Unprocessed Rice?

Carl Zimmer’s blog reports a new paper by some Harvard geneticists proposing that differences in incidence of diabetes of pregnancy between native New Yorkers and immigrant Bangladeshi women reflects adaptation to wheat and a high carb diet in Europeans.

The hypothesis seemed dubious to me, so I wrote to several colleagues about it. Since I haven’t gotten permission to quote them, I will simply call them BC (Bangladeshi colleague) and VC (Vanilla colleague). Here is our correspondence:


Zimmer reports today on a new paper from Sabeti et al., much of which is about gestational diabetes. The suggestion is that Anglo women are adapted to a high carb diet and can keep their blood sugar down during pregnancy, Bangladeshi women can’t. Here are two quotes:

The Harvard researchers suggest that the shift to high-carb agriculture in Europe led to more women dying of gestational diabetes. Women with mutations that lowered their blood sugar level during pregnancy were favored by natural selection. And today, European women enjoy the benefits of that suffering: a low risk of gestational diabetes.

A woman in Bangladesh has a very different history behind her. Her ancestors ate fish, unprocessed rice, and other foods with modest levels of carbohydrates.

Does this ring right to you? Rice according to current tables is as or more glycemic that white bread.


no, i don’t think any of these make much sense. fwiw, it looks like bengalis didn’t live in bengal until the past 2,000 years. and didn’t everyone eat unprocessed carbs until recently?


As I understand it rice, like wheat, must be milled to remove the bran to allow storage. You ever eat much unprocessed rice back home?


unprocessed? no. my parents started eating it in the USA for health. when they go back to bangladesh ppl think they are insane.


Don’t know about India, but across a lot of SE Asia and Indonesia, rice-eating in significant quantities is a very recent status thing. This is a big topic among sustainability-types, who are trying to figure out how to get people to go back to growing some local starches plus manioc and yams, but can’t get good market prices because they’re “low-class” food.

Anyone know about the history or ecology of milling grain?

This entry was posted in Uncategorized. Bookmark the permalink.

51 Responses to Unprocessed Rice?

  1. For whatever it’s worth, as an alternative to the rice —
    Whenever I think of gestational diabetes I think of David Haig’s (et al) work on maternal-fetal conflict and genomic imprinting (going back to the 1990s). (His papers are mostly accessible through his faculty page at Harvard.) To the extent there are regional/genomic differences I’m inclined to at least consider the hypothesis that marriage systems have resulted in a shift (heritable or otherwise) in the default patterns regarding paternity certainty and the associated (still somewhat hypothetical?) sperm genomic imprinting for the levels of fetal extractiveness of maternal resources (hence gestational diabetes). (I notice that paper co-author Maryellen is in the same sub-department at Harvard as is David Haig… so presumably they are at least aware of this possible line of thinking.) (Contexts to possibly consider, the Trivers-Willard Effect, inbreeding, polygamy .. etc.?)

  2. Francis says:

    Zimmer’s piece makes sense to me. Agriculture was very limited in India / Bangladesh. Many people there were hunter-gathers until fairly recently, some even as late as the 20th century. In this respect, Indians / Bangladeshis are much closer to Africans and Amerindians than they are to North Asians.

    • Melykin says:

      I wonder what the prevalence of alcoholism is in India. I’m not aware that it is particularly high. If they have adapted to alcohol, wouldn’t that mean they must have been farming for awhile? Hunter-gathers seem to have very high rates of alcoholism everywhere else.

    • Many people there were hunter-gathers until fairly recently, some even as late as the 20th century.


      • Francis says:

        Sorry. That is not an argument.

        The Indus civilization did have early exposure to agriculture but it was limited and never fully sustained. Many Indians / Bangladeshis were still hunter-gatherers until fairly recently.

        I am actually thinking about writing a book on late 19th century / early 20th century hunter-gatherers in India.

      • Francis says:


        In recently perusing many of the British documents from the 19th and early 20th centuries, there are some very vivid accounts of various Indian / Bangladeshi hunter-gathers — both outright hunter-gatherers and some who had some understanding of agriculture but still engaged in many hunter-gatherer activities. It’s breathtaking how widespread these practices were.

        We mustn’t forget that the British really modernized India and terms of agriculture and afterwards Indian nationalism began to claim “we were always this way,” which was the origin of many of the current myths regarding agriculture.

        By the way, just recently read 10,000 Year Explosion. Excellent book.

        • gcochran9 says:

          It heartens me to to know that silly people can enjoy my book, because there are so many of them.

          If you look at the genetics of Bengalis, using admixture of frappe or whatever, they look like guys from further up the Ganges, only with a bit more stuff from Southeast Asia. I suppose that the Brits, pioneering genetic engineers (centuries ahead of their time), altered the genes of the local hunter-gatherers just to fuck with our colonial minds, or maybe as part of some long range-plan beyond our ken – one of the mysteries of the Occident.

          Or maybe you’re just wrong.

      • Francis says:

        Should read: the British really modernized India in terms of agriculture

      • Discard says:

        Maybe Francis could write a book about hunter/gatherers here in the U.S. at this very moment. They come through my neighborhood every week, at dawn on trash day.
        The presence of primitives does not preclude modernity, let alone agriculture.

      • diana says:

        Did they control for maternal weight gain? White women in NYC are mostly middle class and higher – they watch their diets carefully and try not to gain too much (any) weight at all during pregnancy. Black women start out fat, probably don’t gain a lot. E. Asian women start out normal weight; stay that way. Early pregnancy weight gain is related to GD.

        My guess is that American-based S-Asian women simply eat an awful lot when they get pregnant, gain too much weight, and come down with GD. Just a guess. The article didn’t talk about maternal weight gain. A rather large omission.

        Also, this business of blaming insulin spikes on carbs alone is ridiculous. Protein also raises insulin. The only macronutrient that doesn’t raise insulin is fat.

    • Discard says:

      How did there get to be so many of them without agriculture? How did they develop so elaborate a culture as hunter-gathers?

  3. Andrew says:

    I recall that the middle ages in NW Europe saw a decline in temperature and Europeans switched to a more grain based diet. Todays version of processed wheat has a high glycemic index. Higher than table sugar.

  4. pauljaminet says:

    Milling of grain was uncommon until the industrial era began a few centuries ago. Prior to that, soaking/fermentation was a more common treatment to render the whole grain more digestible and more healthful.

    The argument seems to depend on Bengalis having only recently adopted agriculture. I don’t know the heritage of Bengalis but the Indus civilization was an early adopter of agriculture, so that region was certainly familiar with agriculture at an early date. Northwest Europeans on the other hand were non-agricultural until the Bronze and Iron Ages, and large regions remained pastoral (milk, cheese, mutton, beef) and seafood focused (seafood was a major part of the 9th century English diet). Diets in most of the world become high-carb as soon as grain agriculture made grains the cheapest sources of calories, which it did earlier at more tropical latitudes. Nowadays, poor societies like Bangladesh eat higher-carb diets than Europeans.

    It could be that there are specific adaptations to wheat in Europeans, and Bengalis with a history of consuming other starches may lack those adaptations. Then it would be the wheat, not the carbs, that would cause the disease.

    But the inverse correlation with neural tube defects suggests it may be more a lipid-handling issue. Neural tube defects primarily arise from choline deficiency in the baby; and choline deficiency in adults is a known cause of metabolic syndrome and promotes diabetes. A simple explanation would be that Europeans direct choline more toward the mother than the baby, and Bengalis more toward the baby than the mother (ie sacrifice the mother’s health to make a healthier baby). Perhaps a co-evolution with life history — historically shorter lifespans in Bengal?

    • pauljaminet says:

      Just a note in support of choline transport as the culprit:

      There are multiple choline transporters; SLC44A1 transports choline into nerves (preventing neural tube defects) while SLC44A2-4 transport it into peripheral tissues including the liver (preventing metabolic syndrome). http://www.ncbi.nlm.nih.gov/pubmed/23506897.

      Expression of these transporters is age dependent, suggesting epigenetic regulation. Autoimmunity to one of these transporters is a cause of deafness. http://www.ncbi.nlm.nih.gov/pubmed/21986210.

      So it would be easy for evolution to adjust the relative migration of choline to the baby (which has primarily neuronal needs) or mother (who has primarily peripheral needs) simply by changing the ratio of SLC44A1 to SLC44A2-4 expression.

      A simple story would be that historical choline intake in India has been much lower than in Europe (very likely true), and that led to selection for higher expression of the neuronal carrier in Indians, especially Indian fetuses, since brain function during fetal development is so crucial compared to the peripheral functions of choline in adults. Then if in New York everyone is eating a choline-deficient diet, it manifests as neural tube defects in the Europeans and gestational diabetes in the Bengalis.

    • Greying Wanderer says:

      “Milling of grain was uncommon until the industrial era began a few centuries ago. Prior to that, soaking/fermentation was a more common treatment to render the whole grain more digestible and more healthful.”

      Separate from the diabetes thing and thinking more generally in terms of industrialization leading to changes to traditional food it’s easy to imagine soaking doing something to grain that milling didn’t.

  5. Francis says:

    “but the Indus civilization was an early adopter of agriculture”

    Agriculture was brought there but was always limited and never fully sustained. The overall impact of agriculture on the majority of Indians / Bangladeshis would have been extremely limited.

    • misdreavus says:

      Would you mind explaining to me how the Maurya Empire (4th to 2nd centuries B.C.) sustained a population of over 50 million, then over *one fourth* of the world’s population, when almost nobody scratched a living from the tiling of the soil?

      • Trust me, I'm Indian says:

        Either Francis is saying that an agricultural civilization that survived for two millennia somehow “never fully sustained” agriculture…
        Or, in Francis’ world, the Indus civilization never took root, India was occupied primarily by hunter-gatherers and primitive farmers until the 19th century, with agriculture being introduced only by the British colonists.
        Further reading: http://bit.ly/18m9F84

        • gcochran9 says:

          In principle I welcome visitors from other timelines, but it seems that the transition damages their brains. Which is too bad, because think of what we might learn from them!

    • Greying Wanderer says:


      You need to take into account population density. You can get a lot more farmers per square mile in most places than hunter-gatherers so they can still be a large majority of a total population even if they’re only using a minority of the land.

      • Greying Wanderer says:

        “so they can” = “so the farmers can”

      • misdreavus says:

        This can happen wherever marginal and unproductive lands constitute the vast majority of the landscape. (See: Namibia, most of Siberia, etc.) At the turn of the 20th century, there were millions of Russian farmers in Siberia, some of them as far north as the Yenisei river, and they probably outnumbered hunter-gatherers by a considerable margin.

        None of this is true for India. Not on the Deccan Plateau, not in the Thar Desert, and not even at the foothills of the Himalayas. Just where the hell did all these “hunter gatherers” go during the dissolution of the Mughal Empire?

  6. Candide III says:

    Certainly rice was a luxury until not so long ago in Japan. The commoners ate millets, buckwheat, sorghum and other such lower-quality cereals. Similarly, white bread was a luxury in Europe. The majority ate rye bread (when they could get it).

  7. Violet says:

    From anecdote, unprocessed (unpolished / brown) rice was common until about a hundred years ago. Paddy was de-husked by hand-pounding in small quantities for family consumption. Most farmers in rice-belt in South India became rich by starting rice-mills in early 20th century. Milled rice has been common at least in the last hundred years. Even then, there is a difference in traditional mills which are only hullers, and modern mills which not only hull but also polish rice several times.

    Current survey of paddy processing industries reflect that high-efficiency rice polishing wasn’t the norm until recently. The sharp decline in hand-pounding and use of female labourers was recorded within last 50-60 years.

    1.Dr. Komol Singha, Structure and Performance of Paddy Processing Industry in India: A Case of Karnataka
    2. Women in Rice Farming. International Rice Research Institute

    • Violet says:

      Complete reference for second one:
      Proceedings of a conference on women in rice farming systems, The International Rice Research Institute, Manila, Philippines, 23-26 Sep. 1983

  8. Jaim Jota says:

    The argument seems to depend on Bengalis having only recently adopted agriculture.
    There is no argument that they were eating cereals a thousand years ago. But the machine-processed and polished rice, of the recently introduced high yield varieties, is something totally new.

    • pauljaminet says:

      But the argument is that Europeans are better adapted to carbohydrates than Bengalis, and northwestern Europeans have had (by global standards) dairy, meat, and fish rich diets for thousands of years. Moreover, they consumed a large fraction of their cereal grains as alcoholic beverages, so their carb intake was not high. Milled rice is fairly new, but its character isn’t dramatically different than traditional Indian forms of grain and bean — certainly not its GI or carb content. The argument might work better for southern Europeans, who ate more grain and less meat in the medieval and classical periods, but I suspect Indians were higher in carbs throughout.

      Whereas the argument that Indian diets were lower in lipids, especially phospholipids, than European diets is far more plausible. It is consistent with the much higher population densities in India than in most of Europe. Europeans were able to eat meat and dairy rich diets because their population densities were low enough to support pastoral lifestyles.

  9. I thought it might be useful to provide some more information in the paper I reported on. They write, “Traditionally, Bangladeshis have had high consumption of fish, a low glycemic food; rice, of moderate glycemic index due to little processing; and no dairy.” For these claims they cite the following papers:

    Savitz, D.A. et al. (2008) Ethnicity and gestational diabetes in New
    York City, 1995–2003. BJOG 115, 969–978

    Itan, Y. et al. (2010) A worldwide correlation of lactase persistence
    phenotype and genotypes. BMC Evol. Biol. 10, 36–47

    tkinson, F.S. et al. (2008) International tables of glycemic index and
    glycemic load values: 2008. Diabetes Care 31, 2281–2283. This paper is open access here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2584181/ (Follow links to an appendix table with glycemic values for lots of foods. You’ll see South Asian rice has a much lower glycemic index than, say, a white-flour bagel.)

  10. Francis says:

    I know what I’m saying is a little revisionist but that’s the nature of scholarship.

    Haven’t decided whether I’ll write my dissertation on this (and then a book) but I’m leaning that way. A publisher said I’d need many photos (which I’m trying to find) because a book like this will sell more copies if it has numerous photos.

    My findings are accidental. I was researching something else when I came across vivid descriptions of 19th century hunter gatherers in India.

    Agricultural settlements can be dense and small — small enclaves surrounded by vast lands of hunter gatherers. There’s evidence of this with Amerindians in Latin America.

    My claims are modest:

    – Agriculture in 19th century India was more limited than previously believed

    – Hunter-gatherers were more widespread than previously believed

    We’ll see where this goes…. Thinking about setting up a blog and posting some descriptions from British surveyors.

    • misdreavus says:

      You seem to think it probable that 150 million* people will voluntarily cluster themselves in just a few parcels of a subcontinent that has witnessed over four milllenia of continuous civilization, and agriculture for at least that amount of time. And surely not one where there is hardly a dearth of arable land everywhere.

      So all the farmers of the Gangetic Plain just cozied up together for absolutely no good reason at all, leaving vast tracks of prime real estate to the Stone Age savages for thousands of years. Maybe they were just really nice, for some odd reason.

      Historical records of dozens of empires crisscrossing the landscape, from Sindh to Bangalore, military conquests involving hundreds of thousands of soldiers, war elephants, and millions of civilian casualties — pooh! Yeah that sounds like a primitive wasteland to me. My subjective opinion speaks otherwise.

      (*Estimated population of India ~1750.)

    • misdreavus says:


    • Discard says:

      Hunter-gatherers built the Taj Mahal? There’s a PhD for anyone who can prove that.

  11. The fourth doorman of the apocalypse says:

    As I recall Carl Zimmer seems much taken by the Anthropogenic Global Warming nonsense.

  12. Reader says:

    Don’t know if I buy into Francis’ claim that majority of Indians were still hunter-gatherers in 19th century but there is something genetically interesting going on there in terms of cognitive ability.

    Average IQ of Northeast Asia: 101 – 105
    Average IQ of India: 81

    On PISA, parts of India are not much higher than Sub-Sahara Africa:


    What’s going on here?

  13. RS says:

    > Europeans were able to eat meat and dairy rich diets because their population densities were low enough to support pastoral lifestyles.

    Interesting post, but just this one point seems backward, or likely so. It sounds more like the cereals fared somewhat poorly due to incomplete adaptation to non-Mediterranean Europe (or at least to many tracts thereof), and this caused the relatively primitive/mixed food economy and low pop density. I encountered this idea in Hart’s “Understanding human history” (available for free).

    In Homer there seems to be relatively much talk of pastoralism, compared to our more vegetal image of the classical Mediterranean.

    Alternatively the Euros could also have achieved low density by killing each other a lot, or by suffering especially from infection, but neither really comports with any knowledge we have (other than modern Eastern Slavs being rather violent).

    • Matt says:

      My impression (partly gathered from Nick Szabo’s related blog posts) is that it may have been more of a question of poor adaption (for one reason or another), in Europe North of the alps of legumes, rather than cereals, which were used to provide fat and protein in other agricultural societies, and were also useful for nitrogen fixation.

      Man cannot live by bread alone, thus a greater use of land on animals, to provide fat and protein (through flesh and dairy), and to pass nitrogen back into the soil (manure).

      I think it is a similar story with vegetable fiber crops (more wool, less linen).

  14. Matt says:

    One idea: Although Northwest Europeans are not the population with the highest history of grain carbohydrate consumption as a % of diet, they may have had a longer effective period of adaption to high carbohydrate consumption compared to South Asians, if Europeans underwent a greater degree of population replacement from “first farmer” type cultures. We don’t know how much population replacement went on in Europe, whereas we can be sure substantial amounts of pre-farming ancestry remain in South Asia.

    I think this may not be useful though – Fertile Crescent populations don’t necessarily have lower levels of gd than Europeans and Chinese (who have no pre-farming substratum element) have reportedly high levels of gd.

    Maybe it’s just because Bengalis eat more carbs, but aren’t particularly that more well adapted to it?

    • pauljaminet says:

      Doubtful. The European Neolithic began with the domestication of cattle. Judging by the evolution of lactase persistence a few thousand years later, a significant meat-and-dairy dietary component was present from an early date. In general northern latitudes favor reduced carbohydrate diets compared to lower latitudes.

      • harpend says:

        Hi Paul. Folks of the Colin Renfrew school would have lactase persistence coevolve with cattle in the Neolithic but that hypothesis does not fit the linguistic data very well. There is much to support the old Gimbutas idea that Indo-Europeans were Comanches from the Pontic steppes and that they are the original Eurasian lactase persistent folks.

        A non-LP child gets about 400 Cal from a liter of cow’s milk while a LP child gets 720 Cal, an 80% boost. The energy density of mare’s milk is less. A non-LP child would obtain nearly 200 Cal from a liter but an LP child would obtain 440: the energetic benefit of LP among horse people is 130%. If we are right LP is what we used to call a “pre-adaptation” to dairying.

        My favorite hypothesis is that the first Indo-Europeans were horse people that overran Europe about 4 or 5 kya ago, destroying much of it. The first farmer types, or rather their genes, seem to have persisted in e.g. Corsica but not much anywhere else. Otzi the Iceman, one of the first farmers, is most similar to today’s Corsicans.

        A trendy archaeological site is Botai in northern Kazakstan, where there is the first evidence of bridles on horse teeth. There is a lot of controversy about the finding of course.

        • pauljaminet says:

          Hi Henry,

          When I read David Anthony’s The Horse, the Wheel and Language he persuaded me of the steppe origins theory, but then more recent papers have made me think a multiple-origins model makes more sense, with a Proto-Proto-Indo-European expansion from Anatolia at the time of the domestication of cattle and a second expansion from the steppes a few thousand years later following domestication of the horse, with more population replacement/mixing/migrations following invention of metal weapons. Dienekes has been good on this, see http://dienekes.blogspot.com/2012/08/proto-indo-european-homeland-in.html for a bit of the evidence for an Anatolian origin of PIE, and http://dienekes.blogspot.com/2011/04/indo-european-origins-neolithic.html for a counter-argument to Anthony’s linguistic arguments that claims glottochronology is unreliable and that Bayesian phylogenetics applied to languages points to an Anatolian origin of PIE c. 6000 BC.

          I haven’t been able to keep up with the genetic evidence in detail. The main thing I’ve taken from it is that there was a lot of genetic change between 6000 BC and classical times. Given relatively low population densities and major developments in weapons technology, domesticated animals and transportation technologies, and social structure, such changes aren’t too surprising. The balance of power must have shifted many times, and given the violence of those times — with frequent slaughters of martial-age males of a captured people and enslavement of the rest, think the destruction of the Amalekites or the ten lost tribes of Israel — big genetic shifts shouldn’t surprise us much. It’s not necessary to ascribe the genetic change to a single event like the Indo-European expansion.

          I’m hoping I can ignore the genetic evidence for a few years until we have complete genomes of many ancient skeletons and somebody can write a simple review paper for me.

      • The Bouckaert and others paper certainly represents a lot of computation and the graphics are great. I am not sure what to make of it: I would be much happier if there weren’t a great big broad Turkic wedge right smack in the middle of Eurasia.

        My appeal to linguistics was to the old-fashioned arguments that PIE had bears and beeches and birches and so on.


  15. Greying Wanderer says:

    “None of this is true for India. Not on the Deccan Plateau, not in the Thar Desert, and not even at the foothills of the Himalayas. Just where the hell did all these “hunter gatherers” go during the dissolution of the Mughal Empire?”

    I never said it was. I thought it might nudge Francis.
    “native New Yorkers and immigrant Bangladeshi women”

    I think you need to know which native New Yorkers he’s talking about.

    “Europeans were able to eat meat and dairy rich diets because their population densities were low enough to support pastoral lifestyles.”

    Other way round i think with the ecological niche beyond the meditteranean rim being better suited to cattle-centric farming for a long time.

  16. Greying Wanderer says:

    “The argument seems to depend on Bengalis having only recently adopted agriculture.”

    “There is no argument that they were eating cereals a thousand years ago. But the machine-processed and polished rice, of the recently introduced high yield varieties, is something totally new.”

    Although the processing idea seems equally plausible for there to be a difference between two populations does it have to be recent or does it only have to be *more* recent e.g. if middle easterners had 4000 years of adaptation and Bangladeshis 2000 years couldn’t there still be a difference?

    (just random numbers to illustrate the point)

  17. RS says:

    > To the extent there are regional/genomic differences I’m inclined to at least consider the hypothesis that marriage systems have resulted in a shift (heritable or otherwise) in the default patterns regarding paternity certainty and the associated (still somewhat hypothetical?) sperm genomic imprinting for the levels of fetal extractiveness of maternal resources (hence gestational diabetes).

    You might cite the term ‘maternal-fetal/ maternal-paternal conflict’ since lots of people won’t know what it is.

    I think your idea interesting. Could be the total cause of geographic variance is a composite of that, and some of the other stuff. Anyway, along your lines, what I think of specifically is polygyny rather than paternity certainty. (In fact I am hard pressed to comprehend a connection with the latter.)

    With higher levels of polygyny, genes spend more time inside males who can more readily get another female to replace a compromised one, than genes do under a norm of monogamy. Corresponding defenses ought to evolve one day or another. But as with driving alleles, it may be that aggressor alleles are currently ahead of defender ones in man, due to our recent pop expansion suddenly providing abundant new mutations. Cochran and or Harpending have mentioned that large numbers of new driver alleles might explain the extremely low fecundability rate in normal humans ; they note that the driver (or other aggressor allele) has to become common first in time, because only after it succeeds as an allele is there much positive selection for corresponding defender alleles. Therefore, just after an expansion a species could become relatively more ‘sick’, for some period of time, with yet-uncountered selfish alleles.

    Whether Europeans have actually practiced the monogamy they espoused doctrinally under christendom (and often before? I’m not sure) — is another question; obviously there was lots of covert polygyny (not quite the same thing as nonpaternity), and some seem to think it was very extensive. I don’t know much about polygyny in Bengladash.

  18. Old Pete says:

    Oddly, the Harvard researchers make the link with ultraviolet exposure to explain the difference in neural tube defects but fail to make the connection for gestational diabetes. There is evidence that suggests a causal link between vitamin D deficiency and gestational diabetes. Here is one example, a search will produce many more if anyone is curious.

  19. Toad says:

    There is also a large difference in PUFA consumption.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s