The city that kills you makes you strong!

By Razib Khan | September 24, 2010 12:09 am

ResearchBlogging.orgOver the past day I’ve seen reports in the media of a new paper which claims that long-term urbanization in a region is strongly correlated with genetic variants for disease resistance. I managed to find the paper on Evolution‘s website as an accepted manuscript, ANCIENT URBANISATION PREDICTS GENETIC RESISTANCE TO TUBERCULOSIS:

A link between urban living and disease is seen in recent and historical records, but the presence of this association in prehistory has been difficult to assess. If the transition to urbanisation does result in an increase in disease-based mortality, we might expect to see evidence of increased disease resistance in longer-term urbanised populations, as the result of natural selection. To test this, we determined the frequency of an allele (SLC11A1 1729 + 55del4) associated with natural resistance to intra-cellular pathogens such as tuberculosis and leprosy. We found a highly significantly correlation with duration of urban settlement – populations with a long history of living in towns are better adapted to resisting these infections. This correlation remains strong when we correct for auto-correlation in allele frequencies due to shared population history. Our results therefore support the interpretation that infectious disease loads became an increasingly important cause of human mortality after the advent of urbanisation, highlighting the importance of population density in determining human health and the genetic structure of human populations.

298px-Pericles_Pio-Clementino_Inv269In some ways this seems plausible. There are a priori reasons why we’d expect to see a great deal of evolutionary change in regions of the genome correlated with variations in immune response. Diseases are one of the most likely reasons for why sex exists in complex multicellular species; sex allows a slow-reproducing population to bend with the rapid-fire punches of their pathogens by shuffling their defenses constantly. The results from recent work mapping patterns of variation in relation to natural selection generally indicate that immune related regions show plenty of signs of adaptation. No surprise, a “Red Queen” model whereby pathogens and their hosts constantly co-evolve would imply that immunologically relevant genes would never be at equilibrium frequencies for long, so we’d have a good shot at catching “selective sweeps” on some of the immune loci.

So how do cities play into this picture? I suspect that the picture is more complicated than the presentation in the paper, though I believe that the authors were constrained by considerations of space from evaluating all possibilities in full depth. There are two facts which I think are critical to understanding the pattern of variation here:

– All pre-modern societies were predominantly rural demographically. The difference between an “urban civilization” like Rome and a non-urban one such as Dark Age Ireland was that ~25% of the residents of the Roman Empire lived in urban areas (generously defined) while ~0% of Dark Age Irish lived in urban areas. Rome is generally considered to be a very urban pre-modern society, perhaps the most urban large-scale society before the 17th century.

– I also believe that ancient cities were population sinks. People simply did not replace themselves and cities only perpetuated their massive scale by serving as magnets for excess population from the rural hinterlands. Without appropriate political structures to maintain the population and generate incentives for an inflow migration ancient cities withered away very fast (Rome’s population went from hundreds of thousands to tens to of thousands in the 100 years from the early 6th to early 7th century because of political instability).

Before I go on much, let’s address the results presented in the paper. Below you see the frequencies of the allele which is more protective against tuberculosis in tabular form and on a map, as well as the logistic regressions which show the relationship between time since urbanization and the allele frequencies. Please note that they corrected for genetic relatedness in their regression, so the correlation isn’t just due to population stratification on a world wide scale.

no images were found

Since the allele which confers resistance is at a high frequency everywhere the difference is between those populations where the genotypes are predominantly in a homozygote state (e.g., Iranians), and those where only around half are resistant homozygotes (e.g., Sami). The authors note that because of the high frequency everywhere, including populations with no history of agriculture such as the Sami, one can’t posit a model where positive selection drove the disease resistant alleles from 0 to fixation. Rather, it perturbed the equilibrium frequency. Using the Tajima’s D statistic they do find evidence of balancing selection in both East Asians and Europeans. This would be in keeping with frequency-dependent models of pathogen-host co-evolution.

As I said before there are strong reasons to assume that natural selection reshaped the genomes of populations over the past 10,000 years. It really isn’t if, it’s how and what. The authors present some evidence for a particular variant of the gene SLC11A1 being the target of natural selection. To really accept this specific case I think we’ll need some follow up research. Rather, I want to focus on the narrative which is being pushed in the media that cities were the adaptive environments which really drove the shift in allele frequencies. I don’t think this was the case, I think the cities were essential, but I don’t think ancient urbanites left many descendants. Instead, I think cities, or urbanization, is first and foremost a critical gauge of population density and social complexity. Second, I believe that cities serve as facilitators and incubators for plague. In other words both urbanization and disease adaptation are derived from greater population density, while urbanization also serves a catalytic role in the spread of disease. This could explain the strong correlation we see.

I believe that the Eurasians who may have been subject to natural selection due to the rise of infectious disease are almost all the descendants by and large of ancient rural peasants, or, their rentier elites. These peasants were subject to much greater disease stress even without living in urban areas than hunter-gatherers and pastoralists because their population densities were higher, and quite often they were living a greater proportion of their lives snuggly against the Malthusian lid. Hunter-gatherers may have been healthier on average because of a more diversified diet as well as lower population densities due to endemic warfare. In contrast, agriculturalists lived closely packed together and were far more numerous than hunter-gatherers, and, their immune systems were probably less robust because of the shift away from a mix of meat, nuts and vegetables, to mostly grains.

800px-Republik_Venedig_Handelswege01A downstream consequence of agriculture was the rise of cities through the intermediate result of much higher population densities. I accept the literary depiction of ancient cities as filthy and unhealthful. There’s almost certainly a reason that pre-modern elites idealized rustic life, and had country villas. Additionally, though I assume that both the rural peasantry and urban proletarian led miserable lives, I believe that in terms of reproductive fitness the former were superior to the latter. From what I have read city life only became healthier than rural life in the United States in 1900, in large part due to a massive public health campaign triggered by fear of immigrant contagion. The high mortality rates and low reproductive fitness of urbanites implies that evolutionarily the more important role of cities were as nexus points for trade and the spread of disease. The book Justinian’s Flea chronicles the pandemic in the Roman Empire in the 6th century, in particular its origin in Constantinople from points east. We’re well aware today that a globalized world means that there’s an interconnectedness which can bring us strength through comparative advantage, but also catastrophe through contagion. This is a general dynamic, not simply one applicable to disease, but in the world before modern medicine the utility of trade networks for pathogens would have been of great importance.

One can imagine societies through the organismic lens as if they were cyclical wind-up toys. In the initial stage of expansion and integration political stability and concentration of power results in a peace which allows for the increase in population as more land inputs are thrown into primary production. Eventually diminishing returns kicks in and there’s no more land, so the labor squeezes itself more tightly on fixed land endowments. Their median physiological fitness declines as the pie gets cut into more and more pieces. All the while these massive numbers of peasants serve as the source of revenue for extractive elites, who found and patronize cities where they can signal their status and concentrate their wealth. Most pre-modern cities, like Rome and Constantinople, would have been economic parasites, depending on rents and plunder. As a sidelight cities such as Constantinople which were placed at transportation hubs would also become the focal points for trade, especially if they could be termini themselves for the luxury good trade which was dependent on the demand from rentier elites in residence in the metropolis. Finally, these cities would also be magnets for masses of armies because of the inevitably of sieges.

400px-Plato_Silanion_Musei_Capitolini_MC1377Eventually the combination of factors would result in the outbreak of plague. Social order would collapse, people would flee the cities, and populations would drop as the tightly run ship on the Malthusian margin ran aground. As the population dropped median health and wealth would return, and susceptibility to plague would decrease. And then the cycle of expansion and integration would start anew.

This is I believe the story of the rise and fall of urban societies which reshaped the genomes of people who lived across much of Eurasia. It isn’t a tale of urbanites, rather, urbanites for most of history have almost certainly been epiphenomena in a genetic sense. They’re the excess rural population which finds its way to the polis. Because of the squalor and lack of public health the lot of the urbanite was to consign their genes to oblivion. But for this deal with the devil the urban man had an opportunity to become immortal, and live on in human memory. It is their names which echo down through history, and roll off the tongues of the descendants of the peasants who have long ago forgotten their own genetic forebears.

Citation: Barnes, I., Duda, A., Pybus, O., & Thomas, M. G. (2010). ANCIENT URBANISATION PREDICTS GENETIC RESISTANCE TO TUBERCULOSI Evolution : 10.1111/j.1558-5646.2010.01132.x

Image Credit: Marie-Lan Nguyen, Nikater

CATEGORIZED UNDER: Anthroplogy, Culture, Human Evolution

Comments (26)

  1. bioIgnoramus

    Years ago I read that the British Army knew that healthy ploughboys made worse soldiers than the sweepings of the London slums because the ploughboys were so easily felled by disease.

  2. Don

    Rhazib: A book from you is in order. This topic is your focus, something like, “Ancient Cities as Malthusian Parasites.” I have a bulging file of your brief insights. You need to put something big together.

  3. onur

    Rhazib: A book from you is in order. This topic is your focus, something like, “Ancient Cities as Malthusian Parasites.” I have a bulging file of your brief insights. You need to put something big together.

    It would be nice if Razib devoted a thread exclusively to that issue or to the issue of ancient cities being population sinks.

  4. Paavo Ojala

    “extractive elites, who found and patronize cities where they can signal their status and concentrate their wealth. Most pre-modern cities, like Rome and Constantinople, would have been economic parasites, depending on rents and plunder.”

    It’s not intuitive to me why would these extractive elites want to found and patronize cities. I guess you are not claiming that the cities exist because elites want to signal their status. Why are cities founded and patronized? Why were there pre-modern cities?

    (questions in comment section are often perceived as criticism. But mine are honest lack of knowledge. Any previous post or link or reference to somewhere to find out?)

  5. The Republic of Ireland through most of its existence had a higher rate of TB than had Northern Ireland. I’m sure part of that was due to poorer living conditions, but maybe this contributed to it too.

  6. dave chamberlin

    The book I envision Razib writting starts with a bigger picture and then methodically presents examples that illustrate proofs that it is more than a theory. The bigger picture is the one only begun to be explored in the “10,000 Year Explosion”, that of our recent evolution. There is a pot that cooks us in the stew of human misery, generation by generation, in those circustances where selective pressures push us as a people in specific directions for centuries. Without the misery of malthusian limits the weeding of the weak isn’t possible.
    It won’t be a popular idea for it colides with multiple beliefs which won’t die easily. People still want to keep the human brain in a magic black box along with the soul, to make it just another organ in another species subject to the same laws of evolution will continue to provoke unpredictable negative responses.
    I think Joe Public needs to hear first how recent evolution has worked in less emotionally laden areas like disease, before he will listen to how it has also worked on his intellegence.

  7. Don

    Dave: While Rhazib’s insights combining genetics, demography, and culture would be a great book, probably much greater than “The 10,000 Year Explosion” and its simple selectionism, I do have to agree with your implication that “Joe Public” books sell more copies than seriously thoughtful and original works. Consider how well the shallow, pandering ” The Empathetic Civilization” sold.

  8. Greg Clark argues that it was not the aristocratic elites who left lots of descendants, but upper-middle class farmers. Still perhaps qualifying as rentier elites, but thought I’d add a bit of nuance.

  9. Joe

    I’m with some of your other commenters here. I’ve seen you claim before that cities were population sinks even in ancient times. Is there evidence for this?

  10. Is there evidence for this?

    two points

    1) cosmopolitan ancient cities don’t seem to live a strong genetic impact

    2) the literary evidence suggests that they disappear very quickly when politically enforced rents disappear. one model is that these people scatter to the countryside. i doubt this is so because #1, and, ancient societies were at the malthusian margin. how likely is it that the countryside could support city dwellers?

  11. onur

    cosmopolitan ancient cities don’t seem to live a strong genetic impact

    Is there any evidence for that?

  12. Is there any evidence for that?

    the regions around cities like rome and xian don’t show up as perturbations on the allele frequency maps. they should.

  13. onur

    they should.

    Why assume that?

  14. they were ethnically diverse.

  15. onur

    they were ethnically diverse.

    Italianthro (AKA Racial Reality) objects to the claim that cosmopolitan ancient cities were as ethnically diverse as to have any significant impact on the gene pools:

    Also he specifically devotes this thread to the minimal demographic impact of ancient slavery:

  16. onur. ok, i went over there and he uses the high mortality to support his claim. as for his post on slavery, he makes some of the same points i would have made: slaves did not breed much. the american slaves were very exceptional historically, and even they had much lower fertility than american whites (though that’s a bad comparison as american whites were very fecund). in some parts of the ancient world they weren’t allowed to breed at all (this was common in greece).

    anyway, those points might change my expectations on the margins (he claims that rome’s population was far more native than others would). but he nevertheless supports the literature which indicates cities were population sinks.

  17. onur

    he claims that rome’s population was far more native than others would


    but he nevertheless supports the literature which indicates cities were population sinks.

    But that doesn’t automatically mean that ancient cities usually didn’t have any significant lasting genetic impact. Even in the cases that cities disappeared, city dwellers may really have crowded into the countryside and/or other cities.

  18. since explaining why i weight the different parameters as i do to come to my conclusion is going to take time, and you aren’t being totally clear to me, i’ll leave the discussion for a post where i’ll outline my arguments.

  19. onur

    I think genetic evidence isn’t clear (at least for now) regarding the impact of ancient cities (cosmopolitan or not). So, unlike you, I don’t have a clear opinion on this issue. It seems to me that your opinion is based more on Malthusianism and source-sink dynamics than anything else (including genetics), so I may be missing some important details as, again unlike you, I am not cognizant of the details of these population growth models.

  20. I think genetic evidence isn’t clear (at least for now) regarding the impact of ancient cities (cosmopolitan or not).

    fwiw, it’s not a strong datum in and of itself. you’re right that my confidence is more in malthusianism and source-sink dynamics.

  21. Richard Steckel has assembled a large data set documenting a great decline in health and stature from 3000 years ago until the 19th century. An abstract is in this issue of Science Magazine Gibbons. Civilization’s Cost: The Decline and Fall of Human Health. Science Magazine (2009) vol. 5927 (May) pp. 1-1.

    The city of London could not sustain itself but for immigration.

    A similar pattern of loss of stature and health and the rise of metabolic diseases is emerging in the US since about 1950, coinciding roughly with the birth of MacDonalds.

  22. Nador

    I generally agree with cities being population sinks. Nevertheless share onur’s opinion that it doesn’t mean no genetic impact.
    And one should consider European Jews for example. Most of them were urban dwellers, nevertheless they did not disappear. After the 16th century their numbers increased significantly, so the fact that cities have an overall below replacement level fertility does not necessary imply that there are no groups in cities that flourish.

  23. nador, good point. but what proportion of jews lived in the large cities? i assume that the demographic sink effect would scale with urban-area size. my impression was that many jews lived in towns, or as adjuncts to the estates of polish nobles.

  24. Yawnie

    Urban Jews might have been part of a more cohesive social network, the other inhabitants of towns may have been seperated from extended family support structures.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar