surviving children and grandchildren one has"

Nit pick: What matters is the reproductive success of offspring. Survival is a necessary but not sufficient condition.

Much depends on the environment. Counterexample: Grandmothers in AIDS ravaged Africa raising their grandchildren because their children are dead.

Biologists have known for some time that artificially selecting for early/late reproduction shortens/extends lifespan. Human life spans have increased mainly because we have increasingly delayed reproduction.

I think the uselessness of old people (as measured by decreased productivity) is in part a product of our insulation from the ravages of Nature. Given a harsher existence (e.g. hunter/gatherer in the Arctic), being old was a signal that you were really good at surviving. So it was smart for the younger folks to put up with your lower productivity and keep you around as a kind of insurance e.g. "Geez, there are no seals this year and we're getting hungry. Let's ask the old timer where to find food when the seal hunt fails."

But in a world where old age is basically just a signal that your knowledge and skills are obsolete and that you consumed the most health care, there's significant incentives for the younger generation to turn you into Soylent Green and feed you to their babies.

[Frances Woolley]

Terry - I was delighted that you took the time to comment and respond to my post, and pleased you were able to find my response on this site. I couldn't figure out how to comment on your blog.

So your reading of the science is that the increase in cancer rates is explained by the increase in the number of old people?

Linda - that's an interesting question - what determines the extent of ageism? Some people have done "correspondence testing" or "audit" studies recently along the lines of submitting two almost identical c.v.s, one from a 52 year old and the other from a 25 year old, and seeing who is more likely to get the callback. (Not surprisingly, the 25 year old - it's a very strong effect). I don't remember if those studies have picked up differences across industries/occupations etc.

Patrick: interesting comments. It's not hard to find customs/traditions that involve the killing of older women: witch burning in Europe and North America, Sati (or suttee) http://en.wikipedia.org/wiki/Sati_%28practice%29 in India, though there's always a question about how widespread such practices ever were.

[Patrick]

@Frances: I can easily believe that misogyny often (always?) trumps survival.

[learningjames]

Maybe we shouldn't wait to kill those men. Doesn't history show that people like ghengis khan often had fathers that died while they were young ?

[Frances Woolley]

learningjames - I've heard various theories about the effect of growing up with an absent father on Bill Clinton, but haven't heard the Genghis Khan hypothesis.

[James Splinter]

Hello Frances!

I am not sure if you are suggesting that cancer is a evolutionary advantage (in the traditional sense) or rather cancer could be viewed as an evolutionary advantage. If you are arguing the former I would like to strongly disagree. I venture a third possibility that this is just a satirical exercise in demonstrating the ridiculousness of socio-biology, making the rest of my post rather large waste in time, but as a good economist, I will *assume* this is not the case and proceed.

Humans stopped evolving thousands of years ago, any change in the people from then on would be social. (Increase in lifespan, average height, etc. are functions of improving health care and nutrition).During an evolutionary period, evolution, as you correctly stated, is driven by those able to reproduce. Again, as you correctly suggested, as long as your offspring survive to do the same, everything else doesn't matter. Dieing at age 30 or 40 was common place during that evolutionary time (pre-historic), and to look at our primate brothers, you were most likely to die from environmental causes, starvation, accident or death by attack. High rates of cancer amongst an aging populous needs not be indicative of an evolutionary advantage, it could merely, and in my opinion, more likely, be a spurious effect. (lim (value of genetic predisposition to cancer) tends to zero as t approaches 40, where t is number of years alive.)

I got the impression that you were suggesting that cancer in humans could be responsible for premature death and this death gave an evolutionary advantage to their offspring (whereas I postulate that concepts of inheritance postdate the end of evolution). My position is that people died before cancer was ever an issue (from countless other causes) and only in modern times have we seen cancer rates skyrocket, because, well, we have to die from something.

As a side note, cancer (please forgive my tragic irony) suffers from a survivor bias. It's rampant in today population because so many other diseases have been conquered. If heath care was at its levels it was thousands of years ago, let alone 100 years ago, we would be getting knocked off by the plague, polio, the flu, diarrhea, etc. leaving little room for us to live to die from cancer.

[Ed Seedhouse]

During the period when our basic genome was evolving, people lived a very short time and if, as you say, cancer is a disease mainly of the old then it is hard to see how cancer could be selected for when no one lived long enough to get it! We didn't start living longer lives until we invented agriculture, and that happened, in evolutionary terms, an eyeblink ago. We were pretty much the same then as we are now.

In any event cancers are already well explained in other ways.

[Patrick]

"Humans stopped evolving thousands of years ago"

You can't turn off natural selection. It happens.

[James Splinter]

@Ed, exactly what I am suggesting, I hope I communicated it in my previous post.

@Patrick, for all intents and purposes, the average human is genetically identical to the average human of thousands of years ago. Natural selection, in terms of human evolution, has stopped, and in reality is likely going in reverse. Pure natural selection, as in the need to have superior genetics through random mutation in order to have superior offspring (those that live long enough to have children, and lots of them), has on an aggregate stopped and is starting to reverse. Humans with genetic disorders that would otherwise result in some one dieing, are actually surviving and going on to have children of their own. Having good "survior genes" isn't a requirement for living or having children anymore.
Under "natural" conditions, the conditions conducive to natural selection, those with random mutations that reduce livelihood, die, and those with random mutations that increase livelihood live on, keeping the gene pool of the species contstantly improving (evolving). This really isn't the case right now, and hasn't been for a long time with humans.
Ofcourse, in small cases genes matter quite a bit. But on the aggregate, which constitutes the evolution of a species, it's no longer a factor.

[K]

James: We still select mates. Do you think the ugly reproduce at the same rate as the beautiful? And check out these stats. Single-earner-male couple families with two kids earn almost three times as much as Single-earner-male couple families with no kids. Chalk one up for the alpha males passing on their genes.

[Frances Woolley]

James Splinter, to say that humans stopped evolving thousands of years ago is, I think, overstating the case - I think blonde hair and some variations in ear wax are fairly recent developments - the argument is that blonde hair spread quickly through sexual selection. In fact, in your lifetime and mine there's probably been substantial changes in the frequency of various genetic characteristics within the human population simply as a result of differential population growth rates.

Remember also even when the average life span was relatively short, there was a large variation in that life span - an average life expectancy of 40 was generated by enough people living to a ripe old age to compensate for childhood death.

Still, Ed Seedhouse (and you) may be right to say that the period since the development of agriculture is too short a time frame over which to observe changes of the type hypothesized here, particularly given that any advantage from early inheritance is likely to be pretty small in the cosmic scheme of things.

Although Ed, just because we can explain cancer at the micro level, e.g. in terms of viruses or genetic predispositions, doesn't mean that it's explained at the macro level, i.e. why those genetic predispositions are there.

Are increasing cancer rates better explained by population aging and/or environmental causes than evolution? Almost certainly, yes.

[Patrick]

James Splinter: So long as you have different rates of reproductive success between individuals, you have natural selection and evolution. You may not like what's being selected for in some case, but it's still happening.

[Nick Rowe]

Frances: add in lactose tolerance to blond hair and ear wax.

[Mike Moffatt]

"James Splinter, to say that humans stopped evolving thousands of years ago is, I think, overstating the case - I think blonde hair and some variations in ear wax are fairly recent developments - the argument is that blonde hair spread quickly through sexual selection. In fact, in your lifetime and mine there's probably been substantial changes in the frequency of various genetic characteristics within the human population simply as a result of differential population growth rates. "

'Tis true. The only thing stopping me from signing up for cryogenics is my worry that there may be no attractive redheads in the future. And who would want to live in such a world?

http://science.howstuffworks.com/environmental/life/genetic/redhead-extinction.htm

[Stephen Gordon]

Don't worry; measures are being taken to make sure the genes live on. My wife and I both have red hair, as do our 3 kids.

[Alex Plante]

Many mammals remain fertile into old age. Indeed, in the case of elephants, it's the oldest females that are the most fertile. I think the females of most mammals remain fertile into old age. What needs to be explained is why human females experience menopause at about 2/3 of their lifespan. There was an article about that in Discover or Scientic American a few years ago (Sorry, I'm too lazy to dig them out to find the reference), but I think it had something to do with the fact that humans need to be "educated" for 20 years (and I dont mean classroom education, I mean street smarts and learning from experience).

[James Splinter]

@ Patrick,
"
James Splinter: So long as you have different rates of reproductive success between individuals, you have natural selection and evolution. You may not like what's being selected for in some case, but it's still happening.
"
I don't disagree with the idea, and likewise Frances's and Nick's points about traits we see disappearing or appearing (though lactose tolerance, and similarly celiacs, may continue to consume what they actually lack the genes to digest without undergoing a genetic change). I am suggesting that what we see now is social/sexual selection as opposed to natural selection. Red hair and blonde hair are recessive genes, and will of course disappear if blonde and redheads keep having children with non red heads and blondes. But if that doesn't happen, we will keep seeing them. It's not evolution or natural selection, it's a function of dominant and recessive genes. Blondes and red heads don't seem to have any better ability to produce offspring (or worse) compared to their not-so-colourful haired neighbours. Random mutation can come and go, it's only really natural selection when they have a worse (or better) time reproducing. Patrick, I fully see your point about differing rates of reproduction and how that leads to natural selection and evolution, I think we are looking at two sides of the same coin. If 100 years from now there are no red heads or blondes, I wouldn't exactly say we evolved out of it, since the typical human is still 99.99 + % genetically identical and is no better off in terms of survival. However, if you choose to define changes in traits and phenotypes over time as evolution, then I see where you are coming from.

The specific case of blondes is quite fascinating in my opinion. The tendency is for hair colour is to regress towards to darker coloured hair. (The offspring of a blonde and a brunette is more often than not a brunette or a dirty blonde) If you rule out infanticide, or something very similar, it would have been extraordinarily difficult for blondes to even ever exist (assuming early humans had dark hair). If we accept that blondes came into existence from a random mutation of a handful, if not one or two people, then it follows that the early blondes must have engaged in substantial inbreeding (cousin-cousin offspring are usually fine) and very little mating with the darker haired peoples they came from. It also follows that blondes must be descended from a very small gene pool (but who is to say we all aren't?) that migrated to northern Europe with out any dark haired people's tagging along.

@ K
Intuitively you would assume better looking people have more offspring. However (putting aside the fact that they may not be better at surviving) I would actually suggest that's not necessarily the case. More "plain" looking people tend to have stronger relationships. Looks are highly subjective, but if we accept that there are some generalisations about what's attractive or not we also see that there is an asymmetry when it comes to male-female attractiveness. For example, levels of hair, height, colour of hair, bone structure, facial features, etc. are often opposites of what's attractive in a male versus a female. Allow me to contrast: a broad shouldered, large jawed, cleft chin, hairy man of height 6'3 with a muscular disposition could be considered quite attractive, a female with those same characteristics would not. Like wise, a petite, slender or curvaceous (either or really), with small shoulders and delicate hands, with very little body hair, could be considered attractive, the male with those characteristics would likely not be so attractive. The attractive male-female combo I suggested could have attractive offspring, likewise they could very easily not, the less attractive male-female combo face a similar situation. (Take a look at James Haven-Angelina Jolie, brother-sister)
Looks are a fascinating subject, while we of course see trends, there also exists a tendency of regression towards to average and random attractiveness. I look stinkingly like my brother, but my sisters bear very little resemblance to each other. Why is Brad Pitt so attractive, but his brother, not so much? This applies to almost anyone.
Of course, there are always traits that seem fairly consistently desirable, good teeth, good skin, thick hair (on one's head), a general "healthy look".

Social factors in reproduction, like higher earnings, not surprisingly result in more children for those with the higher earnings. Anecdotally, I seem to see that higher educated people seem to wait longer and have fewer children. And of course it has been shown that poorer countries tend to have much higher birth rates. All of these very fascinating points.

@Alex, Check out this article on reproductive cessation in female mammals, wish I could find the full thing for you, http://cat.inist.fr/?aModele=afficheN&cpsidt=2213174
It also implies that it seems to exist for no particular benefit or detriment, similar to my suggestion of cancer in humans.
You may find topic of neogeny in humans interesting with regards to the human learning timeframe as a contrast to other species.

[Gregory Sokoloff]

Apparently, human genes have evolved as recently as 3,000 years ago among Tibetans adapting to oxygen deprivation at high altitudes. And a little further back, 7,000 to 10,000 years ago, some Asians evolved genes to metabolize alcohol much faster (changes to the alcohol dehydrogenases enzymes), and by the way get red faces in the process.

Still, none of this saves Frances' theory that cancer is an evolutionary adaptation to get rid of useless old people.

[OGT]

I think you have to think of this in more economic terms, what's the opportunity cost of not getting cancer?

For example, why don't we have the same, fairly remarkable, sense of sent that dogs do? There's no really plausible case that having canine powered sniffer wouldn't have been useful to hunter gatherers or farmers, all other things equal. So there's likely some trade of developmental energy, neurological processing, or some such that made the marginal gain to a better sense of smell no longer greater than the marginal cost.

Similarly with cancer, there is likely some cost to further perfecting the process by which cells split to further reduce or eliminate the chance of errors. No doubt there is greater evolutionary costs to childhood cancer than in the elderly. But that doesn't necessarily imply that there is an advantage to the elderly getting cancer, just that the marginal cost of further perfecting the process is too low. In a world where there were other much stronger pressures on reproduction and survival that cost might not have to be too high.

(As I semi-recall from a discussion with a bio PhD friend of mine, the cost may have something to do with telomeres, which govern the number of times cells can divide. The longer they are the slower the aging process, but the higher the chance of malignant cancers... Truly there is no free lunch).

[slagondrayer]

It's important to not think of cancer as being selected for. The idea that the genome would select for things that obsolesce itself in old age doesn't work. There's a proof of this somewhere...

The only way cancer can be 'selected for' is through antagonistic pleiotropy.